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

v>EPA

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

EPA Document# EPA-740-D-23-001

July 2023

United States	Office of Chemical Safety and

Environmental Protection Agency	Pollution Prevention

Draft Supplement to the Risk Evaluation for

1,4-Dioxane

CASRN 123-91-1

July 2023


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

29	TABLE OF CONTENTS

30	ACKNOWLEDGEMENTS	17

31	EXECUTIVE SUMMARY	19

32	1 INTRODUCTION	25

33	1.1 Regulatory Context	25

34	1.2 Scope	27

35	1.3 Use Characterization	28

36	1.3,1 Conceptual Models	28

37	1.3.1.1 1,4-Dioxane as a Byproduct	30

38	1.3.1.2 Occupational Exposures	32

39	1.3.1.3 General Population Exposures	34

40	1.3.1.3.1 Drinking Water	36

41	1.3.1.3.2 Air	37

42	1.3.1.3.3 Aggregate Exposure	37

43	1,3,2 Potentially Exposed or Susceptible Subpopulations	37

44	1.4 Systematic Review	38

45	1.5 Document Outline	38

46	2 RELEASES AND CONCENTRATIONS	40

47	2.1 Approach and Methodology	40

48	2.1,1 Industrial and Commercial Releases	40

49	2.1.1.1 General Approach and Methodology for Environmental Releases	42

50	2.1.1.2 Water Release Estimates	42

51	2.1.1.3 Land Release Estimates	43

52	2.1.1.4 Air Release Estimates	44

53	2.1.1.4.1 Pre-screening Analysis	44

54	2.1.1.4.2 Single-Year Fenceline Analysis	44

55	2.1.1.4.3 Multi-year Analysis	45

56	2.2 Environmental Releases	45

57	2.2,1 Industrial and Commercial Releases	45

58	2.2.1.1 Release Estimates Summary	45

59	2.2.1.2 Weight of the Scientific Evidence Conclusions for Environmental Releases	49

60	2.2.1.3 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the

61	Environmental Release Assessment	50

62	2.3 1,4-Dioxane Environmental Concentrations	52

63	2.3,1 Surface Water Pathway	52

64	2.3.1.1 Monitoring Data	52

65	2.3.1.2 Surface Water and Drinking Water Modeling	58

66	2.3.1.2.1 Modeling Methodology	58

67	2.3.1.2.2 EstimatingDown-the-DrainReleases	60

68	2.3.1.2.3 Hydraulic Fracturing	60

69	2.3.1.2.4 Proximity to Drinking Water Sources	60

70	2.3.1.3 Modeling Results	60

71	2.3.1.3.1 Facility-Specific Results	60

72	2.3.1.3.2 Concentrations from Down-the-Drain Loading	66

73	2.3.1.3.3 Concentrations from Hydraulic Fracturing	67

74	2.3.1.3.4 Aggregate Probabilistic Results	68

Page 2 of 484


-------
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

119

120

121

122

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2.3.1.4	Comparison of Modeled and Monitored Surface Water Concentrations	71

2.3.1.5	Strengths, Limitations, and Sources of Uncertainty in Assessment Results for
Monitored and Modeled Drinking Water and Surface Water Concentrations	72

2.3.2	Land Pathway (Groundwater)	73

2.3.2.1	Groundwater Monitoring Data	73

2.3.2.2	Disposal via Underground Injection	75

2.3.2.2.1	Summary of Assessment for Disposal to Underground Injection	76

2.3.2.2.2	Strengths, Limitations, and Sources of Uncertainty in Assessment of Disposal to
Underground Injection Wells	76

2.3.2.3	Disposal to Landfills	76

2.3.2.3.1	Summary of Assessment for Disposal to Landfills	80

2.3.2.3.2	Strengths, Limitations, and Sources of Uncertainty in Assessment Results for
Disposal to Landfills	80

2.3.2.4	Disposal of Hydraulic Fracturing Produced Water to Surface Impoundments	81

2.3.2.4.1	Summary of Assessment for Disposal of Hydraulic Fracturing Produced Water.... 81

2.3.2.4.2	Strengths, Limitations, and Sources of Uncertainty in Assessment Results for
Disposal from Hydraulic Fracturing Operations	82

2.3.3	Ambient Air Pathway	82

2.3.3.1	Measured Concentrations in Air	83

2.3.3.2	Modeled Concentrations in Air	83

2.3.3.2.1	Ambient Air: Screening Methodology	84

2.3.3.2.2	Ambient Air: Single Year Methodology (AERMOD)	84

2.3.3.2.3	Ambient Air: Multi-Year Analysis (IIOAC)	88

2.3.3.2.4	Ambient Air: IIO AC Methodology for COUs without Site-Specific Data
(Hydraulic Fracturing, Industrial, and Institutional Laundry Facilities)	88

2.3.3.3	Strengths, Limitations, and Sources of Uncertainty for Modeled Air Concentrations.... 89

3 HUMAN EXPOSURES	91

3.1	Occupational Exposures	91

3.1.1	Approach and Methodology	92

3.1.1.1	Process Description, Number of Sites, Number of Workers, and ONUs	92

3.1.1.2	Inhalation Exposures Approach and Methodology	93

3.1.1.3	Dermal Exposures Approach and Methodology	94

3.1.1.4	Engineering Controls and Personal Protective Equipment	94

3.1.2	Occupational Exposure Estimates	95

3.1.2.1	Summary of Inhalation Exposure Assessment	95

3.1.2.2	Summary of Dermal Exposures Assessment	96

3.1.2.3	Weight of the Scientific Evidence Conclusions for Occupational Exposure
Information	97

3.1.2.4	Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
Occupational Exposure Assessment	99

3.1.2.4.1	Number of Workers	99

3.1.2.4.2	Analysis of Inhalation Exposure Monitoring Data	99

3.1.2.4.3	Modeled Inhalation Exposures	100

3.1.2.4.4	Modeled Dermal Exposures	100

3.2	General Population Exposures	101

3.2.1	Approach and Methodology	102

3.2.2	Drinking Water Exposure Assessment	103

3.2.2.1 Surface Water Exposure Assessment	103

Page 3 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

123	3.2.2.1.1 Exposures from Individual Facility Releases	103

124	3.2.2.1.2 Exposures from Down-the-Drain Releases	106

125	3.2.2.1.3 Disposal of Hydraulic Fracturing Produced Waters	106

126	3.2.2.1.4 Aggregate Exposure	107

127	3.2.2.2 Groundwater Exposure Assessment	108

128	3.2.2.2.1 Disposal to Landfills	108

129	3.2.2.2.2 Disposal of Hydraulic Fracturing Produced Waters	109

130	3.2.3 Air Exposure Assessment	110

131	3.2.3.1 Industrial COUs Reported to TRI	110

132	3.2.3.2 Hydraulic Fracturing	112

133	3.2.3.3 Industrial and Institutional Laundry Facilities	115

134	3.3 Weight of the Scientific Evidence Conclusions	116

135	3.3.1 Occupational Exposures	116

136	3.3.1.1 Inhalation Exposure	116

137	3.3.1.2 Dermal Exposure	117

138	3.3,2 Drinking Water	118

139	3.3.2.1 Drinking Water Exposure Estimates Based on Surface Water Concentrations	118

140	3.3.2.2 Drinking Water Exposure Estimates Based on Groundwater Concentrations	122

141	3.3.2.2.1 Groundwater Concentrations Resulting from Disposal to Landfill	123

142	3.3.2.2.2 Groundwater Concentrations Resulting from Disposal of Hydraulic Fracturing

143	Waste	124

144	3.3.3 Air 	124

145	3.3.3.1 Modeled Air Concentrations for Industrial COUs Reported to TRI	125

146	3.3.3.2 Air Concentrations Modeled near Hydraulic Fracturing Operations and

147	Industrial/Institutional Laundries	126

148	4 HUMAN HEALTH HAZARD	128

149	4.1 Summary of Hazard Endpoints Previously Identified in the 2020 Risk Evaluation	128

150	4.2 Summary of Adjustments to Previously Established Hazard Values	128

151	4.2.1 Derivation of Acute/Short-Term Hazard Values	131

152	4.2.1.1 Inhalation HEC	131

153	4.2.1.2 Oral and Dermal HEDs	131

154	4.2.2 Derivation of Chronic Hazard Values	131

155	4.2.2.1 Inhalation HEC	131

156	4.2.2.2 Oral HEDs	132

157	4.2.2.3 Dermal HEDs	132

158	4.2.3 Derivation of Cancer Hazard Values	132

159	4.3 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty in the Hazard and

160	Dose-Response Analysis	133

161	5 HUMAN HEALTH RISK CHARACTERIZATION	134

162	5.1 Risk Characterization Approach	135

163	5,1.1 Estimation of Non-cancer Risks	136

164	5.1.2 Estimation of Cancer Risks	136

165	5.2 Human Health Risk Characterization	136

166	5.2.1 Summary of Risk Estimates for Occupational Exposures	136

167	5.2,2 Summary of Risk Estimates for the General Population	140

168	5.2.2.1 Drinking Water - Surface Water Pathway	140

Page 4 of 484


-------
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

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

5.2.2.1.1 Risks from Exposure to Drinking Water Concentrations Indicated in Finished

Drinking Water Monitoring Data	141

5.2.2.1.2	Risks from Exposures to Water Concentrations Modeled from Industrial
Releases	141

5.2.2.1.3	Risks from Exposures to Water Concentrations Modeled from DTD Releases
(from POTWs), Assuming No Downstream Dilution	146

5.2.2.1.4	Risks from Exposure to Drinking Water Concentrations Modeled from Disposal
of Hydraulic Fracturing Produced Waters to Surface Water, Assuming No
Downstream Dilution	147

5.2.2.1.5	Aggregate Risks from Drinking Water Exposures Modeled from Multiple
Sources Releasing to Surface Water, Assuming No Downstream Dilution	148

5.2.2.1.6	Integrated Summary of Drinking Water Risk Estimates across Multiple Lines of
Evidence for Surface Water	151

5.2.2.2	Drinking Water - Groundwater and Disposal Pathways	152

5.2.2.3	Air Pathway	154

5.2.2.3.1	Industrial COUs Reported to TRI	154

5.2.2.3.2	Hydraulic Fracturing	159

5.2.2.3.3	Industrial and Institutional Laundry Facilities	161

5.2.2.4	Potentially Exposed or Susceptible Subpopulations	162

5.2.2.5	Aggregate and Sentinel Exposures	164

5.2.2.6	Summary of Overall Confidence and Remaining Uncertainties in Human Health Risk
Characterization	165

5.2.2.6.1	Risks from Occupational Exposures	165

5.2.2.6.2	Risks from General Population Exposures through Drinking Water	166

5.2.2.6.3	Risks from General Population Exposures through Groundwater and Land
Disposal Pathways	167

5.2.2.6.4	Risks from General Population Exposures through Air	167

REFERENCES	169

APPENDICES	177

Appendix A ABBREVIATIONS AND ACRONYMS	177

Appendix B LIST OF SUPPLEMENTAL DOCUMENTS	180

Appendix C SYSTEMATIC REVIEW PROTOCOL FOR THE DRAFT SUPPLEMENT TO

THE RISK EVALUATION FOR 1,4-DIOXANE	183

C.l Clarifications and Updates to the 2021 Draft Systematic Review Protocol	184

C.l.l Clarifications and Updates	184

C.l Data Search	187

C.2.1 Multi-Disciplinary Updates to the Data Search	187

C.2.2 Additional Data Sources Identified	188

C.2.2.1 Additional Data Sources Identified for Environmental Release and Occupational

Exposure	188

C.2.2.2 Additional Data Sources Identified for General Population, Consumer, and

Environmental Exposure	189

C.2.3 Search Strings	189

C.2.3.1 Environmental Release and Occupational Exposure Search Strings	189

C.2.3.2 General Population, Consumer, and Environmental Exposure Search Strings	190

C.3 Data Screening	191

Page 5 of 484


-------
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

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

C.3.1 Environmental Release and Occupational Exposure	192

C.3.1.1 Environmental Release and Occupational Exposure Literature Tree	193

C.3.2 General Population, Consumer, and Environmental Exposure	193

C.3.2.1 General Population, Consumer, and Environmental Exposure Literature Tree	194

C.4 Data Evaluation and Data Extraction	194

C.4.1 Environmental Release and Occupational Exposure	195

C.4.2 General Population, Consumer, and Environmental Exposure	195

C.4.2.1 Data Quality Evaluation Metric Updates	196

C.4.2.2 Data Evaluation Criteria for Monitoring Data, as Revised	197

C.4.2.3 Data Evaluation Criteria for Experimental Data, as Revised	204

C.4.2.4 Data Evaluation Criteria for Databases, as Revised	210

C.	5 Evi dence Integrati on	214

C.5.1 Environmental Release and Occupational Exposure	219

C.5.2 General Population	219

C.5.2.1 General Population: Surface Water	219

C.5.2.2 General Population: Groundwater	219

C.5.2.3 General Population Exposure: Ambient Air	219

Appendix D COU-OES MAPPING AND CROSSWALK	221

D.l	COU-OES Mapping	221

D,2	COU-OES Crosswalk	223

Appendix E INDUSTRIAL AND COMMERCIAL ENVIRONMENTAL RELEASES	226

E.l	Estimates of the Number of Industrial and Commercial Facilities with Environmental
Releases	226

E.2 Estimates of Number of Release Days for Industrial and Commercial Releases	229

E.3 Water Release Assessment	231

E.3.1 Assessment Using TRI and DMR	232

E.3.2 Assessment for OES without TRI and DMR	234

E.3.3 Water Release Estimates Summary	237

E.3.4 Summary of Weight of the Scientific Evidence Conclusions in Water Release Estimates. 243

E,4 Land Release Assessment	251

E.4.1 Assessment Using TRI	251

E.4.2 Assessment for OES without TRI	253

E.4.3 Land Release Estimates Summary	259

E.4.4 Summary of Weight of the Scientific Evidence Conclusions in Land Release Estimates.. 263

E.5 Air Release Assessment	271

E.5.1 Assessment Using TRI	271

E.5.2 Assessment for OESd without TRI	273

E.5.3 Air Release Estimates Summary	279

E.5.4 Summary of Weight of the Scientific Evidence Conclusions in Air Release Estimates	285

E.6 Comparison to PET Life Cycle Analysis	291

E.7 Detailed Strengths, Limitations, Assumptions and Key Sources of Uncertainties for the

Environmental Release Assessment	292

E.8 Weight of the Scientific Evidence Conclusions for Environmental Releases	296

E,9 TRI to CDR Crosswalk	300

E, 10 Developing Models that Use Monte Carlo Methods	318

E.10.1 Background on Monte Carlo Methods	318

E.10.2 Implementation of Monte Carlo Methods	318

Page 6 of 484


-------
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

301

302

303

304

305

306

307

308

309

310

319

319

319

323

323

323

323

324

326

329

329

329

329

329

329

329

330

330

330

330

330

330

331

333

339

343

343

343

343

343

344

344

344

344

344

345

345

345

345

345

345

345

346

346

346

347

351

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

E.10.3 Building the Model	

E.10.3.1 Build the Deterministic Model	

E.10.3.2 Define Probability Distributions for Input Parameters	

E.10.3.3 Select Model Outputs for Aggregation of Simulation Results	

E.10.3.4 Select Simulation Settings and Run Model	

E.10.3.5 Aggregate the Simulation Results and Produce Output Statistics	

E, 11 Textile Dye Modeling Approach and Parameters for Estimating Environmental Releases

E. 11.1 Model Equations	

E. 11.2 Model Input Parameters	

E. 11.3 Number of Sites	

E.11.4 Mass Fraction of Dye Containing 1,4-Dioxane	

E. 11.5 Operating Days	

E.l 1.6 Mass Fraction of 1,4-Dioxane in Dye Formulation	

E. 11.7 Textile Production Rate	

E. 11.8 Mass Fraction of Textiles Treated with Dye	

E.l 1.9 Mass Fraction of Dye Used per Mass of Textile Dyed	

E. 11.10 Mass Fraction of the Dye Formulation in the Dyebath	

E. 11.11 Container Size for Dye Formulation	

E. 11.12 Container Residual Fraction for Totes	

E. 11.13 Container Residual Fraction for Drums	

E. 11.14 Container Residual Fraction for Pails	

E. 11.15 Fraction of Dye Product Affixed to Textile during Dyeing Process Substrate	

E, 12 Laundry Detergent Modeling Approach and Parameters for Estimating Environmental

Releases	

E.12.1 Model Equations	

E.12.2 Model Input Parameters	

E.12.3 Operating Days	

E.12.4 Mass Fraction of 1,4-Dioxane in Laundry Detergent	

E.12.5 Daily Use Rate of Detergent	

E.12.6 Container Size	

E.12.7 Indoor Air Speed	

E.12.8 Container Residual Fraction for Totes	

E. 12.9 Container Residual Fraction for Drums	

E. 12.10 Container Residual Fraction for Pails	

E. 12.11 Container Residual Fraction for Powders	

E. 12.12 Fraction of Laundry Detergents Containing 1,4-Dioxane	

E.12.13 Duration of Release for Container Unloading	

E. 12.14 Fraction of Chemical Lost during Transfer of Solid Powders	

E. 12.15 Control Efficiency for Dust Control Methods	

E. 12.16 Capture Efficiency for Dust Capture Methods	

E.12.17 Number of Sites	

E. 12.18 Diameter of Container Opening	

E.12.19 Diameter of Wash Opening	

E.12.20 Dilution Factor	

E. 12.21 Container Fill Rate	

E.l3 Hydraulic Fracturing Modeling Approach and Parameters for Estimating Environmental

Releases	

E.13.1 Model Equations	

E.13.2 Model Input Parameters	

Page 7 of 484


-------
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

354

354

354

354

354

354

354

354

355

355

355

355

355

356

356

357

357

357

360

363

363

369

372

376

380

381

387

392

393

397

401

405

408

411

411

413

416

416

416

416

416

416

417

417

417

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

E.13.3 Number of Sites	

E.13.4 Operating Days	

E.13.5 Container Size	

E.13.6 Diameter of Container Opening	

E.13.7 Diameter of Equipment Opening	

E.13.8 Air Speed during Equipment Cleaning	

E.13.9 Equipment Cleaning Loss Fraction	

E. 13.10 Container Fill Rate	

E. 13.11 Equipment Cleaning Operating Hours	

E.13.12 Annual Use Rate of Fracturing Fluids Containing 1,4-Dioxane	

E.13.13 Mass Fraction of 1,4-Dioxane in Hydraulic Fracturing Additive/Fluid	

E.13.14 Saturation Factor	

E.13.15 Container Residual Fraction for Totes	

E.13.16 Container Residual Fraction for Drums	

E.	13.17 Fraction of Inj ected Fracturing Fluid that Returns to the Surface	

Appendix F OCCUPATIONAL EXPOSURES	

F.l Calculating Acute and Chronic Inhalation Exposures and Dermal Doses	

F.2 Approach for Estimating Number of Workers and Occupational Non-users	

F.3 Occupational Dermal Exposure Assessment Method	

F.4 Occupational Exposure Scenarios	

F.4.1	Textile Dye	

F.4.2 Antifreeze	

F.4.3 Surface Cleaner	

F.4.4 Dish Soap	

F. 4.5 Di shwasher D etergent	

F.4.6 Laundry Detergent (Industrial and Institutional)	

F.4.7 Paint and Floor Lacquer	

F.4.8 Spray Foam Application	

F.4.9 Polyethylene Terephthalate Byproduct	

F.4.10 Ethoxylation Process Byproduct	

F.4.11 Hydraulic Fracturing	

F,5 Summary of Occupational Inhalation Exposures	

F.6 Summary of Weight of the Scientific Evidence Conclusions in Inhalation Exposure

Estimates	

F.7 Antifreeze Modeling Approach and Parameters for Estimating Occupational Inhalation

Exposures	

F.7.1 Model Equations	

F.7.2 Modeling Input Parameters	

F.7.3 Container Size	

F.7.4 Jobs per Day	

F.7.5 Concentration of 1,4-Dioxane in Antifreeze	

F.7.6 Ventilation Rate	

F.7.7 Mixing Factor	

F. 7.8 S aturati on Factor	

F.7.9 Consumer Use Rate of Antifreeze	

F.7.10 Container Fill Rate	

F,8 Laundry Detergent Modeling Approach and Parameters for Estimating Occupational
Inhalation Exposures	

Page 8 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

359	F.8.1 Model Equations	419

360	F.8.2 Model Input Parameters	422

361	F.8.3 Ventilation Rate	424

362	F.8.4 Mixing Factor	424

363	F.8.5 Total Particulate Concentration	424

364	F.8.6 Respirable Particulate Concentration	424

365	F.9 Hydraulic Fracturing Modeling Approach and Parameters for Estimating Occupational

366	Inhalation Exposures	425

367	F.9.1 Model Equations	426

368	F.9.2 Model Input Parameters	430

369	F.9.3 Ventilation Rate	432

370	F.9.4 Mixing Factor	432

371	Appendix G SURFACE WATER CONCENTRATIONS	433

372	G.l Surface Water Monitoring Data	433

373	G. 1.1 Monitoring Data Retrieval and Processing	433

374	G. 1.2 Raw and Finished Drinking Water	434

375	G.2 Surface Water Modeling	435

376	G.2.1 Hydrologic Flow Data	435

377	G.2.2 Facility-Specific Release Modeling	436

378	G.2.3 Aggregate and Probabilistic Modeling	437

379	G.2.3.1 The Fit-For-Purpose Aggregate Surface Water Model	437

380	G.2.3.2 Case Studies to Validate Aggregate Model	440

381	G.2.3.3 The Probabilistic Model	444

382	G.2.3.4 Modeling Ranges of DTD Contributions	446

383	G.2.3.5 Modeling Concentrations in Surface Water from Hydraulic Fracturing	448

384	G.2.4 Assessing Downstream Drinking Water Intakes	450

385	Appendix H GROUNDWATER CONCENTRATIONS AND DISPOSAL PATHWAYS

386	FROM LAND RELEASES	455

387	H. 1 Groundwater Monitoring Data Retrieval and Processing	455

388	H,2 Review of Land Release Permits	455

389	H.3 Landfill Analysis Using DRAS	457

390	H.4 Landfill Analysis Using EPACMTP	460

391	H.5 Surface Impoundment Analysis for the Disposal of Hydraulic Fracturing Produced Water

392	Using DRAS	463

393	Appendix I DRINKING WATER EXPOSURE ESTIMATES	466

394	1.1 Surface Water Sources of Drinking Water	467

395	1.2 Groundwater Sources of Drinking Water	467

396	Appendix J AIR EXPOSURE PATHWAY	468

397	J.l Ambient Air Concentrations and Exposures	468

398	J. 1.1 Ambient Air: Screening Methodologies and Results Summary - Fenceline	468

399	J. 1.2 Ambient Air: IIOAC Methodology and Results for COUs without Site-Specific Data

400	(Hydraulic Fracturing, Industrial, and Institutional Laundry Facilities)	471

401	J. 1.3 Ambient Air: Single Year Methodology (AERMOD)	473

402	J. 1.4 Ambient Air: Multi-Year Analysis Methodology (IIOAC)	476

403	J,2 Inhalation Exposure Estimates for Fenceline Communities	477

404	J.3 Land Use Analysis	478

Page 9 of 484


-------
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

434

435

436

437

438

439

440

441

442

443

444

445

446

447

448

449

450

451

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

J,4 Aggregate Analysis across Facilities	479

LIST OF TABLES

Table 2-1. Additional Categories and Subcategories of COUs and Associated OESs Included in the
Scope of the Draft Supplement Due to the Presence of 1,4-Dioxane Produced as a

Byproduct	41

Table 2-2. Summary of PWS Monitoring Datasets of 1,4-Dioxane Monitoring in PWSs Using

Surface Water as a Source	55

Table 2-3. Summary of Surface Water Concentration Results by OES from Facility-Specific

Modeling of Annual Maximum Releases between 2013 and 2019 for 1 Operating Day

per Year	62

Table 2-4. Summary of Surface Water Concentration Results by OES for Facility-Specific Modeling
of Annual Maximum Releases between 2013 and 2019 for the Maximum Operating

Days per Year	63

Table 2-5. OES-COU Crosswalk for Identified Facilities Releasing to Surface Water	64

Table 2-6. Summary by OES of Data Sources for Releases and Receiving Water Body Flow	65

Table 2-7. Hypothetical Mean Annual Concentrations (|ig/L) for a Range of Annual Release and

Flow Rate Combinations, for a Facility with 250 Days of Release per Year	66

Table 2-8. Occurrence of Facilities for Distributions of Maximum Annual 1,4-Dioxane Release

Amounts and Receiving Water Body Flow	66

Table 2-9. Estimated Surface Water Concentrations (|ig/L) Due to DTD Loading for a Range of

Populations and Hydrologic Flows	67

Table 2-10. Estimated Percent Occurrence of Combinations of Contributing Population to POTWs
and Receiving Water Body Flow, from Combined ICIS-NPDES and 2020 Census

Data	67

Table 2-11. Distribution of Potential Concentrations in Surface Water Resulting from Hydraulic

Fracturing Operations from a Single Site Reporting 1,4-Dioxane as an Ingredient	68

Table 2-12. Aggregate Probabilistic Results Showing Distribution of Total 1,4-Dioxane

Concentration in Surface Water (Release Plus Background)	69

Table 2-13. Potential Groundwater Concentrations (|ig/L) of 1,4-Dioxane Found in Wells within 1

Mile of a Disposal Facility Determined by Using the DRAS Model	79

Table 2-14. Total Annual Release Summary	82

Table 2-15. Summary of Select Statistics for the 95th Percentile Estimated Annual Average
Concentrations from the "Full-Screening" Analysis for 1,4-Dioxane Releases

Reported to TRI	86

Table 3-1. Estimated Dermal Absorbed Dose (mg/day) for Workers in Various Conditions of Use	96

Table 3-2. Summary of the Weight of the Scientific Evidence for Occupational Exposure Estimates

by OES	98

Table 3-3. Adult and Infant Exposures Estimated from Facility-Specific Releases	104

Table 3-4. Adult LADD Exposures (mg/kg/day) Estimated from 1,4-Dioxane DTD Consumer and

Commercial Releases	106

Table 3-5. Adult ADR, ADD, and LADD Exposures Estimated from Disposal of Hydraulic

Fracturing Produced Waters to Surface Water	107

Table 3-6. Adult LADD Exposures from Aggregate Concentrations Estimated Downstream of

Release Sites (Including DTD Releases and Direct and Indirect Industrial Releases)... 107
Table 3-7. Adult LADD Exposures Estimated from Groundwater Contamination from Landfills

under Varying Landfill Conditions	109

Page 10 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Table 3-8. Estimated Exposures Resulting from Groundwater Contamination from Disposal of

Hydraulic Fracturing Produced Water	109

Table 3-9. Lifetime Average Daily Concentrations Estimated within 10 km of 1,4-Dioxane Releases

to Air	Ill

Table 3-10. Exposures from Fugitive Emissions Estimated within 1,000 m of Hydraulic Fracturing

Operations	113

Table 3-11. Exposures from Fugitive Emissions Estimated near Industrial and Institutional Laundry

Facilities	115

Table 4-1. Hazard Values Used for 1,4-Dioxane in this Draft Supplement	130

Table 5-1. Use Scenarios, Populations of Interest, and Toxicological Endpoints Used for Acute and

Chronic Exposures	135

Table 5-2. Lifetime Cancer Risk Estimates for 1,4-Dioxane Concentrations Detected in Finished

Drinking Water	141

Table 5-3. Proximity of Nearest Downstream Drinking Water Intakes to Facilities Resulting in

Cancer Risk Greater than 1 x 10~6	143

Table 5-4. Lifetime Cancer Risk Estimates from DTD Releases Alone (at the Point of Release) under

a Range of Population and Flow Rate Scenarios	147

Table 5-5. Lifetime Cancer Risks Estimated from Hydraulic Fracturing Produced Waters Disposed .. 148
Table 5-6. Lifetime Cancer Risks Estimated for Modeled Groundwater Concentrations Estimated

under Varying Landfill Conditions	153

Table 5-7. Lifetime Cancer Risks Estimated for Modeled Groundwater Concentrations Resulting

from Disposal of Hydraulic Fracturing Produced Water	154

Table 5-8. Inhalation Lifetime Cancer Risksa within 10 km of Industrial Air Releases Based on 95th

Percentile Modeled Exposure Concentrations	156

Table 5-9. Lifetime Cancer Risk Estimates for Fugitive Emissions from Hydraulic Fracturing	160

Table 5-10. Lifetime Cancer Risk Estimates for Fugitive Emissions from Industrial and Institutional

Laundry Facilities	161

Table 5-11. Summary of PESS Considerations Incorporated throughout the Analysis and Remaining

Sources of Uncertainty	162

LIST OF FIGURES

Figure 1-1. 1,4-Dioxane Life Cycle Diagram	29

Figure 1-2. Production of 1,4-Dioxane as a Byproduct and Potential Exposure Pathways	30

Figure 1-3. Conceptual Model for Occupational Exposures from Industrial and Commercial

Activities	33

Figure 1-4. Conceptual Model for Environmental Releases and General Population Exposures	35

Figure 1-5. Overview of Analyses Included in this Draft Supplement to the Risk Evaluation for 1,4-

Dioxane	39

Figure 2-1. Overview of EPA's Approach to Estimate Daily Releases for Each OES	42

Figure 2-2. 1,4-Dioxane Annual Water Releases as Reported to TRI and DMR, 2013-2019	46

Figure 2-3. 1,4-Dioxane Annual Releases to Land as Reported to TRI, 2013-2019	47

Figure 2-4. 1,4-Dioxane Annual Releases to Air as Reported by TRI, 2013-2019	48

Figure 2-5. Locations of Hydraulic Fracturing Operations that Report 1,4-Dioxane in Produced

Waters	49

Figure 2-6. Frequency of Nationwide Measured 1,4-Dioxane Surface Water Concentrations Retrieved

from the Water Quality Portal, 1997-2022	 53

Figure 2-7. Frequency of Detection Limits for Nationwide Non-detect 1,4-Dioxane Surface Water

Samples Retrieved from the Water Quality Portal, 1997-2022 	 54

Page 11 of 484


-------
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

536

537

538

539

540

541

542

543

544

545

546

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Figure 2-8. Detectable Concentrations of 1,4-Dioxane in Surface Water from the Water Quality

Portal, 1997-2022	 54

Figure 2-9. Frequency of 1,4-Dioxane Concentrations Monitored in Raw (Untreated) Drinking Water

Derived from Surface Water	56

Figure 2-10. Frequency of 1,4-Dioxane Concentrations Monitored in Finished (Treated) Drinking

Water Derived from Surface Water	56

Figure 2-11. Map of Counties Containing PWSs that Reported Monitoring of Finished Drinking

Water Drawn from Surface Water for 1,4-Dioxane under UCMR3	57

Figure 2-12. Schematic of the EWISRD-XL Model Inputs and Outputs	59

Figure 2-13. Distributions of Surface Water Concentrations Estimated by Aggregate Probabilistic

Model for Each OES	70

Figure 2-14. Case Study Comparison of Modeled and Monitored Concentrations in Brunswick

County	72

Figure 2-15. Frequency of Nationwide Detected 1,4-Dioxane Groundwater Concentrations (n =

2,284) Retrieved from the Water Quality Portal, 1997-2022	 74

Figure 2-16. Detectable Concentrations of 1,4-Dioxane in Groundwater from the Water Quality

Portal, 1997-2022	 74

Figure 2-17. Groundwater Concentrations of 1,4-Dioxane vs. Sample Collection Date for Data

Collected between 1997 and 2022	75

Figure 2-18. Brief Description of Methodologies and Analyses Used to Estimate Ambient Air

Concentrations and Exposures	83

Figure 3-1. Potential Human Exposure Pathways to 1,4-Dioxane for the General Populationa	102

Figure 5-1. Distribution of Adult Lifetime Cancer Risk across all Facilities, Assuming No Additional
Dilution Occurs between the Point of Release and the Location of Drinking Water

Intakes	142

Figure 5-2. Distribution of Adult Lifetime Cancer Risk across Facilities with High Quality Release
Data, Assuming No Additional Dilution Occurs between the Point of Release and the

Location of Drinking Water Intakes	143

Figure 5-3. Distribution of Dilution of 1,4-Dioxane Concentrations at Downstream Drinking Water

Intakes	144

Figure 5-4. Distribution of Adult Lifetime Cancer Risk across all Facilities, Assuming Dilution to 1%

of Initial Concentrations in the Receiving Water Body	144

Figure 5-5. Histograms of Lifetime Cancer Risk Estimates for Aggregate Water Concentrations

Estimated Downstream of COUs with Vertical Lines Showing the Median and 95th
Percentile (P95) Values	150

LIST OF APPENDIX TABLES

TableApx C-l. Terminology Clarifications between the 2021 Draft Systematic Review Protocol and

the 1,4-Dioxane Systematic Review Protocol	185

Table_Apx C-2. Evaluation Criteria for Sources of Monitoring Data	197

Table Apx C-3. Evaluation Criteria for Sources of Experimental Data	204

Table Apx C-4. Evaluation Criteria for Sources of Database Data	210

Table Apx C-5. Considerations that Inform Evaluations of the Strength of the Evidence	214

Table Apx C-6. Evaluation of the Weight of the Scientific Evidence for Exposure Assessments	216

TableApx D-l. Categories and Subcategories of Conditions of Use Included in the Scope of the

Risk Evaluation	223

Table Apx E-l. Summary of EPA's Estimates for the Number of Facilities for Each OES	226

Page 12 of 484


-------
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

584

585

586

587

588

589

590

591

592

593

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TableApx E-2. Summary of EPA's Estimates for Air and Water Release Days Expected for Each

OES	230

TableApx E-3. Summary of Daily Industrial and Commercial Water Release Estimates for 1,4-

Dioxane	238

Table Apx E-4. Summary of Weight of the Scientific Evidence Conclusions in Water Release

Estimates by OES	243

Table Apx E-5. Summary of Daily Industrial and Commercial Land Release Estimates for 1,4-

Dioxane	260

Table Apx E-6. Summary of Weight of the Scientific Evidence Conclusions in Land Release

Estimates by OES	264

Table Apx E-7 Summary of Daily Industrial and Commercial Air Release Estimates for 1,4-

Dioxane	280

Table Apx E-8 Summary of Weight of the Scientific Evidence Conclusions in Air Release Estimates

by OES	285

Table Apx E-9. Comparison of TRI/DMR Release Data to LCA Study for PET Byproduct	292

Table Apx E-10. Summary of Overall Weight of the Scientific Evidence Conclusions for

Environmental Release Estimates by OES	297

Table_Apx E-l 1. TRI-CDR Use Code Crosswalk	300

Table Apx E-12. Summary of Parameter Values and Distributions Used in the Textile Release

Model	327

Table_Apx E-13. Triangular Distributions Ffixation	331

Table Apx E-14. Summary of Parameter Values and Distributions Used in the Industrial and

Institutional Laundry Release Model	340

TableApx E-15. Summary of Parameter Values and Distributions Used in the Hydraulic Fracturing

Release Model	352

TableApx F-l. Summary of Total Number of Workers and ONUs Potentially Exposed to 1,4-

Dioxane for Each Supplemental OESa	358

Table Apx F-2. Glove Protection Factors for Different Dermal Protection Strategies from ECETOC

TRA v3	360

Table Apx F-3. Textile Dye Worker Exposure Data Evaluation	365

Table Apx F-4. Inhalation Exposures of Workers for the Use of Textile Dye Based on Monitoring

Data	367

Table Apx F-5. Occupational Inhalation Monitoring Data for Textile Dyes	368

Table_Apx F-6. Antifreeze Data Source Evaluation	370

Table Apx F-7. Modeled Occupational Inhalation Exposures for Antifreeze	371

Table Apx F-8. Inhalation Exposures of Workers for the Use of Antifreeze Based on Modeling	371

Table Apx F-9. Surface Cleaner Worker Exposure Data Evaluation	373

Table Apx F-10. Inhalation Exposures of Workers for the Use of Surface Cleaner Based on

Monitoring Data	374

Table Apx F-l 1. Occupational Inhalation Monitoring Data for Surface Cleaner	375

Table Apx F-12. Dish Soap Worker Exposure Data Evaluation	377

TableApx F-13. Inhalation Exposures of Workers for the Use of Dish Soap Based on Monitoring

Data	378

Table Apx F-14. Occupational Inhalation Monitoring Data for Dish Soap	379

Table Apx F-15. Laundry Detergent Worker Exposure Data Evaluation	384

Table Apx F-16. Modeled Occupational Inhalation Exposures for Industrial Laundries	385

Table Apx F-17. Modeled Occupational Inhalation Exposures for Institutional Laundries	385

Page 13 of 484


-------
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

623

624

625

626

627

628

629

630

631

632

633

634

635

636

637

638

639

640

641

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TableApx F-18. Inhalation Exposures of Workers for the Use of Laundry Detergent in Industrial

Laundries Based on Modeling	386

Table Apx F-19. Acute and Chronic Inhalation Exposures of Workers for the Use of Laundry

Detergent in Institutional Laundries Based on Modeling	386

Table Apx F-20. Paint and Floor Lacquer Worker Exposure Data Evaluation	388

Table Apx F-21. Inhalation Exposures of Workers for the Use of Paint and Floor Lacquer Based on

Monitoring Data	389

Table Apx F-22. Occupational Inhalation Monitoring Data for Paint and Floor Lacquer	390

Table Apx F-23. Polyethylene Terephthalate (PET) Byproduct Worker Exposure Data Evaluation ... 394
TableApx F-24. Inhalation Exposures of Workers for PET Byproduct Based on Monitoring Data.... 395
Table Apx F-25. Occupational Inhalation Monitoring Data for Polyethylene Terephthalate (PET)

Byproduct	396

Table Apx F-26. Ethoxylation Process Byproduct Worker Exposure Data Evaluation	398

Table Apx F-27. Inhalation Exposures of Workers for the Ethoxylation Process Byproduct Based on

Monitoring Data	399

Table Apx F-28. Occupational Inhalation Monitoring Data for Ethoxylation Process Byproduct	400

Table Apx F-29. Hydraulic Fracturing Worker Exposure Data Evaluation	403

Table Apx F-30. Modeled Occupational Inhalation Exposures for Hydraulic Fracturing	404

Table Apx F-31. Inhalation Exposures of Workers for Hydraulic Fracturing Based on Modeling	404

Table Apx F-32. Estimated Inhalation Exposure (mg/m3) for Workers during Various Conditions of

Use	406

TableApx F-33. Summary of Weight of the Scientific Evidence Conclusions in Inhalation Exposure

Estimates by OES	408

Table Apx F-34. Summary of Parameter Values and Distributions Used in the Antifreeze Exposure

Modeling	415

TableApx F-35. Summary of Parameter Values and Distributions Used in the Laundry Detergent

Exposure Modeling	423

TableApx F-36. Summary of Parameter Values and Distributions Used in the Hydraulic Fracturing

Exposure Modeling	431

Table Apx G-l. Summary of Community Water Systems with Treatment Processes Capable of

Removing 1,4-Dioxane	435

Table Apx G-2. Summary of per Capita DTD Loading Estimates from SHEDS-HT Modeling	439

Table Apx G-3. Summary of Case Study Locations Including Modeled and Observed Surface Water

Concentrations	440

Table Apx G-4. Distribution of per Capita DTD Loading, in G/Day, by Product, for Non-commercial

Uses Modeled by SHEDS-HT	447

Table Apx G-5. Proportions of Population Expected to Contribute to DTD Loading through

Commercial Activities and Product Uses	447

Table Apx G-6. Summary of Proximity of Downstream Drinking Water Intakes to Releasing

Facilities Resulting in Modeled Risk above 1E-06	452

Table Apx G-l. Ranges of Dilution and Diluted 1,4-Dioxane Concentrations Modeled at Drinking

Water Intakes Downstream of Industrial Releases	453

Table Apx G-8. Ranges of LADD and Adult Lifetime Cancer Risk Estimates for Diluted 1,4-

Dioxane Concentrations Modeled at Drinking Water Intakes Downstream of Industrial

Releases	454

Table Apx H-l. Release Year, TRI Facility ID, Facility Name, State, Registry Number, Disposal
Type, and Disposal Weight for On-Site Class I Underground Injection Wells
According to TRI	455

Page 14 of 484


-------
642

643

644

645

646

647

648

649

650

651

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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TableApx H-2. Release Year, Source TRI Facility ID, Source State, Receiving Facility RCRA ID,
State, Disposal Type, and Disposal Weight for Off-Site Class I Underground Injection

Wells According to TRI and RCRAInfo Databases	456

Table Apx H-3. Release year, TRI Facility ID, Facility Name, State, CERCLIS ID, Disposal Type,

and Disposal Weight for RCRA Subtitle C Landfills According to TRI	456

Table Apx H-4. Release Year, Source TRI Facility ID, Source State, Receiving Facility RCRA ID,

State, Disposal Type, and Disposal Weight for Off-Site Class I Underground Injection

Wells According to TRI and RCRAInfo Databases	456

Table_Apx H-5. Input Variables for Chemical of Concern	458

Table_Apx H-6. Waste Management Unit (WMU) Properties	459

Table Apx H-7. Potential Groundwater Concentrations (mg/L) Based on Disposal of 1,4-Dioxane to

Unlined and Clay-Lined Landfills as Assessed by Applying the EPACMTP Model.... 461

Table Apx H-8. Input Variables for Chemical of Concern	463

Table_Apx H-9. Waste Management Unit	465

Table Apx J-l. Release Estimates from 2019 TRI Used for Ambient Air: Screening Methodology for

1,4-Dioxane	469

Table Apx J-2. Exposure and Risk Estimates from the Ambient Air: Screening Methodology for 1,4-

Dioxane Releases Reported to TRI	471

TableApx J-3. Exposure Scenarios and Inputs Utilized for Pre-screening Analysis of Hydraulic

Fracturing, Industrial Laundry, and Institutional Laundry COU	472

Table Apx J-4. Description of Daily or Period Average and Air Concentration Statistics	476

Table Apx J-5. Summary of Fenceline Community Exposures Expected Near Facilities Where

Modeled Air Concentrations Indicated Risk for 1,4-Dioxane	478

Table_Apx J-6. Summary of Groups of Facilities Considered in Aggregate Analysis	482

LIST OF APPENDIX FIGURES

FigureApx C-l. Overview of the TSCA Risk Evaluation Process with Identified Systematic Review

Steps	183

Figure Apx C-2. Literature Inventory Tree - Environmental Releases and Occupational Exposure

Search Results for 1,4-Dioxane	193

Figure Apx C-3. Literature Inventory Tree - General Population, Consumer, and Environmental

Exposure Search Results for 1,4-Dioxane	194

Figure_Apx D-l. COU and OES Mapping	222

FigureApx E-l. Flowchart of a Monte Carlo Method Implemented in a Microsoft Excel-Based

Model Using a Monte Carlo Add-In Tool	319

Figure Apx E-2. Environmental Release and Occupational Exposure Points during Textile Dying .... 324
Figure Apx E-3. Environmental Release and Occupational Exposure Points during

Industrial/Institutional Laundering Operations	332

Figure Apx E-4. Environmental Release and Occupational Exposure Points during Hydraulic

Fracturing	346

Figure Apx F-l. Environmental Release and Occupational Exposure Points during Textile Dying .... 364
Figure Apx F-2. Environmental Release and Occupational Exposure Points during

Industrial/Institutional Laundering Operation	382

Figure Apx F-3. Environmental Release and Occupational Exposure Points during Hydraulic

Fracturing	402

Figure Apx F-4. Environmental Release and Occupational Exposure Points during

Industrial/Institutional Laundering Operations	418

Page 15 of 484


-------
689

690

691

692

693

694

695

696

697

698

699

700

701

702

703

704

705

706

707

708

709

710

711

712

713

714

715

716

717

718

719

720

721

722

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

FigureApx F-5. Environmental Release and Occupational Exposure Points during Hydraulic

Fracturing	425

Figure Apx G-l. Example Raw and Finished Water Concentrations from a PWS without Processes

to Remove 1,4-Dioxane	434

Figure Apx G-2. Schematic of the General Fit-for-Purpose EWISRD-XL Model	438

Figure_Apx G-3. Map of Brunswick County, NC Model Case Study	441

Figure Apx G-4. Plot Comparing Results from Brunswick County Case Study Modeling with

Observed Concentrations	442

Figure_Apx G-5. Map of the Columbia, TN, Case Study	443

Figure_Apx G-6. Map of the East Liverpool, OH, Case Study	444

Figure Apx G-l. Schematic of the Flow of Data within the EWISRD-XL-R Probabilistic Model	446

Figure Apx G-8. Distribution of Mean Annual Modeled Flow Rates for NHDPlus V2.1 Reaches

Identified within 5 km of Hydraulic Fracturing Wells Reporting 1,4-Dioxane	449

Figure Apx G-9. Distribution of Modeled Ranges of 1,4-Dioxane Concentrations in Streams near

Hydraulic Fracturing Wells Reporting 1,4-Dioxane	450

Figure Apx G-10. Generic Schematic of Hypothetical Release Point with Surface Water Intakes for

Drinking Water Systems Located Downstream	451

FigureApx G-l 1. Summary Distribution of Mean Annual Flow at Stream Reaches Matched with

Drinking Water Intakes	453

Figure Apx J-l. Summary of Methodologies Used to Estimate Ambient Air Concentrations and

Exposures	468

Figure Apx J-2. Exposure Scenarios Modeled for Max and Mean Release Using IIOAC Model for

Ambient Air: Screening Methodology	470

Figure Apx J-3. Modeled Receptor Locations for Finite Distance Rings	474

Figure_Apx J-4. Modeled Receptor Locations for Area Distance	475

Figure Apx J-5. Example of Group of Air Releasing Facilities with Overlapping 10 km Buffers for

Aggregate Air Risk Screening	479

Figure Apx J-6. Decision Tree for Characterizing Aggregate Air Risk for Multiple Facilities	481

Figure_Apx J-7. Map of Aggregated Air Facilities, Group 1	482

Figure_Apx J-8. Map of Aggregated Air Facilities, Group 2	483

Figure_Apx J-9. Map of Aggregated Air Facilities, Group 3	483

Figure_Apx J-10. Map of Aggregated Air Facilities, Group 4	484

Figure_Apx J-l 1. Map of Aggregated Air Facilities, Group 5	484

Page 16 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

ACKNOWLEDGEMENTS

This report was developed by the United States Environmental Protection Agency (U.S. EPA or the
Agency), Office of Chemical Safety and Pollution Prevention (OCSPP), Office of Pollution Prevention
and Toxics (OPPT).

Acknowledgements

The Assessment Team gratefully acknowledges the participation, input, and review comments from
OPPT and OCSPP senior managers and advisors. Acknowledgement is also given for the contributions
of interagency reviewers that included multiple federal agencies and assistance provided from EPA
contractors ERG (Contract No. 68HERD20A0002), ICF (Contract No. EP-W-12-010), and Versar
(Contract No. EP-W-17-006). Special acknowledgement is given for the contributions of technical
experts from EPA's Office of Research and Development, including Daniel Dawson for initial
development of the 1,4-dioxane water model and Caroline Ring for input on probabilistic modeling
approaches for the water pathway.

Docket

Supporting information can be found in the public docket (Docket ID: EPA-HQ-QPPT-2022-0905Y
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

Jeffrey Morris (Division Director), Yvette Selby-Mohamadu (Deputy Division Director), Rochelle
Bohaty (Branch Chief/Management Lead), Susanna Wegner (Assessment Lead), Sarah Au, Rebecca
Feldman, Mark Gibson, Bryan Groza, Franklyn Hall, Lauren Knapp Shannon Rebersak, Shawn Shifflett,
Adam Theising, Jason Todd, Kevin Vuilleumier, Cindy Wheeler, Daniel Whitby

Executive Team

This draft risk evaluation was reviewed by OPPT and OCSPP leadership, including senior advisors Stan
Barone, Jeff Dawson, Anna Lowit, and Ryan Schmit and senior leaders, including Mark Hartman
(Deputy Office Director, OPPT) and Denise Keehner (Office Director, OPPT).

Internal Reviewers

This assessment was provided for review to scientists in EPA's Program and Region Offices. Comments
were submitted by:

•	Office of the Administrator/Office of Children's Health Protection

•	Office of Air and Radiation

•	Office of Chemical Safety and Pollution Prevention/Office of Pesticide Programs

•	Office of General Council

•	Office of Land and Emergency Management

•	Office of Research and Development

•	Office of Water

Page 17 of 484


-------
767

768

769

770

771

772

773

774

775

776

777

778

779

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

External Reviewers

This assessment was provided for review to other federal agencies and Executive Offices of the
President, including:

•	Consumer Product Safety Commission

•	Department of Defense

•	Department of Health and Human Services/National Institute of Environmental Health Sciences

•	Department of Health and Human Services/National Institute for Occupational Safety and Health

•	Department of Justice

•	Department of Labor/Occupational Safety and Health Administration

•	Executive Office of the President/Office of Management and Budget

•	Food and Drug Administration

•	National Aeronautics and Space Administration

•	Small Business Administration Office of Advocacy

Page 18 of 484


-------
780

781

782

783

784

785

786

787

788

789

790

791

792

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Summary of Risk Findings and Support for Risk Determination

1,4-Dioxane is a solvent used in a variety of commercial and industrial applications in the United
States. It is also produced as a byproduct in several manufacturing processes and may remain present
as a byproduct in consumer and commercial products, including soaps, detergents, and cleaning
products. Health effects of concern for 1,4-dioxane include cancer and effects in liver and olfactory
tissue. People may be exposed to 1,4-dioxane through occupational exposure, consumer products, or
contact with water, land, or air where 1,4-dioxane has been released to the environment from
industrial and commercial sources or from consumer and commercial products washed down the
drain or disposed of in landfills.

The risk evaluation for 1,4-dioxane published in 2020 evaluated risks from a range of occupational
and consumer uses of 1,4-dioxane, risks to aquatic species, and risks to the general population
resulting from incidental recreational contact with water. It did not evaluate general population
exposures to 1,4-dioxane in drinking water or air and did not evaluate the full range of exposure that
may result from 1,4-dioxane produced as a byproduct.

This draft supplement completes the Toxics Substances Control Act (TSCA) risk evaluation for 1,4-
dioxane by (1) more comprehensively evaluating risks from 1,4-dioxane present as a byproduct; and
(2) evaluating risks from general population exposures to 1,4-dioxane released to water, air, and
land. This analysis identified cancer risk estimates higher than 1 in 10,000 (1 x 10~4) for a range of
typical and high-end occupational exposures to 1,4-dioxane produced as a byproduct. It also
identified cancer risk estimates higher than 1 in 1 million (1 x 10~6) for a range of general population
exposure scenarios associated with 1,4-dioxane in drinking water sourced downstream of release
sites and in air within 1 km of releasing facilities. Although these risk estimates include inherent
uncertainties and the overall confidence in specific risk estimates varies, the analysis provides
support for the Agency to make a determination about whether 1,4-dioxane poses an unreasonable
risk and to identify drivers of unreasonable risk among exposures for people (1) with occupational
exposure to 1,4-dioxane under some conditions of use, (2) who rely on sources of drinking water
downstream of release sites, and (3) breathing air near release sites.

Subsequent to this draft supplement, EPA is releasing a draft revised risk determination for 1,4-
dioxane. The updated risk determination considers the results presented in this draft supplement as
well as those published in the 2020 Risk Evaluation for 1,4-Dioxane.

EXECUTIVE SUMMARY	

This draft document is a supplement to the Final Risk Evaluation for 1,4-Dioxane. published December
2020 (2020 RE). EPA conducted this supplemental analysis because contrary to the law's requirement
for TSCA risk evaluations to be carried out on the "chemical substance" under the conditions of use
(COUs), the 2020 RE excluded certain known human exposure pathways that are important to
understanding the health implications of exposure to 1,4-dioxane. This supplement effectively completes
EPA's risk evaluation on the chemical substance and positions the Agency to comprehensively address
identified unreasonable risks.

1,4-Dioxane is primarily used as a solvent in commercial and industrial applications. It can also be
produced as a byproduct of several common manufacturing processes, including but not limited to
ethoxylation processes used in the production of surfactants used in soaps and detergents and production

Page 19 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

of polyethylene terephthalate (PET) plastics. Even though it is not intentionally added, 1,4-dioxane
produced as a byproduct may remain present in consumer and commercial products, including soaps and
detergents, cleaning products, antifreeze, textile dyes, and paints/lacquers. 1,4-Dioxane is released to the
environment from industrial and commercial releases and from consumer and commercial products that
are washed down the drain or disposed of in landfills. People may be exposed to 1,4-dioxane through
occupational exposure, consumer products, or contact with water, land, or air where 1,4-dioxane has
been released to the environment.

The 2020 RE did not evaluate risks from two critical areas: (1) general population exposures to 1,4-
dioxane in drinking water or air, and (2) the full range of exposure that may result from 1,4-dioxane
produced as a byproduct. During review of the draft risk evaluation, peer reviewers and public
commenters raised concerns that failure to consider these exposure pathways could leave portions of the
population at risk. These concerns include the fact that 1,4-dioxane has been detected in drinking water
and is not readily removed through traditional treatment. In addition, 1,4-dioxane produced as a
byproduct results in occupational exposures that were not evaluated in the 2020 RE. Finally, 1,4-dioxane
produced as a byproduct also contributes to 1,4-dioxane in drinking water through industrial releases
and down-the-drain (DTD) disposal of consumer and commercial products.

This draft supplement expands on the analysis of COUs in which 1,4-dioxane is present as a byproduct
to include additional COUs for which information is reasonably available and consider associated
occupational exposures. This draft supplement also evaluates risks to the general population—including
potentially exposed or susceptible subpopulations (PESS)—from exposure to 1,4-dioxane through
drinking water or air resulting from all industrial releases (including those resulting from 1,4-dioxane
produced as a byproduct) as well as DTD releases of consumer and commercial products.

EPA's evaluation of these additional human exposure pathways included new methods and novel
applications of existing methods that will be subject to peer review at a Science Advisory Committee on
Chemicals (SACC) meeting in September 2023. Following review by the public and the SACC, the
Agency will finalize this supplemental risk evaluation. At that time, EPA will initiate steps to address
unreasonable risks identified through its complete evaluation of 1,4-dioxane.

Approach

For this draft supplemental risk evaluation, EPA is relying on the physical and chemical properties,
chemical lifecycle information, environmental fate and transport information, and the hazard
identification and dose-response analysis presented in the 2020 RE. EPA evaluated cancer and non-
cancer risks from occupational and general population exposure scenarios using available modeling
and/or monitoring information. The Agency also considered site-specific exposures, including combined
or additive releases from multiple releasing facilities within a single air or water exposure pathway. This
draft supplement considers PESS throughout the human health exposure assessment and risk
characterization.

Exposure

Occupational Exposure: EPA estimated both high-end and central tendency occupational exposures
through inhalation and dermal absorption. High-end exposure estimates were used because they attempt
to capture potential variability in exposure across facilities and individuals and may be representative of
PESS and "sentinel" exposures. In some cases, high-end estimates reflect uncertainty around the extent
of this variability. EPA estimated occupational exposure for most COUs based on available monitoring
data. For COUs without occupational monitoring data, EPA applied Monte Carlo methods to estimate

Page 20 of 484


-------
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

869

870

871

872

873

874

875

876

877

878

879

880

881

882

883

884

885

886

887

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

exposures using generic scenarios and emission scenario documents. These methods are generally
consistent with Monte Carlo approaches used in previous TSCA risk evaluations.

General Population Exposure: EPA evaluated general population exposures to 1,4-dioxane through
drinking water and air that could result from releases to surface water, groundwater, land, and air. The
Agency evaluated a range of exposure scenarios for each pathway, including (but not limited to) high-
end exposure scenarios. To be protective of PESS and sentinel exposures, EPA developed risk estimates
on the scenarios, populations, and life stages with the highest levels of exposure. For drinking water,
EPA evaluated life stage-specific exposures for adults, formula-fed infants, and children. For air
exposures, because the impacts of lifestage differences could not be quantified adequately, air
concentrations are used for all lifestages. To address exposure to fenceline communities, EPA
considered air exposures within 10 km of a release site. For water releases, EPA considered exposures to
communities relying on drinking water sourced near release sites.

Hazard

All hazard values used in this draft supplement were derived from the points of departure (PODs)
previously peer-reviewed by the SACC and published in the 2020 RE. Some of the exposure scenarios
evaluated in this analysis required duration adjustments to the previously established hazard values;
however, the underlying hazard endpoints and PODs remain the same. Health effects of concern for 1,4-
dioxane include cancer and adverse, non-cancer effects in liver and olfactory tissue.

Risk Characterization

EPA evaluated both cancer and non-cancer risks for each exposure pathway. Because cancer is the
primary risk driver for 1,4-dioxane, results presented here are cancer risk estimates. Overall confidence
in the risk estimates varies across exposure pathways and COUs, depending on the data, models, and
assumptions used. For risk estimates, EPA has medium to high confidence in the underlying hazard
PODs used as the basis for risk characterization. Therefore, exposure-related considerations drive
differences in confidence among risk estimates. Differences in central tendency and high-end risk
estimates may reflect variability in exposure across the population (e.g., due to differences in the
frequency or intensity of occupational exposures) and/or uncertainty in the exposure assessment (e.g.,
due to incomplete information on release amounts or variations in flow rates of receiving water bodies
and/or drinking water intake locations).

Estimates of Occupational Risks: EPA estimated cancer and non-cancer risks for a set of new
occupational COUs where 1,4-dioxane is present as a byproduct.

•	Dermal Exposure: Dermal occupational exposure is expected to occur as a result of worker
activities such as transfer operations, application of 1,4-dioxane containing formulations, and the
cleaning of equipment. COU-specific weight fractions and evaporation drive the variability in
results. Cancer risk estimates for dermal exposures range from 8.1 x 10~7 to 8,6/ 10 4 for central
tendency exposure and from 5,Ox 10 6 to 1.5/10 2 for high-end exposures across COUs. Overall
confidence in risk estimates for occupational dermal exposures is medium for all occupational
exposure scenarios.

•	Inhalation Exposure: Inhalation exposure to 1,4-dioxane is expected to occur as a result of the
scenario-specific considerations described in the bullets below. Cancer risk estimates for
inhalation exposure range from 8.3*10~12 to 1.8x10 3 for central tendency exposures and from
5,4/10 " to 2,3/10 2 for high-end exposures across COUs. Occupational exposure scenarios
with the highest estimates of risk from inhalation exposure are summarized below:

Page 21 of 484


-------
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

926

927

928

929

930

931

932

933

934

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

o Polyethylene Terephthalate (PET) Manufacturing. Workers may inhale 1,4-dioxane
generated as a byproduct of PET plastic manufacturing. Cancer risk estimates for
inhalation exposure range from 1.8x10 3 for central tendency exposures to 2.3/10 2 for
high-end exposures. There is uncertainty regarding the risk estimates because the analysis
relied on decades-old monitoring data and the extent to which the monitoring data reflect
current practices is unknown. Overall confidence in risk estimates for PET plastic
manufacturing is medium.

o Hydraulic Fracturing Operations. 1,4-Dioxane inhalation exposures may occur during
hydraulic fracturing operations due to its documented presence in scale inhibitors,
additives, friction reducers, and surfactants used in fracturing fluid formulations. Cancer
risk estimates for inhalation exposure range from 7,Ox 10 5 for central tendency exposures
to 9.5x10 3 for high-end exposures. There is uncertainty regarding the model inputs used
to estimate exposures and the extent to which they reflect the actual distribution of
hydraulic fracturing occupational exposures and workplace practices. Overall confidence
in risk estimates for hydraulic fracturing operations is medium to high.

o Ethoxylation Processes. 1,4-Dioxane may be generated as a byproduct in ethoxylation
reactions during the manufacture of common surfactants that result in worker inhalation
exposure. Cancer risk estimates for inhalation exposure range from 4.6x 10~4 for central
tendency exposures to 5.9xl0~4 for high-end exposures. There are numerous uncertainties
due to the limited monitoring data, unknown concentrations, and the mass of 1,4-dioxane
generated as a byproduct during ethoxylation. There is also uncertainty in the worker
activities covered by the monitoring data and whether all foreseeable activities,
corresponding exposures, and workplace operations are represented. Overall confidence
in risk estimates for ethoxylation processes is low to medium.

o Industrial/Commercial Use of Dish Soap and Dishwasher Detergent. 1,4-Dioxane
inhalation exposures are expected during the use of dish soap and dishwashing detergent
from unloading and transferring detergent formulation, transport container cleaning, and
washing operations due to the presence of 1,4-dioxane as a surfactant byproduct. Cancer
risk estimates for inhalation exposure range from 4,Ox 10 4 for central tendency exposures
to 1.0x10 3 for high-end exposures. There is uncertainty as to the representativeness of
these estimates due to the age of the monitoring data, number of non-detects, and the
limited sample size. Overall confidence in risk estimates for dish soap and dishwasher
detergent use is low to medium.

Estimates of Risk to the General Population Exposure through Drinking Water Sourcedfrom Surface
Water: EPA estimated cancer and non-cancer risks for a range of general population exposures to
surface water used as drinking water. 1,4-Dioxane is not readily removed through typical wastewater or
drinking water treatment processes. Sources of 1,4-dioxane in surface water include direct and indirect
industrial releases from COUs where 1,4-dioxane is manufactured, processed, or used; industrial COUs
where 1,4-dioxane is present due to production as a byproduct (including PET manufacturing,
ethoxylation processes and hydraulic fracturing operations); and DTD releases of 1,4-dioxane present in
consumer and commercial products.

Monitoring data demonstrates that 1,4-dioxane is present in some source water and finished drinking
water samples. Measured concentrations in finished drinking water samples resulted in cancer risk
estimates greater than 1 in 1 million at the high end of the distribution of monitoring samples. However,
available surface water monitoring datasets are not designed to reflect source water impacts of direct and
indirect releases into water bodies. Therefore, EPA estimated concentrations modeled for a range of

Page 22 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

specific release scenarios. The Agency evaluated the performance of the models against monitoring data
from site-specific locations serving as cases studies. This evaluation demonstrated strong concordance
between modeled concentrations and monitoring data, thereby increasing confidence in risk estimates
based on modeled concentrations.

EPA used modeled water concentrations to evaluate risks from a range of sources individually and in
aggregate {i.e., by evaluating risks from water concentrations resulting from multiple sources of 1,4-
dioxane releasing to the same water bodies). For each of the sources assessed, risk estimates from
modeled concentrations in receiving water bodies at the point of release may be greater than 1 in
1,000,000 or 1 in 100,000 under some conditions.

•	Industrial Releases to Surface Water: Risk from individual facilities vary substantially within and
across COUs, with cancer risk estimates ranging from 5,4/10 13 to 0.025. Overall confidence in
risk estimates for specific facilities depends on confidence in facility-specific release data, but
confidence in the overall analysis is medium-high

•	Down-the-drain Releases to Surface Water: EPA evaluated the conditions under which down-
the-drain releases contribute to different levels of risk and identified plausible scenarios in which
risks from down-the-drain releases result in risks greater than 1 in 1 million. Risk estimates from
modeled down-the-drain releases are highest in locations where large populations contribute to
these releases and where they are discharged to streams with low flow. Overall confidence in this
analysis is medium.

•	Hydraulic Fracturing Releases to Surface Water: Cancer risk estimates from modeled hydraulic
fracturing waste releases to surface water are 3,57/10 8 for median modeled releases and

1,45/10 6 for 95th percentile modeled releases. Overall confidence in this analysis is medium.

•	Aggregate Releases to Surface Water: Probabilistic modeling provides a distribution of risk
estimates reflecting a range of drinking water scenarios that account for aggregate sources of 1,4-
dioxane in water. Overall confidence in risk estimates for specific facilities depends on
confidence in facility-specific release data used as model inputs.

The relative contribution from different sources varies under different conditions and is likely to be
driven by site-specific factors including the amounts released from each source, flow rates of receiving
water bodies, and proximity of releases to drinking water intakes. In locations where industrial releases
are particularly high, industrial releases are likely to drive risk. In locations where large populations
contribute to down-the-drain releases and receiving water bodies have relatively low flow rates, down-
the-drain releases on their own may drive risks. Each of the sources evaluated may contribute to
drinking water risks under some conditions. Furthermore, this analysis illustrates the fact that sources of
1,4-dioxane produced as a byproduct, including those from industrial releases and DTD releases, can
contribute to risks from 1,4-dioxane in water. The analyses in this draft supplement describe the
conditions under which different levels of risk may occur.

Estimates of Risk to the General Population Exposure through Drinking Water Sourcedfrom
Groundwater: EPA estimated cancer and non-cancer risks for a range of general population exposures to
groundwater used as drinking water. Sources of 1,4-dioxane in groundwater include leachate from
landfills and disposal of hydraulic fracturing waste. Down the drain releases to septic fields from
consumer and commercial products containing 1,4-dioxane as well as historical disposals of 1,4-dioxane
are other potential sources of groundwater contamination but were not considered in this assessment.
Overall confidence in these risk estimates is low to medium.

Page 23 of 484


-------
980

981

982

983

984

985

986

987

988

989

990

991

992

993

994

995

996

997

998

999

1000

1001

1002

1003

1004

1005

1006

1007

1008

1009

1010

1011

1012

1013

1014

1015

1016

1017

1018

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

•	1,4-Dioxane in Groundwater from Hydraulic Fracturing: Cancer risk estimates for modeled
groundwater concentrations are 4.Ox 10~7 for median modeled releases and 8,6/ 10 6 for 95th
percentile modeled releases.

•	1,4-Dioxane in Groundwater Resulting from Landfill Leachate: Cancer risk estimates increase
under scenarios with higher leachate concentrations and loading rates. Monitoring data for
groundwater contamination surrounding landfills was not readily available for comparison.

Estimates of Risk to the General Population through Air: EPA estimated cancer and non-cancer risks for
a range of general population exposures to 1,4-dioxane in air. 1,4-Dioxane concentrations in air depend
on the facility-specific release amount, stack height(s), topography, and meteorological conditions—not
on specific COUs.

Potential sources of 1,4-dioxane in air include industrial releases reported to Toxics Release Inventory
(TRI), fugitive emissions from hydraulic fracturing, and emissions from institutional and industrial
laundries. The highest estimated risks occurred within 1,000 m of industrial release sites. EPA also
estimated risk from the aggregate exposures from multiple facilities releasing 1,4-dioxane in proximity
to fenceline communities. This aggregate analysis did not identify locations with aggregate cancer risk
greater than 1 x 10~6 that did not already have cancer risk above that level from an individual facility and
therefore did not have a substantial impact on the overall findings.

1,4-Dioxane concentrations in air depend on the facility-specific release amount, stack height(s),
topography, and meteorological conditions.

•	Air Releases Reported to TRI: Cancer risk estimates for 95th percentile modeled air
concentrations with 1,000 m (approximately 0.6 mile) of the highest risk facilities in each COU
range from 1,0x 10~10 to 1.1x 10~4. Cancer risk estimates based on 50th percentile modeled
exposure concentrations within 1,000 m of the highest risk facilities range from 2.5/10 " to
8.3xl0~5. Although individual risk estimates for specific locations should be interpreted with
caution, most estimates are informed by moderate to robust modeling approaches and input data.
Overall confidence in risk estimates for inhalation exposures resulting for air concentrations
modeled for industrial releases ranges from low to high, depending on the level of confidence in
release information underlying risk estimates for specific facilities and COUs.

•	Fugitive Air Emissions from Hydraulic Fracturing Operations: Cancer risk estimates within
1,000 m of hydraulic fracturing operations range from 2.2x 10~8 to 7.1 / 10 5 for a range of model
air model scenarios across a range of high-end (95th percentile) and central tendency release
scenarios. Overall confidence in risk estimates for inhalation exposures resulting for air
concentrations modeled based on releases from hydraulic fracturing operations is medium.

•	Emissions from Industrial and Institutional Laundries: Cancer risk estimates within 1,000 m of
industrial and institutional laundries range from 1.5 xlO-11 to 3.8 xlO-8 across a range of high-end
exposure scenarios. Overall confidence in risk estimates from inhalation exposures resulting
from industrial and institutional laundries is medium.

Page 24 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

1 INTRODUCTION

1,4-Dioxane is one of the first 10 chemicals undergoing the Toxics Substances Control Act (TSCA) risk
evaluation process after passage of the Frank R. Lautenberg Chemical Safety for the 21st Century Act,
which amended TSCA. 1,4-Dioxane is primarily used as a solvent in a variety of commercial and
industrial applications such as the manufacture of other chemicals (e.g., adhesives, sealants) or as a
processing aid or laboratory chemical. It is produced as a byproduct in several manufacturing processes,
including ethoxylation, sulfonation, sulfation, and esterification. Although there are no direct consumer
uses, 1,4-dioxane produced as a byproduct in the aforementioned processes can be present in
commercial and consumer products, including soaps, detergents, and cleaning products. Use of these
products may result in direct occupational and consumer exposures. Disposal of these products down-
the-drain (DTD) may contribute to general population exposure to 1,4-dioxane present in some surface
water.

1.1 Regulatory Context

In the 2019 draft 1,4-dioxane risk evaluation, EPA reviewed the exposures, hazards, and risks of 1,4-
dioxane from occupational exposures and surface water exposures to environmental organisms. It also
included the physical and chemical properties, lifecycle information, environmental fate and transport
information, and hazard identification and dose-response analysis. However, the 2019 draft risk
evaluation excluded general population exposures through drinking water and air and conditions of use
(COUs) in which 1,4-dioxane is present as a byproduct. These exclusions were based in part on an
interpretation that EPA had broad discretionary authority under TSCA to categorically exclude
conditions of use from the scope of its evaluations, and, as described in the 2018 1,4-Dioxane Problem
Formulation} that certain exposure pathways need not be considered if they were under the jurisdiction
of other EPA regulatory programs or analytical processes.

These analyses were reviewed by the Science Advisory Committee on Chemicals (SACC) in 2019. The
SACC raised a number of concerns regarding the evaluation and approach, but particularly noted its
concerns about the Problem Formulation straying from "basic principles of risk assessment," the
omission of well-known exposure routes, and that general lack of comprehensiveness undermining
EPA's ability to protect against risks to human health and the environment. As stated in the meeting
minutes and final report2 from the July 2019 SACC meeting, "there was general dissatisfaction in the
Committee that the human health risk characterization did not extend to the general population since
there was no indication in the Evaluation that other offices in the EPA had plans to conduct such a
characterization." Furthermore, "several committee members also observed that failure to assess 1,4-
dioxane exposure in the general population may leave substantial portions of the population at risk. This
is particularly concerning for drinking water." The SACC also raised concerns about potential risks from
1,4-dioxane produced as a byproduct, recommending that "EPA should provide a detailed discussion of
the scientific basis for the exclusion of impurity or byproduct formation of 1,4-dioxane."

Public stakeholders also raised concerns about water monitoring data demonstrating the presence of 1,4-
dioxane in drinking water. Commenters also identified additional sources of 1,4-dioxane that had not
been assessed, including COUs in which 1,4-dioxane is produced as a byproduct, and raised concerns
that some of these COUs may be an important source of exposure to 1,4-dioxane in water.

1	The 1,4-Dioxane Problem Formulation is available at https://www.epa.. gov/assessing-and-managing-chemicals-under-

tsca/14-dioxane-problem-fonnulation.

2	The SACC July 2019 meeting minutes and final report (Document ID EPA-HQ-OPPT-2019-0237-0064) are available in the
docket at https://www.regulations.gov/document/EPA-HO-OPPT-2Q19-uzj /~wo4.

Page 25 of 484


-------
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

1088

1089

1090

1091

1092

1093

1094

1095

1096

1097

1098

1099

1100

1101

1102

1103

1104

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

In November of 2020, EPA released a supplement to the draft 1,4-dioxane risk evaluation for public
comment. The November 2020 supplement assessed eight additional COUs of 1,4-dioxane as a
byproduct in consumer products and general population exposure from incidental contact with surface
water. The EPA determined that the additional analysis did not warrant SACC review as no novel
science was utilized.

In December 2020, the Agency published the Final Risk Evaluation for 1,4-Dioxane (	)20c)

(referred hereafter as the "2020 RE") (	20c). The 2019 draft and 2020 supplement were

both incorporated into the 2020 RE, which assessed risks for

•	worker and occupational non-user (ONU) exposures to 1,4-dioxane through 16 industrial and
commercial COUs;

•	consumer and bystander exposures to 1,4-dioxane present as a byproduct3 in eight consumer
product categories;

•	general population exposure via incidental/recreational contact with 1,4-dioxane present in
surface water from industrial releases; and

•	aquatic species' exposures to 1,4-dioxane present in surface water.

In January 2021, the White House issued Executive Order 13990 instructing that the federal government
be guided by the best science and be protected by processes that ensure the integrity of Federal decision-
making, and established the Administration's policy of, among other concerns, following the science,
improving public health and protecting the environment, limiting exposure to dangerous chemicals, and
prioritizing environmental justice when delivering on these concerns. Executive Order 13990 also
instructs agencies to (1) review actions issued between January 20, 2017, and January 20, 2021, which
may be inconsistent with or present obstacles to implementing the policy established in the order and;
(2) consider suspending, revising, or rescinding such actions.

Upon further review, EPA determined that the approach taken in the 2020 RE {i.e., the exclusion of
reasonably foreseeable exposures to workers, as well as exposures to the general population from air,
water, and disposal) was inconsistent with the plain language of TSCA section 6 and left potential
risks—including risks to potentially exposed and susceptible subpopulations (PESS)—unaccounted for.
The law's requirement that EPA conduct risk evaluation on a "chemical substance" under the COUs
requires the Agency to determine the chemical's COUs and to not otherwise exclude those COUs from
the scope of the risk evaluation.

In June of 2021, EPA announced that additional analysis was needed to consider critical exposure
pathways not assessed in the final risk evaluations for the first 10 chemicals (including, but not limited
to, ambient air, ambient water, and drinking water). For many of the first 10 risk evaluations, EPA
applied the Draft Screening Level Approach for Assessing Ambient Air and Water Exposures to
Fenceline Communities Version 1.0,4 published in January 2022 to determine whether further analysis
was needed. For 1,4-dioxane, however, EPA determined that a more in-depth analysis was needed to
address concerns about known drinking water contamination (described in Section 1.3.1.3) and to more
fully evaluate COUs in which 1,4-dioxane is present as a byproduct (described in Section 1.3.1.1), and
signaled its intention to re-open and formally supplement the 1,4-dioxane risk evaluation.

3	Byproduct means a chemical substance produced without a separate commercial intent during the manufacture, processing,
use, or disposal of another chemical substance(s) or mixture(s).

4	The draft Screening Level Approach for Assessing Ambient Air and Water Exposures to Fenceline Communities Version 1.0
is available at https://www.epa.gov/assessing~aiMl~managing~chemicals~under~tsca/tsca~screening~level~approach~assessing~
ambient-air-and.

Page 26 of 484


-------
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

1139

1140

1141

1142

1143

1144

1145

1146

1147

1148

1149

1150

1151

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

This draft supplement to the 2020 RE is intended to complete the risk evaluation on the chemical
substance 1,4-dioxane as required under TSCA by (1) expanding the analysis of COUs in which 1,4-
dioxane is present as a byproduct to include additional COUs and consider associated occupational
exposures; and (2) evaluating risks from general population exposures to 1,4-dioxane released to surface
and groundwater, air, and land.

EPA's evaluation of these additional human exposure pathways included new methods and novel
applications of existing methods. This draft supplement is the first under amended TSCA to evaluate:
exposures and risks from a chemical produced as a byproduct, aggregate risks for communities relying
on drinking water sourced from surface water receiving a chemical from multiple sources, risks for
communities relying on drinking water sourced from groundwater, aggregate risks for communities
exposed through air near multiple release sites, and consideration of multiple years of environmental
release data.

1.2 Scope

This draft supplement is intended to evaluate risks from exposure pathways and COUs for 1,4-dioxane
that were not assessed in the 2020 RE. Additional exposure pathways and new COUs included in this
draft supplement were identified based on information submitted in previous public comments and other
reasonably available information. For the current analysis, EPA is relying on the physical and chemical
properties, as well as lifecycle information, environmental fate and transport information, and hazard
identification and dose-response analysis presented in the 2020 RE (Sections 1.1, 1.4, 2.1, and 3.2 of the
2020 RE, respectively). Furthermore, this supplement does not reevaluate the occupational, consumer,
or ecological exposure pathways and risks that were previously assessed in the 2020 RE.

This draft supplement to the 2020 RE evaluates risks for the following exposure pathways:

•	Occupational exposure to

o 1,4-dioxane present as a byproduct in commercial products (corresponding to consumer

products considered in the 2020 RE); and
o 1,4-dioxane produced or present as a byproduct in additional industrial COUs for which
information on the presence of 1,4-dioxane is reasonably available, including
ethoxylation processing, polyethylene terephthalate (PET) manufacturing, and hydraulic
fracturing (Sections 3.1, 5.2.1).

•	General population exposures to

o 1,4-dioxane present in drinking water sourced from surface water as a result of all direct
and indirect industrial releases and DTD releases of consumer and commercial products
(Sections 2.3.1, 3.2.2 and 5.2.2.1);
o 1,4-dioxane present in drinking water sourced from groundwater contaminated as a result

of disposals (Sections 2.3.2, 3.2.2.2 and 5.2.2.1.6); and,
o 1,4-dioxane released to air from industrial and commercial sources (Sections 2.3.3, 3.2.3,
and 5.2.2.3).

Many of the COUs assessed in this draft supplement contribute to more than one exposure pathway. For
example, 1,4-dioxane present as a byproduct of PET manufacturing may contribute to occupational
exposures during manufacturing as well as general population exposures through releases to water and
air. In addition, for many of the exposure pathways assessed, multiple COUs contribute to 1,4-dioxane
exposure. For example, many COUs can contribute to general population exposures to 1,4-dioxane in
surface water, including industrial releases from a range of COUs and DTD releases of consumer and
commercial products. In this draft supplement, EPA evaluated general population exposures resulting
from all known releases, including releases associated with COUs evaluated in the 2020 RE and releases

Page 27 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

1152	associated with new COUs introduced in this draft supplement due to the presence of 1,4-dioxane

1153	produced as a byproduct.

1154	1.3 Use Characterization

1155	1.3.1 Conceptual Models

1156	The life cycle diagram for 1,4-dioxane in Figure 1-1 summarizes the conditions of use that are within the

1157	combined scope of the 2020 RE and the current draft supplement. The life cycle diagram has been

1158	updated from the 2020 RE to highlight additional sources of 1,4-dioxane produced as a byproduct,

1159	including commercial products and industrial uses, releases, and disposals (e.g., PET manufacturing,

1160	ethoxylation byproducts, disposal of hydraulic fracturing produced waters).

Page 28 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

MFG/IMPORT	PROCESSING	INDUSTRIAL, COMMERCIAL,	RELEASES and

CONSUMER USES	WASTE DISPOSAL

Manufacture

(includes import)
(i million lbs.)

Processing as a
Rea eta nt/lnter mediate

(Not reported in 2016 CDR)

Industrial Uses

Repackaging

(270,000 lbs.)

Non-incorporative
Activities

(270,00 lbs.)

Recycling

Processing Aids, Not
Otherwise Listed

(270,000 lbs.)
e.g., wood pulping, etching
of fluropolymers

Functional Fluids
(Open and Closed Systems)
(<150,000 lbs.)
e.g., hydraulic fluid.

Laboratory Chemicals

(<150,000 lbs.)
e.g., laboratory reagent

Adhesive Sealants

e.g., film cement

Other Industrial Uses
e.g., spray polyurethane
foam, printing and printing
compositions; dry film
lubricant

Consumer Uses
{Present only as Byproduct)

Paints and Coatings

e.g., Latex Wall Paint or
Floor Lacquer

Cleaning and Furniture
Care Products

e.g.. Surface Cleaner

Laundry and Dishwashing
Products

e.g., Dish Soap, Dishwasher
Detergent, Laundry
Pete rgent	

Arts, Crafts and Hobby
Materials

e.g.. Textile Dye

Automotive Care Products

e.g., Antifreeze

Other Consumer Uses
e.g., Spray Polyurethane
Foam, Antifreeze

Commercial Uses
(Present only as Byproduct)

Paints and Coatings

e.g.. Latex Wall Paint or
Floor Lacquer

Clea ning a nid Fu miture
Care Products

e.g., Surface Cleaner

Laundry and Dishwashing
Products

e.g.. Dish Soap, Dishwasher
Detergent, Laundry
Detergent	

Arts, Crafts and Hobby
Materials

e.g.. Textile Dye

Automotive Care Products

e.g.. Antifreeze

Other Consumer Uses
e.g.. Spray Polyurethane
Foam, Antifreeze

Industrial uses including 1,4-dioxane as a byproduct

(e.g., ethoxylation processes, PET manufacturing, and hydraulic fracturing operations)

Disposal

Byproduct
U ses from
Manufacturing

1161

1162	Figure 1-1. 1,4-Dioxane Life Cycle Diagram

1163	Note: This life cycle diagram has been expanded from what was published in the 2020 RE to include additional sources of 1,4-dioxane produced as a

1164	byproduct (indicated in blue boxes). See Appendix D for a complete table of COUs considered in the 2020 RE and the current supplement.

1 Assessed in the Supplement to
	I	the Risk Evaluation

Page 29 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

1.3.1.1 1,4-Dioxane as a Byproduct

1,4-Dioxane produced as a byproduct may be a source of exposure for several of the occupational and
general population exposure pathways evaluated in this draft supplement. Byproduct refers to a chemical
substance produced without a separate commercial intent during the manufacture, processing, use, or
disposal of another chemical substance(s) or mixture(s). In this assessment, the term byproduct is used
to refer to 1,4-dioxane produced during manufacturing or industrial processes, including 1,4-dioxane
that remains present in downstream processes or in consumer and commercial products.

In the 2020 RE, EPA evaluated risks to consumers and bystanders from 1,4-dioxane present as a
byproduct in consumer products. In this draft supplement, EPA expanded on the previous evaluation to
consider risks from all other pathways of exposure to 1,4-dioxane produced as a byproduct for which
information is reasonably available. Figure 1-2 summarizes both what is known about the processes that
may result in 1,4-dioxane production and how it may contribute to human exposures through a range of
exposure pathways.

Source
chemicals

Occupational
exposure

Chemical
processes

..in many types of
manufacturing
facilities

(ethoxylation,
esterification,
sulfonation,
sulfation)

Direct and indirect*
industrial releases to
air and water

General population
exposure through air
or drinking water

1,4-dioxane
produced as a
byproduct of
manufacturing

'1

Removal of 1,4-dioxane

prior to product
formulation and releases



Consumer and
occupational
exposure a

t

Consumer and
commercial products

Down-the-drain
releases*

*these releases may also go to wastewater treatment plants

Figure 1-2. Production of 1,4-Dioxane as a Byproduct and Potential Exposure Pathways

1,4-Dioxane is produced as a byproduct in several common manufacturing reactions, including in
manufacturing of PET plastics and in ethoxylation reactions during the manufacture of common
surfactants. In some facilities, additional processing steps may remove 1,4-dioxane produced as a
byproduct prior to product formulation and environmental releases, but the full extent of this practice
across industries is not known. Occupational exposure to 1,4-dioxane produced as a byproduct may
occur at manufacturing facilities and hydraulic fracturing operations. Releases of 1,4-dioxane from
manufacturing and industrial sites may also contribute to general population exposures through drinking
water and air.

1,4-Dioxane produced as a byproduct has also been detected in consumer and commercial products,
resulting in potential exposure to consumers and bystanders (evaluated in the 2020 RE) or workers and
ONUs (evaluated as described in Section 3.1.2). For example, 1,4-dioxane exposures are expected
during the use of dish soap and dishwashing detergent from unloading and transferring detergent
formulation, transport container cleaning, and washing operations due to the presence of 1,4-dioxane as

Page 30 of 484


-------
1197

1198

1199

1200

1201

1202

1203

1204

1205

1206

1207

1208

1209

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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

a surfactant byproduct. In addition, consumer and commercial products containing 1,4-dioxane may
contribute to general population exposures when released down the drain.

In this draft supplement, EPA evaluated pathways of exposure to 1,4-dioxane produced as a byproduct
that were not previously assessed. Specifically, EPA considered 1,4-dioxane present as a byproduct in
commercial products corresponding to the consumer products considered in the 2020 RE. EPA
considered the direct occupational exposures that result from use of these commercial products as well
as the DTD releases of consumer and commercial products, which contribute to general population
exposures through surface water. EPA also identified a new set of COUs where 1,4-dioxane is produced
or present as a byproduct, based on information submitted by public commenters and other reasonably
available information. For each of these new COUs, EPA evaluated occupational exposure as well as
industrial releases that contribute to general population exposures. The available information supporting
inclusion of each of the new COUs is described below. A more detailed list of the new COUs and COU
subcategories considered in this draft supplement is presented in Section 2.1.1. The following COUs are
known to produce 1,4-dioxane as a byproduct based on reasonably available information, but 1,4-
dioxane produced as a byproduct may also be present in other industries that have not yet been
identified:

•	Industrial/Commercial Use of Products Containing 1,4-Dioxane as a Byproduct. 1,4-
Dioxane is present in a range of commercial products (including textile dyes, antifreeze, surface
cleaners, dish soaps, laundry detergents, and paint and floor lacquer) because it is produced as a
byproduct during the manufacture of ingredients such as ethoxylated surfactants. While 1,4-
dioxane present as a byproduct in consumer products was previously assessed, evaluation of 1,4-
dioxane in these corresponding commercial products is new in this draft supplement. 1,4-
Dioxane present in commercial products can result in occupational exposure in commercial
settings as well as DTD releases that contribute to general population exposures.

•	PET Manufacturing. 1,4-Dioxane has been identified as a byproduct in the manufacture of PET
(l c> ' i1 \ 1 ). EPA does not have information on the byproduct concentration of 1,4-
dioxane in PET. PET is produced by the esterification of terephthalic acid to form
bishydroxyethyl terephthalate (BHET) (Forkner et ai. 2004). BHET polymerizes in a
transesterification reaction catalyzed by antimony oxide to form PET (Forkner et ai. 2004). In
2014, 20.6 million metric tons of PET were used in the United States (McDaniel and
DesLauriers. 2015). 1,4-Dioxane produced during PET manufacturing may result in occupational
exposures and may contribute to general population exposures via releases to water and air.

•	Ethoxylation Processing. 1,4-Dioxane may be formed as a byproduct of reactions based on
condensing ethylene oxide or ethylene glycol during manufacture of detergents, shampoos,
surfactants, some food additives, and certain pharmaceuticals (HHS. 2016). In cosmetic
ethoxylated raw materials and ethoxylated alkyl sulfates, 1,4-dioxane has been detected at
concentrations of 0.48 to 1,410 ppm (U.S. EPA. 2020c; Saraii and Shirvani. 2017; Davarani et
al. 2012; Black et al. 2001). Polyethoxylated raw materials are widely used in cosmetic
products as emulsifiers, foaming agents, and dispersants (Black et al.. 2001). They are produced
by polymerizing ethylene oxide, usually with a fatty alcohol, to form polyethoxylated alcohols
which may be used to synthesize other products such as sulfated surface-active agent. During the
ethoxylation process, 1,4-dioxane can be formed as a byproduct by the dimerization of ethylene
oxide (Black et al.. 2001). The volume of 1,4-dioxane produced as a byproduct of ethoxylation
reactions is unknown. 1,4-Dioxane produced during ethoxylation processing may result in
occupational exposures and may contribute to general population exposures via releases to water
and air. It also contributes to the presence of 1,4-dioxane in consumer and commercial products.

Page 31 of 484


-------
1244

1245

1246

1247

1248

1249

1250

1251

1252

1253

1254

1255

1256

1257

1258

1259

1260

1261

1262

1263

1264

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

• Hydraulic Fracturing. Hydraulic fracturing stimulates an existing oil or gas well by injecting a
pressurized fluid containing chemical additives into the well (	I022d). 1,4-Dioxane is

measured in fracturing fluid, a water-based fluid that contains several chemical additives and in
waste fluid (produced waters). FracFocus 3.0 contains self-reported information indicating that
1,4-dioxane is present in hydraulic fracturing fluid additives, as scale inhibitors, additives,
biocides, friction reducers, and surfactants (GWPC and IOGCC. 2022). According to the
FracFocus 3.0 database, 1,4 dioxane is present in weight fractions ranging from 2.3 xl0~n to 0.05
within hydraulic fracturing additives and 1.00/10 12 to 4.30><10"6 in hydraulic fracturing fluids
(GWPC and IOGCC. 2022). 1.4-Dioxane has been documented to have a concentration of 60
|.ig/L in hydraulic fracturing produced waters (Lester et al. ). 1,4-Dioxane present in
fracturing fluid may result in occupational exposures. It may also contribute to general
population exposures via discharge to surface water, groundwater, or fugitive air emissions from
fracturing operations.

1.3.1.2 Occupational Exposures

The conceptual model Figure 1-3 Figure l-3presents the exposure pathways, exposure routes, and
hazards to people from industrial and commercial releases and uses of 1,4-dioxane. Blue shading
highlights the exposures evaluated in this draft supplement. Workers and ONUs may have acute (8 hr)
or chronic (annual to lifetime) exposures to 1,4-dioxane produced as a byproduct during PET
manufacturing, ethoxylation processes, or hydraulic fracturing operations. Workers and ONUs may also
have acute or chronic exposures to 1,4-dioxane present as a byproduct in commercial products,
including detergents, cleaners, and lacquers.

Page 32 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

INDUSTRIAL AND COMMERCIAL ACTIVITIES/USE EXPOSURE PATHWAY EXPOSURE ROUTE & RECEPTORS HAZARDS

DURATION

INDUSTRIAL USES

Manufacture

(Including Import)

Processing:

•	Processing as a
reactant/i ntermediate

•	Repackaging

•	Non-incorporative
activities

Recycling

Processing Aid, Not
Otherwise Listed

Functional Fluids

(Open and Closed Systems)

Laboratory Chemicals
Other Industrial Uses

1,4-Dioxane as a Byproduct

COMMERCIAL USES
(Present only as Byproduct)

Paints and Coatings

e.g., Latex Wall Paint or
Floor Lacquer

Cleaning and Furniture
Care Products

e.g., Surface Cleaner

Laundry and Dishwashing
Products

e.g., Dish Soap, Dishwasher
Detergent, Laundry
Detergent

Arts, Crafts and Hobby
Materials

e.g., Textile Dye

Automotive Care Products

e.g., Antifreeze

Other Consumer Uses

e.g., Spray Polyurethane
Foam, Antifreeze

T

Waste Handling,
Treatment, and Disposal

-~ Liquid Contact

Dermal
(acute dose, chronic
ADD, LADD)

Hazards Potentially
Associated with
Acute and/or
Chronic Exposures

Vapor/Mist

i

L

l Fug
Emis

tive
sions

Inhalation
(8hrTWA, chronic
ADC, LA DC)

1265

1266

Wastewater, Liquid Wastes, Solid
Wastes

Assessed in the Supplement to
the Risk Evaluation

Figure 1-3. Conceptual Model for Occupational Exposures from Industrial and Commercial Activities

Page 33 of 484


-------
1267

1268

1269

1270

1271

1272

1273

1274

1275

1276

1211

1278

1279

1280

1281

1282

1283

1284

1285

1286

1287

1288

1289

1290

1291

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

In this draft supplement, EPA evaluated acute and chronic cancer and non-cancer risks from
occupational inhalation and dermal exposures to 1,4-dioxane produced as a byproduct during PET
manufacturing and ethoxylation processes, hydraulic fracturing fluids and waste containing 1,4-dioxane,
and commercial products containing 1,4-dioxane.

1.3.1.3 General Population Exposures

The conceptual model in Figure 1-4 presents general population exposure pathways and hazards from
environmental releases and wastes associated with COUs (red, blue, and purple shading for each source
in the figure corresponds to the environmental media to which they release). The disposal and release
scenarios illustrated in Figure 1-1, Figure 1-2, and Figure 1-3 all contribute to the releases to air, water
and land that may result in the general population exposures illustrated in Figure 1-4. The general
population may be exposed to 1,4-dioxane released to surface water (blue shading), groundwater (red
shading), and air (purple shading). Drinking water exposures are evaluated based on releases to both
surface water and groundwater and these are each described in the drinking water subsection below.

EPA's evaluation of general population exposures considers potentially exposed or susceptible
subpopulations (PESS). Exposures to 1,4-dioxane through air and water could result in risk to fenceline
communities. As defined in the Draft Screening Level Approach for Assessing Ambient Air and Water
Exposures to Fenceline Communities Version 1.0,5 fenceline communities are members of the general
population that are in proximity to air emitting facilities or a receiving water body, and who therefore
may be disproportionately exposed to a chemical undergoing risk evaluation under TSCA section 6. For
the air pathway, proximity goes out to 10,000 m from an air emitting source. For the water pathway,
proximity does not refer to a specific distance measured from a receiving water body, but rather to those
members of the general population that may interact with the receiving water body and thus may be
exposed.

5 The draft Screening Level Approach for Assessing Ambient Air and Water Exposures to Fenceline Communities Version 1.0
is available at https://www.epa.gov/assessing~aiMl~managing~chemicals~iinder~tsca/tsca~screening~level~approach~assessing~
ambient-air-and.

Page 34 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

1292

1293

RELEASES AND WASTES FROM
INDUSTRIAL / COMMERCIAL / CONSUMER
USES

Industrial Pre-
Treatment or
Industrial WWT

EXPOSURE PATHWAYS

Indirect discharge
~

Wastewater or
Liquid Wastes

POTW

Down-the-Drain
Consumer Product Use

Hydraulic
Fracturing
Produced Water!

Underground
Injection

Hazardous and

Municipal
Waste Landfill

Solid Wastes
Liquid Wastes

Hazardous an
Municipal
Waste

Off-site Waste
T ransfer

Recycling, Other
Treatment

Emissions to Air

EXPOSURE ROUTES

RECEPTORS

HAZARDS

Hazards Potentially
Associated with
Acute and/or Chronic
Exposures

| Land Pathway

| Air Pathway
H Water Pathway

Included in 2020 RE (No Revisit)

Figure 1-4. Conceptual Model for Environmental Releases and General Population Exposures

Page 35 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

The 2020 RE included an assessment of acute incidental/recreational general population exposure to 1,4-
dioxane in surface water resulting from industrial releases. It did not evaluate risks from other general
population exposure pathways such as drinking water or air.

In this draft supplement, EPA evaluated additional general population exposure via air and drinking
water. EPA evaluated acute (24-hour) and chronic (annual to lifetime) cancer and non-cancer risks from
these exposure pathways. Where data were reasonably available, EPA incorporated releases of 1,4-
dioxane produced as a byproduct into these pathways. The Agency also considered aggregate 1,4-
dioxane exposures and risks from multiple releasing facilities or COUs for each pathway.

1.3.1.3.1 Drinking Water

1,4-Dioxane may enter surface water through direct and indirect industrial releases, DTD releases from
consumer and commercial products via wastewater treatment facilities, and releases of wastewater from
hydraulic fracturing sites. Similarly, 1,4-dioxane released or disposed of through various land pathways
may reach groundwater under some conditions. There is potential for general population exposures to
1,4-dioxane if contaminated surface water or groundwater are used as drinking water. 1,4-Dioxane is
mobile in water and does not readily degrade in water. Available data indicate that typical wastewater
treatment and drinking water treatment methods are not effective at removing 1,4-dioxane. The
subsections below explain how general population exposures through surface and groundwater were
considered.

Surface Water Pathway

1,4-Dioxane was included in the third unregulated contaminant monitoring rule (UCMR3) (U.S. EPA.
2017d) published in May of 2012 requiring community water systems to monitor for 1,4-dioxane
between 2013 and 2015. National and state water monitoring programs have detected 1,4-dioxane in
drinking water and drinking water sources (as described in Section 2.3.1.10). In the absence of
monitoring data, estimating 1,4-dioxane surface water concentrations can be complex because in
addition to direct and indirect industrial and commercial releases, upstream sources from releasing
facilities and DTD releases of consumer and commercial products contribute to surface water
contamination.

EPA evaluated surface water concentrations (Section 2.3.1) and drinking water exposures (Section
3.2.2.1) that could result from direct and indirect industrial releases, DTD releases of consumer and
commercial products, and disposal of wastewater from hydraulic fracturing sites. EPA used a novel
aggregate model to predict water concentrations of 1,4-dioxane that could result from multiple sources
that release 1,4-dioxane to the same receiving water bodies. Surface water concentrations estimated for
each source in isolation and from multiple sources in aggregation were used to evaluate potential
exposures (Section 3.2.2.1) and risks (Section 5.2.2.1) from general population oral exposure to 1,4-
dioxane in drinking water.

Land Pathway to Groundwater

Groundwater contamination with 1,4-dioxane presents a potential risk when the chemical substance is
released to landfills, underground injection wells, or surface impoundments. Due to its physical-
chemical properties (e.g., water solubility, Henry's law constant) and fate characteristics (e.g.,
biodegradability, half-life in groundwater), 1,4-dioxane is anticipated to persist in groundwater for
months to years. This persistence has resulted in higher 1,4-dioxane concentrations in groundwater
relative to other media (A.TSD	). EPA considered potential for groundwater contamination

following disposal of waste containing 1,4-dioxane to landfills, underground injection of 1,4-dioxane
waste, and disposals of hydraulic fracturing waste containing 1,4-dioxane. Groundwater concentrations

Page 36 of 484


-------
1342

1343

1344

1345

1346

1347

1348

1349

1350

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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

estimated for each source (Section 2.3.2) were used to evaluate exposure (Section 3.2.2.2) and risks
(Section 5.2.2.1.6) for communities who rely on groundwater as a source of drinking water.

1.3.1.3.2	Air

Industrial releases to air include those from sites where 1,4-dioxane is manufactured intentionally as
well as those where it is produced or present as a byproduct. In this draft supplement, EPA evaluated
exposures and risks for communities located near release sites (fenceline communities) because they are
the members of the general population that are expected to be PESS due to their greater exposure. EPA
applied the methodology presented in the Draft TSCA Screening Level Approach for Assessing Ambient
Air and Water Exposures to Fenceline Communities6 to evaluate risks from industrial air releases to
fenceline communities. In addition to considering risks from individual facilities, EPA evaluated risks
from aggregate exposures in cases where multiple facilities reporting 1,4-dioxane releases to air were in
proximity. The Agency also evaluated potential risks to fenceline communities from air emissions of
1,4-dioxane modeled for hydraulic fracturing operations and industrial and commercial laundries.

1.3.1.3.3	Aggregate Exposure

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 § 702.33)." In this
draft supplement, EPA considered the combined 1,4-dioxane exposure an individual may experience due
to releases to air or water from multiple sources. For general population drinking water exposure
scenarios, EPA evaluated combined exposure and risks from multiple sources of 1,4-dioxane in surface
water, including direct and indirect industrial releases, DTD releases, and upstream background
contamination (Section 5.2.2.1). For general population air exposure scenarios, EPA evaluated
combined exposure and risk across multiple facilities in proximity releasing to air (Section 5.2.2.3 and
Appendix J.4). EPA qualitatively considered aggregate exposures across exposure routes {i.e., across
oral and inhalation) and across exposure pathways {i.e., across air and water) but did not quantitatively
aggregate these exposures due to uncertainties around the additivity of effects across routes. The
rationale for the scope of aggregate analysis in this draft supplement and remaining sources of
uncertainty are further discussed in Section 5.2.2.5.

1.3.2 Potentially Exposed or Susceptible Subpopulations	

TSCA section 6(b)(4)(A) requires that risk evaluations "determine whether a chemical substance
presents an unreasonable risk of injury to health or the environment, without consideration of costs 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 section 3(12) states that "the term 'potentially exposed or susceptible subpopulatioif
[PESS] 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."

Considerations related to PESS can influence the selection of relevant exposure pathways, the sensitivity
of derived hazard values, the inclusion of particular populations, and the discussion of uncertainties
throughout the assessment. Factors that may contribute to increased exposure or biological susceptibility
to a chemical include lifestage, pre-existing disease, lifestyle activities {e.g., smoking, physical activity),
occupational and consumer exposures (including workers and occupational non-users, consumers and

6 The draft Screening Level Approach for Assessing Ambient Air and Water Exposures to Fenceline Communities Version 1.0
is available at https://www.epa.gov/assessing~aiMl~managing~chemicals~iinder~tsca/tsca~screening~level~approach~assessing~
ambient-air-and.

Page 37 of 484


-------
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

1422

1423

1424

1425

1426

1427

1428

1429

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

bystanders), geographic factors (e.g., fenceline communities), socio-demographic factors, nutrition,
genetics, unique activities (e.g., subsistence fishing), aggregate exposures, and other chemical and non-
chemical stressors.

This draft supplement considers PESS throughout the human health exposure assessment and risk
characterization. The hazard assessment and dose-response analysis used in this draft supplement
incorporate all PESS considerations described previously in the 2020 RE. Section 5.2.2.4 provides a
summary of how specific factors contributing to exposure and susceptibility were addressed in this
assessment and identifies remaining sources of uncertainty for PESS.

1.4	Systematic Review

EPA used the TSCA systematic review process described in the Draft Systematic Review Protocol
Supporting TSCA Risk Evaluations for Chemical Substances (	) (hereinafter referred to

as "2021 Draft Systematic Review Protocol") to identify information needed to evaluate additional
COUs and exposure pathways considered in this draft supplement. Appendix C provides additional
information on the literature search strategy, data screening, evaluation, extraction, and evidence
integration steps performed in support of this assessment—including clarifications and updates made to
the 2021 Draft Systematic Review Protocol to better address assessment needs for this draft supplement.

1.5	Document Outline

This draft supplement to the risk evaluation for 1,4-dioxane comprises the following sections and
appendices:

•	Section 1 presents information on the scope of the draft supplement. It also includes an overview
of the systematic review process used in this analysis. Appendix A provides a list of
abbreviations and acronyms used throughout this report while Appendix B provides the full
name and links to all supplemental documents associated with this draft risk evaluation. A more
detailed description of the systematic review protocol for this assessment is presented in
Appendix C while Appendix D provides a crosswalk of COUs with occupational exposure
scenarios.

•	Section 2 presents an overview of releases and concentrations of 1,4-dioxane in the environment.
A more detailed description of the industrial and commercial environmental release assessment
is presented in Appendix E. Methods for estimating environmental concentrations of 1,4-dioxane
are described in more detail in Appendix G (surface water), Appendix H (groundwater), and
Appendix J (air).

•	Section 3 presents the human exposure assessment for occupational and general population
exposure pathways. Details of the occupational exposure assessment are presented in Appendix
F and details of the general population exposure assessment are presented in Appendix I and
Appendix J.

•	Section 4 provides a summary of the human health hazard and dose-response assessment
previously published in the 2020 RE and describes duration adjustments made for the current
analysis.

•	Section 5 presents risk characterization based on the conditions of use and exposure pathways
evaluated in this draft supplement. Details of risk calculations and risk estimates are provided in
the set of supplemental risk calculator files listed in Appendix B. Section 5 also includes a
discussion of PESS based on both greater exposure and susceptibility, a description of aggregate
and sentinel exposures, and a discussion of assumptions and uncertainties and the impact on the
supplemental risk evaluation.

Page 38 of 484


-------
1430

1431

1432

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Figure 1-5 provides an overview of how the analyses presented in each section are integrated into risk
characterization.

Release Assessment

Section 2

Releases to
surface water

Releasesto
land & groundwater

Releasesto air

Exposure Assessment

Section 3

x	N

Occupational

Exposure Scenarios

Inhalation exposure

Dermal exposure

a

General Population
Exposure Scenarios

Oral exposures
through drinking
water

Inhalation
exposure
through air

Risk Characterization

Section 5

Occupational Risk
Characterization

Inhalation risk

Dermal risk

General Population
Risk Characterization

Drinking water risk

Inhalation risk

Hazard Assessment

Section 4

Occupational
hazard values for
acute and chronic
non-cancerand
cancer (consistent
with 2020 RE)

General population
hazard valuesfor
acute, chronic non-
cancer and cancer

hazard values
(derived from PODs
in the 2020 RE, with
some duration
adjustments)

1433

1434	Figure 1-5. Overview of Analyses Included in this Draft Supplement to the Risk Evaluation for

1435	1,4-Dioxane

Page 39 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2 RELEASES AND CONCENTRATIONS
2.1 Approach and Methodology

2.1.1 Industrial and Commercial Releases

Releases to the environment are one component of potential exposure and may be derived from reported
data that are obtained through direct measurement via monitoring, calculations based on empirical data,
and/or assumptions and models.

The original COUs for 1,4-dioxane are summarized in Table 1-4 of the Final Risk Evaluation for 1,4-
Dioxane (	s20c). Additional COUs included in this draft supplement due to 1,4-dioxane

produced as a byproduct are presented in Table 2-1. For general population exposures, this draft
supplement considers releases from all COUs (including the original COUs included in the 2020 RE and
the additional COUs associated with 1,4-dioxane produced as a byproduct). For occupational exposures,
this draft supplement focuses on the additional COUs associated with 1,4-dioxane produced as a
byproduct. For additional information and context on the inclusion of these COUs in the draft
supplement, refer to Sections 1.1 and 1.2. For the full table of COUs, including those previously
assessed in the Final Risk Evaluation for 1,4-Dioxane (	2020c). see Appendix D.

A COU is a combination of life cycle stage, category, and subcategory, as shown in Table 2-1. The COU
subcategory is the most granular description of the use. EPA mapped each COU to an occupational
exposure scenario (OES). The purpose of an OES is to group, where appropriate, COUs based on
similarity of the operations and data availability for each COU. For each OES, EPA estimated air, land,
and water releases and occupational dermal and inhalation exposures. EPA mapped OESs to COUs
using professional judgment based on reasonably available data and information that describe how
releases and exposures take place within an occupational COU. EPA may group multiple COUs into an
OES if the release and exposure potential is similar across the COUs and there is insufficient data to
differentiate the COUs. This grouping minimized repetitive assessments. Alternatively, EPA may assign
multiple OESs to one COU if there are several ways in which release and exposure takes place for the
given COU and sufficient data exist to separately assess the OES. Appendix D.l shows mapping
between COUs and OESs. A crosswalk of the COUs with the OESs assessed is provided in Table 2-1.

As shown in Table 2-1, most COU life cycle stage, category, and subcategory combinations map to a
single OES with a similar or identical name to the COU subcategory. However, for the COU
subcategory of dish soap, dishwasher detergent, and laundry detergent, EPA assigned four OESs: (1)
dish soap, (2) dishwasher detergent, (3) laundry detergent (industrial), and (4) laundry detergent
(institutional). Institutional use of laundry detergent equates to commercial use.

EPA assessed environmental releases (air, water, and land) and occupational exposures (inhalation and
dermal) to 1,4-dioxane for each of the OESs listed in Table Apx D-l. EPA used the environmental
release estimates for each OES for subsequent environmental concentrations and general population
exposure calculations.

Page 40 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

1476	Table 2-1. Additional Categories and Subcategories of COUs and Associated OESs Included in the Scope of the Draft Supplement

1477	Due to the Presence of 1,4-Dioxane Produced as a Byproduct"		

Condition of Use

OES Mapped to COU

Life Cycle Stage

Category''

Subcategory'

Processing

Byproduct

Byproduct produced during the ethoxylation processes

Ethoxylation Process Byproduct

Byproduct produced during the production of
polyethylene terephthalate

Polyethylene Terephthalate (PET)
Byproduct

Industrial use,
commercial use

Other uses

Hydraulic fracturing

Hydraulic Fracturing

Consumer use,
commercial use

Paints and coatings

Latex Wall Paint or Floor Lacquer

Paint and Floor Lacquer

Cleaning and furniture care
products

Surface Cleaner

Surface Cleaner

Laundry and dishwashing products

Dish soap

Dishwasher detergent
Laundry detergent

Dish Soap

Dishwasher Detergent
Laundry Detergent (Industrial)''
Laundry Detergent (Institutional)''

Arts, crafts, and hobby materials

Textile dye

Textile Dye

Consumer use,
commercial use

Automotive care products

Antifreeze

Antifreeze

Disposal

Disposal

Industrial pre-treatment

Disposal

Industrial wastewater treatment

Publicly owned treatment works (POTW)

Underground injection

Municipal landfill

Hazardous landfill

Other land disposal

Municipal waste incinerator

Hazardous waste incinerator

Off-site waste transfer

aNew COUs and associated OESs where 1,4-dioxane is produced as a byproduct.

b These categories of COU reflect CDR codes and broadly represent conditions of use for 1,4-dioxane in industrial and/or commercial settings.
c These subcategories reflect more specific uses of 1,4-dioxane.

dLaundry detergent use may occur in industrial, commercial, or consumer settings. Sufficient information was available to separately assess each use setting.
Thus, there are two OESs—one for industrial and one for institutional, which equates to commercial use.

1478

Page 41 of 484


-------
1479

1480

1481

1482

1483

1484

1485

1486

1487

1488

1489

1490

1491

1492

1493

1494

1495

1496

1497

1498

1499

1500

1501

1502

1503

1504

1505

1506

1507

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2.1.1.1 General Approach and Methodology for Environmental Releases

Data reported to the Toxics Release Inventory (TRI)7 and discharge monitoring reports (DMRs) are the
primary sources of release data that EPA used for the release assessments. Under the Emergency
Planning and Community Right-to-Know Act (EPCRA) section 313, 1,4-dioxane has been a TRI-
reportable substance since 1987. The TRI database includes information on disposal and other releases
of 1,4-dioxane to air, water, and land, in addition to how it is being managed through recycling,
treatment, and burning for energy recovery. Under the Clean Water Act (CWA), surface water
discharges reported in DMRs are based on required monitoring as part of a facility's National Pollutant
Discharge Elimination System (NPDES) permit.

For each OES, a daily air, land, and water release was estimated based on annual releases, release days,
and the number of facilities (Figure 2-1). The blue boxes represent primary sources of release data that
are used to develop annual releases, release days, and number of facilities. The information in the green
boxes is aggregated by OES to provide a daily release estimate.

Figure 2-1. Overview of EPA's Approach to Estimate Daily Releases for Each OES

TRI = Toxics Release Inventory; DMR = discharge monitoring report; ESD = emission scenario
document; GS = generic scenario

2.1.1.2 Water Release Estimates

EPA followed a similar approach for estimating industrial and commercial water releases as it did in the
2020 RE, with one key difference. Here, EPA evaluated multiple years of data using data from 2013 to
2019 TRI (U.S. EPA. 2022g) and 2013 to 2019 DMR (U.S. EPA. 2022c\ as opposed to utilizing 1 year
of data.

Where water releases are expected for an OES but TRI and DMR data were not available, EPA
estimated industrial, and commercial water releases using two approaches. If available, Stochastic
Human Exposure and Dose Simulation for High Throughput (SHEDS-HT) DTD modeling was used.
SHEDS-HT predicts a per capita DTD loading of 1,4-dioxane, which is combined with an estimation of
the population contributing to publicly owned treatment works (POTW) effluent on the modeled water

7 TRI page: https://www.epa.gov/toxics-releasE-inventorv-tri-program.

Page 42 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

body stream to produce an estimated DTD loading. If SHEDS-HT DTD modeling was not available for
an OES without TRI or DMR data, EPA used data from literature, ESDs, and GSs in conjunction with
Monte Carlo simulation where sufficient data were available to vary calculation input parameters to
estimate industrial and commercial water releases (see Appendix E.3 for additional information).

For the following OESs, EPA either could not estimate water releases due to lack of reasonably
available data or information or did not expect water releases based on volatility and use patterns:

•	Functional Fluids (Closed-Systems). Water release data were not available for this OES.
However, EPA expects that the sources of release for this OES to be similar to those for the
Industrial Uses OES (per process information in the 2020 RE, Appendix G.6.4). Therefore, EPA
grouped the water release assessment for Functional Fluids (Closed-Systems) into that for
Industrial Uses. However, there is uncertainty in this assumption of similar release sources
between these OESs.

•	Laboratory Chemical, Film Cement, and Dry Film Lubricant. Wastewater discharges
containing 1,4-dioxane were not expected for these OESs; releases from these OESs are expected
to be to air from volatilizations and landfill/incineration from disposal of empty containers and
other waste (2020 RE, Appendix G).

•	Antifreeze. Wastewater discharges containing 1,4-dioxane were not expected for this OES;
releases from this OES are expected to be to air from volatilizations during antifreeze changeouts
and to landfill/incineration from disposal of empty antifreeze containers and spent antifreeze.

•	Paints and Floor Lacquer. Wastewater discharges containing 1,4-dioxane were not expected
for this OES; releases from this OES are expected to be to air from volatilizations during
painting/drying and to landfill/incineration from disposal of empty paint containers, used paint
brushes/rollers, or solvent washes of paint brushes/rollers.

2.1.1.3 Land Release Estimates

EPA used data from 2019 TRI (	) to estimate industrial and commercial land releases

that were mapped to each OES with the exception of the Disposal OES. For that OES, EPA performed a
more detailed analysis using data from 2013 to 2019 TRI (	Z022g). Where land releases are

expected for an OES, but TRI data were not available, releases were estimated using reasonably
available data from literature, ESDs, and GSs in conjunction with Monte Carlo simulation (Palisade.
2022a) to allow for variability in calculation input parameters where sufficient data were available to
inform such variability.

EPA did not estimate daily land releases due to the high level of uncertainty in the number of release
days. This uncertainty is because facility operating days does not correlate directly to releases. For
example, a facility may wait until a dumpster or other waste receptacle is full before disposing of it.
Because these releases may occur on a daily, weekly, and even monthly scale, EPA used the annual land
releases reported in TRI data or modeled without estimating land releases for a different frequency
(daily, weekly, monthly). See Appendix E.4 for additional information. Annual land release estimates
were used to estimate potential groundwater contamination from landfill releases as described in Section
2.3.2.

For the following OESs, EPA was not able to estimate land releases due to lack of data or information or
did not expect land releases due to physical form, use patterns, and lack of data:

•	Manufacturing, Import and Repackaging, and Functional Fluids (Open-Systems). Data
from 2019 TRI (	22e) indicated that there were no releases of 1,4-dioxane to land

Page 43 of 484


-------
1554

1555

1556

1557

1558

1559

1560

1561

1562

1563

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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

from facilities that EPA mapped to these OESs. EPA did not have additional reasonably
available information to estimate land releases from these OESs.

•	Functional Fluids (Closed-Systems). See previous explanation in the "Water Release
Estimates" section above.

•	3D Printing. Industrial applications of this OES are expected to be accounted for in the
Industrial Uses TRI data. Per Appendix G.6.8 of the 2020 RE, 3D printing ink containing 1,4-
dioxane is used in research labs to print biomedical products. Because the 2019 TRI data for the
Industrial Uses OES include medicinal and pharmaceutical manufacturing NAICS codes,
medical research labs that conduct 3D printing with 1,4-dioxane inks may be captured in that
OES. Therefore, EPA grouped the land release assessment for 3D Printing into that OES for
Industrial Uses. However, the extent to which all potential 3D printing sites that use 1,4-dioxane
are captured in the Industrial Uses TRI data is unknown.

EPA also notes that the Hydraulic Fracturing OES is associated with certain specific land releases that
may not apply to other OESs, such as the releases of wastewater containing 1,4-dioxane to deep well
injection or surface impoundments, which are considered land releases in this assessment. The Agency
estimated these deep well injection and surface impoundment releases, which were used in addition to
landfill releases, to estimate potential groundwater contamination from hydraulic fracturing described in
Section 2.3.2.

2.1.1.4 Air Release Estimates

EPA applied the following tiered approach to developing air release, exposure, and risk estimates:

1.	Pre-screening analysis,

2.	Single-year fenceline analysis, and

3.	Multi-year fenceline analysis.

2.1.1.4.1	Pre-screening Analysis

This analysis is described in the Draft TSCA Screening Level Approach for Assessing Ambient Air and
'Water Exposures to Fenceline Communities and consisted of extracting data for all facilities reporting
1,4-dioxane air releases to the 2019 TRI (	22g). The extracted data were reviewed to

identify the maximum single facility release reported across all reporting facilities. Additionally, the
arithmetic average (mean) value of all reported releases was calculated. These two release values were
used for further analysis to estimate exposure concentrations at select distances from the releasing
facility.

2.1.1.4.2	Single-Year Fenceline Analysis

Where available, EPA used data from 2019 TRI to estimate industrial and commercial air releases in
accordance with the Draft TSCA Screening Level Approach for Assessing Ambient Air and 'Water
Exposures to I7emetine Communities. Facilities are only required to report to TRI if the facility has 10
or more full-time employees; is included in an applicable North American Industry Classification
System (NAICS) code; and manufactures, processes, or uses the chemical in quantities greater than a
certain threshold. Due to these limitations, some sites that manufacture, process, or use 1,4-dioxane may
not report to TRI and are therefore not included in these datasets.

Where air releases are expected for an OES, but TRI data were not available, industrial, and commercial
air releases were estimated using data from literature, ESDs, and GSs in conjunction with Monte Carlo
simulation (Palisade. 2022a) to allow for variability in calculation input parameters where sufficient data
were available to inform such variability. See Appendix E.5 for additional information.

Page 44 of 484


-------
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

1635

1636

1637

1638

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

For the following OESs, EPA was not able to estimate air releases due to lack of data or information:

•	Functional Fluids (Closed-Systems). See previous explanation in the "Water Release
Estimates" section above.

•	3D Printing. Industrial applications of this OES are expected to be accounted for in the
Industrial Uses TRI data. Per Appendix G.6.8 of the 2020 RE, 3D printing ink containing 1,4-
dioxane is used in research labs to print biomedical products. Because the 2019 TRI data for the
Industrial Uses OES include medicinal and pharmaceutical manufacturing NAICS codes,
medical research labs that conduct 3D printing with 1,4-dioxane inks may be captured in that
OES. Therefore, EPA grouped the air release assessment for 3D Printing into that OES for
Industrial Uses. However, the extent to which all potential 3D printing sites that use 1,4-dioxane
are captured in the Industrial Uses TRI data is unknown.

•	Textile Dyes, Dish Soap, and Dishwasher Detergent. EPA did not find relevant reasonably
available 1,4-dioxane or surrogate TRI data, literature sources, sufficient process information,
nor ESD or GS with air release estimation approaches to estimate air releases for these OESs.
Therefore, EPA was not able to estimate air releases for these OESs.

2.1.1.4.3 Multi-Year Analysis

The multi-year analysis incorporates Science Advisory Committee on Chemicals (SACC)8
recommendations on the Draft TSCA Screem	pproach for Assessing Ambient Air and Water

Exposures to Fenceline Communities to evaluate multiple years of chemical release data to estimate
exposures and associated risks to fenceline communities. This is achieved by conducting a facility-by-
facility evaluation of all 1,4-dioxane releases reported to TRI from 2015 through 2020. Data for these 6
years were obtained from the TRI database (TRI basic plus files downloaded on August 5, 2022).

Annual release data for 1,4-dioxane were extracted from the entire TRI data set for all facilities
reporting air releases of 1,4-dioxane for one or more years between 2015 and 2020. Facilities were
categorized into occupational exposure scenarios for exposure modeling purposes and later cross-walked
to COUs for risk management purposes.

2.2 Environmental Releases

2,2.1 Industrial and Commercial Releases

This section summarizes the estimated air, water, and land releases for each OES; the weight of the
scientific evidence conclusions for these estimates; and the strengths, limitations, assumptions, and key
sources of uncertainty for these estimates.

2.2.1.1 Release Estimates Summary

EPA estimated air, water, and land releases of 1,4-dioxane using various methods and information
sources—including TRI and DMR data, GSs and ESDs modeling with and without Monte Carlo,
process information, and SHEDS-HT DTD Modeling. EPA combined its estimates for annual releases,
release days, and number of facilities to estimate a range of daily releases for all OESs, including those
presented in the December 2020 RE. The COUs associated with each OES are summarized in
TableApx D-l. A summary of these industrial and commercial releases for air, water, and land are
presented in Table Apx E-3, Table Apx E-5, and Table Apx E-7, respectively. These release estimates
are for total releases from a facility and may include multiple points of release, such as multiple outfalls
for discharges to surface water or multiple point sources for air emissions. Note that for some release

8 Additional information about SACC is available at https://www.epa.gov/tsca-peer-review/science-advisore-cotninittee-
chemicals-basic-information.

Page 45 of 484


-------
1639

1640

1641

1642

1643

1644

1645

1646

1647

1648

1649

1650

1651

1652

1653

1654

1655

1656

1657

1658

1659

1660

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

estimates, there is uncertainty and variability in the potential media of release. In such cases, EPA did
not have sufficient information to partition the release estimates between all potential media of release
and they are replicated between the air, land, and water subsections if there is overlap in the potential
media of release.

EPA mapped these releases by media, state, and tribal territory for the conterminous United States.
Alaska, American Samoa, Guam, Hawaii, N. Mariana Islands, Puerto Rico, and the U.S. Virgin Islands
are not mapped since no estimated releases are known.

Surface water releases as reported by TRI and DMR are presented in Figure 2-2. For surface water
releases, the data are divided based on the source of data (i.e., DMR, TRI) and whether the release is
from a direct (on-site) source, including on-site wastewater treatment systems, or indirect (offsite)
source where the chemical substance was taken to a different location for potential release, such as a
POTW. The largest releases have been from PET manufacturing in Tennessee (2,512,434 kg in 2019
and 15,168 kg in 2018), Alabama (170,526 kg in 2015; 125,903 kg in 2014; and 111,924 kg in 2017),
and West Virginia (14,134 kg in 2016 and 12,229 kg in 2014).

(7-P Tribal Lands
Source of Water Release

O	Direct Water Releases (TRI)

•	Indirect Water Releases (TRI)

•	Direct Releases (DMR)

Annual Mass Released (kg/yr)

o	0.00-10.0

O	>10.0-100

O	>100- 1,000

O	>1,000- 10,000

O	>10,000 - 100,000

O	>100,000- 1,000,000
o >1,000,000-2,512,434

0 200 400	800 1,200 1,600

¦ Kilometers

Figure 2-2.1,4-Dioxane Annual Water Releases as Reported to TRI and DMR, 2013-2019

Note: Some symbols for individual years may overlap and obscure annual releases at each site.

Alaska, American Samoa, Guam, Hawaii, N. Mariana Islands, Puerto Rico, and the U.S. Virgin Islands are not

shown due to no known modeled or estimated releases.

Page 46 of 484


-------
1661

1662

1663

1664

1665

1666

1667

1668

1669

1670

1671

1672

1673

1674

1675

1676

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Land disposals as reported by TRI are available in Figure 2-3. The largest disposals have been to on-site
Class I Underground Injection Wells in Texas (169,035 kg in 2013; 42,865 kg in 2015; 10,729 kg in
2018), On-site Subtitle C Landfills in Oregon (7,321 kg in 2014; 7,000 kg in 2013; and 6,076 kg in
2015), and Offsite Other Landfills in Indiana (862 kg in 2019; 603 kg in 2018; and 354 kg in 2017). Air
release as reported by TRI are available in Figure 2-4.

Tribal Lands
On-site Disposal

• Disposal to RCRASubtitle C Landfill

¦	Underground I njection to Class 1 Wells
Offsite Disposal

a. Disposal to Other Landfills
c Disposal to RCRASubtitle C Landfills

¦	Underground Injection to Class 1 Wells
Mass of Disposal (kg)

0.001 - 1.00
o >1.00-10.0
O >10.0-100
O >100-862

200 400	800 1,200 1,600

l Kilometers

Figure 2-3.1,4-Dioxane Annual Releases to Land as Reported to TRI, 2013-2019

Note: Some symbols for individual years may overlap and obscure annual releases at each site.

Alaska, American Samoa, Guam, Hawaii, N. Mariana Islands, Puerto Rico, and the U.S. Virgin Islands are not

shown as there are no known modeled or estimated releases.

For air releases, the largest emissions have been in Illinois (9,943 kg/year), South Carolina (3,495
kg/year), and Texas (2,097 kg/year). Collectively, these figures give insight into the spatial distribution
of releases and corresponding amount across the contiguous United States. A full summary of these
estimates can be found in Appendix E.

Page 47 of 484


-------
1677

1678

1679

1680

1681

1682

1683

1684

1685

1686

1687

1688

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Tribal Lands
Total Air Emissions (kg/yr)

O 0.00-10.0

•	>10.0-100

•	>100 - 1,000

•	>1,000 - 10,000

•	>10,000 - 100,000

200 400	800	1,200 1,600

I Kilometers

Figure 2-4. 1,4-Dioxane Annual Releases to Air as Reported by TRI, 2013-2019

Note: Some symbols for individual years may overlap and obscure annual releases at each site.

Alaska, American Samoa, Guam, Hawaii, N. Mariana Islands, Puerto Rico, and the U.S. Virgin Islands are not

shown as there are no known modeled or estimated releases.

In addition to mapping releases from TRI and DMR, EPA also mapped hydraulic fracturing sites
reporting the presence of 1,4-dioxane in hydraulic fracturing operation fluids according to FracFocus 3.0
(GWPC and IOGCC. 2022). These operations are primary sited in a wide range of shale plays across the
country (as indicated by the multi-colored plays mapped in Figure 2-5). The Delaware play in Texas has
the largest number of operations (n = 158) followed by the Niobrara in Colorado (n = 86) and the Utica
play that spreads across Pennsylvania and Ohio (n = 70).

Page 48 of 484


-------
1689

1690

1691

1692

1693

1694

1695

1696

1697

1698

1699

1700

1701

1702

1703

1704

1705

1706

1707

1708

1709

1710

1711

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Figure 2-5. Locations of Hydraulic Fracturing Operations that Report 1,4-Dioxane in
Produced Waters

Note: Some symbols for individual years may overlap and obscure annual releases at each site.

Alaska, American Samoa, Guam, Hawaii, N. Mariana Islands, Puerto Rico, and the U.S. Virgin Islands are
not shown as there are no known modeled or estimated releases.

The basis for overall data quality determinations is also included in the water, air, and land summary
subsections below. Each source is evaluated on multiple metrics based on defined criteria. For air, water,
and land releases, all monitoring data had data quality ratings of medium/high. Modeled data had data
quality ratings of medium or high.

2.2.1.2 Weight of the Scientific Evidence Conclusions for Environmental Releases

EPA's judgment on the weight of the scientific evidence is based on the strengths, limitations, and
uncertainties associated with the release estimates. The Agency considers factors that increase or
decrease the strength of the evidence supporting the release estimate—including quality of the
data/information, applicability of the release data to the COU (including considerations of temporal
relevance, locational relevance) and the representativeness of the estimate for the whole industry. The
best professional judgment is summarized using the descriptors of robust, moderate, slight, or
indeterminant, according to EPA's Application of Systematic Review in TSCA Risk Evaluations (U.S.
EPA. 2018c). For example, a conclusion of moderate weight of the scientific evidence is appropriate
where there is measured release data from a limited number of sources such that there is a limited
number of data points that may not cover most or all of the sites within the COU. A conclusion of slight
weight of the scientific evidence is appropriate where there is limited information that does not

Page 49 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

sufficiently cover all sites within the COU, and the assumptions and uncertainties are not fully known or
documented. See EPA's Application of Systematic Review in TSCA Risk Evaluations (U.S. EPA.. 2018c)
for additional information on weight of the scientific evidence conclusions.

For air, water, and land releases, all monitoring data had data quality ratings of medium/high. Modeled
data had data quality ratings of medium or high. For supplemental releases modeled with TRI/DMR
(PET Byproduct, Ethoxylation Byproduct, Disposal), the weight of the scientific evidence conclusion
was moderate to robust because the reasonably available information relevant for the conditions of use
of 1,4-dioxane at facilities in TRI and DMR is limited. For releases that used SHEDS-HT modeling
(Surface Cleaner, Dish Soap, Dishwasher Detergent), the weight of the scientific conclusion was slight
since there is uncertainty in the application of this modeling for a commercial setting, and this case study
does not represent all sites in this OES. For supplemental releases that used GS/ESDs, the weight of the
scientific conclusion was moderate when used in tandem with Monte Carlo modeling (Textile Dye,
Laundries), and slight/moderate when used alone (Antifreeze, Paint and Floor Lacquer). For Hydraulic
Fracturing, the weight of the scientific conclusion was moderate to robust since FracFocus 3.0, an ESD,
and Monte Carlo modeling were used. See Appendix E.8 for a summary of EPA's overall weight of the
scientific evidence conclusions for its release estimates for each of the assessed OESs.

2.2.1.3 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
Environmental Release Assessment

EPA estimated air, water, and land releases of 1,4-dioxane using various methods and information
sources, including TRI and DMR data, GSs and ESDs modeling with and without Monte Carlo, process
information, and SHEDS-HT DTD Modeling.

TRI and DMR were determined to have the overall data quality determination of medium through
EPA's systematic review process. Uncertainties for using TRI and DMR data are discussed in the Final
Risk Evaluation for 1,4-Dioxane (	:0c). In summary, these uncertainties include

underestimation of the number of sites for a given OES due to reporting thresholds in TRI, the accuracy
of EPA's mapping of sites reporting to TRI and DMR to a specific OES, and quality of the data reported
to TRI and DMR. There is additional uncertainty for sites that report to TRI with Form A because these
sites do not report release quantities if the quantity did not exceed 500 lb for the total annual reportable
release amount. For these sites, EPA assessed a "what-if' scenario, which assumes the entire 500 lb
going to single media of release, noting that the 500 lb should not be added over all release media.
Additional information on TRI uncertainties is provided in Appendix E.5. In addition, as discussed in
Section 2.2.1, EPA used data from the 2019 reporting year to estimate air and land releases. A key
source of uncertainty in the assessment of air and land releases is whether 2019 TRI data is
representative of releases from other reporting years. This does not apply to the water release estimates
because EPA used data from reporting years 2013 to 2019. A strength of using TRI is that it compiles
the best readily available release data for all facilities that reported to EPA. However, not all facilities
are required to report to TRI.

Some uncertainties of using DMR data include the accuracy of EPA's mapping of sites reporting to
DMR to a specific OES, and quality of the data reported to DMR. Also, an uncertainty of using the
ECHO Pollutant Loading Tool Advanced Search option is that average measurements may be reported
as a quantity (kg/day) or a concentration (mg/L). Calculating annual loads from concentrations requires
adding wastewater flow to the equation, which increases the uncertainty of the calculated annual load. In
addition, for facilities that reported having zero pollutant loads to DMR, the EZ Search Load Module
uses a combination of setting non-detects equal to zero and as one-half the detection limit to calculate
the annual pollutant loadings. This method could cause overestimation or underestimation of annual and

Page 50 of 484


-------
1760

1761

1762

1763

1764

1765

1766

1767

1768

1769

1770

1771

1772

1773

1774

1775

1776

1777

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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Daily pollutant loads. A strength of using DMR data and the Pollutant Loading Tool is that the tool
calculates an annual pollutant load by integrating monitoring period release reports provided to the EPA
and extrapolating over the course of the year. However, this approach assumes average quantities,
concentrations, and hydrologic flows for a given period are representative of other times of the year.

Where TRI and DMR data were not reasonably available, EPA used GS and ESDs. One uncertainty for
this method is lack of specific 1,4-dioxane data. Because GS/ESDs are generic, assessed parameter
values may not always be representative of applications specific to 1,4-dioxane use in each OES.
Another uncertainty is lack of consideration for release controls. The GS/ESDs assume that all activities
occur without any release controls, and in an open-system environment where vapor and particulates
freely escape (U.S. EPA. 2022d; QECn 1 , J01 la. b). Actual releases may be less than estimated if
facilities utilize pollution control methods. Although 1,4-dioxane monitoring data are preferred to
modeled data, EPA strengthened modeled estimates by using Monte Carlo modeling to allow for
variation in environmental release calculation input parameters according to the GS/ESD and other
literature sources. However, EPA did not utilize Monte Carlo modeling for all GS/ESD, which is a
limitation of this assessment. TableApx E-8 includes information on which GS/ESDs were used in
tandem with Monte Carlo modeling.

EPA used process information to quantify environmental releases for the film cement and dry film
lubricant OESs. This process information is from the 2020 RE (	20c) and the underlying

sources were determined to have high overall data quality determinations through EPA's systematic
review process. To develop these release estimates, EPA made assumptions on the likely media of
release for various releases sources and, in some cases, used standard EPA models in conjunction with
process information to estimate the release quantity. A source of uncertainty in this approach is the
representativeness of these estimates regarding all sites that use 1,4-dioxane for this OES.

Where no other data or information was reasonably available, EPA used SHEDS-HT down the drain
(DTD) modeling to estimate commercial use environmental releases to surface water or land
(Appendices E.3.2 and E.4.2). The results for this analysis are included in Table Apx E-3. The main
source of uncertainty is that the modeling EPA performed to estimate the total release amounts from
each COU to surface water or land (via disposal to landfills) is based on information for SHEDS-HT in
combination with information from a single case study location, Liverpool, OH. It is uncertain whether
the release estimates generated from this case study are applicable to other areas of the country. EPA
was unable to estimate the number of sites in Liverpool, OH, for the OES where this modeling approach
was used; therefore, the release estimates were presented as totals for all sites as opposed to per-site
estimates. To estimate land release, EPA used the modeled water releases from SHEDS-HT and back-
calculating a 1,4-dioxane use rate based on the expected loss fraction to water for the OES. Finally, a
land release loss fraction was applied to the back-calculated use rate. The uncertainty in this approach is
due to the standard models and assumptions used to estimate loss fractions to water and land.
Additionally, the same uncertainties listed above for the use of SHEDS-HT to estimate water releases
are applicable to the approach for estimating land releases. EPA is unsure whether the use of SHEDS-
HT results in a high-end or typical exposure scenario, so the use of this data may lead to over or
underestimates of releases.

To assess daily air and water discharges, EPA divided annual release loads by the number of facility
release days to estimate the daily release load for the facility. There is uncertainty if the assumed release
duration is applicable to all sites for a given OES; therefore, the average daily releases may be higher if
sites have fewer release days or lower if they have greater release days. Furthermore, 1,4-dioxane
concentrations in air emissions and wastewater release to receiving water bodies at each facility may

Page 51 of 484


-------
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

1848

1849

1850

1851

1852

1853

1854

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

vary from day-to-day such that on any given day the actual daily releases may be higher or lower than
the estimated average daily discharge. Thus, this approach minimizes variations in emissions and
discharges from day to day. EPA did not estimate daily land releases due to the high level of uncertainty
in the number of release days associated with land releases. The Agency expects that sites may not send
waste to landfills every day and are more likely to accumulate waste for periodic shipments to landfills.
However, sites that release to municipal landfills may have more frequent release days based on the
frequency of shipments.

2.3 1,4-Dioxane Environmental Concentrations

2.3.1 Surface Water Pathway

Surface water contamination from 1,4-dioxane can occur from direct releases of wastewater from
industrial operations, discharges from wastewater treatment plants containing DTD releases of 1,4-
dioxane from consumer and commercial product usage {i.e., dish soap, laundry detergent, etc.), and
other activities where 1,4-dioxane may be present as a byproduct, such as in hydraulic fracturing
operations. To understand possible exposure scenarios from these practices, EPA assessed exposures to
the general population from ambient surface and drinking water. These exposures are due to 1,4-dioxane
being directly or indirectly discharged to receiving water bodies.

The evaluation of these exposures considered both the review of reasonably available monitoring data to
both ambient surface water and drinking water as well as the modeling of estimated exposures due to
releases. Although EPA identified a robust set of surface and drinking water monitoring data (Section
2.3.1.1) indicating the presence of 1,4-dioxane in these pathways, it was collected independent of release
data, and cannot be attributed to specific sources (Section 2.2). Therefore, EPA relied primarily on a
series of modeling approaches to estimate concentrations of 1,4-dioxane in surface water near known
release locations (Sections 2.3.1.2 and 2.3.1.3). For this assessment, EPA modeled concentrations
resulting from industrial releases for all COUs releasing to surface water, including those assessed in the
2020 RE, as well as those producing 1,4-dioxane as a byproduct. To the degree possible, the relationship
between monitoring and modeled data is further evaluated in Section 2.3.1.4.

2.3.1.1 Monitoring Data

Environmental concentration data for 1,4-dioxane in ambient surface water {i.e., measured in rivers,
streams, lakes, and ponds, rather than within industrial operations or drinking water systems) across the
country, as well as routine monitoring conducted by public water systems (PWSs) of raw (untreated)
source water and finished (treated) drinking water were collected from readily available public databases
and publications. The methods for retrieving and processing ambient surface water and PWS data are
described in detail in Appendix G. 1.

Ambient Surface Water Monitoring

Data were retrieved from the Water Quality Portal (WQP) to characterize observed concentrations in
ambient surface water (NWQMC. 2022). These monitored values may or may not represent locations
used as a source for drinking water and are analyzed to characterize the observed ranges of 1,4-dioxane
concentrations in ambient surface water—irrespective of the reasons for sample collection—and to
provide context for the modeled surface water concentrations presented in Section 2.3.1.3. Data
retrieved in July 2022 included sampling dates from 1997 to 2022 and resulted in 12,471 available
sample results. Full details of the retrieval and processing of ambient surface water monitoring data from
the WQP are presented in Appendix G. Table 2-2 shows the range of 1,4-dioxane concentrations
detected in surface water samples. Most {i.e., 92.3 percent) of the sample records available had no level
of 1,4-dioxane detected above the reported detection limit for the analysis (referred to as "non-detects"),

Page 52 of 484


-------
1855

1856

1857

1858

1859

1860

1861

1862

1863

1864

1865

1866

1867

1868

1869

1870

1871

1872

1873

1874

1875

1876

1877

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

with limits of detection ranging from 0.001 to 28,000 |ig/L across all samples. The 105 detected values
ranged from 0.016 to 470 |ig/L, with a median of 1.10 |ig/L. Figure 2-6 and Figure 2-7 show the
distribution of detected concentrations and reported detection limits of non-detect samples, respectively.
The highest concentrations reported in this dataset are noted in the metadata to have been collected at
the point of discharge from an industrial facility, while for most samples, the reason for sampling, or
sample location in relation to expected releases is not included in the metadata. Figure 2-8 shows the
spatial distribution of detected samples. For the entire dataset, including non-detects, approximately 70
percent of the samples were collected from the states of North Carolina, New Mexico, and New Jersey.
Of the 105 detected values, 46 percent are located in Pennsylvania, 21 percent in North Carolina, and 14
percent in Illinois. In the absence of a national standardized study of 1,4-dioxane in ambient surface
water (analogous to the UCMR monitoring in drinking water), and without more national coverage and
metadata, it is difficult to characterize the national occurrence of 1,4-dioxane in surface water. It is
apparent from the available monitoring data that certain areas may be more likely to have higher
concentrations, while many others have little or no detected 1,4-dioxane. Over-representation of certain
states or regions may reflect targeted sampling campaigns of specific locations expected to have higher
concentrations, and conclusions about areas without monitoring data cannot be drawn without further
exploration through modeling.

Detected 1,4-Dioxane Concentration (|jg/L)

Figure 2-6. Frequency of Nationwide Measured 1,4-Dioxane Surface Water Concentrations
Retrieved from the Water Quality Portal, 1997-2022

Note: Detectable levels of 1,4-dioxane may vary by location.

Page 53 of 484


-------
1878

1879

1880

1881

1882

1883

1884

1885

1886

1887

1888

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

c



CD

250-

c



o



OJ

?no-

1—



c



CI)



o
c

150 -

o



o



4—



o

100-

>N



o



c



CI)

50-



cr



0



LL



i i 11 in|" i i 1111n| i i i I nn| i i i 111n| i i i I iiii| i i i 11 ii'i| ' i i i 11 iii^ i ill 1111| i ^ i I ii

10~3 10~2 1CT1	10°	101	102	103	104

Detection Limit for 1,4-Dioxane Non-Detect (pg/L)

Figure 2-7. Frequency of Detection Limits for Nationwide Non-detect 1,4-Dioxane
Surface Water Samples Retrieved from the Water Quality Portal, 1997-2022

Figure 2-8. Detectable Concentrations of 1,4-Dioxane in Surface Water from the Water
Quality Portal, 1997-2022

Note: Detectable levels of 1,4-dioxane may vary by sampling location.

Alaska, American Samoa, Guam, Hawaii, N. Mariana Islands, Puerto Rico, and the U.S. Virgin
Islands are not shown as there are no known monitoring data above detection limits.

Page 54 of 484


-------
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

1919

1920

1921

1922

1923

1924

1925

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Drinking Water Monitoring Data

The Safe Drinking Water Act (SDWA) authorizes the United States Environmental Protection Agency
(US EPA) to set national health-based standards for drinking water to protect against both naturally
occurring and man-made contaminants that may be found in drinking water. The National Primary
Drinking Water Regulations (NPDWR) are legally enforceable primary standards and treatment
techniques that apply to PWSs. Although states, tribes or territories that have been approved as the
primary implementation authority for drinking water may require monitoring or impose limits for
contaminants beyond those regulated under SDWA, there are not currently national requirements to
routinely monitor or limit 1,4-dioxane in finished water from PWSs. In support of the SDWA, EPA
often relies on data from the Unregulated Contaminant Monitoring Rule (UCMR) program as the best
available occurrence information to support its regulatory determinations (i.e., to judge whether a
particular contaminant is known to occur or there is substantial likelihood the contaminant will occur in
public water systems with a frequency and at levels of public health concern). UCMR monitoring is
designed to produce a data set that is nationally representative of public water systems (PWSs) across
the country, but its focus is on finished water (rather than source water) and it may not capture worst-
case conditions. PWS monitoring data of finished drinking water were collected for 1,4-dioxane via
EPA's published Third UCMR (UCMR3) dataset from 2013 to 2015, as well as raw and finished
drinking water monitoring from additional individual state databases (CA, MA, and NY) from 2008 to
2022 (CA. Water Board. 2022; NY DOR 2022; Commonwealth of Massachusetts. 201N; 1 ^ \
201 ?d). UCMR3 data were filtered to only include facilities flagged as using surface water, while the
individual state data were filtered down to only those systems with surface water listed as the primary
source in SDWIS. Datasets were processed to ensure that no samples were repeated in multiple datasets.
These PWS monitoring data were collected to assess possible exposures to the general population
through drinking water. Descriptions of the data retrieval and processing methods are presented in
Appendix G.2.

The combined datasets resulted in 16,972 samples from 2,847 PWSs across 50 states (Table 2-2).
Reported detection limits across the PWS datasets ranged from 0.0001 to 3 |ig/L, with 81 percent of
samples reporting a detection limit of 0.07 |ig/L. To the extent that it could be determined from the
database records, samples were separated into raw (untreated) water from the PWS intake or finished
(treated) water being sent to the distribution system. The distribution of raw water monitoring
concentrations is presented in Figure 2-9, and the distribution of finished drinking water concentrations
is presented in Figure 2-10.

Table 2-2. Summary of PWS Monitoring Datasets of 1,4-Dioxane Monitoring in PWSs Using
Surface Water as a Source

Dataset of
Origin

Number of
Samples

Minimum
Concentration
(Hg/L)

Median
Concentration

fag/L)

Maximum
Concentration
(Hg/L)

Start
Year

End
Year

CA

1,797

0.25

0.5

1.5

2011

2022

MA

949

0.049

0.22

3.8

2008

2022

NY

615

7.20E-05

0.035

1

2015

2022

UCMR3

13,611

0.035

0.035

13.3

2013

2016

Note: for the summary presented in this table, results reported as below their respective detection limit were
assigned a value of half of the detection limit.

Page 55 of 484


-------
1926

1927

1928

1929

1930

1931

1932

1933

1934

1935

1936

1937

1938

1939

1940

1941

1942

1943

1944

1945

1946

1947

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Detected 1,4-Dioxane Concentration in Raw Water (ug/L)

Figure 2-9. Frequency of 1,4-Dioxane Concentrations Monitored in Raw (Untreated)
Drinking Water Derived from Surface Water

Data retrieved from state databases (CA, MA, and NY) between 2008-2022.

Detected 1,4-Dioxane Concentration in Finished Drinking Water (ug/L)

Figure 2-10. Frequency of 1,4-Dioxane Concentrations Monitored in Finished (Treated)
Drinking Water Derived from Surface Water.

Data retrieved from the UCMR3 and state databases (CA, MA, and NY) between 2008-2022

Note: the detection limit for the method used in UCMR3, and the most common detection limit reported

in state databases is 0.07 (ig/L.

Water treatment systems may vary widely across the country based on available and utilized water
treatment processes that depend on whether source water is groundwater or surface water. These
processes typically include disinfection, coagulation/flocculation, sedimentation, and filtration (U.S.
EPA. 2006a). In assessing drinking water exposures, the ability to treat and remove or transform
chemicals in possible drinking water supplies should be considered. Typical treatment processes do not
remove 1,4-dioxane from ambient surface water and groundwater prior to possible general population
consumption as drinking water and treatment processes that do effectively remove 1,4-dioxane are
uncommon. EPA therefore assumes zero removal in the following analyses to provide a conservative
estimate of general population drinking water exposures. Even without treatment processes that remove

Page 56 of 484


-------
1948

1949

1950

1951

1952

1953

1954

1955

1956

1957

1958

1959

1960

1961

1962

1963

1964

1965

1966

1967

1968

1969

1970

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

1,4-dioxane, multiple sources of water may be mixed within the same drinking water system which may
result in finished water with lower concentrations than one of the higher contributing source water
concentrations. EPA acknowledges that the surface water concentration at a single intake location may
be higher than the finished drinking water once mixed with other sources. Further discussion of the
prevalence of treatment processes across water systems, and the methodology for identifying raw and
finished drinking water monitoring samples is presented in Appendix G.1.2.

Figure 2-11 shows the spatial distribution of UCMR3 samples at the county level, with 1,4-dioxane
detected in 25 percent or 240 of 943 counties with participating water systems.

Maximum Reported 1,4-dioxane
Concentration

(M9/L)

¦¦ 6.21 - 13.30
2.49 - 6.20
1.21 - 2.48
0.62 - 1.20
0.07 - 0.61
I I < 0.07 (Not Detected)

I I No Data

125 250

500

750

1,000
I Miles

Figure 2-11. Map of Counties Containing PWSs that Reported Monitoring of Finished
Drinking Water Drawn from Surface Water for 1,4-Dioxane under UCMR3

Note: UCMR3 monitoring of 1,4-dioxane required four sampling events, one for each season, to capture
temporal variability. Each county highlighted may include one or multiple PWSs reporting data.

Monitored drinking water data were also included in exposure and risk estimates to assess the human
health implications of drinking water concentrations in this range. Since the UCMR program and state
monitoring datasets are not designed to reflect source water impacts of direct and indirect releases into
water bodies, EPA's TSCA program relied on estimated concentrations modeled for a range of specific
release scenarios to characterize risks from the water pathway. EPA evaluated the performance of the
models used to estimate water concentrations with monitoring data from site-specific locations serving
as cases studies. These case study comparisons demonstrated strong concordance between modeled

Page 57 of 484


-------
1971

1972

1973

1974

1975

1976

1977

1978

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

concentrations and monitoring data, thereby increasing confidence in risk estimates based on modeled
concentrations.

Although monitoring data confirm that 1,4-dioxane is present in drinking water in some locations,
samples collected under the UCMR program are designed to be nationally representative of drinking
water occurrence and not specifically associated with industrial releases of 1,4-dioxane. Since these
monitoring data may not reflect the 1,4-dioxane concentrations that result from industrial releases, EPA
relied on modeling to estimate 1,4-dioxane concentrations that occur near release sites.

2.3.1.2 Surface Water and Drinking Water Modeling

To assess possible general population exposures to 1,4-dioxane via industrial releases to surface water,
concentrations of 1,4-dioxane in surface water were modeled using two separate approaches. First, a
facility-specific approach aimed to quantify the maximum expected aqueous concentrations resulting
from reported 1,4-dioxane discharges from individual facilities in isolation. Second, a probabilistic
model was applied to assess the range of expected aqueous concentrations resulting from reported 1,4-
dioxane discharges across a COU, with consideration of expected ranges of background concentrations
of 1,4-dioxane from DTD loading and other unreported releases.

2.3.1.2.1 Modeling Methodology

A detailed description of modeling methods is presented in Appendix G.2.

As described in Section 2.2, annual releases of 1,4-dioxane to surface water from regulated dischargers
were retrieved from TRI and DMR. To the extent possible, modeled hydrologic flow data {i.e., stream
flow) associated with the receiving water body to which each facility released was retrieved from the
NHDPlus V2.1 dataset (	2016). The receiving water body was identified either

through NPDES permit information for the releasing facility, or the nearest identified NFIDPlus V2.1
flowline. Detailed methods for the retrieval and processing of flow data are presented in Appendix
G.2.1.

Facility-Specific Modeling

Facility-specific modeling was conducted to estimate concentrations in receiving water bodies resulting
from the greatest facility-specific annual release reported between 2013 through 2019. This modeling
approach employed the equations used to model releases from facilities in the E-FAST 2014 model
(I	E014) and is described in Appendix G.2.2. For each facility and annual release amount, three

different scenarios for days of release per year were considered: 1 day, 30 days, and expected number of
days of operation reported in Table Apx E-2 (referred to as the "maximum" number of days and ranges
from 250 to 365 days depending on OES). These additional scenarios with lower numbers of days of
operation provide more conservative estimates of resulting surface water concentrations and are
intended to evaluate the full range of possible facility release patterns based on the best available
information. Two flow metrics were evaluated: the lowest monthly average flow from NHDPlus, and the
harmonic mean flow derived from E-FAST 2014 methodology. The resulting concentrations from the
facility-specific modeling are used in calculations of general population exposure and human health
outcomes.

Probabilistic Modeling

The probabilistic modeling approach was conducted to consider multiple years of release data per
facility and multiple modeled flow metrics from NHDPlus V2.1 (	) per facility

to generate a distribution of potential surface water concentrations resulting from releases across each
COU. The underlying model for the probabilistic approach is a fit-for-purpose model developed by EPA

Page 58 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

in Microsoft Excel, the EWISRD-XL model (Estimating Water Industrial Surface Release and Down
the Drain in Excel). The EWISRD-XL model was designed to model 1,4-dioxane inputs to a stream
segment, including existing in-stream concentrations, DTD loading from consumer and commercial
products, and industrial releases, as a steady-state snapshot of a single point in time (Figure 2-12).
Outputs from the model include the resulting downstream concentration and the relative contribution
from each input source to that concentration. Examples of the EWISRD-XL model applied to three
specific case study locations (Brunswick County, NC, Columbia, TN, and Liverpool, OH) are included
in three Supplemental Information Files (U.S. EPA. 2023 o. p, g).

Inputs: -

POTW Loading:
Down-the-drain
from upstream
population
Indirect industrial
releases

Figure 2-12. Schematic of the EWISRD-XL Model Inputs and Outputs

For the probabilistic 1,4-dioxane COU modeling, an R script (R Core Team. 2022) was developed to
rapidly run multiple iterations of the EWISRD-XL model. In this configuration, called the EWISRD-
XL-R model, the underlying calculations were performed by EWISRD-XL model, and an R script
wrapper managed the processing of input and output data. For the probabilistic COU modeling, the
EWISRD-XL-R model developed to calculate the receiving water body concentrations at the point-of-
release by a facility. The EWISRD-XL-R results include the concentrations due only to releases from
facilities, as well as an estimated background concentration of DTD and unmonitored releases. The full
details of the underlying EWISRD-XL model and the probabilistic implementation are presented in
Appendix G.2.3. Distributions of total concentrations resulting from combinations of facility releases
and background concentrations were used for calculations of general population exposure and human
health outcomes.

A series of case studies was developed with the EWISRD-XL model to evaluate its performance across
various 1,4-dioxane release settings. These cases are presented in Appendix G.2.3.2.

Page 59 of 484


-------
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

2077

2078

2079

2080

2081

2082

2083

2084

2085

2086

2087

2088

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2.3.1.2.2	Estimating Down-the-Drain Releases

To evaluate the anticipated ranges of DTD contributions of 1,4-dioxane to water bodies receiving
POTW effluent, a range of combinations of hydrologic flows and populations served by a POTW were
evaluated using the EWISRD-XL-R model. For this modeling exercise, only contributions from the
DTD component were used to calculate resulting surface water contributions {i.e., no facility releases or
existing background concentrations were included). Hypothetical combinations of hydrologic flows and
populations contributing to wastewater loading derived from the national distribution of hydrologic
flows and populated places were selected to represent a range of results, which were then compared with
concentrations expected from industrial releases and used to calculate ranges of human exposure and
risk. More detailed methodology for this calculation is presented in Appendix G.2.3.4.

2.3.1.2.3	Hydraulic Fracturing

Hydraulic fracturing is a process used to extract oil and gas from shale plays. After hydraulic fracturing
operations inject fluids to extract oil and gas, a substantial volume of water may be produced through
flowback. The composition of these produced waters depends both on the geochemistry of the injected
area and the injected fluids. 1,4-Dioxane has been reported to EPA as one of the chemicals present in
these produced waters by 411 facilities via FracFocus 3.0 (GWPC and IOGCC. 2022). Estimated 1,4-
dioxane loadings of produced water to surface water from hydraulic fracturing activities (described in
Appendix E.9) were evaluated for expected ranges of resulting concentrations in receiving water bodies
using the EWISRD-XL-R model. Hydraulic fracturing wells reporting 1,4-dioxane use by FracFocus 3.0
were mapped, and flow data from nearby water bodies were collected from NHDPlus V2.1. A Monte
Carlo analysis was used to generate loadings to receiving water bodies from the distribution of modeled
releases and to pair them with hydrologic flows, resulting in a distribution of possible surface water
concentrations. Methodology for this analysis is presented in Appendix G.

2.3.1.2.4	Proximity to Drinking Water Sources

Drinking water exposures from facility-specific results assumed that the exposure occurs at the receiving
water body to provide a conservative estimation of drinking water exposures. However, the evaluated
water bodies may not be used as, or proximate to, actual drinking water sources and intakes. To give a
more robust characterization of possible drinking water exposures, known facility-specific releases were
mapped to drinking water sources using public water systems data stored in EPA's Safe Drinking Water
Information System Federal Data Warehouse (U.S. EPA. 2022f). This dataset is updated quarterly, and
the 2nd quarter 2022 version was used for this analysis. Following mapping, the colocation of and
proximity of releases to drinking water sources were evaluated. Locations of raw water intakes for
PWSs are considered sensitive by EPA Office of Water due to public safety concerns. Geospatial
analysis and the NHDPlus V2.1 flowline network were used to assess whether any known drinking
water intakes are located downstream of 1,4-dioxane releasing facilities. Methodology for this analysis
is presented in Appendix G.2.4.

2.3.1.3 Modeling Results

2.3.1.3.1 Facility-Specific Results

The facility-specific results show the expected concentration at the point of release from the facility
discharging 1,4-dioxane to receiving water bodies, without consideration of the contribution from other
sources. The total number of modeled releases within a given OES may be greater than the number of
1,4-dioxane releasing facilities in cases where facilities indirectly dispose of 1,4-dioxane by transferring
to another facility in addition to directly discharging 1,4-dioxane. Surface water concentrations resulting
from facility-specific modeling for one day of release are summarized in Table 2-3 and represent the
highest expected concentrations in receiving water bodies, due to the annual release amount being

Page 60 of 484


-------
2089

2090

2091

2092

2093

2094

2095

2096

2097

2098

2099

2100

2101

2102

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

discharged in a single day. Surface water concentrations resulting from facility-specific modeling for
maximum days of release are summarized in Table 2-4, and represent the lowest expected
concentrations in receiving water bodies due to the annual release spread out over the most days. The
single day release scenario allows consideration of a "worst-case scenario" given the available annual
release information and can inform an upper limit of concentrations resulting from releases. The
maximum days release scenario can inform a lower limit of expected concentrations from the available
annual release data. Full discussion on the evaluation of multiple release days is given in Section
2.3.1.2.1, but the range of evaluated release days is intended to provide to full range of expected surface
water concentrations resulting from possible facility release patterns and available information. As
described in Section 5, these variations in concentration due to days of release do not affect chronic
cancer risk estimates resulting from a particular releasing facility, due to annual averaging of exposure.
Resulting concentrations varied widely, both across and within OESs, due to variability in facility
release amounts as well as receiving water body flow magnitudes. Facility-specific releases are
organized around their identified OES as fully described in Section 2.1 and Appendix D.

Page 61 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2103	Table 2-3. Summary of Surface Water Concentration Results by OES from Facility-Specific Modeling of Annual Maximum Releases

2104	between 2013 and 2019 for 1 Operating Day per Year			

OES

No. of
Releases
Modeled

Sum of

Annual

Releases

Modeled

(kg/year)

Annual Release by Facility
(kg/site-vear)

Surface Water Concentration
(Lowest Monthly Flow)
(jig/L)

Surface Water Concentration
(Harmonic Mean Flow)
(Hg/L)

Min

Mean

Max

Min

Mean

Max

Min

Mean

Max

Disposal

25

16,997

1.36E-04

6.80E02

7.95E03

1.50E-02

6.45E05

9.52E06

1.50E-02

4.77E05

7.34E06

Ethoxylation
Byproduct

8

112,076

4.54E-01

1.40E04

1.12E05

5.39E-03

2.58E06

2.07E07

3.01E-03

1.22E06

9.73E06

Functional
Fluids (Open-
System)

6

17,711

3.80E-01

2.95E03

1.75E04

1.39E01

1.57E03

4.78E03

6.07E00

7.40E02

2.21E03

Import and
Repackaging

12

2,722

2.27E02

2.27E02

2.27E02

1.08E01

8.15E06

9.28E07

4.39E00

1.01E06

7.40E06

Industrial Uses

31

70,343

2.07E-01

2.27E03

2.62E04

1.33E-02

5.11E05

4.64E06

6.52E-03

4.53E05

5.15E06

Manufacture

2

7,034

1.67E03

3.52E03

5.36E03

8.31E04

1.63E06

3.18E06

8.31E04

1.63E06

3.18E06

PET

Manufacturing

19

2,773,355

3.40E-01

1.46E05

2.51E06

2.77E00

1.07E06

1.66E07

1.28E00

1.05E06

1.66E07

Printing Inks

1

5

5.45E00

5.45E00

5.45E00

2.05E03

2.05E03

2.05E03

2.05E03

2.05E03

2.05E03

Remediation

16

46

3.40E-05

2.91E00

2.39E01

1.50E-03

1.83E03

1.79E04

3.54E-04

1.52E03

1.37E04

Overall

120

3,000,290

3.40E-05

2.50E04

2.51E06

1.50E-03

1.45E06

9.28E07

3.54E-04

5.92E05

1.66E07

2105

Page 62 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2106	Table 2-4. Summary of Surface Water Concentration Results by OES for Facility-Specific Modeling of Annual Maximum Releases

2107	between 2013 and 2019 for the Maximum Operating Days per Year		

OES

No. of
Releases
Modeled

Sum of

Annual

Releases

Modeled

(kg/year)

Annual Release by Facility
(kg/site-vear)

Surface Water Concentration
(Lowest Monthly Flow)
(lig/L)

Surface Water Concentration
(Harmonic Mean Flow)

(Hg/L)

Min

Mean

Max

Min

Mean

Max

Min

Mean

Max

Disposal

25

16,997

1.36E-04

6.80E02

7.95E03

5.99E-05

2.57E03

3.81E04

5.99E-05

1.90E03

2.94E04

Ethoxylation
byproduct

8

112,076

4.54E-01

1.40E04

1.12E05

2.16E-05

1.03E04

8.26E04

1.20E-05

4.87E03

3.89E04

Functional
Fluids (Open-
System)

6

17,711

3.80E-01

2.95E03

1.75E04

5.63E-02

6.37E00

1.93E01

2.46E-02

3.00E00

8.95E00

Import and
Repackaging

12

2,722

2.27E02

2.27E02

2.27E02

4.32E-02

3.26E04

3.71E05

1.76E-02

4.04E03

2.96E04

Industrial Uses

31

70,343

2.07E-01

2.27E03

2.62E04

5.31E-05

2.04E03

1.86E04

2.61E-05

1.81E03

2.06E04

Manufacture

2

7,034

1.67E03

3.52E03

5.36E03

3.32E02

6.52E03

1.27E04

3.32E02

6.52E03

1.27E04

PET

Manufacturing

19

2,773,355

3.40E-01

1.46E05

2.51E06

1.11E-02

4.29E03

6.63E04

5.12E-03

4.20E03

6.63E04

Printing Inks

1

5

5.45E00

5.45E00

5.45E00

8.21E00

8.21E00

8.21E00

8.21E00

8.21E00

8.21E00

Remediation

16

46

3.40E-05

2.91E00

2.39E01

4.11E-06

5.01E00

4.90E01

9.69E-07

4.15E00

3.75E01

Overall

120

3,000,290

3.40E-05

2.50E04

2.51E06

4.11E-06

5.80E03

3.71E05

9.69E-07

2.37E03

6.63E04

2108

Page 63 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2109	Table 2-5 describes the crosswalk between identified OESs and relevant COUs under each for the

2110	identified facility releases to surface water. The full facility-specific analysis is included in 1,4-Dioxane

2111	Supplemental Information File: Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Release

2112	to Surface Water from Individual Facilities (U.S. EPA. 2023h).

2113

Table 2-5. OES-COU Crosswalk for Identified Facilities Releasing to Sur

'ace Water"

COU*

OES'

Life Cycle Stage

Category

Subcategory'

Manufacturing

Domestic manufacture

Domestic manufacture

Manufacturing

Import

Import
Repackaging

Import and Repackaging



Processing as a reactant

Polymerization catalyst

Industrial Uses



Non-incorporative

Basic organic chemical manufacturing
(process solvent)



Processing

Byproduct

Byproduct produced during the
ethoxylation process to make
ethoxylated ingredients for personal
care products

Ethoxylation Process
Byproduct





Byproduct produced during the
production of polyethlene terephtalate

PET Byproduct



Intermediate use

Plasticizer intermediate
Catalysts and reagents for anhydrous
acid reactions, brominations, and
sulfonations

Industrial Uses

Industrial Usea

Processing aids, not
otherwise listed

Wood pulping

Extraction of animal and vegetable
oils

Wetting and dispersing agent in

textile processing
Polymerization catalyst
Purification of process intermediates
Etching of fluoropolymers

Industrial Uses



Functional fluids (open
and closed systems)

Polyalkylene glycol lubricant
Synthetic metalworking fluid
Cutting and tapping fluid

Functional Fluids (Open
System)

Industrial Use,
Commercial Use

Other Uses

Spray polyurethane foam
Printing and printing compositions,

including 3D printing
Dry film lubricant
Hydraulic fracturing

Printing Inks (3D)

Disposal

Disposal

Remediation

Remediation

Disposal

Disposal

Industrial pre-treatment
Industrial wastewater treatment
Publicly owned treatment works

(POTW)

Underground injection
Municipal landfill

Disposal

Page 64 of 484


-------
2115

2116

2117

2118

2119

2120

2121

2122

2123

2124

2125

2126

2127

2128

2129

2130

2131

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

COU''

OESrf

Life Cycle Stage

Category

Subcategory'





Hazardous landfill
Other land disposal
Municipal waste incinerator
Hazardous waste incinerator
Off-site waste transfer



" Although EPA has identified both industrial and commercial uses here for purposes of distinguishing scenarios in this
document, the Agency interprets the authority over "any manner or method of commercial use" under TSCA section 6(a)(5)
to reach both.

b As mapped to COU Life Cycle Stage, Category, and Subcategory in Table Apx D-l.
c Evaluated facilities within an OES may not encompass all listed COU subcategories.
d Note that identified OESs can encompass multiple COUs across different life cycle stages and categories.

To put the modeled releases in the context of the underlying data sources for release amounts and
receiving water body flow, Table 2-6 presents the results of the process of assigning the receiving water
body (by reach code in the NHDPlus 2.1 dataset) to each releasing facility. Those facilities with reach
code information in their NPDES permit were regarded as the highest confidence in an accurate match
to the actual discharging water body, followed by facilities matched geospatially to the nearest reach
code within 1 km of the facility. Facilities matched to reaches beyond 1 km from the facility but within 2
km provided lower confidence, and those without reach code matches were substituted with the lowest
non-zero flow within the OES as a conservative estimate. The full details of the flow matching process
are presented in Appendix G.2.1. Due to the assumptions described in Section 2.2.1.3 required to model
releases from facilities reporting only via TRI Form A, the percent of facilities within an OES using
Form A is also reported.

Table 2-6. Summary by OES of Data Sources for Releases and Receiving Water Body Flow



Method of Matching to Receiving Water Body



OES

Total
Number of
Releases

NPDES
Permit
Contains
Reach Code

Nearest
Reach
(within 1
km)

Nearest
Reach
within 2 km

Lowest Non-
zero Flow
within OES
Substituted

%of
Releases
Estimated
from TRI
Form A

Disposal

25

22

0

1

1

8

Ethoxylation
Byproduct

8

1

1

1

0

0

Functional Fluids
(Open-System)

6

5

0

0

0

0

Import and
Repackaging

12

1

2

2

7

100

Industrial Uses

31

11

3

5

7

45

Manufacture

2

1

0

0

0

0

PET Manufacturing

19

11

0

1

0

0

Printing Inks

1

1

0

0

0

0

Remediation

16

14

2

0

0

0

Total

120

67

8

10

15

23

A generic table of annual facility release and average flow rates for the receiving water body is
presented in Table 2-7, which demonstrates the relationship between the facility and water body

Page 65 of 484


-------
2132

2133

2134

2135

2136

2137

2138

2139

2140

2141

2142

2143

2144

2145

2146

2147

2148

2149

2150

2151

2152

2153

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

characteristics with regard to the resulting surface water concentrations. Table 2-8 shows the relative
occurrence of each of the releases modeled for this assessment within binned ranges of releases and
flows. Combined, these tables demonstrate that most facilities releasing 1,4-dioxane are initially
discharging to smaller water bodies, even in some cases where large annual release amounts result in
very high modeled concentrations.

Table 2-7. Hypothetical Mean Annual Concentrations (jig/L) for a Range of Annual Release and
Flow Rate Combinations, for a Facility with 250 Days of Release per Year	



Annual Release Amount (kg)

1

10

100

1,000

10,000

100,000

1,000,000

Mean
Annual
Receiving
Water Body
Flow (cfs)

1

1.6

16

160

1.6E03

1.6E04

1.6E05

1.6E06

10

0.16

1.6

16

160

1.6E03

1.6E04

1.6E05

100

0.016

0.16

1.6

16

160

1.6E03

1.6E04

1,000

1.6E-03

0.016

0.16

1.6

16

160

1.6E03

10,000

1.6E-04

1.6E-03

0.016

0.16

1.6

16

160

100,000

1.6E-05

1.6E-04

1.6E-03

0.016

0.16

1.6

16

Table 2-8. Occurrence of Facilities for Distributions of Maximum Annual 1,4-Dioxane Release
Amounts and Receiving Water Body Flow	



Annual Release Amount (kg)

<10

10 to 100

100 to
1,000

1,000 to
10,000

10,000 to
100,000

>100,000

Mean Annual
Receiving Water
Body Flow (cfs)

<10

14%

8%

11%

6%

2%

<1%

10 to 100

9%

7%

2%

<1%

1%

3%

100 to 1,000

3%

6%

2%

1%

<1%

<1%

1,000 to 10,000

1%

2%

2%

2%

2%

<1%

10,000 to 100,000

3%

1%

3%

4%

2%

<1%

2.3.1.3.2 Concentrations from Down-the-Drain Loading

Water concentrations of 1,4-dioxane resulting from DTD releases depend on the population size (an
indicator of the number of people using products and contributing to the releases) and the stream flows
of the receiving water bodies. The representative per capita DTD loading developed from modeling
results from SHEDS-HT was applied to a range of population sizes (100 to 1,000,000 people) and
stream flows (300 to 30,000 cfs) to develop a distribution of potential surface water concentrations.
Estimated surface water at the point of discharge by POTWs resulting from DTD releases ranged from
less than 0.0001 to 110 |ig/L (Table 2-9). The typical ranges of results from this analysis are comparable
to the range of minimum to mean concentrations calculated from individual facility releases in Section
2.3.1.3.1.

Page 66 of 484


-------
2154

2155

2156

2157

2158

2159

2160

2161

2162

2163

2164

2165

2166

2167

2168

2169

2170

2171

2172

2173

2174

2175

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Table 2-9. Estimated Surface Water Concentrations (jig/L) Due to DTD Loading
for a Range of Populations and Hydrologic Flows	



Population Contributing to DTD Loading

100

1,000

10,000

100,000

1,000,000

Receiving
Water Body
Flow (cfs)

100

0.011

0.11

1.1

11

110

300

3.6E-03

0.036

0.36

3.6

36

1,000

1.1E-03

0.011

0.11

1.1

11

3,000

3.6E-04

3.6E-03

0.036

0.36

3.6

10,000

1.1E-04

1.1E-03

0.011

0.11

1.1

The occurrence of POTWs processing wastewater from various populations and the associated flows of
the receiving water bodies were investigated using data from the ICIS-NPDES database (U.S. EPA.
2013). to inform the interpretation of the above ranges of DTD loading concentrations. For communities
with a single POTW treating wastewater, most fell into the range of 100 to 10,000 people, with the
annual average flow of the receiving water body less than 300 cfs (Table 2-10).

Table 2-10. Estimated Percent Occurrence of Combinations of Contributing Population to



Population Contributing to DTD Loading

<100

100 to 1,000

1,000 to
10,000

10,000 to
100,000

100,000 to
1,000,000

Mean Annual

Receiving
Water Body
Flow (cfs)

<100

5%

44%

26%

4%

<1%

100 to 300

<1%

3%

4%

1%

<1%

300 to 1,000

<1%

2%

2%

1%

<1%

1,000 to 3,000

<1%

1%

2%

<1%

<1%

3,000 to 10,000

<1%

<1%

1%

<1%

<1%

>10,000

<1%

1%

1%

1%

<1%

2.3.1.3.3 Concentrations from Hydraulic Fracturing

The Monte Carlo distribution of potential surface water concentrations resulting from hydraulic
fracturing operations is presented in Table 2-11. Hydrologic flows in water bodies near hydraulic
fracturing wells reporting 1,4-dioxane as a constituent of wastewater ranged from less than 10 to 44,300
cfs. Due to the very low flows in many nearby streams, resulting concentrations were sensitive to the
receiving water body flow rate. The distribution of loading to surface water from hydraulic fracturing
represents the loading from a single site of hydraulic fracturing operations (described in Appendix
G.2.3.5) at the immediate point of discharge to the receiving water body. Concentrations estimated at the
highest end are comparable to mean to high-end facility releases presented in Section 2.3.1.3.1. More
than half of the modeled concentrations fell below the typical detection limit in drinking water of 0.07
Hg/L-

Page 67 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Table 2-11. Distribution of Potential Concentrations in Surface
Water Resulting from Hydraulic Fracturing Operations from a
Single Site Reporting 1,4-Dioxane as an Ingredient	

Monte Carlo Distribution

Concentration (jig/L)

Maximum

198

99th Percentile

7.79

95th Percentile

2.60

Median

0.064

5th Percentile

2.31E-04

Minimum

6.08E-12

2.3.1.3.4 Aggregate Probabilistic Results

The aggregate probabilistic model predicts surface water concentrations at the point of facility releases
when incorporating potential contributions from DTD and other unmonitored sources. The model
incorporates multiple years of release data and was run with 10,000 iterations for each OES using different
combinations of direct and indirect facility releases, DTD releases, flows, and background concentrations.
This results in a more descriptive distribution of the potential releases. At the highest end, the results of
the aggregate probabilistic model are similar to those from the facility-specific modeling. This is due to
both the facility-specific modeling and the highest end of the probabilistic modeling being based on the
maximum reported releases from the modeled facilities. Additionally, the loading from facilities far
outweighs the contribution from background sources at the higher end. The shape of the resulting
distribution can be informative in its representation of the frequency of concentrations exceeding a certain
threshold.

Resulting surface water concentrations ranged from 1,45 x ] 0 4 to 7.34><103 |ig/L. Summaries of the
resulting concentrations by OES are presented in Table 2-12 and Figure 2-13. Overall, releases from
facilities tended to result in greater 1,4-dioxane concentrations in surface water than the expected ranges
of background concentrations. Background concentrations were derived from values of 1,4-dioxane
measured by drinking water systems using surface water as a source that were not downstream of known
1,4-dioxane releases (Figure 2-11). The "% of Releases Greater than Background" column in Table 2-12
refers to the frequency of model runs (out of the 10,000 per OES) in which the resulting concentration
from the facility release was greater than the generated background concentration resulting from DTD
and other unregulated surface water loading. A low percentage for this metric may suggest that releases
by a particular OES are typically outweighed by these other unreported releases with respect to their
contribution to surface water concentrations.

Page 68 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2203	Table 2-12. Aggregate Probabilistic Results Showing Distribution of Total 1,4-Dioxane Concentration in Surface Water (Release

2204	Plus Background)									

OES

Min
(^g/L)

5th
Percentile
(jig/L)

25th
Percentile

(jLLg/L)

Median
(jig/L)

75th
Percentile
(jig/L)

95th
Percentile

(^g/L)

Max
(^g/L)

% of Releases
Greater than
Background

Disposal

1.88E-03

1.51E-01

1.98E-01

3.50E-01

8.52E-01

1.94E00

2.02E01

81

Ethoxylation Process Byproduct

2.25E-03

9.81E-02

1.35E-01

2.74E-01

4.65E-01

2.55E00

1.46E01

72

Functional Fluids (Open-System)

1.78E-04

8.20E-02

1.11 E—01

1.58E-01

2.79E-01

1.60E00

6.10E00

48

Import and Repackaging

5.70E-03

1.32E-01

2.83E-01

6.60E01

3.25E02

1.42E03

2.12E03

90

Industrial Uses

1.45E-04

5.15E-02

8.95E-02

1.22E-01

2.52E-01

1.33E01

2.26E02

44

Manufacture

1.10E02

3.32E02

3.32E02

7.19E02

2.32E03

5.48E03

7.34E03

100

PET Byproduct

5.44E-03

1.42E-01

2.13E-01

4.65E-01

4.36E00

1.30E02

2.23E03

84

Printing Inks (3D)

1.51E-01

3.17E-01

1.10E00

4.00E00

6.41E00

7.48E00

8.26E00

96

Remediation

5.34E-04

5.89E-02

9.04E-02

1.35E-01

2.73E-01

6.66E00

1.46E01

47

Page 69 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2205

2206

2207

50
40
30
20
10-

50
40-

(/>

1 30
0

a:

•5 20

2?

10
0

50-

40-

30

20

10

Disposal

TO
T>

-------
2208

2209

2210

2211

2212

2213

2214

2215

2216

2217

2218

2219

2220

2221

2222

2223

2224

2225

2226

2227

2228

2229

2230

2231

2232

2233

2234

2235

2236

2237

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2.3.1.4 Comparison of Modeled and Monitored Surface Water Concentrations

At the higher end, the modeled concentrations from facility releases are several orders of magnitude
greater than those observed in the 1,4-dioxane monitoring data (Figure 2-6 and Table 2-3). This
difference in concentrations may be due to monitoring data being collected further downstream
(allowing for additional dilution), or on reaches that are not impacted by releasing facilities. Many of the
direct releasing facilities, and POTWs assessed for the DTD component, had a receiving water body
specified on their NPDES permits that was associated with a very small stream or industrial canal. These
small receiving water bodies, combined with larger loading values from the releases, resulted in high
modeled concentrations in surface water at the point of release. As this water travels downstream, it is
expected to eventually join with larger water bodies, where some decrease in concentration due to
dilution would occur.

Because most of the reasonably available monitoring data were generally not co-located with 1,4-
dioxane release sites, EPA relied primarily on modeling to estimate water concentrations that could
result from releases. Where co-located monitoring data were available, EPA compared modeled
concentrations to reasonably available monitoring data in the limited set of specific locations to evaluate
the performance of the model. Comparisons of modeled vs. monitoring water concentrations for this
limited set of "case study" locations demonstrate that modeled mean concentrations are generally
consistent with mean concentrations reported in monitoring data. For example, the Cape Fear River
upstream of the Brunswick County, NC drinking water intake was selected as a case study to test the
model due to abundant monitoring data in the region. Water concentrations modeled based on upstream
releases from an industrial facility in Fayetteville in combination with other upstream sources. As
illustrated in Figure 2-14, modeled surface water concentrations generally fell within the ranges reported
from monitored concentrations. Wide ranges of both monitored and modeled values were noted,
indicating variability among inputs to the system. Details of the case study comparisons for Brunswick
County and other locations are described in Appendix G.2.3.2. The concordance of monitoring and
modeled concentrations increases confidence in the model used to estimate water concentrations from
DTD releases and hydraulic fracturing, and to perform probabilistic modeling of aggregate
concentrations from multiple sources.

Page 71 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Date

Figure 2-14. Case Study Comparison of Modeled and Monitored Concentrations in
Brunswick County

2.3.1.5 Strengths, Limitations, and Sources of Uncertainty in Assessment Results for
Monitored and Modeled Drinking Water and Surface Water Concentrations

The evaluation of general population drinking water exposure scenarios are impacted by uncertainties
and assumptions surrounding inputs and the approaches used for modeling surface water concentrations
and estimation of the drinking water doses. In Appendix E.8, EPA assesses the overall confidence of
estimated releases for various OESs. For those OESs releasing to surface water, confidence is rated as
medium to high depending on an individual OES.

The modeling used and the associated default and user-selected inputs have the ability to affect overall
strength in evaluated general population exposures. The facility-specific releases methodology described
in Section 2.3.1.2.1, and the results in 2.3.1.3.1, rely on a modeling framework that does not consider
downstream fate or transport. However, the physical-chemical properties of 1,4-dioxane are expected to
moderate this limitation due to its likelihood to stay in the water column, and due to the lack of removal
during typical drinking water treatment process. To reduce uncertainties, EPA incorporated an updated
flow network and flow data into this assessment that allowed a more site-specific consideration of
release location and associated receiving water body flows. These facility-specific releases are also
evaluated on a per facility basis that does not account for additional sources of 1,4-dioxane that may be
present in the evaluated waterways. To help address these limitations in this risk evaluation, EPA
conducted additional aggregate and probabilistic approaches, evaluated in Section 2.3.1.2.1 and Section
2.3.1.3.4, that give a more complete overall estimation of possible 1,4-dioxane concentrations. Finally,
drinking water exposures from facility-specific results assume that the exposure occurs at the receiving
water body. The water bodies evaluated may or may not be used as drinking water sources. To address
this limitation, EPA evaluated the proximity of known 1,4-dioxane releases to known drinking water
sources as well as known drinking water intakes as described in Section 2.3.1.2.4.

To evaluate the accuracy of the aggregate model, case studies described in Appendix G.2.3.2 compared
modeled results to observed monitored concentrations. The three evaluated case studies give good
general agreement between available monitoring with modeled values. Overall, this gives strength to the
modeling assumptions, inputs and output calculations for areas that are lacking robust monitoring data.
The model is able to effectively capture the general influences of both DTD loading, facility loading and

Page 72 of 484


-------
2271

2272

2273

2274

2275

2276

2277

2278

2279

2280

2281

2282

2283

2284

2285

2286

2287

2288

2289

2290

2291

2292

2293

2294

2295

2296

2297

2298

2299

2300

2301

2302

2303

2304

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

upstream contributions to create an aggregation of possible ambient surface water concentrations of 1,4-
dioxane. The monitored data encompassed both ambient surface water monitoring as well as drinking
water system monitoring data. For the ambient surface water data, data is limited geographically and
temporally with many states having no reported data and even those areas reporting measured values
having limited samples over time. Monitored concentrations in close proximity to modeled releases were
rare, often making direct comparisons of modeled results unavailable. In most cases, monitoring data
represented water bodies without identified releases of 1,4-dioxane nearby.

The hydraulic fracturing analysis relies on a Monte Carlo distribution of loading values with some level
of uncertainty and is itself a Monte Carlo simulation with potential receiving water body flows. The
precision of such an analysis is lower at the most extreme (minimum and maximum) values.

2.3.2 Land Pathway (Groundwater)

Any activities where chemicals or wastes might be released to the environment has the potential to
pollute groundwater. To understand possible exposure scenarios from these practices, EPA assessed
drinking water exposure resulting from use of 1,4-dioxane contaminated groundwater due to chemical
injection to Underground Class I Wells, leaching from landfills where 1,4-dioxane or products
containing 1,4-dioxane have been disposed, and disposal of hydraulic fracturing produced water to
surface impoundments. Sections 2.3.2.1 through 2.3.2.4 provide a description and an assessment of each
disposal practice. Figure 2-15 and Figure 2-16 provide a visual summary of groundwater monitoring
data available through the WQP fNWOMC. 20221

2.3.2.1 Groundwater Monitoring Data

Measured, field-collected, data from environmental samples representing groundwater 1,4-dioxane
concentrations across the country were collected as direct groundwater monitoring results. These results
are collated by the National Water Quality Monitoring Council and stored in the WQP fNWOMC.
2022). Some monitoring results reported to the WQP included locations expected to be directly
impacted by 1,4-dioxane releases. Data were available from 1997 to 2022, resulting in 8,110 available
sample results. The distribution and detection percentages are presented in Figure 2-15 and mapped in
Figure 2-16. The process for identifying this data is provided in Appendix H.l. This analysis is intended
to characterize the observed ranges of 1,4-dioxane concentrations in groundwater, irrespective of the
reasons for sample collection, and to provide context for the modeled groundwater concentrations
presented in Sections 2.3.2.1 through 2.3.2.4. In order to better understand where highest groundwater
concentrations are occuring, EPA arbitrarily portioned the data based on order of magnitude differences
to best describe where and when data differences could be observed.

Page 73 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Detected 1,4-Dioxane Concentration (|jg/L)

2306	Figure 2-15. Frequency of Nationwide Detected 1,4-Dioxane Groundwater

2307	Concentrations (n = 2,284) Retrieved from the Water Quality Portal, 1997-2022

2308

2309

2310	Figure 2-16. Detectable Concentrations of 1,4-Dioxane in Groundwater from the Water

2311	Quality Portal, 1997-2022

2312	Note: Alaska, American Samoa, Guam, Hawaii, N. Mariana Islands, Puerto Rico, and the U.S. Virgin Islands

2313	are not shown as there are no known monitoring data above detection limits.

2314

Page 74 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Figure 2-15 shows the range of detected concentrations of 1,4-dioxane in groundwater from 1997 to
2022. During this period the detection limits ranged from 0.028 |ig/L to 320 |ig/L. The maximum
detected concentration (31,000 |ig/L) occurred in Westville, IN, in 1997 at a former waste-oil refinery.
This site and many others identified in this monitoring data have ongoing remediation projects to
address these contamination plumes.

Recent changes in industrial activities and disposal may have largely reduced groundwater
contamination with 1,4-dioxane. As shown in Figure 2-17, samples collected prior to 2000 tended to be
substantially higher in concentration relative to those collected after 2003. This finding may be an
artifact of historical uses and industrial practices related to 1,4-dioxane. Although several samples are
still above 10 |ig/L, particularly in 2007, the bulk of data tend to fall between 1 and 10 |ig/L. Without a
thorough investigation of what practices have changed in industry, it is difficult to attribute this decline
to a single event but indicates continued work to prevent groundwater contamination.

Figure 2-17. Groundwater Concentrations of 1,4-Dioxane vs. Sample
Collection Date for Data Collected between 1997 and 2022

Figure 2-16 shows the spatial distribution of detected 1,4-dioxane concentration across the contiguous
states. This map shows nine locations with concentrations of 1,4-dioxane greater than 10 |ig/L. These
tend to be attributed to past industrial activities causing extensive groundwater contamination. In
addition to this monitoring data, groundwater contamination from disposing 1,4-dioxane to landfills has
been documented in Alaska (Li et al.. 2013). California (Li et al.. 2015; Adamson et al.. 20141
Michigan (Mohr and DiGuiseppi. 20101 New York (Lee et al.. 2020). and recently in Ohio
(https://cumulis.epa.gov/supercpad/cursites/csitinfo.cfm?id=0504014). EPA was not able to identify
reasonably available information specific to groundwater concentrations near or around underground
injection sites, landfills, or surface impoundments that received hydraulic fracturing produced water.

2.3.2.2 Disposal via Underground Injection	

Underground injection is a method of disposal for hazardous wastes.9 There are generally six different
classes of underground wells, and only Class I Wells may be permitted to receive hazardous waste.
Oversight of these wells requires that they are designed and constructed to prevent the movement of

9 Additional information about underground injection can be found at https://www.epa.gov/uic.

Page 75 of 484


-------
2346

2347

2348

2349

2350

2351

2352

2353

2354

2355

2356

2357

2358

2359

2360

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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

injected waste streams into drinking water systems. Wells typically consist of three or more concentric
layers of pipe including surface casing, long string casing, and injection tubing. In addition, wells must
be sited at locations with geologies that mitigate any movement of contaminants outside of a confined
layer in case of a well failure. Extensive pre-siting geological tests confirm that the injection zone is of
sufficient lateral extent and thickness and is sufficiently porous so that fluids injected through the well
can enter the rock formation without extensive buildup of pressure or possible displacement of injected
fluids outside of the intended zone.

Potential pathways through which injected fluids can migrate to underground sources of drinking water
include failure of the well or improperly plugged or completed wells near the well. Well failures can be
detected by continuous monitoring systems or mechanical integrity tests, at which point the wells would
be shut-in until they are repaired. EPA's extensive technical requirements for Class I wells (40 CFR
148) are designed to prevent contamination of underground sources of drinking water through these
pathways. Operators must conduct appropriate mechanical integrity tests yearly for hazardous wells and
every 5 years for nonhazardous wells to ensure wells are fit for operation. Note that the loss or failure of
mechanical integrity does not necessarily mean that wastewater will escape the injection zone. This
added security can be attributed to redundant safety systems to protect against loss of waste
confinement.

2.3.2.2.1	Summary of Assessment for Disposal to Underground Injection

According to EPA's TRI database, there are two locations where 1,4-dioxane has been disposed of via
underground injection to Class I Wells. On-site disposals to Class I underground injection wells are
provided in TableApx H-l. On-site Class I underground injection wells may be owned and operated by
the producer of the waste. Off-site disposals to Class I underground injection wells are provided in
Table Apx H-2. Offsite Class I underground injection wells may be secondary entities that own and
operate the well. Both on-site and offsite underground injection wells must be permitted and regularly
inspected. Careful review of the permits and state databases corroborates that both sites are permitted
and compliant. These sites have implemented groundwater migration controls and the Enforcement and
Compliance History Online (ECHO) database (	!022e) indicates the site is currently in

compliance.

In addition to reviewing these permits, EPA reviewed reasonably available groundwater monitoring data
available via state databases as well as via the WQP (see Figure 2-15 and Figure 2-16) and found no
evidence of groundwater contamination near the facilities. Because underground injection is not
expected to result in groundwater contamination based on the reasonably available information, EPA did
not quantitatively estimate groundwater concentrations, exposures, or risks from underground injection.

2.3.2.2.2	Strengths, Limitations, and Sources of Uncertainty in Assessment of
Disposal to Underground Injection Wells

Because EPA did not quantitatively evaluate the potential exposure from disposing 1,4-dioxane via
underground injection, the major source of uncertainty is limited to the accuracy of state databases
providing monitoring data surrounding these wells. EPA believes these databases are reporting
accurately where contaminations are known, but only explored states where the TRI database indicated
there were disposals via underground injection. Disposals below the reporting requirement for TRI may
not be captured.

2.3.2.3 Disposal to Landfills

Landfills may have various levels of engineering controls to prevent groundwater contamination. These
can include industrial liners, leachate capturing systems, and routine integration of waste. However,

Page 76 of 484


-------
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

2432

2433

2434

2435

2436

2437

2438

2439

2440

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

groundwater contamination from disposal of consumer, commercial, and industrial waste streams
continues to be a prominent issue for many landfills throughout the United States (Li et al.. JO I < t
at.. 2013; Mohr and DiGuiseppi. 2010). These contaminations may be attributed to perforations in the
liners, failure of the leachate capturing system, or improper management of the landfills. 1,4-Dioxane
persists in groundwater and can migrate away from landfills into nearby communities at the same rate as
hydraulic flow (Mohr and DiGuiseppi. 2010). When these communities rely on groundwater as their
primary drinking water source, there is a potential for exposure via oral ingestion if that water is
contaminated with 1,4-dioxane and does not undergo treatment. Depending on the distance between the
landfill and a drinking water well, as well as the potential rate of release of landfill leachate into
groundwater, the concentration of this exposure can vary substantially.

Landfills are regulated under the Resource Conservation and Recovery Act (RCRA) or TSCA. RCRA
landfills can be classified as Subtitle C (hazardous waste landfills) or Subtitle D (municipal solid
nonhazardous waste landfills). Subtitle C establishes a federal program to manage hazardous wastes
from cradle to grave. The objective of the Subtitle C program is to ensure that hazardous waste is
handled in a manner that protects human health and the environment. When waste generators produce
greater than 100 kg per month of non-acutely hazardous waste, those hazardous wastes, including 1,4-
dioxane, meeting the U108 waste code description in 40 CFR 261.33, must be treated to meet the land
disposal restriction levels in 40 CFR part 268 and be disposed in RCRA subtitle C landfills. These
disposals are captured partially through the Toxics Release Inventory and are reported for onsite
facilities (TableApx H-3) and offsite facilities (TableApx H-4). Recent violations of permits are
reported in the footnotes of each table.

Review of state databases does not suggest any readily available evidence of groundwater contamination
near or coinciding with these operations that could affect a drinking water supply. Similar review of the
data available via the WQP suggests that there are no known contaminations from RCRA Subtitle C
Landfills as reported to the TRI program (see Figure 2 14 and Figure 2 15). The absence of groundwater
contamination near RCRA Subtitle C Landfills may be attributed to many of the ongoing engineering
controls built into these facilities as well as active monitoring of groundwater wells around facilities. As
a result, EPA did not assess Subtitle C landfills further than understanding their permit violations.

Regulations established under Subtitle D ban open dumping of waste and set minimum federal criteria
for the operation of municipal waste and industrial waste landfills, including design criteria, location
restrictions, financial assurance, corrective action (clean up), and closure requirements. States play a
lead role in implementing these regulations and may set more stringent requirements. National
requirements for Subtitle D landfills are most specific for MSW landfills. MSW landfills built after 1990
must be constructed with composite liner systems and leachate collection systems in place. Composite
landfill liners consist of a minimum of two feet of compacted soil covered by a flexible membrane liner,
which work in concert to create a low hydraulic conductivity barrier and prevent leachate from being
released from the landfill and infiltrating to groundwater. A leachate collection system typically consists
of a layer of higher conductivity material above the composite liner that funnels leachate to centralized
collection points where it is removed from the landfill for treatment and disposal. Despite these controls,
releases may still occur due to imperfections introduced during construction or that form over time (Li et
al., 2015; Li et al., 2013; Mohr and DiGuiseppi, 2010); thus, groundwater monitoring is required to
identify and address any releases before there can be harm to human health and the environment. RCRA
Subtitle D requirements for non-MSW landfills are less stringent. In particular, nonhazardous industrial
landfills and C&D debris landfills do not have specified national requirements for construction and
operation and certain landfills are entirely exempt from RCRA criteria. Under the Land Disposal
Program Flexibility Act of 1996 (Pub.L. 104-119), some villages in Alaska that dispose of less than 20

Page 77 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

tons of municipal solid waste daily (based on an annual average) may dispose of waste in unlined or
clay-lined landfills or waste piles for open burning or incineration.

There are a several potential sources of 1,4-dioxane to Subtitle D landfills. Waste generators that
produce less than 100 kg per month of non-acutely hazardous waste, including 1,4-dioxane meeting the
U108 waste code, may dispose of this waste in these landfills. Nonhazardous industrial wastes also have
the potential to contain 1,4-dioxane at variable concentrations. Consumer and commercial products may
also contain 1,4-dioxane in relatively low amounts. The greatest potential for release of disposed 1,4-
dioxane to groundwater is from landfills that do not have an adequate liner system. Thus, an objective of
this assessment is to evaluate the potential for groundwater contamination in the absence of landfill
controls.

This assessment was completed using the Hazardous Waste Delisting Risk Assessment Software
(DRAS). DRAS was specifically designed to address the Criteria for Listing Hazardous Waste identified
in Title 40 Code of Federal Regulations (40 CFR) Section 261.11(a)(3), a requirement for evaluating
proposed hazardous waste delistings. In this assessment, DRAS is being utilized to determine potential
groundwater concentrations of 1,4-dioxane after they have been disposed of into a non-hazardous waste
landfill. The results of this assessment are found in Table 2-13. This assessment relied on the default
waste loading rates for RCRA Subtitle C Landfills available in DRAS. Similarly, the assessment relied
on the default values for 1,4-dioxane as the chemical of concern. Lastly, leachate concentrations were
estimated for a range of possibilities until no risk could be identified at the lower end of those
concentrations. Because DRAS calculates a weight adjusted dilution attenuation factor (DAF) rather
than a groundwater concentration, a back of the envelop computation was used to convert the DAF to a
potential concentration that people living within 1 mile of a landfill might be exposed if the release were
not identified and remediated.

Page 78 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2466	Table 2-13. Potential Groundwater Concentrations (jig/L) of 1,4-Dioxane Found in Wells within 1 Mile of a Disposal Facility

2467	Determined by Using the DRAS Model	

Leachate Concentration
(jig/L)

Loading Rate (kg)

4.55 E-04

4.55E-03

4.55E-02

4.55E-01

4.55E00

4.55 E01

4.55E02

4.55E03

4.55E04

4.55E05

1.00E-07

7.81E-13

7.46E-12

5.46E-11

5.21E-10

6.49E-09

6.17E-08

5.88E-07

5.62E-06

5.38E-05

5.13E-04

1.00E-06

7.81E-12

7.46E-11

5.46E-10

5.21E-09

6.49E-08

6.17E-07

5.88E-06

5.62E-05

5.38E-04

5.13E-03

1.00E-05

7.81E-11

7.46E-10

5.46E-09

5.21E-08

6.49E-07

6.17E-06

5.88E-05

5.62E-04

5.38E-03

5.13E-02

1.00E-04

7.81E-10

7.46E-09

5.46E-08

5.21E-07

6.49E-06

6.17E-05

5.88E-04

5.62E-03

5.38E-02

5.13E-01

1.00E-03

7.81E-09

7.46E-08

5.46E-07

5.21E-06

6.49E-05

6.17E-04

5.88E-03

5.62E-02

5.38E-01

5.13E00

1.00E-02

7.81E-08

7.46E-07

5.46E-06

5.21E-05

6.49E-04

6.17E-03

5.88E-02

5.62E-01

5.38E00

5.13E01

1.00E-01

7.81E-07

7.46E-06

5.46E-05

5.21E-04

6.49E-03

6.17E-02

5.88E-01

5.62E00

5.38E01

5.13E02

1.00E00

7.81E-06

7.46E-05

5.46E-04

5.21E-03

6.49E-02

6.17E-01

5.88E00

5.62E01

5.38E02

5.13E03

1.00E01

7.81E-05

7.46E-04

5.46E-03

5.21E-02

6.49E-01

6.17E00

5.88E01

5.62E02

5.38E03

5.13E04

Concentrations organized by potential loading rates (kg) and potential leachate concentrations (ng /L).

2468

Page 79 of 484


-------
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

2502

2503

2504

2505

2506

2507

2508

2509

2510

2511

2512

2513

2514

2515

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2.3.2.3.1	Summary of Assessment for Disposal to Landfills

EPA determined through modeling that groundwater concentration of 1,4-dioxane increased with
increasing landfill load rate and increasing leachate concentration. With each progressive iteration of
loading rate or leachate concentration, potential groundwater concentrations increase by an order of
magnitude. When both loading rate and leachate increase by one order of magnitude, potential
groundwater concentration increase by two orders of magnitude. These increases can largely be
attributed to the increasing weight adjusted dilution attenuation factor and are what would be expected
for a chemical substances with 1,4-dioxane's physical-chemical properties (water solubility, Henry's
law constant) and fate characteristics (biodegradability, half-life in groundwater). 1,4-Dioxane migrates
in groundwater at the rate of hydraulic flow and can persist for greater than 30 days in anaerobic
environments (Adamson et al.. 2014; Mohr and DiGuiseppi. 2010) as described in the 2020 RE. Thus,
these concentrations are likely to represent the range of potential groundwater concentrations for
sentinel PESS living within a 1-mile radius of a RCRA Subtitle D landfills and other non-Subtitle C
landfills.

EPA also determined that the modeled concentrations are within the range of concentrations of 1,4-
dioxane found in groundwater monitoring studies. A survey of monitoring studies in California has
demonstrated that 1,4-dioxane concentrations in groundwater can range from 9 |ig/L at 10th percentile
to 13,460 |ig/L at the 90th percentile (Adamson et al.. 2014). Monitoring data from EPA's Third
Unregulated Contaminant Monitoring Rule (UCMR3) reported 1,4-dioxane concentrations in
groundwater ranging from 0.07 |ig/L to 34 |ig/L (Adamson et al. 201 ; 1 c. « i1 \ J017d). Though
many of the corresponding sites in these monitoring surveys are not specifically tied to the disposal of
1,4-dioxane to landfills, they provide context for what concentrations may be expected when
contamination occurs. When focusing on groundwater concentrations of 1,4-dioxane surrounding
landfills based on reasonably available information, EPA found concentrations of 1,4-dioxane ranging
from 6.4 to 25 mg/L (Cordone et al.. 2016). Leaching from unlined lagoons in Michigan resulted in
groundwater concentrations highs ranging from 1,000 to 20,000 |ig/L (Jackson ai ke. 2019; Mohr
and DiGuiseppi. 2010); four decades later concentrations are now reaching 2 mg/L or less after active
treatment and natural attenuation. Mean concentrations of 1,4-dioxane in landfill leachate in the United
States has ranged from 11.8 |ig/L for municipal landfills to 44.6 |ig/L for hazardous waste landfills (as
described in (Mohr and DiGuiseppi. 2010)). These concentrations further support that the modeled
concentrations are within the range of those reported in the literature.

2.3.2.3.2	Strengths, Limitations, and Sources of Uncertainty in Assessment Results
for Disposal to Landfills

Uncertainties and limitations are inherent in the modeling of groundwater concentrations from disposing
chemical substances into RCRA Subtitle D landfills and other non-Subtitle C landfills. These
uncertainties include, but are not limited to, determining the total and leachable concentrations of waste
constituents; estimating the release of pollutants from the waste management units to the environment;
and, estimating transport of pollutants in a range of variable environments by process that often are not
completely understood or are too complex to quantify accurately. To address some of these uncertainties
and add strength to the assessment, EPA considered multiple loading rates and multiple leachate
concentrations. These considerations add value to estimate exposure that falls at an unknown percentile
of the full distribution of exposures.

A strength of the assessment is that the modeled data are within the range of monitoring data that have
been evaluated at both the national scale (Adamson et a! JO I ; 1, c. « ^ \ :0l 
-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

represent current conditions of waste management units in the United States. Both the DRAS model and
EPACMTP are based on a survey of drinking water wells located downgradient from a waste
management unit (	|). Due to the age of the survey, it is unclear how the survey represents

current conditions and proximity of drinking water wells to disposal units. Similarly, it is not clear if the
surveyed waste management units are representative of current waste management practices.

2.3.2.4 Disposal of Hydraulic Fracturing Produced Water to Surface Impoundments

After hydraulic fracturing operations inject fluids to extract oil and gas, a substantial volume of water
may be produced through flowback. Otherwise known as produced waters, the composition of this water
depends both on the geochemistry of the injected area and the injected fluids (	). 1,4-

Dioxane has been reported to EPA as one of the chemicals present in produced waters by 411 facilities
via FracFocus 3.0 (GWPC and IOGCC. 2022) (Table 2-14). A variety of options exist for these
produced waters after use in hydraulic fracturing operations ranging from underground injection,
treatment and subsequent use, treatment and discharge, or evaporation in surface impoundments. Each
of these options are subject to state and federal regulations (	). When produced waters

are released to unlined surface impoundments, there is potential for groundwater contamination and
subsequent human exposure via drinking water. Thus, EPA conducted an assessment to determine the
range of groundwater concentrations within a 1-mile radius of surface impoundments receiving
produced water from hydraulic fracturing operations.

Based on the results of Monte Carlo analysis presented in Table Apx E-5, disposal to these surface
impoundments could account for up to 3 percent of all produced waters. 1,4-Dioxane has been
documented to have a concentration of 60 |ig/L in these produced waters (Lester et at.. 2015). Thus,
EPA assessed the potential for disposing of hydraulic fracturing produced water at the 5th, 50th, 95th,
and 99th percentiles as well as at the min, mean, and max to a managed surface impoundment assuming
these loading rates and concentration using DRAS. The results are presented in Table 2-14.

2.3.2.4.1 Summary of Assessment for Disposal of Hydraulic Fracturing Produced
Water

In general, EPA determined that groundwater concentrations of 1,4-dioxane would increase as more
produced water was released to surface impoundments. The values presented in Table 2-14 represent the
maximum 3 3-year receptor well concentration within a 1-mile radius of a hypothetical surface
impoundment that leaches into groundwater. With each progressive iteration of summary statistic for
loading rate, potential groundwater concentrations increase accordingly. This increase can be attributed
to the decrease in the weight-adjusted dilution attenuation factor. As the mass of 1,4-dioxane entering an
aquifer increases, it is less diluted and higher concentrations will be found downgradient. Due to its
physical-chemical properties (e.g., water solubility, Henry's law constant) and fate characteristics (e.g.,
biodegradability, half-life in groundwater), 1,4-dioxane migrates in groundwater and can persist for
greater than 30 days in anaerobic environments (Adamson et at.. 2014; Mohr and DiGuiseppi. 2010) as
described in the 2020 RE. Thus, these concentrations are likely to represent the range of potential
groundwater concentrations for people living within a 1-mile radius of a surface impoundment,
representing a sentinel PESS exposure.

Page 81 of 484


-------
2557

2558

2559

2560

2561

2562

2563

2564

2565

2566

2567

2568

2569

2570

2571

2572

2573

2574

2575

2576

2577

2578

2579

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Table 2-14. Total Annua

Release Summary

Total Annual Release
(kg/site-year)

Monte Carlo
Analysis Summary
Statistic

Weight Adjusted
Dilution Attenuation
Factor

Potential
Groundwater
Concentration (jig/L)

1.68E00

Max

3.18

1.89E-05

01.87E-01

99th Percentile

3.91

1.54E-05

6.52E-02

95th Percentile

3.91

1.54E-05

1.47E-02

Mean

84

7.10E-07

3.83E-03

50th Percentile

495

1.20E-07

3.24E-05

5th Percentile

495

1.20E-07

1.06E-11

Min

135,000

0.00E00

2.3.2.4.2 Strengths, Limitations, and Sources of Uncertainty in Assessment Results
for Disposal from Hydraulic Fracturing Operations

Although it is well understood that 1,4-dioxane is present in produced waters from hydraulic fracturing
as reported in FracFocus (GWPC and IOGCC. 2022). the number of studies reporting the concentration
of the chemical substances in produced waters is limited (Lester et ai. 2015). FracFocus is generally
considered a moderately reliable source of information as it is based on data from thousands of fracking
wells across the United States. Further, both the release assessment (as discussed in Section 2.2.1.2; see
also Table Apx E-7) and the groundwater concentration assessment (Table 2-14) are modeled using a
Monte Carlo simulation. These conditions lower the confidence in the overall assessment.

2.3.3 Ambient Air Pathway	

EPA developed and applied tiered methodologies and analyses to estimate ambient air concentrations
and exposures to members of the general population. These methodologies and analyses focus on
inhalation exposures to a sub-set of the general population referred to as fenceline communities.
Fenceline communities are defined as a subset of the general population that are in proximity to air
emitting facilities or a receiving water body, and who therefore may be disproportionately exposed to a
chemical undergoing risk evaluation under TSCA section 6(b). For the air pathway, proximity goes out
to 10,000 meters from an air emitting source. The methodology and analyses were first presented in the
Draft TSCA Screem	pproach for Assessing Ambient Air and Water Exposures to Fenceline

Communities referred to here as the 2022 Fenceline Report.1" The specific methodologies used in this
assessment to evaluate general population exposures to 1,4-dioxane in air are briefly described in Figure
2-18. Additional details on the methodologies are provided in Appendix J.

10 The 2022 Fenceline Report is available at https://www.epa.gov/assessing~and~managing~chemieals~iinder~tsea/tsea~

screening~level~approach~assessing~ambient~air~and.

Page 82 of 484


-------
2580

2581

2582

2583

2584

2585

2586

2587

2588

2589

2590

2591

2592

2593

2594

2595

2596

2597

2598

2599

2600

2601

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Figure 2-18. Brief Description of Methodologies and Analyses Used to Estimate Ambient
Air Concentrations and Exposures

EPA used the air release estimates obtained using the methodology described in Section 2.1.1.3 as direct
inputs for the models used to estimate exposure concentrations at various distances from a releasing
facility. EPA expanded upon the methods described in the 2022 Fenceline Report in response to SACC
comments/recommendations by evaluating potential aggregate concentrations from multiple facilities.

2.3.3.1	Measured Concentrations in Air

EPA did not identify quantitative outdoor air monitoring data for 1,4-dioxane.

2.3.3.2	Modeled Concentrations in Air

Because there is no air monitoring data for 1,4-dioxane, the Agency relied upon modeling to estimate
exposure concentrations to fenceline communities at various distances from a releasing facility.
Modeling was used for each analysis described in Figure 2-18 for 1,4-dioxane. For scenarios where the
screening methodology indicated a need for further analysis, EPA performed a full analysis using the
American Meteorological Society/Environmental Protection Agency Regulatory Model (AERMOD)
and/or EOA's Integrated Indoor/Outdoor Air Calculator (IIOAC).11 IIOAC analysis was performed for
three COUs where no site specific data were available (Hydraulic Fracturing, Industrial Laundry
Facilities, Institutional Laundry Facilities) and is briefly described in Section 2.3.3.2.4 with results
presented and discussed in Sections 5.2.2.3.2 and 5.2.2.3.3. An expanded analysis to consider aggregate
exposures was performed for 1,4-dioxane in response to SACC comments/recommendations on the
2022 Fenceline Report.

11 The IIOAC website is available at https://www.epa.gov/tsca-screening-tools/iioac-integrated-indoor-ontdoor-air-calailator.

Page 83 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2.3.3.2.1	Ambient Air: Screening Methodology

The Ambient Air: Screening Methodology utilizes EPA's IIOAC model to estimate high-end and central
tendency (mean) 1,4-dioxane exposure concentrations in ambient air at three distances from an emitting
facility: 100, 100 to 1,000, and 1,000 m. EPA developed and evaluated a range of exposure scenarios for
each of two categorical release amounts12 designed to capture a variety of release types, topography,
meteorological conditions, and release scenarios. A diagram of these exposure scenarios is provided in
Appendix J. Findings from the Ambient Air: Screening Methodology were used to inform the need for a
higher-tier analysis as well as provide insight into whether risk estimates above the benchmarks are or
are not expected for 1,4-dioxane.

The Ambient Air: Screening Methodology design inherently includes both estimates of exposures as
well as estimates of risks to inform the need, or potential need, for further analysis. If findings from the
Ambient Air: Screening Methodology estimate risk (acute non-cancer, chronic non-cancer, or cancer)
for a given chemical above (or below as applicable) typical Agency benchmarks, EPA generally will
conduct a higher-tier analysis of exposures and associated risks for that chemical. If findings from the
Ambient Air: Screening Methodology estimate risks that do not exceed (or fall below as applicable)
benchmarks, EPA may still conduct a limited higher-tier analysis at distances very near a releasing
facility (less than 100 m) to ensure potential risks are not missed.

A more detailed description of the Ambient Air: Screening Methodology for 1,4-dioxane is provided in
Appendix J, along with summarized results. In general, for 1,4-dioxane, the results of this analysis
identified risk estimates above screening benchmarks for cancer at multiple distances and for multiple
releases (max and mean). In accordance with the tiered methodology presented to the SACC in the 2022
Fenceline Report, because risk estimates exceeded the benchmark for cancer based on the Ambient Air:
Screening Methodology, EPA conducted multiple higher-tier analyses of all facilities reporting releases
of 1,4-dioxane to TRI.

2.3.3.2.2	Ambient Air: Single Year Methodology (AERMOD)

The Ambient Air: Single Year Methodology (AERMOD) utilizes the EPA's American Meteorological
Society/Environmental Protection Agency Regulatory Model (AERMOD)13 to estimate 1,4-dioxane
concentrations in ambient air at eight finite distances (5, 10, 30, 60, 100, 2,500, 5,000, and 10,000 m)
and one area distance from an emitting facility.14 EPA modeled two different types of release estimates,
as applicable, for 1,4-dioxane: (1) facility-specific chemical releases with source attribution when 2019
TRI data was available, and (2) alternative release estimates representing a generic facility when 2019
TRI data was not available for an OES. Daily and period average outputs were obtained via modeling,
and post-processing scripts were used to extract a variety of statistics from the modeled concentration
distribution, including the 95th (high-end), 50th (central tendency), and 10th (low-end) percentile 1,4-
dioxane concentrations at each distance modeled.

A summary of the concentration ranges estimated using the Ambient Air: Single Year Methodology
(AERMOD) is provided in Table 2-15. The summary includes 11 OESs and select statistics (maximum,

12	The pre-screening methodology from the 2022 fenceline analysis evaluated two categorical release values across all
facilities reporting releases to the 2019 TRI. The first is the maximum single facility release reported across all facilities; the
second is the mean (arithmetic average) of all releases reported across all facilities reporting.

13	See https://www.epa.gOv/scram/air-qnalitY-dispersion-modeling-preferred-and-recotn.mended-models#aermod for more
information.

14	For the one "area distance" evaluated, receptors are placed in a cartesian grid between approximately 200 and 900 m, at
100 m spacing. This results in a total of 456 receptors. The exposure estimates for the area distance represent the arithmetic
average (mean) exposure concentration across all 456 receptors within the "area distance" for each day.

Page 84 of 484


-------
2642

2643

2644

2645

2646

2647

2648

2649

2650

2651

2652

2653

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

mean, median, and minimum) calculated from the modeled concentration distributions within each OES
at each distance modeled. The associated range of estimated concentrations is based on the maximum
95th percentile annual average exposure concentrations for each distance. Although the range of
concentrations are provided, there are many instances where the range extends as many as 12 orders of
magnitude from minimum to maximum concentration. This occurs because within each OES there are
several individual facilities evaluated and, in most cases, the reported release values from each
individual facility can vary widely (from 500 lb to several hundred thousand pounds), which in turn
affects the range of estimated exposure concentrations at a given distance. Therefore, in trying to
summarize the wide variety of releases into a single range, the variation in estimated concentrations will
also appear extensive. This is not indicative of an inadequate analysis or methodology, but solely based
on the variability of releases across facilities within a given OES.

Page 85 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2654	Table 2-15. Summary of Select Statistics for the 95th Percentile Estimated Annual Average Concentrations from the "Full-

2655	Screening" Analysis for 1,4-Dioxane Releases Reported to TRI	

OES

# Facilities
Evaluated
in OES

Statistic

Annual Average Concentration (ppm) Estimated within 5 to 10,000 m of Releasing Facilities

5

10

30

60

100

100 to

1,000

2,500

5,000

10,000

Disposal

15

Max

4.26E-03

5.05E-03

1.80E-03

6.90E-04

3.15E-04

2.95E-05

1.89E-06

6.28E-07

2.09E-07

Mean

4.76E-04

6.92E-04

2.79E-04

1.12E-04

5.39E-05

6.00E-06

346E-07

1.18E-07

4.12E-08

Median

8.44E-06

1.65E-05

9.35E-06

8.70E-06

5.81E-06

7.64E-07

4.53E-08

1.48E-08

4.81E-09

Min

3.31E-15

9.85E-14

5.17E-11

9.72E-10

2.03E-09

1.10E-09

1.21E-10

3.81E-11

1.22E-11

Dry Film
Lubricant

8

Max

1.61E-10

7.14E-09

5.10E-07

3.88E-06

6.29E-06

9.92E-07

2.79E-08

8.44E-09

3.68E-09

Mean

2.06E-11

9.46E-10

1.90E-07

2.28E-06

4.05E-06

8.14E-07

1.95E-08

5.94E-09

2.45E-09

Median

2.46E-13

3.58E-11

1.59E-07

2.21E-06

4.00E-06

7.75E-07

1.88E-08

6.02E-09

2.66E-09

Min

4.05E-18

2.19E-13

5.64E-08

9.23E-07

2.39E-06

7.39E-07

1.36E-08

4.02E-09

1.40E-09

Ethoxylation
Byproduct

6

Max

6.53E-03

1.36E-02

7.33E-03

3.09E-03

1.64E-03

3.81E-04

2.20E-05

9.00E-06

3.45E-06

Mean

1.74E-03

3.05E-03

149E-03

6.18E-04

3.23E-04

7.03E-05

4.02E-06

1.62E-06

6.13E-07

Median

2.44E-04

4.40E-04

2.08E-04

8.70E-05

440E-05

6.36E-06

3.09E-07

1.03E-07

3.40E-08

Min

4.08E-14

6.32E-13

4.29E-10

5.22E-09

1.15E-08

4.99E-09

7.17E-10

3.39E-10

1.40E-10

Film Cement

1

Max

1.25E-04

1.31E-04

441E-05

2.28E-05

1.25E-05

2.29E-06

1.38E-07

4.60E-08

1.52E-08

Mean

3.90E-05

4.87E-05

2.04E-05

9.36E-06

4.82E-06

7.95E-07

444E-08

1.48E-08

4.87E-09

Median

2.02E-05

2.93E-05

1.74E-05

7.33E-06

3.57E-06

5.68E-07

248E-08

8.10E-09

2.65E-09

Min

3.17E-06

6.36E-06

4.97E-06

1.95E-06

8.99E-07

1.32E-07

444E-09

1.42E-09

4.61E-10

Functional
Fluids (Open-
System)

2

Max

1.28E-05

2.36E-05

1.03E-05

1.08E-05

1.82E-05

7.42E-06

6.78E-07

2.47E-07

8.81E-08

Mean

6.40E-06

1.18E-05

5.74E-06

7.71E-06

1.08E-05

4.24E-06

3.88E-07

1.45E-07

5.28E-08

Median

6.40E-06

1.18E-05

5.74E-06

7.71E-06

1.08E-05

4.24E-06

3.88E-07

1.45E-07

5.28E-08

Min

1.66E-11

1.93E-10

1.18E-06

4.61E-06

3.37E-06

1.06E-06

9.70E-08

4.30E-08

1.74E-08

Import and
Repackaging

1

Single Facility

2.70E-11

5.57E-10

5.52E-08

4.17E-07

8.70E-07

3.21E-07

6.72E-08

4.12E-08

2.23E-08

Page 86 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

# Facilities
Evaluated
in OES

Statistic

Annual Average Concentration (ppm) Estimated within 5 to 10,(MM) m of Releasing Facilities

5

10

30

60

100

100 to

1,000

2,500

5,000

10,000

Industrial Uses

12

Max

4.19E-03

4.78E-03

1.54E-03

5.67E-04

2.80E-04

7.15E-05

8.79E-06

3.22E-06

1.13E-06

Mean

8.76E-04

1.14E-03

4.08E-04

1.70E-04

9.63E-05

1.69E-05

1.48E-06

5.45E-07

1.94E-07

Median

8.76E-05

1.14E-04

3.83E-05

1.65E-05

9.94E-06

2.23E-06

3.19E-07

1.17E-07

4.04E-08

Min

7.75E-13

1.69E-12

2.40E-09

2.50E-08

1.23E-08

1.10E-09

6.36E-11

1.97E-11

6.14E-12

Laboratory
Chemical Use

1

Max

2.06E-03

2.15E-03

7.26E-04

3.75E-04

2.06E-04

3.76E-05

2.27E-06

7.57E-07

2.50E-07

Mean

6.84E-04

8.52E-04

3.58E-04

1.64E-04

8.46E-05

1.39E-05

7.77E-07

2.59E-07

8.55E-08

Median

4.30E-04

5.65E-04

3.15E-04

1.36E-04

6.68E-05

1.08E-05

4.82E-07

1.59E-07

5.24E-08

Min

7.39E-05

1.48E-04

1.16E-04

4.55E-05

2.09E-05

3.08E-06

1.03E-07

3.30E-08

1.07E-08

Manufacturing

1

Single Facility

8.73E-03

1.63E-02

7.69E-03

3.22E-03

1.59E-03

1.42E-04

8.21E-06

2.54E-06

7.92E-07

PET

Manufacturing

13

Max

8.01E-03

9.57E-03

3.50E-03

1.40E-03

6.43E-04

1.07E-04

2.07E-05

1.24E-05

6.58E-06

Mean

1.41E-03

1.89E-03

7.83E-04

3.36E-04

1.85E-04

3.31E-05

4.23E-06

2.08E-06

9.60E-07

Median

8.00E-04

1.64E-03

5.21E-04

2.27E-04

1.42E-04

2.64E-05

2.48E-06

1.09E-06

3.94E-07

Min

6.04E-12

8.54E-11

3.01E-08

2.43E-07

5.56E-07

3.02E-07

4.33E-08

2.07E-08

9.30E-09

Spray Foam
Application

1

Max

7.79E-07

8.40E-07

2.85E-07

1.50E-07

8.55E-08

1.55E-08

1.72E-09

6.30E-10

2.45E-10

Mean

2.68E-07

3.30E-07

1.34E-07

6.21E-08

3.28E-08

5.29E-09

4.78E-10

1.67E-10

5.97E-11

Median

1.41E-07

1.95E-07

1.14E-07

4.88E-08

2.36E-08

3.64E-09

2.25E-10

7.40E-11

2.40E-11

Min

2.51E-08

4.43E-08

3.45E-08

1.36E-08

6.07E-09

8.42E-10

3.26E-11

1.10E-11

3.74E-12

2656

Page 87 of 484


-------
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

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Based on the air concentrations estimated through the Ambient Air: Single Year Methodology, EPA also
estimated potential aggregate air concentrations resulting from the combined releases of multiple
facilities in proximity to each other. Details of the methods used to aggregate exposure and
corresponding risk are presented in Appendix J.4.

2.3.3.2.3	Ambient Air: Multi-Year Analysis (IIOAC)

The multi-year analysis utilizes EPA's IIOAC model to estimate high-end and central tendency (mean)
1,4-dioxane concentrations in ambient air at three distances from an emitting facility: 100, 100 to 1,000,
and 1,000 m. The multi-year analysis incorporates SACC recommendations on the 2022 Fenceline
Report by evaluating multiple years of chemical release data to estimate exposures and associated risks
to fenceline communities. This is achieved for 1,4-dioxane by expanding upon methodologies described
in the 2022 Fenceline Report and conducting a facility-by-facility evaluation of all 1,4-dioxane releases
reported to TRI (2015 through 2020). Data for these 6 years were obtained from the TRI database (TRI
basic plus files downloaded on August 5, 2022). Annual release data for 1,4-dioxane were extracted
from the entire TRI data set for all facilities reporting air releases of 1,4-dioxane for one or more years
between 2015 and 2020. Facilities were categorized into occupational exposure scenarios for modeling
purposes to inform the release scenarios evaluated.

The multi-year analysis provides highlights of the year-to-year variability that exists in the release data
and illustrates the potential impact of considering multiple years of TRI data on exposure and risk
estimates. The findings from the multi-year analysis can also be used in a comparative manner to
determine how representative the single year of data used for the Ambient Air: Single Year
Methodology (AERMOD) presented in the 2022 Fenceline Report is or to provide additional confidence
in the findings from the Ambient Air: Single Year Methodology (AERMOD) described in the 2022
Fenceline Report for purposes of estimating exposures and associated risks to fenceline communities. In
broader terms, the multi-year analysis provides both a broad analysis of multiple years of release data
and enables a general comparison to the Ambient Air: Single Year Methodology (AERMOD) results
described above and in the 2022 Fenceline Report.

2.3.3.2.4	Ambient Air: IIOAC Methodology for COUs without Site-Specific Data
(Hydraulic Fracturing, Industrial, and Institutional Laundry Facilities)

For COUs without site-specific data, EPA's IIOAC model was used to estimate high-end and central
tendency (mean) 1,4-dioxane concentrations in ambient air at three distances from an emitting facility
(100, 100 to 1,000, and 1,000 m). This methodology was applied for three unique COUs (hydraulic
fracturing, and industrial, and institutional laundry facilities) where there was no site-specific data
available for modeling. Environmental releases (fugitive and stack) along with other data (like days of
release) for these COUs were estimated using Monte Carlo modeling. As such, the Ambient Air: IIOAC
Methodology for COUs without site-specific data was developed to allow modeling all possible
iterations of releases provided across eight different exposure scenarios, including consideration of
source attribution as well as actual days of release. Additionally, the product form for laundry detergent
was provided, allowing for analysis of releases associated with detergent in vapor only form, as well as
solid form (particulate) either coarse (PM10) or fine (PM2.5). A description of this methodology is
provided in Appendix J along with a summary of the model inputs and exposure scenarios evaluated. A
full list of the inputs, exposure scenarios, and results is provided in 1,4-Dioxane Supplemental
Information File: Air Exposure and Risk Estimates for 1,4-Dioxane Emissions from Hydraulic
Fracturing Operations (	s23b) and 1,4-Dioxane Supplemental Information File: Air

Exposures and Risk Estimates for Industrial Laundry (	'023c).

Page 88 of 484


-------
2702

2703

2704

2705

2706

2707

2708

2709

2710

2711

2712

2713

2714

2715

2716

2717

2718

2719

2720

2721

2722

2723

2724

2725

2726

2727

2728

2729

2730

2731

2732

2733

2734

2735

2736

2737

2738

2739

2740

2741

2742

2743

2744

2745

2746

2747

2748

2749

2750

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2.3.3.3 Strengths, Limitations, and Sources of Uncertainty for Modeled Air
Concentrations

EPA has medium to high confidence in the air concentrations estimated from TRI release data using
IIOAC and AERMOD.

IIOAC

IIOAC estimates air concentrations at three pre-defined distances (100, 100 to 1,000, and 1,000 m). The
inherent distance limitations of IIOAC do not allow estimation of exposures closer to a facility (less than
100 m from the facility) where we expect to see higher exposures from fugitive releases.

IIOAC uses meteorological data from 14 pre-defined meteorological stations representing large regions
across the United States. This generalizes the meteorological data used to estimate exposure
concentrations where competing conditions can influence the exposure concentrations modeled upwind
and downwind of a releasing facility. To reduce the uncertainties associated with using regional
meteorological data, EPA conducted a sensitivity analysis of all 14 pre-defined meteorological stations
to identify which two within IIOAC tended to result in a high-end and central tendency estimate of
exposure concentrations. This maintained a more conservative exposure concentration estimate which is
then used in calculations to estimate risks. This approach adds confidence to the findings by ensuring,
potential risks would be captured under a high-end exposure scenario, while also providing insight into
potential risks under a less conservative exposure scenario (central tendency).

AERMOD

AERMOD is an EPA regulatory model and has been thoroughly peer reviewed; therefore, the general
confidence in results from the model is high but relies on the integrity and quality of the inputs used and
interpretation of the results. For the full analysis, EPA used releases reported to the 2019 TRI as direct
inputs to AERMOD. Although there is some uncertainty around the representativeness of using only a
single year of data, AERMOD successfully estimated exposure concentrations to fenceline communities.
Furthermore, in response to SACC recommendations to use multiple years of data to estimate exposures
and associated risks, EPA developed the IIOAC and conducted a multi-year analysis using 6 years of
TRI data and compared the results to those of AERMOD and found exposure concentration estimates
from the 2019 data is generally representative of other years.

AERMOD relied upon the latitude/longitude information reported by each facility to TRI as the location
for the point of release. While this may generally be a close approximation of the release point for a
small facility (for example a single building), it may not represent the release point within a much larger
facility. Therefore, there is some uncertainty associated with the modeled distances from each release
point and the associated exposure concentrations to which fenceline communities may be exposed. For
small facilities where the latitude/longitude may closely approximate the release point, there is a less
uncertainty that the estimated exposure at the associated distance is representative of exposure to
fenceline communities at that distance. For larger facilities, where the latitude/longitude may be several
hundred meters away from the actual release point, there is a higher uncertainty that the estimated
exposure at the associated distance is representative of exposure to fenceline communities at that
distance.

The TRI datasets used for both AERMOD and IIOAC do not include source specific stack parameters
that can affect plume characteristics and associated dispersion of the plume. Therefore, EPA used pre-
defined stack parameters within IIOAC to represent stack parameters of all facilities modeled using each
of these methodologies. Those stack parameters include a stack height 10 m above ground with a 2-
meter inside diameter, an exit gas temperature of 300° Kelvin, and an exit gas velocity of 5 m per

Page 89 of 484


-------
2751

2752

2753

2754

2755

2756

2757

2758

2759

2760

2761

2762

2763

2764

2765

2766

2767

2768

2769

2770

2771

2772

2773

2774

2775

2776

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

second (see Table 6 of the IIOAC User Guide). These parameters were selected since they represent a
slow-moving, low-to-the-ground plume with limited dispersion which results in a more conservative
estimate of exposure concentrations at the distances evaluated. As such, these parameters may result in
some overestimation of emissions for certain facilities modeled.

Additionally, the assumption of a 10x 10 area source for fugitive releases may impact the exposure
estimates very near a releasing facility (5 and 10 m from a fugitive release). This assumption places the
receptor at 5 m directly on top of the release point which may result in an over or underestimation of
exposure. This assumption places the 10-meter receptor just off the release point that may again result in
either an over or underestimation of exposure depending on other factors like meteorological data,
release heights, and plume characteristics.

For facilities reporting releases to TRI via a TRI Form A (which is allowed for use by those facilities
releasing less than 500 lb of the chemical reported), EPA assumed the maximum release value of 500 lb
for exposure modeling purposes. TRI Form A reporters do not provide source attribution (fugitive or
stack releases) so EPA modeled each facility associated with a Form A submittal twice—once assuming
all 500 lb of the reporting threshold was fugitive and once assuming all 500 lb of the reporting threshold
was stack. There is no way to attribute a certain portion of the releases to each release type, so this
modeling approach represents a conservative estimate, in terms of total release, but may overestimate
exposure concentrations associated with each release type if a facility did not actually release all 500 lb
via a single release type or even combined release type. To avoid the potential double counting of
facility releases for TRI Form A reporters, when presenting potential exposures EPA presented only the
highest (more conservative) exposure concentration estimated for either of the two release types for
purposes of evaluating potential risks to fenceline communities. Given the exposure scenarios modeled,
this tended to result from the exposure scenario which assumed all 500 lb of the release were fugitive
releases.

Page 90 of 484


-------
2777

2778

2779

2780

2781

2782

2783

2784

2785

2786

2787

2788

2789

2790

2791

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3 HUMAN EXPOSURES

1,4-Dioxane - Human Exposures (Section 3):

Key Points

EPA evaluated reasonably available information for occupational exposures to 1,4-dioxane present
as a byproduct, and general population, including PESS, exposures to 1,4-dioxane present in the
environment.

•	Occupational exposures for workers and ONUs to 1,4-dioxane present as a byproduct occur
during manufacturing, through use of commercial products, or in hydraulic fracturing
operations.

•	General population exposures to 1,4-dioxane occur when 1,4-dioxane is present in potential
drinking water sources or ambient air, particularly in fenceline communities.

•	EPA considered the potential for increased exposures across PESS factors throughout the
exposure assessment. PESS categories incorporated into this supplemental exposure
assessment include

o Lifestage (including formula-fed infant exposures),
o Occupational exposures (including high-end exposure scenarios), and
o Geography/site-specific factors (i.e., fenceline community exposures)

3.1 Occupational Exposures

1,4-Dioxane - Occupational Exposures (Section 3.1):

Key Points

EPA considered the reasonably available information to evaluate occupational exposures.

•	EPA estimated occupational exposures to 1,4-dioxane through air and skin. EPA estimated
both high-end and central tendency exposures for occupational exposure scenarios
associated with each COU.

•	Exposure for most COUs was estimated based on monitoring data. For COUs without
monitoring data, EPA applied Monte Carlo statistical modeling approaches to estimate
exposures.

The following sections describe EPA's approach to assessing occupational exposures for OESs
involving industrial and commercial products containing 1,4-dioxane as a byproduct. The assessed OESs
include textile dye, antifreeze, surface cleaner, dish soap, dishwasher detergent, institutional and
industrial laundries, paints and floor lacquer, PET byproducts, ethoxylation process byproducts, and
hydraulic fracturing. For a crosswalk linking COUs to OESs, see Table 2-1. The remaining OESs have
occupational exposure assessments in Section 2.4.1 of the Final Risk Evaluation for 1,4-Dioxane (U.S.
EPA. 2020c).

EPA distinguishes between exposures to workers and exposures to ONUs. Normally, workers may
handle 1,4-dioxane and have direct contact with the chemical, while ONUs are working in the general

Page 91 of 484


-------
2792

2793

2794

2795

2796

2797

2798

2799

2800

2801

2802

2803

2804

2805

2806

2807

2808

2809

2810

2811

2812

2813

2814

2815

2816

2817

2818

2819

2820

2821

2822

2823

2824

2825

2826

2827

2828

2829

2830

2831

2832

2833

2834

2835

2836

2837

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

vicinity of workers but do not handle 1,4-dioxane and do not have direct contact with 1,4-dioxane being
handled by the workers.

EPA evaluated acute and chronic inhalation exposures to workers and ONUs, and dermal exposures to
workers. EPA did not assess dermal exposures to ONUs as EPA does not expect ONUs to have routine
dermal exposures in the course of their work.

The occupational exposure assessment for each COU comprises the following components:

•	Process Description of the COU, including the role of the chemical 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.

•	Number of Sites that use the chemical for the given COU.

•	Number of Workers and ONUs potentially exposed to the chemical for the given COU. Unless
mentioned otherwise in this report, the total number of workers and ONUs are number of
personnel per site per day. The details on estimation of the number of workers and ONUs are
discussed below for each COU.

•	Central Tendency and High-End Estimates of Inhalation Exposure to workers and
occupational non-users. See "General Approach and Methodology for Environmental Releases"
for a discussion of EPA's statistical analysis approach for assessing inhalation exposure.

•	Dermal Exposure estimates for multiple scenarios, accounting for simultaneous absorption and
evaporation, and different protection factors of glove use.

•	Users include adult workers (greater than 16 years old) exposed to 1,4-dioxane for 8-hour
exposure.

•	ONUs include female and male adult workers (greater thanl6 years old) exposed to 1,4-dioxane
indirectly by being in the same work area of the building.

3.1.1 Approach and Methodology

EPA developed occupational exposure values representative of central tendency (50th percentile, mean)
conditions and high-end (90th and 99.9th percentiles). Additional explanation of central tendency and
high-end conditions are described in the Final Risk Evaluation for 1,4-Dioxane (	2020c).

3.1.1.1 Process Description, Number of Sites, Number of Workers, and ONUs

EPA performed a literature search to find descriptions of processes involving 1,4-dioxane and
worker activities that could potentially result in occupational exposures. This literature search was
specific to the scope of this draft supplement and is described in Section 1.4. A summary of the data
quality evaluation results for the 1,4-dioxane occupational exposure sources are presented in the
attachment Systematic Review Supplemental File: Data Quality Evaluation and Data Extraction
Information for Environmental Release and Occupational Exposure (	2023t).

EPA used a variety of sources to supplement the data found through the Systematic Review
process. The additional sources included relevant NIOSH Health Hazard Evaluations, Generic
Scenarios, and ESDs. These sources were sometimes used to provide process descriptions of the COUs
as well as estimates for the number of sites and number of workers. Because CDR data were not
available for the COUs included in this occupational exposure assessment, EPA used data from the
Bureau of Labor Statistics (BLS) and the U.S. Census' Statistics of US Businesses (SUSB) to estimate
the number of sites, workers, and ONUs for each OES. This approach involved the identification of
relevant Standard Occupational Classification (SOC) codes within the BLS data for the identified
NAICS codes for each OES. First, EPA identified the affected NAICS codes. Then, EPA reviewed

Page 92 of 484


-------
2838

2839

2840

2841

2842

2843

2844

2845

2846

2847

2848

2849

2850

2851

2852

2853

2854

2855

2856

2857

2858

2859

2860

2861

2862

2863

2864

2865

2866

2867

2868

2869

2870

2871

2872

2873

2874

2875

2876

2877

2878

2879

2880

2881

2882

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

occupation descriptions to designate which SOC codes contained potentially exposed workers and
ONUs. The Agency refined the estimates by using U.S. Census Bureau data. Next, EPA estimated the
percentage of workers using 1,4-dioxane instead of other chemicals to calculate number of workers per
site. Finally, this data was separated by COU. Additional details on this approach can be found in
Appendix G.5 of the Final Risk Evaluation for 1,4-Dioxane (	020c). These estimates were

utilized in Monte Carlo modeling of environmental releases and occupational exposures.

3.1.1.2 Inhalation Exposures Approach and Methodology

EPA used inhalation monitoring data from literature sources having high or medium data quality ratings
during data evaluation. EPA used modeling approaches to estimate potential inhalation exposures where
inhalation monitoring data were not available.

The Agency reviewed workplace inhalation monitoring data collected by government agencies such as
OSHA and NIOSH, and monitoring data found in published literature {i.e., personal exposure
monitoring data and area monitoring data). Central tendency and high-end exposure values were
calculated from the monitoring data provided in the sources depending on the size of the dataset {i.e.,
number of data points). Where discrete sampling points were not provided in the source and EPA was
unable to calculate central tendency and high-end values, the Agency used values of central tendency
and high-end that were provided in the source. EPA's approach for evaluating central tendency and
high-end estimates from inhalation monitoring data is further discussed in the Final Risk Evaluation for
1,4-Dioxane (	)20c). EPA used the following types of monitoring data of 1,4-dioxane from

various sources to estimate occupational inhalation exposure:

•	Personal sample monitoring data from directly applicable scenarios {e.g., personal breathing
zone [PBZ]). This type of monitoring data was used for the textile dye, surface cleaner, dish
soap, paint and floor lacquer, PET byproduct, and the Ethoxylation Process Byproduct OESs.

•	Personal sample monitoring data from potentially applicable or similar scenarios. Specifically,
PBZ data from the dish soap OES was also used for the dishwasher detergent OES because these
OESs are expected to be similar.

EPA used the following models and modeling approaches to estimate occupational inhalation exposure
where no monitoring data were found:

•	Monte-Carlo statistical modeling approaches, which was used for the antifreeze, laundry
detergent, and hydraulic fracturing OES. EPA developed these models for the purposes of this
assessment. The models and the associated sources of data used in the modeling are described in
detail in Appendices F.7, F.8, and F.9, respectively.

•	Additional modeling approaches, including the use of surrogate data and fundamental modeling
approaches for the spray polyurethane foam OES in the Final Risk Evaluation for 1,4-Dioxane
(U.S. EPA. 2020c). While this OES is included in the scope of this draft supplement, EPA
evaluated occupational exposure estimates for this OES in the published risk evaluation and
these estimates remain unchanged in this draft supplement.

•	EPA AP-42 Loading Model estimates vapor releases that occur when vapor is displaced by liquid
during container loading. It calculates a vapor generation rate (G) using the physio-chemical
properties of the chemical.

•	EPA Mass Balance Inhalation Model estimates occupational inhalation exposures assuming the
air immediately around the source of exposure behaves as a well-mixed zone. EPA used the
vapor generation rate (G), calculated using the EPA AP-42 Loading Model, in conjunction with
this model to develop estimates of inhalation exposure.

Page 93 of 484


-------
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

2915

2916

2917

2918

2919

2920

2921

2922

2923

2924

2925

2926

2927

2928

2929

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

• Generic Model for Central Tendency and High-End Inhalation Exposure to Total and Respirable
Particulates Not Otherwise Regulated (PNOR) estimates occupational inhalation exposures to
particulates containing the chemical using OSHA PNOR data.

EPA did not utilize occupational exposure limits to estimate occupational inhalation exposures in this
assessment because sufficient monitoring data or modeling approaches were available for all OES.

EPA then used measured or modeled air concentrations to calculate exposure concentration metrics
essential for risk assessment. These exposures are presented as 8-hour time weighted averages (TWAs)
and used to calculate average daily concentrations (ADCs) and lifetime average daily concentrations
(LADCs). The ADC is used to estimate chronic, non-cancer risks and the LADC is used to estimate
chronic, cancer risks. These calculations required additional parameter inputs, such as years of exposure,
exposure duration and frequency, and lifetime years. See Appendix F.l for more information about
parameters and equations used to calculate acute and chronic exposures.

3.1.1.3	Dermal Exposures Approach and Methodology

EPA modeled dermal doses using the EPA Dermal Exposure to Volatile Liquids Model. This model
determines a dermal potential dose rate based on an assumed amount of liquid on skin during one
contact event per day and the steady-state fractional absorption for 1,4-dioxane. The amount of liquid on
the skin is adjusted by the weight fraction of 1,4-dioxane in the liquid to which the worker is exposed.
This is the same approach that EPA used in the Final Risk Evaluation for 1,4-Dioxane (U.S. EPA.
2020c).

3.1.1.4	Engineering Controls and Personal Protective Equipment

EPA's occupational exposure estimates do not assume the use of engineering controls or PPE.
Additionally, as discussed in Section 5.2.1, the risk estimates are based on exposures to workers in the
absence of PPE such as gloves or respirators. Reasonably available monitoring data or information on
effectiveness of engineering control and PPE for reducing occupational exposures to 1,4-dioxane during
the assessed OES were not available. This section presents a general discussion on engineering controls
and PPE for informative purposes only.

OSHA recommends employers utilize the hierarchy of controls for reducing or removing hazardous
exposures. The most effective controls are elimination, substitution, or engineering controls. Respirators,
and any other personal protective equipment (PPE), are the last means of worker protection in the
hierarchy of controls and should only be considered when process design and engineering controls
cannot reduce workplace exposure to acceptable levels. OSHA's Respiratory Protection Standard (29
CFR 1910.134) provides a summary of respirator types by their assigned protection factor (APF). OSHA
defines the APF to mean the workplace 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 according to the requirements of the OSHA Respiratory Protection
Standard. Exposure limits, respirator requirements, worker respirator use rates, and a table of APFs for
different types of respirators are provided in the 2020 RE (	2020c).

OSHA's hand protection standard (29 CFR 1910.138) states that employers must select and require
employees to use appropriate hand protection when employees are expected to be exposed to hazards
such as those from skin absorption of harmful substances; severe cuts or lacerations; severe abrasions;
punctures; chemical burns; thermal burns; and harmful temperature extremes. Dermal protection
selection provisions are provided in § 1910.138(b) and require that appropriate hand protection is
selected based on the performance characteristics of the hand protection relative to the task(s) to be
performed, conditions present, duration of use, and the hazards to which employees will be exposed.

Page 94 of 484


-------
2930

2931

2932

2933

2934

2935

2936

2937

2938

2939

2940

2941

2942

2943

2944

2945

2946

2947

2948

2949

2950

2951

2952

2953

2954

2955

2956

2957

2958

2959

2960

2961

2962

2963

2964

2965

2966

2967

2968

2969

2970

2971

2972

2973

2974

2975

2976

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Unlike respiratory protection, OSHA standards do not provide protection factors (PFs) associated with
various hand protections. Data regarding the frequency of effective glove use in industrial and
commercial settings is limited. While there are no regulatory standards containing dermal protection
factors, the European Center for Ecotoxicity and Toxicology of Chemicals (ECETOC) targeted risk
assessment (TRA) tool includes fixed, assigned protection factors equal to 5, 10, or 20 for various
dermal protection strategies. These are discussed in Appendix F.3 and further explained in the 2020 RE
(I	>020cY

3.1.2 Occupational Exposure Estimates

In this section, EPA provides a summary of the exposure estimates for each OES, including estimates
for number of workers and ONUs, inhalation exposures, and dermal exposures. For the crosswalk
linking COU to OES, see Table 2-1. Note that EPA assessed dermal exposures for all OESs with the
same methodology, which is described at the end of this section.

3.1.2.1 Summary of Inhalation Exposure Assessment

EPA estimated central tendency and high-end occupational inhalation exposures using various methods
and information sources—including OSHA data, NIOSH health hazard evaluation data, and GSs and
ESDs with Monte Carlo modeling. EPA estimated inhalation exposures as 8-hour TWA values for the
COUs included in this draft supplement per Table 2-1. Using the estimated central tendency and high-
end inhalation exposures with the estimated exposure frequency, EPA then calculated the cancer and
non-cancer exposures using the calculations described in Appendix F.l.

A summary of the occupational inhalation exposures is presented Table Apx F-32. EPA used
monitoring data to estimate occupational inhalation exposures to workers for the textile dye, surface
cleaner, dish soap, dishwasher detergent, paint and floor lacquer, PET byproduct, and ethoxylation
byproduct conditions of use. This monitoring data was found to be relevant to these scenarios and based
on medium to high data quality. However, several of the scenarios had a low number of samples and
may have preceded changes in current industry practices. Additionally, sufficient representation of the
entire industry is uncertain due to the limited number of sites. For the remaining conditions of use
included in this draft supplement, which are antifreeze, laundry detergent, and hydraulic fracturing, EPA
did not find reasonably available monitoring data and estimated worker inhalation exposure using GSs
and ESDs with Monte Carlo modeling. The applied models are directly relevant to these conditions of
use, but the underlying distributions may not sufficiently capture variability across entire industry
sectors. For both measured and modeled data, the degree of certainty to which these data represent the
true distribution of exposure and the potential over- or underestimation of exposure is unknown.

Monitoring data and modeling approaches were not available to estimate occupational inhalation
exposures for ONUs. The ONU exposures are anticipated to be lower than worker exposures since
ONUs do not typically directly handle the chemical.

The PET byproduct and textile dyes conditions of use had the highest central tendency and high-end
worker inhalation exposure values, respectively. For PET byproduct, worker inhalation exposures were
estimated using OSHA monitoring data, which resulted in central tendency exposure of 4.7 mg/m3 and
high-end exposure of 47 mg/m3. For textile dyes, worker inhalation exposures were also estimated using
OSHA monitoring data, which resulted in central tendency exposure of 0.066 mg/m3 and high-end
exposure of 74 mg/m3.

The monitoring data sources and GSs and ESDs used to estimate occupational inhalation exposures all
had overall data quality determinations of either medium or high. The basis for determining overall data

Page 95 of 484


-------
2977

2978

2979

2980

2981

2982

2983

2984

2985

2986

2987

2988

2989

2990

2991

2992

2993

2994

2995

2996

2997

2998

2999

3000

3001

3002

3003

3004

3005

3006

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

quality is described in this section. In summary, each source is evaluated on multiple metrics based on
defined criteria. The individual metric ratings are used to obtain an overall study rating. All metrics have
an equal weight in determining the overall study rating. The resulting values are converted to an overall
data quality determination of "high," "medium," "low," "critically deficient," or "not rated/not
applicable." For more details on this process, see Section 5 of the 2021 Draft Systematic Review
Protocol

This section also includes information on the weight of the scientific evidence conclusions for these
estimates, and a summary of the strengths, limitations, assumptions, and key sources of uncertainty for
these estimates.

3.1.2.2 Summary of Dermal Exposures Assessment

Table 3-1 presents the estimated dermal absorbed dose for workers in various OES. The dose estimates
assume one dermal exposure event (applied dose) per workday and that approximately 78 or 86 percent
of the applied dose is absorbed through the skin (depending on whether the OES is industrial or
commercial). The exposure estimates are provided for each OES, where the OES are "binned" based on
characteristics known to effect dermal exposure such as the maximum weight fraction of 1,4-dioxane
that could be present in that OES, open or closed system use of 1,4-dioxane, and large or small-scale
use. For a more detailed description of EPA's dermal assessment approach and each bin, see Appendix
F.3.

As shown in the Table 3-1, the calculated dermal absorbed dose for workers is lower in comparison to
those presented in the December 2020 Final Risk Evaluation for 1,4-Dioxane (	)20c). This is

due to the relatively lower concentrations of 1,4-dioxane found for the OES included in this draft
supplement than for those included in the 2020 RE. As noted previously, EPA did not assess dermal
exposures to ONUs as the Agency does not expect ONUs to have routine dermal exposures in the course
of their work. Depending on the OES, ONUs may have incidental dermal exposures due to surface
contamination. However, data (e.g., frequency and amount of liquid on the skin after contact) were not
identified to assess this exposure.

Table 3-1. Estimated Dermal Absorbed Dose (mg/day) for Workers in Various Conditions of Use







Weight
Fraction

(Max
Yilm,,)

No Gloves
(PF = 1)

Exposures Due to Glove Permeation/Chemical
Breakthrough (mg/day)

OES

Bin

Use Setting

Protective
Gloves
(PF = 5)

Protective

Gloves
(PF= 10)

Protective Gloves
(Industrial Uses
Only, PF = 20)

Textile Dye

7

Industrial
and

Commercial

4.7E-06

0.003 (CT)
0.009 (HE)

0.001 (CT)
0.002 (HE)

3.0E-4 (CT)
0.001 (HE)

1.5E-4 (CT)
4.5E-4 (HE)

Antifreeze

8

Commercial

8.6E-05

0.055 (CT)
0.165 (HE)

0.011 (CT)
0.033 (HE)

0.006 (CT)
0.017 (HE)

N/A

Surface
Cleaner

9

Commercial

9.0E-06

0.006 (CT)
0.017 (HE)

0.001 (CT)
0.003 (HE)

0.001 (CT)
0.002 (HE)

N/A

Dish Soap

10

Commercial

2.04E-04

0.131 (CT)
0.393 (HE)

0.026 (CT)
0.079 (HE)

0.013 (CT)
0.039 (HE)

N/A

Dishwasher
Detergent

11

Commercial

5.1E-05

0.033 (CT)
0.098 (HE)

0.007 (CT)
0.020 (HE)

0.003 (CT)
0.010 (HE)

N/A

Page 96 of 484


-------
3007

3008

3009

3010

3011

3012

3013

3014

3015

3016

3017

3018

3019

3020

3021

3022

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023







Weight
Fraction

(Max

Yilcrm)

No Gloves
(PF = 1)

Exposures Due to Glove Permeation/Chemical
Breakthrough (mg/dav)

OES

Bin

Use Setting

Protective
Gloves
(PF = 5)

Protective

Gloves
(PF= 10)

Protective Gloves
(Industrial Uses
Only, PF = 20)

Laundry
Detergent
(Industrial
and

Institutional)

12

Industrial
and

Commercial

1.4E-05

0.009 (CT)
0.027 (HE)

0.002 (CT)
0.005 (HE)

0.001 (CT)
0.003 (HE)

4.5E-4 (CT)
0.001 (HE)

Paint and

Floor

Lacquer

13

Industrial
and

Commercial

3.0E-05

0.019 (CT)
0.058 (HE)

0.004 (CT)
0.012 (HE)

0.002 (CT)
0.006 (HE)

0.001 (CT)
0.003 (HE)

Polyethylene
Terephthalate
(PET)
Byproduct















14

Industrial

1.4E-03

0.827 (CT)
2.48 (HE)

0.165 (CT)
0.496 (HE)

0.083 (CT)
0.248 (HE)

0.041 (CT)
0.124 (HE)

Ethoxylation

Process

Byproduct







Hydraulic
Fracturing

15

Industrial
and

Commercial

0.05

32.1	(CT)

96.2	(HE)

6.41 (CT)
19.2 (HE)

3.21 (CT)
9.62 (HE)

1.60 (CT)
4.81 (HE)

CT = central tendency; HE = high-end; PF = protection factor

3.1.2.3 Weight of the Scientific Evidence Conclusions for Occupational Exposure
Information

Table 3-2 provides a summary of EPA's overall weight of the scientific evidence conclusions for its
occupational exposure estimates for each of the assessed OES. These determinations are OES-specific.
For a description of overall confidence in all inhalation exposures, see Section 3.3.1.1. For an
explanation of EPA's judgement on the weight of the scientific evidence conclusion, see Section 2.2.1.2.
Factors that increase and decrease the strength of the weight of the scientific evidence are listed in
Table Apx C-5.

Due to a lack of data, EPA was not able to estimate ONU inhalation exposure from monitoring data or
models, so a qualitative assessment of potential ONU exposures was made. Similarly, EPA did not
assess dermal exposures to ONUs as EPA does not expect ONUs to have routine dermal exposures in
the course of their work. Depending on the COU, ONUs may have incidental dermal exposures due to
surface contamination. However, data (e.g., frequency and amount of liquid on the skin after contact)
were not identified to assess this exposure. Finally, due to the absence of dermal monitoring data, these
columns were omitted from Table 3-2.

Page 97 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3023 Table 3-2. Summary of the Weight of the Scientific Evidence for Occupational Exposure Estimates by PES

OES

Inhalation Exposure

Dermal Exposure

Monitoring

Modeling

Weight of the
Scientific Evidence

Modeling

Weight of the
Scientific Evidence

Worker

# Data
Points

Data Quality
Rating"

Worker

Data
Quality
Rating"

Worker

Worker6

Worker

Textile Dye

~

14

H

X

N/A

Moderate

~

Moderate

Antifreeze

X

N/A

N/A

~

H

Moderate

~

Moderate

Surface Cleaner

~

49

H

X

N/A

Moderate to Robust

~

Moderate

Dish Soap

~

29

M

X

N/A

Slight to Moderate

~

Moderate

Dishwasher Detergent

~

29

M

X

N/A

Slight to Moderate

~

Moderate

Laundry Detergent
(Industrial and
Institutional)

X

N/A

N/A

~

M

Moderate

~

Moderate

Paint and Floor
Lacquer

~

17

H

X

N/A

Moderate

~

Moderate

PET Byproduct

~

11

H

X

N/A

Moderate

~

Moderate

Ethoxylation Process
Byproduct

~

1

H

X

N/A

Slight to Moderate

~

Moderate

Hydraulic Fracturing

X

N/A

N/A

~

M

Moderate to Robust

~

Moderate

"Data quality ratings of modeling approaches are based on the GS/ESD that was used in tandem with Monte Carlo modeling.

b Data quality ratings are not applicable for the dermal modeling approach because this modeling was conducted with an already-developed EPA model.

3024

Page 98 of 484


-------
3025

3026

3027

3028

3029

3030

3031

3032

3033

3034

3035

3036

3037

3038

3039

3040

3041

3042

3043

3044

3045

3046

3047

3048

3049

3050

3051

3052

3053

3054

3055

3056

3057

3058

3059

3060

3061

3062

3063

3064

3065

3066

3067

3068

3069

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3.1.2.4 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
Occupational Exposure Assessment

3.1.2.4.1	Number of Workers

There are uncertainties surrounding the estimated number of workers potentially exposed to 1,4-dioxane.
First, BLS employment data for each industry/occupation combination are only available at the 3-, 4-, or
5-digit NAICS level, rather than at the full 6-digit NAICS level. This lack of specificity could result in
an overestimate of the number of exposed workers if some 6-digit NAICS are included in the less
granular BLS estimates but are not likely to use 1,4-dioxane for the assessed applications. EPA
addressed this issue by refining the OES estimates using total employment data from the U.S. Census'
SUSB. However, this approach assumes that the distribution of occupation types (SOC codes) in each 6-
digit NAICS is equal to the distribution of occupation types at the parent 5-digit NAICS level. If the
distribution of workers in occupations with 1,4-dioxane exposure differs from the overall distribution of
workers in each NAICS, then this approach will result in inaccuracy. The effects of this uncertainty on
the number of worker estimates are unknown, as the uncertainties may result in either over or
underestimation of the estimates depending on the actual distribution.

Second, EPA's determinations of industries (represented by NAICS codes) and occupations (represented
by SOC codes) that are associated with the OES assessed in this report are based on EPA's
understanding of how 1,4-dioxane is used in each industry. The designations of which industries and
occupations have potential exposures is a matter of professional judgement; therefore, the possibility
exists for the erroneous inclusion or exclusion of some industries or occupations. This may result in
inaccuracy but would be unlikely to systematically either overestimate or underestimate the count of
exposed workers.

3.1.2.4.2	Analysis of Inhalation Exposure Monitoring Data

The principal limitation of the monitoring data is the uncertainty in the representativeness of the data
due to some scenarios having limited exposure monitoring data in literature. Where few data are
available, the assessed exposure levels are unlikely to be representative of worker exposure across the
entire job category or industry. This may particularly be the case when monitoring data were available
for only one site. Differences in work practices and engineering controls across sites can introduce
variability and limit the representativeness of monitoring data. Age of the monitoring data can also
introduce uncertainty due to differences in workplace practices and equipment used at the time the
monitoring data were collected compared those currently in use. Therefore, older data may overestimate
or underestimate exposures, depending on these differences. The effects of these uncertainties on the
occupational exposure assessment are unknown, as the uncertainties may result in either overestimation
or underestimation of exposures depending on the actual distribution of 1,4-dioxane air concentrations
and the variability of work practices among different sites.

In some scenarios where monitoring data were available, EPA did not find sufficient data to determine
complete statistical distributions. Ideally, EPA will present 50th and 95th percentiles for each exposed
population. In the absence of percentile data for monitoring, the mean or midpoint of the range may
serve as a substitute for the 50th percentile of the actual distributions. Similarly, the highest value of a
range may serve as a substitute for the 95th percentile of the actual distribution. However, these
substitutes are uncertain. The effects of these substitutes on the occupational exposure assessment are
unknown, as the substitutes may result in either overestimation or underestimation of exposures
depending on the actual distribution.

Page 99 of 484


-------
3070

3071

3072

3073

3074

3075

3076

3077

3078

3079

3080

3081

3082

3083

3084

3085

3086

3087

3088

3089

3090

3091

3092

3093

3094

3095

3096

3097

3098

3099

3100

3101

3102

3103

3104

3105

3106

3107

3108

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3.1.2.4.3	Modeled Inhalation Exposures

EPA addressed variability in inhalation models by identifying key model parameters to apply a
statistical distribution that mathematically defines the parameter's variability. EPA defined statistical
distributions for parameters using documented statistical variations where available. Where the
statistical variation was unknown, assumptions were made to estimate the parameter distribution using
available literature data, such as GSs and ESDs. However, there is uncertainty as to the
representativeness of the parameter distributions with respect to the modeled scenario because the data
are often not specific to sites that use 1,4-dioxane. In general, the effects of these uncertainties on the
exposure estimates are unknown, as the uncertainties may result in either overestimation or
underestimation on exposures depending on the actual distributions of each of the model input
parameters.

There is also uncertainty as to whether the model equations generate results that represent actual
workplace air concentrations. Some activity-based modeling does not account for exposures from other
activities. Another uncertainty is lack of consideration for engineering controls. The GS/ESDs assume
that all activities occur without any engineering controls or PPE, and in an open-system environment
where vapor and particulates freely escape and can be inhaled. Actual exposures may be less than
estimated depending on engineering control and PPE use.

A strength of the assessment is the variation of the model input parameters as opposed to using a single
static value. This parameter variation increases the likelihood of true occupational inhalation exposures
falling within the range of modeled estimates. An additional strength is that all data that EPA used to
inform the modeling parameter distributions have overall data quality determinations of either high or
medium from EPA's systematic review process.

3.1.2.4.4	Modeled Dermal Exposures

The Dermal Exposure to Volatile Liquids Model used to estimate dermal exposure to 1,4-dioxane in
occupational settings assumes a fixed fractional absorption of the applied dose; however, fractional
absorption may be dependent on skin loading conditions. The model also assumes a single exposure
event per day based on existing framework of the EPA/OPPT 2-HandDermal Exposure to Liquids
Model and does not address variability in exposure duration and frequency. Additionally, dermal
exposures to 1,4-dioxane vapor that may penetrate clothing and the potential for associated direct skin
contact with clothing saturated with 1,4-dioxane vapor are not included in quantifying exposures, which
could potentially result in underestimates of exposures. Although the extent of saturation of clothing
with 1,4-dioxane vapors is unknown, it is expected to be minimal given the low concentrations of 1,4-
dioxane in formulations for the conditions of use in the draft supplement.

A strength of the dermal assessment approach is the estimation of two different fractional absorption
values specific to industrial and commercial use settings as opposed to applying only one fractional
absorption value to both settings.

Page 100 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3109	3.2 General Population Exposures

3110

1,4-Dioxane - General Population Exposures (Section 3.2):

Key Points

•	EPA estimated oral and inhalation exposures to the concentrations of 1,4-dioxane in drinking
water and air estimated in Section 2.3 using equations and exposure factors described in
Appendix G.l and Appendix J.2.

•	EPA estimated general population exposures to 1,4-dioxane in air and water with a particular
focus on populations that may be highly exposed

0 Fenceline communities. For exposures through air, EPA considered potential

exposures for communities within 10 km of a release site. For drinking water, EPA
considered potential exposures for communities relying on drinking water collected
downstream of release sites.

0 Lifestage. For drinking water, EPA evaluated lifestage-specific exposures for adults,
formula-fed infants, and children. For air exposures, the impacts of lifestage
differences were not able to be adequately quantified and so the air concentrations are
used for all lifestages.

0 High-end exposure estimates. EPA evaluated exposures based on high-end exposure
scenarios (e.g., air exposures include a range of modeled concentration predictions
[low-end, central tendency, and high-end]), although only high-end model predictions
of air concentrations are presented in this section).

3111

3112	General population exposures occur when 1,4-dioxane is released into the environment and the media is

3113	then a pathway for exposure. Figure 3-1 below provides a graphic representation of where and in which

3114	media 1,4-dioxane may be found and the corresponding route of exposure.

3115

Page 101 of 484


-------
3116

3117

3118

3119

3120

3121

3122

3123

3124

3125

3126

3127

3128

3129

3130

3131

3132

3133

3134

3135

3136

3137

3138

3139

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Ambient Air

Inhalation

Landfills
(Industrial or
Muncipal)

Drinking
Water

Oral

Drinking

Water
Treatment

Wastewater
Facility

>undw£

Water
Recreation
Oral, Dermal

Figure 3-1. Potential Human Exposure Pathways to 1,4-Dioxane for the General Population"

" The diagram presents the media (white text boxes) and routes of exposure (italics for oral, inhalation, or dermal)
for the general population. Sources of drinking water from surface or water pipes is depicted w ith grey arrows.

3.2.1 Approach and Methodology

EPA used modeled concentrations of 1,4-dioxane in air, surface water, and groundwater estimated in
Section 2.3 to estimate acute and chronic general population exposures that could result from contact
with environmental media. These acute and chronic exposure estimates are used to evaluate cancer and
non-cancer risk described in Section 5. To estimate oral exposures to 1,4-dioxane in drinking water
(including groundwater used as drinking water), EPA used equations and exposure factors described in
Appendix G.l. To estimate inhalation exposures from 1,4-dioxane in air, EPA used equations and
exposure factors described in Appendix J.2. Where possible, available monitored data within these
environmental media were used to provide context for modeled results.

To estimate potential acute and chronic exposures through drinking water EPA calculated Acute Dose
Rates (ADR) and Average Daily Doses (ADD) for adults, formula-fed infants, and children. To estimate
lifetime exposures through drinking water, EPA calculated a Lifetime Average Daily Dose (LADD)
based on 33 years1' of exposure starting from birth or 33 years of exposure as an adult, averaged over a
78-year lifetime. Assumptions about drinking water intake and body weight for each age group were
based on information in the Exposure Factors Handbook. EPA calculated ADs, ADDs, and LADDs
based on the drinking water concentrations estimated under a range of conditions in Section 2.3.1.3.
Details of these calculations are presented in Appendix I and 1,4-Dioxane Supplemental Information
File: Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Release to Surface Water fr om
Individual Facilities (U.S. EPA. 2023h).

Thirty-three years is the 95th percentile residential occupancy period (U.S. EPA Exposure Factors Handbook. Chapter 16,
Table 16-5).

Page 102 of 484


-------
3140

3141

3142

3143

3144

3145

3146

3147

3148

3149

3150

3151

3152

3153

3154

3155

3156

3157

3158

3159

3160

3161

3162

3163

3164

3165

3166

3167

3168

3169

3170

3171

3172

3173

3174

3175

3176

3177

3178

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

To estimate potential acute and chronic exposures through air, EPA calculated ADs and ADCs based on
modeled air concentrations described in Section 2.3.3 To estimate potential lifetime exposures, EPA
calculated LADCs based on 33 years of exposure. Methods adequate to quantify the impact of lifestage
differences on 1,4-dioxane exposure are not available (see Section 4.3) and air concentration is used as
the exposure metric for all lifestages per EPA guidance (I v << \	i Mb). Specific equations,

inputs and assumptions are described in detail in Appendix IJ.2.

3.2.2 Drinking Water Exposure Assessment	

EPA assessed general population drinking water exposures that could result from surface water or
groundwater used as drinking water. Exposures estimates presented below are based on surface water
concentrations modeled in Section 2.3.1 or groundwater concentrations modeled in Section 2.3.2.
Exposure estimates presented throughout this section focus on adults and formula-fed infants because
these are lifestages with the greatest drinking water intake relative to body weights and therefore the
greatest potential exposures.

3.2.2.1 Surface Water Exposure Assessment

EPA evaluated acute, chronic, and lifetime general population, exposures to 1,4-dioxane that may be
present in drinking water based on modeled surface water concentrations estimated in Section 2.3.1. For
drinking water exposures in this assessment, EPA focused on exposures in fenceline communities,
defined in this context as members of the general population who rely on drinking water from water
bodies receiving 1,4-dioxane releases from any industrial or DTD source.

Drinking water exposures were evaluated using a series of parallel analyses that provide information
about the individual contributions of specific COUs as well as information about aggregate exposures
that could result from multiple sources releasing to the same water body.

3.2.2.1.1 Exposures from Individual Facility Releases

To evaluate the individual contributions of releases associated with specific industrial and commercial
COUs to general population exposures, EPA calculated ADRs, ADDs and LADDs based on modeled
water concentrations estimated in Section 2.3.1.3.1 (Table 3-3). A total of 125 release scenarios were
evaluated based on water concentrations estimated for annual releases that occur over a single day (a
peak exposure scenario), over 30 days of release, or over 250 to 365 days of release. Exposure estimates
are presented for both adults and formula-fed infants because these are lifestages with greatest drinking
water intake relative to body weights and therefore greatest exposures. ADRs based on a single day
release scenario range from 6.Ox 10~8 to 3,730 mg/kg for adults and 2.1 /10 1 to 1.3x 104 mg/kg for
infants. ADDs range from 1.1/10 " to 0.5 mg/kg/day for adults and 2,7/10 " to 1.3 mg/kg/day for
infants and are not influenced by the days of release. LADDS range from 4.2x 10~12 to 0.2 mg/kg/day for
adults exposed for 33 years and 3,5x10 13 to 1,6x 10 2 mg/kg/day for infants exposed for 1 year.
Complete exposure calculations are available in 1,4-Dioxane Supplemental Information File: Drinking
Water Exposure and Risk Estimates for 1,4-Dioxane Release to Surface Water from Individual Facilities
(U.S. EPA. 2023h\

Page 103 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Table 3-3. Adult and Infant Exposures Estimated from Facility-Specific Releases

OES

No. of
Releases
Modeled

Age Group'1

ADR (mg/kg-dav)

ADD (mg/kg-dav)

LADD (mg/kg-day)

Min
Exposure'

Mean
Exposure''

Max
Exposure1'

Min
Exposure'

Mean
Exposure''

Max
Exposure1'

Min
Exposure'

Mean
Exposure''

Max
Exposure1'

Disposal

25

Adult (21+
years)

6.03E-07

2.60E01

3.83E02

4.51E-10

1.44E-02

2.21E-01

1.91E—10

6.08E-03

9.36E-02

Infant (birth to
<1 year)

2.11E-06

9.11E01

1.34E03

1.15E-09

3.67E-02

5.65E-01

1.48E-11

4.71E-04

7.25E-03

Ethoxylation
Byproduct

8

Adult (21+
years)

2.17E-07

1.04E02

8.31E02

9.06E-11

3.66E-02

2.93E-01

3.83E-11

1.55E-02

1.24E-01

Infant (birth to
<1 year)

7.61E-07

3.65E02

2.92E03

2.31E-10

9.36E-02

7.48E-01

2.97E-12

1.20E-03

9.59E-03

Functional
Fluids (Open-
System)

6

Adult (21+
years)

5.59E-04

6.33E-02

1.92E-01

1.83E-07

2.23E-05

6.66E-05

7.73E-08

9.44E-06

2.82E-05

Infant (birth to
<1 year)

1.96E-03

2.22E-01

6.75E-01

4.67E-07

5.70E-05

1.70E-04

5.99E-09

7.30E-07

2.18E-06

Import and
Repackaging

12

Adult (21+
years)

4.35E-04

3.28E02

3.73E03

1.32E-07

3.05E-02

2.23E-01

5.59E-08

1.29E-02

9.43E-02

Infant (birth to
<1 year)

1.53E-03

1.15E03

1.31E04

3.38E-07

7.78E-02

5.69E-01

4.33E-09

9.97E-04

7.30E-03

Industrial

Uses

31

Adult (21+
years)

5.34E-07

2.05E01

1.87E02

1.97E-10

1.36E-02

1.55E-01

8.3 IE—11

5.77E-03

6.56E-02

Infant (birth to
<1 year)

1.87E-06

7.21E01

6.55E02

5.02E-10

3.48E-02

3.96E-01

6.44E-12

4.47E-04

5.08E-03

Manufacture

2

Adult (21+
years)

3.35E00

6.56E01

1.28E02

2.50E-03

4.91E-02

9.57E-02

1.06E-03

2.08E-02

4.05E-02

Infant (birth to
<1 year)

1.17E01

2.30E02

4.48E02

6.40E-03

1.25E-01

2.44E-01

8.20E-05

1.61E-03

3.13E-03

PET

Manufacturing

19

Adult (21+
years)

1.11 E—04

4.32E01

6.67E02

3.86E-08

3.16E-02

5.00E-01

1.63E-08

1.34E-02

2.11 E—01

Infant (birth to
<1 year)

3.91E-04

1.52E02

2.34E03

9.86E-08

8.07E-02

1.28E00

1.26E-09

1.04E-03

1.64E-02

Printing Inks

1

Adult (21+
years)

8.26E-02

8.26E-02

8.26E-02

6.18E-05

6.18E-05

6.18E-05

2.62E-05

2.62E-05

2.62E-05

Infant (birth to
<1 year)

2.90E-01

2.90E-01

2.90E-01

1.58E-04

1.58E-04

1.58E-04

2.02E-06

2.02E-06

2.02E-06

Remediation

16

Adult (21+
years)

6.04E-08

7.36E-02

7.19E-01

1.07E-11

4.56E-05

4.12E-04

4.51E-12

1.93E-05

1.74E-04

Infant (birth to
<1 year)

2.12E-07

2.58E-01

2.52E00

2.72E-11

1.17E-04

1.05E-03

3.49E-13

1.49E-06

1.35E-05

Page 104 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

No. of
Releases
Modeled

Age Group*

ADR (mg/kg-dav)

ADD (mg/kg-dav)

LADD (mg/kg-day)

Min
Exposure1'

Mean
Exposure''

Max
Exposure1'

Min
Exposure1'

Mean
Exposure''

Max
Exposure1'

Min
Exposure1'

Mean
Exposure''

Max
Exposure"

Overall

120

Adult (21+
years)

6.04E-08

5.84E01

3.73E03

1.07E-11

1.78E-02

5.00E-01

4.51E-12

7.55E-03

2.11E-01

Infant (birth to
<1 year)

2.12E-07

2.05E02

1.31E04

2.72E-11

4.56E-02

1.28E00

3.49E-13

5.84E-04

1.64E-02

Adult LADDs presented in this table were used to derive cancer risk estimates presented in Figure 5-1, Figure 5-2 and Figure 5-4.

ADRs presented here are calculated based on the assumption that all releases could occur on a single day of release (peak exposure scenario); ADDs and LADDs are based on
chronic exposure scenarios and are the same regardless of the number of days of release assumed. LADDs for adults are based on 33 years of exposure averaged over a 78-year
lifetime while LADDs for infant-specific exposures are based on 1 year of exposure averaged over a 78-year lifetime.

"Adult refers to 21+years; infant refers to birth to <1 year.

4 These COUs are added since the 2020 RE was published.

c The minimum exposure for the identified days of release, within the identified OES, and for the identified age group.
dThe arithmetic mean exposure for the identified days of release, within the identified OES, and for the identified age group.
e The maximum exposure for the identified days of release, within the identified OES, and for the identified age group.

3180

Page 105 of 484


-------
3181

3182

3183

3184

3185

3186

3187

3188

3189

3190

3191

3192

3193

3194

3195

3196

3197

3198

3199

3200

3201

3202

3203

3204

3205

3206

3207

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3.2.2.1.2 Exposures from Down-the-Drain Releases

To evaluate the potential contribution of DTD consumer and commercial releases, EPA calculated
ADRs, ADDs, and LADDs using modeled water concentrations estimated as described in Section
2.3.1.2.2. Water concentrations of 1,4-dioxane resulting from DTD releases depend on the population
size (an indicator of the number of people using products and contributing the releases) and the stream
flows of the receiving water bodies. Therefore, the adult LADDs presented in Table 3-4 are based on the
range of water concentrations estimated by Monte Carlo modeling of DTD release scenarios with
varying population size and stream flows. LADDs range from 1.7/10 10 to 5,1 / ] 0 4 mg/kg/day for
adults exposed for 33 years. Complete exposure calculations for adults and infants are available in 1,4-
Dioxane Supplemental Information File: Drinking Water Exposure and Risk Estimates for 1,4-Dioxane
Surface Water Concentrations Predicted with Probabilistic Modeling (U.S. EPA. 20230.

Table 3-4. Adult LADD Exposures (mg/kg/day) Estimated from 1,4-Dioxane DTD Consumer and
Commercial Releases



Population Contributing to DTD Releases

100

1,000

10,000

100,000

1,000,000

Stream Flow
(cfs)

100

5.1E-08

5.1E-07

5.1E-06

5.1E-05

5.1E-04

300

1.7E-08

1.7E-07

1.7E-06

1.7E-05

1.7E-04

1,000

5.1E-09

5.1E-08

5.1E-07

5.1E-06

5.1E-05

3,000

1.7E-09

1.7E-08

1.7E-07

1.7E-06

1.7E-05

10,000

5.1E-10

5.1E-09

5.1E-08

5.1E-07

5.1E-06

30,000

1.7E-10

1.7E-09

1.7E-08

1.7E-07

1.7E-06

The frequencies of each of these combinations of population size and flow rate are presented Table 2-10. Adult
LADDs presented in this table were used to derive the cancer risk estimates presented in Table 5-4.

LADDs for adults are based on 33 years of exposure averaged over a 78-year lifetime.

3.2.2.1.3 Disposal of Hydraulic Fracturing Produced Waters

To evaluate the potential contribution of disposal of hydraulic fracturing produced waters to surface
water, EPA calculated ADRs, ADDs, and LADDs using the range of modeled water concentrations
estimated in Section 2.3.1.2.2. (Table 3-5). Water concentrations of 1,4-dioxane resulting from disposal
of hydraulic fracturing produced water vary substantially across sites. The estimated exposures
presented here are based on the range of water concentrations estimated by Monte Carlo modeling for a
range of site-specific factors. For this range of estimated surface water concentrations, Adult ADRs
range from 2.45xl0~16 to 7.96xl0~3 mg/kg and adult ADDs range from 6.69xl0~17 to 2.18><10~3
mg/kg/day. LADDs for adults exposed over 33 years over a 78-year lifetime range from 2,8x10 17 to
9.2xl0~4 mg/kg/day. Complete exposure calculations for adults and infants are available in 1,4-Dioxane
Supplemental Information File: Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Surface
Water Concentrations Predicted with Probabilistic Modeling (U.S. EPA. 2023i).

Page 106 of 484


-------
3208

3209

3210

3211

3212

3213

3214

3215

3216

3217

3218

3219

3220

3221

3222

3223

3224

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Table 3-5. Adult ADR, ADD, and LADD Exposures Estimated from Disposal of Hydraulic
Fracturing Produced Waters to Surface Water		

Monte Carlo
Distribution

Adult Acute Dose
Rate (mg/kg)

Adult Average Daily
Dose (mg/kg/day)

Adult Lifetime Average
Daily Dose (mg/kg/day)

Maximum

7.96E-03

2.18E-03

9.2E-04

99th percentile

3.13E-04

8.56E-05

3.6E-05

95th percentile

1.05E-04

2.86E-05

1.2E-05

Median

2.57E-06

7.03E-07

3.0E-07

5th percentile

9.31E-09

2.54E-09

1.1E-09

Minimum

2.45E-16

6.69E-17

2.8E-17

Adult LADDs presented in this table were used to derive cancer risk estimates presented in Table 5-5.

3.2.2.1.4 Aggregate Exposure

Because multiple sources of 1,4-dioxane contribute to surface water and drinking water concentrations,
EPA also estimated aggregate general population exposures that could occur because of combined
contributions from DTD releases from consumer and commercial uses, upstream sources, and direct and
indirect industrial releases. EPA calculated ADRs, ADDs, and LADDs based on modeled water
concentrations estimated in Section 2.3.1.3.4 using probabilistic modeling of aggregate 1,4-dioxane
surface water concentrations that could occur downstream of industrial release sites for each COU.
LADDs estimated for adults exposed over 33 years over a 78-year lifetime range from 8.07x 10~7 to
7.4x10 3 mg/kg/day based on median modeled water concentrations across COUs (Table 3-6). Complete
exposure calculations for adults and infants are available in 1,4-Dioxane Supplemental Information File:
Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Surface Water Concentrations Predicted
with Probabilistic Modeling (	Z0231).

Table 3-6. Adult LADD Exposures from Aggregate Concentrations Estimated Downstream of
Release Sites (Including DTD Releases and Direct and Indirect Industrial Releases)	

COU

LADDs (mg/kg/day) Based on Modeled Aggregate Surface Water Concentrations
Estimated across the Monte Carlo Distribution

Min

5th
Percentile

25th
Percentile

Median

75th
Percentile

95th
Percentile

Max

Disposal

6.00E-09

6.82E-07

1.63E-06

3.93E-06

9.00E-06

9.64E-04

1.21E-01

Ethoxylation
byproduct

7.17E-09

3.72E-07

1.05E-06

1.93E-06

8.11E-06

1.98E-02

2.63E-01

Functional
Fluids (Open-
System)

5.62E-10

2.58E-07

4.99E-07

8.91E-07

7.20E-06

4.13E-05

6.22E-05

Import and
Repackaging

1.82E-08

4.21E-07

2.10E-04

1.03E-03

4.53E-03

1.34E-02

1.18E00

Industrial Uses

4.61E-10

1.65E-07

3.90E-07

8.07E-07

4.66E-05

4.90E-03

5.91E-02

Page 107 of 484


-------
3225

3226

3227

3228

3229

3230

3231

3232

3233

3234

3235

3236

3237

3238

3239

3240

3241

3242

3243

3244

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

cou

LADDs (mg/kg/day) Based on Modeled Aggregate Surface Water Concentrations
Estimated across the Monte Carlo Distribution

Min

5th
Percentile

25th
Percentile

Median

75th
Percentile

95th
Percentile

Max

Manufacture

3.51E-04

1.06E-03

2.29E-03

7.40E-03

1.75E-02

4.04E-02

4.04E-02

PET

Manufacturing

1.73E-08

4.54E-07

1.48E-06

1.43E-05

6.06E-04

2.53E-02

2.11E-01

Printing Inks

4.80E-07

1.01E-06

1.27E-05

2.04E-05

2.38E-05

2.66E-05

2.71E-05

Remediation

2.48E-09

2.74E-07

6.29E-07

1.27E-06

3.10E-05

9.61E-05

2.29E-04

Adult LADDs presented in this table were used to derive cancer risk estimates presented in Figure 5-5.
Percentiles reflect concentrations estimated at various points in the Monte Carlo distribution.

3.2.2.2 Groundwater Exposure Assessment

EPA evaluated general population exposures that could occur from disposals of 1,4-dioxane that
contaminate groundwater used as a primary source of drinking water. To estimate chronic exposures
through this drinking water pathway, EPA calculated ADDs and LADDs for adults and formula-fed
infants based on modeled groundwater concentrations of 1,4-dioxane estimated in Section 2.3.2. EPA
did not evaluate acute exposures because methods used to estimate groundwater concentrations provide
an indication of potential concentrations occuring over many years, rather than peak concentrations.

3.2.2.2.1 Disposal to Landfills

To evaluate general population exposure, EPA calculated ADDs and LADDs based on modeled
groundwater concentrations estimated in Section 2.3.2.3. Potential groundwater concentrations resulting
from disposal of 1,4-dioxane to municipal solid waste landfills vary across landfill loading rates and
concentrations of 1,4-dioxane in leachate. Estimated exposures presented here are therefore based on the
range of groundwater concentrations estimated under varying landfill conditions. Table 3-7 summarizes
LADD exposure estimates estimated for 33 years of exposure as an adult. Under the range of landfill
scenarios considered, adult LADDs range from 2.5><106 to 2,4/ 10 2 mg/kg/day. The highest LADDs
occur when leachate concentrations are above 100 mg/L and loading rates are above 10,000 lb. The
complete set of exposure estimates for adults and infants relying on groundwater as a primary drinking
water source are presented in 1,4-Dioxane Supplemental Information File: Drinking Water Exposure
and Risk Estimates for 1,4-Dioxane Land Releases to Landfills (	)23f).

Page 108 of 484


-------
3245

3246

3247

3248

3249

3250

3251

3252

3253

3254

3255

3256

3257

3258

3259

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Table 3-7. Adult LADD Exposures Estimated from Groundwater Contamination from Landfills
under Varying Landfill Conditions



Loading Rate (lb)

Leachate
Concentration
(mg/L)

0.1

1

10

100

1,000

10,000

100,000

1,000,000

0.0001

2.5E-16

2.4E-15

3.0E-14

2.9E-13

2.7E-12

2.6E-11

2.5E-10

2.4E-09

0.001

2.5E-15

2.4E-14

3.0E-13

2.9E-12

2.7E-11

2.6E-10

2.5E-09

2.4E-08

0.01

2.5E-14

2.4E-13

3.0E-12

2.9E-11

2.7E-10

2.6E-09

2.5E-08

2.4E-07

0.1

2.5E-13

2.4E-12

3.0E-11

2.9E-10

2.7E-09

2.6E-08

2.5E-07

2.4E-06

1

2.5E-12

2.4E-11

3.0E-10

2.9E-09

2.7E-08

2.6E-07

2.5E-06

2.4E-05

10

2.5E-11

2.4E-10

3.0E-09

2.9E-08

2.7E-07

2.6E-06

2.5E-05

2.4E-04

100

2.5E-10

2.4E-09

3.0E-08

2.9E-07

2.7E-06

2.6E-05

2.5E-04

2.4E-03

1,000

2.5E-09

2.4E-08

3.0E-07

2.9E-06

2.7E-05

2.6E-04

2.5E-03

2.4E-02

10,000

2.5E-08

2.4E-07

3.0E-06

2.9E-05

2.7E-04

2.6E-03

2.5E-02

2.4E-01

Adult LADDs presented in this table were used to derive cancer risk estimates presented in Table 5-6.

3.2.2.2.2 Disposal of Hydraulic Fracturing Produced Waters

To evaluate general population exposure resulting from disposal of hydraulic fracturing produced waters
to groundwater, EPA calculated ADDs and LADDs estimated in Section 2.3.2.4 (Table 3-8). Potential
groundwater concentrations resulting from disposal of hydraulic fracturing produced waters vary
substantially across sites. Estimated exposures presented here are based on the range of groundwater
concentrations estimated through Monte Carlo modeling. Under the range of hydraulic fracturing
scenarios considered, adult LADDs range from 4,9/ 10 9 to 2,1 /10 4 mg/kg/day. The complete set of
exposure estimates for adults and infants relying on groundwater as a primary drinking water source are
presented in 1,4-Dioxcme Supplemental Information File: Drinking Water Exposure and Risk Estimates
for 1,4-Dioxane Land Releases to Surface Impoundments (U.S. EPA. 2023 g).

Table 3-8. Estimated Exposures Resulting from Groundwater Contamination from Disposal of

Monte Carlo
Distribution

Modeled
Groundwater
Concentration (mg/L)

Adult ADD

(mg/kg/day)

Adult LADD

(mg/kg/day)

Infant ADD

(mg/kg/day)

Max

1.9E-02

2.1E-04

8.8E-05

5.3E-04

99th

1.5E-02

1.7E-04

7.1E-05

4.3E-04

95th

1.5E-02

1.7E-04

7.1E-05

4.3E-04

Mean

7.1E-04

7.9E-06

3.3E-06

2.0E-05

50th

1.2E-04

1.3E-06

5.6E-07

3.4E-06

5th

1.2E-04

1.3E-06

5.6E-07

3.4E-06

Min

4.4E-07

4.9E-09

2.1E-09

1.2E-08

Adult LADDs presented in this table were used to derive cancer risk estimates presented in Table 5-7.

Page 109 of 484


-------
3260

3261

3262

3263

3264

3265

3266

3267

3268

3269

3270

3271

3272

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3,2,3 Air Exposure Assessment

EPA evaluated acute, chronic and lifetime general population, exposures to 1,4-dioxane in air. This
analysis focuses on potential fenceline community exposures that may occur within 10 km of release
sites.

3.2.3.1 Industrial COUs Reported to TRI

To evaluate general population exposures from industrial fugitive and stack emissions, EPA calculated
ACs, ADCs, and LADCs based on modeled air concentrations estimated in Section 2.3.3. The LADCs
presented in Table 3-9 are based on the maximum 95th percentile air concentrations estimated for the
facilities within each COU. LADCs within 10 km of release types considered here range from 1.1x10 "
to 6.9x10 3 ppm. These lifetime exposure estimates are based on 33 years of exposure over a 78-year
lifetime and are relevant to all lifestages. The complete set of inhalation exposure estimates for fenceline
communities are presented in 1,4-Dioxane Supplemental Information File: Air Exposures and Risk
Estimates for Single Year Analysis (	023 e).

Page 110 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3273 Table 3-9. Lifetime Average Dai y Concentrations Estimated within 10 km of 1,4-Dioxane Releases to Air

1,4-Dioxane OES

#

Facilities

LADCs Estimated within 5-10,000 m of Facilities with Greatest Exposures (ppm)

5 m

III in

30 m

60 m

Kill in

100 to
1,000 m

2,500 m

5,000 m

10,000 m

Disposal

15

1.8E-03

2.1E-03

7.6E-04

2.9E-04

1.3E-04

1.3E-05

8.0E-07

2.7E-07

8.8E-08

Dry Film Lubricant

8

6.8E-11

3.0E-09

2.2E-07

1.6E-06

2.7E-06

4.2E-07

1.2E-08

3.6E-09

1.6E-09

Ethoxylation
Byproduct

6

2.8E-03

5.8E-03

3.1E-03

1.3E-03

6.9E-04

1.6E-04

9.3E-06

3.8E-06

1.5E-06

Film Cement

1

5.3E-05

5.5E-05

1.9E-05

9.7E-06

5.3E-06

9.7E-07

5.8E-08

2.0E-08

6.4E-09

Functional Fluids
(Open-System)

2

5.4E-06

1.0E-05

4.4E-06

4.6E-06

7.7E-06

3.1E-06

2.9E-07

1.1E-07

3.7E-08

Import and
Repackaging

1

1.1E-11

2.4E-10

2.3E-08

1.8E-07

3.7E-07

1.4E-07

2.8E-08

1.7E-08

9.4E-09

Industrial Uses

12

1.8E-03

2.0E-03

6.5E-04

2.4E-04

1.2E-04

3.0E-05

3.7E-06

1.4E-06

4.8E-07

Laboratory Chemical
Use

1

8.7E-04

9.1E-04

3.1E-04

1.6E-04

8.7E-05

1.6E-05

9.6E-07

3.2E-07

1.1E-07

Manufacturing

1

3.7E-03

6.9E-03

3.3E-03

1.4E-03

6.7E-04

6.0E-05

3.5E-06

1.1E-06

3.4E-07

PET Manufacturing

13

3.4E-03

4.0E-03

1.5E-03

5.9E-04

2.7E-04

4.5E-05

8.8E-06

5.3E-06

2.8E-06

Spray Foam
Application

1

3.3E-07

3.6E-07

1.2E-07

6.4E-08

3.6E-08

6.6E-09

7.3E-10

2.7E-10

1.0E-10

LADCs are based on the maximum 95th percentile air concentration predictions for the facility in each COU with the greatest exposures. Adult LADCs
presented in this table were used to derive the cancer risk estimates presented in Table 5-8.

3274

3275

Page 111 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3276	3.2.3.2 Hydraulic Fracturing

3277	To evaluate general population exposures to fugitive emissions from hydraulic fracturing operations,

3278	EPA calculated ACs, ADCs, and LADCs based on modeled air concentrations estimated in Section

3279	2.3.3.2.4 under a range of different release scenarios and topographical conditions (Table 3-10). LADCs

3280	within 1,000 m of hydraulic fracturing operations range from 8.7/10 4 to 5.2 ppm. These lifetime

3281	exposure estimates are based on 33 years of exposure over a 78-year lifetime and are relevant to all

3282	lifestages. The complete set of inhalation exposure estimates from fugitive emissions of hydraulic

3283	fracturing operations are presented in 1,4-Dioxane Supplemental Information File: Air Exposure and

3284	Risk Estimates for 1,4-Dioxane Emissions from Hydraulic Fracturing Operations (U.S. EPA. 2023b).

Page 112 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3285 Table 3-10. Exposures from Fugitive Emissions Estimated within 1,000 in of Hydraulic Fracturing Operations"





Air Concentrations for 95th Percentile Modeled Releases

Air Concentrations for 50th Percentile Modeled Releases

Fugitive







(ppm)









(ppm)































Emissions

Exposure

High-End Modeled Air

Central Tendency (Mean)

High-End Modeled Air

Central Tendency (Mean)

Release

Duration

Concentrations

Modeled Air Concentrations

Concentrations

Modeled Air Concentrations

Scenario



100 m

1,000 m

100 to
1,000 m

100 m

1,000 m

100 to
1,000 m

100 m

1,000 m

100 to
1,000 m

100 m

1,000 m

100 to
1,000 m



AC

5.3E-02

2.4E-03

6.2E-03

4.2E-02

1.8E-03

4.8E-03

3.1E-03

1.4E-04

3.6E-04

2.4E-03

1.0E-04

2.7E-04

South





















































(Coastal)-

ADC

1.1E-02

4.8E-04

1.2E-03

8.3E-03

3.6E-04

9.4E-04

6.0E-04

2.7E-05

7.0E-05

4.7E-04

2.0E-05

5.4E-05

Rural-24



























LADC



























4.5E-03

2.0E-04

5.2E-04

3.5E-03

1.5E-04

4.0E-04

2.6E-04

1.2E-05

3.0E-05

2.0E-04

8.6E-06

2.3E-05



AC

4.1E-02

2.4E-03

5.5E-03

3.1E-02

1.5E-03

3.7E-03

2.3E-03

1.3E-04

3.1E-04

1.8E-03

8.3E-05

2.1E-04

West North





















































Central-

ADC

8.0E-03

4.6E-04

1.1E-03

6.1E-03

2.9E-04

7.3E-04

4.6E-04

2.7E-05

6.2E-05

3.5E-04

1.6E-05

4.2E-05

Rural-24



























LADC



























3.4E-03

2.0E-04

4.6E-04

2.6E-03

1.2E-04

3.1E-04

1.9E-04

1.1E-05

2.6E-05

1.5E-04

6.9E-06

1.8E-05



AC

2.6E-02

5.8E-04

1.8E-03

2.3E-02

5.0E-04

1.6E-03

1.5E-03

3.3E-05

1.0E-04

1.3E-03

2.9E-05

9.0E-05

South





















































(Coastal)-

ADC

5.1E-03

1.1E-04

3.5E-04

4.5E-03

9.9E-05

3.1E-04

2.9E-04

6.5E-06

2.0E-05

2.6E-04

5.7E-06

1.8E-05

Urban-24



























LADC



























2.2E-03

4.8E-05

1.5E-04

1.9E-03

4.2E-05

1.3E-04

1.2E-04

2.8E-06

8.5E-06

1.1E-04

2.4E-06

7.5E-06



AC

2.4E-02

6.2E-04

1.9E-03

1.9E-02

4.6E-04

1.4E-03

1.4E-03

3.6E-05

1.1E-04

1.1E-03

2.6E-05

8.1E-05

West North





















































Central-

ADC

4.8E-03

1.2E-04

3.7E-04

3.8E-03

9.0E-05

2.8E-04

2.7E-04

7.0E-06

2.1E-05

2.2E-04

5.2E-06

1.6E-05

Urban-24



























LADC



























2.0E-03

5.2E-05

1.6E-04

1.6E-03

3.8E-05

1.2E-04

1.2E-04

3.0E-06

8.9E-06

9.3E-05

2.2E-06

6.7E-06



AC

9.6E-03

8.8E-05

3.8E-04

8.3E-03

6.8E-05

3.2E-04

5.5E-04

5.1E-06

2.2E-05

4.8E-04

3.9E-06

1.8E-05

South





















































(Coastal)-

ADC

1.9E-03

1.7E-05

7.6E-05

1.6E-03

1.3E-05

6.2E-05

1.1E-04

1.0E-06

4.3E-06

9.4E-05

7.6E-07

3.6E-06

Rural-8



























LADC



























8.0E-04

7.4E-06

3.2E-05

7.0E-04

5.6E-06

2.6E-05

4.6E-05

4.2E-07

1.8E-06

4.0E-05

3.2E-07

1.5E-06



AC

2.0E-02

7.7E-04

1.9E-03

1.1E-02

2.1E-04

6.6E-04

6.1E-04

1.2E-05

3.8E-05

6.1E-04

1.2E-05

3.8E-05

West North





















































Central-

ADC

4.0E-03

1.5E-04

3.7E-04

2.1E-03

4.2E-05

1.3E-04

1.2E-04

2.4E-06

7.4E-06

1.2E-04

2.4E-06

7.4E-06

Rural-8



























LADC



























1.7E-03

6.4E-05

1.5E-04

8.9E-04

1.8E-05

5.5E-05

5.1E-05

1.0E-06

3.1E-06

5.1E-05

1.0E-06

3.1E-06

Page 113 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Fugitive
Emissions
Release
Scenario

Exposure
Duration

Air Concentrations for 95th Percentile Modeled Releases

(ppm)

Air Concentrations for 50th Percentile Modeled Releases

(ppm)

High-End Modeled Air
Concentrations

Central Tendency (Mean)
Modeled Air Concentrations

High-End Modeled Air
Concentrations

Central Tendency (Mean)
Modeled Air Concentrations

100 m

1,000 m

100 to
1,000 m

100 m

1,000 m

100 to
1,000 m

100 m

1,000 m

100 to
1,000 m

100 m

1,000 m

100 to
1,000 m

South

(Coastal)-

Urban-8

AC

8.7E-03

6.8E-05

3.2E-04

8.0E-03

6.0E-05

2.9E-04

4.6E-04

3.4E-06

1.7E-05

4.6E-04

3.4E-06

1.7E-05

ADC

1.7E-03

1.3E-05

6.4E-05

1.6E-03

1.2E-05

5.8E-05

9.0E-05

6.8E-07

3.3E-06

9.0E-05

6.8E-07

3.3E-06

LADC

7.3E-04

5.6E-06

2.7E-05

6.7E-04

5.0E-06

2.4E-05

3.8E-05

2.9E-07

1.4E-06

3.8E-05

2.9E-07

1.4E-06

West North

Central-

Urban-8

AC

1.5E-02

2.9E-04

9.4E-04

9.2E-03

1.2E-04

4.5E-04

5.3E-04

6.8E-06

2.6E-05

5.3E-04

6.8E-06

2.6E-05

ADC

2.9E-03

5.8E-05

1.8E-04

1.8E-03

2.4E-05

8.9E-05

1.0E-04

1.4E-06

5.1E-06

1.0E-04

1.4E-06

5.1E-06

LADC

1.2E-03

2.4E-05

7.8E-05

7.7E-04

1.0E-05

3.8E-05

4.4E-05

5.7E-07

2.2E-06

4.4E-05

5.7E-07

2.2E-06

" Lifetime Average Daily Concentrations (LADCs) presented in this table correspond to the cancer risk estimates presented in Table 5-7.
AC = Acute Concentration; ADC = Average Daily Concentration; LADC = Lifetime Average Daily Concentration

3286

Page 114 of 484


-------
3287

3288

3289

3290

3291

3292

3293

3294

3295

3296

3297

3298

3299

3300

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3.2.3.3 Industrial and Institutional Laundry Facilities

To evaluate exposures to emissions from industrial and institutional laundry facilities, EPA calculated
ACs, ADCs, and LADCs based on vapor and particular air concentrations estimated in Section 2.3.3.2.4.
High-end and central tendency air exposures estimated under the more conservative exposure scenario
evaluated (rural south coastal topography, assuming 24 hours of releases each day) are presented for
each type of laundry in Table 3-11. LADCs estimated within 1,000 m of laundry facilities operations
range from 8.7><10~4 to 2.4><10H5 ppm. These lifetime exposure estimates are based on 33 years of
exposure over a 78-year lifetime and are relevant to all lifestages. The complete set of inhalation
exposure estimates from fugitive emissions of commercial laundry facilities are presented in 1,4-
Dioxane Supplemental Information File: Air Exposures and Risk Estimates for Industrial Laundry (U.S.
I 23c).

Table 3-11. Exposures from Fugitive Emissions Estimated near Industrial and Institutional
Laundry Facilities" 		

Facility
Type

Detergent

and
Emission
Type

Exposure
Duration

Modeled Air Concentrations for Maximum Release Estimates

(ppm)

High-End

Central Tendency (Mean)

100 m

1,000 m

100 to
1,000 m

100 m

1,000 m

100 to
1,000 m

Industrial
laundry

Liquid-
Vapor

AC

6.9E-06

3.3E-07

8.2E-07

5.0E-06

2.1E-07

5.6E-07

ADC

5.4E-06

2.5E-07

6.3E-07

4.9E-06

2.1E-07

5.6E-07

LADC

2.3E-06

1.1E-07

2.7E-07

2.1E-06

9.0E-08

2.4E-07

Powder-
Vapor

AC

6.9E-06

3.3E-07

8.2E-07

4.9E-06

2.1E-07

5.6E-07

ADC

5.4E-06

2.5E-07

6.2E-07

4.9E-06

2.1E-07

5.6E-07

LADC

2.3E-06

1.1E-07

2.6E-07

2.1E-06

8.9E-08

2.3E-07

Powder-
PM10

AC

7.2E-06

1.6E-07

5.9E-07

5.1E-06

1.2E-07

4.2E-07

ADC

5.6E-06

1.3E-07

4.7E-07

5.0E-06

1.2E-07

4.2E-07

LADC

2.4E-06

5.5E-08

2.0E-07

2.1E-06

4.9E-08

1.8E-07

Powder-
PM2.5

AC

6.9E-06

3.1E-07

8.0E-07

4.9E-06

2.0E-07

5.5E-07

ADC

5.4E-06

2.4E-07

6.1E-07

4.9E-06

2.0E-07

5.4E-07

LADC

2.3E-06

1.0E-07

2.6E-07

2.1E-06

8.5E-08

2.3E-07

Institutional
laundry

Liquid-
Vapor

AC

3.6E-06

1.6E-07

4.1E-07

3.1E-06

1.3E-07

3.5E-07

ADC

3.4E-06

1.6E-07

4.0E-07

3.1E-06

1.3E-07

3.5E-07

LADC

1.4E-06

6.7E-08

1.7E-07

1.3E-06

5.7E-08

1.5E-07

Powder-
vapor

AC

1.1E-07

4.8E-09

1.2E-08

9.2E-08

4.0E-09

1.0E-08

ADC

1.0E-07

4.7E-09

1.2E-08

9.2E-08

3.9E-09

1.0E-08

LADC

4.2E-08

2.0E-09

4.9E-09

3.9E-08

1.7E-09

4.4E-09

Page 115 of 484


-------
3301

3302

3303

3304

3305

3306

3307

3308

3309

3310

3311

3312

3313

3314

3315

3316

3317

3318

3319

3320

3321

3322

3323

3324

3325

3326

3327

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023



Detergent



Modeled Air Concentrations for Maximum Release Estimates

(ppm)

Facility
Type

and
Emission

Exposure
Duration

High-End

Central Tendency (Mean)

Type

100 m

1,000 m

100 to
1,000 m

100 m

1,000 m

100 to
1,000 m





AC

1.1E-07

2.5E-09

8.9E-09

9.4E-08

2.2E-09

7.9E-09



Powder-
PM10

ADC

1.0E-07

2.4E-09

8.7E-09

9.4E-08

2.2E-09

7.9E-09





LADC

4.4E-08

1.0E-09

3.7E-09

4.0E-08

9.2E-10

3.3E-09





AC

1.1E-07

4.6E-09

1.2E-08

9.2E-08

3.8E-09

1.0E-08



Powder-
PM2.5

ADC

1.0E-07

4.5E-09

1.1E-08

9.2E-08

3.8E-09

1.0E-08





LADC

4.2E-08

1.9E-09

4.8E-09

3.9E-08

1.6E-09

4.3E-09

" LADCs presented in this table correspond to the cancer risk estimates presented in Table 5-10.

AC = Acute Concentration; ADC = Average Daily Concentration; LADC = Lifetime Average Daily Concentration

3.3 Weight of the Scientific Evidence Conclusions

As described in the 2021 Draft Systematic Review Protocol (	), the weight of the

scientific evidence supporting exposure assessments is evaluated based on the availability and strength
of exposure scenarios and exposure factors, measured and monitored data, estimation methodology and
model input data, and, if appropriate, comparisons of estimated and measured exposures. The strength of
each of these evidence streams can be ranked as either robust, moderate, slight, or indeterminate. For
each component of this exposure assessment, EPA evaluated the weight of the scientific evidence for
individual evidence streams and then used that information to evaluate the overall weight of evidence
supporting each set of exposure estimates. General considerations for evaluating the strength of evidence
for each evidence stream are summarized in TableApx C-5. Specific examples of how these
considerations are applied to overall weight of evidence conclusions are provided in Table Apx C-6.

Overall confidence descriptions of high, medium, or low are assigned to the exposure assessment based
on the strength of the underlying scientific evidence. When the assessment is supported by robust
evidence, overall confidence in the exposure assessment is high; when supported by moderate evidence,
overall confidence is medium; when supported by slight evidence, overall confidence is low.

3.3.1 Occupational Exposures

The weight of the scientific evidence for occupational exposure estimates is determined by several
different evidence streams, including:

•	Evidence supporting the exposure scenarios (Section 3.1.1 and Appendix F.4)

•	The quality and representativeness of available monitoring data (Appendix F.4)

•	Evidence supporting modeling approaches (Section 3.1.1 and Appendix F.4)

•	Evidence supporting model input data (Appendix F.4)

3.3.1.1 Inhalation Exposure

Occupational inhalation exposure estimates are supported by moderate to robust evidence (see Appendix
F.6).

•	Exposure Scenarios and Exposure Factors. The exposure scenarios and exposure factors
underlying the inhalation assessment are supported by moderate to robust evidence.

Page 116 of 484


-------
3328

3329

3330

3331

3332

3333

3334

3335

3336

3337

3338

3339

3340

3341

3342

3343

3344

3345

3346

3347

3348

3349

3350

3351

3352

3353

3354

3355

3356

3357

3358

3359

3360

3361

3362

3363

3364

3365

3366

3367

3368

3369

3370

3371

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Occupational inhalation exposure scenarios and exposure factors, including duration of exposure,
body weight, and breathing rate, were informed by sources of data with medium to high data
quality ratings, increasing the strength of evidence. For most OESs/COUs, EPA used
information directly relevant to the evaluated exposure scenarios; however, for some
OESs/COUs, EPA used information from surrogate scenarios, decreasing the strength of
evidence for those scenarios. Additionally, there is uncertainty in the extent to which the entire
population of workers within an OES/COU are represented by the available data.

•	Measured and Monitored Data. Measured/monitored data are supported by moderate to robust
evidence. EPA used sources of data such as OSHA and NIOSH, which have medium to high data
quality ratings, increasing the strength of the evidence. For the OESs/COUs with available
monitoring data, the data was directly applicable to the assessed exposure scenario, as opposed to
from a surrogate exposure scenario. However, the available monitoring data was limited to a
single source for each OES/COU and often consisted of a small or dated dataset. Additionally,
these data often only included one or a limited number of sites at which the data were measured,
decreasing the strength of evidence for those OESs/COUs.

•	Modeling Methodologies. The modeling methodologies are supported by moderate to robust
evidence. Modeling was implemented to assess occupational inhalation exposures for three of
the OESs/COUs, using methodologies from GS/ESD that are generally well described. The
modeling incorporates Monte Carlo simulation to allow for variation in the model input data,
which increases the representativeness of the approach towards the true population of potentially
exposed workers and increases the strength of the evidence. However, EPA was unable to
develop distributions for all input parameters, increasing the uncertainty in the parameterization
and applicability.

•	Model Input Data. Model input data are supported by moderate to robust evidence. For some
model input data, EPA used 1,4-dioxane-specific data from sources such as process information,
product concentration information, and FracFocus 3.0. For other model input parameters, generic
data from the GS/ESD used for the modeling methodology was used due to lack of 1,4-dioxane
data.

•	Comparison of Modeled and Monitored Data. The comparison of modeled and measured
occupational inhalation exposures is not rated because no comparisons between modeled and
measured exposures were made.

Overall Confidence in Occupational Inhalation Exposure Estimates

The overall confidence in the occupational inhalation exposure estimates (Section 3.3.1.1) ranges from
low to high, depending on the OES/COU. Measured/monitored data are supported by moderate to robust
evidence. Additionally, the modeling methodologies and underlying model input data is supported by
moderate to robust evidence. However, there is uncertainty in the representativeness of the assessed
exposure scenarios towards all potential exposures for the given OES/COU, limitations in the amount
and age of monitoring data, and limitations in the modeling approaches towards 1,4-dioxane-specific use
within the OES/COU. Therefore, while the underlying data and methods used to estimate occupational
inhalation exposures is supported by moderate to robust evidence, the overall confidence of these
estimates is low to high depending on the OES/COU. OES/COU-specific discussions of the available
inhalation exposure data and overall confidence are presented in Appendix F.6.

3.3.1.2 Dermal Exposure

Occupational dermal exposure estimates are supported by slight to robust evidence (see Appendix F.3).

Page 117 of 484


-------
3372

3373

3374

3375

3376

3377

3378

3379

3380

3381

3382

3383

3384

3385

3386

3387

3388

3389

3390

3391

3392

3393

3394

3395

3396

3397

3398

3399

3400

3401

3402

3403

3404

3405

3406

3407

3408

3409

3410

3411

3412

3413

3414

3415

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

•	Exposure Scenarios and Exposure Factors. The exposure scenarios and exposure factors
underlying the dermal assessment are supported by moderate to robust evidence. Dermal
exposure scenarios were informed by process information and GS/ESD with medium to high
data quality ratings, increasing the strength of evidence. Exposure factors, including amount of
material on skin, surface area of skin exposed, and absorption of 1,4-dioxane through the skin,
were informed by literature sources, the ChemSTEER User Guide (x v ii \ i) for
standard exposure parameters, and a European model, which have medium to high data quality
ratings. EPA used information directly relevant to the evaluated exposure scenarios; however,
there is uncertainty in the extent to which the entire population of workers within an OES/COU
are represented by the available data.

•	Measured and Monitored Data. No measured/monitored dermal exposure data were used in
the occupational dermal exposure assessment. EPA did use measured data on 1,4-dioxane
concentrations in various products from process information and other literature sources, which
have medium to high data quality ratings, depending on the data source.

•	Modeling Methodologies. The modeling methodologies are supported by moderate evidence.
EPA used the EPA Dermal Exposure to Volatile Liquids to calculate the dermal retained dose for
each OES/COU. This model modifies the EPA/OPPT 2-HandDermal Exposure to Liquids
Model by incorporating a "fraction absorbed (fabs)" parameter to account for the evaporation of
volatile chemicals and a "protection factor (PF)" to account for glove use. These modifications
improve the modeling methodology and allow EPA to differentiate dermal exposures between
commercial and industrial settings by varying the absorption and dermal protection factors.
However, the modeling approach is still limited by the low variability for different worker
activities/exposure scenarios.

•	Model Input Data. Model input data are supported by slight to moderate evidence. As discussed
above, model parameters were informed by sources with medium to high data quality ratings.
However, a limitation is that some of the model input data was generic and not specific to 1,4-
dioxane OES/COU.

•	Comparison of Modeled and Monitored Data. All occupational dermal exposures were
modeled, and no measured dermal exposures were used in this assessment, therefore there is no
comparison.

Overall Confidence in Occupational Dermal Exposure Estimates

The overall confidence in the occupational dermal exposure estimates (Section 3.1.2.2) is medium for all
OES/COU because the same modeling approach was used for all OES/COU. The modeling
methodology is supported by moderate evidence, with model input parameters from literature sources, a
European model, standard defaults from the ChemSTEER User Guide (	i), and 1,4-

dioxane product concentration data from process information. These sources range from slight to robust,
depending on factors such as age and applicability to OES/COU. The modeling is limited by the use of
standard input parameters that are not specific to 1,4-dioxane and a lack of variability in dermal
exposure for different worker activities. Therefore, EPA's overall confidence in the occupational dermal
exposure estimates is medium.

3.3.2 Drinking Water

3.3.2.1 Drinking Water Exposure Estimates Based on Surface Water Concentrations

The weight of evidence for drinking water exposure estimates is determined by several different
evidence streams, including:

Page 118 of 484


-------
3416

3417

3418

3419

3420

3421

3422

3423

3424

3425

3426

3427

3428

3429

3430

3431

3432

3433

3434

3435

3436

3437

3438

3439

3440

3441

3442

3443

3444

3445

3446

3447

3448

3449

3450

3451

3452

3453

3454

3455

3456

3457

3458

3459

3460

3461

3462

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

•	Evidence supporting the general population exposure scenarios (Section 3.2.1)

•	The quality and representativeness of available surface water and drinking water monitoring
data (Section 2.3.1.1)

•	Evidence supporting modeling approaches (Section 2.3.1.3 and Appendix G.2)

•	Evidence supporting release data used as model input data (Section 2.2 and Appendix E.3)

•	Concordance between modeled and monitored water concentrations (Section 2.3.1.4)

As described in Section 2.3.1, multiple approaches were used to predict surface water concentrations
resulting from several sources. These included the evaluation of facility-specific releases, down the
drain releases to surface water, hydraulic fracturing releases and aggregation of surface water releases.
The associated strengths, limitations and confidence in these estimated environmental concentrations are
described in Section 2.3.1.4. The general population drinking water exposure scenarios and exposure
factors used to estimate exposures that could result from estimated water concentrations are described in
Section 3.2.

Drinking water exposure estimates based on modeled surface water concentrations are supported by
overall moderate to robust evidence, with the strength of the evidence varying across analysis
approaches and COUs/OESs.

•	Exposure Scenarios and Exposure Factors. The exposure scenarios and exposure factors
underlying all drinking water exposure estimates are supported by moderate to robust evidence.
Exposure factors for drinking water are based on robust data on drinking water intake rates and
body weights as derived from exposure factors from the EPA's Exposure Factors Handbook

(	2011). However, the drinking water exposure scenarios generally rely on the

assumption that little or no dilution occurs prior to drinking water intakes. That assumption may
not be representative of exposures at all locations. While there are locations where this
assumption is expected to the accurate, the extent of downstream dilution that occurs prior to
drinking water intakes is highly variable across locations. The proximity of facility releases to
actual drinking water intakes is evaluated in Section 2.3.1.2.4 and Appendix G.2.4. Uncertainties
related to downstream dilution decrease the overall strength of evidence for these exposure
scenarios. However, EPA has performed several analyses that calculate exposures and risks
under alternate assumptions about downstream dilution and illustrate the quantitative impact of
those assumptions (see Section 5.2.2.1.2), increasing the overall strength of evidence. Drinking
water exposure scenarios also rely on the data-driven assumption that 1,4-dioxane is not removed
through treatment. Moderate to robust data provide support for this assumption under many
treatment scenarios. These assumptions may over-estimate exposure for some locations, but
provide an overall distribution that is generally expected to be representative of exposure
scenarios.

•	Measured and Monitored Data. The measured/monitored data are supported by moderate
evidence. The high number of monitoring data points for surface water and drinking water from
high quality sources in multiple locations over multiple years increases the strength of the
evidence from monitoring data. Monitoring data confirm that 1,4-dioxane is present in some
surface water and drinking water in some locations. However, evidence from monitoring data
may not be representative of all sites where 1,4-dioxane is released to surface water from TSCA
sources, decreasing the strength of evidence from monitoring data. The lack of temporal and/or
spatial concurrence between most monitoring data and reported release locations makes direct
comparison challenging for most locations. However, a limited number of sites with monitoring
data are co-located with sites where 1,4-dioxane releases are reported, supporting comparisons of
monitoring and modeled estimates that increase the overall strength of the evidence. In addition,

Page 119 of 484


-------
3463

3464

3465

3466

3467

3468

3469

3470

3471

3472

3473

3474

3475

3476

3477

3478

3479

3480

3481

3482

3483

3484

3485

3486

3487

3488

3489

3490

3491

3492

3493

3494

3495

3496

3497

3498

3499

3500

3501

3502

3503

3504

3505

3506

3507

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

as described in Section 2.3.1.4, monitoring data for surface water directly downstream from
releases show concentrations multiple orders of magnitude greater than typical ambient surface
water concentrations, aligning with patterns of modeled results.

•	Modeling Methodologies. The modeling methodologies are supported by moderate to robust
evidence.

o The methodology for deriving exposure estimates for facility releases is moderate and is
applicable to the populations included in the exposure scenarios. This approach makes
some conservative assumptions about flow rates and release frequency and amount.
Additionally, the modeling does not take into account downstream fate or transport, but
the physical chemical properties of 1,4-dioxane are expected to moderate the impact
these influences could have on the modeled instream concentrations. The model is
designed to estimate possible higher end water concentrations expected at specific
locations.

o The probabilistic methodology used for deriving exposure estimates for DTD releases,
hydraulic fracturing releases, and aggregate releases from all sources is robust. This
approach incorporates the full distribution of facility releases over multiple years and
corresponding instream flow rate data rather than relying on the most conservative model
inputs. It is designed to provide a nationally representative distribution of estimated water
concentrations under varying conditions.

•	Model Input Data. Model input data are supported by slight to robust evidence, with the
strength of the evidence varying across individual COUs/OESs. The strength of evidence
supporting modeled water concentrations relies heavily on the quality of the facility or OES-
specific release data used as inputs for the model, including both the amount of release, location
of the release, and the corresponding flow in the receiving water body. A summary of sources of
flow and release data for facility release modeling is presented in Table 2-6. A more detailed
OES-specific discussion of the confidence in sources of release information is presented in
Appendix E.3.4.

o For overall distributions of industrial releases across sites, model input data are supported
by robust evidence. As illustrated in Section 5.2.2.1.2, EPA estimated exposures and risks
across the full distribution of facility releases both for the whole dataset and for a subset
of facilities with high quality reporting information. Comparison of these distributions
demonstrates that inclusion of locations relying on more limited release information had
limited impact on the overall distributions of exposures.

o For COUs/OESs that rely primarily on release data reported to TRI via Form R, or
reported to ICIS-NPDES via DMR, site-specific release estimates are supported by
moderate to robust evidence. As described in Appendix E.3.1, these release estimates are
based on release amounts reported by facilities. Most COUs/OESs are included in this
group.

o For COUs/OESs that rely primarily on release data reported to TRI via Form A, site-
specific release estimates are supported by moderate evidence. As described in Appendix
E.3.1, Form A simply indicates that releases are below the reporting thresholds and
specific release estimates require assumptions about amounts, locations, and media of
release. The Import and Repackaging OES releases used in this analysis are entirely
based on Form A reporting of releases, and just under half of the Industrial Uses OES
releases were reported via Form A.

Page 120 of 484


-------
3508

3509

3510

3511

3512

3513

3514

3515

3516

3517

3518

3519

3520

3521

3522

3523

3524

3525

3526

3527

3528

3529

3530

3531

3532

3533

3534

3535

3536

3537

3538

3539

3540

3541

3542

3543

3544

3545

3546

3547

3548

3549

3550

3551

3552

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

o For COUs/OESs that rely primarily on other sources of release information or generic
scenarios, site-specific release estimates are supported by slight to moderate evidence.
For these scenarios, EPA estimated daily wastewater discharges by using various
modeling approaches, including the use of surrogate TRI and DMR data and modeling
using data from literature, GSs, and ESDs.

¦	For DTD sources, release information is supported by slight to moderate
evidence. Although confidence in the individual contribution from some specific
COUs (ie specific consumer or commercial product categories) is lower,
confidence in estimates of overall DTD releases is moderate. The presented model
is intended to inform the total contribution of DTD releases to overall aggregate
instream concentration as well as providing evidence of individual COUs that
may be most influential. Presented results should be taken in relation to one
another qualitatively rather than discrete quantitative values. Distributions of
DTD releases of consumer and commercial products were estimated for each
COU on a per capita basis using the SHEDS-HT model.

¦	For hydraulic fracturing releases, release information is supported by moderate
evidence. Releases were estimated using Monte Carlo modeling with information
from the Draft ESD on Hydraulic Fracturing and FracFocus 3.0.

•	Comparison of Modeled and Monitored Data. Comparisons of estimated and measured
exposures provide moderate evidence. Because most of the available monitoring data are not
collected in locations with known releases temporally or spatially, it is difficult to make direct
comparisons in most locations. However, in case study locations where monitoring data re
located near release sites, comparisons demonstrate that there is concordance between measured
and/or reported and modeled estimates (Section 2.3.1.4), increasing the overall strength of the
evidence. Monitoring data confirm that 1,4-dioxane is present in some surface water and
drinking water. Uncertainty as to whether trends observed in case study locations are
representative of all of the sites decreases overall confidence in these comparisons.

Overall Confidence in Exposure Estimates

Overall confidence in drinking water exposure estimates for surface water concentrations modeled from
facility releases (Section 3.2.2.1.1) is high across the overall distribution, particularly when limited to
sites with high quality sources of release data. For individual facilities and COUs, overall confidence in
exposure estimates varies depending on the confidence in source-specific release data. The modeling
methodology used for this analysis is supported by moderate evidence. This approach makes some
conservative assumptions about flow rates and release frequency and amount. It is designed to estimate
water concentrations expected at specific locations. Available monitoring data confirm that 1,4-dioxane
is present in some surface water and drinking water, though most of the available data were not collected
near release sites are therefore not directly comparable. The overall level of confidence in OES/COU-
specific exposure estimates depends on the source of OES/COU-specific release data described in
Appendix E.3:

•	Overall confidence in drinking water exposure estimates is medium to high for OESs/COUs that
rely primarily on site-specific release data reported to DMR or to TRI via Form R.

•	Overall confidence in site-specific drinking water exposure estimates is medium for OESs/COUs
for which site-specific release estimates are based on reporting to TRI via Form A

•	Overall confidence in drinking water exposure estimates is low to medium for OESs/COUs for
which site-specific release estimates are based on surrogate or modeled information.

Page 121 of 484


-------
3553

3554

3555

3556

3557

3558

3559

3560

3561

3562

3563

3564

3565

3566

3567

3568

3569

3570

3571

3572

3573

3574

3575

3576

3577

3578

3579

3580

3581

3582

3583

3584

3585

3586

3587

3588

3589

3590

3591

3592

3593

3594

3595

3596

3597

3598

3599

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Overall confidence in drinking water exposure estimates for DTD releases under varying conditions
(Section 3.2.2.1.2), is medium. The modeling methodology used for this analysis is supported by robust
evidence. This analysis is designed to provide a nationally representative distribution of estimated water
concentrations under varying conditions. This analysis defines the conditions under which exposures are
higher, but is not designed to predict the specific levels of exposure resulting from DTD releases at
specific locations with precision. Exposure estimates rely on estimated distributions of DTD releases of
specific consumer and commercial products categories associated with each COU. Distributions of DTD
releases of consumer and commercial products were estimated for each COU on a per capita basis using
the SHEDS-HT model. While confidence in the individual contribution from some specific COUs is
lower, confidence in estimates of overall DTD releases is moderate.

Overall confidence in drinking water exposure estimates for hydraulic fracturing releases (Section
3.2.2.1.3) is medium. The modeling methodology used for this analysis is supported by robust evidence
and is designed to provide a nationally representative distribution of estimated water concentrations
under varying conditions. Releases used as inputs in the model were estimated using Monte Carlo
modeling that captures variability across sites. However, the modeled exposure estimates are not directly
tied to specific releases at known locations, decreasing the strength of the evidence related to the
representativeness of the exposure estimates for actual exposures.

Overall confidence in drinking water exposure estimates for aggregate surface water concentrations
predicted by probabilistic modeling (Section 3.2.2.1.4) is high across the overall distribution. For
individual facilities and COUs, overall confidence in exposure estimates varies depending on the
confidence in source-specific release data. The modeling methodology used for this analysis is
supported by robust evidence and is designed to provide a nationally representative distribution of
estimated water concentrations under varying conditions. The estimated drinking water concentrations
modeled in this analysis incorporate contributions from direct and indirect industrial releases, DTD
releases, and other upstream sources. Available monitoring data confirm that 1,4-dioxane is present in
some surface water and drinking water, though most of the available data were not collected near release
sites and are therefore not directly comparable. The overall level of confidence in resulting exposure
estimates depends on the source of OES/COU-specific release data described in Appendix E.3:

•	Overall confidence in drinking water exposure estimates is medium to high for OESs/COUs that
rely primarily on release data reported to DMR or to TRI via Form R. Most COUs/OESs are
included in this group

•	Overall confidence in drinking water exposure estimates is medium for OESs/COUs for which
release estimates are based on reporting to TRI via Form A. The Import and Repackaging OES
releases used in this analysis are entirely based on Form A reporting of releases, and just under
half of the Industrial Uses OES releases were reported via Form A.

•	Overall confidence in drinking water exposure estimates is low to medium for OESs/COUs for
which release estimates are based on surrogate or modeled information.

3.3.2.2 Drinking Water Exposure Estimates Based on Groundwater Concentrations

The weight of evidence for exposure estimates presented in this section is determined by several
different evidence streams, including:

•	Evidence supporting the exposure scenarios (Section 3.2.1)

•	The quality and representativeness of available groundwater monitoring data (Section 2.3.2.1),

•	Evidence supporting modeling approaches and input data (Sections 2.3.2.3.1 and 2.3.2.4.1)

•	Evidence supporting release data used as model input data (Section 2.2 and Appendix E.4)

•	Concordance between modeled and monitored water concentrations.

Page 122 of 484


-------
3600

3601

3602

3603

3604

3605

3606

3607

3608

3609

3610

3611

3612

3613

3614

3615

3616

3617

3618

3619

3620

3621

3622

3623

3624

3625

3626

3627

3628

3629

3630

3631

3632

3633

3634

3635

3636

3637

3638

3639

3640

3641

3642

3643

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3.3.2.2.1 Groundwater Concentrations Resulting from Disposal to Landfill

Drinking water exposure estimates based on groundwater concentrations modeled for landfill disposal
scenarios are supported by overall slight to moderate evidence.

•	Exposure Scenarios and Exposure Factors. The exposure scenarios and exposure factors
underlying these drinking water exposure estimates are supported by slight to moderate
evidence. Exposure factors for drinking water are based on robust data on drinking water intakes,
body weight, and other standard exposure factors from the EPA's Exposure Factors Handbook
(\ v < < \ 1 I). However, the drinking water exposure scenario relies on the assumption that
the groundwater concentrations estimated with the DRAS model may occur in locations where
groundwater is used as a primary drinking water source. While there is uncertainty around this
assumption, this scenario represents a sentinel exposure scenario

•	Measured and Monitored Data. Measured/monitored data are supported by moderate evidence.
Monitoring data were available to sufficiently cover most or all of the population groups
included within the exposure scenarios but there are a limited number of studies to corroborate
findings. Since little data is readily available on the concentration of 1,4-dioxane near or around
landfills in groundwater, some caution is required when interpreting monitoring data as it may
not be fully representative of conditions around all landfills.

•	Modeling Methodology. The modeling methodology is supported by robust evidence. The
DRAS methodology for deriving the estimate is well described. The underlying computational or
scientific basis is robust and has an empirical basis considering chemical specific properties.

•	Model Input Data. The release data relied on as a model input is supported by slight evidence.
Model inputs for the DRAS model include chemical properties of 1,4-dioxane that are well-
defined and reviewed and therefore supported by robust evidence. However, model inputs for
leachate concentrations and loading rates are more uncertain. EPA does not have reasonably
available information on actual concentrations of 1,4-dioxane in leachate for most landfills and
therefore selected landfill leachate concentrations are based on potential for risk to human health.
Loading rates are based on the range reported in TRI for RCRA subtitle C landfills and therefore
may not be representative of nonhazardous landfills evaluated in this analysis. These
uncertainties around landfill leachate concentrations and loading rates decrease the strength of
the evidence for model input data.

•	Comparison of Modeled and Monitored Data. Comparison of estimated and measured
exposures provides moderate evidence because monitoring data confirm the presence of 1,4-
dioxane in groundwater in some locations and modeled estimates and measured exposure values
are comparable, however differences in methodology, collection, or context make it difficult to
arrive at full concordance.

Overall Confidence in Exposure Estimates

Overall confidence in drinking water exposure estimates resulting from disposal to landfills (Section
3.2.2.2.1) is low to medium. The modeling methodology is robust. However, the release information
relied on as model input data is supported by slight to moderate evidence, decreasing overall confidence.
In addition, this drinking water exposure scenario relies on the assumption that the groundwater
concentrations estimated with the DRAS model may occur in locations where groundwater is used as a
primary drinking water source. While the substantial uncertainty around the extent to which these
exposures occur decreases overall confidence in the exposure scenario, this scenario represents a
sentinel exposure.

Page 123 of 484


-------
3644

3645

3646

3647

3648

3649

3650

3651

3652

3653

3654

3655

3656

3657

3658

3659

3660

3661

3662

3663

3664

3665

3666

3667

3668

3669

3670

3671

3672

3673

3674

3675

3676

3677

3678

3679

3680

3681

3682

3683

3684

3685

3686

3687

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3.3.2.2.2 Groundwater Concentrations Resulting from Disposal of Hydraulic
Fracturing Waste

Drinking water exposure estimates based on modeled groundwater concentrations estimated under a
range of hydraulic fracturing waste disposal scenarios are supported by slight to moderate evidence.

•	Exposure Scenarios and Exposure Factors. The exposure scenario factors underlying these
exposure estimates are supported by slight to moderate evidence. Exposure factors for drinking
water are based on robust data on drinking water intakes, body weight, and other standard
exposure factors from the Exposure Factors Handbook. However, the drinking water exposure
scenario relies on the assumption that the estimated groundwater concentrations may occur in
locations where groundwater is used as a primary drinking water source.

•	Measured and Monitored Data. The measured/monitored data are supported by indeterminate
evidence. Available groundwater monitoring data are not located near hydraulic fracturing
operations and do not provide information about the potential for hydraulic fracturing operations
to contribute to groundwater contamination.

•	Modeling Methodologies. The modeling methodology and input data are supported by robust
evidence. The methodology for deriving the estimate is well described, the underlying
computational or scientific basis is robust, and has an empirical basis considering chemical
specific properties.

•	Model Input Data. Hydraulic fracturing releases are supported by moderate evidence. As
described in Appendix E.4.4, releases were estimated using Monte Carlo modeling with
information from the Draft ESD on Hydraulic Fracturing and FracFocus 3.0. DRAS modeling
was based on very limited data on concentrations of 1,4-dioxane in produced water as reported in
the literature. Reliance on limited data and uncertainty around the representativeness of that data
decrease the strength of the evidence for model input data.

•	Comparison of Modeled and Monitored Data. The comparison of estimated and measured
exposures is not rated because no comparisons between estimated and measured exposures were
made.

Overall Confidence in Exposure Estimates

Overall confidence in drinking water exposure estimates resulting from disposal of hydraulic fracturing
waste (Section 3.2.2.2.2) is low to medium. The modeling methodology is robust and the release
information relied on as model input data is supported by moderate evidence. However, no monitoring
data are available to confirm detection of 1,4-dioxane in groundwater near hydraulic fracturing
operations. This drinking water exposure scenario relies on the assumption that the estimated
groundwater concentrations may occur in locations where groundwater is used as a primary drinking
water source. While the substantial uncertainty around the extent to which these exposures occur
decreases overall confidence in the exposure scenario, this scenario represents a sentinel exposure.

3,3.3 Air	

The weight of the scientific evidence for exposure estimates presented in this section is determined by
several different evidence streams, including:

•	Evidence supporting the exposure scenarios (Section 3.2.1)

•	The quality and representativeness of available groundwater monitoring data (Section 2.3.3.1),

•	Evidence supporting modeling approaches and input data (Section 2.3.3.2)

•	Evidence supporting release data used as model input data (Section 2.3.3.2)

•	Concordance between modeled and monitored water concentrations.

Page 124 of 484


-------
3688

3689

3690

3691

3692

3693

3694

3695

3696

3697

3698

3699

3700

3701

3702

3703

3704

3705

3706

3707

3708

3709

3710

3711

3712

3713

3714

3715

3716

3717

3718

3719

3720

3721

3722

3723

3724

3725

3726

3727

3728

3729

3730

3731

3732

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

As described in Section 2.3.3, 1,4-dioxane concentrations in air were estimated for areas around
industrial COUs reported to TRI, hydraulic fracturing operations, and institutional and industrial laundry
facilities. The associated strengths and limitations of these estimated environmental concentrations are
described in Section 2.3.3.3. The general population air exposure scenarios and exposure factors used to
estimate exposures are described in Section 3.2.3.

3.3.3.1 Modeled Air Concentrations for Industrial COUs Reported to TRI

Inhalation exposure estimates resulting from 1,4-dioxane releases for industrial COUs reported to TRI
are supported by overall moderate evidence.

•	Exposure Scenarios and Exposure Factors. Exposure scenarios underlying these exposure
estimates are supported by moderate evidence. The exposure factors used to build the exposure
scenarios are directly relevant to general population exposures for communities living in close
proximity to releasing facilities. While the long-term exposure scenarios are most directly
relevant for individuals who reside in fenceline communities for many years, these scenarios are
expected to be within the range of normal habits and exposure patterns expected in the general
population. However, there is uncertainty around the extent to which people actually live and
work around the specific facilities where exposures are highest, decreasing the overall strength of
evidence for these exposure scenarios, particularly at the distances nearest to facilities.

•	Measured and Monitored Data. The measured/monitored data are supported by indeterminate
evidence. No measured or monitored data were available.

•	Modeling Methodologies. The modeling methodology used to estimate exposure concentrations
via the ambient air pathway is supported by robust evidence. Air concentrations were estimated
using AERMOD and IIOAC. AERMOD is EPA's regulatory model and has been thoroughly
peer reviewed; therefore, the general confidence in results from the model is high but reliant on
the integrity and quality of the inputs used and interpretation of the results. Confidence in
modeled air concentrations resulting from stack releases is lower at distances less than 100 m of
release sites, but confidence in modeled concentrations for fugitive emissions is higher near
release sites. While this is a source of uncertainty, air concentrations from fugitive emissions
tend to peak within 10m of release sites while stack releases were found to peak around 100 m,
indicating that air concentrations modeled at distances less than 100 m of release sites are
generally driven by fugitive emissions. IIOAC is an Excel-based model with results based on
pre-run AERMOD exposure scenarios under a variety of environmental and release conditions.
There is a moderate to high confidence in air concentrations estimated using IIOAC because,
although IIOAC results are based on pre-run AERMOD exposure scenarios (high confidence),
some key sources of uncertainty identified in Section 2.3.3.3 (like limited set of distances
evaluated (100, 100 to 1,000, and 1,000 m) and assumptions made about meteorological
conditions necessary to provide a more conservative exposure estimate) can lead to a slightly
lower confidence (moderate).

•	Model Input Data. Model input data on air releases are supported by slight to robust evidence,
with the strength of the evidence varying across COUs/OESs. A more detailed OES-specific
discussion of the confidence in sources of release information is presented in Appendix E.5.4.

o For COUs/OESs that rely primarily on release data reported to TRI via Form R, site-
specific release estimates are supported by moderate to robust evidence. As described in
Appendix E.5.4, these release estimates are based on specific release amounts and other
source-specific information reported by facilities as a regulatory requirement.

Page 125 of 484


-------
3733

3734

3735

3736

3737

3738

3739

3740

3741

3742

3743

3744

3745

3746

3747

3748

3749

3750

3751

3752

3753

3754

3755

3756

3757

3758

3759

3760

3761

3762

3763

3764

3765

3766

3767

3768

3769

3770

3771

3772

3773

3774

3775

3776

3777

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

o For COUs/OESs that rely primarily on release data reported to TRI via Form A, site-
specific release estimates are supported by moderate evidence. As described in Appendix
E.5.4, Form A simply indicates that releases are below the reporting thresholds and
specific release estimates require assumptions about exact amounts and locations of
releases.

o For COUs/OESs that rely primarily on other sources of release information or generic
scenarios, release estimates are supported by evidence ranging from slight to moderate
evidence. For these scenarios, EPA estimated daily and annual air releases using various
modeling approaches, including the use of surrogate TRI data and modeling using data
from literature, GSs, and ESDs.

•	Comparison of Modeled and Monitored Data. Comparison of estimated and measured
exposures provide indeterminate evidence. No measured or monitored data were available for
comparison.

Overall Confidence in Exposure Estimates

Overall confidence in inhalation exposure estimates resulting for air concentrations modeled based on
industrial releases (Section 3.2.3.1) varies across COUs. The AERMOD modeling methodology used for
this analysis is robust and considers contributions from both stack and fugitive emissions. The exposure
scenarios considered are most relevant to long-term residents in fenceline communities. There is
uncertainty around the extent to which people live and work in the specific locations where exposures
are highest, decreasing confidence in the exposure scenarios, particularly at distances nearest to
facilities. Overall confidence varies due to variable levels of confidence in underlying release
information used to the support the analysis:

•	Overall confidence in site-specific inhalation exposure estimates is medium to high for
OESs/COUs that rely primarily on release data reported to TRI via Form R.

•	Overall confidence in site-specific inhalation exposure estimates is medium for OESs/COUs for
which release estimates are based on data reported to TRI via Form A.

•	Overall confidence in inhalation exposure estimates is low to medium for OESs/COUs for which
release estimates are based on surrogate or modeled information.

3.3.3.2 Air Concentrations Modeled near Hydraulic Fracturing Operations and
Industrial/Institutional Laundries

Inhalation exposure estimates resulting from 1,4-dioxane released to air from hydraulic fracturing
operations and industrial/institutional laundries are supported by overall moderate evidence.

•	Exposure Scenarios and Exposure Factors. Exposure scenarios underlying these exposure
estimates are supported by moderate evidence. The factors used to build the exposure scenarios
are directly relevant to general population exposures for communities living in close proximity to
releasing facilities. While the long-term exposure scenarios are most directly relevant for
individuals who reside in fenceline communities for many years, these scenarios are expected to
be within the range of normal habits and exposure patterns expected in the general population.
However, there is some uncertainty around the extent to which people actually live and work
around the specific locations where exposures are highest, decreasing the overall strength of
evidence for these exposure scenarios.

•	Measured and Monitored Data. The measured/monitored data are supported by indeterminate
evidence. No measured or monitored data were available.

•	Modeling Methodologies. The modeling methodology used to estimate exposure concentrations
via the ambient air pathway is supported by robust evidence. Air concentrations were estimated

Page 126 of 484


-------
3778

3779

3780

3781

3782

3783

3784

3785

3786

3787

3788

3789

3790

3791

3792

3793

3794

3795

3796

3797

3798

3799

3800

3801

3802

3803

3804

3805

3806

3807

3808

3809

3810

3811

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

using IIOAC. IIOAC is an Excel-based model with results based on pre-run AERMOD exposure
scenarios under a variety of environmental and release conditions. There is a moderate to high
confidence in air concentrations estimated using IIOAC because, although IIOAC results are
based on pre-run AERMOD exposure scenarios (high confidence), some key sources of
uncertainty identified in Section 2.3.3.3 (like limited set of distances evaluated (100, 100 to
1,000, and 1,000 m) and assumptions made about meteorological conditions necessary to provide
a more conservative exposure estimate) can lead to a slightly lower confidence (moderate).

•	Model Input Data. Input data used for modeling exposures from hydraulic fracturing operations
and industrial/institutional laundries are supported by moderate evidence. As described in
Appendix E.5.4, these modeled exposure estimates are based on alternative release estimates and
scenario conditions found in the literature and derived with Monte Carlo models of release
estimate, some of which have been peer reviewed, others which may not be peer reviewed. Since
the modeled exposures are based on alternative release estimates, which in turn are based on
modeled data and outputs, there is a lower overall confidence in the modeled exposures from
such input data. Additionally, exposure estimates using this input data requires certain
assumptions which can lead to a lower overall confidence in the estimated exposure
concentrations.

•	Comparison of Modeled and Monitored Data. Comparison of estimated and measured
exposures provide indeterminate evidence. No measured or monitored data were available for
comparison.

Overall Confidence in Exposure Estimates

Overall confidence in inhalation exposure estimates resulting for air concentrations modeled based on
releases from hydraulic fracturing operations (Section 3.2.3.2) is medium. The modeling methodologies
used to estimate air concentrations are robust. The distribution of air releases used as model input data
were estimated using Monte Carlo modeling and rely on assumptions. No air monitoring data were
available to confirm detection of 1,4-dioxane in air near hydraulic fracturing operations. There is
uncertainty around the extent to which people live and work in the specific locations where exposures
are highest, decreasing confidence in the exposure scenarios.

Overall confidence in inhalation exposure estimates resulting from air concentrations modeled based on
releases from industrial and institutional laundries ion 0) is medium. The modeling methodologies are
robust. The distribution of air releases used as model input data were estimated using Monte Carlo
modeling and rely on assumptions. No air monitoring data were available to determine whether 1,4-
dioxane is detected near industrial and institutional laundry facilities.

Page 127 of 484


-------
3812

3813

3814

3815

3816

3817

3818

3819

3820

3821

3822

3823

3824

3825

3826

3827

3828

3829

3830

3831

3832

3833

3834

3835

3836

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

4 HUMAN HEALTH HAZARD

1,4-Dioxane - Human Health Hazards (Section 4):

Key Points

EPA previously evaluated reasonably available information for human health hazards and
identified hazard endpoints for non-cancer effects and cancer effects following acute and chronic
exposures. This section describes adjustments made to previously published hazard values to
align with the exposure scenarios evaluated in this supplemental evaluation.

•	EPA considered the potential for increased susceptibility across PESS factors throughout
the hazard assessment and dose-response analysis. PESS categories identified in the
assessment include lifestage, genetics, and preexisting disease.

•	The primary acute/short-term, non-cancer endpoint for 1,4-dioxane is liver toxicity
following inhalation exposure.

•	The primary chronic, non-cancer endpoints for 1,4-dioxane are liver toxicity and systemic
effects on the olfactory epithelium.

•	Inhalation cancer endpoint for 1,4-dioxane is based on combined tumor risk at multiple
sites.

•	Oral and dermal cancer endpoints for 1,4-dioxane are based on liver tumors following
oral exposures.

4.1	Summary of Hazard Endpoints Previously Identified in the 2020 Risk

Evaluation	

This draft supplement relies on the Hazard Identification and Dose-Response Assessment that was
previously described in the 2020 RE. All hazard values used to calculate risks for 1,4-dioxane in this
draft supplement were derived from the previously peer-reviewed PODs published in the 2020 RE and
amended in the recent correction memo.

Hazard values used in the 2020 RE include human equivalent concentrations (HECs) and human
equivalent doses (HEDs) for non-cancer endpoints. Additionally, an inhalation unit risk (IUR) and
cancer slope factor (CSF) for lifetime cancer risk were derived for both occupational and consumer
scenarios for COUs where it was applicable. The hazard values published in the 2020 RE and used as
the basis for hazard values in this draft supplement were developed with consideration for potentially
susceptible subpopulations. Several potential sources of susceptibility were discussed qualitatively
including lifestage, genetic variability, liver disease, and other chronic diseases that may influence
metabolism or target organ susceptibility. EPA applied a 10x uncertainty factor to non-cancer hazard
values to account for these sources of human variability.

4.2	Summary of Adjustments to Previously Established Hazard Values

For many of the exposure scenarios evaluated in this draft supplement, the previously established peer-
reviewed hazard values were applied without modification. For example, risks from occupational
exposures to products containing 1,4-dioxane as a byproduct can be evaluated using the acute, chronic,
and cancer hazard values previously developed for OESs.

Some of the exposure scenarios included in this draft supplement require duration adjustments to the
previously established PODs. For example, to evaluate risks from ambient air exposures for fenceline

Page 128 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3837	communities, EPA assumes continuous exposure to air for 24 hours/day, 7 days/week. As described in

3838	more detail below, EPA adjusted the previously established HEC and IUR values (originally developed

3839	for 8 hours/day, 5 days/week exposures) to identify hazard values appropriate for continuous exposure

3840	scenarios.

3841

3842	In addition, acute and chronic non-cancer oral and dermal HEDs extrapolated from occupational HECs

3843	were corrected to apply consistent breathing rates assumptions.

3844

3845	The full set of hazard values used to evaluate risk from the exposure scenarios in this draft supplement

3846	are presented in Table 4-1.

Page 129 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3847 Table 4-1. Hazard Values Used for 1,4-Dioxane in this Draft Supplement 	

Scenario
(Population)

Endpoints

Inhalation
HEC/IUR

Dermal
HED/CSF

Oral HED/CSF

T otal
Uncertainty
Factors

Reference(s)

Acute non-cancer
(general population)

Systemic liver
toxicity

26.2 ppm
(94.5 mg/m3)
24 hours

17.4 mg/kg-d
(extrapolated from HEC)

17.4 mg/kg-d
(extrapolated from HEC)

300

(Pntz et a )

(Mattic et al, 2012)

Acute

non-cancer

(occupational)

Systemic liver
toxicity

78.7 ppm
(284 mg/m3)
8 hours

17.4 mg/kg-d
(extrapolated from HEC)a

17.4 mg/kg-d
(extrapolated from
HEC)fl

300

(Mattic et al., 2012)

Chronic
non-cancer
(general population)

Olfactory epithelium
effects attributed to
systemic delivery
(inhalation) a; liver
toxicity (oral)

0.846 ppm
(3 mg/m3)
24 hours, 7
days/week

0.56 mg/kg-d
(extrapolated from HEC)

2.6 mg/kg-d

30

(Kano et al., 2009;
Kasai et al., 2009)

Chronic

non-cancer

(occupational)

Olfactory epithelium
effects attributed to
systemic delivery
(inhalation)a; liver
toxicity (oral)

3.6 ppm
(12.8 mg/m3)
8 hours, 5
days/week

0.56 mg/kg-d
(extrapolated from HEC)''

2.6 mg/kg-d

30

(Kano et al., 2009;
Kas£u_etjyLv2009)

Cancer

(general population)

Inhalation cancer
risk based on
combined tumor risk
at multiple sites;
oral/dermal cancer
risk based on liver
tumors

IUR:

1.6E-02 per ppm
4.3E-06
(Hg/m3)"1
24 hours, 365
days/ year

CSF:

1.2E-01 (mg/kg-d)1
(extrapolated from oral
CSF)

CSF:

1.2E-01 (mg/kg-d)1



(Kano et al., 2009;
Kasai et al., 2009;
NIP. 1986)

Cancer

(occupational)

Inhalation cancer
risk based on
combined tumor risk
at multiple sites;
oral/dermal cancer
risk based on liver
tumors

IUR:

3.7E-03 per ppm
1.0E-06
(Hg/m3)-1
8 hours, 5
days/week

CSF: 1.2E-01 (mg/kg-d)1
(extrapolated from oral
CSF)

CSF:

1.2E-01 (mg/kg-d)1



(Kano et al., 2009;
Kasai et al.. 2009)

" Due to the uniform distribution of lesions (rather than a distribution consistent with airflow), EPA concluded that effects in the olfactory epithelium may be due to

systemic delivery rather than portal of entry effects due to the (see discussion in the 2020 RE on p 183, p. 188).

b Occupational HEDs extrapolated from occupational HECs were corrected as described in the correction memo.

HEC = Human Equivalent Concentration; HED = Human Equivalent Dose; CSF = Cancer Slope Factor; 1UR = Inhalation Unit Risk

3848

Page 130 of 484


-------
3849

3850

3851

3852

3853

3854

3855

3856

3857

3858

3859

3860

3861

3862

3863

3864

3865

3866

3867

3868

3869

3870

3871

3872

3873

3874

3875

3876

3877

3878

3879

3880

3881

3882

3883

3884

3885

3886

3887

3888

3889

3890

3891

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

4,2,1 Derivation of Acute/Short-Term Hazard Values

4.2.1.1	Inhalation HEC

The acute/short-term HECs are based on the lowest observed adverse effect concentration (LOAEC) for
systemic liver toxicity observed in a short-term inhalation toxicity study in rats exposed 6 hours/day for
5 days/week. In the 2020 RE, EPA derived an HEC for 8-hour occupational exposures by applying a
duration adjustment for an 8-hour exposure and a dosimetric adjustment factor of 1 (the default value
when the calculated ratio of animal to human blood:air partition coefficients is greater than 1 (
1994b). The occupational HEC derived in the 2020 RE is based on default breathing rate assumptions
and did not use adjustments for occupational breathing rates. For this draft supplement, EPA also
derived an HEC for continuous general population exposures by applying a 24-hour duration adjustment
to the original HEC.

4.2.1.2	Oral and Dermal HEDs

In the absence of acute oral or dermal toxicity studies, the acute/short-term HED was derived from the
acute HEC using route-to-route extrapolation. An acute HED for the general population was derived
from the duration-adjusted 24-hour HEC using the following equation:

dermal or oral HED (mg/kg-d) = PODhec (mg/m3) x inhalation volume x 100% inhalation
absorption ^ body weight

where the inhalation volume for the general population is 14.7 m3/day and body weight is 80 kg, based
on EPA's Exposure Factors Handbook (	2011). Inhalation absorption was estimated based on

experimental data from inhalation exposures in humans (Young et ai. 1977; Young et ai. 1976) that
indicated that 1,4-dioxane is readily absorbed; however, the available studies did not measure the
parameters needed to generate a quantitative estimate of the fraction absorbed. Given this qualitative
indication of rapid systemic uptake and the absence of quantitative inhalation absorption data, 100
percent inhalation absorption is assumed.

In the 2020 RE, an occupational acute HED was derived from the occupational HEC using the same
equation but with an inhalation volume for workers based on higher breathing rates. As described in the
correction memo, that derivation was incorrect. Because the occupational HEC was derived based on a
normal general population breathing rate, the HED derivation should apply the same breathing rate
assumptions. This draft supplement for 1,4-dioxane uses the revised acute occupational HED, which is
equal to the general population HED.

4.2,2 Derivation of Chronic Hazard Values

4.2.2.1 Inhalation HEC

The chronic HECs are based on BMCLio {i.e., the lower confidence limit of the benchmark
concentrations associated with a benchmark response of 10 percent) for effects in the olfactory
epithelium following inhalation exposures to rats for 6 hours/day, 5 days/week for 2 years. In the 2020
RE, EPA derived an HEC for chronic worker exposures by applying a duration adjustment for 8
hours/day and a dosimetric adjustment factor of 1 (the default value using the RGDR approach for
systemic effects when the calculated ratio of animal to human blood:air partition coefficients is greater
than 1 (U.S. EPA. 1994b)). The occupational HEC derived in the 2020 RE used default breathing rate
assumptions and did not adjust for occupational breathing rates. For this draft supplement, EPA also

Page 131 of 484


-------
3892

3893

3894

3895

3896

3897

3898

3899

3900

3901

3902

3903

3904

3905

3906

3907

3908

3909

3910

3911

3912

3913

3914

3915

3916

3917

3918

3919

3920

3921

3922

3923

3924

3925

3926

3927

3928

3929

3930

3931

3932

3933

3934

3935

3936

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

derived an HEC for continuous general population exposures by applying an alternate duration
adjustment for 24 hours/day, 7 days/week.

4.2.2.2	Oral HEDs

A chronic oral HED was calculated based on a benchmark dose level (BMDL) for liver toxicity
observed following chronic drinking water exposures to male rats and a nearly identical NOAEL value
for liver toxicity in male rats in a similar chronic toxicity study. In the 2020 RE, EPA derived an HED
by multiplying the nearly identical rodent BMDL and NOAEL values by (BWa/BWh)0'25, where BWa is
the bioassay-specific rodent body weight, and BWh is the default human body weight of 70 kg. Because
the chronic HED is based on a daily dose rate (as opposed to an intermittent exposure concentration), it
is equally applicable to both occupational and general population exposures and no additional
conversion is required.

4.2.2.3	Dermal HEDs

In the absence of chronic dermal toxicity studies, chronic dermal HEDs were derived from both the
chronic HEC and from the oral HED using route-to-route extrapolation. In the 2020 RE, the dermal
HED used for occupational risk calculations was extrapolated from the chronic worker HEC. For this
draft supplement, EPA also derived an HED from the HEC for continuous general population exposure.
The duration-adjusted chronic HEC for general populations was converted to a chronic HED for the
general population using the following equation:

dermal HED (mg/kg-d) = inhalation BMDLhec (mg/m3) x inhalation volume x 100% inhalation
absorption ^ body weight

where the inhalation volume for the general population is 14.7 m3/day (	) for a 24-hour

general population exposure and the body weight is 80 kg. As described above for the acute hazard
values, EPA assumed 100 percent inhalation absorption. In the 2020 RE, an occupational HED was
derived from the occupational HEC using the same basic equation but with an inhalation volume for
workers based on higher breathing rates. The difference in the HEDs derived from occupational and
general population HECs reflect differences in breathing rate assumptions for the two populations.

In the 2020 RE, an occupational chronic dermal HED was derived from the occupational HEC using the
same equation but with an inhalation volume for workers based on higher breathing rates. As described
in the correction memo, that derivation was incorrect. Because the occupational HEC was derived based
on a normal general population breathing rate, the HED derivation should apply the same breathing rate
assumptions. This assessment uses the revised occupational chronic dermal HED, which is equal to the
general population HED.

4.2.3 Derivation of Cancer Hazard Values

For cancer, the inhalation unit risk (MR) value was derived using the MS-Combo model to evaluate the
combined cancer risk for multiple tumor sites observed in male rats following inhalation exposure for 6
hours/day, 5 days/week for 2 years. Tumor types included in the MS-Combo model include nasal cavity
squamous cell carcinoma, Zymbal gland adenoma, hepatocellular adenoma or carcinoma, renal cell
carcinoma, peritoneal mesothelioma, mammary gland fibroadenoma, and subcutis fibroma. In the 2020
RE, EPA derived an IUR for chronic worker exposures by applying a dosimetric adjustment factor of 1
and a duration adjustment for 8 hours/day. The occupational IUR derived in the 2020 RE applied default
breathing rate assumptions and did not use adjustments for occupational breathing rates. The
occupational IUR was rounded to 1 x 10~6 (|ig/m3) 1 for application in risk calculations. For this draft

Page 132 of 484


-------
3937

3938

3939

3940

3941

3942

3943

3944

3945

3946

3947

3948

3949

3950

3951

3952

3953

3954

3955

3956

3957

3958

3959

3960

3961

3962

3963

3964

3965

3966

3967

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

supplement, EPA used that rounded occupational IUR to derive an IUR for continuous general
population exposures to 1,4-dioxane by applying a duration adjustment for 24 hours/day, 7 days/week.

The oral and dermal cancer slope factor was derived using the Multistage Weibull Model for the liver
tumors in female mice that had been exposed continuously via drinking water. In the 2020 RE, EPA
calculated an HED for each tumor type by multiplying rodent doses by (BWa/BWh)0,25, where BWa is
the bioassay-specific rodent body weight and BWh is the default human body weight of 70 kg. The CSF
was then calculated by dividing the benchmark response rate (0.5) by the HED. This CSF was applied to
both occupational and consumer/general population scenarios using scenario-specific risk benchmarks
and lifetime exposure estimates.

4.3 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty
in the Hazard and Dose-Response Analysis

All assumptions or uncertainties inherent to the human health hazard assessment and dose-response
analysis peer-reviewed in the 2020 RE are still applicable for this draft supplement. As described in the
2020 RE, EPA has medium confidence in the acute non-cancer PODs and high confidence in the chronic
non-cancer PODs for oral, dermal, and inhalation exposures. EPA has high confidence in the cancer
inhalation unit risk and medium-high confidence in the oral and dermal cancer slope factor. These
conclusions are based on the fact that there is a robust set of high quality chronic and sub-chronic
inhalation and oral exposure studies in rats and mice. The available evidence demonstrates consistent
systemic toxicity and tumor formation in rats exposed via inhalation and in both rats and mice exposed
via drinking water. Key sources of uncertainty include limited data on some sensitive reproductive and
developmental endpoints, reliance on route-to-route extrapolation, uncertainty around the mode of
action for 1,4-dioxane carcinogenicity, and the potential for subpopulations or lifestages with increased
biological susceptibility to 1,4-dioxane. Available methods indicate potential higher inhaled doses in
young children than adults, consistent with 1,4-dioxane specific studies integrating lifestage differences
in ventilation, anatomy and metabolism via CYP2E1 (	). The preferred method to

quantify these lifestage differences is a 1,4-dioxane specific PBPK model; however, the available PBPK
models for 1,4-dioxane are not adequate and there are not generally accepted default methods not
specific to 1,4-dioxane. Therefore, the air concentration is used as the exposure metric for all lifestages
and the 10x uncertainty factor accounts for these lifestage differences per EPA guidance (

2012. 1994bY

Page 133 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3968 5 HUMAN HEALTH RISK CHARACTERIZATION	

1,4-Dioxane - Human Health Risk Characterization (Section 5):

Key Points

EPA estimated cancer and non-cancer risks for each exposure pathway for a range of central tendency and high-
end exposure scenarios. Overall confidence in risk estimates varies across exposure pathways and COUs,
depending on the data and assumptions used to derive exposure and risk estimates. Differences in estimates
between central tendency and high-end exposure scenarios may reflect both variability across the population and
uncertainty in the exposure assessment.

•	Cancer and non-cancer risks were evaluated for occupational inhalation and dermal exposures to 1,4-
dioxane present as a byproduct.

O Cancer risk estimates for inhalation exposure range from 8.3x 10 12 to 1,8Ex 10 3 for central
tendency exposures and 5.4x 10" to 2.3x 10 2 for high-end exposures.

O Cancer risk estimates for dermal exposure range from 8.1 x 10 7 to 8.6x 1 () 4 for central tendency
exposures and from 5.Ox 10 " to 1,5x 10 2 for high-end exposures.

•	Cancer and non-cancer risks were evaluated for drinking water exposures resulting from releases to
surface water, including facility releases, down-the-drain releases, hydraulic fracturing releases, and
aggregate releases from multiple sources.

O Risk from individual facilities vary substantially within and across COUs, with cancer risk
estimates ranging from 5.4x10 12 to 0.025.

O Cancer risk estimates from modeled do wn-the-drain releases are highest in locations where large
populations are contributing to these releases and where they are ultimately discharged to
streams with low flow.

O Cancer risk estimates from modeled hydraulic fracturing waste releases to surface water are
3.57x 10~8 for median modeled releases and 1.45x 10~6 for 95th percentile modeled releases.

O Probabilistic modeling provides a distribution of risk estimates reflecting a range of drinking
water scenarios that account for aggregate sources of 1,4-dioxane in water.

•	Cancer risks were evaluated for drinking water exposures resulting from releases to land with potential
to reach groundwater.

O Risk estimates from landfill leachate are highest under disposal scenarios resulting in higher 1,4-
dioxane concentrations in leachate and higher landfill loading rates.

O Cancer risk estimates for drinking water exposures resulting from hydraulic fracturing waste
released to land/groundwater range from 4.Ox 10 7 for median modeled releases to 8.6x 10 " for
95th percentile modeled releases.

•	Cancer and non-cancer risks were evaluated for general population exposure to 1,4-dioxane in air.

O Cancer risk estimates for industrial air releases reported to TRI were generally highest within
1,000 m of the facilities and lower at greater distances.

O Cancer risk estimates within 1,000 m of hydraulic fracturing operations range from 0.2x 10~8 to
7.1x10 5 for a range of model scenarios across a range of high-end and central tendency release
scenarios.

O Cancer risk estimates within 1,000 m of industrial and institutional laundries range from
1.5xl0~n to 3.8 xlO-8 across a range of high-end and central tendency air concentrations
modeled for maximum release scenarios.

3969

Page 134 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3970	5.1 Risk Characterization Approach

3971	The exposure scenarios, populations of interest, and toxicological endpoints used for evaluating risks

3972	from acute and chronic exposures are summarized below in Table 5-1. To estimate risks from

3973	occupational and general population exposure scenarios evaluated in this draft supplement, EPA used

3974	the same methods described in the 2020 RE, as summarized below.

3975

3976	Table 5-1. Use Scenarios, Populations of Interest, and Toxicological Endpoints Used for Acute and

3977	Chronic Exposures	

Workers"

Acute - Adolescent (>16 years old) and adult workers exposed to 1,4-dioxane for a single 8-hour exposure
Chronic - Adolescent (>16 years old) and adult workers exposed to 1,4-dioxane for the entire 8-hour
workday for 260 days per year for 40 working years

General Population Drinking Water Exposures b

Acute - Adults, children, and formula-fed infants exposed to 1,4-dioxane through drinking water over a 24-
hour period

Chronic - Adults, children, and formula-fed infants exposed to 1,4-dioxane through drinking water up to 33

years d	

General Population Ambient Air Exposurec

Acute - People exposed to 1,4-dioxane through ambient air over a 24-hour period

Chronic - People exposed to 1,4-dioxane through ambient air continuously up to 33 years d	

Non-cancer Acute/Short-term Hazard Values
Sensitive acute/short-term health effect: liver toxicity

Acute Uncertainty Factors (Benchmark MOE) = 300 (UFA = 3; UFH = 10; UFL = 10)

•	8-hour HEC (occupational exposure) = 78.7 ppm

•	24-hour HEC (continuous general population exposure) = 26.2 ppm

•	Acute Oral and Dermal HED (occupational and general population exposure) = 17.4 mg/kg
Non-cancer Chronic Hazard Values
Sensitive chronic health effects:

•	Liver toxicity (oral)

•	Effects on the olfactory epithelium due to systemic exposures (inhalation and dermal)

Chronic Uncertainty Factors (Benchmark MOE) = 30 (UFa = 3; UFh = 10)

•	HEC (8-hour occupational exposure) = 3.6 ppm

•	HEC (continuous exposure general population exposure) = 0.846 ppm

•	Oral HED (for both occupational and general population scenarios) = 2.6 mg/kg/day

•	Dermal HED (extrapolated from HECs for both occupational and general population scenarios) =
0.56 mg/kg/day

Cancer Hazard Values

Inhalation cancer hazard for 1,4-dioxane is based on combined tumor hazard at multiple sites

•	IUR (occupational) = 3.7E-03 per ppm

•	IUR (continuous) = 1.6E-02 per ppm
Oral and dermal cancer hazards for 1,4-dioxane are based on liver tumors following oral exposures

•	Oral/dermal slope factor = 1.2E-01 (mg/kg/day)'1	

" Adult workers (>16 years old) include both female and male workers. Risks to ONUs were not calculated separately because
exposure data were not available for ONUs for the OESs being evaluated. Risks to ONUs are assumed to be equal to or less
than risks to workers who handle materials containing 1,4-dioxane as part of their job.

b These scenarios are used to evaluate potential risks from 1,4-dioxane in surface water, drinking water sources and
groundwater that may be used as drinking water.

c Inhalation exposures are described in terms of air concentrations and do not include lifestage-specific adjustments; risk
estimates based on air concentrations are intended to address risks to all lifestages (see Section 4.3).
d33 years is the 95th percentile residential occupancy period. U.S. EPA Exposure Factors Handbook (U.S. EPA, 2011).
Chapter 16, Table 16-5.

MOE = margin of exposure; UFA = Interspecies uncertainty factor for animal-to-human extrapolation; UFH = Intraspecies
uncertainty factor for human variability; UFL = LOAEC-to-NOAEC uncertainty factor for reliance on a LOAEC as the POD

3978

Populations
of Interest
and Exposure
Scenarios

Health
Effects,
Hazard
Values and
Benchmarks

Page 135 of 484


-------
3979

3980

3981

3982

3983

3984

3985

3986

3987

3988

3989

3990

3991

3992

3993

3994

3995

3996

3997

3998

3999

4000

4001

4002

4003

4004

4005

4006

4007

4008

4009

4010

4011

4012

4013

4014

4015

4016

4017

4018

4019

4020

4021

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

.1.1 Estimation of Non-cancer Risks

EPA used a margin of exposure (MOE) approach to identify potential non-cancer risks. The MOE is the
ratio of the non-cancer POD divided by a human exposure dose. Acute and chronic MOEs for non-
cancer inhalation and dermal risks were calculated using the following equation:

Non — cancer Hazard value (POD)

MOEacute or chronic —

Human Exposure

Where:

MOE = Margin of exposure (unitless)

Hazard value (POD)= HEC (ppm) or HED (mg/kg-d)

Human Exposure = Exposure estimate (in ppm or mg/kg-d)

MOE risk estimates may be interpreted in relation to benchmark MOEs. Benchmark MOEs are typically
the total UF for each non-cancer POD. The MOE estimate is interpreted as indicating a human health
risk if the MOE estimate is less than the benchmark MOE {i.e., the total UF). On the other hand, if the
MOE estimate is equal to or exceeded the benchmark MOE, risk is not indicated. Typically, the larger
the MOE, the more unlikely it is that a non-cancer adverse effect occurs relative to the benchmark.

When determining whether a chemical substance presents unreasonable risk to human health or the
environment, calculated risk estimates are not "bright-line" indicators of unreasonable risk, and EPA has
discretion to consider other risk-related factors apart from risks identified in risk characterization.

5.1.2 Estimation of Cancer Risks

Extra cancer risks for repeated exposures to a chemical were estimated using the following equations:

Inhalation Cancer Risk = Human Exposure x IUR

or

Dermal/Oral Cancer Risk = Human Exposure x CSF

Where:

Risk = Extra cancer risk (unitless)

Human exposure = Exposure estimate (LADC in ppm)

IUR. = Inhalation unit risk

CSF = Cancer slope factor

Estimates of extra cancer risks are interpreted as the incremental probability of an individual developing
cancer over a lifetime following exposure {i.e., incremental, or extra individual lifetime cancer risk).

5.2 Human Health Risk Characterization

5.2,1 Summary of Risk Estimates for Occupational Exposures

EPA estimated cancer and non-cancer risks for workers exposed to 1,4-dioxane based on the
occupational exposure estimates that were described in Section 3.1. Risks to ONUs were not calculated
separately because exposure data were not available for ONUs for the OESs being evaluated. Risks to
ONUs are assumed to be equal to or less than risks to workers who handle materials containing 1,4-
dioxane as part of their job.

Page 136 of 484


-------
4022

4023

4024

4025

4026

4027

4028

4029

4030

4031

4032

4033

4034

4035

4036

4037

4038

4039

4040

4041

4042

4043

4044

4045

4046

4047

4048

4049

4050

4051

4052

4053

4054

4055

4056

4057

4058

4059

4060

4061

4062

4063

4064

4065

4066

4067

4068

4069

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Below are summaries of the cancer risk estimates for the inhalation and dermal exposures as well as key
sources of uncertainty for all occupational exposure scenarios assessed in this draft supplement. These
risk estimates are based on exposures to workers in the absence of PPE such as gloves or respirators.
Section 3.1.2.4 contains an overall discussion on strengths, limitations, assumptions, and key sources of
uncertainty for the occupational exposure assessment. Additionally, Appendix F contains a
comprehensive weight of the scientific evidence summary table which presents an OES-by-OES
discussion of the key factors that contributed to each weight of the scientific evidence conclusion.
Complete risk calculations and results for occupational OES/COUs from the current analysis as well as
those previously presented in the 2020 RE are available in 1,4-Dioxane Supplemental Information File:
Occupational Exposure and Risk Estimates (U.S. EPA. 2023r).

Risk estimates vary across OES/COUs. Because cancer risk is the primary risk driver in most exposure
scenarios, this summary of results focuses on cancer risk estimates. For 7 of the 10 COU subcategories
evaluated, high-end cancer risk estimates were above 1 in 10,000. For many of those COUs, acute
and/or chronic non-cancer risk estimates were below the corresponding benchmark MOEs, indicating
that non-cancer risks may also be a concern. Cancer risk estimates for inhalation exposure range from
8.3/10 12to 1.8/10 3 for central tendency exposures and 5.4/10 " to 2.3/10 2 for high end exposures.
Cancer risk estimates for dermal exposure range from 8.1 x 10~7 to 8.6 / 10 4 for central tendency
exposures and from 5.Ox 10~6 to 1.5/10 2 for high end exposures. Risks are highest for PET
manufacture, hydraulic fracturing operations, ethoxylation processes, and dish soap/dishwashing
detergent. For these OESs, cancer risk estimates were greater than 1 in 10,000 for both central tendency
and high-end exposures. For these OESs, the key uncertainties include limited exposure monitoring data,
age of data, representativeness of key modeling parameters, and the extent to which the data collected
under past practices and operations are representative of modern practice and operations.

Overall confidence in risk estimates for occupational inhalation exposures ranges from low to high,
depending on the confidence in exposure assessment for each OES/COU. As described in Section 4.3,
overall confidence in the cancer inhalation unit risk underlying these risk estimates is high. As described
in Section 3.3.1.1, the measured and monitored inhalation exposure data are supported by moderate to
robust evidence. Additionally, the exposure modeling methodologies and underlying model input data is
supported by moderate to robust evidence. However, there is uncertainty in the representativeness of the
assessed exposure scenarios towards all potential exposures for the given OES/COU, limitations in the
amount and age of monitoring data, and limitations in the modeling approaches towards 1,4-dioxane-
specific use within the OES/COU. Therefore, while the underlying data and methods used to estimate
occupational inhalation risk is supported by moderate to robust evidence, the overall confidence of these
estimates ranges from low to high depending on the OES/COU. Key exposure considerations along with
the corresponding risk estimates are below.

•	Industrial/Commercial Use of Textile Dye. Risk estimates were derived using personal
breathing zone and area monitoring data collected from 1991 to 2010 at four facilities linked to
the use of textile dyes. Cancer risk estimates for inhalation exposure range from 1.6/10 5 for
central tendency exposures to 3.6* 10~2 for high-end exposures. However, there is uncertainty in
the risk estimates. The monitoring data used in this analysis are limited {i.e., 14 samples from
four sites). It also is not known how manufacturing processes and workplace conditions have
changed since the 1990s, when approximately half of the data was collected. For instance, EPA
does not have information available about the actual activities of the sampled workers and the
representativeness of the facility engineering controls to the modern practice.

•	Industrial/Commercial Use of Antifreeze. Risk estimates were derived from occupational
exposures modeled using Monte Carlo simulations for the worker activity of container

Page 137 of 484


-------
4070

4071

4072

4073

4074

4075

4076

4077

4078

4079

4080

4081

4082

4083

4084

4085

4086

4087

4088

4089

4090

4091

4092

4093

4094

4095

4096

4097

4098

4099

4100

4101

4102

4103

4104

4105

4106

4107

4108

4109

4110

4111

4112

4113

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

unloading. Cancer risk estimates for inhalation exposure range from 8.3x10 12 for central
tendency exposures to 5.4xl0~n for high-end exposures. However, there is uncertainty in the
risk estimates. Specifically, there is uncertainty as to the representativeness of some of the model
input data and, therefore, subsequent calculated exposures to the actual distribution of antifreeze
occupational exposures. This is due to limitations of using generic industry values identified for
the automotive industry. Also contributing to the uncertainty is that EPA used consumer
antifreeze use rates that were scaled up for commercial/industrial use in the Monte Carlo
modeling.

•	Industrial/Commercial Use of Surface Cleaner. Risk estimates were derived using 49 personal
breathing zone samples taken in 2019 during the use of surface cleaners in domestic kitchens and
bathrooms. Cancer risk estimates for inhalation exposure range from 1.1 x 10~7 for central
tendency exposures to 1,8x 10~6 for high-end exposures. However, there is uncertainty in the risk
estimates. Specifically, the monitoring data summary did not provide discrete monitoring points
and only provided summary statistics such as the geometric mean and maximum. Therefore,
EPA could not calculate the 50th and 95th percentile exposures. Also, it is uncertain the extent to
which the cleaning activities captured in this study reflect all occupational surface cleaning
scenarios, as they were measured in a consumer setting.

•	Industrial/Commercial Use of Dish Soap and Dishwasher Detergent. Risk estimates were
derived using 29 personal breathing zone and area exposure monitoring samples from 1980 taken
during the transfer of detergents to/from storage, liquid mixing, and detergent bottling. Cancer
risk estimates for inhalation exposure range from 4,Ox 10 4 for central tendency exposures to
1.0x10 3 for high-end exposures. However, there is uncertainty in these risk estimates. Although
the percent of 1,4 dioxane in the detergents used at the site where monitoring was conducted was
up to 0.423 percent, all monitoring samples were non-detect for 1,4-dioxane. Therefore, EPA
used the estimated LOD and LOD/2 for the worker high-end and central-tendency exposure
estimates. These values are two to four orders of magnitude (for dish soap and dishwasher
detergent, respectively) greater than the 2020 RE consumer exposure inhalation estimate. This
difference can reasonably be expected considering occupational users of dish soap are potentially
exposed to higher concentrations of 1,4-dioxane in concentrated commercial formulations and
use these dish soaps at higher frequencies for longer durations than consumers. However, there is
uncertainty as to the representativeness of these estimates due to the age of the monitoring data,
number of non-detects, and the limited sample size.

•	Industrial/Commercial Use of Laundry Detergent. Risk estimates were derived from
occupational exposures modeled using Monte Carlo simulations for the worker activity of
unloading detergent into machines, container cleaning, and laundry operations. For industrial
laundries, cancer risk estimates for vapor inhalation exposure range from 2.Ox 10~7 for central
tendency exposures to 9.2x10 7 for high-end exposures. For institutional laundries, cancer risk
estimates for vapor inhalation exposure range from 1,6x 10~7 for central tendency exposures to
7.1 x 10~7 for high-end exposures. In both cases, cancer risk estimates for total particulates
inhalation range from 4.0xl0~8 for central tendency exposures to 9.8x ] 0 8 for high-end
exposures. Cancer risk estimates for respirable particulates inhalation range from 1.3 x 10~8 for
central tendency exposures to 3.3x10-8 for high-end exposures. However, there is uncertainty in
the risk estimates. Specifically, there is uncertainty as to the representativeness of some of the
model inputs and, therefore, subsequent calculated exposures to the actual distribution of laundry

Page 138 of 484


-------
4114

4115

4116

4117

4118

4119

4120

4121

4122

4123

4124

4125

4126

4127

4128

4129

4130

4131

4132

4133

4134

4135

4136

4137

4138

4139

4140

4141

4142

4143

4144

4145

4146

4147

4148

4149

4150

4151

4152

4153

4154

4155

4156

4157

4158

4159

4160

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

detergent occupational exposures. This is due to limitations of using generic industry values
identified for institutional and industrial laundries.

•	Industrial/Commercial Use of Paint and Floor Lacquer. Risk estimates were derived using 17
personal breathing zone samples collected by NIOSH in 1987 at a military vehicle painting site.
Cancer risk estimates for inhalation exposure range from 8.Ox 10~5 for central tendency exposures
to 5.9x 10~4 for high-end exposures. However, there is uncertainty in the risk estimates. The
monitoring data used in this analysis are limited {i.e., 17 samples taken at one site). It also is not
known how processes and workplace conditions have changed since 1987. For instance, EPA
does not have information available about the actual activities of the sampled workers and the
representativeness of the facility engineering controls to modern practice.

•	Polyethylene Terephthalate (PET) Manufacturing. Risk estimates were derived using
personal breathing zone monitoring data collected from 1985 to 1994 at five facilities linked to
PET manufacturing. Cancer risk estimates for inhalation exposure range from 1.8x10 3 for
central tendency exposures to 2.3xl0~2 for high-end exposures. However, there is uncertainty in
the risk estimates. The monitoring data used in this analysis are limited {i.e., 11 samples from
five sites). It also is not known how manufacturing processes and workplace conditions have
changed since 1994. For instance, EPA does not have information available about the actual
activities of the sampled workers and the representativeness of the facility engineering controls
to modern practice.

•	Ethoxylation Processes. Risk estimates were derived using one composite 8-hour time-weighted
average personal breathing zone sample collected from one worker in 2000 at a soap and
detergent manufacturing facility. Cancer risk estimates for inhalation exposure range from

4,6x 10 4 for central tendency exposures to 5,9x 10 4 for high-end exposures. However, there is
uncertainty in the risk estimates. Specifically, EPA is unable to model these occupational
exposures and cannot determine the statistical representativeness of the one monitoring data
point {e.g., high-end, central tendency) towards potential exposures from this OES. Additionally,
there is uncertainty as to the worker activities covered by this monitoring data and whether all
foreseeable activities, corresponding exposures, and workplace operations are represented.

•	Hydraulic Fracturing Operations. Risk estimates were derived from occupational exposures
modeled using Monte Carlo simulations for the worker activities of container unloading,
container cleaning, and equipment cleaning. Cancer risk estimates for inhalation exposure range
from 7,Ox 10 5 for central tendency exposures to 9,5x ] 0 3 for high-end exposures. However,
there is uncertainty in the risk estimates. Specifically, there is uncertainty as to the
representativeness of some of the model input data and, therefore, the subsequent calculated
exposures to the actual distribution of hydraulic fracturing occupational exposures. This is due to
limitations of using generic industry values identified for the hydraulic fracturing industry as
well as self-reported values from FracFocus as model parameters. FracFocus data may not fully
represent operations across multiple sites throughout the United States as only certain sites
volunteered to submit data.

Overall confidence in risk estimates for occupational dermal exposures is medium for all OES/COUs
because the same modeling approach was used for all OES/COUs. As described in Section 4.3 overall
confidence in the oral and dermal cancer slope factor underlying these risk estimates is medium-high. As
described in Section 3.3.1.2, the dermal exposure modeling methodology is supported by moderate
evidence, with model input parameters from literature sources, a European model, standard defaults
from the ChemSTEER User Guide (	015a). and 1,4-dioxane product concentration data from

process information. These sources range from slight to robust, depending on factors such as age and
applicability to OES/COU. The modeling is limited by the use of standard input parameters that are not

Page 139 of 484


-------
4161

4162

4163

4164

4165

4166

4167

4168

4169

4170

4171

4172

4173

4174

4175

4176

4177

4178

4179

4180

4181

4182

4183

4184

4185

4186

4187

4188

4189

4190

4191

4192

4193

4194

4195

4196

4197

4198

4199

4200

4201

4202

4203

4204

4205

4206

4207

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

specific to 1,4-dioxane and a lack of variability in dermal exposure for different worker activities.
Differences in the dermal exposure modeling across COUs are driven primarily by COU-specific weight
fractions of 1,4 dioxane and the independent assessment of evaporative impacts in commercial and
industrial settings. Therefore, EPA's overall confidence in the occupational dermal risk estimates is
medium.

5.2.2 Summary of Risk Estimates for the General Population

5.2.2.1 Drinking Water - Surface Water Pathway

Risks from drinking water exposures were evaluated using a series of analyses that provide information
about the specific contributions of releases associated with individual OESs as well as information about
aggregate exposures and risks that could result from multiple sources releasing to the same water body.
Because most reasonably available surface water and drinking water monitoring data are not co-located
with 1,4-dioxane release sites, this analysis relies primarily on drinking water concentrations modeled
based on reasonably available release information. Risks predicted based on reasonably available
monitoring data are presented in 5.2.2.1.1.

EPA estimated cancer and non-cancer risks for adults, children, and formula-fed infants exposed to 1,4-
dioxane in drinking water. All risk estimates presented in this summary focus on the scenario with the
greatest potential exposure and risk. Because adult drinking water exposures relative to body weight are
greater than exposures relative to body weight averaged over the course of childhood (as illustrated in
1,4-Dioxane Supplemental Information File: Drinking Water Exposure and Risk Estimates for 1,4-
Dioxane Release to Surface Water from Individual Facilities (	2023hV) the drinking water

exposure scenario with the greatest lifetime cancer risk is 33 years of drinking water exposure as an
adult. Similarly, because cancer risk is the primary risk driver in most exposure scenarios, this summary
of results focuses on cancer risk estimates. More comprehensive sets of risk estimates for non-cancer
effects and other exposure scenarios are presented in the supplemental files referenced throughout this
section.

Drinking water exposure and risk estimates are highly dependent on the amount of 1,4-dioxane released
and the flow of the receiving water body. Both of these factors vary substantially across facilities within
each COU/OES, making release amount and flow much more important predictors of risk than a
facility's identified COU/OES. Exposure and risk estimates are also influenced by whether there is a
drinking water intake downstream of a release and the degree of dilution that occurs between the point
of release and the drinking water intake. Many of the risk estimates presented in the sections that follow
(for facility-specific releases, DTD, hydraulic fracturing, and aggregate modeling) assume that no
additional downstream dilution occurs prior to reaching drinking water intakes. This represents an upper
end estimate of exposure and risk based on the available data and the potential for intakes to be directly
downstream of a releasing facility. EPA conducted further analysis of the facility-specific releases to
consider the potential impact of downstream dilution on actual concentrations at drinking water intakes
and resulting risk estimates. Even when accounting for dilution between known releases and identified
drinking water intake locations, water concentrations estimated at drinking water intakes, instances of
cancer risks greater than 1 in 1 million for some public water systems are identified. Proximity of
releases to drinking water intakes and dilution are further discussed in Section 2.3.1.2.4/Appendix G.2.4
and Section 5.2.2.1.2.

1,4-Dioxane is not readily removed through typical wastewater or drinking water treatment processes.
Therefore, the drinking water risk estimates presented below are derived based on the assumptions that
drinking water intakes are located near 1,4-dioxane release sites and that no 1,4-dioxane is removed by

Page 140 of 484


-------
4208

4209

4210

4211

4212

4213

4214

4215

4216

4217

4218

4219

4220

4221

4222

4223

4224

4225

4226

4227

4228

4229

4230

4231

4232

4233

4234

4235

4236

4237

4238

4239

4240

4241

4242

4243

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

POTWs or through drinking water treatment. Use of source water estimated concentrations of 1,4-
dioxane to calculate cancer risk estimates is considered protective of all systems. These assumptions are
further discussed in Section 2.3.1.1/Appendix G.1.2).

5.2.2.1.1 Risks from Exposure to Drinking Water Concentrations Indicated in
Finished Drinking Water Monitoring Data

EPA evaluated risks for 1,4-dioxane concentrations reported in the reasonably available finished
(treated) drinking water monitoring data. Monitoring data included in this analysis were from
generalized, broad monitoring strategies, rather than targeted efforts to assess areas of known
contamination. As previously illustrated in Figure 2-10, 1,4-dioxane was below limits of detection for 89
percent of finished drinking water samples included in UCMR3 and state databases. Table 5-2
summarizes the distribution of lifetime cancer risk estimates from 1,4-dioxane concentrations detected
in finished drinking water reported in these databases (described in Section 2.3.1.1). This drinking water
monitoring data provides evidence that 1,4-dioxane is present in some finished drinking water and may
contribute to cancer risks in locations at the high end of monitored drinking water concentrations.

Monitoring data may not include the full range of 1,4-dioxane concentrations that result from industrial
releases. As discussed in Section 2.3.1.1, available drinking water monitoring data do not necessarily
capture locations that are most impacted by releases temporally or spatially and they often reflect
concentrations at a single point in time rather than average concentrations. However, as described in
Appendix G.2.3.2, in locations where monitoring data are available near release sites, comparisons
demonstrate strong concordance between modeled concentrations and monitoring data. EPA's
evaluation of drinking water risks therefore primarily relied on modeled estimates of 1,4-dioxane
concentrations that occur near release sites.

Table 5-2. Lifetime Cancer Risk Estimates for 1,4-Dioxane Concentrations Detected in Finished
Drinking Water	



Percentile Drinking Water Monitoring Data

Min

5%

25%

Median

75%

90%

95%

Max

Water conc.
(Mg/L)

2.00E-03

3.50E-02

3.50E-02

3.50E-02

3.50E-02

7.93E-02

0.16

13.3

Lifetime
Cancer Risk

4.02E-11

1.95E-08

1.95E-08

1.95E-08

8.37E-08

2.79E-07

3.46E-07

7.42E-06

Lifetime cancer risks based on 33 years of oral exposure through drinking water as an adult. Percentiles reflect
concentrations across the distribution of available drinking water monitoring data (this distribution includes non-
detects as half the detection limit).

5.2.2.1.2 Risks from Exposures to Water Concentrations Modeled from Industrial
Releases

To estimate the contribution of industrial releases to general population risks from drinking water, EPA
calculated cancer and non-cancer risk estimates based on modeled surface water concentrations in
receiving water bodies described in Section 2.3.1.3.1 and the resulting drinking water exposures
calculated as described in Section 3.2.2. Because there is substantial variation and uncertainty around
the extent of dilution that may occur in the receiving water body between the point of release and the
locations of drinking water intakes, EPA calculated cancer risk estimates under a range of reasonable
downstream dilution assumptions.

Page 141 of 484


-------
4244

4245

4246

4247

4248

4249

4250

4251

4252

4253

4254

4255

4256

4257

4258

4259

4260

4261

4262

4263

4264

4265

4266

4267

4268

4269

4270

4271

4272

4273

4274

4275

4276

4277

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Figure 5-1 shows the distribution of cancer risk estimates for industrial releases reported to TRI and
DMR, assuming that concentrations at drinking water intakes are the same as concentrations that occur
at the point of release after initial mixing in the receiving water body. Based on available data, this is a
plausible scenario in some locations. Lifetime cancer risk estimates are based on 33 years of exposure as
an adult and range from 5.41 x 10~13 to 2.54x 10~2. The median cancer risk estimate for these modeled
concentrations is 2.32/ 10 6 and the 95th percentile risk estimate is 4.92/ 10 3. Acute and chronic non-
cancer risk estimates for some facilities (data not shown) also indicate potential for non-cancer risk
relative to benchmark MOEs. This analysis represents an upper bound drinking water exposure scenario
in which intakes are located near the point of release or in which minimal additional dilution occurs
downstream. Complete cancer and non-cancer risk estimates for facility and OES-specific releases are
presented in 1,4-Dioxcme Supplemental Information File: Drinking Water Exposure and Risk Estimates
for 1,4-Dioxane Release to Surface Water from Individual Facilities (U.S. EPA. 2023h).



1









c
.2









1
1
1
1

1 IO









CD











1 ~

1
1
1
1

.

































1
1
1
1
1























1
1
1
1
1



























1
1
1
1
1











m





























i	1	1	1	1	1	1	1		1	1	1	—i	r

io~13 icr12 io_" icr10 10"S io_s 10"7 icr6 icr5 10"4 «r3 io~2 io~1

Adult Lifetime Cancer Risk

Figure 5-1. Distribution of Adult Lifetime Cancer Risk across all Facilities, Assuming No
Additional Dilution Occurs between the Point of Release and the Location of Drinking
Water Intakes

There is substantial variation in cancer risk estimates both within and across OESs. The large ranges of
modeled water concentrations and corresponding risk estimates reflect the large differences in the
amount of 1,4-dioxane released from facilities, the magnitude of flow within the receiving water body or
both.

For facilities where specific release amounts or locations are not reported, release amounts and flow
rates are based on conservative assumptions that may result in high risk estimates. There is uncertainty
around risk estimates for those facilities with limited release information, but facility-specific
information on release amounts and locations was available for most facilities. Therefore, while facility-
specific risk estimates based on facilities with limited information should be interpreted with caution,
most estimates are informed by moderate to robust modeling approaches and input data. To determine
the extent to which inclusion of facilities with limited release information influences the overall
distribution, EPA repeated this risk estimate analysis presented in Figure 5-1 using only facilities for
which high quality release data are available (Figure 5-2). Specifically, this additional analysis is limited
to facilities for which the annual release amount was sourced from either TRI Form R or DMR, and the
receiving water body reach code was identified in the facility's NPDES permit. Out of the 120 total

Page 142 of 484


-------
4278

4279

4280

4281

4282

4283

4284

4285

4286

4287

4288

4289

4290

4291

4292

4293

4294

4295

4296

4297

4298

4299

4300

4301

4302

4303

4304

4305

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

direct and indirect releases evaluated in this section, 80 met these strict data criteria. The resulting
distribution of risk estimates are similar to the results of the analysis including all facilities, ranging
from 5.41 xlO-13 to 2.54x10-2, with a median of 8.51><10~7 and 95th percentile of 4.92xl0~3.













c

.2









LO













CD

5









O-











































n

































































i



10 12 10 11 1(T10 1CT9 1CTa 1(T7 10" 10 s 10~4 10"s 10~2 10"'
Adult Lifetime Cancer Risk

Figure 5-2. Distribution of Adult Lifetime Cancer Risk across Facilities with High
Quality Release Data, Assuming No Additional Dilution Occurs between the Point of
Release and the Location of Drinking Water Intakes

The risk estimates summarized in Figure 5-1 and Figure 5-2 rely on the assumption that concentrations
at drinking water intakes are the same as concentrations estimated near the point of release. To evaluate
the validity of that assumption, EPA considered the proximity of release sites to downstream drinking
water intake locations for community and non-community non-transient PWSs. As shown in Table 5-3,
of the 69 facilities with cancer risk greater than 1 x 10~6, 22 (32 percent) have a downstream drinking
water intake within 250 km and 4 of those have a drinking water intake within 10 km. A detailed
description of this analysis is provided in Appendix G.2.4.

Table 5-3. Proximity of Nearest Downstream Drinking Water Intakes to Facilities Resulting in
Cancer Risk Greater than lxlO-6

Total
Facilities
Evaluated

Facilities with
Cancer Risk
above 1E-06

DWI within
250 km

DWI within
100 km

DWI within
50 km

DWI within
25 km

DWI within
10 km

120

69

22

17

11

7

4

The portion of 1,4-dioxane that remains after the additional dilution that occurs as it travels downstream
is highly variable based on site-specific characteristics, ranging from less than 1 percent to nearly 100
percent of the original concentrations (Figure 5-3). The site-specific factors that influence this additional
downstream dilution may not be fully captured in a national-scale assessment. Based on available site-
specific information for each facility, the mean modeled dilution predicted at downstream drinking
water intakes is diluted to 1 percent of original concentrations estimated in receiving water bodies near
the point of release.

Page 143 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

I ^<9

LO LJ ^

CM _c 0 40

•E | g

(/J o .<£
d) (0 (Y

1 0

£ CO
o £ CO 20

- rod

0 0 t_

-Q £= 0

E <2 ro
Q

4306

4307

4308

4309

4310

4311

4312

4313

4314

4315

4316

4317

4318

4319

4320

4321

4322

4323

4324

0	25	50	75	100

Diluted Concentration at Intake as a Percent
of the Concentration at the Point of Release

Figure 5-3. Distribution of Dilution of 1,4-Dioxane Concentrations at Downstream
Drinking Water Intakes

Figure 5-4 shows the distribution of cancer risk estimates for industrial releases, assuming that
concentrations at drinking water intake locations are diluted to 1 percent of the original 1,4-dioxane
concentrations in surface water estimated at the point of release. Lifetime cancer risk estimates for these
modeled concentrations range from 5.41 x 10~15 to 2.54x 10~4. The median cancer risk estimate is
8.51><10~9 and the 95th percentile risk estimate is 4.92xl0~5. This represents a plausible drinking water
exposure scenario consistent with mean modeled downstream dilution predicted across all facilities
based on available site-specific information.

20-

16-

cn
0
u)
cu
0

0 .£
Cd CQ

O W

c
0
3

4-

0-

10

-14

10",J 10",z 10 " 10~IU 10"M 10"° 10" 10"° 10"° 10
Adult Lifetime Cancer Risk

-4 "10 3

Figure 5-4. Distribution of Adult Lifetime Cancer Risk across all Facilities,

Assuming Dilution to 1% of Initial Concentrations in the Receiving Water Body

In addition to estimating how the overall distribution of cancer risk estimates would shift based on a
standard assumption of downstream dilution to 1 percent of original concentrations (as shown in Figure
5-4), EPA also estimated water concentrations and risks that may occur at specific drinking water

Page 144 of 484


-------
4325

4326

4327

4328

4329

4330

4331

4332

4333

4334

4335

4336

4337

4338

4339

4340

4341

4342

4343

4344

4345

4346

4347

4348

4349

4350

4351

4352

4353

4354

4355

4356

4357

4358

4359

4360

4361

4362

4363

4364

4365

4366

4367

4368

4369

4370

4371

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

intakes located downstream of releasing facilities, based on location-specific estimates of dilution. For
the 22 facilities with cancer risk greater than 1 in 1 million and drinking water intakes located within
250 km downstream, EPA identified downstream intakes associated with 73 distinct PWSs.

Even when accounting for site-specific influences on dilution, EPA modeled concentrations that would
result in adult lifetime cancer risk in excess of 1 in 1 million at intakes for 20 of the PWSs identified
through this assessment, serving a combined population of 2,124,000 people. Adult lifetime cancer risk
estimates were greater than 1 in 100,000 for 5 of these public water systems, serving a combined
population of 834,000 people. This analysis also identified locations with multiple releasing facilities
upstream of the same drinking water intake; however, in all such cases the aggregated adult lifetime
cancer risk calculated at the intake for the aggregated diluted concentration was less than 1 in a million.
A detailed description of this analysis is provided in Appendix G.2.4. Overall confidence in these
dilution-adjusted risk estimates is high for drinking water intakes located at or near the point of release,
but confidence decreases substantially with increasing distance downstream. This analysis does not
provide a comprehensive survey of modeled 1,4-dioxane concentrations at all drinking water intakes.
There may be additional drinking water intakes downstream of facilities releasing 1,4-dioxane that are
not accounted for in the intake database used in this analysis.

Overall, these analyses indicate that in many locations, downstream dilution may be expected to
substantially reduce 1,4-dioxane concentrations at the point of drinking water intakes. However, even
when accounting for dilution, upstream industrial releases reported to TRI or DMR contribute to cancer
risk estimates greater than 1 in a million or 1 in 100,000 at known drinking water intake locations.

The set of distributions presented in Figure 5-1, Figure 5-2, and Figure 5-4 indicate that high risks can
occur in specific locations downstream of release sites due to factors such as the size of the releasing
event(s), stream flow volume, proximity of the release site to drinking water intake, and limited drinking
water treatment removal from typical treatment methods.

Overall confidence in the overall distribution of risk estimates for drinking water exposures resulting
from facility releases is medium-high. Overall confidence in site-specific risk estimates for individual
facility releases varies both within and across OES, depending on the confidence in the source-specific
release data. As described in Section 4.3 overall confidence in the oral and dermal cancer slope factor
underlying these risk estimates is medium-high. As described in Section 3.3.2.1, the overall exposure
modeling methodology used for this analysis is supported by moderate evidence. It is designed to
estimate water concentrations expected at specific locations. Exposure estimates for this scenario are
based on some conservative assumptions about flow rates and release frequency and amount. For most
COUs, this analysis is limited to facilities that report via TRI and/or DMR. Other sources releasing
smaller amounts of 1,4-dioxane are not directly captured. Available monitoring data confirm that 1,4-
dioxane is present in some surface water and drinking water, though most of the available data were not
collected near release sites are therefore not directly comparable.

The overall level of confidence in facility-specific release estimates and resulting risk estimates depends
on the source of the release data described in Appendix E.3:

•	Overall confidence in drinking water exposure estimates is medium to high for OESs/COUs that
rely primarily on release data reported to DMR or to TRI via Form R. Most COUs/OESs are
included in this group.

•	Overall confidence in drinking water exposure estimates is medium for OESs/COUs for which
release estimates are based on reporting to TRI via Form A. The Import and Repackaging OES

Page 145 of 484


-------
4372

4373

4374

4375

4376

4377

4378

4379

4380

4381

4382

4383

4384

4385

4386

4387

4388

4389

4390

4391

4392

4393

4394

4395

4396

4397

4398

4399

4400

4401

4402

4403

4404

4405

4406

4407

4408

4409

4410

4411

4412

4413

4414

4415

4416

4417

4418

4419

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

releases used in this analysis are entirely based on Form A reporting of releases, and just under

half of the Industrial Uses OES releases were reported via Form A.

While confidence in facility-specific risk estimates varies, estimates for most facilities are informed by
moderate to robust modeling approaches and input data. Furthermore, the overall distribution is not
meaningfully altered by exclusion of facility-specific data based on more limited release information (as
illustrated in Figure 5-1 and Figure 5-2). There is some uncertainty around the proximity of releases to
drinking water intake locations and the extent to which 1,4-dioxane is further diluted prior to reaching
intake locations. EPA therefore estimated distributions of cancer risk estimates under a range of
assumptions about downstream dilution, reflecting the range of plausible drinking water intake
scenarios, as indicated by available site-specific information.

5.2.2.1.3 Risks from Exposures to Water Concentrations Modeled from DTD
Releases (from POTWs), Assuming No Downstream Dilution

EPA evaluated the potential contribution of DTD releases of consumer and commercial products to
drinking water exposure and risk. Surface water concentrations at the point of DTD releases via POTWs
are primarily determined by the size of the population contributing to DTD releases and the flow rates of
receiving water bodies. Risk estimates presented in this section are not tied to known releases at specific
locations. Rather, this analysis defines the conditions under which DTD releases would result in varying
levels of risk.

Cancer risk estimates shown in Table 5-4 were calculated based on drinking water exposure estimates
presented in Section 3.2.2.1.2, which correspond to surface water concentrations estimated by
probabilistic modeling of DTD releases under varying population sizes and stream flows. The resulting
risk estimates indicate that risk is highest in locations where large populations are contributing to DTD
releases and those releases are ultimately discharged to streams with low flow. Cancer risk estimates
greater than 1 in a million were seen in combinations of population size and flow rates that are likely to
occur in some locations. As described in Section 2.3.1.3.2, EPA considered the frequency of the varying
combinations of population sizes and flow rates. For communities with single POTWs treating
wastewater, most fell into the range of 100 to 10,000 people, with the annual average flow of the
receiving water body less than 300 cfs (Table 2-10). Cancer risk estimates for communities in this range
of population sizes are as low as 2.04x 10~8 at flows of 300 cfs and increase at lower flows. For example,
cancer risk estimates for releases from a population size of 10,000 could be as high as 2.04x 10~6 at a
flow of 30 cfs. Acute and chronic non-cancer risk estimates for these scenarios do not indicate non-
cancer risk relative to benchmark MOEs. For reference, stream flows of 100 cfs might be considered a
small river, while anything less than 100 cfs would be considered a stream or creek. Complete cancer
and non-cancer risk estimates for the range of water concentrations from DTD releases estimated under
varying conditions using probabilistic modeling are presented in 1,4-Dioxane Supplemental Information
File: Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Surface Water Concentrations
Predicted with Probabilistic Modeling (	,0231).

Overall confidence in risk estimates for drinking water exposures resulting from DTD releases under
varying population and stream flow conditions is medium. As described in Section 4.3 overall
confidence in the oral and dermal cancer slope factor underlying these risk estimates is medium-high. As
described in Section 3.3.2.1, the exposure modeling methodology used for this analysis is supported by
robust evidence and is designed to provide a nationally representative distribution of estimated water
concentrations under varying conditions. Exposure estimates rely on estimated distributions of DTD
releases of consumer and commercial products for each COU. Distributions of DTD releases of
consumer and commercial products were estimated for each COU on a per capita basis using the

Page 146 of 484


-------
4420

4421

4422

4423

4424

4425

4426

4427

4428

4429

4430

4431

4432

4433

4434

4435

4436

4437

4438

4439

4440

4441

4442

4443

4444

4445

4446

4447

4448

4449

4450

4451

4452

4453

4454

4455

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

SHEDS-HT model. Because this analysis is not tied to specific sites, there is uncertainty around the
proximity of releases to drinking water intake locations and the extent to which 1,4-dioxane is further
diluted prior to reaching intake locations. For this analysis, EPA assumed that no additional dilution
occurs prior to reaching drinking water intakes. While confidence in the individual contribution from
some specific COUs is lower, confidence in estimates of overall DTD releases is moderate.

Table 5-4. Lifetime Cancer Risk" Estimates from DTD Releases Alone (at the Point of Release)
under a Range of Population and Flow Rate Scenarios	



Population Contributing to Down-the-Drain Releases

100

1,000

10,000

100,000

1,000,000

Stream Flow
(cfs)

100

6.1 1E 09

6.1 1E-08

6.1 1E-07

6.1 1E 06

6.1 1E-05

300

2.04E-09

2.04E-08

2.04E-07

2.04E-06

2.04E-05

1,000

6.11E-10

6.11E-09

6.11E-08

6.11E-07

6.11E-06

3,000

2.04E-10

2.04E-09

2.04E-08

2.04E-07

2.04E-06

10,000

6.11 E— 11

6.11E-10

6.11E-09

6.11E-08

6.11E-07

30,000

2.04E-11

2.04E-10

2.04E-09

2.04E-08

2.04E-07

"Lifetime cancer risks based on 33 years of oral exposure through drinking water as an adult.

The frequencies of each of these combinations of population size and flow rate are presented in Table 2-10.

5.2.2.1.4 Risks from Exposure to Drinking Water Concentrations Modeled from
Disposal of Hydraulic Fracturing Produced Waters to Surface Water,
Assuming No Downstream Dilution

EPA evaluated the potential contribution of the disposal of hydraulic fracturing produced waters to
surface water by aggregating exposures and risks. The range of water concentrations that may result
from releases of hydraulic fracturing waste to surface water were estimated using probabilistic
modeling. Risk estimates presented in this section are not tied to known releases at specific locations.
Rather, this analysis defines the conditions under which releases from hydraulic fracturing would result
in varying levels of risk. These risk estimates are based on the assumption that 1,4-dioxane is not
removed by POTWs or through drinking water treatment.

Cancer risk estimates across the full distribution of modeled releases are presented in Table 5-5. Cancer
risk estimates are 3.57><10~8 for median modeled releases and 1.45><10~6 for 95th percentile modeled
releases. Acute and chronic non-cancer risk estimates based on 95th percentile modeled releases do not
indicate risk relative to benchmark MOEs. The maximum water concentration estimated by the model
reflects a scenario in which waste is released to a stream with very low flow. EPA does not have site-
specific information to indicate that such a scenario actually occurs. Complete risk estimates for the
range of water concentrations from DTD releases estimated under varying conditions using probabilistic
modeling are presented in 1,4-Dioxane Supplemental Information File: Drinking Water Exposure and
Risk Estimates for 1,4-Dioxane Surface Water Concentrations Predicted with Probabilistic Modeling
(	>02311

Overall confidence in risk estimates for drinking water exposures resulting from hydraulic fracturing
releases is medium. As described in Section 4.3 overall confidence in the oral and dermal cancer slope
factor underlying these risk estimates is medium-high. As described in Section 3.3.2.1, the exposure
modeling methodology used for this analysis is supported by robust evidence and is designed to provide
a nationally representative distribution of estimated water concentrations under varying conditions.
Releases used as inputs in the model were estimated using Monte Carlo modeling that captures

Page 147 of 484


-------
4456

4457

4458

4459

4460

4461

4462

4463

4464

4465

4466

4467

4468

4469

4470

4471

4472

4473

4474

4475

4476

4477

4478

4479

4480

4481

4482

4483

4484

4485

4486

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

variability across sites. However, the modeled exposure estimates are not directly tied to specific
releases at known locations, decreasing the strength of the evidence related to the representativeness of
the exposure estimates for actual exposures. There is some uncertainty around the proximity of releases
to drinking water intake locations and the extent to which 1,4-dioxane is further diluted prior to reaching
intake locations. For this analysis, EPA assumed that no additional dilution occurs prior to reaching
drinking water intakes.

Table 5-5. Lifetime Cancer Risks Estimated from
Hydraulic Fracturing Produced Waters Disposed
to Surface Water under a Range of Scenarios

Monte Carlo
Distribution

Adult Lifetime Cancer Risk

Maximum

1.10E-04

99th Percentile

4.35E-06

95th Percentile

1.45E-06

Median

3.57E-08

5th Percentile

1.29E-10

Minimum

3.40E-18

LADDs used to calculate these cancer risk estimates are
presented in Table 3-5.

5.2.2.1.5 Aggregate Risks from Drinking Water Exposures Modeled from Multiple
Sources Releasing to Surface Water, Assuming No Downstream Dilution

Multiple sources may contribute to 1,4-dioxane concentrations in drinking water sourced from surface
water in a single location. EPA therefore estimated aggregate general population exposures and risks
that could occur as a result of combined contributions from multiple sources. As described in Section
2.3.1.3.4, EPA used probabilistic modeling to predict aggregate surface water concentrations that could
occur when accounting for DTD releases, indirect releases, and other upstream sources. EPA estimated
cancer and non-cancer risks for the drinking water exposure estimates in Section 3.2.2, which
correspond to the modeled aggregate surface water concentrations described in Section 2.3.1.3.4 and
assume that no 1,4-dioxane is removed through treatment. This analysis also assumes that
concentrations at drinking water intakes are not further diluted from the concentrations modeled near the
point of release. There is wide variation in both cancer and non-cancer risk within and across
OESs/COUs when taking into account aggregate contributions from other sources. This variation is
illustrated in the cancer risk estimates shown in the distributions of cancer risk estimates for exposures
modeled for each OES/COU in Figure 5-5. The large ranges of risk estimates for some OESs/COUs
reflect substantial variation in releases and characteristics of receiving water bodies across the set
facilities associated with those OESs. High-end cancer risk estimates in this analysis are very similar to
high-end risk estimates for individual facility releases alone, indicating that high-end estimates are
driven primarily by high-end industrial releases. Complete cancer and non-cancer risk estimates for the
range of aggregate water concentrations estimated for each COU using probabilistic modeling are
presented in 1,4-Dioxane Supplemental Information File: Drinking Water Exposure and Risk Estimates

Page 148 of 484


-------
4487

4488

4489

4490

4491

4492

4493

4494

4495

4496

4497

4498

4499

4500

4501

4502

4503

4504

4505

4506

4507

4508

4509

4510

4511

4512

4513

4514

4515

4516

4517

4518

4519

4520

4521

4522

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

for 1,4-Dioxane Surface Water Concentrations Predicted with Probabilistic Modeling (

20231)-

Overall confidence in distributions of risk estimates for drinking water exposures resulting from for
aggregate surface water concentrations predicted by probabilistic modeling varies across OES/COU.
While confidence is not uniform for all facilities within an OES, overall confidence ratings for each
OES are intended to communicate how the factors that contribute to confidence and uncertainty vary
across COUs. As described in Section 4.3 overall confidence in the oral and dermal cancer slope factor
underlying these risk estimates is medium-high. As described in Section 3.3.2.1, the exposure modeling
methodology used for this analysis is supported by robust evidence and is designed to provide a
nationally representative distribution of estimated water concentrations under varying conditions. For
most COUs, this analysis is limited to facilities that report releases via TRI and/or DMR. Other sources
releasing smaller amounts of 1,4-dioxane are not directly captured, though the distribution of surface
water monitoring data used to represent background concentrations in the model is intended to capture
these other upstream sources. Available monitoring data confirm that 1,4-dioxane is present in some
surface water and drinking water, though most of the available data were not collected near release sites
and are therefore not directly comparable. In release locations where monitoring data are available, case
studies demonstrate strong concordance between modeled estimates and measured surface water
concentrations.

There is some uncertainty around the proximity of releases to drinking water intake locations and the
extent to which 1,4-dioxane is further diluted prior to reaching intake locations. For this analysis, EPA
assumed that no additional dilution occurs prior to reaching drinking water intakes. The characterization
of downstream dilution presented in 5.2.2.1.2 for individual facility releases illustrates the extent to
which downstream dilution may impact overall risk estimates.

The overall level of confidence in resulting exposure estimates depends on the source of OES/COU-
specific release data described in Appendix E.3:

•	Overall confidence in drinking water exposure estimates is medium to high for OESs/COUs that
rely primarily on release data reported to DMR or to TRI via Form R. Most COUs/OESs are
included in this group.

•	Overall confidence in drinking water exposure estimates is medium for OESs/COUs for which
release estimates are based on reporting to TRI via Form A. The Import and Repackaging OES
releases used in this analysis are entirely based on Form A reporting of releases, and just under
half of the Industrial Uses OES releases were reported via Form A.

•	Overall confidence in drinking water exposure estimates is low to medium for OESs/COUs for
which release estimates are based on surrogate or modeled information.

Page 149 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

4523

4524

4525

4526

30-

20-

10-

40-

w 30 -

3

v.)


CL

Ethoxylation byproduct

c
gj

a>

-=d



lO

CJ>

a.

Functional Fluids (Open-System)

c
<5
TJ
CD

J

LO

o
Q.

illli]

Import and Repackaging

_dn	tH

LO
CT>

Q_

Industrial Uses





LO

o
CL

Manufacture

LO
CT>
CL

PET Manufacturing

~u

0)



LO

a>
CL

Printing Inks
		



Remediation

-=~£1

LO
O)
CL

imlrL

10~J 10"^ 10"1 10"1U 10"3 10"B 10"' 10~b 10"& 10"4 10"J 10"* 10"1 10~1U 10"
Modeled Adult Lifetime Cancer Risk from Total 1,4-Dioxane (Releases + Background)

Figure 5-5. Histograms of Lifetime Cancer Risk Estimates for Aggregate Water Concentrations Estimated Downstream of COUs
with Vertical Lines Showing the Median and 95th Percentile (P95) Values

Page 150 of 484


-------
4527

4528

4529

4530

4531

4532

4533

4534

4535

4536

4537

4538

4539

4540

4541

4542

4543

4544

4545

4546

4547

4548

4549

4550

4551

4552

4553

4554

4555

4556

4557

4558

4559

4560

4561

4562

4563

4564

4565

4566

4567

4568

4569

4570

4571

4572

4573

4574

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

5.2.2.1.6 Integrated Summary of Drinking Water Risk Estimates across Multiple
Lines of Evidence for Surface Water

Risks from drinking water exposures were evaluated using a series of analyses that provided information
about the specific contributions of releases associated with individual sources as well as aggregate
exposures and risks to the general population. This analysis finds cancer risk estimates greater than 1 in
1 million from drinking water exposures informed by both monitoring data and modeled surface water
concentrations. Modeled concentrations result in cancer risk estimates greater than 1 in 1 million across
a range of individual sources and aggregate sources utilizing plausible drinking water exposure
scenarios.

Monitoring data demonstrates that 1,4-dioxane is present in some source water and finished drinking
water samples. Measured concentrations in finished drinking water samples resulted in cancer risk
estimates greater than 1 in 1 million at the high end of the distribution of monitoring samples. Most
drinking water treatment systems are not expected to remove 1,4-dioxane from water, suggesting that
concentrations detected in source water can also be an indication of concentrations in drinking water.

Available monitoring data provided information about general population exposures but did not capture
high concentrations occuring in specific locations or at specific times from direct and indirect releases
into water bodies. Therefore, EPA relied on estimated concentrations modeled for a range of specific
release scenarios, including direct and indirect industrial releases, DTD releases, disposal of hydraulic
fracturing waste, and aggregate concentrations resulting from varying combinations of multiple sources
to characterize risks from the water pathway. EPA evaluated the performance of the models used to
estimate water concentrations with monitoring data from site-specific locations serving as cases studies.
These case study comparisons demonstrated strong concordance between modeled concentrations and
monitoring data, thereby increasing confidence in risk estimates based on modeled concentrations.

Across all modeled scenarios, 1,4-dioxane concentrations in water are primarily determined by the
amount of release from varying sources and the flow of the receiving water body. These two factors are
highly location and source-specific, resulting in very wide ranges of modeled water concentrations and
risk estimates for each set of analyses presented in the previous section above.

Risk estimates based on 1,4-dioxane concentrations modeled in the receiving water bodies at the point of
release show potential for risk greater than 1 in 1 million or 1 in 100,000 from each of the sources
assessed.

As described in Section 5.2.2.1.2, dilution that occurs between the point of release and drinking water
intake locations may be expected to reduce 1,4-dioxane concentrations in some locations. However,
even when accounting for dilution, upstream releases contribute to cancer risk estimates greater than 1 in
a million or 1 in 100,000 at some drinking water intake locations. EPA evaluated risks based on
modeled water concentrations for a sample of drinking water intake locations downstream of releases
where risk was greater than 1 in 1 million. After accounting for additional dilution, cancer risk estimates
remained greater than 1 in 1 million for 27 percent of the public water systems evaluated, serving a
combined population of over 2 million people.

The potential relative contribution from different sources varies under different conditions and is likely
to be site-specific. For example, high-end risk estimates in the aggregate model (presented in Section
5.2.2.1.5) are very similar to high end risk estimates for facility releases alone (presented in Section
5.2.2.1.2), suggesting that in cases where industrial releases are high, those releases will be the dominant

Page 151 of 484


-------
4575

4576

4577

4578

4579

4580

4581

4582

4583

4584

4585

4586

4587

4588

4589

4590

4591

4592

4593

4594

4595

4596

4597

4598

4599

4600

4601

4602

4603

4604

4605

4606

4607

4608

4609

4610

4611

4612

4613

4614

4615

4616

4617

4618

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

source of 1,4-dioxane in water. On the other hand, under some conditions (presented in Section
5.2.2.1.3), plausible DTD release scenarios may present risk greater than 1 in 1 million in the absence of
industrial releases. Taken together, the analyses presented throughout this section demonstrate that each
of these sources may contribute to drinking water risks under some conditions. These analyses define the
conditions under which different levels of risk may occur.

5.2.2.2 Drinking Water - Groundwater and Disposal Pathways

EPA estimated risks from general population exposures that could occur if groundwater containing 1,4-
dioxane is used as a source of drinking water. Risk estimates presented in this section are not tied to
known releases at specific locations. Rather, this analysis defines the conditions under which 1,4-
dioxane disposal to landfills or from hydraulic fracturing operations could result in varying levels of
risk.

Cancer and non-cancer risk estimates were calculated based on modeled groundwater concentrations
described in Section 2.3.1.4 and corresponding drinking water exposures estimates described in Section
3.2.2.2. All risk estimates presented in this summary focus on the scenario with the greatest potential
exposure and risk. Because adult drinking water exposures relative to body weight are greater than
exposures relative to body weight that occur over the course of childhood (as illustrated in 1,4-Dioxane
Supplemental Information File: Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Release
to Surface Water from Individual Facilities (	2023h)), the scenario with the greatest lifetime

cancer risk is 33 years of exposure as an adult.

For potential groundwater concentrations resulting from landfill leachate, EPA estimated cancer and
non-cancer risks for adults and formula-fed infants at concentrations estimated under varying
hypothetical combinations of leachate concentrations and loading rates. As shown in Table 5-6, lifetime
cancer risk estimates increase under scenarios with higher leachate concentrations and loading rates.
Chronic non-cancer risk estimates (not shown) indicate risk relative to the benchmark MOE only at the
highest leachate concentrations and loading rates. These concentrations and loading rates represent a
scenario where 1,4-dioxane is either delisted and released to a municipal solid waste landfill or when
trace concentrations present in consumer and commercial products are disposed to those same landfills.
Though the higher concentrations of 1,4-dioxane in leachate and higher loading rates are less likely, they
may represent a sentinel PESS exposure. Complete results for cancer and non-cancer risk are available
in 1,4-Dioxane Supplemental Information File: Drinking Water Exposure and Risk Estimates for 1,4-
Dioxane Land Releases to Landfills (	2023f).

Overall confidence in risk estimates for drinking water exposures resulting from disposal to landfills is
low to medium. As described in Section 4.3 overall confidence in the oral and dermal cancer slope
factor underlying these risk estimates is medium-high. As described in Section 3.3.2.2.1 the modeling
methodology is robust. However, the release information relied on as model input data is supported by
slight to moderate evidence, decreasing overall confidence. In addition, this drinking water exposure
scenario relies on the assumption that the estimated groundwater concentrations may occur in locations
where groundwater is used as a primary drinking water source. While the substantial uncertainty around
the extent to which these exposures occur decreases overall confidence in the exposure scenario, this
scenario represents a sentinel exposure.

Page 152 of 484


-------
4619

4620

4621

4622

4623

4624

4625

4626

4627

4628

4629

4630

4631

4632

4633

4634

4635

4636

4637

4638

4639

4640

4641

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Table 5-6. Lifetime Cancer Risks" Estimated for Modeled Groundwater Concentrations Estimated

under Varying Landfill Conditions

Leachate
Concentration
(mg/L)

Loading Rate (lb)

0.1

1

10

100

1,000

10,000

100,000

1,000,000

0.0001

3.1E-17

2.9E-16

3.6E-15

3.4E-14

3.3E-13

3.1E-12

3.0E-11

2.9E-10

0.001

3.1E-16

2.9E-15

3.6E-14

3.4E-13

3.3E-12

3.1E-11

3.0E-10

2.9E-09

0.01

3.1E-15

2.9E-14

3.6E-13

3.4E-12

3.3E-11

3.1E-10

3.0E-09

2.9E-08

0.1

3.1E-14

2.9E-13

3.6E-12

3.4E-11

3.3E-10

3.1E-09

3.0E-08

2.9E-07

1

3.1E-13

2.9E-12

3.6E-11

3.4E-10

3.3E-09

3.1E-08

3.0E-07

2.9E-06

10

3.1E-12

2.9E-11

3.6E-10

3.4E-09

3.3E-08

3.1E-07

3.0E-06

2.9E-05

100

3.1E-11

2.9E-10

3.6E-09

3.4E-08

3.3E-07

3.1E-06

3.0E-05

2.9E-04

1,000

3.1E-10

2.9E-09

3.6E-08

3.4E-07

3.3E-06

3.1E-05

3.0E-04

2.9E-03

10,000

3.1E-09

2.9E-08

3.6E-07

3.4E-06

3.3E-05

3.1E-04

3.0E-03

2.9E-02

a Lifetime cancer risks based on 33 years of oral exposure through drinking water as an adult.

For groundwater concentrations resulting from disposal of hydraulic fracturing produced water, EPA
calculated cancer and non-cancer risks for adults and formula-fed infants. Cancer risk estimates across
the full distribution of modeled releases are presented in Table 5-7. Cancer risk estimates are 4.0E-07 for
median modeled releases and 8.6xl0~6 for 95th percentile modeled releases. Chronic non-cancer risk
estimates are above the corresponding benchmark MOE for all modeled groundwater concentrations,
indicating lower non-cancer risk from non-cancer effects. Complete cancer and noncancer risk
calculations are available in 1,4-Dioxane Supplemental Information File: Drinking Water Exposure and
Risk Estimates for 1,4-Dioxane Land Releases to Surface Impoundments (U.S. EPA. 2023e).

Overall confidence in risk estimates for drinking waters resulting from disposal of hydraulic fracturing
waste is low to medium. As described in Section 4.3 overall confidence in the oral and dermal cancer
slope factor underlying these risk estimates is medium-high. As described in Section 3.3.2.2.2, the
modeling methodology is robust and the release information relied on as model input data is supported
by moderate evidence. However, no monitoring data are available to confirm detection of 1,4-dioxane in
groundwater near hydraulic fracturing operations. This drinking water exposure scenario relies on the
assumption that the estimated groundwater concentrations may occur in locations where groundwater is
used as a primary drinking water source. While the substantial uncertainty around the extent to which
these exposures occur decreases overall confidence in the exposure scenario, this scenario represents a
sentinel exposure.

Page 153 of 484


-------
4642

4643

4644

4645

4646

4647

4648

4649

4650

4651

4652

4653

4654

4655

4656

4657

4658

4659

4660

4661

4662

4663

4664

4665

4666

4667

4668

4669

4670

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Table 5-7. Lifetime Cancer Risks" Estimated for Modeled Groundwater Concentrations Resulting

from Disposal of Hydraulic Fracturing Produced Water

Monte Carlo
Distribution

Modeled Groundwater
Concentration (mg/L)

Adult LADD

(mg/kg/day)

Adult Cancer Risk
Estimate

Max

1.9E-02

8.8E-05

1.1E-05

99th

1.5E-02

7.1E-05

8.6E-06

95th

1.5E-02

7.1E-05

8.6E-06

Mean

7.1E-04

3.3E-06

4.0E-07

50th

1.2E-04

5.6E-07

6.8E-08

5th

1.2E-04

5.6E-07

6.8E-08

Min

4.4E-07

2.1E-09

2.5E-10

a Lifetime cancer risks based on 33 years of oral exposure through drinking water as an adult.

5.2.2.3 Air Pathway

EPA estimated risks from general population exposures to 1,4-dioxane released to air, with a focus on
exposures in fenceline communities. Risks were evaluated for air releases from industrial COUs,
hydraulic fracturing operations, and industrial and institutional laundry facilities based on exposure
estimates in Section 3.2.3.

5.2.2.3.1 Industrial COUs Reported to TRI

EPA estimated risks from general population exposures that could occur in communities neighboring
industrial releases associated with stack and fugitive emissions. Cancer and non-cancer risk estimates for
general population exposures within 10,000 m of industrial releases were calculated for the 10th, 50th,
and 95th percentiles of modeled exposure concentrations estimated in Section 3.2.3.1. Table 5-8
summarizes the cancer risk estimates for 95th percentile exposure concentrations within 1,000 m of the
facilities with the greatest risk in each OES/COU, ranging from 1,05/10 10 to 1.1 x ] o 4, Cancer risk
estimates based on 50th percentile modeled exposure concentrations within 1,000 m of the highest risk
facilities range from 2,5x 10 " to 8.3 x 10 5 (data not shown). Risk estimates were generally highest
within 1,000 m of the facilities and lower at greater distances. As discussed in Section 2.3.3.3, exposure
estimates very near facilities (5 to 10 m) may be impacted by assumptions made for modeling around an
area source (10x10 area source places people at 5 m on top of the release point). This, in combination
with other factors like meteorological data, release heights, and plume characteristics can result in lower
or higher exposures at 5 m than just off the release point at 10 m. Air concentrations from fugitive
emissions tend to peak within 10 m of release sites while contributions from stack releases generally
peak around 100 m, meaning that risks nearest to release sites are often driven by fugitive releases.

Acute and chronic non-cancer risk estimates (not shown) do not indicate risk relative to benchmark
MOES for any of the estimated exposure concentrations at any facilities evaluated. Complete cancer and
non-cancer risk results for air concentrations modeled from stack, fugitive and combined air emissions
are provided in 1,4-Dioxane Supplemental Information File: Air Exposures and Risk Estimates for
Single Year Analysis (	:023e).

Page 154 of 484


-------
4671

4672

4673

4674

4675

4676

4677

4678

4679

4680

4681

4682

4683

4684

4685

4686

4687

4688

4689

4690

4691

4692

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Air exposure and risk estimates are dependent on release amounts, stack heights, contributions from
stack releases and fugitive emissions, topography, and meteorological conditions. These factors vary
substantially across facilities within each OES/COU, making release amount, stack height, and
meteorological conditions more important predictors of risk than a facility's identified OES/COU.

Overall confidence in site-specific risk estimates for inhalation exposure resulting from industrial
releases varies across OES/COUs. As described in Section 4.3, overall confidence in the cancer
inhalation unit risk underlying these risk estimates is high. As described in Section 3.3.3.1, the
AERMOD modeling methodology used for this analysis is robust and accounts for both stack and
fugitive emissions. The exposure scenarios considered are most relevant to long-term residents in
fenceline communities. There is some uncertainty around the extent to which people actually live and
work around the specific facilities where risks are highest, decreasing overall confidence in the exposure
scenario, particularly at distances nearest release sites. Overall confidence varies due to variable levels
of confidence in underlying release information used to estimate exposures. An OES-specific discussion
of the confidence in sources of release information is presented in Appendix E.5E.5.4, but in general
terms:

•	Overall confidence in risk estimates is medium to high for OESs/COUs that rely primarily on
release data reported to TRI via Form R.

•	Overall confidence in risk estimates is medium for OESs/COUs for which release estimates are
based on data reported to TRI via Form A.

•	Overall confidence in risk estimates is low to medium for OESs/COUs for which release
estimates are based on surrogate or modeled information.

Page 155 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

4693	Table 5-8. Inhalation Lifetime Cancer Risks" within 10 km of Industrial Air Releases Based on 95th Percentile Modeled Exposure

4694	Concentrations

OES

Corresponding COUs

# Facilities

Distance from Facility with Greatest Risk (m)6

Overall
Confidence

Life Cvelc
Stage

Category

Subcategory

Total

Risk
>1E—06

5

10

30

60

100

100-
1,000

Disposal

Disposal

Disposal

Hazardous waste

incinerator
Off-site waste

transfer
Underground

injection
Hazardous landfill

15

5

2.88E-05

3.42E-05

1.22E-05

4.67E-06

2.13E-06

2.00E-07

Medium
to High

Dry Film
Lubricant

Industrial use,

commercial

use

Other uses

Dry film lubricant

8

0

1.09E-12

4.83E-11

346E-09

2.62E-08

4.26E-08

6.72E-09

Low to
Medium

Ethoxylation
Byproduct

Processing

Byproduct

Byproduct
produced during
the ethoxylation
process to make
ethoxylated
ingredients for
personal care
products

6

3

4.42E-05

9.21E-05

4.96E-05

2.09E-05

1.11E-05

2.58E-06

Medium
to High

Film Cement

Industrial use,

commercial

use

Adhesives
and sealants

Film cement

1

0

846E-07

8.86E-07

2.99E-07

1.54E-07

8.46E-08

1.55E-08

Low to
Medium

Functional
Fluids
(Open-
System)

Industrial use

Functional
fluids (open
and closed
systems)

Polyalkylene

glycol lubricant
Synthetic
metalworking
fluid
Cutting and
tapping fluid

2

0

8.67E-08

1.60E-07

6.98E-08

7.31E-08

1.23E-07

5.02E-08

Medium
to High

Import and
Repackaging

Manufacturing

Import

Import
Repackaging

1

0

1.82E-13

3.78E-12

3.74E-10

2.82E-09

5.89E-09

2.18E-09

Medium
to High

Page 156 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023



Corresponding COUs

# Facilities

Distance from Facility with Greatest Risk (m)6

Overall

OES

Life Cvelc

Stage

Category

Subcategory

Total

Risk
>1E—06

5

10

30

60

100

100-
1,000

Confidence



Processing

Processing as
a reactant

Polymerization
Catalyst



















Industrial

Uses

Processing

Non-

incorporative

Basic organic
chemical
manufacturing
(process solvent)

















Medium
to High

Industrial use

Intermediate
use

Plasticizer

intermediate
Catalysts and
reagents for
anhydrous acid
reactions,
brominations,
and sulfonations

12

6

2.84E-05

3.24E-05

1.04E-05

3.84E-06

1.89E-06

4.85E-07

Laboratory

Chemical

Use

Industrial use,

commercial

use

Laboratory
chemicals

Chemical reagent
Reference material
Spectroscopic and
photometric
measurement
Liquid scintillation
counting medium
Stable reaction

medium
Cryoscopic solvent
for molecular
mass

determinations
Preparation of
histological
sections for
microscopic
examination

1

1

1.40E-05

1.46E-05

4.91E-06

2.54E-06

1.40E-06

2.54E-07

Low to
Medium

Manufactur-
ing

Manufacturing

Domestic
manufacture

Domestic
manufacture

1

1

5.91E-05

1.10E-04

5.20E-05

2.18E-05

1.08E-05

9.62E-07

Medium
to High

Page 157 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Corresponding COUs

# Facilities

Distance from Facility with Greatest Risk (m)6

Overall
Confidence

Life Cvelc

Stage

Category

Subcategory

Total

Risk
>1E—06

5

10

30

60

100

100-
1,000

PET

Manufactur-
ing

Processing

Byproduct

Byproduct
produced during
the production of
polyethlene
terephtalate

13

10

5.42E-05

5.48E-05

2.37E-05

9.47E-06

4.35E-06

7.25E-07

Medium
to High

Spray Foam
Application

Industrial use,

commercial

use

Other uses

Spray polyurethane
foam

1

0

5.28E-09

5.68E-09

1.94E-09

1.02E-09

5.79E-10

1.05E-10

Low to
Medium

"Lifetime cancer risks based on 33 years of continuous inhalation exposure averaged over a 78-year lifetime.
h Cancer risks were also calculated at 2,500, 5,000 and 10,000 m from all facilities.

4695

Page 158 of 484


-------
4696

4697

4698

4699

4700

4701

4702

4703

4704

4705

4706

4707

4708

4709

4710

4711

4712

4713

4714

4715

4716

4717

4718

4719

4720

4721

4722

4723

4724

4725

4726

4727

4728

4729

4730

4731

4732

4733

4734

4735

4736

4737

4738

4739

4740

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Land Use Analysis

For locations where lifetime cancer risk is greater than 1 x 10~6, EPA evaluated land use patterns to
determine whether fenceline community exposures may be reasonably anticipated. Detailed results of
this analysis are described in Appendix J.3 and are consistent with the methods described in the draft
2022 Fenceline Report. In short, EPA determined whether residential, industrial/commercial businesses,
or other public spaces are present within those radial distances where cancer risk estimates for 95th
percentile modeled air concentrations are greater than 1 x 10~6 for each facility. In all cases, risks greater
than 1 x 10~6 were within 1,000 m or less of releasing facilities. This analysis was limited to facilities that
could be mapped to a GIS location. Based on this characterization of land use patterns, fenceline
community exposures have the potential to occur at 50 percent of facilities (11 of 22 GIS-mapped
facilities) where cancer risk is greater than 1 x 10~6 based on modeled fenceline air concentrations.

Aggregate Risk

EPA also evaluated potential risks from aggregate exposures from multiple neighboring facilities using a
conservative screening methodology. EPA identified five groups of two to four facilities reporting 1,4-
dioxane releases in proximity to each other, i.e., within 10 km. Aggregating risks estimated for these
groups of facilities were generally dominated by the facility with the greatest risk. This aggregate
analysis did not identify locations with cancer risk greater than 1 x 10~6 that did not already have cancer
risk above that level from an individual facility. Details of the methods and results of this aggregate
analysis are described in Appendix J.4.

5.2.2.3.2 Hydraulic Fracturing

Cancer and non-cancer risk estimates for potential general population exposures within 1,000 m of
hydraulic fracturing operations were calculated for a range of air concentrations modeled across the
distribution of release estimates, as described in Section 3.3.3.2. Table 5-9 presents lifetime cancer risk
estimates for exposure to high-end and central tendency air concentrations modeled for both high end
(95th percentile) and central tendency (50th percentile) modeled releases for a range of topographical
and meteorological scenarios. Lifetime cancer risk estimates for distances within 1,000 m of hydraulic
fracturing operations range from 3.9x 10~7 to 7.1 x 10 5 for high end release estimates and 2,2/ 10 8 to
4.1xio-6 for central tendency release estimates across a range of model scenarios. Acute and chronic
non-cancer risk estimates (not shown) do not indicate risk relative to benchmark MOEs for any exposure
concentrations estimated for hydraulic fracturing operations. Complete results are provided in 1,4-
Dioxane Supplemental Information File: Air Exposure and Risk Estimates for 1,4-Dioxane Emissions
from Hydraulic Fracturing Operations (U.S. EPA. 2023b).

Overall confidence in risk estimates for inhalation exposures resulting for air concentrations modeled
based on releases from hydraulic fracturing operations is medium. As described in Section 4.3, overall
confidence in the cancer inhalation unit risk underlying these risk estimates is high. As described in
Section 3.3.3.2 the modeling methodologies used to estimate air concentrations are robust. The
distribution of air releases used as model input data were estimated using Monte Carlo modeling and
rely on assumptions. No air monitoring data were available to confirm detection of 1,4-dioxane is air
near hydraulic fracturing operations. Because the air concentrations underlying this analysis are based
on releases estimated using probabilistic modeling, they are not tied to specific locations that can be
evaluated for land use patterns. There is therefore substantial uncertainty around the extent to which
people actually live and work around the specific locations where risks are highest, decreasing overall
confidence in the exposure scenario.

Page 159 of 484


-------
4741

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Fugitive
Emissions Release
Scenario

Cancer Risk Estimates for 95th Percentile Modeled Releases

Cancer Risk Estimates for 50th Percentile Modeled Releases

High-End Modeled Air
Concentrations

Central Tendency Modeled Air
Concentrations

High-End Modeled Air
Concentrations

Central Tendency Modeled
Air Concentrations

100 m

1,000 m

100 to
1,000 m

100 m

1,000 m

100 to
1,000 m

100 m

1,000 m

100 to
1,000 m

100 m

1,000 m

100 to
1,000 m

South (Coastal)-
Rural-24

7.1E-05

3.2E-06

8.3E-06

5.6E-05

2.4E-06

6.4E-06

4.1E-06

1.8E-07

4.7E-07

3.2E-06

1.4E-07

3.6E-07

West North
Central-Rural-24

5.4E-05

3.1E-06

7.3E-06

4.1E-05

1.9E-06

4.9E-06

3.1E-06

1.8E-07

4.2E-07

2.4E-06

1.1E-07

2.8E-07

South (Coastal)-
Urban-24

3.4E-05

7.7E-07

2.4E-06

3.0E-05

6.7E-07

2.1E-06

2.0E-06

4.4E-08

1.4E-07

1.7E-06

3.8E-08

1.2E-07

West North
Central-Urban-24

3.2E-05

8.3E-07

2.5E-06

2.6E-05

6.1E-07

1.9E-06

1.8E-06

4.8E-08

1.4E-07

1.5E-06

3.5E-08

1.1E-07

South (Coastal)-
Rural-8

1.3E-05

1.2E-07

5. IE—07

1.1E-05

9.0E-08

4.2E-07

7.3E-07

6.8E-09

2.9E-08

6.4E-07

5.2E-09

2.4E-08

West North
Central-Rural-8

2.7E-05

1.0E-06

2.5E-06

1.4E-05

2.9E-07

8.8E-07

1.5E-06

5.9E-08

1.4E-07

8.2E-07

1.6E-08

5.0E-08

South (Coastal)-
Urban-8

1.2E-05

9.0E-08

4.3E-07

1.1E-05

8.0E-08

3.9E-07

6.7E-07

5.2E-09

2.5E-08

6.1E-07

4.6E-09

2.2E-08

West North
Central-Urban-8

1.9E-05

3.9E-07

1.2E-06

1.2E-05

1.6E-07

6.0E-07

1.1E-06

2.2E-08

7.1E-08

7.0E-07

9. IE—09

3.4E-08

" Lifetime cancer risks based on 33 years of continuous inhalation exposure averaged over a 78-year lifetime.
h Cancer risk estimates shown here are based on modeled releases and air concentrations estimated for 72 days of release.

¦.ab

4742

Page 160 of 484


-------
4743

4744

4745

4746

4747

4748

4749

4750

4751

4752

4753

4754

4755

4756

4757

4758

4759

4760

4761

4762

4763

4764

4765

4766

4767

4768

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

5.2.2.3.3 Industrial and Institutional Laundry Facilities

Cancer and non-cancer risk estimates for potential general population exposures within 1,000 m of
industrial and institutional laundry facilities were calculated for a range of air concentrations modeled
for a range of releases, as described in Section 3.2.3.3. Table 5-10 presents lifetime cancer risk estimates
for exposures estimated from both high-end and central tendency air concentrations modeled based on
the maximum release scenario for each type of laundry under the most conservative exposure scenario
evaluated (rural south coastal topography, assuming 24 hours of releases each day). Lifetime cancer risk
estimates for distances within 1,000 m of laundry facilities range from 1.5/10 " to 3,8/10 8 across a
range of high end and central tendency exposure scenarios. Acute and chronic non-cancer risk estimates
(not shown) do not indicate risk for any estimated exposure concentrations for laundries relative to the
benchmark MOEs. Complete results are provided in 1,4-Dioxane Supplemental Information File: Air
Exposures and Risk Estimates for Industrial Laundry 0 1 S IT \ 2023c).

Overall confidence in risk estimates from inhalation exposures resulting from industrial and institutional
laundries is medium. As described in Section 4.3, overall confidence in the cancer inhalation unit risk
underlying these risk estimates is high. As described in Section 3.3.3.2, the modeling methodologies are
robust. The distribution of air releases used as model input data were estimated using Monte Carlo
modeling and rely on assumptions. No air monitoring data were available to determine whether 1,4-
dioxane is detected near industrial and institutional laundry facilities. Because the air concentrations
underlying this analysis are based on probabilistic modeling, they are not tied to specific locations that
can be evaluated for land use patterns. There is therefore substantial uncertainty around the extent to
which people actually live and work around the specific locations where risks are highest, decreasing
overall confidence in the exposure scenario.

Table 5-10. Lifetime Cancer Risk Estimates for Fugitive Emissions from Industrial and
Institutional Laundry Facilities"	

Facility Type

Detergent and
Emissions Type

Cancer Risk Estimates for Maximum Modeled Releases

High-End Modeled Air
Concentrations

Central Tendency Modeled Air
Concentrations

100 m

1,000 m

100 to
1,000 m

100 m

1,000 m

100 to
1,000 m

Industrial
Laundry

Liquid - vapor

3.7E-08

1.7E-09

4.3E-09

3.3E-08

1.4E-09

3.8E-09

Powder - vapor

3.6E-08

1.7E-09

4.2E-09

3.3E-08

1.4E-09

3.8E-09

Powder - PM10

3.8E-08

8.8E-10

3.2E-09

3.4E-08

7.9E-10

2.9E-09

Powder - PM2.5

3.6E-08

1.6E-09

4.1E-09

3.3E-08

1.4E-09

3.7E-09

Institutional
Laundry

Liquid - vapor

2.3E-08

1.1E-09

2.7E-09

2.1E-08

9.0E-10

2.4E-09

Powder - vapor

6.8E-10

3.2E-11

7.9E-11

6.2E-10

2.7E-11

7.0E-11

Powder - PM10

7.1E-10

1.6E-11

5.9E-11

6.4E-10

1.5E-11

5.3E-11

Powder - PM2.5

6.8E-10

3.0E-11

7.7E-11

6.2E-10

2.6E-11

6.9E-11

a Lifetime cancer risks based on 33 years of continuous inhalation exposure averaged over a 78-year lifetime.

Page 161 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

4769	5.2.2.4 Potentially Exposed or Susceptible Subpopulations

4770	EPA considered PESS throughout the exposure assessment presented in this draft supplement and

4771	throughout the hazard identification and dose-response analysis described in the 2020 RE. Table 5-11

4772	summarizes how PESS were incorporated into the draft supplement through consideration of increased

4773	exposures and/or increased biological susceptibility. The table also summarizes the remaining sources of

4774	uncertainty related to consideration of PESS.

4775

4776	Table 5-11. Summary of PESS Considerations Incorporated throughout the Analysis and

4777	Remaining Sources of Uncertainty		

PESS
Categories

Potential Exposures Identified in Specific
Subpopulations and Incorporated into
Exposure Assessment

Potential Sources of Biological
Susceptibility Identified and
Incorporated into Hazard Assessment

Lifestage

General population drinking water exposure
scenarios include lifestage-specific exposure
factors for adults, children, and formula-fed
infants (Section 5.2.2.1); Inhalation exposures
are based on air concentrations and are therefore
consistent across lifestages (Section 5.2.2.3).
Based on pchem properties and a lack of studies
evaluating potential for accumulation in milk,
EPA did not quantitatively evaluate the milk
pathway and this is a source of uncertainty.

EPA qualitatively described the potential
for biological susceptibility due to lifestage
differences and developmental toxicity but
did not identify quantitative evidence of
lifestage-specific susceptibilities to 1,4-
dioxane; A 10/ UF was applied for human
variability. The magnitude of potential
lifestage differences in metabolism and
toxicity are not well quantified and are a
remaining source of uncertainty.

Pre-existing
Disease

EPA did not identify health conditions that may
influence exposure. The potential for pre-
existing disease to influence exposure (due to
altered metabolism, behaviors, or treatments
related to the condition) is a source of
uncertainty.

EPA qualitatively described the potential
for pre-existing health conditions, such as
liver disease, to increase susceptibility or
alter toxicokinetics, but did not identify
direct quantitative evidence. A 10* UF
was applied for human variability. The
potential impact of pre-existing diseases on
susceptibility to 1,4-dioxane is a remaining
source of uncertainty.

Lifestyle
Activities

EPA did not identify specific lifestyle activities
that expected to increase 1,4-dioxane exposure.
This is a remaining source of uncertainty.

EPA did not identify lifestyle factors that
influence biological susceptibility to 1,4-
dioxane. This is a remaining source of
uncertainty.

Occupational
Exposures

EPA evaluated a range of occupational exposure
scenarios in manufacturing, hydraulic fracturing
and use of commercial products that increase
exposure to 1,4-dioxane present as a byproduct.
EPA evaluated risks for high-end exposure
estimates for each of these scenarios (Section
5.2.1).

EPA did not identify occupational factors
that increase biological susceptibility to
1,4-dioxane. This is a remaining source of
uncertainty.

Geographic
Factors

EPA evaluated risks to fenceline communities
from 1,4-dioxane in ambient air (Section 5.2.2.3)
and in drinking water downstream of release
sites (Section 5.2.2.1). EPA mapped tribal lands
in relation to air, surface water and ground water
releases of 1,4-dioxane to identify potential for

EPA did not identify geographic factors
that increase biological susceptibility to
1,4-dioxane. This is a remaining source of
uncertainty.

Page 162 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

PESS
Categories

Potential Exposures Identified in Specific
Subpopulations and Incorporated into
Exposure Assessment

Potential Sources of Biological
Susceptibility Identified and
Incorporated into Hazard Assessment



increased exposures for tribes due to geographic
proximity (Section 2.3).



Socio-

demographic
Factors

EPA did not identify specific sociodemographic
factors that influence exposure to 1,4-dioxane.
This is a remaining source of uncertainty.

EPA did not identify sociodemographic
factors that influence biological
susceptibility to 1,4-dioxane. This is a
remaining source of uncertainty.

Nutrition

EPA did not identify nutritional factors
influencing exposure to 1,4-dioxane. This is a
remaining source of uncertainty.

EPA did not identify nutritional factors
that influence biological susceptibility to
1,4-dioxane. This is a remaining source of
uncertainty.

Genetics

EPA did not identify genetic factors influencing
exposure to 1,4-dioxane. This is a remaining
source of uncertainty.

Indirect evidence that genetic variants may
increase susceptibility of the target organ
was addressed through a 10* UF for
human variability. The magnitude of the
impact of genetic variants is unknown and
is a source of uncertainty.

Unique
Activities

Some tribes may have increased exposure to
drinking water due to tribal activities such as
sweat lodges. EPA has identified upper bound
drinking water estimates of 2 to 4 L/day
associated with tribal lifeways for some tribes
(Harper, : iarper and Ranco, 2009; Harper
et aL 2007; Harper et aL 2002). Risk
calculations in this draft supplement assume an
acute adult drinking water intake of 3.2 L/day
and a chronic drinking water intake of 0.88
L/day. Other potential sources of increased
exposure to 1,4-dioxane due to specific tribal
lifeways or other unique activity patterns are a
source of uncertainty.

EPA did not identify unique activities that
influence susceptibility to 1,4-dioxane.
This is a remaining source of uncertainty.

Aggregate
Exposures

EPA evaluated risk from aggregate sources of
exposure contributing to 1,4-dioxane in water
(Section 5.2.2.1) or from multiple sources in
proximity releasing to air (Section 5.2.2.3,
Appendix J.4). Risks from aggregate exposures
across routes or pathways were evaluated
qualitatively and are a remaining source of
uncertainty.

EPA does not identify ways that aggregate
exposures would influence susceptibility to
1,4-dioxane. This is a remaining source of
uncertainty.

Other

Chemical and
Non-chemical
Stressors

EPA did not identify chemical and nonchemical
factors influencing exposure to 1,4-dioxane. This
is a remaining source of uncertainty.

EPA did not identify chemical or
nonchemical factors that influence
susceptibility to 1,4-dioxane. There is
insufficient data to quantitatively address
potential increased susceptibility due to
chemical or nonchemical stressors and this
is a remaining source of uncertainty.

Page 163 of 484


-------
4778

4779

4780

4781

4782

4783

4784

4785

4786

4787

4788

4789

4790

4791

4792

4793

4794

4795

4796

4797

4798

4799

4800

4801

4802

4803

4804

4805

4806

4807

4808

4809

4810

4811

4812

4813

4814

4815

4816

4817

4818

4819

4820

4821

4822

4823

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

5.2.2.5 Aggregate and Sentinel Exposures

In this draft supplement, EPA considers the combined 1,4-dioxane exposure an individual may experience
due to releases from multiple facilities in proximity releasing to air or multiple releases contributing to
drinking water concentrations in a particular location. For general population drinking water exposure
scenarios, EPA evaluated combined exposure and risks from multiple sources of 1,4-dioxane in surface
water, including direct and indirect industrial releases, DTD releases, and upstream background
contamination (Section 5.2.2.1). For general population air exposure scenarios, EPA evaluated
combined exposure and risk across multiple facilities in proximity releasing to air (Section 5.2.2.3 and
Appendix J.4).

EPA considered aggregating cancer risks across inhalation, oral, and/or dermal routes of exposure.
There is uncertainty around the extent to which cancer risks across routes are additive for 1,4-dioxane.
Liver tumors are the primary site of cancer risk from oral exposures. Inhalation exposure in rats is
associated with multiple tumor types, including liver. The IUR used to calculate inhalation cancer risk
reflects combined risks from multiple tumor types. While EPA concluded that nasal cavity lesions are
likely to be primarily the result of systematic delivery (as discussed on p. 192 of the 2020 RE), there is
uncertainty around the degree to which those effects could be partially due to portal of entry effects
following inhalation exposure. It is therefore unclear the extent to which it is appropriate to
quantitatively aggregate cancer risks based on the IUR with liver tumor risks associated with oral or
dermal exposures. EPA considers the potential aggregate cancer risk across routes to be a source of
uncertainty for 1,4-dioxane cancer risk estimates.

EPA also considered aggregating cancer risks across dermal and oral exposures. The dermal cancer
slope factor is derived from the oral cancer slope factor by route-to-route extrapolation. Because the
systemic effect is assumed to be the same for both routes, EPA determined that it could be biologically
appropriate to aggregate risk from dermal and oral exposures. General population scenarios included
inhalation and oral not dermal exposures and occupational and consumer exposure scenarios included
inhalation and dermal not oral exposures. However, this draft supplement does not include COUs or
pathways in which both oral and dermal exposure routes are considered.

EPA also considered potential for aggregate exposures across groups. For example, there may be some
individuals who are exposed at work as well as through general population air and drinking water
pathways or through consumer product use. This as a source of uncertainty. These types of aggregate
risks were not quantified and risks for individual exposure scenarios should be interpreted with an
appreciation for potential aggregate exposures and risks.

Section 2605(b)(4)(F)(ii) of TSCA requires 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. 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 § 702.33)." In this draft supplement, EPA considered sentinel
exposures by considering risks to populations who may have upper bound exposures, including workers
and ONUs who perform activities with higher exposure potential and fenceline communities. EPA
characterized high-end exposures in evaluating exposure using both monitoring data and modeling
approaches. Where statistical data are available, EPA typically uses the 95th percentile value of the
available dataset to characterize high-end exposure for a given COU.

Page 164 of 484


-------
4824

4825

4826

4827

4828

4829

4830

4831

4832

4833

4834

4835

4836

4837

4838

4839

4840

4841

4842

4843

4844

4845

4846

4847

4848

4849

4850

4851

4852

4853

4854

4855

4856

4857

4858

4859

4860

4861

4862

4863

4864

4865

4866

4867

4868

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

5.2.2.6 Summary of Overall Confidence and Remaining Uncertainties in Human
Health Risk Characterization

The overall level of confidence in each set of risk estimates depends on the level of confidence in the
underlying hazard values summarized in Section 4.3 and the level of confidence in exposure estimates
described in more detail in Section 3.3.

For all risk estimates, EPA has medium to high confidence in the underlying hazard PODs used as the
basis for this risk characterization. Sources of confidence in each of the hazard values were described in
the 2020 RE and are summarized in Section 4.3. Cancer risk is the primary risk driver for each of the
scenarios evaluated in this draft supplement and is therefore the basis of overall confidence levels
described herein. There is remaining uncertainty for all risk estimates around the potential impact of 1,4-
dioxane on potentially exposed or susceptible subpopulations (as discussed in Section 5.2.2.4). EPA
applied an intraspecies uncertainty factor of 10 to all non-cancer PODs to account for variation across
gender, age, health status, or genetic makeup, and other factors that may increase susceptibility, but the
actual magnitude of the impact of these factors on susceptibility is unknown. Similarly, EPA evaluated
risks from exposure scenarios that are intended to reflect the most highly exposed populations (including
fenceline communities and highly exposed workers), but the potential for other highly exposed
populations that were not identified in this analysis is a source of uncertainty. Potential for aggregate
risks across routes or pathways that are not quantitatively evaluated in this assessment is another source
of uncertainty.

5.2.2.6.1 Risks from Occupational Exposures

Overall confidence in risk estimates for occupational inhalation exposures ranges from low to high,
depending on the confidence in exposure assessment for each OES/COU. As described in Section 4.3,
overall confidence in the cancer inhalation unit risk underlying these risk estimates is high. As described
in Section 3.3.1.1, the measured and monitored inhalation exposure data are supported by moderate to
robust evidence. Additionally, the exposure modeling methodologies and underlying model input data is
supported by moderate to robust evidence. However, there is uncertainty in the representativeness of the
assessed exposure scenarios towards all potential exposures for the given OES/COU, limitations in the
amount and age of monitoring data, and limitations in the modeling approaches towards 1,4-dioxane-
specific use within the OES/COU. Therefore, while the underlying data and methods used to estimate
occupational inhalation risk is supported by moderate to robust evidence, the overall confidence of these
estimates ranges from low to high depending on the OES/COU. OES/COU-specific discussions of the
available inhalation exposure data and overall confidence are presented in Appendix F.6 and
summarized in Table 3-2.

Overall confidence in risk estimates for occupational dermal exposures is medium for all OES/COUs
because the same modeling approach was used for all OES/COUs. As described in Section 4.3 overall
confidence in the oral and dermal cancer slope factor underlying these risk estimates is medium-high. As
described in Section 3.3.1.2, the dermal exposure modeling methodology is supported by moderate
evidence, with model input parameters from literature sources, a European model, standard defaults
from the ChemSTEER User Guide (	015a). and 1,4-dioxane product concentration data from

process information. These sources range from slight to robust, depending on factors such as age and
applicability to OES/COU. The modeling is limited by the use of standard input parameters that are not
specific to 1,4-dioxane and a lack of variability in dermal exposure for different worker activities.
Therefore, EPA's overall confidence in the occupational dermal risk estimates is medium

Page 165 of 484


-------
4869

4870

4871

4872

4873

4874

4875

4876

4877

4878

4879

4880

4881

4882

4883

4884

4885

4886

4887

4888

4889

4890

4891

4892

4893

4894

4895

4896

4897

4898

4899

4900

4901

4902

4903

4904

4905

4906

4907

4908

4909

4910

4911

4912

4913

4914

4915

4916

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

5.2.2.6.2 Risks from General Population Exposures through Drinking Water

Overall confidence in the overall distribution of risk estimates for drinking water exposures resulting
from facility releases is medium-high. Overall confidence in site-specific risk estimates for individual
facility releases varies both within and across OES, depending on the confidence in the source-specific
release data. As described in Section 4.3 overall confidence in the oral and dermal cancer slope factor
underlying these risk estimates is medium-high. As described in Section 3.3.2.1, the exposure modeling
methodology used for this analysis is supported by moderate evidence. It is designed to estimate water
concentrations expected at specific locations. Exposure estimates for this scenario are based on some
conservative assumptions about flow rates and release frequency and amount. A summary of sources of
flow and release data for facility release modeling is presented in Table 2-6. Available monitoring data
confirm that 1,4-dioxane is present in some surface water and drinking water, though most of the
available data were not collected near release sites are therefore not directly comparable. The overall
level of confidence depends on the source of OES/COU-specific release data described in Appendix E.3:

•	Overall confidence in drinking water exposure estimates is medium to high for OESs/COUs that
rely primarily on release data reported to DMR or to TRI via Form R. Most COUs/OESs are
included in this group.

•	Overall confidence in drinking water exposure estimates is medium for OESs/COUs for which
release estimates are based on reporting to TRI via Form A. The Import and Repackaging OES
releases used in this analysis are entirely based on Form A reporting of releases, and just under
half of the Industrial Uses OES releases were reported via Form A.

•	Overall confidence in drinking water exposure estimates is low to medium for OESs/COUs for
which release estimates are based on surrogate or modeled information.

Overall confidence in risk estimates for drinking water exposures resulting from DTD releases under
varying population and stream flow conditions is medium. As described in Section 4.3 overall
confidence in the oral and dermal cancer slope factor underlying these risk estimates is medium-high. As
described in Section 3.3.2.1, the exposure modeling methodology used for this analysis is supported by
robust evidence and is designed to provide a nationally representative distribution of estimated water
concentrations under varying conditions. Exposure estimates rely on estimated distributions of DTD
releases of consumer and commercial products for each COU.

Overall confidence in risk estimates for drinking water exposures resulting from hydraulic fracturing
releases is medium. As described in Section 4.3 overall confidence in the oral and dermal cancer slope
factor underlying these risk estimates is medium-high. As described in Section 3.3.2.1, the exposure
modeling methodology used for this analysis is supported by robust evidence and is designed to provide
a nationally representative distribution of estimated water concentrations under varying conditions.
Releases used as inputs in the model were estimated using Monte Carlo modeling that captures
variability across sites. However, the modeled exposure estimates are not directly tied to specific
releases at known locations, decreasing the strength of the evidence related to the representativeness of
the exposure estimates for actual exposures.

Overall confidence in risk estimates for drinking water exposures resulting from for aggregate surface
water concentrations predicted by probabilistic modeling varies across OES/COU. As described in
Section 4.3 overall confidence in the oral and dermal cancer slope factor underlying these risk estimates
is medium-high. As described in Section 3.3.2.1, the exposure modeling methodology used for this
analysis is supported by robust evidence and is designed to provide a nationally representative
distribution of estimated water concentrations under varying conditions. Available monitoring data
confirm that 1,4-dioxane is present in some surface water and drinking water, though most of the
available data were not collected near release sites and are therefore not directly comparable. In release

Page 166 of 484


-------
4917

4918

4919

4920

4921

4922

4923

4924

4925

4926

4927

4928

4929

4930

4931

4932

4933

4934

4935

4936

4937

4938

4939

4940

4941

4942

4943

4944

4945

4946

4947

4948

4949

4950

4951

4952

4953

4954

4955

4956

4957

4958

4959

4960

4961

4962

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

locations where monitoring data are available, case studies demonstrate strong concordance between
modeled estimates and measured surface water concentrations. There is some uncertainty around the
proximity of releases to drinking water intake locations and the extent to which 1,4-dioxane is further
diluted prior to reaching intake locations. The characterization of downstream dilution presented in
5.2.2.1.2 for individual facility releases illustrates the extent to which downstream dilution may impact
overall risk estimates. The overall level of confidence in resulting exposure estimates depends on the
source of OES/COU-specific release data described in Appendix E.3:

•	Overall confidence in drinking water exposure estimates is medium to high for OESs/COUs that
rely primarily on release data reported to DMR or to TRI via Form R. Most COUs/OESs are
included in this group.

•	Overall confidence in drinking water exposure estimates is medium for OESs/COUs for which
release estimates are based on reporting to TRI via Form A. The Import and Repackaging OES
releases used in this analysis are entirely based on Form A reporting of releases, and just under
half of the Industrial Uses OES releases were reported via Form A.

•	Overall confidence in drinking water exposure estimates is low to medium for OESs/COUs for
which release estimates are based on surrogate or modeled information.

5.2.2.6.3	Risks from General Population Exposures through Groundwater and
Land Disposal Pathways

Overall confidence in risk estimates for drinking water exposures resulting from disposal to landfills is
low to medium. As described in Section 4.3 overall confidence in the oral and dermal cancer slope
factor underlying these risk estimates is medium-high. As described in Section 3.3.2.2.1 the modeling
methodology is robust. However, the release information relied on as model input data is supported by
slight to moderate evidence. In addition, this drinking water exposure scenario relies on the assumption
that the estimated groundwater concentrations may occur in locations where groundwater is used as a
primary drinking water source. While the substantial uncertainty around the extent to which these
exposures occur decreases overall confidence in the exposure scenario, this scenario represents a
sentinel exposure.

Overall confidence in risk estimates for drinking waters resulting from disposal of hydraulic fracturing
waste is low to medium. As described in Section 4.3 overall confidence in the oral and dermal cancer
slope factor underlying these risk estimates is medium-high. As described in Section 3.3.2.2.2, the
modeling methodology is robust, and the release information relied on as model input data is supported
by moderate evidence. However, no monitoring data are available to confirm detection of 1,4-dioxane in
groundwater near hydraulic fracturing operations. This drinking water exposure scenario relies on the
assumption that the estimated groundwater concentrations may occur in locations where groundwater is
used as a primary drinking water source. While the substantial uncertainty around the extent to which
these exposures occur decreases overall confidence in the exposure scenario, this scenario represents a
sentinel exposure.

5.2.2.6.4	Risks from General Population Exposures through Air

Overall confidence in risk estimates for inhalation exposure resulting from industrial releases varies
across COUs. As described in Section 4.3, overall confidence in the cancer inhalation unit risk
underlying these risk estimates is high. As described in Section 3.3.3.1, the AERMOD modeling
methodology used for this analysis is robust. The exposure scenarios considered are most relevant to
long-term residents in fenceline communities. There is some uncertainty around the extent to which
people actually live and work around the specific facilities where risks are highest, decreasing overall
confidence in the exposure scenario. Overall confidence varies due to variable levels of confidence in

Page 167 of 484


-------
4963

4964

4965

4966

4967

4968

4969

4970

4971

4972

4973

4974

4975

4976

4977

4978

4979

4980

4981

4982

4983

4984

4985

4986

4987

4988

4989

4990

4991

4992

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

underlying release information used to estimate exposures. An OES-specific discussion of the
confidence in sources of release information is presented in Appendix E.5.4, but in general terms:

•	Overall confidence in risk estimates is medium to high for OESs/COUs that rely primarily on
release data reported to TRI via Form R.

•	Overall confidence in risk estimates is medium for OESs/COUs for which release estimates are
based on data reported to TRI via Form A.

•	Overall confidence in risk estimates is low to medium for OESs/COUs for which release
estimates are based on surrogate or modeled information.

Overall confidence in risk estimates for inhalation exposures resulting for air concentrations modeled
based on releases from hydraulic fracturing operations is medium. As described in Section 4.3, overall
confidence in the cancer inhalation unit risk underlying these risk estimates is high. As described in
Section 3.3.3.2 the modeling methodologies used to estimate air concentrations are robust. The
distribution of air releases used as model input data were estimated using Monte Carlo modeling and
rely on assumptions. No air monitoring data were available to confirm detection of 1,4-dioxane is air
near hydraulic fracturing operations. Because the air concentrations underlying this analysis are based
on probabilistic modeling, they are not tied to specific locations that can be evaluated for land use
patterns. There is therefore substantial uncertainty around the extent to which people actually live and
work around the specific locations where risks are highest, decreasing overall confidence in the
exposure scenario.

Overall confidence in risk estimates from inhalation exposures resulting from industrial and institutional
laundries is medium. As described in Section 4.3, overall confidence in the cancer inhalation unit risk
underlying these risk estimates is high. As described in Section 3.3.3.2, the modeling methodologies are
robust. The distribution of air releases used as model input data were estimated using Monte Carlo
modeling and rely on assumptions. No air monitoring data were available to determine whether 1,4-
dioxane is detected near industrial and institutional laundry facilities. Because the air concentrations
underlying this analysis are based on probabilistic modeling, they are not tied to specific locations that
can be evaluated for land use patterns. There is therefore substantial uncertainty around the extent to
which people actually live and work around the specific locations where risks are highest, decreasing
overall confidence in the exposure scenario.

Page 168 of 484


-------
4993

4994

4995

4996

4997

4998

4999

5000

5001

5002

5003

5004

5005

5006

5007

5008

5009

5010

5011

5012

5013

5014

5015

5016

5017

5018

5019

5020

5021

5022

5023

5024

5025

5026

5027

5028

5029

5030

5031

5032

5033

5034

5035

5036

5037

5038

5039

5040

5041

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

REFERENCES

Adam son. DT; Mahendra. S; Walker. KL. Jr.; Ranch. SR; Sengupta. S; Newell. CI. (2014). A multisite
survey to identify the scale of the 1,4-dioxane problem at contaminated groundwater sites.
Environ Sci Technol Lett 1: 254-258. http://dx.doi.ore/10.1021/ez500092u
Adam sob iH Pifia. EA; Cartwrigln \« Kauch. SR; Anderson. RH; Moln « t minor. J A. (2017). 1.4-
Dioxane drinking water occurrence data from the third unregulated contaminant monitoring rule.
Sci Total Environ 596-597: 236-245. http://dx.doi.o	/i.scitotenv.2017.04.085

Analvtica. (2015). Choosing an appropriate distribution [Website],

http://wiki.analytica.com/Choosine an appropriate distribution
AT SDR. (2012). Toxicological profile for 1,4 dioxane [ATSDR Tox Profile], Atlanta, GA: U.S.
Department of Health and Human Services, Public Health Service.
http://www.atsdr.cdc. eov/toxprofiles/tp.asp?id=955&tid=l 99
Baldwin. PE; Mayn	(1998). A survey of wind speed in indoor workplaces. Ann Occup Hyg 42:

303-313. http://dx.doi.c	5/80003-4878(98)00031-3

Bel anger. PL. (1980). Health hazard evaluation report no. HHE-80-21-721, Colgate-Palmolive Co.,
Berkeley, California. (HHE-80-21-721). Cincinnati, OH: National Institue for Occupational
Safety and Health.

Black. RE; Hurley. FJ; Havery. DC. (2001). Occurrence of 1,4-dioxane in cosmetic raw materials and
finished cosmetic products. J AOAC Int 84: 666-670.

Brouehtoii Sepulveili \ I < ter. K; Kail I' x ickelsen. MG; Gillan. M; Mohr. TKG. (2019). 1.4-
Dioxane: Emerging technologies for an emerging contaminant. Remediation Journal 29: 49-63.
http://dx.doi.ore/10.1002/rem.. I I
Brunswick County. (2022). Brunswick County, North Carolina: Annual reports/facilities information
[Website], https://www.bmnswickcountync.eov/utilities/annual-report.s-facilities-information/
CA Water Board. (2022). EDT library and water quality analyses data and download page [Website],

https://www.waterboards.ca.eov/drinkine water/certlic/drinkinewater/EDTlibrary.html
Commonwealth of Massachusetts. (2018). Energy & Environmental Affairs Data Portal: Search for

drinking water [Website], https://eeaonline.eea.state.ma.us/portal#!/search/drinkine-water
Cordone. L; Carlson (*, Plaehn. W; Shanera^\ « Wilmoth. D. (2016). Case study and retrospective:
Aerobic fixed film biological treatment process for 1,4-dioxane at the Lowry landfill superfund
site. Remediation Journal 27: 159-172. http://dx.doi.ore	em.21502

Davarani. SSH; Masoomi. L; Banitaba. IV id. HRLZ; Sadeehi. O; Samiei. A. (2012). A New

Aluminium Hydroxide Coating on Fused Silica Fiber for the Determination of 1,4-Dioxane in
Surfactants and Detergents Using HS-SPME-GC. Chromatographia 75: 371-377.
http://dx.doi.orE

Forkner. MW; Robson. J.H; Snellines. WM; Martin. AE; Mum	rsons. TE. (2004). Kirk-

Othmer Encyclopedia of Chemical Technology
Glycols. Hoboken, New Jersey: John Wiley & Sons, Inc.

http://dx.doi.ore 10 100. 0471238961.05200V-0 hi\| j.a0i.pub2
Franklin Associates. (2020). Cradle-to-resin life cycle analysis of polyethylene terephthalate resin (final
revised report). Charlotte, NC: The National Association for PET Container Resources
(NAPCOR).

Franz nnett. S; DeLeo. PC; Collatz. M; Kelly. K; Nekoomaram. J; Wieroniey. S. (2015).
Comments of the Adhesive and Sealant Council, the American Coatings Association, the
American Chemistry Council, the American Cleaning Institute, the Consumer Specialty Products
Association, and Waste Management on the 1,4-dioxane problem formulation and initial
assessment. (EPA-HQ-OPPT-2015-0078). Franz, C; Bennett, S; DeLeo, PC; Collatz, M; Kelly,
K; Nekoomaram, J; Wieroniey, S. https://arasp.americanchemistry.com/Media-Center/Industry-
Comm ents-June-2015 .pdf

Page 169 of 484


-------
5042

5043

5044

5045

5046

5047

5048

5049

5050

5051

5052

5053

5054

5055

5056

5057

5058

5059

5060

5061

5062

5063

5064

5065

5066

5067

5068

5069

5070

5071

5072

5073

5074

5075

5076

5077

5078

5079

5080

5081

5082

5083

5084

5085

5086

5087

5088

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Gillespie. BW; Chen. Q; Reichca IS, 1'irmzblau. A; Hedeeman 1,1 epkowski > \f08
Godri Pollitt. KJ; Kim. J.H; Peccia. J; Elimelech. M; Zhang. t barkoftali u, Hodges. B; Zucker. I;
Huang. H; Deziel. NC: Murphy. K; Ishii. M; Johnson. CH; Boissevairi \ U'Keet'1 I \iustas.
PT; Orlickv. D; Thompson. DC; Vasiliou. ¥. (2019). 1,4-Dioxane as an emerging water
contaminant: State of the science and evaluation of research needs [Review], Sci Total Environ
690: 853-866. http://dx.doi.org/10.1016/i.scitotenv.20 r" 0 > i i;

Harlev. KG: Calderon. L; Nolan. JES; Maddalena. R; Russell. M; Roman. K; Mayo-Burgos. S: Cabrera.
J: Morga. N: Bradman. A. (2021). Changes in Latina women's exposure to cleaning chemicals
associated with switching from conventional to "green" household cleaning products: The
LUCIR intervention study. Environ Health Perspect 129: 97001.
http://dx.doi.orE 39/EHP8831
Harpi (2017). Shoshone-Bannock exposure scenario for use in risk assessment. Harper, B.

https://superfund.oregonstate.edu/sites/superfund.oregonstate.edu/files/shoshone bannock scena
rio 2017.pdf

Harptn Tn kanc<> 1 * (2009). Wabanaki traditional cultural lifeways exposure scenario. U.S.
Environmental Protection Agency and the Tribal Nations in Maine.

https://www.epa.gov/sites/default/files/2015-08/documents/ditca.pdf
Harptn r% 1 i _ H, I. Harris. S: Abevta. C: Kirschner. F. (2002). The Spokane Tribe's multipathway
subsistence exposure scenario and screening level RME. Risk Anal 22: 513-526.
http://dx.doi.orE	>272-4332.00047

Harpi	'ding. AK: Waterhous. T: Harris. SG. (2007). Traditional tribal subsistence exposure

scenario and risk assessment guidance manual. (Grant Number EPA-STAR-J1-R831046).
Corvallis, OR: Oregon State University Department of Public Health and Confederated Tribes of
the Umatilla Indian Reservation.

https://health.oregonstate.edu/sites/health.oregonstate.edu/files/research/pdf/tribal-
grant/exposure scenario a : guidance manual v2.pdf
HHS. (2016). 1,4-Dioxane: CAS No. 123-91-1. In Report on Carcinogens (14th ed.). Research Triangle
Park, NC: National Toxicology Program, Department of Health and Human Services.

https://ntp.niehs.nih.gov/ntp/roc/content/profiles/dioxane.pdf
Hills. B: Ktincewicz. S: Blade. LM: Sack. I ) (1989). Health hazard evaluation report no. HETA-87-
367-1987, BMY Corporation, A Division of Harsco Corporation, York, Pennsylvania. (HETA-
87-367-1987). Cincinnati, OH: National Institue for Occupational Safety and Health.
https://www.cdc.gov/niosh/hhe/reports/pdfs/1987-0. 87.pdf
Jackson. LE: Lemkt (2019). Evidence for natural attenuation of 1,4-dioxane in a glacial aquifer

system. Hydrogeology Journal v. 27: 3009-3024. http://dx.doi.< 10 100 1 100 10 0 r"-02028-6
Kano. H: Umeda. Y: Kasai. T: Sasaki. T: Matsumoto. M: Yamazaki. K: Nagano. to. H:

Fukushima. S. (2009). Carcinogenicity studies of 1,4-dioxane administered in drinking-water to
rats and mice for 2 years. Food Chem Toxicol 47: 2776-2784.
http://dx.doi.orE 10 101 | ict.2009.0N 01 J
Kasai. T: Kano. H: Umeda. Y: Sasaki wa. N: Nishizawa. T: Nagano. K: Arito. H: Nagashima. H:
Fukushima. S. (2009). Two-year inhalation study of carcinogenicity and chronic toxicity of 1,4-
dioxane in male rats. Inhal Toxicol 21: 889-897. http://dx.doi.org/10 10SQ/0895837080262( \ h1
Kasti	iller. MA. (2006). Kinetics of finite dose absorption through skin 2: Volatile

compounds. J Pharm Sci 95: 268-280. http://dx.doij	>s.20497

Page 170 of 484


-------
5089

5090

5091

5092

5093

5094

5095

5096

5097

5098

5099

5100

5101

5102

5103

5104

5105

5106

5107

5108

5109

5110

5111

5112

5113

5114

5115

5116

5117

5118

5119

5120

5121

5122

5123

5124

5125

5126

5127

5128

5129

5130

5131

5132

5133

5134

5135

5136

5137

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Lee. CS; Venkatesan. A.K; Walker. H )ler. CI. (2020). Impact of groundwater quality and

associated byproduct formation during UV/hydrogen peroxide treatment of 1,4-dioxane. Water
Res 173: 115534. http://dx.doi.ore/10.1016/j.watres.2020.115534
Leste	inn an. EM; Sitterlev. oral	¦ inden. KG. (2015).

Characterization of hydraulic fracturing flowback water in Colorado: Implications for water
treatment. Sci Total Environ 512-513: 637-644.
http://dx.doi.org 10 101 i.scitotenv.20I 01 0 l'<

Li. M; Mathieu. J; Yarn \ t torenza. S: Dem \ lie. Z; Zhou j,_ \lvare (2013). Widespread

distribution of soluble di-iron monooxygenase (SDIMO) genes in Arctic groundwater impacted
by 1,4-dioxane. Environ Sci Technol 47: 9950-9958. http://dx.doi.org/10.102 1 /es402228x
Li. M; Van Orden. ET; Devries. DJ; Xione. Z: Hinchee. R; Ah o I (2015). Bench-scale

biodegradation tests to assess natural attenuation potential of 1,4-dioxane at three sites in
California. Biodegradation 26: 39-50. http://dx.doi.c
Lin. WT; Chen. WL; Cheng. V ang. H.C; Tsai. SW. (2017). Determining the residual

characteristics of alkylphenols, arsenic, and lead as well as assessing the exposures of 1,4-
dioxane from household food detergents. J AOAC Int 100: 1086-1093.
http://dx.doi.ore/10.5740/iaoacint. 16-0380
Makino. R; Kawasaki. H; Kishimoto. A; Gamo. M; Nakanishi. J. (2006). Estimating health risk from

exposure to 1,4-dioxane in Japan. Environ Sci 13: 43-58.

Marquart. H; Franken. R; Goei	isman. W; Schinkel. J. (2017). Validation of the dermal

exposure model in ECETOC TRA. Ann Work Expo Health 61: 854-871.
http://dx.doi.ore/10.1093/annweh/wxx059
Mattie. PR; Bucher. TW; Carter. AL; Stoffreeen. DE; Reboulet. IE. (2012). Acute inhalation toxicity

study of 1, 4-dioxane in rats (Rattus norvegicus). GRA and I: 29.

McDaniel. MP; DesLauriei (2015). Kirk-Othmer encyclopedia of chemical technology
Ethylene polymers, HDPE. Hoboken, NJ: Wiley-Interscience.

http://dx.doi.ore 10 1002/0 i i. 3961.08090708 I I0"l"l" jQl wb3
Millard. SP. (2013). EnvStats: An R package for environmental statistics. New York, NY: Springer.

http://dx.doi.ore/10J007/9 • I I ti - h I
Minitab. (2022). A comparison of the Pearson and Spearman correlation methods [Website],
https://support.minitab.com/en-us/minitab/21/help-and~how-to/statistics/basic-
statistics/supporting-topics/correlation-and-covariance/a-comparison-of-the-pearson-and-
spearman-correlation-methods/

Mohi	seppi. WH. (2010). Case Studies of 1,4-Dioxane Releases, Treatment, and Drinking

Water Contamination. In T Mohr; B Diguiseppi; JK Anderson; JW Hatton (Eds.), (2nd ed., pp.
369-420). Boca Raton, FL: CRC Press. huf; 'ivdoi.oa	s < 06629-c8

N	einer. E. (2020). Kirk-Othmer Encyclopedia of Chemical Technology

Detergency and detergents [Encyclopedia], New York, NY: John Wiley & Sons.

http://dx.doi.ore 10 1002/0 i i. 3961.04052005122 M II UOt |':iib3
NIOSH. (1987). NIOSH Guide to Industrial Respiratory Protection (pp. 87-1 16). (NIOSH/00177545).
Anonymous.

NTP. (1986). NTP Toxicology and carcinogenesis studies of dichloromethane (methylene chloride)
(CAS No. 75-09-2) in F344/N rats and B6C3F1 mice (inhalation studies). 306: 1-208.
https://heronet.epa.gov/heronet/index.cfm?action=search.view&reference id=l 03686800berle,
W. (2015). Monte Carlo simulations: Number of iterations and accuracy. (ARL-TN-0684).
Aberdeen Proving Ground, MD: US Army Research Laboratory.
https://apps.dtic.mil/sti/pdl0 \Vi \ _| |.i!f
OECD. (2009). Emission scenario documents on coating industry (paints, lacquers and varnishes).
(JT03267833). Paris, France.

Page 171 of 484


-------
5138

5139

5140

5141

5142

5143

5144

5145

5146

5147

5148

5149

5150

5151

5152

5153

5154

5155

5156

5157

5158

5159

5160

5161

5162

5163

5164

5165

5166

5167

5168

5169

5170

5171

5172

5173

5174

5175

5176

5177

5178

5179

5180

5181

5182

5183

5184

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

http://www.oecd. org/officialdocuments/publicdisplavdocumentpdf/?cote=env%20/im/mono(200

9)24& docl an guage=en

OECD. (201 la). Emission scenario document on coating application via spray-painting in the

automotive refinishing industry. In OECD Series on Emission Scenario Documents No 11.
(ENV/JM/MONO(2004)22/REV1). Paris, France: Organization for Economic Co-operation and
Development.

http://www.oecd. org/officialdocuments/publicdisplaydocumentpdf/?cote=env/im/mono(2004)22/
rev 1& docl an guage=en

OECD. (201 lb). Emission scenario document on the chemicals used in water-based washing operations
at industrial and institutional laundries. (JT03304937). Paris, France, https://www.oecd-
ilibrary.org/docserver/9y892642209y3~

en.pdf?e:	50&id=id&accname=guest&checksum=B8Cyi4E510603EB565B89A

8AE6A24C14

OECD. (2015). Emission scenario document (ESD) on industrial use of industrial cleaners.
(JT03373556). Paris, France.

http://www.oecd. org/officialdocuments/publicdisplaydocumentpdf/?cote=ENV/JM/MQNQ(2015
)3 & docl an guage=en

OECD. (2017). Emission scenario document (ESD) on the use of textile dyes. In Emission Scenario

Document, Series No 36. (ENV/JM/MONO(2015)50). Paris, France: OECD, Inter-Organization
Programme for the Sound Management of Chemicals (IOMC).

http://www.oecd.org/chemicalsafety/risk-assessment/emissionscenariodocuments.htm
OECD. (2020). Emission scenario document on chemical additives used in automotive lubricants. Paris,
France.

Oracle. (2017). Oracle® Crystal Ball User's Guide; 2017 (Release 11.1.2.4)

Selecting probability distributions. Austin, TX.

https ://docs. oracl e. com/cd/	^pasOS.html

Otto. M; Nagaraia. S. (2007). Treatment technologies for 1,4-Dioxane: Fundamentals and field

applications. Remediation Journal 17: 81-88. http://dx.doi.org/10.1002/rem.201 '< ?

Palisade. (2015a). Palisade Knowledge Base: 2.6. How many iterations do I need? [Website],

https://kb.palisade.com/index.php?pg=kb.page&id=125
Palisade. (2015b). Palisade Knowledge Base: 2.19. Random number generators [Website],

https://kb.palisade.com/index.php?pg=kb.page&id=25
Palisade. (2018). Palisade Knowledge Base: 2.18. Latin Hypercube versus Monte Carlo sampling

[Website], https://kb.palisade.com/index.php?pg=kb.page&id=28
Palisade. (2022a). Define correlation matrix [Website], https://help.palisade.eom/v8/en/@RISK/l-

Define/5-Correlation/Define-Correlation-Matrix.htm
Putz. VR; Johnson. BL; Setzer. JV. (1979). A comparative study of the effects of carbon monoxide and
methylene chloride on human performance. J Environ Pathol Toxicol 2: 97-112.
https://heronet.epa.gov/heronet/index.cfm?action=search.view&reference id=l 0626648 Saraii.
M; Shirvani. N. (2017). Determination of residual 1,4-dioxane in surfactants and cleaning agents
using headspace single-drop microextraction followed by gas chromatography-flame ionization
detection. Int. J Cosmet Sci 39: 36-41. http://dx.doi.org/10.1 1 1 1/ics. 12345
Stefl.	ge. KF. (2014). Kirk-Othmer Encyclopedia of Chemical Technology

Antifreezes and deicing fluids [Encyclopedia], New York, NY: John Wiley & Sons.

http://dx.doi.org/iOJ002/0 i l _ • x mm l >200919200506 jQl |':iib2
US JitSl (2016). May 2016 Occupational Employment and Wage Estimates: National Industry-
Specific Estimates [Website], http://www.bls.gov/oes/tables.htm

Page 172 of 484


-------
5185

5186

5187

5188

5189

5190

5191

5192

5193

5194

5195

5196

5197

5198

5199

5200

5201

5202

5203

5204

5205

5206

5207

5208

5209

5210

5211

5212

5213

5214

5215

5216

5217

5218

5219

5220

5221

5222

5223

5224

5225

5226

5227

5228

5229

5230

5231

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

IJ.S JitSl (2022). Labor Force Statistics from the Current Population Survey: Household data, Annual
averages, 11. Employed persons by detailed occupation, sex, race, and Hispanic or Latino
ethnicity [Numbers in thousands] [Website], https://www.bls.eov/cps/cpsaatl 1 .htm

sus Bureau. (2015). Statistics of U.S. Businesses (SUSB).
https://www.censiis.eov/data/tables/2015/econ/susb/2015-siisb-aimiial.html

(1988). National survey of solid waste (municipal) landfill facilities [EPA Report],
(EPA/530-SW88-034). Washington, DC: Office of Solid Waste and Emergency Response, U.S.
Environmental Protection Agency.

(1991). Chemical engineering branch manual for the preparation of engineering assessments.
Volume I. Ceb Engineering Manual. Washington, DC: Office of Pollution Prevention and
Toxics, US Environmental Protection Agency.

(1994a). Guidelines for Statistical Analysis of Occupational Exposure Data: Final. United
States Environmental Protection Agency :: U.S. EPA.

(1994b). Methods for derivation of inhalation reference concentrations and application of
inhalation dosimetry [EPA Report], (EPA600890066F). Research Triangle Park, NC.
https://crpub.epa.eov/ncea/risk/fecofdisplay.crm?deid=yi993«S	329&CFTOKEN=2

300b3 1 7

U.S. EPA. (2006a). 2006 community water system survey - Volume I: Overview [EPA Report], (EPA
815-R-09-001). Washington, DC: U.S. Environmental Protection Agency, Office of Water.
https://nepis.epa. eov/Exe/ZyPDF.cei?Dockey=P 1009JJI.txt
U.S. EPA. (2006b). Treatment technologies for 1,4-dioxane: Fundamentals and field applications. (EPA-
542-R-06-009). Washington, DC: Office of Solid Waste and Emergency Response.
http://costperformance.Org/remediation/pdf/I'<1 \ IV ^ ument of 1.4~Dioxane.pdf
U.S. EPA. (201 1). Exposure factors handbook: 201 1 edition [EPA Report], (EPA/600/R-090/052F).

Washington, DC: U.S. Environmental Protection Agency, Office of Research and Development,
National Center for Environmental Assessment.
https://nepis.epa. gov/Exe/ZyPURL.cgi?Dockey=P 100F2QS.txt
U.S. EPA. (2012). Advances in inhalation gas dosimetry for derivation of a reference concentration
(RfC) and use in risk assessment (pp. 1-140). (EPA/600/R-12/044). Washington, DC.
https://cfpub.epa.gov/ncea/risk/recordisplav.cfm?deid=244650	>	, CFTOKEN=

U.S. EPA. (2015a). ChemSTEER user guide - Chemical screening tool for exposures and environmental
releases. Washington, D C. https://www.epa.gov/sites/production/files/2Q15-

05/documents/user guide.pdf
U.S. EPA. (2015b). TSCA. work plan chemical problem formulation and initial assessment. 1,4-
Dioxane. (740-R1-5003). Washington, DC: Environmental Protection Agency, Office of
Chemical Safety and Pollution Prevention.
http://nepis.epa. eov/Exe/ZyPURL. cei?Dockev=P 1OOMDC1. TXT
U.S. EPA. (2016a). Hydraulic fracturing for oil and gas: Impacts from the hydraulic fracturing water
cycle on drinking water resources in the United States [EPA Report], (EPA/600/R-16/236F).
Washington, DC. https://cfpub.epa.eov/ncea/hfstudy/recordisplay.cfm?deid=332990
U.S. EPA. (2016b). Public database 2016 chemical data reporting (May 2017 release). Washington, DC:
US Environmental Protection Agency, Office of Pollution Prevention and Toxics.
https://www.epa.gov/chemical-data-reporting
U.S. EPA. (2017a). Information on the various spray polyurethane foam products. U.S. Environmental
Protection Agency, Design for the Environment.

https://www.epa.eov/sites/prodiiction/files/2015-08/documents/spfprodi: ;s.pdf

Page 173 of 484


-------
5232

5233

5234

5235

5236

5237

5238

5239

5240

5241

5242

5243

5244

5245

5246

5247

5248

5249

5250

5251

5252

5253

5254

5255

5256

5257

5258

5259

5260

5261

5262

5263

5264

5265

5266

5267

5268

5269

5270

5271

5272

5273

5274

5275

5276

5277

5278

5279

5280

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

U.S. EPA. (2017b). Preliminary information on manufacturing, processing, distribution, use, and

disposal: 1,4-dioxane. https://www.regulations.gov/document?D=] " \ ^ iiQ~QPPT~201 0 23-

0003

U.S. EPA. (2017c). Technical fact sheet 1,4-dioxane [EPA Report], (EPA-505/F-17/01 1). Office of
Land and Emergency Management, U.S. EPA.
https://nepis.epa. gov/Exe/ZyPURL.cgi?Dockev=P 100THWG.txt
U.S. EPA. (2018a). Application of spray polyurethane foam insulation - Generic scenario for estimating
occupational exposures and environmental releases -Draft. Washington, DC.
https://www.epa.gov/tsca-screening-tools/using-predictive-methods-assess-exposure-and-fate-
under-tsca

U.S. EPA. (2018b). Application of spray polyurethane foam insulation - Generic scenario for estimating
occupational exposures and environmental releases -Methodology review draft, revised.
https://www.epa.gov/tsca-screening-tools/chemsteer-chemical-screening-tool-exposures-and-
environmental-releases#genericscenarios
U.S. EPA. (2018c). 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/2Q18-
06/documents/final application of sr in tsc	if

U.S. EPA. (2018d). User's Guide for the AMS/EPA Regulatory Model (AERMOD). (EPA Document
Number: EPA-454/B-18-001). U.S. EPA.

(2019a). Exposure factors handbook chapter 3 (update): Ingestion of water and other select
liquids [EPA Report], (EPA/600/R-18/259F). Washington, DC.
https://cfpub. epa.gov/ncea/efp/recordisplav. cfm?deid=343661
U.S. EPA. (2019b). User's Guide: Integrated Indoor-Outdoor Air Calculator (IIOAC). Washington, DC:
U.S. EPA.

https://heronet.epa.gov/heronet/index.cfm?action=search.view&refere

(2020b). Final scope of the risk evaluation for 4,4'-(l-Methylethylidene)bis[2, 6-dibromophenol]
(TBBPA); CASRN 79-94-7. (EPA 740-R-20-008). Washington, DC: Office of Chemical Safety
and Pollution Prevention. https://www.regulations.gov/document/EPA-HQ-OPPT-2018-Q462-

0047

U.S. EPA. (2020c). Risk evaluation for 1,4-dioxane (CASRN: 123-91-1) [EPA Report], (EPA-740-R1-
8007). Washington, DC: Office of Chemical Safety and Pollution Prevention.

https://www.regiilations.gov/dociiment/EPA-H.Q-0]	0238-0092

U.S. EPA. (202 la). Draft systematic review protocol supporting TSCA risk evaluations for chemical
substances, Version 1.0: A generic TSCA systematic review protocol with chemical-specific
methodologies. (EPA Document #EPA-D-20-031). Washington, DC: Office of Chemical Safety
and Pollution Prevention. https://www.regiilations.gov/dociiment/EPA-HQ-OPPT-2'

0005

U.S. EPA. (202 lb). Use of additives in plastic compounding - Generic scenario for estimating

occupational exposures and environmental releases (Revised draft) [EPA Report], Washington,
DC: Office of Pollution Prevention and Toxics, Risk Assessment Division.

U.S. EPA. (2022a). Chemicals used in furnishing cleaning products - Generic scenario for estimating
occupational exposures and environmental releases (revised draft). Washington, DC: Office of
Pollution Prevention and Toxics.

U.S. EPA. (2022b). Commercial use of automotive detailing products - Generic scenario for estimating
occupational exposures and environmental releases (Methodology review draft) [EPA Report],
Washington, DC: U.S. Environmental Protection Agency, Office of Pollution Prevention and
Toxics, Risk Assessment Division.

https://heronet.epa.gov/heronet/index.cfm?action=search.view&reference id=l 048047211.8.

Page 174 of 484


-------
5281

5282

5283

5284

5285

5286

5287

5288

5289

5290

5291

5292

5293

5294

5295

5296

5297

5298

5299

5300

5301

5302

5303

5304

5305

5306

5307

5308

5309

5310

5311

5312

5313

5314

5315

5316

5317

5318

5319

5320

5321

5322

5323

5324

5325

5326

5327

5328

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

EPA. (2022d). Emission scenario document on chemicals used in hydraulic fracturing (draft). In
OECD Environmental Health and Safety Publications, Series on Emission Scenario Documents.
Paris, France: Organization for Economic Co-operation and Development.
https://heronet.epa.gov/heronet/index.cfm?action=search.view&reference id=10480474U.S.
EPA. (2022h). Use of laboratory chemicals - Generic scenario for estimating occupational
exposures and environmental releases (Revised draft generic scenario) [EPA Report],
Washington, DC: U.S. Environmental Protection Agency, Office of Pollution Prevention and
Toxics, Existing Chemicals Risk Assessment Division.

U.S. EPA. (2023a). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: 1,4-Dioxane Supplemental Information Files EWISRD-XL-R probabilistic
model code. Washington, DC: Office of Pollution Prevention and Toxics, Office of Chemical
Safety and Pollution Prevention.

U.S. EPA. (2023b). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: Air Exposure and Risk Estimates for 1,4-Dioxane Emissions from Hydraulic
Fracturing Operations. Washington, DC: Office of Pollution Prevention and Toxics, Office of
Chemical Safety and Pollution Prevention.

U.S. EPA. (2023c). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: Air Exposures and Risk Estimates for Industrial Laundry. Washington, DC:
Office of Pollution Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.

U.S. EPA. (2023d). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: Air Exposures and Risk Estimates for Multi-Year Analysis. Washington, DC:
Office of Pollution Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.

U.S. EPA. (2023e). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: Air Exposures and Risk Estimates for Single Year Analysis. Washington, DC:
Office of Pollution Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.

U.S. EPA. (2023f). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Land Releases
to Landfills. Washington, DC: Office of Pollution Prevention and Toxics, Office of Chemical
Safety and Pollution Prevention.

U.S. EPA. (2023g). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Land Releases
to Surface Impoundments. Washington, DC: Office of Pollution Prevention and Toxics, Office
of Chemical Safety and Pollution Prevention.

U.S. EPA. (2023h). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Release to
Surface Water from Individual Facilities. Washington, DC: Office of Pollution Prevention and
Toxics, Office of Chemical Safety and Pollution Prevention.

U.S. EPA. (2023i). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Surface Water
Concentrations Predicted with Probabilistic Modeling. Washington, DC: Office of Pollution
Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.

U.S. EPA. (2023j). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental
Information File: Environmental Releases to Air. Washington, DC: Office of Pollution
Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.

U.S. EPA. (2023k). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: Environmental Releases to Land for all OES Except Disposal. Washington,
DC: Office of Pollution Prevention and Toxics, Office of Chemical Safety and Pollution
Prevention.

Page 175 of 484


-------
5329

5330

5331

5332

5333

5334

5335

5336

5337

5338

5339

5340

5341

5342

5343

5344

5345

5346

5347

5348

5349

5350

5351

5352

5353

5354

5355

5356

5357

5358

5359

5360

5361

5362

5363

5364

5365

5366

5367

5368

5369

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

U.S. EPA. (20231). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: Environmental Releases to Land for the Disposal OES. Washington, DC:
Office of Pollution Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.

U.S. EPA. (2023m). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: Environmental Releases to Water for OES with TRI and DMR. Washington,
DC: Office of Pollution Prevention and Toxics, Office of Chemical Safety and Pollution
Prevention.

U.S. EPA. (2023n). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental
Information File: Environmental Releases to Water for OES without TRI or DMR data.
Washington, DC: Office of Pollution Prevention and Toxics, Office of Chemical Safety and
Pollution Prevention.

U.S. EPA. (2023o). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: EWISRDXL BrunswickCountyNC Case Study. Washington, DC: Office of
Pollution Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.

U.S. EPA. (2023p). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: EWISRDXL ColumbiaTN Case Study. Washington, DC: Office of Pollution
Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.

U.S. EPA. (2023q). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: EWISRDXL Liverpool OH Case Study. Washington, DC: Office of Pollution
Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.

U.S. EPA. (2023r). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: Occupational Exposure and Risk Estimates. Washington, DC: Office of
Pollution Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.

U.S. EPA. (2023s). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental

Information File: WQP Processed Surface Water Data. Washington, DC: Office of Pollution
Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.

U.S. EPA. (2023t). Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Systematic Review

Supplemental File: Data Quality Evaluation and Data Extraction Information for Environmental
Release and Occupational Exposure. Washington, DC: Office of Pollution Prevention and
Toxics, Office of Chemical Safety and Pollution Prevention.

https://heronet.epa. gov/heronet/index.cfm?action=search.view&referenceid=3419938W al a-
Jerzykiewicz, A; Szymanowski, J. (1998). Headspace gas chromatography analysis of toxic
contaminants in ethoxylated alcohols and alkylamines. Chromatographia 48: 299-304.

http://dx.doi.ore 10 100 n 0._ I S6

Youtiv dt i tin. WH; Gehriin P < Ucrvath. BS; Daniel. RL. (1976). 1,4-Dioxane and beta-

hydroxyethoxyacetic acid excretion in urine of humans exposed to dioxane vapors. Toxicol Appl
Pharmacol 38: 643-646. http://dx.doi.org/i0 J0 i6/00 II 008X(76)90195-2

Yotffly 11 * Braun. WH; Ramp\ 1 \\ t lienowetli > 'Lm. Gl". (1977). Pharmacokinetics of 1,4-
dioxane in humans. J Toxicol Environ Health 3: 507-520.
http://dx.doi.orE 30/15287397709529583

Page 176 of 484


-------
5370

5371

5372

5373

5374

5375

5376

5377

5378

5379

5380

5381

5382

5383

5384

5385

5386

5387

5388

5389

5390

5391

5392

5393

5394

5395

5396

5397

5398

5399

5400

5401

5402

5403

5404

5405

5406

5407

5408

5409

5410

5411

5412

5413

5414

5415

5416

5417

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

APPENDICES

Appendix A ABBREVIATIONS AND ACRONYMS

30Q5

Lowest 30-day average flow that occurs in a 5-year period

7Q10

Lowest 7-day average flow that occurs in a 10-year period

AC

Acute concentrations

ACA

American Coatings Association

ACGM

American Conference of Governmental Industrial Hygienists

ADC

Average daily concentration

ADD

Average daily dose

ADR

Acute Dose Rate

AEC

Acute Exposure Concentration

APF

Assigned protection factor

ASTDR

Agency for Toxic Substances and Disease Registry

BHET

Bishydroxyethyl terephthalate

BLS

Bureau of Labor Statistics

BMD

Benchmark dose

BMDL

Benchmark dose level

CASRN

Chemical Abstracts Service Registry Number

CDR

Chemical Data Reporting

CEB

Chemical Engineering Branch

CEHD

Chemical Exposure Health Data

CERCLA

Comprehensive Environmental Response, Compensation and Liability Act

CFR

Code of Federal Regulations

COU

Condition of use

CSF

Cancer slope factor

CT

Central tendency

CWA

Clean Water Act

DAF

Dilution attenuation factor

DHHR

Department of Health and Human Services

DIY

Do-it-yourself

DMR

Discharge monitoring report

DRAS

Delisting Risk Assessment Software

DTD

Down-the-drain

DWI

Drinking water intake

DWT

Drinking water treatment

ECHA

European Chemicals Agency

ECHO

Environmental Compliance History Online database

EPA

Environmental Protection Agency

EPACMTP

Environmental Protection Agency Composite Model for Leachate Migration with



Transformation Products

EPCRA

Emergency Planning and Community Right-to-Know Act

EROM

Enhanced Runoff Method (database)

ESD

Emission Scenario Document

FRS

Facility Registry Service

FT

Full-text (screening)

GS

Generic Scenario

HAWC

Health Assessment Workplace Collaborative (tool)

Page 177 of 484


-------
5418

5419

5420

5421

5422

5423

5424

5425

5426

5427

5428

5429

5430

5431

5432

5433

5434

5435

5436

5437

5438

5439

5440

5441

5442

5443

5444

5445

5446

5447

5448

5449

5450

5451

5452

5453

5454

5455

5456

5457

5458

5459

5460

5461

5462

5463

5464

5465

5466

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

HE

High-end

HEC

Human equivalent concentration

HED

Human equivalent dose

HERO

Health and Environmental Research Online (EPA Database)

HHE

Health hazard evaluation

HSDB

Hazardous Substances Data Bank

ICIS

Integrated Compliance Information System

IFC

Industrial Function Category

IIOAC

Integrated Indoor/Outdoor Air Calculator (EPA)

IRIS

Integrated Risk Information System

IUR

Inhalation unit risk

Koc

Soil organic carbon: water partitioning coefficient

Kow

Octanol: water partition coefficient

LADC

Lifetime Average Daily Concentration

LADD

Lifetime Average Daily Dose

LOAEC

Lowest-observed-adverse-effect-concentration

LOD

Limit of detection

Log Koc

Logarithmic organic carbon: water partition coefficient

Log Kow

Logarithmic octanol: water partition coefficient

LOQ

Limit of quantitation

MLD

Million liters per day

MOE

Margin of exposure

MRD

Methodology Review Draft (EPA)

MW

Molecular weight

NAICS

North American Industry Classification System

ND

Non-detect

NEI

National Emissions Inventory

NIOSH

National Institute for Occupational Safety and Health

NOAEL

No-observed-adverse-effect-level

NPDES

National Pollutant Discharge Elimination System

OAQPS

Office of Air Quality Planning and Standards

OCF

One-component foam

OCSPP

Office of Chemical Safety and Pollution Prevention

OD

Operating days

OECD

Organisation for Economic Co-operation and Development

OEHHA

Office of Environmental Health Hazard Assessment

OES

Occupational exposure scenario

ONU

Occupational non-user

OPPT

Office of Pollution Prevention and Toxics

OSHA

Occupational Safety and Health Administration

PBZ

Personal breathing zone

PECO

Population, exposure, comparator, and outcome

PEL

Permissible exposure limit

PESS

Potentially exposed or susceptible subpopulations

PET

Polyethylene terephthalate

PF

Protection factor

PNOR

Particulates not otherwise regulated

POD

Point of departure

POTW

Publicly owned treatment works (wastewater)

Page 178 of 484


-------
5467

5468

5469

5470

5471

5472

5473

5474

5475

5476

5477

5478

5479

5480

5481

5482

5483

5484

5485

5486

5487

5488

5489

5490

5491

5492

5493

5494

5495

5496

5497

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

PPE

Personal protective equipment

PV

Production volume

PWS

Public water system

QA/QC

Quality assurance/quality control

QE

NHDPlus V2.1 flow values representing "the best EROM estimate of actual mean flow"

RE

(2020 RE) Risk Evaluation

RCRA

Resource Conservation and Recovery Act

REACH

Registration, Evaluation, Authorisation and Restriction of Chemicals (European Union)

RESO

Receptors, exposure, setting or scenario, and outcomes

SACC

Science Advisory Committee on Chemicals

SDS

Safety data sheet

SDWA

Safe Drinking Water Act

SHEDS-HT

Stochastic Human Exposure and Dose Simulation-High Throughput

SIC

Standard Industrial Classification

SOC

Standard Occupational Classification

SpERC

Specific Environmental Release Categories

SPF

Spray polyurethane foam

STORET

STOrage and RETrieval and Water Quality exchange

SUSB

Statistics of United States Businesses

SWIFT

Sciome Workbench for Interactive Computer-Facilitated Text-mining

TIAB

Title/abstract (screening)

TRI

Toxics Release Inventory

TSCA

Toxic Substances Control Act

TWA

Time-weighted average

UCMR

Unregulated Contaminant Monitoring Rule

U.S.

United States

USGS

U.S. Geological Survey

VOC

Volatile organic compound

VP

Vapor pressure

WQP

Water Quality Portal

WWT

Wastewater treatment

Page 179 of 484


-------
5498

5499

5500

5501

5502

5503

5504

5505

5506

5507

5508

5509

5510

5511

5512

5513

5514

5515

5516

5517

5518

5519

5520

5521

5522

5523

5524

5525

5526

5527

5528

5529

5530

5531

5532

5533

5534

5535

5536

5537

5538

5539

5540

5541

5542

5543

5544

5545

5546

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Appendix B LIST OF SUPPLEMENTAL DOCUMENTS

Associated Systematic Review Data Quality Evaluation and Data Extraction Documents - Provide
additional detail and information on individual study evaluations and data extractions including criteria
and data quality results.

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Systematic Review Supplemental
File: Data Quality Evaluation and Data Extraction Information for Environmental Release and
Occupational Exposure - Provides a compilation of tables for the data extraction and data
quality evaluation information for 1,4-dioxane. Each table shows the data point, set, or
information element that was extracted and evaluated from a data source that has information
relevant for the evaluation of environmental release and occupational exposure (

2023t). This supplemental file may also be referred to as the 1,4-Dioxane Supplement to the Risk
Evaluation Data Quality Evaluation and Data Extraction Information for Environmental Release
and Occupational Exposure.

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Systematic Review Supplemental
File: Data Quality Evaluation Information for General Population, Consumer, and
Environmental Exposure - Provides a compilation of tables for the data quality evaluation
information for 1,4-dioxane. Each table shows the data point, set, or information element that
was evaluated from a data source that has information relevant for the evaluation of general
population, consumer, and environmental exposure (	23r). This supplemental file

may also be referred to as the 1,4-Dioxane Supplement to the Risk Evaluation Data Quality
Evaluation Information for General Population, Consumer, and Environmental Exposure.

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Systematic Review Supplemental
File: Data Extraction Information for General Population, Consumer, and Environmental
Exposure - Provides a compilation of tables for the data extraction for 1,4-dioxane. Each table
shows the data point, set, or information element that was extracted from a data source that has
information relevant for the evaluation of general population, consumer, and environmental
exposure (	023h). This supplemental file may also be referred to as the 1,4-Dioxane

Supplement to the Risk Evaluation Data Extraction Information for General Population,
Consumer, and Environmental Exposure.

Associated Supplemental Information Files - Provide additional details and information on exposure,
hazard, and risk assessments.

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File:
Environmental Releases to Air - Provides a summary of stack and fugitive air emissions for each
occupational exposure scenario (OES) in the 1,4-Dioxane Supplemental Risk Evaluation (

23j).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File:
Environmental Releases to Landfor all OES Except Disposal - Provides a summary of land
releases for each 1,4-dioxane OES except for the Disposal OES (	:023k).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File:
Environmental Releases to Landfor the Disposal OES - This spreadsheet contains a summary of
land releases for the 1,4-dioxane occupational Disposal OES (U.S. EPA. 20231).

Page 180 of 484


-------
5547

5548

5549

5550

5551

5552

5553

5554

5555

5556

5557

5558

5559

5560

5561

5562

5563

5564

5565

5566

5567

5568

5569

5570

5571

5572

5573

5574

5575

5576

5577

5578

5579

5580

5581

5582

5583

5584

5585

5586

5587

5588

5589

5590

5591

5592

5593

5594

5595

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File:
Environmental Releases to Water for OES without TRI or DMR data - Provides a summary of
direct and indirect water releases for each 1,4-dioxane OES for which Toxics Release Inventory
(TRI) and Discharge Monitoring Report (DMR) data were not available (	2023m).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File:
Environmental Releases to Water for OES with TRI and DMR - Provides a summary of direct
and indirect water releases for each 1,4-dioxane OES for which TRI or DMR data were available

(	>023m).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File:
Occupational Exposure and Risk Estimates - Provides a summary of occupational exposures and
risks estimated for all conditions of use (COUs; (including those evaluated in this supplemental
evaluation as well as those previously evaluated in the 2020 RE) (U.S. EPA. 2023r).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File:
Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Release to Surface Water from
Individual Facilities - Provides water concentrations estimated from individual facility releases
reported to TRI and calculates corresponding drinking water exposures and risks (

2023h).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File:
Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Surface Water Concentrations
Predicted with Probabilistic Modeling - Provides water concentrations estimated by
probabilistic modeling for DTD releases, disposal of hydraulic fracturing waste to surface water,
and for aggregate concentrations estimated downstream of industrial release sites; calculates
corresponding drinking water exposures and risks. This file also calculates drinking water
exposures and risks estimated from drinking water monitoring data (U.S. EPA. 20231).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File:
EWISRDXL BrunswickCountyNC Case Study - Provides the Excel workbook file for the
Brunswick County, NC surface water case study (	)23o).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File:
EWISRDXL ColumbiaTN Case Study - Provides the Excel workbook file for the Columbia, TN
surface water case study (U.S. EPA. 2023p).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File:
EWISRDXL LiverpoolOH Case Study - Provides the Excel workbook file for the Liverpool, OH
surface water case study (	A. 2023 q).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane Supplemental Information File:
EWISRD-XL-Rprobabilistic model code - Provides the R script used to perform the probabilistic
surface water modeling by OES/COU (	323a).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File: WQP
Processed Surface Water Data - Provides the processed monitoring data in surface water
retrieved from the Water Quality Portal (	2023s).

Page 181 of 484


-------
5596

5597

5598

5599

5600

5601

5602

5603

5604

5605

5606

5607

5608

5609

5610

5611

5612

5613

5614

5615

5616

5617

5618

5619

5620

5621

5622

5623

5624

5625

5626

5627

5628

5629

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File:
Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Land Releases to Landfills -
Provides calculations of groundwater concentration derived from the waste adjusted dilution
attenuation factor extracted from Delisting Risk Assessment Software (DRAS) for Landfills and
the corresponding risk calculations (	,023f).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File:
Drinking Water Exposure and Risk Estimates for 1,4-Dioxane Land Releases to Surface
Impoundments - Provides calculations of groundwater concentration derived from the waste
adjusted dilution attenuation factor extracted from DRAS for release of hydraulic fluid produced
water to surface impoundments and the corresponding risk calculations (	).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File: Air
Exposures and Risk Estimates for Single Year Analysis - Provides air concentrations estimated
by American Meteorological Society/Environmental Protection Agency Regulatory Model
(AERMOD) for air releases reported to TRI in 2019 and calculates corresponding exposure
concentrations and risk estimates (	23e).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File: Air
Exposure and Risk Estimates for 1,4-Dioxane Emissions from Hydraulic Fracturing Operations
- Provides air concentrations estimated by Integrated Indoor/Outdoor Air Calculator (IIOAC)
based on Monte Carlo modeling of air releases from hydraulic fracturing operations and
calculates corresponding exposure concentrations and risks (	)23b).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File: Air
Exposures and Risk Estimates for Industrial Laundry - Provides air concentrations estimated by
IIO AC based on Monte Carlo modeling of air releases from hydraulic fracturing operations and
calculates corresponding exposure concentrations and risks (	)23c).

Draft Supplement to the Risk Evaluation for 1,4-Dioxane - Supplemental Information File: Air
Exposures and Risk Estimates for Multi-Year Analysis - Provides air concentrations estimated by
IIO AC for 6 years (2015 to 2020) of air releases reported to TRI and calculates the
corresponding exposure concentrations and risk estimates (	3d).

Page 182 of 484


-------
5630

5631

5632

5633

5634

5635

5636

5637

5638

5639

5640

5641

5642

5643

5644

5645

5646

5647

5648

5649

5650

5651

5652

5653

5654

5655

5656

Appendix C

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

SYSTEMATIC REVIEW PROTOCOL FOR THE
DRAFT SUPPLEMENT TO THE RISK EVALUATION
	FOR 1,4-DIOXANE	

The U.S. EPA's Office of Pollution Prevention and Toxics (EPA/OPPT) applies systematic review
principles in the development of risk evaluations under the amended TSCA. TSCA section 26(h)
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 CFR 702.33).

To meet the TSCA section 26(h) science standards, EPA used the TSCA systematic review process
described in the Draft Systematic Review Protocol Supporting TSCA Risk Evaluations for Chemical
Substances (U.S. EPA. 2021a) (hereinafter referred to as "2021 Draft Systematic Review Protocol").
Table Apx C-l (Section 3 of the 2021 Draft Systematic Review Protocol) depicts the steps in which
information is identified and whether it undergoes the formal systematic review process (U.S. EPA.
2021a). Information attained via the systematic review process is integrated with information attained
from sources of information that do not undergo systematic review (e.g., EPA-generated model outputs)
to support a weight of the scientific evidence analysis.

Scope

Risk Evaluation



Systematic Review

1

Literature Searching
and Screening

Test
Order/
Rule Data

Data Evaluation

Data Extraction
(in scope chemicals)

Data Gap filling from sources
outside of the Systematic Review
process
(i.e., systematic approaches using
model outputs, analogue, qualitative
information on a COU)

Weight of the Scientific Evidence
Analysis'/or each discipline and across
disciplines

Evidence Integration of
Systematically Reviewed
data

Evidence Integration of data
obtained outside of
Systematic Review

Conclusions from
the Weight of the
Scientific Evidence
Analysis

Risk
Characterization

across disciplines

Legend

TSCA Process/Product

Systematic Review Step

Non-Systematic Review
Step (may encompass
systematic approaches)

Step utilizing Systematic
Review and non-
Systematic Review
Results

'Weight of the
Scientific Evidence
(WoSE) considers
the results of the
Systematic Review
method and
additional evidence
integrated from non-
systematic review
methods. Analysis of
the WoSE may
include evidence
integration across
disciplines.

FigureApx C-l. Overview of the TSCA Risk Evaluation Process with Identified Systematic
Review Steps

Page 183 of 484


-------
5657

5658

5659

5660

5661

5662

5663

5664

5665

5666

5667

5668

5669

5670

5671

5672

5673

5674

5675

5676

5677

5678

5679

5680

5681

5682

5683

5684

5685

5686

5687

5688

5689

5690

5691

5692

5693

5694

5695

5696

5697

5698

5699

5700

5701

5702

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

The process complements the risk evaluation process in that 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).

C.l Clarifications and Updates to the 2021 Draft Systematic Review
Protocol

In 2021, EPA released the Draft Systematic Review Protocol Supporting TSCA Risk Evaluations for
Chemical Substances (U.S. EPA. 2021a). a framework of systematic review approaches under TSCA, to
address comments received on a precursor systematic review approaches framework, the Application of
Systematic Review in TSCA Risk Evaluations (	2018c). In April 2022, the SACC provided

comments on the 2021 Draft Systematic Review Protocol while additional comments on OPPT's
systematic review approaches were garnered during the public comment period. In lieu of an update to
the 2021 Draft Systematic Review Protocol, this systematic review protocol for the Draft Supplement to
the Risk Evaluation for 1,4-Dioxane describes some clarifications and different approaches that were
implemented than those described in the 2021 Draft Systematic Review Protocol in response to (1)

SACC comments, (2) public comments, or (3) to reflect chemical-specific risk evaluation needs.

C.l.l Clarifications and Updates	

Throughout the 2021 Draft Systematic Review Protocol, there were some terms used that were not
explicitly defined, resulting in their different uses within the document (U.S. EPA. 2021a). Table Apx
C-l lists the terms that were updated to resolve some of the confusion expressed by the public and
SACC comments regarding the implementation of the respective systematic review-related step. One
main clarification is that all references that undergo systematic review are consideredfor use in the risk
evaluation, even those that do not meet the various discipline and sub-discipline screening criteria {i.e.,
RESO, PESO, PECO) or that are categorized as supplemental information at title and abstract (TIAB) or
full-text screening.

Section 4.2.5 of the 2021 Draft Systematic Review Protocol describes how data sources {e.g., individual
references, databases) may be tagged and linked in epidemiological cohort studies when information is
present in multiple studies (	21a). References will generally undergo data quality evaluation

and extraction if there are data that pass screening criteria; however, to prevent the same data from being
represented multiple times and conflating the amount of available information on a subject area, EPA
selects the reference(s) that most appropriately describes the extractable results (indicated as the parent
reference in DistillerSR). For example, if two references portray the same information from the same
dataset, only one is counted in the overall dataset {i.e., deduplication). If two references contain
information about the same dataset, but only one provides additional contextual information or summary
statistics {e.g., mean), both data sources are linked but the extractable information from both may be
combined in DistillerSR. This allows the capture of key information while avoiding double counting the
data of interest, which may be the case whether or not one reference contains original or extractable data
that passes screening criteria. The linked reference containing the majority of the data, which are
evaluated and extracted, is identified in DistillerSR as the parent reference; the "complementary child
reference" in DistillerSR does not undergo data evaluation and extraction. Linking the references in
DistillerSR allows the reference with more limited information or only contextual information to be
tracked and utilized to evaluate the extracted data in the other related studies. The child reference may
undergo data quality evaluation and extraction if there are additional unique and original data that pass
screening criteria. One clarification is that this procedure of identifying potential duplicative information
applies to all information that is considered in a risk evaluation under TSCA (not just epidemiological

Page 184 of 484


-------
5703

5704

5705

5706

5707

5708

5709

5710

5711

5712

5713

5714

5715

5716

5717

5718

5719

5720

5721

5722

5723

5724

5725

5726

5727

5728

5729

5730

5731

5732

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

cohort studies). Also, this procedure may apply when there is duplicative information in two references,
even if it is more than just "contextual."

Section 5 of the 2021 Draft Systematic Review Protocol describes how EPA conducts data quality
evaluation of data/information sources considered for a respective chemical risk evaluation, with Section
5.2 specifically explaining the terminology used to describe both metric and overall data/information
source quality determinations (	21a). To respond to both SACC and public comments

regarding the inappropriate use of quantitative methodologies to calculate both "Metric Rankings" and
"Overall Study Rankings", EPA decided to not implement quantitative methodologies to attain either
metric and overall data/information source quality determinations and therefore updated the
terminology used for both metric ("Metric Ranking") and overall data/information source ("Overall
Study Ranking") quality determinations (TableApx C-l). Specifically, metric and overall
data/information source quality determination terminology have been updated to "Metric Rating" and
"Overall Quality Determination", respectively. The word "level" was also often used synonymously and
inconsistently with the word "ranking" in the 2021 Draft Systematic Review Protocol; that inconsistency
has been rectified, resulting in the word "level" no longer being used to indicate either metric or overall
data/information source quality determinations (	2021a).

Sections 4.3.2.1.3 and 6 of the 2021 Draft Systematic Review Protocol describes when EPA may reach
out to authors of data/information sources to obtain raw data or missing elements that are important to
support the data evaluation and data integration steps (	:021a). In such cases, the request(s)

for additional data/information, number of contact attempts, and responses from the authors are
documented. EPA's outreach is considered unsuccessful if those contacted do not respond to email or
phone requests within 1 month of initial attempt(s) of contact. One important clarification to this
guidance is that EPA may reach out to authors anytime during the systematic review process for a given
data/information source or reference, and that contacting authors does not explicitly happen during the
data quality evaluation or extraction steps.

Table Apx C-l. Terminology Clarifications between the 2021 Draft Systematic Review Protocol
and the 1,4-Dioxane Systematic Review Protocol 	

2021 Draft Systematic
Review Protocol Term

1,4-Dioxane Systematic
Review Protocol Term Update

Clarification

"Title and abstract" or
"Title/abstract"

Title and abstract (TIAB)

To increase consistency, the term "title and abstract"
will be used to refer to information specific to "title
and abstract" screening.

Variations of how
"include," "on topic" or
"PECO/PESO/RESO
relevant" implied a
reference was
considered for use in the
risk evaluation, whereas
"exclude," "off topic" or
"not PECO/PESO/RESO
relevant" implied a
reference was not
considered for use in the
risk evaluation.

Meets/does not meet
PECO/PESO/RESO screening
criteria

The term "include" or "exclude" falsely suggests
that a reference was or was not, respectively,
considered in the risk evaluation. There was also
confusion regarding whether "on topic" and
"PECO/PESO/RESO relevant" were synonymous
and suggested those references were explicitly
considered for use in the risk evaluation (and by
default, "off topic" and "not PECO/PESO/RESO
relevant" references were not). References that meet
the screening criteria (e.g., PECO, PESO, RESO)
proceed to the next systematic review step;
however, all references that undergo systematic
review at any time are considered in the risk
evaluation. Information that is categorized as

Page 185 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2021 Draft Systematic
Review Protocol Term

1,4-Dioxane Systematic
Review Protocol Term Update

Clarification





supplemental or does not meet screening criteria are
generally less relevant for quantitative use in the
risk evaluation but may be considered if there is a
data need identified. For instance, mechanistic
studies are generally categorized as supplemental
information at either title and abstract or full-text
screening steps but may undergo the remaining
systematic review steps if there is a relevant data
need for the risk evaluation (e.g., dose response,
mode of action).

Database source not
unique to a chemical

Database

Updated term and definition of "Database": Data
obtained from databases that collate information for
the chemical of interest using methods that are
reasonable and consistent with sound scientific
theory and/or accepted approaches and are from
sources generally using sound methods and/or
approaches (e.g., state or federal governments,
academia). Example databases include STORET
and the Massachusetts Energy and Environmental
Affairs Data Portal.

The term in the 2021 Draft Systematic Review
Protocol (Table_ApxN-l) incorrectly suggested
that databases that contain information on a singular
chemical are not considered (U.S. EPA, 2021a).
Furthermore, the wording "large" was removed to
prevent confusion and the incorrect suggestion that
there is a data size requirement for databases that
contain information that may be considered for
systematic review.

Metric Ranking or Level

Metric Rating

As explained above, EPA is not implementing
quantitative methodologies to indicate metric
quality determinations, therefore the term "ranking"
is inappropriate. The term "level" was inconsistently
used to indicate metric quality determinations
previously, therefore EPA is removing the use of
this term to reduce confusion when referring to
metric quality determinations. The term "Rating" is
more appropriate to indicate the use of professional
judgement to determine a quality level for
individual metrics.

Overall Study Ranking
or Level

Overall Quality Determination
(OQD)

As explained above, EPA is not implementing
quantitative methodologies to indicate overall
data/information source quality determinations,
therefore the term "ranking" is inappropriate. The
term "level" was inconsistently used to indicate
overall data/information source quality
determinations previously, therefore EPA is
removing the use of this term to reduce confusion

Page 186 of 484


-------
5733

5734

5735

5736

5737

5738

5739

5740

5741

5742

5743

5744

5745

5746

5747

5748

5749

5750

5751

5752

5753

5754

5755

5756

5757

5758

5759

5760

5761

5762

5763

5764

5765

5766

5767

5768

5769

5770

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

2021 Draft Systematic
Review Protocol Term

1,4-Dioxane Systematic
Review Protocol Term Update

Clarification





when referring to overall data/information source
quality determinations. The term "Rating" is more
appropriate to indicate the use of professional
judgement to determine a quality level for the
overall data/information source quality
determination.

C.2 Data Search

To expand upon the previous analysis conducted in the 2020 RE, this draft Supplement to the Risk
Evaluation for 1,4-Dioxane addresses additional COUs in which 1,4-dioxane is present as a byproduct of
the manufacturing process and evaluates risks from general population exposures to 1,4-dioxane
released to water, air, and land. This draft supplement focuses on evaluating additional exposure
pathways that were not addressed in the Final Risk Evaluation for 1,4-Dioxane (	2020c).

Therefore, the data search focused on prioritizing updated literature search results to characterize
environmental releases and occupational exposure, and general population, consumer, and
environmental exposure information to evaluate the exposure pathways in scope for this draft
supplement. Data sources may also contain information that may be used to evaluate exposure pathways
already addressed in the 2020 RE (	20c) {i.e., consumer exposure). Below are the four

additional exposure pathways being assessed in this Draft Supplement to the Risk Evaluation for 1,4-
Dioxane (Section 1.2).

•	Occupational exposure to

o 1,4-dioxane present as a byproduct in commercial products during ethoxylation

processing or polyethylene terephthalate (PET) manufacturing and in hydraulic fracturing
waste (Sections 3.1, 5.2.1)

•	General population exposures to

o 1,4-dioxane present in drinking water sourced from surface water as a result of direct and
indirect industrial releases and DTD releases of consumer and commercial products
(Sections 2.3.1, 3.2.2 and 5.2.2.1);
o 1,4-dioxane present in drinking water sourced from groundwater contaminated as a result

of disposals (Sections 2.3.2, 3.2.2.2 and 5.2.2.1.6); and,
o 1,4-dioxane released to air from industrial and commercial sources (Sections 2.3.3, 3.2.3,
and 5.2.2.3).

C.2.1 Multi-Disciplinary Updates to the Data Search

For this Draft Supplement to the Risk Evaluation for 1,4-Dioxane, the updated literature search was
conducted as described in Section 4 of the 2021 Draft Systematic Review Protocol (1	2021a).

where the peer-reviewed and gray literature updated search followed the approach outlined in Sections
4.2 and 4.3 of the 2021 Draft Systematic Review Protocol, respectively (	). The updated

search for peer-reviewed and gray literature relevant references was completed in October 2021 and
January 2022, respectively, which also considered information found for the Final Risk Evaluation for
1,4-Dioxane (	)20c). Occasionally additional data sources relevant for the risk evaluation

may be identified after the initial search for peer-reviewed and gray literature; these data sources will
then undergo systematic review for the relevant discipline(s). Additionally, each discipline utilizes
different strategies (e.g., search strings) to attain their discipline-specific pools of data sources that
undergo systematic review

Page 187 of 484


-------
5771

5772

5773

5774

5775

5776

5777

5778

5779

5780

5781

5782

5783

5784

5785

5786

5787

5788

5789

5790

5791

5792

5793

5794

5795

5796

5797

5798

5799

5800

5801

5802

5803

5804

5805

5806

5807

5808

5809

5810

5811

5812

5813

5814

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

As mentioned in Section 4.2.2 of the 2021 Draft Systematic Review Protocol, a supplemental literature
search is conducted to fill data gaps, but in this draft supplement, the supplemental search was
conducted to update the literature search conducted to identify any potentially relevant environmental
release and occupation exposure and general population, consumer, and environmental exposure
information (U.S. EPA. 2021a). Rather than utilizing positive and negative seed references as described
in Section 4.2.4.2 of the 2021 Draft Systematic Review Protocol, search strings were used in SWIFT16-
Review to better identify relevant references to evaluate exposure pathways addressed in this draft
supplement (	21a). The language describing the new exposure pathways and COUs that are

in scope for this draft supplement was used to derive the search strings listed below in Sections C.2.3.1
and C.2.3.2. When the search strings are identified in the title, abstract, keyword, or Medical Subject
Heading (MeSH) fields of a given reference in SWIFT-Review, those references proceeded with TIAB
screening.

The evaluation of physical and chemical properties, fate properties and environmental and human health
hazard information did not differ from the respective information provided in the Final Risk Evaluation
for 1,4-Dioxane (	?20c) to address the additional exposure pathways in this draft

supplement, therefore no additional references were identified for these respective topics or underwent
systematic review for these disciplines. One minor clarification to what was described in the 2021 Draft
Systematic Review Protocol is that the PECO statement used to screen general population, consumer,
and environmental exposure information considered for this draft supplement, currently resides in
Appendix Section H.5 (which was intended to encompass PECO statements regarding environmental
and human health hazard information), rather than in Appendix Section H.4 (	). Please

see Appendix C.3.2 below for additional updates specific to the implementation of the PECO statement.

C.2.2 Additional Data Sources Identified

As mentioned above in Appendix C.2, additional data sources containing potentially relevant
information for a respective risk evaluation may be identified. For this draft supplement, additional gray
literature data sources were identified for the characterization of environmental release and occupational
exposure and general population, consumer, and environmental exposure, as explained below in
Sections C.2.2.1 and C.2.2.2, respectively.

C.2.2.1 Additional Data Sources Identified for Environmental Release and
Occupational Exposure

As explained in Appendix E of the 2021 Draft Systematic Review Protocol (	), generic

scenarios and emission scenarios documents are listed as part of the initial gray literature sources. Some
generic scenarios and a draft emission scenario document became available after the gray literature
search was completed in January 2022 and were considered for the environmental release and
occupational exposure assessment. This includes the Draft OECD ESD on Hydraulic Fracturing (U.S.

22d), Draft GS on Furnishing Cleaning Products (	2a), EPA Methodology Review

Draft (MRD) on Commercial Use of Automotive Detailing Products ( 'A. 2022b). and Draft GS
on Use of Laboratory Chemicals (	!2h). The updated sources were added to EPA's Health

and Environmental Research Online (HERO) database in 2022 as well as the systematic review process.

In addition to the gray literature sources listed above, an online database called FracFocus 3.0 (GWPC
and IOGCC. 2022) was included in the pool of references EPA considered for environmental release and
occupational exposure through backward searching. These are described in Section 4.4 of the 2021 Draft

16 SWIFT is an acronym for "Sciome Workbench for Interactive Computer-Facilitated Text-mining." SWIFT-Active
Screeneruses machine learning approaches to save screeners' time and effort.

Page 188 of 484


-------
5815

5816

5817

5818

5819

5820

5821

5822

5823

5824

5825

5826

5827

5828

5829

5830

5831

5832

5833

5834

5835

5836

5837

5838

5839

5840

5841

5842

5843

5844

5845

5846

5847

5848

5849

5850

5851

5852

5853

5854

5855

5856

5857

5858

5859

5860

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Systematic Review Protocol (	2021a). Backward searching from the Draft OECD ESD on

Hydraulic Fracturing (U.S. EPA. 2022d) led to EPA's identification of the FracFocus data. The Agency
gathered the data directly from the source and only pulled data specific to sites that reported using 1,4-
dioxane in fracturing fluids. This source was added to the HERO database as well as the systematic
review process.

C.2.2.2 Additional Data Sources Identified for General Population, Consumer, and
Environmental Exposure

In addition to the gray literature sources listed in Appendix E of the 2021 Draft Systematic Review
Protocol (	'2la), several other gray literature sources were considered for inclusion the

general population, consumer, and environmental exposure assessment and added to the HERO database
in 2022. The Water Quality Portal (WQP) database, the successor of EPA's STORET (STOrage and
RETrieval) database, was incorporated because it includes a large variety of chemical-specific data.

Also, WQP is a portal that combines data from multiple databases—not just STORET—such as the U.S.
Geological Survey's National Water Information System. Information from WQP was collected in July
2022.

A few additional gray literature sources (databases) were included in the pool of references EPA
considered on general population, consumer, and environmental exposure through backwards searching,
which is described in Section 4.3.3 of the 2021 Draft Systematic Review Protocol (U.S. EPA. 2021a).
Backwards searching from the Third Unregulated Contaminant Monitoring Rule (UCMR3) database
(\ c. < ^ \ 201 (!) led to EPA's identification of data from a few states collecting data on 1,4-dioxane
for longer periods of time than reported in UCMR3. EPA was able to secure data from three state
databases, which as added. In addition, elevated levels of 1,4-dioxane in samples from UCMR3 and a
reference found in the pool of peer-reviewed articles led to addition to databases with data on 1,4-
dioxane levels measured in drinking water and surface water in the Cape Fear Watershed in North
Carolina. Six of the highest concentrations of 1,4-dioxane in the UCMR3 database were reported in
water from this watershed, which were also discussed in one of the references found in the literature
search for peer-reviewed sources. All gray literature database sources added to the search, including the
datasets from North Carolina, were added to HERO and the systematic review process (see Section G. 1
for further information).

C.2.3 Search Strings

As explained above in Section C.2.1, the search strings below were used to identify references relevant
to evaluating environmental releases and occupational exposure, as well as general population,
consumer, and environmental exposure.

C.2.3.1 Environmental Release and Occupational Exposure Search Strings

Life Cycle:

TIAB: ("1,4-dioxane" AND ("MFG" OR "import" OR "processing" OR "manufactur*" OR "releases"
OR "waste disposal" OR "reaction product" OR "repackaging" OR "recycling" OR "throughput" OR
"operating days" OR "batch" OR "production speed"))

Treatment Efficiencies:

TIAB: ("1,4-dioxane" AND ("GAC" OR "granular activated carbon" OR "reverse osmosis" OR
"advanced oxidation" OR "hydrogen peroxide with ultraviolet" OR ("hydrogen peroxide" AND "UV")
OR "hydrogen peroxide with ozone" OR ("hydrogen peroxide" AND "ozone") OR "AOP" OR
"Fenton's reagent" OR "bioremediation"))

Page 189 of 484


-------
5861

5862

5863

5864

5865

5866

5867

5868

5869

5870

5871

5872

5873

5874

5875

5876

5877

5878

5879

5880

5881

5882

5883

5884

5885

5886

5887

5888

5889

5890

5891

5892

5893

5894

5895

5896

5897

5898

5899

5900

5901

5902

5903

5904

5905

5906

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Occupational Workers:

TIAB: ("1,4-dioxane" AND ("janitor*" OR "mechanic" OR "laborer" OR "custodia*" OR "painter*"
OR "laboratory technician" OR "laboratory employee*" OR ("pharmaceutical" AND ("employee" OR
"worker" OR "technician")) OR "residential construction" OR "industrial construction"))

General:

TIAB: ("1,4-dioxane" AND ("surfactant" OR "ethoxylat*" OR "nonylphenol ethoxylate" OR
"alkylphenol ethoxylate" OR "sulfated" OR "industrial laundr*" OR "commercial laundr*" OR
"institutional laundr*" OR "institutional laundr*" OR "advanced oxidation" OR "ozone-peroxide
advanced oxidation" OR "low dioxane" OR "low dioxane ether sulfates" OR "low dioxane ethoxylated
surfactants" OR "low 1,4 dioxane ether sulfates" OR "low 1,4 dioxane ethoxylated surfactants" OR
"safety data sheet" OR "material safety data sheet"))

Process Uses:

TIAB: ("1,4-dioxane" AND ("stabilizer" AND ("chlorinated solvents" OR "degreasing" OR
"electronics manufacturing" OR "metal finishing")) OR ("solvent" AND ("histology" OR "cellulose
acetate membrane" OR "microscopy" OR "organic chemical manufacturing" OR "organic chemical"))
OR ("textile" AND ("wetting" OR "dispersing")) OR ("esterification" AND ("by-product" OR
"byproduct")))

Product Uses:

TIAB: ("1,4-dioxane" AND ("solvent" AND ("paint*" OR "lacquer*" OR "varnish remover" OR
"stain" OR "printing" OR "scintillation" OR "resin*" OR "oil*" OR "rubber chemicals" OR "rubber"
OR "sealant*" OR "adhesive*" OR "wax*" OR "cement*")))

TIAB: ("1,4-dioxane" AND ("artificial leather" OR "purifying agent" OR "antifreeze" OR "de-icing"
OR "pesticide*" OR "fumigant*"))

CASRNs of Ethoxylated Chemicals:

TIAB: ("9005-65-6" OR "3088-31-1" OR "68081-98-1" OR "68439-50-9" OR "68551-12-2" OR
"68439-49-6" OR "9043-30-5" OR "26183-52-8" OR "9002-92-0" OR "9004-82-4" OR "9005-64-5"
OR "68131-40-8" OR "68991-48-0" OR "37251-67-5" OR "5274-68-0" OR "864529-51-1" OR
"84133-50-6" OR "68439-45-2" OR "68987-81-5" OR "9003-11-6" OR "61791-29-5" OR "9005-08-7"
OR "61791-13-7" OR "166736-08-9" OR "3055-99-0" OR "66455-14-9" OR "68131-39-5" OR
"68213-23-0" OR "68951-67-7" OR "66455-15-0" OR "61791-26-2" OR "9004-95-9" OR "9005-00-9"
OR "61827-42-7" OR "68081-91-4" OR "68585-40-0" OR "68815-56-5" OR "61788-85-0" OR "3055-
97-8" OR "120313-48-6" OR "68439-46-3" OR "69227-22-1" OR "68002-97-1")

C.2.3.2 General Population, Consumer, and Environmental Exposure Search Strings

Population:

TIAB: ("general population" OR "bystanders" OR "near-facility" OR "industrial facilit*" OR
"commercial facilit*" OR "employee" OR "employees" OR "worker*" OR "manufacturer" OR "near-
disposal" OR "near surface disposal" OR "child*" OR "teenage*" OR "susceptible population" OR
"immunocompromised" OR "preschool" OR "senior*" OR "older adults" OR "elderly" OR "pregnant
women" OR "preexisting condition*" OR "lactating women" OR "childbearing" OR "prenatal" OR
"infant*" OR "adolescen*")

Landfills:

Page 190 of 484


-------
5907

5908

5909

5910

5911

5912

5913

5914

5915

5916

5917

5918

5919

5920

5921

5922

5923

5924

5925

5926

5927

5928

5929

5930

5931

5932

5933

5934

5935

5936

5937

5938

5939

5940

5941

5942

5943

5944

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TIAB: ("dioxane" AND ("landfill" OR "leach*" OR "incineration" OR "wastewater" OR"GAC" OR
"granular activated carbon" OR "reverse osmosis" OR "waste site" OR "land disposal" OR "waste
disposal" OR "landfill leach*"))

Indoor Air and Water:

TIAB: ("dioxane" AND ("inhal*" OR "tap water" OR "water well" OR "indoor air" OR "surface water"
OR "groundwater" OR "outdoor air" OR "ambient air" OR "drinking water" OR ("biomonitoring" OR
"monitoring" AND ("air" OR "water")) OR "drinking" OR "aquifer" OR "leach*" OR "municipal
water")) NOT ("spill")

Consumer and Industrial Use:

TIAB: ("1,4-dioxane" AND ("ingest*" OR "swallow*" OR "showering" OR "bathing" OR "swimming"
OR "wading" OR "inhal*" OR "paint*" OR "industrial manufactur*" OR "residential construction" OR
"commercial construction" OR "cleaning" OR "dishwasher" OR "printing" OR "food supplement*" OR
"packaging" OR "breast milk" OR "human milk" OR "intake rates" OR "launder*" OR "surface
cleaner" OR "automotive"))

Concentration and Dose:

TIAB: ("reference concentration" OR "RfC" OR "NOAEL" OR "LOAEL" OR "benchmark
concentration" OR "reference dose" OR "RfD" OR "chronic oral" OR "chronic inhalation" OR "oral
slope factor" OR "soil screening level" OR "PEL" OR "permissible exposure limit" OR "weighted
average" OR "weight fraction" OR "emission rate*" OR "inhalation unit risk" OR "IUR" OR "dose-
response" OR "reverse dosimetry" OR "biomonitoring" OR "media concentration*" OR ("estimate*"
AND ("acute" OR "subchronic" OR "chronic")) OR "single-dose" OR "repeated-dose" OR "daily
intake")

C.3 Data Screening

Sections 4.2.5 and 4.3.2 of the 2021 Draft Systematic Review Protocol describe how title and abstract
(TIAB) and full-text (FT) screening, respectively, are conducted to identify references that may contain
relevant information for use in risk evaluations under TSCA using discipline-specific screening criteria
(defined below in Sections C.3.1.1 and C.3.2.1 (	2021a). Specifically, TIAB screening efforts

may be conducted using the specialized web-based software programs DistillerSR17 and SWIFT-Active-
Screener18; however, for this draft supplement, EPA used SWIFT-Active-Screener exclusively.
Additional details on how SWIFT Active-Screener utilizes a machine-learning algorithm to
automatically compute which unscreened documents are most likely to be relevant19 are available in
Section 4.2.5 of the 2021 Draft Systematic Review Protocol (U.S. EPA. 2021a). When TIAB screening
is completed, references that meet screening criteria {i.e., PECO/RESO/PESO statements) then undergo
FT screening if the full reference is able to be retrieved and generated into a Portable Document Format
(PDF).

17	As noted on the DistillerSR web page, this systematic review software "automates the management of literature collection,
triage, and assessment using AI and intelligent workflows...to produce transparent, audit ready, and compliant literature
reviews." EPA uses DistillerSR to manage the workflow related to screening and evaluating references; the literature search
is conducted external to DistillerSR.

18	SWIFT-Active Screener is another systematic review software that EPA is adopting in the TSCA systematic review
process. From Sciome's SWIFT-Active Screener web page: "As screening proceeds, reviewers include or exclude articles
while an underlying statistical model in SWIFT-Active Screener automatically computes which of the remaining unscreened
documents are most likely to be relevant. This 'Active Learning' model is continuously updated during screening, improving
its performance with each reference reviewed. Meanwhile, a separate statistical model estimates the number of relevant
articles remaining in the unscreened document list."

19	Description comes from the SWIFT-Active Screener web page.

Page 191 of 484


-------
5945

5946

5947

5948

5949

5950

5951

5952

5953

5954

5955

5956

5957

5958

5959

5960

5961

5962

5963

5964

5965

5966

5967

5968

5969

5970

5971

5972

5973

5974

5975

5976

5977

5978

5979

5980

5981

5982

5983

5984

5985

5986

5987

5988

5989

5990

5991

5992

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Literature inventory trees were introduced in the scoping process for the risk evaluations that began
systematic review in 2019 in response to comments received from the SACC and public to better
illustrate how references underwent various systematic review steps (e.g., TIAB and FT screening). As
explained in various final scope documents (e.g., Section 2.1.2 in the Final Scope of the Risk Evaluation
for 4,4'-(I-Me/ln letin liilene) bis/2, 6-dibromophenol] (	:0b)). literature inventory trees

demonstrate how references that meet screening criteria progress to the next systematic review step.
EPA used the Health Assessment Workplace Collaborative (HAWC) tool to develop web-based
literature inventory trees to enhance the transparency of the decisions resulting from the screening
processes. Updates made to the available literature considered for the draft supplement that are made
between publishing the draft and final Supplement to the Risk Evaluation for 1,4-Dioxane (e.g.,
additional references may be provided to the EPA through public comment) will be reflected in HAWC
(see also hyperlinks to HAWC in the figure captions below for each respective literature inventory tree).

The web-based literature inventory trees in HAWC also allow users to directly access the references in
the HERO database (more details available in Section 1 of the 2021 Draft Systematic Review Protocol)
by selecting appropriate nodes, which indicate whether a reference has met screening criteria at different
screening steps and/or types of content that may be discerned at that respective systematic review step
(	2021a). Furthermore, as mentioned in the various final scope documents, the sum of the

numbers for the various nodes in the literature inventory trees may be smaller or larger than the
preceding node because some studies may have unclear relevance or be relevant for many categories of
information. The screening process for each discipline varies and the nodes in the literature tree indicate
the screening decisions determined for each reference and whether specific content could be determined;
if no references had a specific screening decision and/or contained specific content relevant for a
respective discipline, a node will not be present on the literature tree to depict this.

In the literature inventory trees below, which depict systematic review search results used to evaluate the
new exposure pathways in this draft supplement, some references were unattainable for FT screening.
The "PDF not available" node refers to references or sources of information for which EPA was unable
to obtain the entire reference or source of data/information but were identified in the literature search
because of the availability of the title and abstract. For the references considered to evaluate
environmental release and occupational exposure, all references that passed screening criteria were
found and underwent FT screening. For the references considered to evaluate general population,
consumer, and environmental exposure, one PDF could not be obtained through interlibrary loan or
through other channels available to EPA to obtain reprints of published sources.

As mentioned in Section C.l, although all information contained in references that enter systematic
review are considered for use in the risk evaluation, the references that satisfy the screening criteria are
generally deemed to contain the most relevant and useful information for characterizing the uses,
exposure, and hazard of a chemical of interest and are generally utilized in the risk evaluation (and can
be used later on to identify further data needs). On the other hand, data or information sources that do
not satisfy the screening criteria outlined below may undergo data quality evaluation and extraction
should a data need arise for the risk evaluation.

C.3.1 Environmental Release and Occupational Exposure

During data screening, EPA followed the process described in Appendix H, Section H-3 of the 2021
Draft Systematic Review Protocol (	021a) to conduct TIAB and FT screening for 1,4-

dioxane literature search results guided by the RESO statement. RESO stands for Receptors, Exposure,
Setting or Scenario, and Outcomes. Data or information sources that comply with the screening criteria

Page 192 of 484


-------
5993

5994

5995

5996

5997

5998

5999

6000

6001

6002

6003

6004

6005

6006

6007

6008

6009

6010

6011

6012

6013

6014

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

specified in the RESO statement then undergo data quality evaluation and extraction. Figure_Apx C-2
presents the number of references that report general engineering data, environmental release, and
occupational exposure data that passed RESO screening criteria at TIAB and FT screening.

C.3.1.1 Environmental Release and Occupational Exposure Literature Tree

46

Meets RESO Criteria at FT
Screening

©

General Engineering
Assessment

®

Peer-Reviewed Literature
Search

0

Occupational Exposure

®

Environmental Release

©

Meets RESO Criteria at TIAB
Screening

0 0

TSCA Engineering 1,4-Dioxane - Does Not Meet RESO Criteria at
Part 2 (2021)	TIAB Screening

Ooes Not Meet RESO Criteria at
FT Screening

©

Supplemental From FT
Screening

©

Gray Literature Search

©

General Engineering
Assessment

©

Occupational Exposure

©

Supplemental From TIAB
Screening

Environmental Release

FigureApx C-2. Literature Inventory Tree - Environmental Releases and Occupational
Exposure Search Results for 1,4-Dioxane

View the interactive literature inventory tree in HAWC. Data in this figure represent all references obtained from
the publicly available databases and gray literature references searches that were included in systematic review as
of March 30, 2023. Additional data may be added to the interactive version as they become available.

C.3.2 General Population, Consumer, and Environmental Exposure
The TIAB and FT screening process was consistent with what EPA previously outlined in Sections 4.2.5
and 4.3.2 of the 2021 Draft Systematic Review Protocol ( J.S. EPA. 2021a). PECO stands for
Population, Exposure, Comparator or Scenario, and Outcomes for Exposure Concentration or Dose. The
PECO statement, as depicted in Appendix H.5.14 of the 2021 Draft Systematic Review Protocol (U.S.
EPA. 202la), was refined to better identify references that may contain information relevant for this
draft supplement. Specifically, data that are relevant for characterizing exposure to 1,4-dioxane in food,
including biota that humans consume, was not evaluated and extracted because 1,4-dioxane is not
expected to bioaccumulate in organi sms likely to be consumed by humans. During TIAB screening, if it
is unclear if a reference will meet the PECO screening criteria without having the full reference to
review, or if a reference is determined to meet the PECO screening criteria, that reference will advance
to FT screening. Studies containing potentially relevant supplemental material were also tracked and

Page 193 of 484


-------
6015

6016

6017

6018

6019

6020

6021

6022

6023

6024

6025

6026

6027

6028

6029

6030

6031

6032

6033

6034

6035

6036

6037

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

categorized during the literature screening process. Relevant supplemental material may be reviewed,
evaluated for data quality, and incorporated into risk evaluations, as needed. For example, references
were considered supplemental if they contained data from countries outside of North America on 1,4-
dioxane levels associated with landfills because different countries have very different waste
management policies (including requirements for landfills), and local hydrogeology in other regions
may not be relevant to sites in the United States. FigureApx C-3 presents the number of references that
report general population, consumer, and environmental exposure data that passed PECO screening
criteria at TIAB and FT screening.

C.3.2.1 General Population, Consumer, and Environmental Exposure Literature Tree

©

Monitoring Study

(-)

©

Peer-Reviewed Literature	Modeling Study

Search

©

Experimental Study

Meets PECO Criteria at FT
Screening

9	@

Gray Literature Search	Database

0

1,4-Dioxane Supplement:	Does	PECO Criteria at

_	_	TIAB Screening

General Population, Consumer,	"

Does Not Meet PECO Criteria at
FT Screening

©

and Environmental Exposure	Supplemental From FT

Screening
©

Supplemental From TIAB
Screening

Figure Apx C-3. Literature Inventory Tree - General Population, Consumer, and Environmental
Exposure Search Results for 1,4-Dioxane

View the interactive literature inventory tree in HAWC. Data in this figure represent all references obtained from
the publicly available databases and gray literature references searches that were included in systematic review as
of November 28, 2022. Additional data may be added to the interactive version as they become available.

C.4 Data Evaluation and Data Extraction	

Data evaluation and extraction for this draft supplement are as described in Sections 5 and 6 of the 2021
Draft Systematic Review Protocol (U.S. EPA. 2021a). Data evaluation is the systematic review step in
which EPA assesses quality of the individual data sources using the evaluation strategies and criteria for
each topic area (e.g., physical and chemical property data, fate and transport data, occupational exposure
and environmental release data, general population, consumer, and environmental exposure data). The
evaluation method uses a structured framework with predefined criteria for each type of

Page 194 of 484


-------
6038

6039

6040

6041

6042

6043

6044

6045

6046

6047

6048

6049

6050

6051

6052

6053

6054

6055

6056

6057

6058

6059

6060

6061

6062

6063

6064

6065

6066

6067

6068

6069

6070

6071

6072

6073

6074

6075

6076

6077

6078

6079

6080

6081

6082

6083

6084

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

data/information source. The goal of the method used by EPA is to provide transparency, consistency,
and as much objectivity as possible to the evaluation process along with meeting the TSCA science
standards. Data extraction is the systematic review step in which EPA identifies quantitative and
qualitative data/information from data/information sources that meet screening criteria and extract the
data/information using structured forms or templates.

As explained above in Section C.l, terminology updates were made regarding the description of both
metric and overall data/information source quality determinations from what was originally described in
the 2021 Draft Systematic Review Protocol (U.S. EPA. 2021a). Specifically, metric and overall
data/information source quality determination terminology have been updated to "Metric Rating" and
"Overall Quality Determination", respectively. For additional clarifications regarding these updates,
please see TableApx C-l.

Although references that meet screening criteria following FT screening will generally proceed to data
quality evaluation and extraction steps, one clarification to the procedures outlined in Section 6 of the
2021 Draft Systematic Review Protocol is that in situations where EPA is unable to extract
data/information from sources that meet screening criteria (e.g., formatting prohibits accurate
extraction), such sources may not have extracted data to present in the risk evaluation or the respective
supplemental documents. Systematic Review Supplemental Files for the draft supplement contain results
from the data quality evaluation and extraction systematic review steps. Also, the template used to
display the data may be modified from those that were provided in the 2021 Draft Systematic Review
Protocol (	21a) because the purpose of these supplemental documents is to accommodate

the data needs for each respective risk evaluation. The following sections provide specific information
about the data quality and extraction process followed to address the exposure pathways in scope for this
draft supplement and any clarifications or updates regarding these systematic review steps as described
in the 2021 Draft Systematic Review Protocol (	1021a).

C.4.1 Environmental Release and Occupational Exposure	

As described in the 2021 Draft Systematic Review Protocol, evaluation and extraction followed the
steps outlined in Sections 5, 6, and 6.2 (	021a). The data extraction and data quality results

are summarized in Table Apx E-8 for air, Table Apx E-4 for water, Table Apx E-6 for land, and
Table Apx F-33 for occupational exposure. The 1,4-Dioxane Supplement to the Risk Evaluation Data
Quality Evaluation and Data Extraction Information for Environmental Release and Occupational
Exposure provides the results from the data extraction and quality evaluation, including metric rating
and the overall quality determination for each data source (U.S. EPA. 2023t).

C.4.2 General Population, Consumer, and Environmental Exposure

As described in the 2021 Draft Systematic Review Protocol, evaluation and extraction generally
followed the steps outlined in Section 5 and 6 (	021a). However, a few updates were made to

the data quality evaluation metrics for a few evidence streams since the metrics were published in the
2021 Draft Systematic Review Protocol. Most of the changes were editorial or minor clarifications,
including the standardization of some metrics that apply to multiple evidence streams, where
appropriate. For example, in the Quality Assurance/Quality Control (QA/QC) metric for evaluating
monitoring and experimental evidence streams, the acronym QA/QC was defined and replaced all
references to quality assurance and quality control when occurring separately or together, and the term
QA/QC techniques was changed to QA/QC measures, which already appeared in the metrics.

A few metrics applicable to multiple evidence streams were modified slightly to better fit some of the
unique situations that frequently arise for a certain type of evidence stream (e.g., databases). For

Page 195 of 484


-------
6085

6086

6087

6088

6089

6090

6091

6092

6093

6094

6095

6096

6097

6098

6099

6100

6101

6102

6103

6104

6105

6106

6107

6108

6109

6110

6111

6112

6113

6114

6115

6116

6117

6118

6119

6120

6121

6122

6123

6124

6125

6126

6127

6128

6129

6130

6131

6132

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

example, some metrics were updated to clarify the intent of the metric and better account for variation in
types of evidence included in one grouping (e.g., experiments involving chamber studies vs. product
concentration assessments). The domains did not change; however, see below for the changes and
updates made to the data evaluation metrics for the respective evidence types (i.e., monitoring,
experimental studies and databases) as presented in Sections C.4.2.2, C.4.2.3 and C.4.2.4. No changes
were made to the data evaluation metrics for modeling data, as described in Appendix N Section N.6.2
in the 2021 Draft Systematic Review Protocol. The 1,4-Dioxane Supplement to the Risk Evaluation
Data Quality Evaluation Information for General Population, Consumer, and Environmental Exposure
provides details of the data quality evaluation results, including metric rating and the overall quality
determination for each data source (	023 r).

Data extraction is the process in which quantitative and qualitative data/information are identified from
each relevant data/information source and extracted using structured forms or templates. Data extraction
was conducted as described in Section 6 of the 2021 Draft Systematic Review Protocol for all evidence
streams relevant for this supplement. However, with respect to information stored within databases, EPA
does not conduct a separate data extraction because the data are more accessible and have additional
context in the original database format. Both the date and data present in the database when the database
underwent FT screening are available in the HERO database (HERO IDs: 10365582, 10365609,
10365665, 10365667, 10365696, 10365698, 10368680, 10410586, and 10501014). If a reference or
data/information source (e.g., a peer-reviewed reference) presents data from a database that did not
undergo systematic review (e.g., a foreign database that is not publicly accessible), the data would be
extracted from the reference or data/information source to the extent possible; this did not apply to
references or sources of data or information that underwent systematic review for this supplement.

As mentioned above in Section C.4, references may not undergo data extraction, regardless of data
quality rating, if they contain no extractable data points (e.g., values are contained in a non-digitizable
figure or are representative of unspecified media or treatment processes). This constitutes an update to
Section 6 of the 2021 Draft Systematic Review Protocol (	021a). Extraction forms and

templates are tailored to fit the data extraction needs for each risk evaluation.

The types of fields extracted vary by evidence stream and generally followed Section 6.3 of the 2021
Draft Systematic Review Protocol with regard to the data characteristics captured (	021a).

Examples of types of data extracted and the extraction formats for the four evidence streams identified
through systematic review to evaluate environmental, general population, and consumer exposure data
are listed in the extraction tables provided in the 1,4-Dioxane Supplement to the Risk Evaluation Data
Extraction Information for General Population, Consumer, and Environmental Exposure (

2023h).

C.4.2.1 Data Quality Evaluation Metric Updates

Shown below are the data evaluation metrics for three evidence streams, presenting which data
evaluation metrics changed since the publication of the 2021 Draft Systematic Review Protocol (]j _S

21a). For evidence streams not listed below, there were no changes to the data evaluation
metrics since the 2021 Draft Systematic Review Protocol was published. Other data quality criteria for
studies on consumer, general population, and environmental exposure appear in Appendix N of the 2021
Draft Systematic Review Protocol (	)21a). For example, the criteria for modeling studies

appear in Table Apx N9. Data quality criteria for other types of studies (e.g., environmental release and
occupational exposure assessment) are published in other appendices to the 2021 Draft Systematic
Review Protocol (	021a).

Page 196 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

6133	For the below tables in Sections C.4.2.2, C.4.2.3 and C.4.2.4, in order to make it easier for the reader to

6134	see what the changes were to the data evaluation metrics, the following convention is used: text inserted

6135	is underlined, and text deleted is in strikeout.

6136

6137

6138

C.4.2.2 Data Evaluation Criteria for Monitoring Data, as Revised
Table Apx C-2. Evaluation Criteria for Sources of Monitoring Data	

Quality Rating

Description

Domain I Rcliabilil\

Mclnc I Sam pi i ny nvlhoi.loloij\

High

Samples were collected according to publicly available SOPs that are scientifically sound and
widely accepted (i.e., from a source generally using known to use sound methods and/or
approaches) for the chemical and media of interest. Example SOPs include U.S. Geological
Survey (USGS') "National Field Manual for the Collection of Water-Quality Data," EPA's
"Ambient Air Sampling" (SESDPROC-303-R5), etc.

OR

The sampling protocol used was not a publicly available SOP from a source generally known to
use using sound methods and/or approaches, but the sampling methodology is clear, appropriate
(i.e., scientifically sound), and similar to widely accepted protocols for the chemical and media
of interest. All pertinent sampling information is provided in the data source or companion
source. Examples include:

•	Sampling equipment

•	Sampling procedures/regime

•	Sample storage conditions/duration

•	Performance/calibration of sampler

•	Study site characteristics

•	Matrix characteristics

Medium

Sampling methodology is discussed in the data source or companion source and is generally
appropriate (i.e., scientifically sound) for the chemical and media of interest; however, one or
more pieces of sampling information is not described. The missing information is unlikely to
have a substantial impact on results.

OR

Standards, methods, protocols, or test guidelines may not be widely accepted, but a successful
validation study for the new/unconventional procedure was conducted prior to the sampling
event and is consistent with sound scientific theory and/or accepted approaches. Or a review of
information indicates the methodology is acceptable and differences in methods are not
expected to lead to lower quality data.

Low

Sampling methodology is only briefly discussed; therefore, most sampling information is

missing and likely to have a substantial impact on results.

AND/OR

The sampling methodology does not represent best sampling methods, protocols, or
guidelines for the chemical and media of interest (e.g., outdated [but still valid] sampling
equipment or procedures, long storage durations).

AND/OR

There are some inconsistencies in the reporting of sampling information (e.g., differences
between text and tables in data source, differences between standard method and actual
procedures reported to have been used, etc.) that led to a low confidence in the sampling
methodology used.

Page 197 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Critically
Deficient

The sampling methodology is not discussed in the data source or companion source.

AND/OR

Sampling methodology is not scientifically sound or is not consistent with widely accepted
methods/approaches for the chemical and media being analyzed (e.g., inappropriate sampling
equipment, improper storage conditions).

AND/OR

There are numerous inconsistencies in the reporting of sampling information, resulting in high
uncertainty in the sampling methods used.

Not Rated/Not
Applicable



Reviewer's
communis

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Mel lie 2 AikiK lical mclhodolo-jN

High

Samples were analyzed according to publicly available analytical methods that are scientifically
sound and widelv accepted (i.e., from a source generally using known to use sound methods
and/or approaches) and are appropriate for the chemical and media of interest. Examples
include EPA SW-846 Methods, NIOSH Manual of Analytical Methods 5th Edition, etc.

OR

The analytical method used was not a publicly available method from a source generally using
known to use sound methods and/or approaches, but the methodoloav is clear and appropriate
(i.e., scientifically sound) and similar to widely accepted protocols for the chemical and media
of interest. All pertinent sampling information is provided in the data source or companion
source. Examples include:

•	Extraction method

•	Analytical instrumentation (required)

•	Instrument calibration

•	Limit of quantitation (LOQ), LOD, detection limits, and/or reporting limits

•	Recovery samples

•	Biomarker used (if applicable)

•	Matrix-adjustment method (i.e., creatinine, lipid, moisture)



Medium

Analytical methodology is discussed in detail and is clear and appropriate (i.e., scientifically
sound) for the chemical and media of interest; however, one or more pieces of analytical
information is not described. The missing information is unlikely to have a substantial impact
on results.

AND/OR

The analytical method may not be standard/widely accepted, but a method validation
study was conducted prior to sample analysis and is expected to be consistent with sound
scientific theory and/or accepted approaches.

AND/OR

Samples were collected at a site and immediately analyzed using an on-site mobile laboratory,
rather than shipped to a stationary laboratory.

Low

Analytical methodology is only briefly discussed. Analytical instrumentation is provided and
consistent with accepted analytical instrumentation/methods. However, most analytical
information is missing and likely to have a substantial impact on results.

AND/OR

Analytical method is not standard/widely accepted, and method validation is limited or not

available.

AND/OR

Samples were analyzed using field screening techniques.

AND/OR

Page 198 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023



LOQ, LOD, detection limits, and/or reporting limits not reported.

AND/OR

There are some inconsistencies or possible errors in the reporting of analytical information
(e.g., differences between text and tables in data source, differences between standard method
and actual procedures reported to have been used, etc.) which leads to a lower confidence in the
method used.

Critically
Deficient

Analytical methodology is not described, including analytical instrumentation (i.e., HPLC,
GC).

AND/OR

Analytical methodology is not scientifically appropriate for the chemical and media being
analyzed (e.g., method not sensitive enough, not specific to the chemical, out of date).
AND/OR

There are numerous inconsistencies in the reporting of analytical information, resulting in high
uncertainty in the analytical methods used.

Not Rated/
Not Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

\lelrie 3 Seleelion of biomarker of exposure

High

Biomarker in a specified matrix is known to have an accurate and precise quantitative
relationship with external exposure, internal dose, or target dose (e.g., previous studies (or the
current study) have indicated the biomarker of interest reflects external exposures).

AND

Biomarker (parent chemical or metabolite) is derived from exposure to the chemical of interest.

Medium

Biomarker in a specified matrix has accurate and precise quantitative relationship with external

exposure, internal dose, or target dose.

AND

Biomarker is derived from multiple parent chemicals, not only the chemical of interest, but
there is a stated method to apportion the estimate to only the chemical of interest

Low

Biomarker in a specified matrix has accurate and precise quantitative relationship with external

exposure, internal dose, or target dose.

AND

Biomarker is derived from multiple parent chemicals, not only the chemical of interest, and
there is NOT an accurate method to apportion the estimate to only the chemical of interest.

OR

Biomarker in a specified matrix is a poor surrogate (low accuracy and precision) for
exposure/dose.

Critically
Deficient

Not applicable. A study will not be deemed critically deficient based on the use of biomarker of
exposure.

Not rated/
applicable

Metric is not applicable to the data source.

Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Domain 2 Represenlali\e

Melric 4 ( ieouranhic area

Page 199 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

High

Geographic location(s) is reported, discussed, or referenced.

Medium

Not applicable. This metric is dichotomous (i.e., high vs. critically deficient).

Low

Not applicable. This metric is dichotomous (i.e., high vs. critically deficient).

Critically
Deficient

Geographic location is not reported, discussed, or referenced.

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

\lclnc 5 Tcmnoralih

High

Timing of sample collection for monitoring data is consistent with current or recent exposures
(within 5 years) may be expected.

Medium

Timing of sample collection for monitoring data is less consistent with current or recent
exposures (>5 to 15 years) may be expected.

Low

Timing of sample collection for monitoring data is not consistent with when current exposures
(>15 years old) may be expected and likely to have a substantial impact on results.

Critically
Deficient

Timing of sample collection for monitoring data is not reported, discussed, or referenced.

Not Rated/
Not Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Melnc 10 or more samples for a single scenario).

•	Use of replicate samples.

•	Use of systematic or continuous monitoring methods.

•	Sampling over a sufficient period of time to characterize trends.

•	For urine, 24-hour samples are collected (vs. first morning voids or spot).

•	For biomonitoring studies, the timing of sample collected is appropriate based on
chemical properties (e.g., half-life), the pharmacokinetics of the chemical (e.g., rate of
uptake and elimination), and when the exposure event occurred.

Medium

Sampling approach likely captures variability of environmental contamination in
population/scenario/media of interest based on the heterogeneity/homogeneity and
dynamic/static state of the environmental system. Some uncertainty may exist, but it is unlikely
to have a substantial impact on results. For example:

•	Moderate sample size (i.e., 5-10 samples for a single scenario), or

•	Use of judgmental (non-statistical) sampling approach, or

•	No replicate samples.

•	For urine, first morning voids or pooled spot samples.

Low

Sampling approach poorly captures variability of environmental contamination in
population/scenario/media of interest. For example:

Page 200 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023



•	Small sample size (i.e., <5 samples), or

•	Use of haphazard sampling approach, or

•	No replicate samples, or

•	Grab or spot samples in single space or time, or

•	Random sampling that does not include all periods of time or locations, or

•	For urine, un-pooled spot samples.

Critically
Deficient

Sample size is not reported.

Single sample collected per data set.

For biomonitoring studies, the timing of sample collected is not appropriate based on chemical
properties (e.g., half-life), the pharmacokinetics of the chemical (e.g., rate of uptake and
elimination), and when the exposure event occurred.

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Mcli'ic 7 IAf>osuiv scenario

High

The data closely represent relevant exposure scenario (i.e., the population/scenario/media of
interest). Examples include:

•	Amount and type of chemical/product used

•	Source of exposure

•	Method of application or by-stander exposure

•	Use of exposure controls

•	Microenvironment (location, time, climate)

Medium

The data likely represent the relevant exposure scenario (i.e., population/scenario/media of
interest). One or more key pieces of information may not be described but the deficiencies
are unlikely to have a substantial impact on the characterization of the exposure scenario.
AND/OR

If surrogate data, activities seem similar to the activities within scope.

Low

The data lack multiple key pieces of information, and the deficiencies are likely to have a

substantial impact on the characterization of the exposure scenario.

AND/OR

There are some inconsistencies or possible errors in the reporting of scenario information
(e.g., differences between text and tables in data source, differences between standard method
and actual procedures reported to have been used, etc.) which leads to a lower confidence in the
scenario assessed.

AND/OR

If surrogate data, activities have lesser similarity but are still potentially applicable to the
activities within scope.

Critically
Deficient

If reported, the exposure scenario discussed in the monitored study does not represent the
exposure scenario of interest for the chemical.

Not rated/
Not applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Domain 3 Acccssilnlih clanl\

\Iclric N. Renoilintj ol'ivsulls

Page 201 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

High

Supplementary or raw data (i.e., individual data points) are reported, allowing summary

statistics to be calculated or reproduced.

AND

•	Description of data set summarized (i.e., location, population, dates, etc.)

•	Range of concentrations or percentiles

•	Number of samples in data set

•	Frequency of detection

•	Measure of variation (coefficient of variation [CV], standard deviation)

•	Measure of central tendency (mean, geometric mean, median)

•	Test for outliers (if applicable)

Summary statistics are detailed and complete. Example parameters include:

AND

Both adjusted and unadjusted results are provided (i.e., correction for void completeness in
urine biomonitoring, whole-volume or lipid adjusted for blood biomonitoring, wet or dry
weight for environmental tissue samples or soil samples) [only if applicable].

Medium

Supplementary or raw data (i.e., individual data points) are not reported, and therefore

summary statistics cannot be reproduced.

AND/OR

Summary statistics are reported but are missing one or more parameters (see description for

high).

AND/OR

Only adjusted or unadjusted results are provided, but not both [only if applicable].

Low

Supplementary data are not provided, and summary statistics are missing most parameters (see

description for high).

AND/OR

There are some inconsistencies or errors in the results reported, resulting in low confidence in
the results reported (e.g., differences between text and tables in data source, less appropriate
statistical methods).

Critically
Deficient

There are numerous inconsistencies or errors in the calculation and/or reporting of results,
resulting in highly uncertain reported results.

Not Rated/
Not Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Mel l ie ^ Oualil\ assurance

High

The study quality assurance/quality control (QA/QC) measures and all pertinent QA/QC
information is provided in the data source or companion source. Examples include:

•	Field, laboratory, and/or storage recoveries

•	Field and laboratory control samples

•	Baseline (pre-exposure) samples

•	Biomarker stability

•	Completeness of sample (i.e., creatinine, specific gravity, osmolality for urine
samples)

AND

No QA/QC issues were identified, or any identified issues were minor and adequately
addressed (i.e., correction for low recoveries, correction for completeness).

Medium

The study applied and documented QA/QC measures; however, one or more pieces of QA/QC
information is not described. Missing information is unlikely to have a substantial impact on
results.

Page 202 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023



AND

No QA/QC issues were identified, or any identified issues were minor and addressed (i.e.,
correction for low recoveries, correction for completeness).

Low

QA/QC measures and results were not directly discussed but are implied through the study's

use of standard field and laboratory protocols.

AND/OR

Deficiencies were noted in QA/QC control measures that are likely to have a substantial impact

on results.

AND/OR

There are some inconsistencies in the QA/QC measures reported, resulting in low confidence in
the QA/QC measures taken and results (e.g., differences between text and tables in data
source).

Critically
Deficient

QA/QC issues have been identified which significantly interfere with the overall reliability of
the study.

Not Rated/
Not Applicable



Reviewer's
commcnls

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Domain 4 Variabilis and unccrlains

\lclric In Variabilis and unccrlains

High

The slud) characterizes \ ariabi hL\ in Lhc population, media sludicd.
AND

Key uncertainties, limitations, and data gaps have been identified.
AND

The uncertainties are minimal and have been characterized.

Medium

The study has limited characterization of variability in the population/media studied.
AND/OR

The study has limited discussion of key uncertainties, limitations, and data gaps.

AND/OR

Multiple uncertainties have been identified but are unlikely to have a substantial impact on
results.

Low

The characterization of variability is absent.

AND/OR

Key uncertainties, limitations, and data gaps are not discussed.

AND/OR

Uncertainties identified may have a substantial impact on the exposure the exposure assessment

Critically
Deficient

Estimates are highly uncertain based on characterization of variability and uncertainty.

Not Rated/
Not Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Page 203 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

6139	C.4.2.3 Data Evaluation Criteria for Experimental Data, as Revised

6140

6141	Table Apx C-3. Evaluation Criteria for Sources of Experimental Data	

Data Quality
Rating

Metric Description

Domain 1 Rcliabilil\

\lclnc 1 Sam pi i nu \lclhodoloij\ and ( oiidilions

High

Samples were collected according to publicly available SOPs, methods, protocols, or test
guidelines that are scientifically sound and widely accepted from a source generally known to
use sound methods and/or approaches such as EPA, NIST, American Society for Testing and
Materials, ISO, and ACGIH.

OR

The sampling protocol used was not a publicly available SOP from a source generally known
to use sound methods and/or approaches, but the sampling methodology is clear, appropriate
(i.e.. scientifically sound), and similar to widely accepted protocols for the chemical and
media of interest. All pertinent sampling information is provided in the data source or
companion source. Examples include:

•	Sampling conditions (e.g., temperature, humidity)

•	Sampling equipment and procedures

•	Sample storage conditions/duration

•	Performance/calibration of sampler

Medium

Sampling methodology is discussed in the data source or companion source and is generally
appropriate (i.e.. scientifically sound) for the chemical and media of interest, however, one or
more pieces of sampling information is not described. The missing information is unlikely to
have a substantial impact on results.

OR

Standards, methods, protocols, or test guidelines may not be widely accepted, but a successful
validation study for the new/unconventional procedure was conducted prior to the sampling
event and is consistent with sound scientific theory and/or accepted approaches.

Low

Sampling methodology is only briefly discussed. Therefore, most sampling information is

missing and likely to have a substantial impact on results.

AND/OR

The sampling methodology does not represent best sampling methods, protocols, or guidelines
for the chemical and media of interest (e.g., outdated (but still valid) sampling equipment or
procedures, long storage durations).

AND/OR

There are some inconsistencies in the reporting of sampling information (e.g., differences
between text and tables in data source, differences between standard method and actual
procedures reported to have been used, etc.) which lead to a low confidence in the sampling
methodology used.

Critically
Deficient

The sampling methodology is not discussed in the data source or companion source.

AND/OR

Sampling methodology is not scientifically sound or is not consistent with widely accepted
methods/approaches for the chemical and media being analyzed (e.g., inappropriate sampling
equipment, improper storage conditions).

AND/OR

There are numerous inconsistencies in the reporting of sampling information, resulting in high
uncertainty in the sampling methods used.

Page 204 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

High

Samples were analyzed according to publicly available analytical methods that are
scientifically sound and widely accepted (i.e., from a source generally using sound methods
and/or approaches) and are appropriate for the chemical and media of interest. Examples
include EPA SW-846 Methods, NIOSH Manual of Analytical Methods 5th Edition, etc.

OR

The analytical method used was not a publicly available method from a source generally
known to use sound methods and/or approaches, but the methodology is clear and appropriate
(/. e., scientifically sound) and similar to widely accepted protocols for the chemical and media
of interest. All pertinent analytical sampling information is provided in the data source or
companion source. Examples include:

•	Extraction method

•	Analytical instrumentation (required)

•	Instrument calibration

•	LOQ, LOD, detection limits, and/or reporting limits

•	Recovery samples

•	Biomarker used (if applicable)

•	Matrix-adjustment method (B creatinine, lipid, moisture)



Medium

Analytical methodology is discussed in detail and is clear and appropriate (i.e.. scientifically
sound) for the chemical and media of interest; however, one or more pieces of analytical
information is not described. The missing information is unlikely to have a substantial impact
on results.

AND/OR

The analytical method may not be standard/widely accepted, but a method validation study
was conducted prior to sample analysis and is expected to be consistent with sound scientific
theory and/or accepted approaches.

AND/OR

Samples were collected at a site and immediately analyzed using an on-site mobile laboratory,
rather than shipped to a stationary laboratory.

Low

Analytical methodology is only briefly discussed. Analytical instrumentation is provided and
consistent with accepted analytical instrumentation/methods. However, most analytical
information is missing and likely to have a substantial impact on results.

AND/OR

Analytical method is not standard/widely accepted, and method validation is limited or not

available.

AND/OR

Samples were analyzed using field screening techniques.

AND/OR

LOQ, LOD, detection limits, and/or reporting limits not reported.

AND/OR

There are some inconsistencies or possible errors in the reporting of analytical information
(e.g., differences between text and tables in data source, differences between standard method
and actual procedures reported to have been used, etc.) which leads to a lower confidence in
the method used.

Critically
Deficient

Analytical methodology is not described, including analytical instrumentation (/. e., HPLC,
GC).

AND/OR

Page 205 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023



Analytical methodology is not scientifically appropriate for the chemical and media being
analyzed (e.g., method not sensitive enough, not specific to the chemical, out of date).
AND/OR

There are numerous inconsistencies in the reporting of analytical information, resulting in high
uncertainty in the analytical methods used.

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Melnc 3 Selection ol'luomaiker ol'e\nosure

High

Biomarker in a specified maln\ is known lo ha\ e an accurate and precise quantilaln e
relationship with external exposure, internal dose, or target dose (e.g., previous studies (or the
current study) have indicated the biomarker of interest reflects external exposures).

AND

Biomarker (parent chemical or metabolite) is derived from exposure to the chemical of
interest.

Medium

Biomarker in a specified matrix has accurate and precise quantitative relationship with

external exposure, internal dose, or target dose.

AND

Biomarker is derived from multiple parent chemicals, not only the chemical of interest, but
there is a stated method to apportion the estimate to only the chemical of interest

Low

Biomarker in a specified matrix has accurate and precise quantitative relationship with

external exposure, internal dose, or target dose.

AND

Biomarker is derived from multiple parent chemicals, not only the chemical of interest, and
there is NOT a stated method to apportion the estimate to only the chemical of interest.

OR

Biomarker in a specified matrix is a poor surrogate (low accuracy and precision) for
exposure/dose.



Critically
Deficient

Not applicable. A study will not be deemed critically deficient based on the use of biomarker
of exposure.

Not Rated/Not
Applicable

Metric is not applicable to the data source.

Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Metric 4 Testing

scenario

High

Testing conditions closely represent relevant exposure scenarios (i.e.,
population/scenario/media of interest). Examples include:

•	Amount and type of chemical/product used

•	Source of exposure/test substance

•	Method of application or by-stander exposure

•	Use of exposure controls

•	Microenvironment (location, time, climate, temperature, humidity, pressure, airflow)
AND

Testing conducted under a broad range of conditions for factors such as temperature,
humidity, pressure, airflow, and chemical mass/weight fraction (if appropriate).

Page 206 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Medium

The data likely represent the relevant exposure scenario (i.e., population/scenario/media of
interest). One or more key pieces of information may not be described but the deficiencies are
unlikely to have a substantial impact on the characterization of the exposure scenario.
AND/OR

If surrogate data, activities seem similar to the activities within scope.

Low

The data lack multiple key pieces of information and the deficiencies are likely to have a

substantial impact on the characterization of the exposure scenario.

AND/OR

There are some inconsistencies or possible errors in the reporting of scenario information (e.g.,
differences between text and tables in data source, differences between standard method and
actual procedures reported to have been used, etc.) which leads to a lower confidence in the
scenario assessed.

AND/OR

If surrogate data, activities have lesser similarity but are still potentially applicable to the

activities within scope.

AND/OR

Testing conducted under a single set of conditions^except for experiments to determine a
weight fraction or concentration in a product.

Critically
Deficient

Testing conditions are not relevant to the exposure scenario of interest for the chemical.

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Melnc 5 Saninle si/.e and \analnlil\

High

Sample size is reported and large enough (i.e., >10 samples) to be reasonably assured that the

samples represent the scenario of interest.

AND

Replicate tests performed and variability across tests is characterized (if appropriate).

Medium

Sample size is moderate (i.e., 5 to <10 samples), thus the data are likely to represent the

scenario of interest.

AND

Replicate tests performed and variability across tests is characterized (if appropriate).

Low

Sample size is small (i.e., <5 samples for most types of experiments or 1 per product for
experiments to determine a weight fraction or concentration in a product), thus the data are
likely to poorly represent the scenario of interest.

AND/OR

Replicate tests were not performed.

Critically
Deficient

Sample size is not reported.

AND/OR

Single sample collected per data set,.except for experiments to determine a weight fraction or

concentration in a product.

AND/OR

For biomonitoring studies, the timing of sample collected is not appropriate based on chemical
properties (e.g., half-life), the pharmacokinetics of the chemical (e.g., rate of uptake and
elimination), and when the exposure event occurred.

Not Rated/Not
Applicable



Page 207 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

\lclnc (¦> Taii|iomlii\

High

Source(s) of tested items appears to be current (within 5 years).

Medium

Source(s) of tested items is less consistent with when current or recent exposures (>5 to 15
years) are expected.

Low

Source(s) of tested items is not consistent with when current or recent exposures (>15 years)
are expected or is not identified.

Critically
Deficient

Temporality of tested items is not reported, discussed, or referenced.

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Domain 3 Acccssilnlil\ clanl\

Mcli'ic 7. Kqioriinu nl'ivsulls

High

Supplementary or raw data (i.e.. individual data points) are reported, allowing summary

statistics to be calculated or reproduced.

AND

Summary statistics are detailed and complete. Example parameters include:

•	Description of data set summarized (i.e., location, population, dates, etc.)

•	Range of concentrations or percentiles

•	Number of samples in data set

•	Frequency of detection

•	Measure of variation (CV, standard deviation)

•	Measure of central tendency (mean, geometric mean, median)

•	Test for outliers (if applicable)

AND

Both adjusted and unadjusted results are provided (i.e., correction for void completeness in
urine biomonitoring, whole-volume or lipid adjusted for blood biomonitoring) [only if
applicable].

Medium

Supplementary or raw data (i.e.. individual data points) are not reported, and therefore

summary statistics cannot be reproduced.

AND/OR

Summary statistics are reported but are missing one or more parameters (see description for

high).

AND/OR

Only adjusted or unadjusted results are provided, but not both [only if applicable].

Low

Supplementary data are not provided, and summary statistics are missing most parameters (see

description for high).

AND/OR

There are some inconsistencies or errors in the results reported, resulting in low confidence in
the results reported (e.g., differences between text and tables in data source, less appropriate
statistical methods).

Critically
Deficient

There are numerous inconsistencies or errors in the calculation and/or reporting of results,
resulting in highly uncertain reported results.

Page 208 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Not Rated/Not
Applicable



Reviewer's
commenls

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Mel lie S OualiS assurance

High

The study applied quality assurance/quality control (QA/QC) measures and all pertinent
QA/QC information is provided in the data source or companion source. Examples include:

•	Laboratory, and/or storage recoveries.

•	Laboratory control samples.

•	Baseline (pre-exposure) samples.

•	Biomarker stability

•	Completeness of sample (i.e., creatinine, specific gravity, osmolality for urine
samples)

AND

No QA/QC issues were identified, or any identified issues were minor and adequately
addressed (i.e., correction for low recoveries, correction for completeness).

Medium

The study applied and documented QA/QC measures; however, one or more pieces of QA/QC
information is not described. Missing information is unlikely to have a substantial impact on
results.

AND

No QA/QC issues were identified, or any identified issues were minor and addressed (i.e.,
correction for low recoveries, correction for completeness).

Low

QA/QC measures and results were not directly discussed but are implied through the study's

use of standard field and laboratory protocols.

AND/OR

Deficiencies were noted in QA/QC measures that are likely to have a substantial impact on

results.

AND/OR

There are some inconsistencies in the QA/QC measures reported, resulting in low confidence
in the QA/QC measures taken and results (e.g., differences between text and tables in data
source).

Critically
Deficient

QA/QC issues have been identified which significantly interfere with the overall reliability of
the study.

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Domain 4 Variabilis and unceilains

\Iclnc Variabilis and unceilaum

High

The study characterizes variability in the population/media studied.
AND

Key uncertainties, limitations, and data gaps have been identified.
AND

The uncertainties are minimal and have been characterized.

Medium

The study has limited characterization of variability in the population/media studied.
AND/OR

Page 209 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023



The study has limited discussion of key uncertainties, limitations, and data gaps.

AND/OR

Multiple uncertainties have been identified but are unlikely to have a substantial impact on
results.

Low

The characterization of variability is absent.

AND/OR

Key uncertainties, limitations, and data gaps are not discussed.

AND/OR

Uncertainties identified may have a substantial impact on the exposure the exposure
assessment

Critically
Deficient

Estimates are highly uncertain based on characterization of variability and uncertainty.

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

6142	C.4.2.4 Data Evaluation Criteria for Databases, as Revised

6143

Table Apx C-4

. Evaluation Criteria for Sources of Database Data

Data Quality
Rating

Description

Domain 1 Rcliubilih

Mclnc 1. Samnlinu nvlhodolou\

High

Widelv accepted sampling methodologies (i.e., from a source generally known to use using
sound methods and/or approaches) were used to generate the data presented in the database.
Example SOPs include USGS's "National Field Manual for the Collection of Water-Quality
Data," EPA's "Ambient Air Sampling" (SESDPROC-303-R5), etc.

Medium

One or more pieces of sampling methodology information is not described, but missing
information is unlikely to have a substantial impact on results.

OR

The sampling methodologies were consistent with sound scientific theory and/or accepted
approaches based on the reported sampling information but may not have followed published
procedures from a source generally known to use sound methods and/or approaches.



Low

The sampling methodologv was not reported in data source or readilv available companion data
source.

Critically
Deficient

The sampling methodologies used were not appropriate for the chemical/media of interest in
the database (e.g., inappropriate sampling equipment, improper storage conditions).

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

\lclnc 2 AnaK Heal mclhodolou\

Page 210 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

High

Widely accepted analytical methodologies (i.e., from a source generally using sound methods
and/or approaches) were used to generate the data presented in the database. Example SOPs
include EPA SW-846 Methods, NIOSH Manual of Analytical Methods 5th Edition, etc.

Medium

The analytical methodologies were consistent with sound scientific theory and/or accepted
approaches based on the reported analytical information but may not have followed published
procedures from a source generally known to use sound methods and/or approaches.

Low

The analytical methodology was not reported in data source or companion data source.

Critically
Deficient

The analytical methodologies used were not appropriate for the chemical/media of interest in
the database (e.g., method not sensitive enough, not specific to the chemical, out of date).

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Domain 2 Rqnvsailali\c

\lclnc 3 ( ieouianhic aiva

High

Geographic location(s) is reported, discussed, or referenced.

Medium

Not applicable. This metric is dichotomous (i.e., high vs. critically deficient).

Low

Not applicable. This metric is dichotomous (i.e., high vs. critically deficient).

Critically
Deficient

Geographic location is not reported, discussed, or referenced.

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

\lclnc4 Temporal

High

The data reflect current conditions (within 5 years)

AND/OR

Database contains robust historical data for spatial and temporal analyses (if applicable).

Medium

The data are less consistent with current or recent exposures (>5 to 15 years)

AND/OR

Database contains sufficient historical data for spatial and temporal analyses (if applicable).

Low

Data are not consistent with when current exposures (>15 years old) may be expected
AND/OR

Database does not contain enough historical data for spatial and temporal analyses (if
applicable).

Critically
Deficient

Timing of sample data is not reported, discussed, or referenced.

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Page 211 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Mclnc 5 1 \nnsuiv scenario

High

The data closely represent relevant exposure scenario (i.e., the population/scenario/media of
interest). Examples include:

•	Amount and type of chemical/product used

•	Source of exposure

•	Method of application or by-stander exposure

•	Use of exposure controls

•	Microenvironment (location, time, climate)

Medium

The data likely represent the relevant exposure scenario (i.e., population/scenario/media of
interest). One or more key pieces of information may not be described but the deficiencies
are unlikely to have a substantial impact on the characterization of the exposure scenario.
AND/OR

If surrogate data, activities seem similar to the activities within scope.

Low

The data lack multiple key pieces of information and the deficiencies are likely to have a

substantial impact on the characterization of the exposure scenario.

AND/OR

There are some inconsistencies or possible errors in the reporting of scenario information
(e.g., differences between text and tables in data source, differences between standard method
and actual procedures reported to have been used, etc.) which leads to a lower confidence in the
scenario assessed.

AND/OR

If surrogate data, activities have lesser similarity but are still potentially applicable to the
activities within scope.

Critically
Deficient

If reported, the exposure scenario discussed in the monitored study does not represent the
exposure scenario of interest for the chemical.

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Domain 3 \ccessilnlil\ clanl\

\lclnc ^ A\ailaliilil\ ol'dalalxise and sunnorlinu documenls

High

Database is widely accepted and/or from a source generally known to use sound methods
and/or approaches (e.g., raw data from NHANES. STORET).

Medium

The database may not be widely known or accepted (e.g., state-maintained databases), but the
database is adequately documented with most or all of the following information:

•	Within the database, metadata is present (sample identifiers, annotations, flags, units,
matrix descriptions, etc.) and-data fields are generally clear and defined.

•	A user manual and other supporting documentation is available, or there is sufficient
documentation in the data source or companion source.

•	Database quality assurance and data quality control measures are defined and/or a
QA/QC protocol was followed.

Low

The database may not be widely known or accepted, and only limited database documentation
is available (see the medium rating).

Critically
Deficient

No information is provided on the database source or availability to the public.

Page 212 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Not Rated/
Applicable



Reviewer's
communis

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Mclnc 7 Rcinu'linu of ivsulls

High

The database or information source reporting the analvsis of the database data is well organized

and understandable by the target audience.

AND

Summary statistics in the data source are detailed and complete. Example parameters include:

•	Description of data set summarized (i.e., location, population, dates, etc.)

•	Range of concentrations or percentiles

•	Number of samples in data set

•	Frequency of detection

•	Measure of variation (CV, standard deviation)

•	Measure of central tendency (mean, geometric mean, median)

•	Test for outliers (if applicable)

Medium

The database or information source reporting the analysis of the database data is well organized

and understandable by the target audience.

AND/OR

Summary statistics are missing one or more parameters (see description for high).

Low

The database or information source reporting the analysis of the database data is unclear or not

well organized.

AND/OR

Summary statistics are missing most parameters (see description for high)

AND/OR

There are some inconsistencies or errors in the results reported, resulting in low confidence in
the results reported (e.g., differences between text and tables in data source, less appropriate
statistical methods).

Critically
Deficient

There are numerous inconsistencies or errors in the calculation and/or reporting of results,

resulting in highly uncertain reported results.

AND/OR

The information source reporting the analysis of the database data is missing key sections or
lacks enough organization and clarity to locate and extract necessary information.

Not Rated/Not
Applicable



Reviewer's
Comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

Domain 4 \ anabilih and uiKvrlainl\

\lclnc S Yanalnlih and uiKvi'lanm

High

Variabilitv. kev uncertainties, limitations, and/or data saps have been identified.
AND/OR

The uncertainties are minimal and have been characterized.

Medium

The studv has limited discussion of variabilitv. kev uncertainties, limitations, and/or data saps.
AND/OR

Multiple uncertainties have been identified but are unlikely to have a substantial impact on
results.

Page 213 of 484


-------
6145

6146

6147

6148

6149

6150

6151

6152

6153

6154

6155

6156

6157

6158

6159

6160

6161

6162

6163

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Low

Variabilitv. kev uncertainties, limitations, and data saps are not discussed.

AND/OR

Uncertainties identified may have a substantial impact on the exposure the exposure assessment

Critically
Deficient

Estimates are highly uncertain based on characterization of variability and uncertainty.

Not Rated/Not
Applicable



Reviewer's
comments

[Document concerns, uncertainties, limitations, and deficiencies and any additional comments
that may highlight study strengths or important elements such as relevance]

C.5 Evidence Integration

As described in Section 7 of the 2021 Draft Systematic Review Protocol (U.S. EPA.. 2021a). evidence
integration refers to the consideration of evidence obtained from systematic review and scientific
information obtained from sources that did not undergo systematic review to implement a weight of the
scientific evidence approach. 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 CFR
702.33).The consideration of the quality and relevance of the data, while taking into account the
strengths and limitations of the data, to appropriately evaluate the evidence for this supplement, is
described in Section 7 of the 2021 Draft Systematic Review Protocol (U.S. EPA. 2021a).

TableApx C-5 and TableApx C-6, originally from Section 7.3 of the 2021 Draft Systematic Review
Protocol, provide general considerations and examples of factors that contribute to the strength of
evidence for each evidence stream and example weight of the scientific evidence judgments based on
these general considerations, respectively, when evaluating potentially relevant exposure data for this
draft supplement (	021a).

Table Apx C-5. Considerations that Inform Evaluations of the Strength of the Evidence

Considerations

Factors that
Increase Strength

Factors that
Decrease Strength

The overall weight of the scientific evidence judgment considers the general considerations below as well as
chemical-specific considerations to designate each exposure scenario as robust, moderate, slight, or
indeterminate. The designation is a measure of the weight of the evidence supporting the representativeness of
the exposure estimates toward the true distribution of exposure (and releases) for the scenario.

Relevance to
exposure scenario

• Directly relevant to evaluated
exposure scenario

• Data used is for an alternative or surrogate
scenario

For modeled
estimates

• Model used has been peer-

reviewed and is being applied in a
manner appropriate to its design
and objective

•	Evidence demonstrating implausibility

•	Model has not been peer-reviewed and no
ground-truthing has been performed

•	Parameterization is not well described,
documented or is not appropriate to the
evaluated scenario

Data quality

• Medium or high data quality
rating (via Data Evaluation)

•	Low data quality rating (via Data Evaluation)

•	Imprecision or inaccuracy

Page 214 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Considerations

Factors that
Increase Strength

Factors that
Decrease Strength

Data points

• High number of data points

•	Low number of data points

•	High proportion of data sampled prior to
changes in industry or other relevant
conditions (e.g., OSHA PEL)

Representative of
the whole industry
(for occupational
scenarios)

• Large proportion of sites included
within the exposure scenario were
measured

• Evidence may not be sufficiently
representative of all of the sites for the
exposure scenario

Representative of
the sub-population

• Applicable to most or all of the
different population groups
included within the exposure
scenario

• Information was not available to sufficiently
cover most or all of population groups
included within the exposure scenarios

Consistency

• Consistency and replication
within a study and across studies

• Inexplicable contradictory findings across
studies

Variability

•	Variability is accounted in
estimates

•	Full distributions of input
parameters

• Variability unaccounted in estimates

Uncertainties

• Uncertainties are low and the
uncertainties are unlikely to
significantly impact exposure
estimates

• Uncertainties that are likely to over- or under-
estimate exposure from the actual exposures
for the exposure scenario

6164

Page 215 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Table Apx C-6. Evaluation of the Weight of the Scientific Evidence for

Exposure Assessments

Category

Robust

Moderate

Slight

Indeterminate

Overall Weight of
the Scientific
Evidence

Exposure Scenario
Factors

(e.g., habits, worker
activities, exposure
factors)

•	Directly relevant to
evaluated exposure
scenario

•	Applicable to most or all of
the different population
groups included within the
exposure scenario

•	Full distributions of input
parameters

•	High or medium quality
data ratings

•	The habits, worker
activities, and/or use
patterns are accounted for,
are current

•	Uncertainties are low and
the uncertainties are
unlikely to significantly
impact exposure estimates

•	Surrogate scenarios from
similar chemicals are used
to infer similar exposures
or emissions.

•	Some distribution of input
parameters

•	High or medium quality
data ratings

•	There is some, but not
complete, documentation
or description of
assumptions, limitations
and uncertainties

•	Surrogate scenarios from
similar uses are used to
infer similar use patterns or
habits and practices

•	Medium or low quality
data ratings

•	Partially supported by
assumptions

•	Uncertainties are not
fully known or
documented

•	Habits and practices are
not fully known and
there is a high degree of
uncertainty in defining
use patterns

•	Qualitative
descriptions of
exposure without
additional context.

•	No supporting data
on habits and
practices are
available

The consideration
factors and the
categories to the left
result in an overall
weight of the
scientific evidence
judgment as one of
the following:

•	Robust

•	Moderate

•	Slight

•	Indeterminate

Measured/Monitored
Data

There is measured information
and the temporal and spatial
aspect of the measurements
are well described, relevant
and reflect current conditions

•	Medium or high data
quality rating (via Data
Evaluation)

•	High number of data points

•	Multiple studies or a large
number of data points
which indicate similar
findings

•	Large proportion of sites
included within the
exposure scenario were
measured

There is measured information
which does not reflect current
environmental conditions or
does not correspond to current
activities but provides
evidence of exposure

•	Limited number of studies
or limited number of data
points which indicate similar
findings

•	Information was not
available to sufficiently
cover most or all of
population groups included
within the exposure
scenarios

•	There is some, but not
complete, documentation or

There is limited measured
information and
information and does not
reflect exposure conditions
and does not correspond to
known activities

•	Information was not
available to sufficiently
cover most or all of
population groups
included within the
exposure scenarios

•	Assumptions and
uncertainties are not
known or documented

No measured or
monitored data are
available

Page 216 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Category

Robust

Moderate

Slight

Indeterminate

Overall Weight of
the Seientifie
Evidcnee



•	Consistency and
replication within a study
and across studies

•	Uncertainties are low and
the uncertainties are
unlikely to significantly
impact exposure estimates

•	Sensitivity of the exposure
estimates has been
described and quantified
incorporating assumptions,
limitations, and
uncertainties

description of assumptions,
limitations, and uncertainties





The consideration
factors and the
categories to the left
result in an overall
weight of the
scientific evidence
judgment as one of
the following:

•	Robust

•	Moderate

•	Slight

•	Indeterminate

The consideration
factors and the
categories to the left
results in an overall
weight of the
scientific evidence
judgment as one of
the following:

•	Robust

•	Moderate

•	Slight

•	Indeterminate

Estimation
Methodology/Data

•	The methodology for
deriving the estimate is
well described and the
underlying computational
and/or scientific basis is
robust, has an empirical
basis or well documented
mathematical basis and
considers chemical
specificity (e.g., physical
and chemical properties
and fate)

•	Applicable to most or all
of the different population
groups included within the
exposure scenario
(representative)

•	Sensitivity of the exposure
estimates has been
described and quantified
incorporating assumptions,
limitations, and
uncertainties

•	The methodology for
deriving the estimate is
well described and the
underlying computational
and/or scientific basis is
robust, however there is
uncertainty in the
parameterization or
applicability

•	There is some, but not
complete, documentation
or description of
assumptions, limitations
and uncertainties.

•	Modeling approach used
to estimate exposures is
not rooted in scientific
rigor or does not
mathematically
represent the exposure
scenario;

parameterization is not
complete or does not
utilize the best available
science.

•	Assumptions and
uncertainties are not
known or documented

• Modeling approach
is not available for
the scenario or lack
of information on
parameters
prohibits use of
available models.

Comparison of
Estimated and
Measured Exposures (if

• There are comparable
estimates using alternate
approaches

• Modeled estimates and
measured exposure values
are comparable, however

• There is a lack of
correspondence between
measured exposures and

• Category does not
have indeterminate
criterion.

Page 217 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Category

Robust

Moderate

Slight

Indeterminate

Overall Weight of
the Seientifie
Evidcnee

both estimated and
measured estimates are
used)

•	There is concordance
between measured and/or
reported and modeled
estimates/predictions for
the same exposure scenario

•	Sensitivity of the exposure
estimates has been
described and quantified
incorporating assumptions,
limitations, and
uncertainties

differences in
methodology, collection,
or context make it difficult
to arrive at full
concordance
• There is some, but not
complete, documentation
or description of
assumptions, limitations
and uncertainties

modeled exposure
estimates even when
uncertainty and
variability are accounted
for.

• Assumptions and
uncertainties are not
known or documented





6166

Page 218 of 484


-------
6167

6168

6169

6170

6171

6172

6173

6174

6175

6176

6177

6178

6179

6180

6181

6182

6183

6184

6185

6186

6187

6188

6189

6190

6191

6192

6193

6194

6195

6196

6197

6198

6199

6200

6201

6202

6203

6204

6205

6206

6207

6208

6209

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

C.5.1 Environmental Release and Occupational Exposure

EPA evaluated environmental releases based on reported release data, modeling approaches, and
industry sector information from standard engineering sources such as TRI and DMR. As described in
Appendix E, EPA estimated COU-specific releases where supporting data existed and documented
uncertainties where an absence of such data required a broader application of release estimates.

EPA evaluated occupational exposures based on monitoring data, modeling approaches, and worker
activity information from standard engineering sources and systematic review as described in Appendix
F. EPA used COU-specific assessment approaches where supporting data existed and documented
uncertainties where supporting data were only applicable for broader assessment approaches.

C.5.2 General Population

General population exposures were evaluated for each exposure pathway based on environmental
release data identified as described above in Section C.4.1, environmental monitoring data identified
through available databases or as described in Section C.4.2, and any other relevant information
identified through systematic review. As described in Section 1, all physical and chemical and fate
properties evaluated in the 2020 RE were used to evaluate the in-scope exposure pathways of the
supplement.

C.5.2.1 General Population: Surface Water

To evaluate the surface water pathway, EPA relied on modeled surface water concentrations based on
environmental release data reported to TRI and DMR (Appendix E.3.1) and releases modeled for other
release types, including DTD and hydraulic fracturing (Appendix E.3.2).

EPA identified ambient surface water monitoring data through the WQP, drinking water monitoring
from PWSs through the UCMR3 database and three state-specific databases (Section 2.3.1.1). EPA used
available surface water monitoring data to confirm the accuracy of model predictions in location-
specific case-studies (Appendix G.2.3.2). In addition, available drinking water monitoring data (see 1,4-
Dioxane Supplement to the Risk Evaluation Data Extraction Information for General Population,
Consumer, and Environmental Exposure) were used to provide context and a point of reference for
modeled drinking water concentrations and risk estimates (Section 5.2.2.1.5) (	23h).

C.5.2.2 General Population: Groundwater

To evaluate the land pathway (groundwater) releases, EPA relied on environmental release data reported
to TRI (Section 2.2.1.1 and Appendix E.4.1) and releases modeled for hydraulic fracturing operations
(Appendix E.5.2).

EPA identified groundwater monitoring data for 1,4-dioxane through the WQP as presented in Section
2.3.2.1 and described in Appendix H.l. Furthermore, EPA contextualized potential groundwater
concentrations identified in the literature through systematic review (see 1,4-Dioxane Supplement to the
Risk Evaluation Data Extraction Information for General Population, Consumer, and Environmental
Exposure) using search terms identified in Appendix C.2 (	IK).

C.5.2.3 General Population Exposure: Ambient Air

EPA did not identify quantitative outdoor air monitoring data for 1,4-dioxane through systematic
review. To evaluate the air pathway, EPA relied on modeled air concentrations based on industrial
releases reported to TRI (Section 2.3.3.2.2 and Appendix E.5.1), releases modeled for laundry facilities

Page 219 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

6210	(Section 2.3.3.2.4 and Appendix E.5.2), and releases modeled for hydraulic fracturing operations

6211	(Section 2.3.3.2.4 and Appendix E.5.2).

Page 220 of 484


-------
6212

6213

6214

6215

6216

6217

6218

6219

6220

6221

6222

6223

6224

6225

6226

6227

6228

6229

6230

6231

6232

6233

6234

6235

6236

6237

6238

6239

6240

6241

6242

6243

6244

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Appendix D COU-OES MAPPING AND CROSSWALK

This appendix contains additional information about the relationship between the COUs and OESs
determined for 1,4-dioxane.

Condition of Use (CPU): TSCA § 3(4) defines COUs 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'. COUs included in the scope of
EPA's risk evaluations are typically tabulated in scope documents and risk evaluation documents as
summaries of life cycle stages, categories, and subcategories of use. Therefore, a COU is composed of a
combination of life cycle stage, category, and subcategory. COU development may include Chemical
Data Reporting (CDR) information, market profile information, and literature sources. Early in the risk
evaluation process, EPA maps each COU to an occupational exposure scenario for the environmental
release and occupational exposure assessment.

Occupational Exposure Scenario (PES): This term is intended to describe the grouping or segmenting
of COUs for assessment of releases and exposures. For example, EPA may assess a group of multiple
COUs together as one OES due to similarities in release and exposure sources, worker activities, and use
patterns. Alternatively, EPA may assess multiple OES for one COU because there are different release
and exposure potentials for a given COU. OES determinations are also largely driven by the availability
of data and modeling approaches to assess occupational releases and exposures. For example, even if
there are similarities between multiple COUs, if there is sufficient data to separately assess releases and
exposures for each COU, EPA would not group them into the same OES.

D.l COU-OES Mapping

The details of an identified COU will determine the number of associated OES(s). Mapping OES to
COUs may come in many forms, including a direct one-to-one mapping of a single OES to a single
COU, mapping of one OES to multiple COUs, or mapping of multiple OES to a single COU, as shown
in Figure Apx D-l. The OES mapping is driven by similarities and differences in the expected
occupational exposures and releases for a COU and the reasonably available data to estimates such
exposures and releases, as discussed in Section 2.1.1. Further, there may be differences in the name of
an OES from the name of the COU to which it is mapped. This is because OES names are intended to be
succinct, capture all COUs where one OES is mapped to multiple COUs, and distinct enough to
represent the specific occupational exposure and release scenario.

Page 221 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

COU1

OES1

One COU may map to one OES

There may be differences in the COU and OES names because the OES name is intended to be
succinct and specific to the assessed occupational releases and exposures

For example, the 1,4-dioxane COU for "Byproduct produced during the production of
polyethylene terephthalate" maps only to the OES named "PET byproduct" (see excerpt from
crosswalk Table 2-1 andApxD-1 below)

6245

Condition of Use

OES

Life Cycle
Stage

Category

Subcategory

Processing

Byproduct

Byproduct produced during
production of PET

PET Byproduct

Multiple COUs may be mapped to the same OES

Multiple COUs may be mapped to one OES when the COUs have similar activities and exposure
potentials, and exposures and releases can be assessed for the COUs using a single approach

There may be differences between the COU and OES names because the OES is name is intended to
be succinct and encompass all COUs grouped therein under a general name

For example, the 1,4-dioxane COUs for "Industrial wastewater treatment", "underground injection",
"municipal landfill", and others were assessed together under the OES named "disposal" (see excerpt
from crosswalk Table 2-1 and Apx D-l below)

COU liCOU2lCOU3

OES1

6246

Condition of Use



Life Cycle
Stage

Category

Subcategory

OES





Industrial pre-treatment







Industrial wastewater treatment







Publicly owned treatment
works (POTW)







Underground injection



Disposal

Disposal

Municipal landfill

Disposal

Hazardous landfill





Other land disposal







Municipal waste incinerator







Hazardous waste incinerator







Off-site waste transfer



COU1

OESliOES2iOES3

One COU may be mapped to multiple OES

Mapping a COU to multiple OES allows for the assessment of distinct scenarios that are not
expected to result in similar releases and exposures

There may be differences between the COU and OES names because the OES capture more distinct
scenarios of occupational release and exposure than the COU

For example, the 1,4-dioxane COU for "dish soap, dishwasher detergent, laundry detergent" (which
is a single COU) was assessed as separate OES named "dish soap", "dishwasher detergent",
"laundry detergent (industrial)", and "laundry detergent (institutional)" (see excerpt from crosswalk
Table 2-1 and Apx D-l below)

Condition of Use

OES

Life Cycle
Stage

Category

Subcategory

Consumer use,

commercial

use

Laundry and
Dishwashing
Products

Dish soap

Dishwasher detergent
Laundry detergent

Dish Soap

Dishwasher Detergent
Laundry Detergent

(Industrial)
Laundry Detergent
(Institutional)

6248	Figure Apx D-1. COU and OES Mapping

6249

Page 222 of 484


-------
6250

6251

6252

6253

6254

6255

6256

6257

6258

6259

6260

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

D.2 CQU-QES Crosswalk

A crosswalk of the COU with the OES assessed is provided in TableApx D-l. As discussed in Section
2.1.1, a COU is a combination of life cycle stage, category, and subcategory and EPA mapped each
COU to an OES. The purpose of an OES is to group, where appropriate, COUs based on similarity of
the operations and data availability for each COU. EPA assessed environmental releases (air, water, and
land) and occupational exposures (inhalation and dermal) to 1,4-dioxane for each of the OES listed in
TableApx D-l. As noted in this table, some of these OESs were in scope of the Final Risk Evaluation
for 1,4-Dioxane (	320c) while others were in scope of this supplemental risk evaluation.

Table Apx D-l. Categories and Subcategories of Conditions of Use Included in the Scope of the
Risk Evaluation

Condition of Use

OES

Risk Evaluation in Which
Occupational Exposures
Were Assessed

Life Cycle
Stage

Category"

Subcategory''

Manufacturing

Domestic
Manufacture

Domestic Manufacture

Manufacturing

2020 RE

Import

Import
Repackaging

Import and Repackaging

2020 RE

Processing

Processing as a
Reactant

Polymerization catalyst

Industrial Uses

2020 RE

Non-

incorporative

Basic organic chemical
manufacturing (process
solvent)

Byproduct

Byproduct produced
during processes

Ethoxylation Process
Byproduct

Supplemental RE

Byproduct produced
during production of PET

PET Byproduct

Supplemental RE

Recycling

Recycling

Disposal

2020 RE

Distribution in
commerce

Distribution

Distribution

Distribution activities
(e.g., loading) considered
throughout life cycle,
rather than using a single
distribution scenario

N/A

Industrial use

Intermediate use

Plasticizer intermediate

Catalysts and reagents for
anhydrous acid reactions,
brominations, and
sulfonations

Industrial Uses

2020 RE

Processing aids,
not otherwise
listed

Wood pulpingc
Extraction of animal and
vegetable oilsc

Wetting and dispersing
agent in textile
processing17

Polymerization catalyst

Purification of process
intermediates

Etching of fluoropolymers

Page 223 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Condition of Use

OES

Risk Evaluation in Which
Occupational Exposures
Were Assessed

Life Cycle
Stage

Category"

Subcategory''



Functional
fluids (open and
closed system)

Polyalkylene glycol
lubricant

Functional Fluids (Open-
System)

2020 RE

Synthetic metalworking
fluid

Cutting and tapping fluid

Hydraulic fluid

Functional Fluids
(Closed-System)

2020 RE

Industrial use,

commercial

use

Laboratory
Chemicals

Chemical reagent

Laboratory Chemicals

2020 RE

Reference material

Spectroscopic and
photometric measurement

Liquid scintillation
counting medium

Stable reaction medium

Cryoscopic solvent for
molecular mass
determinations

Preparation of histological
sections for microscopic
examination

Adhesives and
Sealants

Film cement

Film Cement

2020 RE

Other Uses

Spray polyurethane foam;
Printing and printing
compositions, including
3D printing; dry film
lubricant; Hydraulic
fracturing

Spray Foam Application

2020 RE

Printing Inks (3D)

2020 RE

Dry Film Lubricant

2020 RE

Hydraulic Fracturing

Supplemental RE

Consumer
use,

commercial
use

Paints and
Coatings

Latex wall paint or floor
lacquer

Paint and Floor Lacquer

Supplemental RE

Cleaning and
Furniture Care
Products

Surface cleaner

Surface Cleaner

Supplemental RE

Laundry and
Dishwashing
Products

Dish soap

Dishwasher detergent
Laundry detergent

Dish Soap

Dishwasher Detergent
Laundry Detergent

(Industrial)
Laundry Detergent
(Institutional)

Supplemental RE

Arts, Crafts, and

Hobby

Materials

Textile dye

Textile Dye

Supplemental RE

Automotive
Care Products

Antifreeze

Antifreeze

Supplemental RE

Other Consumer
Uses

Spray polyurethane foam

Spray Foam Application

2020 RE

Disposal

Disposal

Industrial pre-treatment

Disposal

2020 RE

Page 224 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Condition of Use

OES

Risk Evaluation in Which
Occupational Exposures
Were Assessed

Life Cycle
Stage

Category"

Subcategory''





Industrial wastewater
treatment

Publicly owned treatment
works (POTW)

Underground injection

Municipal landfill

Hazardous landfill

Other land disposal

Municipal waste
incinerator

Hazardous waste
incinerator

Off-site waste transfer

" These categories of conditions of use reflect Chemical Data Reporting (CDR) rule codes and broadly represent conditions
of use for 1,4-dioxane in industrial and/or commercial settings.
b These subcategories reflect more specific uses of 1,4-dioxane.

6261

Page 225 of 484


-------
6262

6263

6264

6265

6266

6267

6268

6269

6270

6271

6272

6273

6274

6275

6276

6277

6278

6279

6280

6281

6282

6283

6284

6285

6286

6287

6288

6289

6290

6291

6292

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Appendix E INDUSTRIAL AND COMMERCIAL
ENVIRONMENTAL RELEASES

This appendix contains additional information relevant to the assessment of industrial and commercial
environmental releases.

E.l Estimates of the Number of Industrial and Commercial Facilities with
Environmental Releases

As a part of the assessment of industrial and commercial environmental releases, EPA estimated the
number of facilities with releases for each OES. Where available, EPA used 2013 to 2019 TRI (

12e) and 2013-2019 DMR (	:022c) data to provide a basis to estimate the number of

sites using 1,4-dioxane within an OES. Additional information on how EPA utilized TRI and DMR to
estimate the number of sites using 1,4-dioxane within a COU can be found in Section 2.2.1.2.2 of the
December 2020 Final Risk Evaluation for 1,4-Dioxane (	20c).

Where the number of sites could not be determined using TRI or DMR or where these data were
determined to not capture the entirety of sites within an OES, EPA supplemented the available data with
U.S. economic data using the following methods:

•	Identify the North American Industry Classification System (NAICS) codes for the industry
sectors associated with these uses.

•	Estimate total number of sites using the U.S. Census" Statistics of US Businesses (SUSB) (U.S.
Census Bureau. ) data on total establishments by 6-digit NAICS.

•	Review available ESDs and GSs for established facility estimates for each occupational exposure
scenario.

•	Combine the data generated in bullets 1 through 3 to produce an estimate of the number of sites
using 1,4-dioxane in each 6-digit NAICS code and sum across all applicable NAICS codes for
the COU, augmenting as needed with data from the ESDs and GSs, to arrive at a total estimate of
the number of sites within the COU.

A summary of the number of facilities EPA determined for each OES and each type of release is shown
in Table Apx E-l. The number of facilities may be different for each type of release within the same
OES if sufficient data were available to make this differentiation.

Table Apx E-l. Summary of EPA's Estimates for the Number of Facilities for Each OES

OES

Type of Release

Number of
Facilities

Notes

Manufacturing

Air, Land

1

Based on 2019 TRI reDortina (U.S. EPA. 2022s).

Surface Water

1

Based on 2013-2019 DMR and TRI reporting (U.S.

EPA. 2022c. e).

POTW or Non-
POTWWWT

1

Based on 2013-2019 TRI reDortina (U.S. EPA. 2022s).

Import and
Repackaging

Air, Land

1

Based on 2019 TRI reDortina (U.S. EPA. 2022s).

Surface Water

6

Based on 2013-2019 DMR and TRI reporting (U.S.

EPA. 2022c. s).

POTW or Non-
POTWWWT

6

Based on 2013-2019 TRI reDortina (U.S. EPA. 2022s).

Page 226 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Type of Release

Number of
Facilities

Notes

Industrial Uses

Air, Land

12

Based on 2019 TRI reporting (U.S. EPA. 2022s).

Surface Water

24

Based on 2013-2019 DMR and TRI reporting (U.S.

EPA. 2022c. a).

POTW or Non-
POTWWWT

17

Based on 2013-2019 TRI reoortina (U.S. EPA. 2022s).

Functional
Fluids (Open-
System)

Air, Land

2

Based on 2019 TRI reDortina (U.S. EPA. 2022s).

Surface Water

6

Based on 2013-2019 DMR and TRI reporting (U.S.

EPA. 2022c. a).

POTW or Non-
POTWWWT

1

Based on 2013-2019 TRI reDortina (U.S. EPA. 2022a).

Functional
Fluids (Closed-
System)

All

N/A

Assessed as a part of Industrial Uses OES.

Laboratory
Chemical

All

132

Calculated using the GS on Use of Laboratory
Chemicals (U.S. EPA. 2022h) and the amount of 1.4-
dioxane used in laboratory uses per the December 2020
Final Risk Evaluation for 1,4-Dioxane (

2020c).

Film Cement

All

211

Based on the number of sites for this OES in the
December 2020 Final Risk Evaluation for 1,4-Dioxane
(U.S. EPA, 2020c). which is a bounding estimate based
on U.S. Census Bureau data forNAICS code 512199,
Other Motion Picture and Video Industries.

Spray Foam
Application

All

1,553,559

Based on the number of sites for this OES in the
December 2020 Final Risk Evaluation for 1,4-Dioxane
(U.S. EPA. 2020c). which is a bounding estimate based
on U.S. Census Bureau data forNAICS code 238310,
Dry wall and Insulation Contractors.

Printing Inks
(3D)

Air, Land

N/A

Assessed as a part of Industrial Uses OES.

Surface Water,
POTW, non-POTW
WWT

1

Based on 2013-2019 DMR and TRI reporting (U.S.

EPA. 2022c. a).

Dry Film
Lubricant

All

8

Based on the number of sites for this OES in the
December 2020 Final Risk Evaluation for 1,4-Dioxane
(U.S. EPA, 2020c). which is based on conversations
with the Kansas City National Security Campus
(manufacturer and uses of dry film lubricants).

Disposal

Air

15

Based on 2019 TRI reporting (U.S. EPA, 2022s).

Surface Water

24

Based on 2013-2019 DMR and TRI reporting (U.S.

EPA, 2022c. e).

POTW or Non-
POTW WWT,
Land

4

Based on 2013-2019 TRI reDortina (U.S. EPA. 2022s).

Page 227 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Type of Release

Number of
Facilities

Notes

Textile Dye

All

783

Bounding estimate based on U.S. Census Bureau data
for NAICS code 313310, Textiles and Fabric Finishing
Mills.

Antifreeze

All

84,383

Bounding estimate based on U.S. Census Bureau data
for NAICS codes 811111, General Automotive Repair,
and 811198, All Other Automotive Repair and
Maintenance.

Surface Cleaner

All

Unknown
within

Liverpool OH;

55,998

(industry

bounding

estimate)

Land release estimates for this OES were developed for
the Liverpool, OH case study and the number of sites
within this locality is unknown (e.g., the release
estimates are not per site but for the entire locality).
Bounding estimate for the industry is based on U.S.
Census Bureau data for NAICS code 561720, Janitorial
Services.

Dish Soap

All

Unknown
within

Liverpool OH;

773,851

(industry

bounding

estimate)

Land release estimates for this OES were developed for
the Liverpool, OH, case study and the number of sites
within this locality is unknown (e.g., the release
estimates are not per site but for the entire locality).
Bounding estimate for the industry is based on U.S.
Census Bureau data for NAICS codes 623300, 713900,
721100, 721300, 722300, 722400, and 722500.

Dishwasher
Detergent

All

Unknown
within

Liverpool OH;

773,851

(industry

bounding

estimate)

Land release estimates for this OES were developed for
the Liverpool, OH, case study and the number of sites
within this locality is unknown (e.g., the release
estimates are not per site but for the entire locality).
Bounding estimate for the industry is based on U.S.
Census Bureau data for NAICS codes 623300, 713900,
721100, 721300, 722300, 722400, and 722500.

Laundry

Detergent

(Institutional)

All

95,533

Bounding estimate based on industry information as
described in the ESD on Water Based Washing
operations at Industrial and Institutional Laundries
(OECD.: ).

Laundry

Detergent

(Industrial)

All

2,453

Bounding estimate based on U.S. Census Bureau data
for NAICS code 812330, Linen and Uniform Supply.

Paints and Floor
Lacquer

All

33,648

Bounding estimate based on U.S. Census Bureau data
for NAICS code 811121, Automotive Body, Paint, and
Interior Repair and Maintenance.

Polyethylene
Terephthalate
(PET)
Byproduct

Air, Land

13

Based on 2019 TRI reporting (IIS. EPA. 2022s).

Surface Water

19

Based on 2013-2019 DMR and TRI reporting (U.S.

EPA. 2022c. a).

POTW or Non-
POTWWWT

14

Based on 2013-2019 TRI reDortina (U.S. EPA. 2022s).

Ethoxylation

Process

Byproduct

Air, Land

8

Based on 2019 TRI reDortine (U.S. EPA. 2022s).

Surface Water

7

Based on 2013-2019 DMR and TRI reporting (U.S.

EPA. 2022c. s).

Page 228 of 484


-------
6293

6294

6295

6296

6297

6298

6299

6300

6301

6302

6303

6304

6305

6306

6307

6308

6309

6310

6311

6312

6313

6314

6315

6316

6317

6318

6319

6320

6321

6322

6323

6324

6325

6326

6327

6328

6329

6330

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Type of Release

Number of
Facilities

Notes



POTW orNon-
POTWWWT

6

Based on 2013-2019 TRI reDortine (TJ.S. EPA. 2022a).

Hydraulic
Fracturing

All

411

Based on the number of sites that reported using 1,4-
dioxane to FracFocus 3.0 (GWPC and IOGCC. 2022).

E.2 Estimates of Number of Release Days for Industrial and Commercial
Releases

As a part of the assessment of industrial and commercial environmental releases, EPA also estimated the
number of release days for each OES. EPA referenced the December 2020 Final Risk Evaluation for
1,4-Dioxane (U.S. EPA. 2020c). Generic Scenarios (GS), Emission Scenario Documents (ESDs), or
made assumptions when estimating release days for each OES. In summary, EPA estimated the number
of operating days using the below sources of data:

1.	Facility-Specific Data: Use facility-specific data if available. If facility-specific data is not
available, estimate the days/year using one of the following approaches:

a.	If facilities have known or estimated average daily use rates, calculate the days/year as:
Days/year = Estimated Annual Use Rate for the Site (kg/year) / average daily use rate
from sites with available data (kg/day).

b.	If sites with days/year data do not have known or estimate average daily use rates, use the
average number of days/year from the sites with such data.

2.	Industry-Specific Data: Industry-specific data may be available in the form of GSs, ESDs, trade
publications, or other relevant literature. In such cases, these estimates should take precedent
over other approaches, unless facility-specific data are available.

3.	Manufacture of Lower-PV Specialty Chemicals: For the manufacture of lower-PV specialty
chemicals like 1,4-dioxane, the chemical is not expected to be manufactured continuously
throughout the year. Therefore, a value of 250 days/year should be used. This assumes the plant
manufactures the chemical 5 days/week and 50 weeks/year (with 2 weeks down for turnaround).
For the manufacture of 1,4-dioxane as a byproduct (e.g., ethoxylation process, PET
manufacturing), 250 days/year is also used, assuming these industrial manufacturing facilities
have a similar operating schedule of 5 days/week and 50 weeks/year.

4.	Processing as Reactant (Intermediate Use) in the Manufacture of Specialty Chemicals:

Similar to #3, the manufacture of specialty chemicals is not expected to occur continuously
throughout the year. Therefore, a value of 250 days/year can be used.

5.	Other Chemical Plant OES (e.2.. Industrial Uses): For these OESs, it is reasonable to assume
that 1,4-dioxane is not always in use at the facility, even if the facility operates 24/7. Therefore,
in general, a value of 300 days/year can be used based on the "SpERC [Specific Environmental
Release Categories] fact sheet - Formulation & (re)packing of substances and mixtures -
Industrial (Solvent-borne)" which uses a default of 300 days/year for the chemical industry.
However, in instances where the OES uses a low volume of the chemical of interest, 250
days/year can be used as a lower estimate for the days/year.

6.	POTWs: Although POTWs are expected to operate continuously over 365 days/year, the
discharge frequency of 1,4-dioxane from a POTW will be dependent on the discharge patterns of
the chemical from the upstream facilities discharging to the POTW. However, there can be
multiple upstream facilities (possibly with different OES) discharging to the same POTW and

Page 229 of 484


-------
6331

6332

6333

6334

6335

6336

6337

6338

6339

6340

6341

6342

6343

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

information to determine when the discharges from each facility occur on the same day or
separate days is typically not available. Therefore, an exact number of days/year the 1,4-dioxane
is discharged from the POTW cannot be determined and a value of 365 days/year should be
used.

7. All Other OESs: Regardless of the facility operating schedule, other OESs are unlikely to use
1,4-dioxane every day. Therefore, a value of 250 days/year should be used for these OESs.

A summary along with a brief explanation is presented in TableApx E-2 below. These estimates of
release days are applicable to the air and water release estimates for each OES; however, there is a high
level of variability and uncertainty associated with the number of days of release associated with land
releases. Therefore, EPA could not estimate the number of days of release for land releases.

Table Apx E-2. Summary of EPA's Estimates for Air and Water Release Days Expected for Each

OES

OES

Release
Days

Notes

Manufacturing

250

Assumed 5 days per week and 50 weeks per year with 2 weeks per year for
shutdown activities, consistent with the 2020 RE (U.S. EPA. 2020c).

Import and
Repackaging

250

Assumed 5 days per week and 50 weeks per year with 2 weeks per year for
shutdown activities, consistent with the December 2020 Final Risk Evaluation
for 1.4-Dioxane ("U.S. EPA. 2020c).

Industrial Uses

250

Assumed 5 days per week and 50 weeks per year with 2 weeks per year for
shutdown activities, consistent with the December 2020 Final Risk Evaluation
for 1.4-Dioxane (U.S. EPA. 2020c).

Functional
Fluids (Open-
System)

247

Per the 2011 OECD Emission Scenario Document on the Use of Metalworking
Fluids, consistent with the December 2020 Final Risk Evaluation for 1,4-
Dioxane (U.S. EPA. 2020c).

Functional
Fluids (Closed-
System)

N/A

Assessed as a part of Industrial Uses OES.

Laboratory
Chemical

250

Assumed 5 days per week and 50 weeks per year with 2 weeks per year for
shutdown activities, consistent with the December 2020 Final Risk Evaluation
for 1.4-Dioxane (U.S. EPA. 2020c).

Film Cement

250

Assumed 5 days per week and 50 weeks per year with 2 weeks per year for
shutdown activities, consistent with the December 2020 Final Risk Evaluation
for 1.4-Dioxane (U.S. EPA. 2020c).

Spray Foam
Application

3

Per the 2018 EPA generic scenario Application of Spray Polyurethane Foam
Insulation, consistent with the December 2020 Final Risk Evaluation for 1,4-
Dioxane (U.S. EPA. 2020c).

Printing Inks
(3D)

250

Assumed 5 days per week and 50 weeks per year with 2 weeks per year for
shutdown activities, consistent with the December 2020 Final Risk Evaluation
for 1.4-Dioxane (U.S. EPA. 2020c).

Dry Film
Lubricant

48

Per process description information provided in the December 2020 Final Risk
Evaluation for 1.4-Dioxane (U.S. EPA, 2020c).

Page 230 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Release
Days

Notes

Disposal

250

Assumed 5 days per week and 50 weeks per year with 2 weeks per year for
shutdown activities, consistent with the December 2020 Final Risk Evaluation
for 1.4-Dioxane (TJ.S. EPA. 2020c).

Textile Dye

31 to 295

Based on the 2015 OECD on Textile Dves (OECD. 2017) and Monte Carlo
Modeling.

Antifreeze

250

Assumed 5 days per week and 50 weeks per year with 2 weeks per year for
shutdown activities.

Surface Cleaner

250

Assumed 5 days per week and 50 weeks per year with 2 weeks per year for
shutdown activities.

Dish Soap

250

Assumed 5 days per week and 50 weeks per year with 2 weeks per year for
shutdown activities.

Dishwasher
Detergent

250

Assumed 5 days per week and 50 weeks per year with 2 weeks per year for
shutdown activities.

Laundry

Detergent

(Institutional)

250 to
365

Based on the 2011 OECD ESD on Industrial and Institutional Laundries
( M CD. JO I 1and Monte Carlo Modeling.

Laundry

Detergent

(Industrial)

20 to 365

Based on the 2011 OECD ESD on Industrial and Institutional Laundries
(¦ M CD, .'01 11») and Monte Carlo Modeling.

Paints and Floor
Lacquer

250

Based on the 2011 OECD ESD on Coating Application via Spray Painting in
the Automotive Refinishine Industry (OECD, 201 la).

Polyethylene
Terephthalate
(PET)
Byproduct

250

Assumed 5 days per week and 50 weeks per year with 2 weeks per year for
shutdown activities.

Ethoxylation

Process

Byproduct

250

Assumed 5 days per week and 50 weeks per year with 2 weeks per year for
shutdown activities.

Hydraulic
Fracturing

1 to 72

Based on the reported number of days for sites that use 1,4-dioxane in
FracFocus 3.0 (GWPC and IOGCC. 2022). This ranae of release davs refers to
only the hydraulic fracturing and not post-fracturing production stages. EPA's
estimates for flowback and produced water releases during production stages
occur over 350 davs/vear ("U.S. EPA. 2022d).

6344	E.3 Water Release Assessment

6345	This section describes EPA's methodology for estimating daily wastewater discharges from industrial

6346	and commercial facilities manufacturing, processing, or using 1,4-dioxane. Facilities report wastewater

6347	discharges either via Discharge Monitoring Reports (DMRs) under the NPDES or TRI. EPA used 2013

6348	to 2019 DMR (	22c) and 2013 to 2019 TRI (	2g) data to estimate daily

6349	wastewater discharges for the OES where available; however, EPA did not have these data for every

6350	OES. For OES without DMR and TRI data, EPA used alternate assessment approaches to estimate

6351	wastewater discharges. Both approaches, that for OESs with DMR and TRI data and that for OESs

6352	without these data, are described below.

Page 231 of 484


-------
6353

6354

6355

6356

6357

6358

6359

6360

6361

6362

6363

6364

6365

6366

6367

6368

6369

6370

6371

6372

6373

6374

6375

6376

6377

6378

6379

6380

6381

6382

6383

6384

6385

6386

6387

6388

6389

6390

6391

6392

6393

6394

6395

6396

6397

6398

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

E.3.1 Assessment Using TRI and DMR

EPA found 2013 to 2019 DMR and/or 2013 to 2019 TRI data for facilities within the following OESs:

•	Manufacturing,

•	Import and repackaging,

•	Industrial uses,

•	Functional fluids (open-system),

•	3D printing,

•	Disposal,

•	PET byproduct, and

•	Ethoxylation byproduct.

The 2013 to 2019 TRI data were rated "medium" in EPA's systematic review process and the 2013 to
2019 DMR were rated "medium." EPA estimated daily discharges using TRI and DMR data for these
OESs, with the following general stages as described in the rest of this section:

1.	Collect wastewater discharge data from 2013 to 2019 DMR and TRI data,

2.	Map wastewater discharge data to occupational exposure scenarios,

3.	Estimate the number of facility operating days per year, and

4.	Estimate daily wastewater discharges and summarize wastewater discharges for each OES.

Note that EPA compared the TRI and DMR data used to estimate water releases for the PET byproduct
OES in this risk evaluation to information from a life cycle analysis on the PET manufacturing process
in Appendix E.6.

Step 1: Collect Wastewater Discharge Data from DMR and TRI

The first step in estimating daily releases was to obtain 2013 through 2019 DMR and TRI data. Under
the CWA, EPA regulates the discharge of pollutants into receiving waters through NPDES. A NPDES
permit authorizes discharging facilities to discharge pollutants to specified limits. NPDES permits apply
pollutant discharge limits to each outfall at a facility. For risk evaluation purpose, EPA is interested only
in the outfalls to surface water body. NPDES permits also include internal outfalls, but they aren't
included in this analysis. This is because these outfalls are internal monitoring points within the facility
wastewater collection or treatment system, so they do not represent discharges from the facility. The
permits require facilities to monitor their discharges and report the results to EPA and the state
regulatory agency. Facilities report these results in DMRs. EPA makes these reported data publicly
available via EPA's ECHO system and EPA's Water Pollutant Loading Tool (Loading Tool). The
Loading Tool is a web-based tool that obtains DMR data through ECHO, presents data summaries and
calculates pollutant loading (mass of pollutant discharged). EPA queried the ECHO Loading Tool to
pull data for each of years 2013 through 2019. EPA removed facilities reporting zero discharges for 1,4-
dioxane in DMR from the analysis because EPA cannot confirm if the pollutant is present at the facility.

Each facility subject to the TRI reporting rule must report annually the volume of chemical released to
the environment and/or managed through recycling, energy recovery, and treatment. Unlike DMR, TRI
includes both reports of annual direct discharges to surface water and annual indirect discharges to off-
site publicly owned treatment works (POTW) and wastewater treatment (WWT) facilities (non-POTW
WWT). Similar to the air release assessment, EPA included both TRI reporting Form R and TRI
reporting Form A submissions in the water release assessment. Where sites reported to TRI with Form
A, EPA used the Form A threshold for total releases of 500 lb/year. EPA used the entire 500 lb/year for
both direct and indirect wastewater discharges; however, since this threshold is for total site releases,
these 500 lb/year are attributed either to direct discharges or indirect discharges for this analysis, not

Page 232 of 484


-------
6399

6400

6401

6402

6403

6404

6405

6406

6407

6408

6409

6410

6411

6412

6413

6414

6415

6416

6417

6418

6419

6420

6421

6422

6423

6424

6425

6426

6427

6428

6429

6430

6431

6432

6433

6434

6435

6436

6437

6438

6439

6440

6441

6442

6443

6444

6445

6446

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

both (since that would double count the releases and exceed the total release threshold for Form A). EPA
pulled the TRI Basic Plus Data Files for each of years 2013 through 2019.

In summary, wastewater discharges reported to DMR and TRI include the following:

•	DMR:

o	On-site releases to surface water (direct discharges).

•	TRI:

o	On-site releases to surface water (direct discharges),

o	Off-site transfers to POTWs (indirect discharges), and

o	Off-site transfers to non-POTW WWT (indirect discharges).

Note that the two datasets are not updated concurrently. The Loading Tool automatically and
continuously checks ICIS-NPDES for newly submitted DMRs. The Loading Tool processes the data
weekly and calculates pollutant loading estimates; therefore, water discharge data (DMR data) are
available on a continual basis. Although the Loading Tool process data weekly, each permitted
discharging facility is only required to report their monitoring results for each pollutant at a frequency
specified in the permit (e.g., monthly, every two months, quarterly). TRI data is reported annually for
the previous calendar year and is typically released in October (i.e., 2020 TRI data is released in October
2021).

Step 2: Map Wastewater Discharge Data to Occupational Exposure Scenarios
The next step in estimating daily releases was to map 2013 through 2019 DMR and TRI data to the 1,4-
dioxane OES. EPA used the same mapping methodology for the water assessment as that described in
Appendix E.5.1. EPA ensured consistency in the OES mapping for sites that reported to both TRI and
DMR. EPA also ensured consistency in the OES mapping between the air, water, and land assessments.

Step 3: Estimate the Number of Facility Operating Days per Year

EPA then estimated the number of operating days (days/year) for each facility reporting wastewater
discharges to DMR and TRI. EPA generally used the same number of operating days for the same OES
for both the air and water analysis, which is based on the general methodology described previously in
Appendix E.2.

Step 4: Estimate Daily Wastewater Discharges and Summarize Wastewater Discharges for each OES
After the initial steps of selecting and mapping of the water discharge data and estimating the number of
facility operating days/year were completed, the next step was to summarize annual and daily
wastewater discharges for each OES. EPA summarized annual wastewater discharges reported in DMR
and TRI for each facility. EPA estimated daily wastewater discharges separately for direct and indirect
discharges, as discussed below.

EPA estimated the median and maximum daily direct wastewater discharges at each facility, using the
steps below. EPA presented the calculated median and maximum daily direct wastewater discharged
separately for the DMR and TRI datasets because these data do not always agree/match.

1.	Obtained total annual loads calculated from the Loading Tool and reported annual surface water
discharges in TRI for years 2013 through 2019.

2.	Divided the annual direct discharge over the number of estimated operating days for the OES to
which the facility has been mapped. The number of operating days differ for each OES, as
summarized in Appendix E.2.

3.	Calculated the median daily direct wastewater discharge across all years of data for each facility,
separately for both DMR and TRI data.

Page 233 of 484


-------
6447

6448

6449

6450

6451

6452

6453

6454

6455

6456

6457

6458

6459

6460

6461

6462

6463

6464

6465

6466

6467

6468

6469

6470

6471

6472

6473

6474

6475

6476

6477

6478

6479

6480

6481

6482

6483

6484

6485

6486

6487

6488

6489

6490

6491

6492

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

4. Identified the maximum daily direct wastewater discharge across all years of data for each
facility. EPA also noted which reporting year had this maximum daily direct wastewater
discharge, separately for both DMR and TRI data.

For indirect discharges to POTW or non-POTW WWT, EPA estimated the average daily indirect
discharges for each facility and each reporting year (2013 through 2019) in TRI using steps #1 and #2
above. DMR data do not include indirect discharges. EPA did not estimate the median or maximum
daily indirect discharges across all years.

A summary of the estimated daily discharges using 2013 to 2019 DMR and TRI is included in 1,4-
Dioxane Supplemental Information File: Environmental Releases to Water for OES with TRI and DMR

(U.S. EPA. 2023ml

E.3.2 Assessment for OES without TRI and DMR

EPA did not find DMR or TRI data for any of the years included in this analysis for the following OESs:

•	Functional Fluids (Closed-Systems)

•	Laboratory Chemicals

•	Film Cement

•	Spray Polyurethane Foam

•	Dry Film Lubricant

•	Textile Dye

•	Antifreeze

•	Surface Cleaner

•	Dish Soap

•	Dishwasher Detergent

•	Laundry Detergent,

•	Paints and Floor Lacquer

•	Hydraulic Fracturing

For these OESs, EPA estimated daily wastewater discharges by using various modeling approaches—
including the use of surrogate TRI and DMR data and modeling using data from literature, GSs, and
ESDs. EPA's assessment of daily wastewater discharges for each of these OESs is described below.

Functional Fluids (Closed-Systems)

Wastewater discharge data were not available for this OES and EPA did not find any information to
model wastewater discharges for this OES using literature, GSs, or ESDs. EPA expects that the sources
of release for this OES to be similar to those for the Industrial Uses OES, based on the process
information in the Final Risk Evaluation for 1,4-Dioxane (	2020c). Therefore, EPA grouped

the water release assessment for Functional Fluids (Closed-Systems) into that for Industrial Uses.
However, there is uncertainty in this assumption of similar release sources between these OESs.

Laboratory Chemicals

EPA estimated daily wastewater discharges for facilities within the laboratory chemicals OESs using the
Draft GS on Use of Laboratory Chemicals (U.S. EPA. 2022h). The GS on Use of Laboratory Chemicals
was rated "high" during EPA's systematic review process.

Per the GS on Use of Laboratory Chemicals, water releases are not expected for hazardous chemicals.
Because 1,4-dioxane is considered a hazardous substance under CERCLA (40 CFR Part 302.4) and the
PubChem Hazardous Substances Data Bank (HSDB), there are no water releases for this OES. This is

Page 234 of 484


-------
6493

6494

6495

6496

6497

6498

6499

6500

6501

6502

6503

6504

6505

6506

6507

6508

6509

6510

6511

6512

6513

6514

6515

6516

6517

6518

6519

6520

6521

6522

6523

6524

6525

6526

6527

6528

6529

6530

6531

6532

6533

6534

6535

6536

6537

6538

6539

6540

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

consistent with the water release assessment for this OES in the Final Risk Evaluation for 1,4-Dioxane
(	1020c). which indicates that water releases are not expected for laboratory uses of 1,4-

dioxane.

Film Cement

EPA estimated daily wastewater discharges for facilities within the Film Cement OES using process
information from the Final Risk Evaluation for 1,4-Dioxane (U.S. EPA. 2020c). The underlying process
information for this assessment was rated "high" during EPA's systematic review process.

Per the risk evaluation, EPA does not expect water releases of 1,4-dioxane for this OES. EPA expects
the glue bottles to be disposed of as solid waste without. There is some uncertainty as to whether and
how much 1,4-dioxane may remain in the glue bottles when disposed. However, due to the small
quantities of the glue and high volatility of the 1,4-dioxane, EPA expects any residual 1,4-dioxane to
evaporate to the air or remain in the solid waste stream (	20c).

Spray Polyurethane Foam

EPA estimated daily wastewater discharges for facilities within the Spray Polyurethane Foam OES
using the same approach described for this OES in Appendix E.5.2, which is the use of the GS on
Application of Spray Polyurethane Foam Insulation (	»20c). The GS on the Application of

Spray Polyurethane Foam Insulation was rated "medium" during EPA's systematic review process.

The GS indicates that there are six release points:

1.	Releases to fugitive air for volatile chemicals during unloading of raw materials from transport
containers;

2.	Releases to water, incineration, or landfill from cleaning or disposal of transport containers;

3.	Releases to fugitive air for volatile chemicals during transport container cleaning;

4.	Releases to incineration or landfill from spray polyurethane foam application equipment
cleaning;

5.	Releases to fugitive air for volatile chemicals during equipment cleaning; and

6.	Releases to landfill of scrap foam from trimming applied foam.

Based on the GS, only release point #2 has the potential for wastewater discharges. To estimate this
release, EPA used the equations specified in the GS (	»20c). Apart from weight fraction in

spray polyurethan foam, EPA did not find any data specific to 1,4-dioxane in this OES. Therefore, the
calculation of releases using this GS are for a "generic site," using the default input parameter values
from the GS. Specifically, EPA used the same input parameter values that were used in the original risk
evaluation for estimates of occupational exposure; see Appendix G of the Final Risk Evaluation for 1,4-
Dioxane (U.S. EPA. 2020c).

Using this methodology, EPA calculated a range of wastewater releases for this OES. For the low-end,
EPA assumed there are no water releases, which is consistent with the GS explanation that containers
may be disposed of without rinsing. For the high-end, EPA assumed the containers may be rinsed /
poured down drains such that the entire release point #2 is to POTW. Direct water discharges are not
likely given the setting (construction/ renovation sites).

EPA's calculation of wastewater discharges for this OES, including all calculation inputs, can be found
in 1,4-Dioxane Supplemental Information File: Environmental Releases to Water for OES without TRI
or DMR data (11.S. EPA. 2023n).

Page 235 of 484


-------
6541

6542

6543

6544

6545

6546

6547

6548

6549

6550

6551

6552

6553

6554

6555

6556

6557

6558

6559

6560

6561

6562

6563

6564

6565

6566

6567

6568

6569

6570

6571

6572

6573

6574

6575

6576

6577

6578

6579

6580

6581

6582

6583

6584

6585

6586

6587

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Dry Film Lubricant

EPA estimated daily wastewater discharges for facilities within the Dry Film Lubricant OES using
process information from the Final Risk Evaluation for 1,4-Dioxane (	2020c). The underlying

process information for this assessment was rated "high" during EPA's systematic review process.

Per the risk evaluation, EPA does not expect water releases of 1,4-dioxane for this OES. Based on
conversations the with only known user, EPA expects wastes to be drummed and sent to a waste handler
with residual wastes releasing to air or being disposed to landfill. (	2020c).

Textile Dye

EPA estimated daily wastewater discharges for facilities within the Textile Dye OES using the OECD
ESD on Textile Dyes (	) and Monte Carlo modeling. The ESD on Textile Dyes was rated

"medium" during EPA's systematic review process. The use of Montel Carlo modeling allows for
variation of calculation input parameters such that a distribution of environmental releases can be
calculated, from which EPA can estimate the 50th and 95th percentile releases. An explanation of this
modeling approach is included in Appendix E. 11.

Antifreeze

EPA did not find any information to model wastewater discharges for this OES using literature, GSs, or
ESDs, nor does EPA expect this OES to be similar to other OES such that surrogate data may be used.
EPA evaluated the potential for releases using the OECD ESD on Chemical Additives used in
Automotive Lubricants (OECD. 2020) and the EPA MRD on Commercial Use of Automotive Detailing
Products (U.S. EPA. 2022b). The ESD and MRD were both rated "high" during EPA's systematic
review process.

For the use of antifreeze, EPA expects releases may occur from volatilizations of 1,4-dioxane, disposal
or cleaning of empty antifreeze containers, and spent antifreeze. Both the ESD and MRD indicate that
containers of automotive maintenance fluids are typically small and are not rinsed, but rather disposed of
as solid waste ( v << \ 2022b; OEl O . '20). Additionally, the ESD on Chemical Additives used in
Automotive Lubricants indicates that spent lubricants are disposed of via incineration, which EPA
expects is similarly done for spent antifreeze (OECD. 2020). Therefore, based on this information, EPA
does not expect water releases of 1,4-dioxane for this OES.

Surface Cleaner

EPA estimated daily wastewater discharges for facilities within the Surface Cleaner OES using the
SHEDs-HT model, which is described in Section 2.1.1.2. This modeling was completed for one case
study location (Liverpool OH) and only estimates indirect wastewater discharges. EPA does not expect
direct wastewater discharges to surface water from the types of commercial facilities within this OES
(e.g., restaurants, office buildings, other locations with janitorial services).

Dish Soap

EPA estimated daily wastewater discharges for facilities within the Surface Cleaner OES using the
SHEDs-HT model, which is described in Section 2.1.1.2. This modeling was completed for one case
study location (Liverpool OH) and only estimates indirect wastewater discharges. EPA does not expect
direct wastewater discharges to surface water from the types of commercial facilities within this OES
(e.g., restaurants, assisted living facilities, amusement, and recreation facilities).

Page 236 of 484


-------
6588

6589

6590

6591

6592

6593

6594

6595

6596

6597

6598

6599

6600

6601

6602

6603

6604

6605

6606

6607

6608

6609

6610

6611

6612

6613

6614

6615

6616

6617

6618

6619

6620

6621

6622

6623

6624

6625

6626

6627

6628

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Dishwasher Detergent

EPA estimated daily wastewater discharges for facilities within the Surface Cleaner OES using the
SHEDs-HT model, which is described in Section 2.1.1.2. This modeling was completed for one case
study location (Liverpool OH) and only estimates indirect wastewater discharges. EPA does not expect
direct wastewater discharges to surface water from the types of commercial facilities within this OES
(e.g., restaurants, assisted living facilities, amusement, and recreation facilities).

Laundry Detergent

EPA estimated daily wastewater discharges for facilities within the Laundry Detergent OES using the
OECD ESD on Industrial and Institutional Laundries (OECD. ^ ) and Monte Carlo modeling. The
ESD on Industrial and Institutional Laundries was rated "medium" during EPA's systematic review
process. The use of Montel Carlo modeling allows for variation of calculation input parameters such that
a distribution of environmental releases can be calculated, from which EPA can estimate the 50th and
95th percentile releases. An explanation of this modeling approach is included in Appendix E.12.

Paints and Floor Lacquer

EPA estimated daily wastewater discharges for facilities within the Paints and Floor Lacquers OES
using the OECD ESD on Coating Application via Spray-Painting in the Automotive Refinishing
Industry (OECD. 201 la). The ESD was rated "medium" during EPA's systematic review process.

As described in the process description in Appendix F.4.7, 1,4-dioxane was identified by a public
comment as present in automotive refinishing products, thereby allowing EPA to identify the above
ESD as the most applicable GS/ESD available. This ESD indicates that releases are expected from

1.	Releases to incineration or landfill from container cleaning/disposal,

2.	Releases to incineration or landfill from equipment cleaning,

3.	Releases to incineration or landfill from over sprayed coating that is captured by emission
controls, and

4.	Releases to air from over sprayed coating that is not captured by emission controls.

None of these releases are expected to water (OECD. 201 la). Therefore, based on this ESD, EPA does
not expect water releases of 1,4-dioxane for this OES.

Hydraulic Fracturing

EPA estimated daily wastewater discharges for facilities within the Hydraulic Fracturing OES using the
Draft OECD ESD on Hydraulic Fracturing (	2022d) and Monte Carlo modeling. The Draft

ESD on Hydraulic Fracturing was rated "high" during EPA's systematic review process. The use of
Montel Carlo modeling allows for variation of calculation input parameters such that a distribution of
environmental releases can be calculated, from which EPA can estimate the 50th and 95th percentile
releases. An explanation of this modeling approach is included in Appendix E.13.

E.3.3 Water Release Estimates Summary	

A summary of industrial and commercial water releases estimated using the above methods is presented
in Table Apx E-3 below.

Page 237 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

6629 Table Apx E-3. Summary of Daily Industrial and Commercial Water Release Estimates for 1,4-Dioxane

OES

Type of Water
Discharge

Number of
Facilities with
Releases"

Estimated Daily Release Range
across Sites
(kg/site-day)

Estimated

Release
Frequency
Range
(days)''

Overall Data

Quality
Determination

Sources'

Min

Max

Manufacturing

Surface Water

1

1.21

21.4

250

Medium

TRI, DMR

POTW or
Industrial WWT

1

0

6.69

Medium

TRI

Import and
Repackaging

Surface Water

6

0.91™

250

Medium

TRI, DMR

POTW or
Industrial WWT

6

0

0.91

Medium

TRI

Industrial Uses

Surface Water

24

0

24.5

250

Medium

TRI, DMR

POTW or
Industrial WWT

17

0

105

Medium

TRI

Functional
Fluids (Open-
System)

Surface Water

6

0

0.67

247

Medium

TRI, DMR

POTW or
Industrial WWT

1

4.67

70.9

Medium

TRI

Functional
Fluids (Closed-
System)

All

Assessed as a part of Industrial Uses OES

N/A

N/A

Laboratory
Chemical

Surface Water,
POTW, or
Industrial WWT

132

0 (water releases not expected)

250

High

GSd

Film Cement

Surface Water,
POTW, or
Industrial WWT

211

0 (water releases not expected)

250

High

Process
information®

Spray Foam
Application

Surface Water

1,553,559

0 (surface water releases not
expected)

3

Medium

GS^

POTW

1,553,559

0

0.0036

Medium

GS^

Printing Inks
(3D)

Surface Water

1

0.018

0.022

250

Medium

TRI, DMR

POTW or
Industrial WWT

1

0 (no indirect releases per TRI)

Medium

Medium

Page 238 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE
		July 2023		

OES

Type of Water
Discharge

Number of
Facilities with
Releases"

Estimated Daily Release Range
across Sites
(kg/site-day)

Estimated

Release
Frequency
Range
(days)''

Overall Data

Quality
Determination

Sources'

Min

Max

Dry Film
Lubricant

Surface Water,
POTW, or
Industrial WWT

8

0 (water releases not expected)

48

High

Process
information®

Disposal

Surface Water

24

0

31.8

250

Medium

TRI, DMR

POTW or
Industrial WWT

4

0

0.91

Medium

TRI

Textile Dye

POTW

783

1.50E-05

0.001

31 to 295

Medium

ESDg and
Modeling''

Land (unknown
landfill type) or
POTW (unknown
partitioning)

783

2.09E-07

9.72E-05

Medium

ESDg and
Modeling''

Antifreeze

Surface water,
POTW, or
Industrial WWT

84,383

0 (water releases not expected)

250

High

Process

information®

and

Modeling''

Surface Cleaner

POTW

Unknown

0.072 (single value for all sites in
Liverpool OH case study)

250

N/A

SHEDS-HT'

Land (unknown
landfill) or
POTW

Unknown

18"

High

SHEDS-HT,
Process
information®
Modeling''

Dish Soap

POTW

Unknown

0.064"

250

N/A

SHEDS-HT"

Dishwasher
Detergent

POTW

Unknown

0.00144"

250

N/A

SHEDS-HT'

Laundry
Detergent

Fugitive air, stack
air, or POTW

95,533

1.51 E— 10

0.00714

250 to 365

Medium

ESD! and
Modeling''

Page 239 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE
		July 2023		

OES

Type of Water
Discharge

Number of
Facilities with
Releases"

Estimated Daily Release Range
across Sites
(kg/site-day)

Estimated

Release
Frequency
Range
(days)''

Overall Data

Quality
Determination

Sources'

Min

Max

(Institutional) -

Liquid

Detergents

(unknown
partitioning)













Land (unknown
landfill),
incineration, or
POTW (unknown
partitioning)

95,533

4.05E-12

3.95E-05

Medium

ESD! and
Modeling''

Laundry
Detergent
(Institutional) -
Powder
Detergents

Fugitive air, stack
air, or POTW
(unknown
partitioning)

95,533

3.05E-08

2.10E-04

250 to 365

Medium

ESD! and
Modeling''

Land (unknown
landfill),
incineration, or
POTW (unknown
partitioning)

95,533

5.36E-08

0.0018

Medium

ESD! and
Modeling''

Laundry
Detergent
(Industrial) -
liquid detergents

Fugitive air, stack
air, or POTW
(unknown
partitioning)

2,453

5.48E-12

0.011

20 to 365

Medium

ESD! and
Modeling''

Land (unknown
landfill),
incineration, or
POTW (unknown
partitioning)

2,453

4.78E-12

1.46E-04

Medium

ESD! and
Modeling''

Page 240 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE
		July 2023		

OES

Type of Water
Discharge

Number of
Facilities with
Releases"

Estimated Daily Release Range
across Sites
(kg/site-day)

Estimated

Release
Frequency
Range
(days)''

Overall Data

Quality
Determination

Sources'

Min

Max

Laundry
Detergent
(Industrial) -
powder
detergents

Fugitive air, stack
air, or POTW
(unknown
partitioning)

2,453

1.76E-09

0.0112

20 to 365

Medium

ESD! and
Modeling''

Land (unknown
landfill),
incineration, or
POTW (unknown
partitioning)

2,453

2.92E-11

3.92E-04

Medium

ESD1 and
Modeling''

Paints and Floor
Lacquer

Surface water,
POTW, or
Industrial WWT

33,648

0 (water releases not expected)

250

Medium

ESD> and

process

information®

PET Byproduct

Surface water

19

0

10,050

250

Medium

TRI, DMR

POTW or
Industrial WWT

14

0

682

Medium

TRI

Ethoxylation

Process

Byproduct

Surface water

7

0

0.25

250

Medium

TRI, DMR

POTW or
Industrial WWT

6

0

448

Medium

TRI

Hydraulic
Fracturing

Surface water,
incineration, or
landfill (unknown
partitioning)

411

3.61E-10

4.59

1 to 72

Medium

ESD' and
Modeling''

Recycle/Reuse
(48%),
underground
injection (43%),
Surface water
(6%), or land
(3%)

411

1.85E-10

1.12

Medium

ESD'' and
Modeling''

"Where available. EPA used 2013-2019 TRI (U.S. EPA, 2022g) and 2013-2019 DMR ( 2022c) data to provide a basis to estimate the
number of sites using 1,4-dioxane within a COU.

Page 241 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE
	July 2023	

OES

Type of Water
Discharge

Number of
Facilities with
Releases"

Estimated Daily Release Range
across Sites
(kg/site-day)

Min

Max

Estimated

Release
Frequency
Range
(days)''

Overall Data

Quality
Determination

Sources'

Where available. EPA used the December 2020 Final Risk Evaluation for 1.4-Dioxane (U.S. EPA. 20200). generic scenarios, and emission scenario
documents to provide a basis to estimate the number of release days of 1,4-dioxane within a COU.
c Narrative descriptions of all release estimate sources are provided in Appendix E.3.2.
d The generic scenario used for this COU is the GS on Use of Laboratory Chemicals (U.S. EPA. 202211).

'' For this COU, EPA used process information, which is further described in Appendix E.3.2.

' The generic scenario used for this COU is the GS on Application of Spray PoK urethane Foam Insulation (U.S. EPA. 2018b).
g The emission scenario document used for this COU is the ESD on Textile Dyes (OECD. 2017).
h For this COU, EPA used various models and literature for model input parameters as described in Appendix E.3.2.

1 The emission scenario document used for this COU is the ESD on Industrial and Institutional Laundries (OECD. 201 lb).

1 The emission scenario document used for this COU is the ESD on Coating Application via Spray Painting in the Automotive Refinishing Industry

(OECD. 2011a).

k The emission scenario document used for this COU is the Draft ESD on Hydraulic Fracturing (U.S. EPA. 2022d).

' This value is the Commercial Upstream POTW releases estimated from the SHEDS-HT Down the Drain Model for the Liverpool OH case study (see
Section 2.1.1.2).

m All sites for this OES reported under Form A in TRI.

" A single annual value was provided for all sites in the Liverpool OH case study.

6630

Page 242 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

6631	E.3.4 Summary of Weight of the Scientific Evidence Conclusions in Water Release Estimates	

6632	TableApx E-4 provides a summary of EPA's weight of the scientific evidence conclusions in its water release estimates for each of the OES.

6633	Detailed descriptions of non-OES specific strengths, limitations, assumptions, and uncertainties (e.g., general limitations for TRI, DMR, etc.)

6634	are provided in Appendix E.6.

6635

6636	Table Apx E-4. Summary of Weight of the Scientific Evidence Conclusions in Water Release Estimates by OES	

OES

Weight of the Scientific Evidence Conclusion in Release Estimates

Manufacturing

Wastewater discharges are assessed using reported discharges from 2013-2019 TRI and DMR. Factors that increase the
strength of evidence for this OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI
and DMR have medium overall data quality determinations, and consistency within the dataset (all reporters use the same or
similar reporting forms). EPA included 7 years of TRI and DMR data in the analysis, which increases the variability of the
dataset. A strength of TRI data is that TRI compiles the best readily available release data for all reporting facilities.

Strengths of DMR data are that it is based on monitoring data collected by facilities and the annual pollutant load is
calculated by integrating release reports over shorter timeframes (e.g., monthly, quarterly) and extrapolating over the course
of the year. Factors that decrease the strength of the evidence for this OES include the low number of data points, uncertainty
in the accuracy of reported releases, and the limitations in representativeness to all sites because TRI may not capture all
relevant sites. Additionally, EPA made assumptions on the number of operating days to estimate daily releases, which
introduces additional uncertainty. Based on this information, EPA has concluded that the weight of the scientific evidence
for this assessment is moderate to robust and provides a plausible estimate of releases in consideration of the strengths and
limitations of reasonably available data.

Import and Repackaging

Wastewater discharges are assessed using reported discharges from 2013-2019 TRI and DMR. Factors that increase the
strength of evidence for this OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI
and DMR have medium overall data quality determinations, and consistency within the dataset (all reporters use the same or
similar reporting forms). EPA included seven years of TRI and DMR data in the analysis, which increases the variability of
the dataset. A strength of TRI data is that TRI compiles the best readily available release data for all reporting facilities.
Strengths of DMR data are that it is based on monitoring data collected by facilities and the annual pollutant load is
calculated by integrating release reports over shorter timeframes (e.g., monthly, quarterly) and extrapolating over the course
of the year. Factors that decrease the strength of the evidence for this OES include the low number of data points, uncertainty
in the accuracy of reported releases, uncertainty in EPA's use of Form A submissions, and the limitations in
representativeness to all sites because TRI may not capture all relevant sites. Some facilities within this OES reported to TRI
using a Form A, which does not include any details on chemical release quantities. When a facility has submitted a Form A,
there is no way to discern the quantity released. Therefore, where facilities reported to TRI with a Form A, EPA used the
Form A threshold for total releases of 500 lb/year for each release media; however, there is uncertainty in this because the
actual release quantity is unknown. Furthermore, the threshold represents an upper limit on total releases from the facility;
therefore, assessing releases at the threshold value may overestimate actual releases from the facility. Additionally,
uncertainty is introduced from EPA's assumptions on the number of operating days to estimate daily releases and in the

Page 243 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates



mapping of DMR-reporting facilities to this OES. Based on this information, EPA has concluded that the weight of the
scientific evidence for this assessment is moderate to robust and provides a plausible estimate of releases in consideration of
the strengths and limitations of reasonably available data.

Industrial Uses

Wastewater discharges are assessed using reported discharges from 2013-2019 TRI and DMR. Factors that increase the
strength of evidence for this OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI
and DMR have medium overall data quality determinations, and consistency within the dataset (all reporters use the same or
similar reporting forms). EPA included seven years of TRI and DMR data in the analysis, which increases the variability of
the dataset. A strength of TRI data is that TRI compiles the best readily available release data for all reporting facilities.
Strengths of DMR data are that it is based on monitoring data collected by facilities and the annual pollutant load is
calculated by integrating release reports over shorter timeframes (e.g., monthly, quarterly) and extrapolating over the course
of the year. Factors that decrease the strength of the evidence for this OES include uncertainty in the accuracy of reported
releases, uncertainty in EPA's use of Form A submissions, and the limitations in representativeness to all sites because TRI
may not capture all relevant sites. Some facilities within this OES reported to TRI using a Form A, which does not include
any details on chemical release quantities. When a facility has submitted a Form A, there is no way to discern the quantity
released. Therefore, where facilities reported to TRI with a Form A, EPA used the Form A threshold for total releases of 500
lb/year for each release media; however, there is uncertainty in this because the actual release quantity is unknown.
Furthermore, the threshold represents an upper limit on total releases from the facility; therefore, assessing releases at the
threshold value may overestimate actual releases from the facility. Additionally, uncertainty is introduced from EPA's
assumptions on the number of operating days to estimate daily releases and in the mapping of DMR-reporting facilities to
this OES. Based on this information, EPA has concluded that the weight of the scientific evidence for this assessment is
moderate to robust and provides a plausible estimate of releases in consideration of the strengths and limitations of
reasonably available data.

Functional Fluids (Open-
System)

Wastewater discharges are assessed using reported discharges from 2013-2019 TRI and DMR. Factors that increase the
strength of evidence for this OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI
and DMR have medium overall data quality determinations, and consistency within the dataset (all reporters use the same or
similar reporting forms). EPA included seven years of TRI and DMR data in the analysis, which increases the variability of
the dataset. A strength of TRI data is that TRI compiles the best readily available release data for all reporting facilities.
Strengths of DMR data are that it is based on monitoring data collected by facilities and the annual pollutant load is
calculated by integrating release reports over shorter timeframes (e.g., monthly, quarterly) and extrapolating over the course
of the year. Factors that decrease the strength of the evidence for this OES include the low number of data points, uncertainty
in the accuracy of reported releases, and the limitations in representativeness to all sites because TRI may not capture all
relevant sites. The assessment includes data from only two sites that reported to TRI (one of which reported zero water
releases) and four that reported to DMR. Additionally, uncertainty is introduced from EPA's assumptions on the number of
operating days to estimate daily releases and in the mapping of DMR-reporting facilities to this OES. Based on this
information, EPA has concluded that the weight of the scientific evidence for this assessment is moderate to robust and
provides a plausible estimate of releases in consideration of the strengths and limitations of reasonably available data.

Page 244 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates

Functional Fluids (Closed-
System)

No data was available to estimate releases for this OES. For the water release assessment, EPA grouped this OES with the
Industrial Uses OES because the sources of release are expected to be similar between these OESs. Factors that increase the
strength of evidence for this OES are that TRI and DMR have medium overall data quality determinations and consistency
within the dataset (all reporters use the same or similar reporting forms). Additionally, EPA included seven years of TRI and
DMR data in the analysis, which increases the variability of the dataset. Factors that decrease the strength of evidence for
this OES are that the Industrial Releases OES release data are use as surrogate for this OES, uncertainty in the accuracy of
reported releases, and the limitations in representativeness to all sites because TRI may not capture all relevant sites.

Refer to the Industrial Uses OES discussion for additional discussion. Based on this information, EPA has concluded that the
weight of the scientific evidence for this assessment is slight and provides a plausible estimate of releases in consideration of
the strengths and limitations of reasonably available data.

Laboratory Chemicals

Wastewater discharges are assessed using the Draft GS on Use of Laboratory Chemicals. Per the GS, water releases are not
expected for hazardous chemicals. Because 1,4-dioxane is considered a hazardous chemical under CERCLA, no water
releases are expected for this OES according to the GS. Factors that increase the strength of evidence for this OES are that
the release estimates are directly relevant to the OES (as opposed to surrogate), the Draft GS on Use of Laboratory
Chemicals has a high overall data quality determination, and there is a low level of uncertainty in the data. Factors that
decrease the strength of the evidence for this OES include the that the GS has not been peer-reviewed, uncertainty in the
representativeness of the GS towards all sites in this OES, and a lack of variability in the analysis. Specifically, because the
default values in the ESD are generic, there is uncertainty in the representativeness of generic site estimates of actual releases
from real-world sites that use 1,4-dioxane. Another uncertainty is lack of consideration for release controls. The ESD
assumes that all activities occur without any release controls. Actual releases may be less than estimated if facilities utilize
pollution control methods, contributing to uncertainty. Based on this information, EPA has concluded that the weight of the
scientific evidence for this assessment is slight to moderate and provides a plausible estimate of releases in consideration of
the strengths and limitations of reasonably available data.

Film Cement

Wastewater discharges are assessed using process information from the Final Risk Evaluation for 1,4-Dioxane. Per the
process information, EPA does not expect water releases of 1,4-dioxane for this OES because 1,4-dioxane volatilizes to air
after application of the film cement and empty cement bottles are disposed of as solid waste without rinsing. Factors that
increase the strength of evidence for this OES are that the release estimates are directly relevant to the OES (as opposed to
surrogate), the underlying data sources for the process information have a high overall data quality determination, and there
is a low level of uncertainty in the data because the process information comes directly from actual users of 1,4-dioxane in
film cement. Factors that decrease the strength of the evidence for this OES include the uncertainty in the representativeness
of evidence to all sites in this OES and a lack of variability. Specifically, the process information for the production and use
of film cement is based on information from three use sites, one from Australia and two from the U.S. Based on this
information, EPA has concluded that the weight of the scientific evidence for this assessment is slight to moderate and
provides a plausible estimate of releases in consideration of the strengths and limitations of reasonably available data.

Page 245 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates

Spray Foam Application

Wastewater discharges are assessed using the GS on Application of Spray Polyurethane Foam Insulation. Factors that
increase the strength of evidence for this OES are that the release estimates are directly relevant to the OES (as opposed to
surrogate), the underlying data sources for the process information have a medium overall data quality determination, and
there is a low level of uncertainty in the data. Factors that decrease the strength of the evidence for this OES include
uncertainty in the representativeness of the GS to all sites since it is generic and not specific to sites that use 1,4-dioxane and
a lack of variability. Based on this information, EPA has concluded that the weight of the scientific evidence for this
assessment is slight to moderate and provides a plausible estimate of releases in consideration of the strengths and limitations
of reasonably available data.

Printing Inks (3D)

Wastewater discharges are assessed using reported discharges from 2013-2019 DMR. Factors that increase the strength of
evidence for this OES are that the release data are directly relevant to the OES (as opposed to surrogate), that DMR has a
medium overall data quality determination, and consistency within the dataset (all reporters use the same or similar reporting
forms). Additionally, EPA used DMR data for seven years, which increases the variability of the dataset. Strengths of DMR
data are that it is based on monitoring data collected by facilities and the annual pollutant load is calculated by integrating
release reports over shorter timeframes (e.g., monthly, quarterly) and extrapolating over the course of the year. Factors that
decrease the strength of the evidence for this OES include the low number of data points, uncertainty in the accuracy of
reported releases, and the limitations in representativeness to all sites. Additionally, no TRI data is available for this OES,
EPA made assumptions on the number of operating days, and there is uncertainty in the mapping of DMR-reporting facilities
to this OES. Based on this information, EPA has concluded that the weight of the scientific evidence for this assessment is
moderate to robust and provides a plausible estimate of releases in consideration of the strengths and limitations of
reasonably available data.

Dry Film Lubricant

Wastewater discharges are assessed using process information from the Final Risk Evaluation for 1,4-Dioxane. Based on
conversations the with only known user who supplied this process information, EPA expects wastes to be drummed and sent
to a waste handler with residual wastes releasing to air or being disposed to landfill, such that there are no water releases.
Factors that increase the strength of evidence for this OES are that the release estimates are directly relevant to the OES (as
opposed to surrogate), the underlying data sources for the process information have a high overall data quality determination,
and there is a low level of uncertainty in the data. Additionally, the process information comes directly from an actual user of
1,4-dioxane in dry film lubricants. Factors that decrease the strength of the evidence for this OES include uncertainty in the
representativeness of evidence to all sites and a lack of variability. Based on this information, EPA has concluded that the
weight of the scientific evidence for this assessment is slight to moderate and provides a plausible estimate of releases in
consideration of the strengths and limitations of reasonably available data.

Disposal

Wastewater discharges are assessed using reported discharges from 2013-2019 TRI and DMR. Factors that increase the
strength of evidence for this OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI
and DMR have medium overall data quality determinations, and consistency within the dataset (all reporters use the same or
similar reporting forms). EPA included seven years of TRI and DMR data in the analysis, which increases the variability of
the dataset. A strength of TRI data is that TRI compiles the best readily available release data for all reporting facilities.
Strengths of DMR data are that it is based on monitoring data collected by facilities and the annual pollutant load is

Page 246 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates



calculated by integrating release reports over shorter timeframes (e.g., monthly, quarterly) and extrapolating over the course
of the year. Factors that decrease the strength of the evidence for this OES include uncertainty in the accuracy of reported
releases, uncertainty in EPA's use of Form A submissions, and the limitations in representativeness to all sites because TRI
may not capture all relevant sites. Some facilities within this OES reported to TRI using a Form A, which does not include
any details on chemical release quantities. When a facility has submitted a Form A, there is no way to discern the quantity
released. Therefore, where facilities reported to TRI with a Form A, EPA used the Form A threshold for total releases of 500
lb/year for each release media; however, there is uncertainty in this because the actual release quantity is unknown.
Furthermore, the threshold represents an upper limit on total releases from the facility; therefore, assessing releases at the
threshold value may overestimate actual releases from the facility. Additionally, uncertainty is introduced from EPA's
assumptions on the number of operating days to estimate daily releases and in the mapping of DMR-reporting facilities to
this OES. Based on this information, EPA has concluded that the weight of the scientific evidence for this assessment is
moderate to robust and provides a plausible estimate of releases in consideration of the strengths and limitations of
reasonably available data.

Textile Dye

Wastewater discharges are assessed using Monte Carlo modeling with information from the ESD on Textile Dyes. Factors
that increase the strength of evidence for this OES are that the release estimates are directly relevant to the OES (as opposed
to surrogate), that the ESD on Textile Dyes has a medium overall data quality determination and was peer reviewed, the high
number of data points (simulation runs), consistency within the dataset, and full distributions of input parameters. The Monte
Carlo modeling accounts for the entire distribution of input parameters, calculating a distribution of potential release values
that represents a larger proportion of sites than a discrete value. Factors that decrease the strength of the evidence for this
OES include uncertainties and limitations in the representativeness of the estimates for sites that specifically use 1,4-dioxane
because the default values in the ESD are generic. Another uncertainty is lack of consideration for release controls. The ESD
assumes that all activities occur without any release controls. Actual releases may be less than estimated if facilities utilize
pollution control methods, contributing to uncertainty. Based on this information, EPA has concluded that the weight of the
scientific evidence for this assessment is moderate and provides a plausible estimate of releases in consideration of the
strengths and limitations of reasonably available data.

Antifreeze

Wastewater discharges are assessed using the OECD ESD on Chemical Additives used in Automotive Lubricants and the
EPA MRD on Commercial Use of Automotive Detailing Products. Factors that increase the strength of evidence for this
OES are that the ESD and MRD used have high overall data quality determinations, consistency within the sources used, and
a low number of uncertainties. Both sources indicate that containers of automotive maintenance fluids are not typically
rinsed, but rather disposed of as solid waste or incinerated, such that there are no water releases, contributing to consistency
and a low level of uncertainty in the data. Factors that decrease the strength of the evidence for this OES include that the
ESD and MRD are not directly applicable to antifreeze uses (used as surrogate), uncertainty in the representativeness of the
ESD and MRD to all sites and sites that specifically use 1,4-dioxane since these documents contain generic values, and a
lack of variability. Based on this information, EPA has concluded that the weight of the scientific evidence for this
assessment is slight to moderate and provides a plausible estimate of releases in consideration of the strengths and limitations
of reasonably available data.

Page 247 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates

Surface Cleaner

Wastewater discharges are assessed using the SHEDS-HT model. Factors that increase the strength of evidence for this OES
include that the release estimates are directly relevant to the OES (as opposed to surrogate) and variability in the model input
parameters. Factors that decrease the strength of the evidence for this OES include uncertainty in the representativeness to all
sites because the estimate is based on one case study for Liverpool, OH and because the estimate is not site-specific (the
release estimate is a total for all sites in Liverpool, OH). Based on this information, EPA has concluded that the weight of the
scientific evidence for this assessment is slight and provides a plausible estimate of releases in consideration of the strengths
and limitations of reasonably available data.

Dish Soap

Wastewater discharges are assessed using the SHEDS-HT model. Factors that increase the strength of evidence for this OES
include that the release estimates are directly relevant to the OES (as opposed to surrogate) and variability in the model input
parameters. Factors that decrease the strength of the evidence for this OES include uncertainty in the representativeness to all
sites because the estimate is based on one case study for Liverpool, OH and because the estimate is not site-specific (the
release estimate is a total for all sites in Liverpool, OH). Based on this information, EPA has concluded that the weight of the
scientific evidence for this assessment is slight and provides a plausible estimate of releases in consideration of the strengths
and limitations of reasonably available data.

Dishwasher Detergent

Wastewater discharges are assessed using the SHEDS-HT model. Factors that increase the strength of evidence for this OES
include that the release estimates are directly relevant to the OES (as opposed to surrogate) and variability in the model input
parameters. Factors that decrease the strength of the evidence for this OES include uncertainty in the representativeness to all
sites because the estimate is based on one case study for Liverpool, OH and because the estimate is not site-specific (the
release estimate is a total for all sites in Liverpool, OH). Based on this information, EPA has concluded that the weight of the
scientific evidence for this assessment is slight and provides a plausible estimate of releases in consideration of the strengths
and limitations of reasonably available data.

Laundry Detergent

Wastewater discharges are assessed using Monte Carlo modeling with information from the ESD on Industrial and
Institutional Laundries. Factors that increase the strength of evidence for this OES are that the release estimates are directly
relevant to the OES (as opposed to surrogate), that the ESD on Industrial and Institutional Laundries has a medium overall
data quality determination and was peer reviewed, there are a high number of data points (simulation runs), consistency
within the dataset, and full distributions of input parameters. Monte Carlo modeling accounts for the entire distribution of
input parameters, calculating a distribution of potential release values that represents a larger proportion of sites than a
discrete value. Additionally, EPA was able to separately estimate releases for industrial and institutional laundry settings.
Factors that decrease the strength of the evidence for this OES include uncertainties and limitations in the representativeness
of the estimates for sites that specifically use 1,4-dioxane because the default values in the ESD are generic. Another
uncertainty is lack of consideration for release controls. The ESD assumes that all activities occur without any release
controls. Actual releases may be less than estimated if facilities utilize pollution control methods. Based on this information,
EPA has concluded that the weight of the scientific evidence for this assessment is moderate and provides a plausible
estimate of releases in consideration of the strengths and limitations of reasonably available data.

Page 248 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates

Paint and Floor Lacquer

Wastewater discharges are assessed using OECD ESD on Coating Application via Spray-Painting in the Automotive
Refinishing Industry. According to the ESD, no releases are expected to water. Factors that increase the strength of evidence
for this OES are that the release estimates are directly relevant to the OES (as opposed to surrogate), the ESD has a medium
overall data quality determination, and a low number of uncertainties. F Factors that decrease the strength of the evidence for
this OES include a lack of variability and uncertainty in the representativeness of the ESD to all sites and sites that
specifically use 1,4-dioxane since the ESD is generic. Based on this information, EPA has concluded that the weight of the
scientific evidence for this assessment is slight to moderate and provides a plausible estimate of releases in consideration of
the strengths and limitations of reasonably available data.

Polyethylene
Terephthalate (PET)
Byproduct

Wastewater discharges are assessed using reported discharges from 2013-2019 TRI and DMR. Factors that increase the
strength of evidence for this OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI
and DMR have medium overall data quality determinations, consistency within the dataset (all reporters use the same or
similar reporting forms), and consistency with the emission data from the related life cycle analysis discussed in Appendix
E.6. EPA included seven years of TRI and DMR data in the analysis, which increases the variability of the dataset. A
strength of TRI data is that TRI compiles the best readily available release data for all reporting facilities. Strengths of DMR
data are that it is based on monitoring data collected by facilities and the annual pollutant load is calculated by integrating
release reports over shorter timeframes (e.g., monthly, quarterly) and extrapolating over the course of the year. Factors that
decrease the strength of the evidence for this OES include the uncertainty in the accuracy of reported releases and the
limitations in representativeness to all sites because TRI may not capture all relevant sites. Additionally, EPA made
assumptions on the number of operating days to estimate daily releases, which introduces additional uncertainty. Based on
this information, EPA has concluded that the weight of the scientific evidence for this assessment is moderate to robust and
provides a plausible estimate of releases in consideration of the strengths and limitations of reasonably available data.

Ethoxylation Process
Byproduct

Wastewater discharges are assessed using reported discharges from 2013-2019 TRI and DMR. Factors that increase the
strength of evidence for this OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI
and DMR have medium overall data quality determinations, and consistency within the dataset (all reporters use the same or
similar reporting forms). EPA included seven years of TRI and DMR data in the analysis, which increases the variability of
the dataset. A strength of TRI data is that TRI compiles the best readily available release data for all reporting facilities.
Strengths of DMR data are that it is based on monitoring data collected by facilities and the annual pollutant load is
calculated by integrating release reports over shorter timeframes (e.g., monthly, quarterly) and extrapolating over the course
of the year. Factors that decrease the strength of the evidence for this OES include uncertainty in the accuracy of reported
releases and the limitations in representativeness to all sites because TRI may not capture all relevant sites. Additionally,
EPA made assumptions on the number of operating days to estimate daily releases, which introduces additional uncertainty.
Based on this information, EPA has concluded that the weight of the scientific evidence for this assessment is moderate to
robust and provides a plausible estimate of releases in consideration of the strengths and limitations of reasonably available
data.

Hydraulic Fracturing

Wastewater discharges are assessed using Monte Carlo modeling with information from the Draft ESD on Hydraulic
Fracturing and FracFocus 3.0. Factors that increase the strength of evidence for this OES are that the release estimates are

Page 249 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates



directly relevant to the OES (as opposed to surrogate), that the Draft ESD on Hydraulic Fracturing and FracFocus 3.0 have
medium overall data quality determinations, the high number of data points (simulation runs), and full distributions of input
parameters. Monte Carlo modeling accounts for the entire distribution of input parameters, calculating a distribution of
potential release values that represents a larger proportion of sites than a discrete value. Factors that decrease the strength of
the evidence for this OES include that the Draft ESD has not been peer reviewed, uncertainties that may result in over-
estimates of releases, and limitations in the representativeness of the estimates for all sites. Specifically, EPA used some
input values from the Draft ESD; because the default values in the ESD are generic, there is uncertainty in the
representativeness of the generic site estimates of real-world sites that use 1,4-dioxane. Another uncertainty is lack of
consideration for release controls. The ESD assumes that all activities occur without any release controls. Actual releases
may be less than estimated if facilities utilize pollution control methods, contributing to uncertainty. Based on this
information, EPA has concluded that the weight of the scientific evidence for this assessment is moderate to robust and
provides a plausible estimate of releases in consideration of the strengths and limitations of reasonably available data.

6637

Page 250 of 484


-------
6638

6639

6640

6641

6642

6643

6644

6645

6646

6647

6648

6649

6650

6651

6652

6653

6654

6655

6656

6657

6658

6659

6660

6661

6662

6663

6664

6665

6666

6667

6668

6669

6670

6671

6672

6673

6674

6675

6676

6677

6678

6679

6680

6681

6682

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

E.4 Land Release Assessment

This section describes EPA's methodology for estimating annual land releases from industrial and
commercial facilities manufacturing, processing, or using 1,4-dioxane. EPA did not estimate daily land
releases due to the high level of uncertainty in the number of release days associated with land releases.
Facilities report annual land releases to the Toxics Release Inventory (TRI), which include a variety of
release mechanisms, including but not limited to underground injection, RCRA Subtitle C landfills,
other landfills, surface impoundments, and land treatment. EPA used 2019 TRI (	2022e) data

to estimate annual land releases for the OES where available; however, EPA did not have these data for
every OES. For OES without TRI data, EPA used alternate assessment approaches to estimate annual
land releases.

In addition, EPA did a more in-depth analysis of TRI for sites within the disposal OES. Specifically,
EPA did an analysis of 2013 to 2019 TRI data for this OES. Operations at disposal sites are expected to
be more complex than those at sites in other OES, which typically generate waste for land disposal off
site. Additionally, the disposal OES includes the sites of ultimate disposal {i.e., they are the landfills
themselves) and EPA considered the impact of continuous years of land releases of 1,4-dioxane at these
sites on general population and ecological exposures.

E.4.1 Assessment Using TRI

EPA found 2019 TRI data for facilities within the following OESs:

•	Manufacturing

•	Import and Repackaging

•	Industrial Uses

•	Functional Fluids (Open-System)

•	Disposal

•	PET Byproduct

•	Ethoxylation Byproduct

The TRI data were rated "medium" in EPA's systematic review process. EPA estimated annual land
releases using TRI for these OESs, with the following general stages as described in the rest of this
section.

1.	Collect land release data from the 2013 to 2019 TRI for the Disposal OES and 2019 TRI data for
all other OES,

2.	Map land release data to occupational exposure scenarios,

3.	Analyze 2013 to 2019 TRI data for the disposal OES, and

4.	Summarize 2019 annual land releases for the other OES.

Step 1: Collect Land Release Data from TRI

The first step in estimating land releases was to obtain TRI data. As previously discussed in Appendix
E.3.1, each facility subject to the TRI reporting rule must report annually the volume of chemical
released to the environment and/or managed through recycling, energy recovery, and treatment. Similar
to the air release assessment, EPA included both TRI reporting Form R and TRI reporting Form A
submissions in the land release assessment. Where sites reported to TRI with Form A, EPA used the
Form A threshold for total releases of 500 lb/year. EPA used the entire 500 lb/year for each type of land
release; however, since this threshold is for total site releases, these 500 lb/year are attributed one type
of land release at a time (since assessing it for more than one land release media at a time would double
count the releases and exceed the total release threshold for Form A). EPA pulled the TRI Basic Plus
Data Files for each of years 2013 through 2019.

Page 251 of 484


-------
6683

6684

6685

6686

6687

6688

6689

6690

6691

6692

6693

6694

6695

6696

6697

6698

6699

6700

6701

6702

6703

6704

6705

6706

6707

6708

6709

6710

6711

6712

6713

6714

6715

6716

6717

6718

6719

6720

6721

6722

6723

6724

6725

6726

6727

6728

6729

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI data include both on- and off-site land releases. In summary, TRI includes the following land
release media:

•	On-site releases:

o	Underground injection

o	RCRA subtitle C landfills

o	Other landfills

o	Land treatment

o	RCRA surface impoundments

o	Other surface impoundments

o	Other land disposal

o	Waste rock

•	Off-site releases:

o	Underground injection

o	RCRA subtitle C landfills

o	Other landfills

o	Land treatment

o	RCRA surface impoundments

o	Other surface impoundments

o	Other land disposal

o	Transfer to waste broker for disposal

o	Solidification/stabilization

Step 2: Map Land Release Data to Occupational Exposure Scenarios

The next step in estimating land releases was to map the 2013 to 2019 TRI data to the 1,4-dioxane OES.
EPA used the same mapping methodology as that used for both the air and water assessments, which is
described in Appendix E.5.1. EPA ensured consistency in the OES mapping between the air, water, and
land assessments.

Step 3: Analyze and Summarize 2013 to 2019 TRI Data for the Disposal OES

For the sites that EPA mapped to the disposal OES in the 2013 to 2019 TRI data, EPA analyzed and
summarized the land release data as follows:

•	EPA summarized which of the reporting years that each disposal facility submitted data to TRI.
This summary allows for visualization of which sites report recurring land disposal of 1,4-
dioxane between 2013 and 2019.

•	EPA differentiated between disposal sites that transferred 1,4-dioxane to other sites for disposal
and the receiving sites that disposed of 1,4-dioxane on site. For the receiving sites at which 1,4-
dioxane was disposed of to land, EPA summarized the number of unique sites from which the
receiving sites received 1,4-dioxane for land disposal and the total amount of 1,4-dioxane
received for land disposal between 2013 and 2019.

•	EPA summarized the total amount of 1,4-dioxane released to each land release media between
2013 and 2019. In summary, 1,4-dioxane was disposed of from disposal OES sites via on-site
and off-site RCRA subtitle C landfills, on-site and off-site underground injection, and off-site
other landfills between 2013 and 2019.

EPA's analysis and summary of land releases for 2013 to 2019 TRI sites in the disposal OES can be
found in 1,4-Dioxane Supplemental Information File: Environmental Releases to landfor the Disposal
OES (	231).

Page 252 of 484


-------
6730

6731

6732

6733

6734

6735

6736

6737

6738

6739

6740

6741

6742

6743

6744

6745

6746

6747

6748

6749

6750

6751

6752

6753

6754

6755

6756

6757

6758

6759

6760

6761

6762

6763

6764

6765

6766

6767

6768

6769

6770

6771

6772

6773

6774

6775

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Step 4: Summarize Annual Land Releases for Other OES with 2019 TRI data

For the remaining OES for which 2019 TRI data were available, EPA summarized the annual land
releases by media type (e.g., underground injection, RCRA subtitle C landfills, other landfills, land
treatment) and site information, including site identity, city, state, zip code, TRI facility ID, and FRS ID.
EPA did not estimate daily land releases due to the high level of uncertainty in the number of release
days associated with land releases.

EPA's summary of land release for these OESs is included in 1,4-Dioxane Supplemental Information
File: Environmental Releases to Landfor all OES Except Disposal (	023k).

E.4.2 Assessment for OES without TRI

EPA did not find 2019 TRI data for the following OES:

•	Functional Fluids (Closed-Systems)

•	Laboratory Chemicals

•	Film Cement

•	Spray Polyurethane Foam

•	3D Printing

•	Dry Film Lubricant

•	Textile Dye

•	Antifreeze

•	Surface Cleaner

•	Dish Soap

•	Dishwasher Detergent

•	Laundry Detergent

•	Paints and Floor Lacquer

•	Hydraulic Fracturing

For these OESs, EPA estimated land releases by using various modeling approaches, including the use
of surrogate TRI data and modeling using data from literature, GSs, and ESDs. EPA's assessment of
land releases for each of these OESs is described below.

Functional Fluids (Closed-Systems)

Land release data were not available for this OES and EPA did not find any information to model land
release for this OES using literature, GSs, or ESDs. EPA expects that the sources of release for this OES
to be similar to those for the Industrial Uses OES, based on the process information in the Final Risk
Evaluation for 1,4-Dioxane (U, S. EPA. 2020c). Therefore, EPA grouped the land release assessment for
Functional Fluids (Closed-Systems) into that for Industrial Uses. However, there is uncertainty in this
assumption of similar release sources between these OESs.

Laboratory Chemicals

EPA estimated land releases for facilities within the Laboratory Chemicals OES using the Draft GS on
Use of Laboratory Chemicals (	22h). The GS on Use of Laboratory Chemicals was rated

"high" during EPA's systematic review process.

The GS indicates that there are eight release points:

1.	Release to air from transferring volatile chemicals from transport containers.

2.	Release to air, water, incineration, or landfill from transferring solid powders.

3.	Release to water, incineration, or land from cleaning or disposal of transport containers.

Page 253 of 484


-------
6776

6777

6778

6779

6780

6781

6782

6783

6784

6785

6786

6787

6788

6789

6790

6791

6792

6793

6794

6795

6796

6797

6798

6799

6800

6801

6802

6803

6804

6805

6806

6807

6808

6809

6810

6811

6812

6813

6814

6815

6816

6817

6818

6819

6820

6821

6822

6823

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

4.	Release to air from cleaning containers used for volatile chemicals.

5.	Labware equipment cleaning residuals released to water, incineration, or landfill.

6.	Release to air during labware equipment cleaning for volatile chemicals.

7.	Release to air from laboratory analyses for volatile chemicals.

8.	Release to water, incineration, or landfill from laboratory waste disposal.

Based on the GS, release points #2, 3, 5, and 8 have the potential for land releases; however, release
point #2 is not applicable because 1,4-dioxane is not a solid powder. To estimate the remaining land
releases, EPA used the equations specified in the Draft GS (	|22h). EPA did not find any

data specific to 1,4-dioxane in this OES. Therefore, the calculation of releases using this GS are for a
"generic site," using the default input parameter values from the GS.

Using this methodology, EPA calculated high-end and low-end annual land releases for this OES. The
low- and high-end estimates are based on the low-end or typical and high-end or worst-case calculation
input parameter defaults from the GS. EPA's calculation of land releases for this OES, including all
calculation inputs, can be found in 1,4-Dioxane Supplemental Information File: Environmental Releases
to Landfor all OES Except Disposal (U.S. EPA. 2023k).

Film Cement

EPA estimated land releases for facilities within the Film Cement OES using process information from
the Final Risk Evaluation for 1,4-Dioxane (	2020c). The underlying process information for

this assessment was rated "high" during EPA's systematic review process.

The process of using film cement involves applying the cement onto edges of photographic films by
hand using a small brush, then joining the pieces of film by pressing and heating to dry the cement.
Based on this process information, EPA expects land releases may result from disposal of empty film
cement bottles that contain residual amounts of film cement containing 1,4-dioxane. EPA estimated this
land release as a range, using a film cement use rate of 2.5 to 12 L/site-year and a concentration of 1,4-
dioxane in the film cement of 45 to 50 percent from the process information in the Final Risk Evaluation
for 1,4-Dioxane (U.S. EPA. 2020c). and the EPA/OPPT Small Container Residual Model central
tendency loss fraction of 0.3 percent and high-end loss fraction of 0.6 percent. EPA is uncertain of the
specific type of land disposal for the empty film cement bottles.

EPA's calculation of land releases for this OES, including all calculation inputs and assumptions, can be
found in 1,4-Dioxane Supplemental Information File: Environmental Releases to Landfor all OES
Except Disposal (U.S.. 023k).

Spray Polyurethane Foam

EPA estimated land releases for facilities within the Spray Polyurethane Foam OES using the GS on
Application of Spray Polyurethane Foam Insulation (U.S. EPA. 2020c). The GS on the Application of
Spray Polyurethane Foam Insulation was rated "medium" during EPA's systematic review process.

The GS indicates that there are six release points:

1.	Releases to fugitive air for volatile chemicals during unloading of raw materials from transport
containers;

2.	Releases to water, incineration, or landfill from cleaning or disposal of transport containers;

3.	Releases to fugitive air for volatile chemicals during transport container cleaning;

4.	Releases to incineration or landfill from spray polyurethane foam application equipment
cleaning;

Page 254 of 484


-------
6824

6825

6826

6827

6828

6829

6830

6831

6832

6833

6834

6835

6836

6837

6838

6839

6840

6841

6842

6843

6844

6845

6846

6847

6848

6849

6850

6851

6852

6853

6854

6855

6856

6857

6858

6859

6860

6861

6862

6863

6864

6865

6866

6867

6868

6869

6870

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

5.	Releases to fugitive air for volatile chemicals during equipment cleaning; and

6.	Releases to landfill of scrap foam from trimming applied foam.

Based on the GS, release points #2, 4, and 6 have the potential for land releases. To estimate these
releases, EPA used the equations specified in the GS (	1020c). Apart from weight fraction in

spray polyurethan foam, EPA did not find any data specific to 1,4-dioxane in this OES. Therefore, the
calculation of releases using this GS are for a "generic site," using the default input parameter values
from the GS. Specifically, EPA used the same input parameter values that were used in the original risk
evaluation for estimates of occupational exposure; see Appendix G of the Final Risk Evaluation for 1,4-
Dioxane (U.S. EPA. 2020c).

Using this methodology, EPA calculated high-end and low-end annual land releases for this OES. The
low- and high-end estimates are based on the low-end or typical and high-end or worst-case calculation
input parameter defaults from the GS. EPA's calculation of land releases for this OES, including all
calculation inputs, can be found 1,4-Dioxane Supplemental Information File: Environmental Releases to
Landfor all OES Except Disposal (U.S. EPA. 2023k).

3D Printing

Land release data were not available for this OES and EPA did not find any information to model land
releases for this OES using literature, GSs, or ESDs. EPA expects that industrial applications of this
OES to be accounted for in the Industrial Uses TRI data. Per the December 2020 Final Risk Evaluation
for 1,4-Dioxane (	320c). 3D printing ink containing 1,4-dioxane is used in research labs to

print biomedical products. Because the 2019 TRI data for the Industrial Uses OES include medicinal and
pharmaceutical manufacturing NAICS codes, medical research labs that conduct 3D printing with 1,4-
dioxane inks may be captured in that OES. Therefore, EPA grouped the land release assessment for 3D
Printing into that for Industrial Uses. However, there is uncertainty in whether 3D printing sites are truly
captured in the Industrial Uses TRI data.

Dry Film Lubricant

EPA estimated land releases for facilities within the Dry Film Lubricant OES using process information
from the Final Risk Evaluation for 1,4-Dioxane (	'20c). The underlying process information

for this assessment was rated "high" during EPA's systematic review process.

The process for the production and use of dry film lubricant is described in the 2020 RE and is based on
information provided to EPA by the one known user. In summary, the process entails first producing the
concentrated dry film lubricant by mixing 1,4-dioxane with other additives, followed by the dilution of
the concentrated dry film lubricant with additional 1,4-dioxane and the use of the dry film lubricant. The
use involves spray application onto substrates in a vented paint booth and the subsequent curing in a
vented oven and cleaning of the dried parts in a 1,4-dioxane bath (	10c). Based on this

process description, EPA expects land releases may result from

1.	Residuals in empty containers of pure 1,4-dioxane used for mixing of the concentrated dry film
lubricant,

2.	Cleaning residuals for equipment used for mixing of the concentrated dry film lubricant,

3.	Residuals in empty containers of pure 1,4-dioxane used for diluting the concentrated dry film
lubricant,

4.	Residuals in empty containers of concentrated dry film lubricant, and

5.	Waste from cleaning spray application equipment and the parts onto which the dry film lubricant
was applied.

Page 255 of 484


-------
6871

6872

6873

6874

6875

6876

6877

6878

6879

6880

6881

6882

6883

6884

6885

6886

6887

6888

6889

6890

6891

6892

6893

6894

6895

6896

6897

6898

6899

6900

6901

6902

6903

6904

6905

6906

6907

6908

6909

6910

6911

6912

6913

6914

6915

6916

6917

6918

6919

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

EPA estimated land releases using 1,4-dioxane use rates derived from the process information and
standard EPA models. Specifically, EPA estimated land releases from release points #1, 3 and 4 using
the EPA/OPPT Small Container Residual Model, which has a central tendency loss fraction of 0.3
percent and a high-end loss fraction of 0.6 percent of the container volume. EPA used container volumes
specified in the process information, which are either 1-gallon or 0.5-pint containers (	?20c).

EPA estimated releases from release point #2 using the EPA/OPPT Single Process Vessel Residual
Model, which has a central tendency loss fraction of 0.2 percent and a high-end loss fraction of 1 percent
of the 1,4-dioxane throughput. EPA estimated land releases from release point #5 by assuming the entire
volume of the cleaning bath used for equipment and parts is released to landfill. This is consistent with
the process information, which indicates that spent 1,4-dioxane is disposed of as chemical waste, which
EPA assumes may be to either incineration or RCRA subpart C landfills (U.S. EPA. 2020c).

EPA's calculation of land releases for this OES, including all calculation inputs and assumptions, can be
found in 1,4-Dioxane Supplemental Information File: Environmental Releases to Landfor all OES
Except Disposal ( BP A. 2023k).

Textile Dye

EPA estimated land releases for facilities within the Textile Dye OES using the OECD ESD on Textile
Dyes (OECD. 2017) and Monte Carlo modeling. The ESD on Textile Dyes was rated "medium" during
EPA's systematic review process. The use of Montel Carlo modeling allows for variation of calculation
input parameters such that a distribution of environmental releases can be calculated, from which EPA
can estimate the 50th and 95th percentile releases. An explanation of this modeling approach is included
in Appendix E. 11.

Antifreeze

EPA did not find any directly applicable GS/ESD or literature sources for this OES; however, EPA
evaluated the potential for releases using the OECD ESD on Chemical Additives used in Automotive
Lubricants (<	20) and the EPA MRD on Commercial Use of Automotive Detailing Products

(1 c. i i1 \ AV2b). The ESD and MRD were both rated "high" during EPA's systematic review
process.

For the use of antifreeze, EPA expects releases may occur from volatilizations of 1,4-dioxane during
unloading/ pouring antifreeze into vehicles, disposal, or cleaning of empty antifreeze containers, and
spent antifreeze. Both the ESD and MRD indicate that containers of automotive maintenance fluids are
typically small and are not rinsed, but rather disposed of as solid waste (I v < < \ .022 k * ^ CD.
2020). Additionally, the ESD on Chemical Additives used in Automotive Lubricants indicates that spent
lubricants are disposed of via incineration by blending with fuel oil (OECD. 2020). However, since
spent antifreeze is unlikely to be blended with fuel oil, EPA expects spent antifreeze may be disposed of
via incineration or landfills that take chemical waste. Therefore, EPA expects land releases result from
disposal of empty antifreeze containers and spent antifreeze.

To estimate the use rate of 1,4-dioxane for this OES, EPA used the consumer use rate of antifreeze (0.15
kg antifreeze/job) from the Final Risk Evaluation for 1,4-Dioxane (	'20c) and scaled this

value up to a commercial use rate based on a range of the number of cars serviced at auto shops from the
Near-Field/Far-Field Brake Model and Automotive Detailing MRD (1 to 9 jobs/day). EPA used a range
of concentration of 1,4-dioxane in antifreeze from the process description in Appendix F.4.2 and
assumed antifreeze container sizes ranging from 16 ounces to 5 gallons per the default container sizes in
the MRD and ESD, respectively (I v « « \ 202 J., * i n :020).

Page 256 of 484


-------
6920

6921

6922

6923

6924

6925

6926

6927

6928

6929

6930

6931

6932

6933

6934

6935

6936

6937

6938

6939

6940

6941

6942

6943

6944

6945

6946

6947

6948

6949

6950

6951

6952

6953

6954

6955

6956

6957

6958

6959

6960

6961

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

To estimate the land release from container disposal, EPA used the calculated 1,4-dioxane throughput
based on the above batch parameters and the EPA/OPPT Small Container Residual Model, which has a
central tendency loss fraction of 0.3 percent and a high-end loss fraction of 0.6 percent. To estimate the
land release from spent antifreeze, EPA used the 1,4-dioxane throughput and a mass balance assuming
100 percent release minus upstream losses from container disposal and volatilizations during unloading
(estimated with the EPA/OAQPS AP-42 Loading Model).

EPA's calculation of land releases for this OES, including all calculation inputs and assumptions, can be
found in the supplemental attachment 1,4-Dioxane Supplemental Information File: Environmental
Releases to Landfor all OES Except Disposal (U.S. EPA. 2023k).

Surface Cleaner

EPA did not find any directly applicable GS/ESD or literature sources for this OES; however, EPA
estimated land releases using the SHEDs-HT modeling conducted for the one case study location
(Liverpool OH) and the assumptions described here. EPA expects that the main release points from the
use of surface cleaners are from

1.	Disposal of empty containers containing residual cleaning solution,

2.	Application of the cleaning solution, and

3.	Disposal of cleaning solution by rinsing or wiping.

Because EPA did not find any directly applicable GSs or ESDs, EPA used the Draft GS on Furnishing
Cleaning (	322a) to inform these releases due to the similarities in surface cleaning and

furnishing cleaning. The Draft GS on Furnishing Cleaning was rated "high" during EPA's systematic
review process. Per this Draft GS, empty containers may be rinsed out in sinks or disposed of without
rinsing, such that releases may be to wastewater or landfill; the GS uses the EPA/OPPT Small Container
Residual Model to estimate this release. Application losses are to fugitive air from spray application; the
GS uses literature data to estimate this release. Once applied, the cleaner may be rinsed off or wiped off
with rags or towels, such that releases may be to wastewater or landfill; the GS assumes 100 percent
release scenario, estimating this release by subtracting the upstream losses from the cleaner use rate
(	>022a).

The SHEDs-HT modeling estimated wastewater discharges of 72 g of 1,4-dioxane per day for
commercial uses of surface cleaners containing 1,4-dioxane in Liverpool OH. As described previously,
because both release point #1 and #3 may also be to either wastewater or landfills, EPA assumes the
same quantity of 72 g of 1,4-dioxane per day from the SHEDs-HT model may be released to unknown
landfills for this OES. EPA notes that these 72 g is either entirely to wastewater or landfill or some split
between the two media. The 72 g is not to both wastewater and landfill because that would double count
the release,

EPA's calculation of land releases for this OES, including all calculation inputs and assumptions, can be
found in 1,4-Dioxane Supplemental Information File: Environmental Releases to Landfor all OES
Except Disposal ( BP A. 2023k).

Page 257 of 484


-------
6962

6963

6964

6965

6966

6967

6968

6969

6970

6971

6972

6973

6974

6975

6976

6911

6978

6979

6980

6981

6982

6983

6984

6985

6986

6987

6988

6989

6990

6991

6992

6993

6994

6995

6996

6997

6998

6999

7000

7001

7002

7003

7004

7005

7006

7007

7008

7009

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Dish Soap

EPA did not find any directly applicable GS/ESD or literature sources for this OES; however, EPA
estimated land releases using the SHEDs-HT modeling conducted for the one case study location
(Liverpool OH) and the assumptions described here. EPA expects that the main release points from the
use of dish soaps are from:

1.	Disposal of empty containers containing residual dish soap, and

2.	Cleaning and rinsing of dishes.

EPA expects that empty containers may be rinsed out in sinks or disposed of without rinsing, such that
releases may be to wastewater or landfill. Further, EPA expects that the entire amount of dish soap used
for cleaning dishes is rinsed down the drain of sinks during the cleaning and rinsing process. EPA uses
the SHEDs-HT modeled estimated of wastewater discharges for this OES (64 g 1,4-dioxane per day for
Liverpool OH) and back calculates a 1,4-dioxane use rate using the EPA/OPPT Small Container
Residual Model central tendency and high-end loss fractions and an assumption of 100 percent release.

Using this back-calculated 1,4-dioxane use rate, EPA then applied the EPA/OPPT Small Container
Residual Model to estimate land releases for the Liverpool OH case study. EPA expects that this land
release is to unknown landfills. EPA's calculation of land releases for this OES, including all calculation
inputs and assumptions, can be found in 1,4-Dioxane Supplemental Information File: Environmental
Releases to Landfor all OES Except Disposal (	A. 2023k).

Dishwasher Detergent

EPA did not find any directly applicable GS/ESD or literature sources for this OES; however, EPA
estimated land releases using the SHEDs-HT modeling conducted for the one case study location
(Liverpool OH) and the same assumptions as that described for the dish soap OES above. EPA's
calculation of land releases for this OES, including all calculation inputs and assumptions, can be found
in 1,4-Dioxane Supplemental Information File: Environmental Releases to Landfor all OES Except
Disposal (	)23k).

Laundry Detergent

EPA estimated land releases for facilities within the Laundry Detergent OES using the OECD ESD on
Industrial and Institutional Laundries (OI	) and Monte Carlo modeling. The ESD on

Industrial and Institutional Laundries was rated "medium" during EPA's systematic review process. The
use of Montel Carlo modeling allows for variation of calculation input parameters such that a
distribution of environmental releases can be calculated, from which EPA can estimate the 50th and 95th
percentile releases. An explanation of this modeling approach is included in Appendix E.12.

Paints and Floor Lacquer

EPA estimated land releases for facilities within the Paints and Floor Lacquers OES using the OECD
ESD on Coating Application via Spray-Painting in the Automotive Refinishing Industry (OECD.
201 la). The ESD was rated "medium" during EPA's systematic review process.

As described in the process description in Appendix F.4.7, 1,4-dioxane was identified by a public
comment as present in automotive refinishing products, thereby allowing EPA to identify the above
ESD as the most applicable GS/ESD available. This ESD indicates that releases are expected from

1.	Releases to incineration or landfill from container cleaning/disposal,

2.	Releases to incineration or landfill from equipment cleaning,

3.	Releases to incineration or landfill from over sprayed coating that is captured by emission
controls, and

Page 258 of 484


-------
7010

7011

7012

7013

7014

7015

7016

7017

7018

7019

7020

7021

7022

7023

7024

7025

7026

7027

7028

7029

7030

7031

7032

7033

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

4. Releases to stack air from over sprayed coating that is not captured by emission controls.

Based on the GS, release points #1 through 3 have the potential for land releases. To estimate these
releases, EPA used the equations specified in the ESD (OB	). Apart from weight fraction in

coatings (see Appendix F.4.7), EPA did not find any data specific to 1,4-dioxane in this OES. Therefore,
the calculation of releases using this GS are for a "generic site," using the default input parameter values
from the ESD.

Using this methodology, EPA calculated the low-end and high-end land releases for this OES, which are
expected to be to unknown landfills per the ESD (OECD. 201 la). The low- and high-end estimates are
based on the low- and high-end calculation input parameter defaults from the ESD. EPA's calculation of
land releases for this OES, including all calculation inputs and assumptions, can be found in 1,4-
Dioxane Supplemental Information File: Environmental Releases to Landfor all OES Except Disposal
(U.S. EPA. 2023kY

Hydraulic Fracturing

EPA estimated land releases for facilities within the Hydraulic Fracturing OES using the Draft OECD
ESD on Hydraulic Fracturing (	22d) and Monte Carlo modeling. The Draft ESD on

Hydraulic Fracturing was rated "high" during EPA's systematic review process. The use of Montel
Carlo modeling allows for variation of calculation input parameters such that a distribution of
environmental releases can be calculated, from which EPA can estimate the 50th and 95th percentile
releases. An explanation of this modeling approach is included in Appendix E.13.

E.4.3 Land Release Estimates Summary

A summary of industrial and commercial land releases estimated using the above methods is presented
in Table Apx E-5 below.

Page 259 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

7034 Table Apx E-5. Summary of Daily Industrial and Commercial Land Release Estimates for 1,4-Dioxane





Number of

Estimated Daily Release Range
across Sites

Estimated
Release

Overall Data



OES

Type of Land Release

Facilities with
Releases"

(kg/site-day)

Frequency
Range
(days)''

Quality
Determination

Sources'





Min

Max



Manufacturing

Land (all types)

1

0

250

Medium

TRI

Import and
Repackaging

Land (all types)

1

0

250

Medium

TRI

Industrial Uses

Land (all types)

12

0

227 (annually)

250

Medium

TRI

Functional Fluids

Land (all types)

2

0

0

247

Medium

TRI

(Open-System)















Functional Fluids

All

Assessed as a part of Industrial Uses OES



N/A

N/A

(Closed-System)















Laboratory
Chemical

Land (unknown type)

132

0

489 (annually)

250

High

GSd

Film Cement

Land (unknown type)

211

0.0035
(annually)

0.037
(annually)

250

High

Process
information®

Spray Foam
Application

Land (unknown type)

1,553,559

0.032 (annually)

0.047
(annually)

3

Medium

GS^

Printing Inks (3D)

Fugitive Air, Stack

Assessed as a part of Industrial Uses OES

250

N/A

N/A



Air, and Land (all















types)













Dry Film
Lubricant

Land (hazardous waste
landfill)

8

0

188 (annually)

48

High

Process
information®

Page 260 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Type of Land Release

Number of
Facilities with
Releases"

Estimated Daily Release Range
across Sites
(kg/site-day)

Estimated

Release
Frequency
Range
(days)''

Overall Data

Quality
Determination

Sources'

Min

Max

Disposal

Land (RCRA Sub C
landfill)

18

0

7,307
(annually)

250

Medium

TRI

Land (Underground
injection)

18

0

331,980
(annually)

Medium

TRI

Land (Non-RCRA
landfills)

18

0

890 (annually)

Medium

TRI

Land (all other types)

18

0

0

Medium

TRI

Textile Dye

Land (unknown
landfill type) or
POTW (unknown
partitioning)

783

2.09E-07

9.72E-05

31 to 295

Medium

ESDg and
Modeling''

Antifreeze

Land (unknown
landfill)

84,383

3.75E-07
(annually)

0.029
(annually)

250

High

Process

information® and
Modeling''

Surface Cleaner

Land (unknown
landfill) or POTW

Unknown

18™

250

High

SHEDS-HT',
Process
information®
Modeling''

Dish Soap

Land (unknown
landfill)

Unknown

0.048 (annual
value for all
sites in

Liverpool OH
case study)

0.097 (annual
value for all
sites in

Liverpool OH
case study)

250

High

SHEDS-HT',
Process
information®
Modeling''

Dishwasher
Detergent

Land (unknown
landfill)

Unknown

1.08E-03
(annual value for
all sites in
Liverpool, OH
case study)

2.17E-03
(annual value
for all sites in
Liverpool, OH
case study)

250

High

SHEDS-HT',
Process
information®
Modeling''

Laundry Detergent
(Institutional) -
liquid detergents

Land (unknown
landfill), incineration,

95,533

4.05E-12

3.95E-05

250 to 365

Medium

ESD! and
Modeling''

Page 261 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Type of Land Release

Number of
Facilities with
Releases"

Estimated Daily Release Range
across Sites
(kg/site-day)

Estimated

Release
Frequency
Range
(days)''

Overall Data

Quality
Determination

Sources'





Min

Max





or POTW (unknown
partitioning)













Laundry Detergent
(Institutional) -
powder detergents

Land (unknown
landfill), incineration,
or POTW (unknown
partitioning)

95,533

5.36E-08

0.0018

250 to 365

Medium

ESD! and
Modeling''

Laundry Detergent
(Industrial) -
liquid detergents

Land (unknown
landfill), incineration,
or POTW (unknown
partitioning)

2,453

4.78E-12

1.46E-04

20 to 365

Medium

ESD! and
Modeling''

Laundry Detergent
(Industrial) -
powder detergents

Land (unknown
landfill), incineration,
or POTW (unknown
partitioning)

2,453

2.92E-11

3.92E-04

20 to 365

Medium

ESD! and
Modeling''

Paints and Floor
Lacquer

Land (unknown
landfill)

33,648

3.04E-06
(annually)

0.010
(annually)

250

Medium

ESD7

Polyethylene
Terephthalate
(PET) Byproduct

Land (Land treatment)

13

0

45.4 (annually)



Medium

TRI

Land (Non-RCRA
landfills)

13

0

0.10 (annually)

250

Medium

TRI

Land (all other types)

13

0

0



Medium

TRI

Ethoxylation
Process Byproduct

Land (underground
injection)

8

0

197,714
(annually)

250

Medium

TRI

Land (all other types)

8

0

0



Medium

TRI

Page 262 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Type of Land Release

Number of
Facilities with
Releases"

Estimated Daily Release Range
across Sites
(kg/site-day)

Estimated

Release
Frequency
Range
(days)''

Overall Data

Quality
Determination

Sources'

Min

Max

Hydraulic
Fracturing

Surface water,
incineration, or landfill
(unknown
partitioning)

411

3.61E-10

4.59

1 to 72

Medium

ESD' and
Modeling11

Land (underground
injection)

411

5.35E-09

108

Medium

ESD' and
Modeling''

Recycle/Reuse (48%),
underground injection
(43%), Surface water
(6%), or land (3%)

411

1.85E-10

1.12

Medium

ESD'' and
Modeling''

"Where available. EPA used 2013-2019 TRI (U.S. EPA, 2022g) and 2013-2019 DMR (U.S. EPA, 2022c) data to provide a basis to estimate the number of sites
using 1,4-dioxane within a COU.

h Where available. EPA used the December 2020 Final Risk Evaluation for 1,4-Dioxane (U.S. EPA, 2020c). generic scenarios, and emission scenario documents
to provide a basis to estimate the number of release days of 1,4-dioxane within a COU.
c Narrative descriptions of all release estimate sources are provided in Appendix E.4.2.
d The generic scenario used for this COU is the GS on Use of Laboratory Chemicals (U.S. EPA. 2022h).

'' For this COU, EPA used process information, which is further described in Appendix E.4.2.

' The generic scenario used for this COU is the GS on Application of Sprav Polvurethane Foam Insulation (U.S. EPA. 2018b).

8 The emission scenario document used for this COU is the ESD on Textile Dves (OECD. 2017).
h For this COU, EPA used various models and literature for model input parameters as described in Appendix E.4.2.

1 The emission scenario document used for this COU is the ESD on Industrial and Institutional Laundries (OECD, 201 lb).

1 The emission scenario document used for this COU is the ESD on Coating Application via Sprav Painting in the Automotive Refinishing Industrv (OECD.

2011a).

k The emission scenario document used for this COU is the Draft ESD on Hvdraulic Fracturing (U.S. EPA, 2022d).

' EPA used the down the drain water release estimates from the SHEDs-HT model for the Liverpool OH case study (see Section 2.1.1.2) to estimate air and land
releases by back calculating 1,4-dioxane use rates and applying loss fractions for air and land releases using literature and standard models described in Appendix
E.4.2.

m A single annual value was provided for all sites in the Liverpool, OH case study.

7035	E.4.4 Summary of Weight of the Scientific Evidence Conclusions in Land Release Estimates

7036	TableApx E-6 provides a summary of EPA's weight of the scientific evidence conclusions in its land release estimates for each of the

7037	Occupational Exposure Scenarios assessed. Detailed descriptions of non-OES specific strengths, limitations, assumptions, and uncertainties

7038	{e.g., general limitations for TRI, DMR, etc.) are provided in Appendix E.6.

Page 263 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

7039 Table Apx E-6. Summary of Weight of the Scientific Evidence Conclusions in Land Release Estimates by PES

OES

Weight of the Scientific Evidence Conclusion in Release Estimates

Manufacturing

Land releases are assessed using reported releases from 2019 TRI. Factors that increase the strength of evidence for this
OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium overall data
quality determination, and consistency within the dataset (all reporters use the same or similar reporting forms). A strength
of TRI data is that TRI compiles the best readily available release data for all reporting facilities. Factors that decrease the
strength of the evidence for this OES include the low number of data points, lack of variability (only one year of data
used), uncertainty in the accuracy of reported releases, and the limitations in representativeness to all sites because TRI
may not capture all relevant sites. Additionally, EPA could not estimate the number of release days per year associated
with land releases. Based on this information, EPA has concluded that the weight of the scientific evidence for this
assessment is moderate to robust and provides a plausible estimate of releases in consideration of the strengths and
limitations of reasonably available data.

Import and Repackaging

Land releases are assessed using reported releases from 2019 TRI. Factors that increase the strength of evidence for this
OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium overall data
quality determination, and consistency within the dataset (all reporters use the same or similar reporting forms). A strength
of TRI data is that TRI compiles the best readily available release data for all reporting facilities. Factors that decrease the
strength of the evidence for this OES include the low number of data points, lack of variability (only one year of data
used), uncertainty in the accuracy of reported releases, and the limitations in representativeness to all sites because TRI
may not capture all relevant sites. Additionally, the land release assessment is based on one reporting site that reported no
land releases and EPA did not have additional sources to estimate land releases for other sites in this OES. Additionally,
EPA could not estimate the number of release days per year associated with land releases. Based on this information, EPA
has concluded that the weight of the scientific evidence for this assessment is moderate to robust and provides a plausible
estimate of releases in consideration of the strengths and limitations of reasonably available data.

Industrial Uses

Land releases are assessed using reported releases from 2019 TRI. Factors that increase the strength of evidence for this
OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium overall data
quality determination, and consistency within the dataset (all reporters use the same or similar reporting forms). A strength
of TRI data is that TRI compiles the best readily available release data for all reporting facilities. Factors that decrease the
strength of the evidence for this OES include the lack of variability (only one year of data used), uncertainty in the
accuracy of reported releases, uncertainty in EPA's use of Form A submissions, and the limitations in representativeness to
all sites because TRI may not capture all relevant sites. Some facilities within this OES reported to TRI using a Form A,
which does not include any details on chemical release quantities. When a facility has submitted a Form A, there is no way
to discern the quantity released. Therefore, where facilities reported to TRI with a Form A, EPA used the Form A threshold
for total releases of 500 lb/year for each release media; however, there is uncertainty in this because the actual release
quantity is unknown. Furthermore, the threshold represents an upper limit on total releases from the facility; therefore,
assessing releases at the threshold value may overestimate actual releases from the facility. Additionally, EPA could not
estimate the number of release days per year associated with land releases. Based on this information, EPA has concluded

Page 264 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates



that the weight of the scientific evidence for this assessment is moderate to robust and provides a plausible estimate of
releases in consideration of the strengths and limitations of reasonably available data.

Functional Fluids (Open-
System)

Land releases are assessed using reported releases from 2019 TRI. Factors that increase the strength of evidence for this
OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium overall data
quality determination, and consistency within the dataset (all reporters use the same or similar reporting forms). A strength
of TRI data is that TRI compiles the best readily available release data for all reporting facilities. Factors that decrease the
strength of the evidence for this OES include the low number of data points, lack of variability (only one year of data
used), uncertainty in the accuracy of reported releases, and the limitations in representativeness to all sites because TRI
may not capture all relevant sites. The land release assessment is based on two reporting sites that both reported no land
releases and EPA did not have additional sources to estimate land releases for sites in this OES. Additionally, EPA could
not estimate the number of release days per year associated with land releases. Based on this information, EPA has
concluded that the weight of the scientific evidence for this assessment is moderate to robust and provides a plausible
estimate of releases in consideration of the strengths and limitations of reasonably available data.

Functional Fluids (Closed-
System)

No data was available to estimate releases for this OES. For the land release assessment, EPA grouped this OES with the
Industrial Uses OES because the sources of release are expected to be similar between these OESs. Factors that increase
the strength of evidence for this OES are that TRI has a medium overall data quality determination and consistency within
the dataset (all reporters use the same or similar reporting forms). Factors that decrease the strength of evidence for this
OES are that the Industrial Releases OES release data are use as surrogate for this OES, uncertainty in the accuracy of
reported releases, limitations in representativeness to all sites because TRI may not capture all relevant sites, and lack of
variability (only one year of data used). Refer to the Industrial Uses OES discussion for additional discussion. Based on
this information, EPA has concluded that the weight of the scientific evidence for this assessment is slight and provides a
plausible estimate of releases in consideration of the strengths and limitations of reasonably available data.

Laboratory Chemicals

Land releases are assessed using the Draft GS on Use of Laboratory Chemicals. Factors that increase the strength of
evidence for this OES are that the release estimates are directly relevant to the OES (as opposed to surrogate), the Draft GS
on Use of Laboratory Chemicals has a high overall data quality determination, and the low level of uncertainty in the data.
Factors that decrease the strength of the evidence for this OES include the that the GS has not been peer-reviewed,
uncertainty in the representativeness of the GS towards all sites in this OES, and a lack of variability in the analysis.
Specifically, because the default values in the ESD are generic, there is uncertainty in the representativeness of generic site
estimates of actual releases from real-world sites that use 1,4-dioxane. Another uncertainty is lack of consideration for
release controls. The ESD assumes that all activities occur without any release controls. Actual releases may be less than
estimated if facilities utilize pollution control methods, contributing to uncertainty. Based on this information, EPA has
concluded that the weight of the scientific evidence for this assessment is slight to moderate and provides a plausible
estimate of releases in consideration of the strengths and limitations of reasonably available data.

Film Cement

Land releases are assessed using process information from the Final Risk Evaluation for 1,4-Dioxane and EPA/OPPT
models. Factors that increase the strength of evidence for this OES are that the release estimates are directly relevant to the

Page 265 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates



OES (as opposed to surrogate), the underlying data sources for the process information have a high overall data quality
determination, and the low level of uncertainty in the data because the process information comes directly from actual
users of 1,4-dioxane in film cement. Factors that decrease the strength of the evidence for this OES include uncertainty in
the representativeness of evidence to all sites in this OES and a lack of variability in the input parameters for the used
models. Specifically, the process information for the production and use of film cement is based on information from three
use sites, one from Australia and two from the U.S. Based on this information, EPA has concluded that the weight of the
scientific evidence for this assessment is slight to moderate and provides a plausible estimate of releases in consideration
of the strengths and limitations of reasonably available data.

Spray Foam Application

Land releases are assessed using the GS on Application of Spray Polyurethane Foam Insulation. Factors that increase the
strength of evidence for this OES are that the release estimates are directly relevant to the OES (as opposed to surrogate),
the underlying data sources for the process information have a medium overall data quality determination, and a low level
of uncertainty in the data. Factors that decrease the strength of the evidence for this OES include uncertainty in the
representativeness of the GS to all sites since it is generic and not specific to sites that use 1,4-dioxane and a lack of
variability. Based on this information, EPA has concluded that the weight of the scientific evidence for this assessment is
slight to moderate and provides a plausible estimate of releases in consideration of the strengths and limitations of
reasonably available data.

Printing Inks (3D)

No data was available to estimate releases for this OES. For the land release assessment, EPA grouped this OES with the
Industrial Uses OES because the sources of release are expected to be similar between these OESs. Factors that increase
the strength of evidence for this OES are that TRI has a medium overall data quality determination and consistency within
the dataset (all reporters use the same or similar reporting forms). Factors that decrease the strength of evidence for this
OES are that the Industrial Releases OES release data are use as surrogate for this OES, uncertainty in the accuracy of
reported releases, limitations in representativeness to all sites because TRI may not capture all relevant sites or smaller
commercial 3D printing uses, and lack of variability (only one year of data used). Refer to the Industrial Uses OES
discussion for additional discussion. Based on this information, EPA has concluded that the weight of the scientific
evidence for this assessment is slight and provides a plausible estimate of releases in consideration of the strengths and
limitations of reasonably available data.

Dry Film Lubricant

Land releases are assessed using process information from the Final Risk Evaluation for 1,4-Dioxane. Factors that increase
the strength of evidence for this OES are that the release estimates are directly relevant to the OES (as opposed to
surrogate), that the underlying data sources for the process information have a high overall data quality determination, and
a low level of uncertainty in the data because the process information comes directly from an actual user of 1,4-dioxane in
dry film lubricants. Factors that decrease the strength of the evidence for this OES include uncertainty in the
representativeness of evidence to all sites and a lack of variability. Based on this information, EPA has concluded that the
weight of the scientific evidence for this assessment is slight to moderate and provides a plausible estimate of releases in
consideration of the strengths and limitations of reasonably available data.

Page 266 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates

Disposal

Land releases are assessed using reported discharges from 2013-2019 TRI. Factors that increase the strength of evidence
for this OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium
overall data quality determination, high number of data points, and consistency within the dataset (all reporters use the
same or similar reporting forms). Additionally, EPA included seven years of TRI data in the analysis, which increases the
variability of the dataset. A strength of TRI data is that TRI compiles the best readily available release data for all reporting
facilities. Factors that decrease the strength of the evidence for this OES include uncertainty in the accuracy of reported
releases, uncertainty in EPA's use of Form A submissions, and the limitations in representativeness to all sites because TRI
may not capture all relevant sites. Some facilities within this OES reported to TRI using a Form A, which does not include
any details on chemical release quantities. When a facility has submitted a Form A, there is no way to discern the quantity
released. Therefore, where facilities reported to TRI with a Form A, EPA used the Form A threshold for total releases of
500 lb/year for each release media; however, there is uncertainty in this because the actual release quantity is unknown.
Furthermore, the threshold represents an upper limit on total releases from the facility; therefore, assessing releases at the
threshold value may overestimate actual releases from the facility. Based on this information, EPA has concluded that the
weight of the scientific evidence for this assessment is moderate to robust and provides a plausible estimate of releases in
consideration of the strengths and limitations of reasonably available data.

Textile Dye

Land releases are assessed using Monte Carlo modeling with information from the ESD on Textile Dyes. Factors that
increase the strength of evidence for this OES are that the release estimates are directly relevant to the OES (as opposed to
surrogate), the ESD on Textile Dyes has a medium overall data quality determination and was peer reviewed, the high
number of data points (simulation runs), consistency within the dataset, and full distributions of input parameters. The
Monte Carlo modeling accounts for the entire distribution of input parameters, calculating a distribution of potential
release values that represents a larger proportion of sites than a discrete value. Factors that decrease the strength of the
evidence for this OES include uncertainties and limitations in the representativeness of the estimates for sites that
specifically use 1,4-dioxane because the default values in the ESD are generic. Another uncertainty is lack of consideration
for release controls. The ESD assumes that all activities occur without any release controls. Actual releases may be less
than estimated if facilities utilize pollution control methods, contributing to uncertainty. Based on this information, EPA
has concluded that the weight of the scientific evidence for this assessment is moderate and provides a plausible estimate
of releases in consideration of the strengths and limitations of reasonably available data.

Antifreeze

Land releases are assessed using the OECD ESD on Chemical Additives used in Automotive Lubricants, the EPA MRD on
Commercial Use of Automotive Detailing Products, and EPA/OPPT models. Factors that increase the strength of evidence
for this OES are that the ESD and MRD used have high overall data quality determinations and consistency within the
sources used. Factors that decrease the strength of the evidence for this OES include that the ESD and MRD are not
directly applicable to antifreeze uses (used as surrogate), uncertainty in the representativeness of the ESD and MRD to all
sites and sites that specifically use 1,4-dioxane since these documents contain generic values, and a lack of variability.
Additionally, EPA scaled up a consumer antifreeze use rate to a commercial use rate based on information in the ESD and
MRD, which increases uncertainty. Based on this information, EPA has concluded that the weight of the scientific

Page 267 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates



evidence for this assessment is slight to moderate and provides a plausible estimate of releases in consideration of the
strengths and limitations of reasonably available data.

Surface Cleaner

Land releases are assessed using SHEDS-HT modeled water releases in conjunction with the Draft GS on Furnishing
Cleaning. Factors that increase the strength of evidence for this OES include that the release estimates are directly relevant
to the OES (as opposed to surrogate), that the Draft GS used has a high overall data quality determination, and variability
in the model input parameters. Factors that decrease the strength of the evidence for this OES include uncertainty in the
representativeness to all sites because the SHEDS-HT estimate is based on one case study for Liverpool, OH and because
the estimate is not site-specific (the release estimate is a total for all sites in Liverpool, OH). Additionally, the Draft GS
describes potential release points for this OES, identifying releases that may be to either water or land depending on site
practices (e.g., surface cleaner may be rinsed down drains or wiped off with rags that are disposed of as trash). Because
there is no information to determine the quantity released specifically to land, EPA assumed that the entire quantity
modeled to water with the SHEDS-HT model may also be released to land, which introduces uncertainty. Based on this
information, EPA has concluded that the weight of the scientific evidence for this assessment is slight and provides a
plausible estimate of releases in consideration of the strengths and limitations of reasonably available data.

Dish Soap

Land releases are assessed using the SHEDS-HT modeled water releases, the expected sources of release for this OES, and
EPA/OPPT models. Factors that increase the strength of evidence for this OES include that the release estimates are
directly relevant to the OES (as opposed to surrogate) and variability in the model input parameters. Factors that decrease
the strength of the evidence for this OES include uncertainty in the representativeness to all sites because the SHEDS-HT
estimate is based on one case study for Liverpool, OH and because the estimate is not site-specific (the release estimate is a
total for all sites in Liverpool, OH). Additionally, based on the quantity released to water and the expected loss fraction for
water releases, EPA back-calculated a 1,4-dioxane use rate and applied the expected land release loss fraction to estimate
land releases, which introduces uncertainty. Based on this information, EPA has concluded that the weight of the scientific
evidence for this assessment is slight and provides a plausible estimate of releases in consideration of the strengths and
limitations of reasonably available data.

Dishwasher Detergent

EPA used the same approach as described above for the dish soap OES to estimate land releases for the dishwasher
detergent OES.

Laundry Detergent

Land releases are assessed using Monte Carlo modeling with information from the ESD on Industrial and Institutional
Laundries. Factors that increase the strength of evidence for this OES are that the release estimates are directly relevant to
the OES (as opposed to surrogate), that the ESD on Industrial and Institutional Laundries has a medium overall data
quality determination and was peer reviewed, there are high number of data points (simulation runs), consistency within
the dataset, and full distributions of input parameters. The Monte Carlo modeling accounts for the entire distribution of
input parameters, calculating a distribution of potential release values that represents a larger proportion of sites than a
discrete value. Additionally, EPA was able to separately estimate releases for industrial and institutional laundry settings.
Factors that decrease the strength of the evidence for this OES include uncertainties and limitations in the
representativeness of the estimates for sites that specifically use 1,4-dioxane because the default values in the ESD are

Page 268 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates



generic. Another uncertainty is lack of consideration for release controls. The ESD assumes that all activities occur without
any release controls. Actual releases may be less than estimated if facilities utilize pollution control methods. Based on this
information, EPA has concluded that the weight of the scientific evidence for this assessment is moderate and provides a
plausible estimate of releases in consideration of the strengths and limitations of reasonably available data.

Paint and Floor Lacquer

Land releases are assessed using OECD ESD on Coating Application via Spray-Painting in the Automotive Refinishing
Industry. Factors that increase the strength of evidence for this OES are that the release estimates are directly relevant to
the OES (as opposed to surrogate), the ESD has a medium overall data quality determination and has been peer reviewed,
consistency within the sources used, and a low amount of uncertainties. Factors that decrease the strength of the evidence
for this OES include a lack of variability and uncertainty in the representativeness of the ESD to all sites and sites that
specifically use 1,4-dioxane since the ESD is generic. Based on this information, EPA has concluded that the weight of the
scientific evidence for this assessment is slight to moderate and provides a plausible estimate of releases in consideration
of the strengths and limitations of reasonably available data.

PET Byproduct

Land releases are assessed using reported discharges from 2019 TRI. Factors that increase the strength of evidence for this
OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium overall data
quality determination, and consistency within the dataset (all reporters use the same or similar reporting forms). A strength
of TRI data is that TRI compiles the best readily available release data for all reporting facilities. Factors that decrease the
strength of the evidence for this OES include uncertainty in the accuracy of reported releases, lack of variability (only one
year of data used), and the limitations in representativeness to all sites because TRI may not capture all relevant sites. The
land release assessment is based on 13 reporting sites, 11 of which reported no land releases. EPA did not have additional
sources to estimate land releases for site in this OES that may not report to TRI. Based on this information, EPA has
concluded that the weight of the scientific evidence for this assessment is moderate to robust and provides a plausible
estimate of releases in consideration of the strengths and limitations of reasonably available data.

Ethoxylation Process
Byproduct

Land releases are assessed using reported discharges from 2019 TRI. Factors that increase the strength of evidence for this
OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium overall data
quality determination, and consistency within the dataset (all reporters use the same or similar reporting forms). A strength
of TRI data is that TRI compiles the best readily available release data for all reporting facilities. Factors that decrease the
strength of the evidence for this OES include uncertainty in the accuracy of reported releases, lack of variability (only one
year of data used), and the limitations in representativeness to all sites because TRI may not capture all relevant sites. The
land release assessment is based on eight reporting sites, seven of which reported no land releases. EPA did not have
additional sources to estimate land releases for site in this OES that may not report to TRI. Based on this information, EPA
has concluded that the weight of the scientific evidence for this assessment is moderate to robust and provides a plausible
estimate of releases in consideration of the strengths and limitations of reasonably available data.

Hydraulic Fracturing

Land releases are assessed using Monte Carlo modeling with information from the Draft ESD on Hydraulic Fracturing and
FracFocus 3.0. Factors that increase the strength of evidence for this OES are that the release estimates are directly relevant
to the OES (as opposed to surrogate), that the Draft ESD on Hydraulic Fracturing and FracFocus 3.0 have medium overall

Page 269 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates



data quality determinations, high number of data points (simulation runs), consistency within the dataset, and full
distributions of input parameters. The Monte Carlo modeling accounts for the entire distribution of input parameters,
calculating a distribution of potential release values that represents a larger proportion of sites than a discrete value. Factors
that decrease the strength of the evidence for this OES include that the Draft ESD has not been peer reviewed and the
uncertainties and limitations in the representativeness of the estimates for sites that specifically use 1,4-dioxane because
the default values from the Draft ESD on Hydraulic Fracturing. Another uncertainty is lack of consideration for release
controls. The ESD assumes that all activities occur without any release controls. Actual releases may be less than estimated
if facilities utilize pollution control methods, contributing to uncertainty. Based on this information, EPA has concluded
that the weight of the scientific evidence for this assessment is moderate to robust and provides a plausible estimate of
releases in consideration of the strengths and limitations of reasonably available data.

7040

Page 270 of 484


-------
7041

7042

7043

7044

7045

7046

7047

7048

7049

7050

7051

7052

7053

7054

7055

7056

7057

7058

7059

7060

7061

7062

7063

7064

7065

7066

7067

7068

7069

7070

7071

7072

7073

7074

7075

7076

7077

7078

7079

7080

7081

7082

7083

7084

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

E.5 Air Release Assessment

This section describes EPA's methodology for estimating daily fugitive and stack air emissions from
industrial and commercial facilities manufacturing, processing, or using 1,4-dioxane. Facilities report air
emissions to the TRI. EPA used 2019 TRI (I v « « 4.. 2022e) data to estimate daily air emissions for
each OES where available; however, EPA did not have these data for every OES. For OES without TRI
data, EPA used alternate assessment approaches to estimate air emissions. These approaches are
described below.

E.5.1 Assessment Using TRI

EPA found 2019 TRI data for facilities within the following OESs:

•	Manufacturing

•	Import and repackaging

•	Industrial uses

•	Functional fluids (open-system)

•	Disposal

•	PET byproduct

•	Ethoxylation byproduct

The 2019 TRI data were rated "medium" in EPA's systematic review process. EPA estimated daily air
emissions using TRI data for these OESs, with the following general steps as described in the rest of this
section.

1.	Collect air emission data from 2019 TRI data,

2.	Map air emission data to occupational exposure scenarios,

3.	Estimate the number of facility operating days per year, and

4.	Estimate daily air emissions and prepare a summary of the air emissions for each OES.

Note that EPA compared the TRI data used to estimate air releases for the PET byproduct OES in this
risk evaluation to information from a life cycle analysis on the PET manufacturing process in Appendix
E.6.

Step 1: Collect Air Emission Data TRI

The first step in the methodology for estimating air emissions was to obtain 2019 TRI data for the
chemical from EPA's Basic Plus Data Files. TRI requires U.S. facilities in various industry sectors to
report the annual release volumes to the environment through air emissions, water discharges, and land
disposal, and/or managed through recycling, energy recovery, and treatment, including by off-site
transfers. TRI reporters may report either with a Form R or a Form A. Facilities must report with a Form
R, which requires reporting of release quantities and uses/sub-uses of the chemical, among other
information, unless they meet the alternate threshold requirements for submitting a Form A.

Specifically, facilities may submit a Form A if the volume of chemical manufactured, processed, or
otherwise used does not exceed 1,000,000 lb per year (lb/year) and the total annual reportable releases
do not exceed 500 lb/year. Facilities do not need to report release quantities or uses/sub-uses on Form A.
EPA included both TRI reporting Form R and TRI reporting Form A submissions in the air release
assessment.

Air emissions in TRI are reported separately for stack air and fugitive air and always occur on-site at the
facility. Where sites reported to 2019 TRI with Form A, EPA used the Form A threshold for total
releases of 500 lb/year. EPA used the entire 500 lb/year for both the fugitive and stack air release

Page 271 of 484


-------
7085

7086

7087

7088

7089

7090

7091

7092

7093

7094

7095

7096

7097

7098

7099

7100

7101

7102

7103

7104

7105

7106

7107

7108

7109

7110

7111

7112

7113

7114

7115

7116

7117

7118

7119

7120

7121

7122

7123

7124

7125

7126

7127

7128

7129

7130

7131

7132

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

estimates; however, since this threshold is for total site releases, these 500 lb/year are attributed either to
fugitive air or stack air for this analysis—not both (to avoid double counting the releases and exceeding
the total release threshold for Form A).

Step 2: Map Air Emission Data to Occupational Exposure Scenarios

In the next step of air release assessment, EPA mapped the chemical's 2019 TRI data to the 1,4-dioxane
OES. EPA used the following procedure to map 2019 TRI data to OES:

1.	Compile TRI uses/sub-uses: EPA first compiled all the reported TRI uses/sub-uses for each
facility into one column.

2.	Map TRI uses/sub-uses to Chemical Data Reporting (CDR) IFC codes: EPA then mapped the
TRI uses/sub-uses for each facility to one or more 2016 CDR Industrial Function Category (IFC)
codes using the TRI-to-CDR Use Mapping crosswalk (see Appendix E.9).

3.	Map OES to CDR IFC codes: EPA then mapped each COU/OES combination to a 2016 CDR
IFC code and description. The basis for this mapping was generally the COU category and
subcategory.

4.	Map TRIfacilities to an OES: Using the CDR IFC codes from Step 2 and the COU-CDR
Mapping from Step 3, EPA mapped each TRI facility to an OES. EPA's rationale for the OES
determination is generally described below.

o In some cases, the facility mapped to only one OES and the mapping appeared to be
correct given the facility name and NAICS code. For these, the OES as mapped from
Steps 2 and 3 was used without adjustment,
o There were instances where a facility mapped to multiple OESs which required some
engineering judgement to identify a primary OES. EPA documented the rationale for
these determinations for each facility in 1,4-Dioxane Supplemental Information File:
Environmental Releases to Air (	023]). In summary, these determinations

were made with the following considerations:

•	Industry and NAICS codes reported in TRI (e.g., for a facility that reported TRI
uses for both waste treatment and ancillary use, EPA assigned the Disposal OES
if the NAICS code was 562211, Hazardous Waste Treatment and Disposal);

•	Internet research of the types of products manufactured at the facility (e.g., if a
facility's website indicates the facility manufactures PET, the facility is likely to
produce 1,4-dioxane as a byproduct in PET manufacturing); and

•	Grouping of similar OES (e.g., for facilities that reported the sub-use of chemical
processing aid, process solvent, or processing as a reactant), EPA assigned the
Industrial Uses OES because this includes multiple processes as described in the
2020 RE (	020cV).

o In some cases, EPA identified that there were instances where the preliminary OES
mapping from the TRI use/sub-use - CDR IFC code required re-mapping. This re-
mapping is a result of limitations of the TRI-to-CDR Use Mapping crosswalk. For
example, the crosswalk maps the TRI use/sub-use of "Otherwise Use as Manufacturing
Aid (Other)" to only CDR IFC codes U013 (closed-system functional fluids) and U023
(plating agents and surface treating agents); however, this TRI use/sub-use may
encompass multiple other uses that are not captured in these CDR IFC codes. In these
cases, EPA reviewed the reported NAICS codes and conducted internet research on the
types of products manufactured at the facility to determine the likely OES.
o Additionally, EPA reviewed 2016 CDR (U.S. EPA. 2016b) for sites that reported
manufacturing (including importing) of 1,4-dioxane. If the sites that reported to 2016
CDR also reported in 2019 TRI, EPA assigned the OES according to 2016 CDR. Note

Page 272 of 484


-------
7133

7134

7135

7136

7137

7138

7139

7140

7141

7142

7143

7144

7145

7146

7147

7148

7149

7150

7151

7152

7153

7154

7155

7156

7157

7158

7159

7160

7161

7162

7163

7164

7165

7166

7167

7168

7169

7170

7171

7172

7173

7174

7175

7176

7177

7178

7179

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

that some sites that reported to 2019 TRI may not be in 2016 CDR (e.g., sites that
manufacture the chemical as a byproduct). In these cases, EPA determined the OES using
only the above bulleted steps.

5. Form A's: For Form A submissions, there were no reported TRI uses/sub-uses. To determine the
COU for these facilities, EPA used 2016 CDR as described above, the NAICS codes, and
internet searches to determine the type of products and operations at the facility.

The specific rationale for the OES mapping for each facility is described in 1,4-Dioxane Supplemental
Information File: Environmental Releases to Air (U.S. EPA. 2023D.

Step 3: Estimate the Number of Facility Operating Days per Year

EPA then estimated the number of operating days (days/year) for each facility reporting air emissions to
TRI. For the OES that were included in the Final Risk Evaluation for 1,4-Dioxane (	2020c\

EPA used the number of operating days from that risk evaluation. For the additional OES included in
this supplemental risk evaluation, EPA estimated the number of operating days using the methodology
described in Appendix E.2.

Step 4: Estimate Daily Air Emissions and Summarize Air Emissions for each OES
The final step was to prepare a summary of the fugitive and stack releases. For each OES and facility
mapped to that OES, EPA summarized the annual fugitive and stack air emissions reported in 2019 TRI
and daily fugitive and stack air emissions that EPA estimated by dividing the annual emissions by the
number of operating days determined for the OES in Step 3. EPA also summarized site information,
including site identity, city, state, zip code, TRI facility ID, and Facility Registry Service (FRS) ID
because the subsequent exposure modeling is site and location specific. Latitude and longitude
coordinates are included in 1,4-Dioxane Supplemental Information File: Environmental Releases to Air
(1 c. i i1 \ 2023D but not in the summary tables.

To accompany the summary table for each OES, EPA also provided any reasonably available
information on the release duration and pattern, which are needed for the exposure modeling. Release
duration is the expected amount of time per day during which the air releases may occur. Release pattern
is the temporal variation of the air release, such as over consecutive days throughout the year, over
cycles that occur intermittently throughout the year, or in a puff/instantaneous release that occurs over a
short duration. The TRI dataset does not include release pattern or duration; therefore, EPA used
information from models or literature, where available. For release pattern, EPA provided the number of
release days with the associated basis as described in Step 3. However, for most OES, no information
was found on release duration and pattern. In such cases, EPA listed the release duration and pattern as
"unknown."

EPA's summary of air releases for each OES is included in 1,4-Dioxane Supplemental Information File:
Environmental Releases to Air (	323i).

E.5.2 Assessment for OESs without TRI

EPA did not find TRI data for any of the years included in this analysis for the following OESs:

•	Functional fluids (closed-systems)

•	Laboratory chemicals

•	Film cement

•	Spray polyurethane foam

•	3D printing

•	Dry film lubricant

Page 273 of 484


-------
7180

7181

7182

7183

7184

7185

7186

7187

7188

7189

7190

7191

7192

7193

7194

7195

7196

7197

7198

7199

7200

7201

7202

7203

7204

7205

7206

7207

7208

7209

7210

7211

7212

7213

7214

7215

7216

7217

7218

7219

7220

7221

7222

7223

7224

7225

7226

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

•	Textile dye

•	Antifreeze

•	Surface cleaner

•	Dish soap

•	Dishwasher detergent

•	Laundry detergent

•	Paints and floor lacquer

•	Hydraulic fracturing

For these OESs, EPA estimated air emissions by using various modeling approaches, including the use
of surrogate TRI data and data from literature, GSs, and ESDs. EPA's assessment of air emissions for
each of these OESs is described below.

Functional Fluids (Closed-Systems)

Air emission data were not available for this OES and EPA did not find any information to model air
emissions for this OES using literature, GSs, or ESDs. EPA expects that the sources of release for this
OES to be similar to those for the Industrial Uses OES, based on the process information in the Final
Risk Evaluation for 1,4-Dioxane (U.	2020c). Therefore, EPA grouped the air release assessment

for Functional Fluids (Closed-Systems) into the OES for Industrial Uses. However, there is uncertainty
in the assumption of similar release sources between these OESs.

Laboratory Chemicals

EPA estimated air emissions for facilities within the Laboratory Chemicals OES using the Draft GS on
Use of Laboratory Chemicals (	22h). The Draft GS on Use of Laboratory Chemicals was

rated "high" during EPA's systematic review process.

The GS indicates that there are eight release points:

1.	Release to air from transferring volatile chemicals from transport containers

2.	Release to air, water, incineration, or landfill from transferring solid powders

3.	Release to water, incineration, or land from cleaning or disposal of transport containers

4.	Release to air from cleaning containers used for volatile chemicals

5.	Labware equipment cleaning residuals released to water, incineration, or landfill

6.	Release to air during labware equipment cleaning for volatile chemicals

7.	Release to air from laboratory analyses for volatile chemicals

8.	Release to water, incineration, or landfill from laboratory waste disposal

Based on the GS, release points #1, 2, 4, 6, and 7 have the potential for air emissions; however, release
point #2 is not applicable because 1,4-dioxane is not a solid powder. To estimate the remaining air
releases, EPA used the equations specified in the Draft GS (	22h). EPA did not find any

data specific to 1,4-dioxane in this OES. Therefore, the calculation of releases using this GS are for a
"generic site," using the default input parameter values from the GS.

Using this methodology, EPA calculated the "typical" and "worst-case" air emissions for this OES.
These characterizations are based on the GS, which provides default "typical" and "worst-case" input
parameters for the release calculations. EPA's calculation of air emissions for this OES, including all
calculation inputs, can be found in 1,4-Dioxane Supplemental Information File: Environmental Releases
to Air (	023iY

Page 274 of 484


-------
7227

7228

7229

7230

7231

7232

7233

7234

7235

7236

7237

7238

7239

7240

7241

7242

7243

7244

7245

7246

7247

7248

7249

7250

7251

7252

7253

7254

7255

7256

7257

7258

7259

7260

7261

7262

7263

7264

7265

7266

7267

7268

7269

7270

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Film Cement

EPA estimated air emissions for facilities within the Film Cement OES using process information from
the Final Risk Evaluation for 1,4-Dioxane (	1020c). The underlying process information for

this assessment was rated "high" during EPA's systematic review process.

The process of using film cement involves applying the cement onto edges of photographic films by
hand using a small brush, then joining the pieces of film by pressing and heating to dry the cement.
Based on this process information, EPA assumes that the 1,4-dioxane within film cement is volatilized
to air during the drying process and that 1,4-dioxane residual within empty film cement bottles may also
be volatilized to air. EPA estimated these air releases for this OES as a range, using a film cement use
rate of 2.5 to 12 L/site-year and a concentration of 1,4-dioxane in the film cement of 45 to 50 percent,
from the process information in the Final Risk Evaluation for 1,4-Dioxane (	020c). These

releases may be to fugitive air or stack air depending on site-specific engineering controls.

EPA's calculation of air emissions for this OES, including all calculation inputs and assumptions, can be
found in 1,4-Dioxane Supplemental Information File: Environmental Releases to Air (	023i).

Spray Polyurethane Foam

EPA estimated air emissions for facilities within the spray polyurethane foam OES using the GS on
Application of Spray Polyurethane Foam Insulation (U.S. EPA. 2020c). The GS on the Application of
Spray Polyurethane Foam Insulation was rated "medium" during EPA's systematic review process.

The GS indicates that there are six release points:

1.	Releases to fugitive air for volatile chemicals during unloading of raw materials from transport
containers

2.	Releases to water, incineration, or landfill from cleaning or disposal of transport containers

3.	Releases to fugitive air for volatile chemicals during transport container cleaning

4.	Releases to incineration or landfill from spray polyurethane foam application equipment cleaning

5.	Releases to fugitive air for volatile chemicals during equipment cleaning

6.	Releases to landfill of scrap foam from trimming applied foam

Based on the GS, release points #1, 3, and 5 have the potential for air emissions. To estimate these
releases, EPA used the equations specified in the GS (	020c). Apart from weight fraction in

spray polyurethan foam, EPA did not find any data specific to 1,4-dioxane in this OES. Therefore, the
calculation of releases using this GS are for a "generic site," using the default input parameter values
from the GS. Specifically, EPA used the input parameter values that were presented in the original risk
evaluation for estimates of occupational exposure; see Appendix G of the Final Risk Evaluation for 1,4-
Dioxane (U.S. EPA. 2020c).

Using this methodology, EPA calculated the "typical" and "worst-case" air emissions for this OES.
These characterizations are based on the GS, which provides default "typical" and "worst-case" input
parameters for the release calculations. EPA's calculation of air emissions for this OES, including all
calculation inputs and assumptions, can be found in 1,4-Dioxane Supplemental Information File:
Environmental Releases to Air Q x \ \\ ,023i).

Page 275 of 484


-------
7271

7272

7273

7274

7275

7276

7277

7278

7279

7280

7281

7282

7283

7284

7285

7286

7287

7288

7289

7290

7291

7292

7293

7294

7295

7296

7297

7298

7299

7300

7301

7302

7303

7304

7305

7306

7307

7308

7309

7310

7311

7312

7313

7314

7315

7316

7317

7318

7319

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

3D Printing

Air emission data were not available for this OES and EPA did not find any information to model air
emissions for this OES using literature, GSs, or ESDs. EPA expects that industrial applications of this
OES to be accounted for in the Industrial Uses TRI data. Per the December 2020 Final Risk Evaluation
for 1,4-Dioxane (	320c). 3D printing ink containing 1,4-dioxane is used in research labs to

print biomedical products. Because the 2019 TRI data for the Industrial Uses OES include medicinal and
pharmaceutical manufacturing NAICS codes, medical research labs that conduct 3D printing with 1,4-
dioxane inks may be captured in that OES. Therefore, EPA grouped the air release assessment for 3D
Printing into that for Industrial Uses. However, there is uncertainty in whether 3D printing sites are truly
captured in the Industrial Uses TRI data.

Dry Film Lubricant

EPA estimated air emissions for facilities within the Dry Film Lubricant OES using process information
from the Final Risk Evaluation for 1,4-Dioxane (	'20c). The underlying process information

for this assessment was rated "high" during EPA's systematic review process.

The process for the production and use of dry film lubricant is described in the 2020 RE and is based on
information provided to EPA by the one known user. In summary, the process entails producing the
concentrated dry film lubricant by mixing 1,4-dioxane with other additives, followed by the dilution of
the concentrated dry film lubricant with additional 1,4-dioxane and finally the use of the dry film
lubricant. The use involves spray application onto substrates in a vented paint booth and the subsequent
curing in a vented oven and cleaning of the dried parts in a 1,4-dioxane bath (	20c). Based

on this process description, EPA assumes that 100 percent of the 1,4-dioxane in the applied dry film
lubricant is released to stack air from the paint booth and the oven. EPA estimated this release quantity
using batch parameters from the process description, including 5 percent 1,4-dioxane in the dry film
lubricant, 48 dry film lubricant applications per year, 0.5-pints of concentrated dry film lubricant, and
1.5-pints of pure 1,4-dioxane per application (U.S. EPA. 2020c).

EPA's calculation of air emissions for this OES, including all calculation inputs and assumptions, can be
found in 1,4-Dioxane Supplemental Information File: Environmental Releases to Air (]j_S IT \ 2023i).

Textile Dye

EPA used the OECD ESD on Textile Dyes (OE(	) to estimate water and land releases for this

OES; however, this ESD does not include approaches for estimating air releases. EPA did not find any
other GS/ESD, literature sources, or process information to model air releases for this OES. In addition,
EPA does not expect this OES to be sufficiently similar to other OES such that surrogate TRI data can
be used to estimate air emissions for this OES. Therefore, EPA was not able to estimate air releases for
these OESs.

Antifreeze

EPA did not find any directly applicable GS/ESD or literature sources for this OES; however, EPA
evaluated the potential for releases using the OECD ESD on Chemical Additives used in Automotive
Lubricants (OB 20) and the EPA MRD on Commercial Use of Automotive Detailing Products
(	'A. 2022b). The ESD and MRD were both rated "high" during EPA's systematic review

process.

For the use of antifreeze, EPA expects releases may occur from volatilizations of 1,4-dioxane during
unloading/ pouring antifreeze into vehicles, disposal, or cleaning of empty antifreeze containers, and
spent antifreeze. Both the ESD and MRD indicate that containers of automotive maintenance fluids are

Page 276 of 484


-------
7320

7321

7322

7323

7324

7325

7326

7327

7328

7329

7330

7331

7332

7333

7334

7335

7336

7337

7338

7339

7340

7341

7342

7343

7344

7345

7346

7347

7348

7349

7350

7351

7352

7353

7354

7355

7356

7357

7358

7359

7360

7361

7362

7363

7364

7365

7366

7367

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

typically small and are not rinsed, but rather disposed of as solid waste (1, c. < i1 \ J022U oU U
2020). Additionally, the ESD on Chemical Additives used in Automotive Lubricants indicates that spent
lubricants are disposed of via incineration, which EPA expects is similarly done for spent antifreeze
(OECD. 2020). Therefore, EPA expects the main source of air emissions to be from volatilizations of
1,4-dioxane during unloading/ pouring antifreeze into vehicles. EPA estimated this release using the
EPA/OAQPS AP-42 Loading Model and batch parameters from the ESD, MRD, and other sources.

Specifically, EPA used the consumer use rate of antifreeze (0.15 kg antifreeze/job) from the Final Risk
Evaluation for 1,4-Dioxane (U.S. EPA. 2020c) and scaled this value up to a commercial use rate based
on a range of the number of cars serviced at auto shops from the Near-Field/Far-Field Brake Model and
Automotive Detailing MRD (1 to 9 jobs/day). EPA used a range of concentrations of 1,4-dioxane in
antifreeze from the process description in Appendix F.4.2 and assumed antifreeze container sizes
ranging from 16 ounces to 5 gallons per the default container sizes in the MRD and ESD, respectively
(\ v i i \	t (i . '20). Using these batch parameters and the default parameters for the

EPA/OAQPS AP-42 Loading Model, EPA estimated low-end and high-end air emissions. EPA expects
these air emissions to be to fugitive air based on the use setting (e.g., outdoors, maintenance garages).

EPA's calculation of air emissions for this OES, including all calculation inputs and assumptions, can be
found in 1,4-Dioxane Supplemental Information File: Environmental Releases to Air (	023i).

Surface Cleaner

EPA did not find any directly applicable GS/ESD or literature sources for this OES; however, EPA
estimated air releases using the SHEDs-HT modeling conducted for the one case study location
(Liverpool OH) and the assumptions described herein. EPA expects that the main release points from the
use of surface cleaners are from

1.	Disposal of empty containers containing residual cleaning solution,

2.	Application of the cleaning solution, and

3.	Disposal of cleaning solution by rinsing or wiping.

Because EPA did not find any directly applicable GSs or ESDs, EPA used the Draft GS on Furnishing
Cleaning (	322a) to inform these releases due to the similarities in surface cleaning and

furnishing cleaning. The Draft GS on Furnishing Cleaning was rated "high" during EPA's systematic
review process. Per this Draft GS, empty containers may be rinsed out in sinks or disposed of without
rinsing, such that releases may be to wastewater or landfill; the GS uses the EPA/OPPT Small Container
Residual Model to estimate this release. Application losses are to fugitive air from spray application; the
GS uses literature data to estimate this release. Once applied, the cleaner may be rinsed off or wiped off
with rags or towels, such that releases may be to wastewater or landfill; the GS assumes 100 percent
release scenario, estimating this release by subtracting the upstream losses from the cleaner use rate
(	2a).

The SHEDs-HT modeling estimated wastewater discharges of 72 g of 1,4-dioxane per day for
commercial uses of surface cleaners containing 1,4-dioxane in Liverpool OH. EPA used this quantity
and the above release information and models from the Draft GS on Furnishing Cleaning to back-
calculate a 1,4-dioxane use rate. EPA then applied the loss fraction to fugitive air from release point #2
to estimate air releases for this OES. EPA's calculation of air releases for this OES, including all
calculation inputs and assumptions, can be found in 1,4-Dioxane Supplemental Information File:
Environmental Releases to Air Q x \ \\ ,023i).

Page 277 of 484


-------
7368

7369

7370

7371

7372

7373

7374

7375

7376

7377

7378

7379

7380

7381

7382

7383

7384

7385

7386

7387

7388

7389

7390

7391

7392

7393

7394

7395

7396

7397

7398

7399

7400

7401

7402

7403

7404

7405

7406

7407

7408

7409

7410

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Dish Soap

EPA did not find relevant or surrogate TRI data, literature sources, sufficient process information, nor
ESD or GS with air release estimation approaches to estimate air releases for these OESs. Therefore,
EPA was not able to estimate air releases for these OESs.

Dishwasher Detergent

EPA did not find relevant or surrogate TRI data, literature sources, sufficient process information, nor
ESD or GS with air release estimation approaches to estimate air releases for these OESs. Therefore,
EPA was not able to estimate air releases for these OESs.

Laundry Detergent

EPA estimated air emissions for facilities within the Laundry Detergent OES using the OECD ESD on
Industrial and Institutional Laundries (OI	) and Monte Carlo modeling. The ESD on

Industrial and Institutional Laundries was rated "medium" during EPA's systematic review process. The
use of Monte Carlo modeling allows for variation of calculation input parameters such that a distribution
of environmental releases can be calculated, from which EPA can estimate the 50th and 95th percentile
releases. An explanation of this modeling approach is included in Appendix E.12.

Paints and Floor Lacquer

EPA estimated air emissions for facilities within the Paints and Floor Lacquers OES using the OECD
ESD on Coating Application via Spray-Painting in the Automotive Refinishing Industry (OECD.
201 la). The ESD was rated "medium" during EPA's systematic review process.

As described in the process description in Appendix F.4.7, 1,4-dioxane was identified by a public
comment as present in automotive refinishing products, thereby allowing EPA to identify the above
ESD as the most applicable GS/ESD available. This ESD indicates that releases are expected from

1.	Releases to incineration or landfill from container cleaning/disposal,

2.	Releases to incineration or landfill from equipment cleaning,

3.	Releases to incineration or landfill from over sprayed coating that is captured by emission
controls, and

4.	Releases to stack air from over sprayed coating that is not captured by emission controls.

Based on the GS, release point #4 has the potential for air emissions. To estimate this release, EPA used
the equations specified in the ESD (OECD. 201 la). Apart from weight fraction in coatings (see
Appendix F.4.7), EPA did not find any data specific to 1,4-dioxane in this OES. Therefore, the
calculation of releases using this GS are for a "generic site," using the default input parameter values
from the ESD.

Using this methodology, EPA calculated the low-end and high-end air emissions for this OES, which are
expected to be to stack air per the ESD (	). The low- and high-end estimates are based on

the low- and high-end calculation input parameter defaults from the ESD. EPA's calculation of air
emissions for this OES, including all calculation inputs and assumptions, can be found in 1,4-Dioxane
Supplemental Information File: Environmental Releases to Air (	)23i).

Page 278 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

7411	Hydraulic Fracturing

7412	EPA estimated air emissions for facilities within the Hydraulic Fracturing OES using the Draft OECD

7413	ESD on Hydraulic Fracturing ( v «« \ . 022d) and Monte Carlo modeling. The Draft ESD on

7414	Hydraulic Fracturing was rated "high" during EPA's systematic review process. The use of Monte Carlo

7415	modeling allows for variation of calculation input parameters such that a distribution of environmental

7416	releases can be calculated, from which EPA can estimate the 50th and 95th percentile releases. An

7417	explanation of this modeling approach is included in Appendix E.13.

7418	E.5.3 Air Release Estimates Summary

7419	A summary of industrial and commercial air releases estimated using the above methods is presented in

7420	Table_Apx E-7 below.

Page 279 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

7421 Table Apx E-7 Summary of Daily Industrial and Commercial Air Release Estimates for 1,4-Dioxane

OES

Type of Air Release

Number of
Facilities with
Releases"

Estimated Daily Release
Range across Sites
(kg/site-day)

Estimated

Release
Frequency
Range
(days)''

Overall Data

Quality
Determination

Sources'

Min

Max

Manufacturing

Fugitive Air

1

2.62

250

Medium

TRI

Stack Air

1

0.0018

Medium

TRI

Import and
Repackaging

Fugitive Air

1

0

250

Medium

TRI

Stack Air

1

0.091

Medium

TRI

Industrial Uses

Fugitive Air

12

0

0.91

250

Medium

TRI

Stack Air

12

0

8.14

Medium

TRI

Functional Fluids
(Open-System)

Fugitive Air

2

0

0.009

247

Medium

TRI

Stack Air

2

0.19

1.38

Medium

TRI

Functional Fluids
(Closed-System)

All

Assessed as a part of Industrial Uses OES

N/A

N/A

Laboratory
Chemical

Fugitive Air or Stack
Air (-Unknown)

132

0.11 (typical)

0.41 (worst-
case)

250

High

GSd

Film Cement

Fugitive Air or Stack
Air (Unknown)

211

0.0046

0.025

250

High

Process
information®

Spray Foam
Application

Fugitive Air

1,553,559

0.0024
(typical)

0.012 (worst-
case)

3

Medium

GSf

Stack Air

1,553,559

0 (all air releases assessed to
fugitive)

Medium

GSf

Printing Inks (3D)

Fugitive Air, Stack
Air, and Land (all
types)

Assessed as a part of Industrial Uses OES

250

N/A

N/A

Dry Film Lubricant

Fugitive Air

8

0 (no fugitive releases per
process information)

48

High

Process
information®

Stack Air

8

0.75 (single value estimated
from process information)

High

Process
information®

Page 280 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Type of Air Release

Number of
Facilities with
Releases"

Estimated Daily Release
Range across Sites
(kg/site-day)

Estimated

Release
Frequency
Range
(days)''

Overall Data

Quality
Determination

Sources'





Min

Max



Disposal

Fugitive Air

15

0

0.91

250

Medium

TRI

Stack Air

15

0

0.91

Medium

TRI

Textile Dye

Fugitive Air and
Stack Air



Not assessed



31 to 295

Medium

ESDg and
Modeling''

Antifreeze

Fugitive Air and
Stack Air

84,383

7.26E-16

1.80E-07

250

High

Process

information® and
Modeling''

Surface Cleaner

Fugitive Air

Unknown

0.0071
(typical; for
all sites in
Liverpool
OH, case
study)

0.013 (worst
case; for all
sites in

Liverpool, OH,
case study)

250

High

SHEDS-HT,'
Process
information®
Modeling''

Dish Soap

Fugitive air and
stack air



Not assessed



250

N/A

N/A

Dishwasher

Fugitive air and



Not assessed



250

N/A

N/A

Detergent

stack air















Fugitive air

95,533

1.83E-10

6.52E-07



Medium

ESD! and
Modeling''



Fugitive air, stack

95,533

1.51 E— 10

0.00714



Medium

ESD! and

Laundry Detergent
(Institutional) -

air, or POTW

(unknown

partitioning)







250 to 365



Modeling''

liquid detergents

Land (unknown
landfill),
incineration, or
POTW (unknown
partitioning)

95,533

4.05E-12

3.95E-05



Medium

ESD! and
Modeling''

Page 281 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Type of Air Release

Number of
Facilities with
Releases"

Estimated Daily Release
Range across Sites
(kg/site-day)

Estimated

Release
Frequency
Range
(days)''

Overall Data

Quality
Determination

Sources'





Min

Max





Fugitive air

95,533

3.42E-12

2.77E-07



Medium

ESD! and
Modeling''



Stack air

95,533

1.40E-11

3.75E-06



Medium

ESD! and
Modeling''

Laundry Detergent
(Institutional) -
powder detergents

Fugitive air, stack
air, or POTW
(unknown
partitioning)

95,533

3.05E-08

2.10E-04

250 to 365

Medium

ESD! and
Modeling''



Land (unknown
landfill),
incineration, or
POTW (unknown
partitioning)

95,533

5.36E-08

0.0018



Medium

ESD! and
Modeling''



Fugitive air

2,453

6.25E-10

1.93E-06



Medium

ESD! and
Modeling''

Laundry Detergent
(Industrial) - liquid

Fugitive air, stack
air, or POTW
(unknown
partitioning)

2,453

5.48E-12

0.011

20 to 365

Medium

ESD! and
Modeling''

detergents

Land (unknown
landfill),
incineration, or
POTW (unknown
partitioning)

2,453

4.78E-12

1.46E-04



Medium

ESD! and
Modeling''

Page 282 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Type of Air Release

Number of
Facilities with
Releases"

Estimated Daily Release
Range across Sites
(kg/site-day)

Estimated

Release
Frequency
Range
(days)''

Overall Data

Quality
Determination

Sources'





Min

Max





Fugitive air

2,453

3.13E-13

1.47E-05



Medium

ESD! and
Modeling''



Stack air

2,453

1.68E-12

1.82E-04



Medium

ESD! and
Modeling''

Laundry Detergent
(Industrial) -
powder detergents

Fugitive air, stack
air, or POTW
(unknown
partitioning)

2,453

1.76E-09

0.0112

20 to 365

Medium

ESD! and
Modeling''



Land (unknown

2,453

2.92E-11

3.92E-04



Medium

ESD! and



landfill),











Modeling''



incineration, or















POTW (unknown
partitioning)













Paints and Floor

Stack air

33,648

4.68E-10

1.60E-06

250

Medium

ESD7

Lacquer















Polyethylene

Fugitive Air

13

0

1.57



Medium

TRI

Terephthalate
(PET) Byproduct

Stack Air

13

0.0049

13.8

250

Medium

TRI

Ethoxylation

Fugitive Air

8

0

7.4

250

Medium

TRI

Process Byproduct

Stack Air

8

0

32

Medium

TRI



Fugitive air

411

1.99E-07

5482



Medium

ESD^ and
Modeling''



Stack air

411

0 (all air releases assessed to



Medium

ESD" and

Hydraulic





fugitive)

1 to 72



Modeling''

Fracturing

Surface water,
incineration, or
landfill (unknown
partitioning)

411

3.61E-10

4.59

Medium

ESD'' and
Modeling''

Page 283 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Type of Air Release

Number of
Facilities with
Releases"

Estimated Daily Release
Range across Sites
(kg/site-day)

Min

Max

Estimated

Release
Frequency
Range
(days)''

Overall Data

Quality
Determination

Sources'

"Where available, EPA used 2013-2019 TRI (U.S. EPA. 2022g) and 2013-2019 DMR (U.S. EPA. 2022c) data to provide a basis to estimate the number of sites
using 1,4-dioxane within a COU.

* Where available. EPA used the December 2020 Final Risk Evaluation for 1.4-Dioxane (U.S. EPA. 20200). generic scenarios, and emission scenario documents
to provide a basis to estimate the number of release days of 1,4-dioxane within a COU.
c Narrative descriptions of all release estimate sources are provided in Appendix E.5.
d The generic scenario used for this COU is the GS on Use of Laboratory Chemicals (U.S. EPA. 202211).
e For this COU, EPA used process information, which is further described in Appendix E.5.2.

' The generic scenario used for this COU is the GS on Application of Spray PoK urethane Foam Insulation (U.S. EPA. 2018b).
g The emission scenario document used for this COU is the ESD on Textile Dyes (OECD. 2017).
h For this COU, EPA used various models and literature for model input parameters as described in Appendix E.5.2.

1 The emission scenario document used for this COU is the ESD on Industrial and Institutional Laundries (OECD. 201 lb).

1 The emission scenario document used for this COU is the ESD on Coating Application via Spray Painting in the Automotive Refinishing Industry (OECD.

2011a).

The emission scenario document used for this COU is the Draft ESD on Hydraulic Fracturing (U.S. EPA. 2022d).

' EPA used the down the drain water release estimates from the SHEDs-HT model for the Liverpool OH case study (see Section 2.1.1.2) to estimate air releases
by back calculating 1,4-dioxane use rates and applying loss fractions for air releases using literature and standard models described in Appendix E.5.2.

7422

Page 284 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

7423	E.5.4 Summary of Weight of the Scientific Evidence Conclusions in Air Release Estimates	

7424	Table Apx E-8 provides a summary of EPA's weight of the scientific evidence conclusions in its air release estimates for each of the

7425	Occupational Exposure Scenarios assessed. Detailed descriptions of non-OES specific strengths, limitations, assumptions, and uncertainties

7426	{e.g., general limitations for TRI, DMR, etc.) are provided in Appendix E.6.

7427

7428	Table Apx E-8 Summary of Weight of the Scientific Evidence Conclusions in Air Release Estimates by PES	

OES

Weight of the Scientific Evidence Conclusion in Release Estimates

Manufacturing

Air releases are assessed using reported releases from 2019 TRI. Factors that increase the strength of evidence for this OES
are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium overall data
quality determination, and consistency within the dataset (all reporters use the same or similar reporting forms). A strength
of TRI data is that TRI compiles the best readily available release data for all reporting facilities. Factors that decrease the
strength of the evidence for this OES include the low number of data points, lack of variability (only one year of data
used), uncertainty in the accuracy of reported releases, and the limitations in representativeness to all sites because TRI
may not capture all relevant sites. Additionally, EPA made assumptions on the number of operating days to estimate daily
releases. Based on this information, EPA has concluded that the weight of the scientific evidence for this assessment is
moderate to robust and provides a plausible estimate of releases in consideration of the strengths and limitations of
reasonably available data.

Import and Repackaging

Air releases are assessed using reported releases from 2019 TRI. Factors that increase the strength of evidence for this OES
are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium overall data
quality determination, and consistency within the dataset (all reporters use the same or similar reporting forms). A strength
of TRI data is that TRI compiles the best readily available release data for all reporting facilities. Factors that decrease the
strength of the evidence for this OES include the low number of data points, lack of variability (only one year of data
used), uncertainty in the accuracy of reported releases, uncertainty in EPA's use of Form A submissions, and the
limitations in representativeness to all sites because TRI may not capture all relevant sites. Some facilities within this OES
reported to TRI using a Form A, which does not include any details on chemical release quantities. When a facility has
submitted a Form A, there is no way to discern the quantity released. Therefore, where facilities reported to TRI with a
Form A, EPA used the Form A threshold for total releases of 500 lb/year for each release media; however, there is
uncertainty in this because the actual release quantity is unknown. Furthermore, the threshold represents an upper limit on
total releases from the facility; therefore, assessing releases at the threshold value may overestimate actual releases from
the facility. Additionally, EPA made assumptions on the number of operating days to estimate daily releases. Based on this
information, EPA has concluded that the weight of the scientific evidence for this assessment is moderate to robust and
provides a plausible estimate of releases in consideration of the strengths and limitations of reasonably available data.

Industrial Uses

Air releases are assessed using reported releases from 2019 TRI. Factors that increase the strength of evidence for this OES
are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium overall data
quality determination, and consistency within the dataset (all reporters use the same or similar reporting forms). A strength
of TRI data is that TRI compiles the best readily available release data for all reporting facilities. Factors that decrease the

Page 285 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates



strength of the evidence for this OES include a lack of variability (only one year of data used), uncertainty in the accuracy
of reported releases, uncertainty in EPA's use of Form A submissions, and the limitations in representativeness to all sites
because TRI may not capture all relevant sites. Some facilities within this OES reported to TRI using a Form A, which
does not include any details on chemical release quantities. When a facility has submitted a Form A, there is no way to
discern the quantity released. Therefore, where facilities reported to TRI with a Form A, EPA used the Form A threshold
for total releases of 500 lb/year for each release media; however, there is uncertainty in this because the actual release
quantity is unknown. Furthermore, the threshold represents an upper limit on total releases from the facility; therefore,
assessing releases at the threshold value may overestimate actual releases from the facility. Based on this information, EPA
has concluded that the weight of the scientific evidence for this assessment is moderate to robust and provides a plausible
estimate of releases in consideration of the strengths and limitations of reasonably available data.

Functional Fluids (Open-
System)

Air releases are assessed using reported releases from 2019 TRI. Factors that increase the strength of evidence for this OES
are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium overall data
quality determination, and consistency within the dataset (all reporters use the same or similar reporting forms). A strength
of TRI data is that TRI compiles the best readily available release data for all reporting facilities. Factors that decrease the
strength of the evidence for this OES include the low number of data points (only two reporting sites), lack of variability
(only one year of data used), uncertainty in the accuracy of reported releases, and the limitations in representativeness to all
sites because TRI may not capture all relevant sites. Additionally, EPA made assumptions on the number of operating days
to estimate daily releases. Based on this information, EPA has concluded that the weight of the scientific evidence for this
assessment is moderate to robust and provides a plausible estimate of releases in consideration of the strengths and
limitations of reasonably available data.

Functional Fluids (Closed-
System)

No data was available to estimate releases for this OES. For the air release assessment, EPA grouped this OES with the
Industrial Uses OES because the sources of release are expected to be similar between these OESs. Factors that increase
the strength of evidence for this OES are that TRI has a medium overall data quality determination and consistency within
the dataset (all reporters use the same or similar reporting forms). Factors that decrease the strength of evidence for this
OES are that the Industrial Releases OES release data are use as surrogate for this OES, uncertainty in the accuracy of
reported releases, limitations in representativeness to all sites because TRI may not capture all relevant sites, and lack of
variability (only one year of data used). Refer to the Industrial Uses OES discussion for additional discussion. Based on
this information, EPA has concluded that the weight of the scientific evidence for this assessment is slight and provides a
plausible estimate of releases in consideration of the strengths and limitations of reasonably available data.

Laboratory Chemicals

Air releases are assessed using the Draft GS on Use of Laboratory Chemicals. Factors that increase the strength of
evidence for this OES are that the release estimates are directly relevant to the OES (as opposed to surrogate), the Draft GS
on Use of Laboratory Chemicals has a high overall data quality determination, and the low level of uncertainty in the data.
Factors that decrease the strength of the evidence for this OES include the that the GS has not been peer-reviewed,
uncertainty in the representativeness of the GS towards all sites in this OES, and a lack of variability in the analysis.
Specifically, because the default values in the ESD are generic, there is uncertainty in the representativeness of generic site
estimates of actual releases from real-world sites that use 1,4-dioxane. Another uncertainty is lack of consideration for

Page 286 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates



release controls. The ESD assumes that all activities occur without any release controls. Actual releases may be less than
estimated if facilities utilize pollution control methods, contributing to uncertainty. Based on this information, EPA has
concluded that the weight of the scientific evidence for this assessment is slight to moderate and provides a plausible
estimate of releases in consideration of the strengths and limitations of reasonably available data.

Film Cement

Air releases are assessed using process information from the Final Risk Evaluation for 1,4-Dioxane. Factors that increase
the strength of evidence for this OES are that the release estimates are directly relevant to the OES (as opposed to
surrogate), the underlying data sources for the process information have a high overall data quality determination, and the
low level of uncertainty in the data because the process information comes directly from actual users of 1,4-dioxane in film
cement. Factors that decrease the strength of the evidence for this OES include uncertainty in the representativeness of
evidence to all sites in this OES and a lack of variability. Specifically, the process information for the production and use
of film cement is based on information from three use sites, one from Australia and two from the U.S. Based on this
information, EPA has concluded that the weight of the scientific evidence for this assessment is slight to moderate and
provides a plausible estimate of releases in consideration of the strengths and limitations of reasonably available data.

Spray Foam Application

Air releases are assessed using the GS on Application of Spray Polyurethane Foam Insulation. Factors that increase the
strength of evidence for this OES are that the release estimates are directly relevant to the OES (as opposed to surrogate),
the underlying data sources for the process information have a medium overall data quality determination, and the low
level of uncertainty in the data. Factors that decrease the strength of the evidence for this OES include uncertainty in the
representativeness of the GS to all sites since it is generic and not specific to sites that use 1,4-dioxane and a lack of
variability. Based on this information, EPA has concluded that the weight of the scientific evidence for this assessment is
slight to moderate and provides a plausible estimate of releases in consideration of the strengths and limitations of
reasonably available data.

Printing Inks (3D)

No data was available to estimate releases for this OES. For the air release assessment, EPA grouped this OES with the
Industrial Uses OES because the sources of release are expected to be similar between these OESs. Factors that increase
the strength of evidence for this OES are that TRI has a medium overall data quality determination and consistency within
the dataset (all reporters use the same or similar reporting forms). Factors that decrease the strength of evidence for this
OES are that the Industrial Releases OES release data are use as surrogate for this OES, uncertainty in the accuracy of
reported releases, limitations in representativeness to all sites because TRI may not capture all relevant sites or smaller
commercial 3D printing uses, and lack of variability (only one year of data used). Refer to the Industrial Uses OES
discussion for additional discussion. Based on this information, EPA has concluded that the weight of the scientific
evidence for this assessment is slight and provides a plausible estimate of releases in consideration of the strengths and
limitations of reasonably available data.

Dry Film Lubricant

Air releases are assessed using process information from the Final Risk Evaluation for 1,4-Dioxane. Factors that increase
the strength of evidence for this OES are that the release estimates are directly relevant to the OES (as opposed to
surrogate), that the underlying data sources for the process information have a high overall data quality determination, and
a low level of uncertainty in the data because the process information comes directly from an actual user of 1,4-dioxane in

Page 287 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates



dry film lubricants. Factors that decrease the strength of the evidence for this OES include uncertainty in the
representativeness of evidence to all sites and a lack of variability. Based on this information, EPA has concluded that the
weight of the scientific evidence for this assessment is slight to moderate and provides a plausible estimate of releases in
consideration of the strengths and limitations of reasonably available data.

Disposal

Air releases are assessed using reported discharges from 2019 TRI. Factors that increase the strength of evidence for this
OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium overall data
quality determination, high number of data points, and consistency within the dataset (all reporters use the same or similar
reporting forms). A strength of TRI data is that TRI compiles the best readily available release data for all reporting
facilities. Factors that decrease the strength of the evidence for this OES include lack of variability (only one year of data
used), uncertainty in the accuracy of reported releases, uncertainty in EPA's use of Form A submissions, and the
limitations in representativeness to all sites because TRI may not capture all relevant sites. Some facilities within this OES
reported to TRI using a Form A, which does not include any details on chemical release quantities. When a facility has
submitted a Form A, there is no way to discern the quantity released. Therefore, where facilities reported to TRI with a
Form A, EPA used the Form A threshold for total releases of 500 lb/year for each release media; however, there is
uncertainty in this because the actual release quantity is unknown. Furthermore, the threshold represents an upper limit on
total releases from the facility; therefore, assessing releases at the threshold value may overestimate actual releases from
the facility. Additionally, uncertainty is introduced from EPA's assumptions on the number of operating days to estimate
daily releases. Based on this information, EPA has concluded that the weight of the scientific evidence for this assessment
is moderate to robust and provides a plausible estimate of releases in consideration of the strengths and limitations of
reasonably available data.

Textile Dye

EPA used the ESD on Textile Dyes to estimate land and water releases; however, this ESD does not include approaches for
estimating air releases. EPA did not find any other GS/ESD, literature sources, or process information to model air releases
for this OES. Furthermore, EPA does not expect this OES to be sufficiently similar to other OES such that surrogate TRI
data can be used to estimate air emissions for this OES. Therefore, EPA was not able to estimate air releases for this OES
and concluded that the weight of the scientific evidence is indeterminant.

Antifreeze

Air releases are assessed using the OECD ESD on Chemical Additives used in Automotive Lubricants, the EPA MRD on
Commercial Use of Automotive Detailing Products, and EPA/OPPT models. Factors that increase the strength of evidence
for this OES are that the ESD and MRD used have high overall data quality determinations and consistency within the
sources used. Factors that decrease the strength of the evidence for this OES include that the ESD and MRD are not
directly applicable to antifreeze uses (used as surrogate), uncertainty in the representativeness of the ESD and MRD to all
sites and sites that specifically use 1,4-dioxane since these documents contain generic values, and a lack of variability.
Additionally, EPA scaled up a consumer antifreeze use rate to a commercial use rate based on information in the ESD and
MRD, which increases uncertainty. Based on this information, EPA has concluded that the weight of the scientific
evidence for this assessment is slight to moderate and provides a plausible estimate of releases in consideration of the
strengths and limitations of reasonably available data.

Page 288 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates

Surface Cleaner

Air releases are assessed using the SHEDS-HT model and the Draft GS on Furnishing Cleaning. To estimate air releases,
EPA used loss fractions for water releases from the GS and the modeled water release from SHEDS-HT to back-calculate a
1,4-dioxane use rate. EPA then applied loss fractions for air releases from the GS to estimate air releases for this OES.
Factors that increase the strength of evidence for this OES include that the release estimates are directly relevant to the
OES (as opposed to surrogate), that the Draft GS used has a high overall data quality determination, and variability in the
model input parameters. Factors that decrease the strength of the evidence for this OES include uncertainty in the
representativeness to all sites because the SHEDS-HT estimate is based on one case study for Liverpool, OH and because
the estimate is not site-specific (the release estimate is a total for all sites in Liverpool, OH). Based on this information,
EPA has concluded that the weight of the scientific evidence for this assessment is slight and provides a plausible estimate
of releases in consideration of the strengths and limitations of reasonably available data.

Dish Soap

EPA did not find relevant or surrogate TRI data, literature sources, sufficient process information, nor ESD or GS with air
release estimation approaches to estimate air releases for these OESs. Therefore, EPA was not able to estimate air releases
for this OES and concluded that the weight of the scientific evidence is indeterminant.

Dishwasher Detergent

EPA did not find relevant or surrogate TRI data, literature sources, sufficient process information, nor ESD or GS with air
release estimation approaches to estimate air releases for these OESs. Therefore, EPA was not able to estimate air releases
for this OES and concluded that the weight of the scientific evidence is indeterminant.

Laundry Detergent

Air releases are assessed using Monte Carlo modeling with information from the ESD on Industrial and Institutional
Laundries. Factors that increase the strength of evidence for this OES are that the release estimates are directly relevant to
the OES (as opposed to surrogate), that the ESD on Industrial and Institutional Laundries has a medium overall data
quality determination and was peer reviewed, high number of data points (simulation runs), consistency within the dataset,
and full distributions of input parameters. The Monte Carlo modeling accounts for the entire distribution of input
parameters, calculating a distribution of potential release values that represents a larger proportion of sites than a discrete
value. Additionally, EPA was able to separately estimate releases for industrial and institutional laundry settings. Factors
that decrease the strength of the evidence for this OES include uncertainties and limitations in the representativeness of the
estimates for sites that specifically use 1,4-dioxane because the default values in the ESD are generic. Another uncertainty
is lack of consideration for release controls. The ESD assumes that all activities occur without any release controls. Actual
releases may be less than estimated if facilities utilize pollution control methods. Based on this information, EPA has
concluded that the weight of the scientific evidence for this assessment is moderate and provides a plausible estimate of
releases in consideration of the strengths and limitations of reasonably available data.

Paint and Floor Lacquer

Air releases are assessed using OECD ESD on Coating Application via Spray-Painting in the Automotive Refinishing
Industry. Factors that increase the strength of evidence for this OES are that the release estimates are directly relevant to
the OES (as opposed to surrogate), the ESD has a medium overall data quality determination, consistency within the
sources used, and a low amount of uncertainties. Factors that decrease the strength of the evidence for this OES include a
lack of variability and uncertainty in the representativeness of the ESD to all sites and sites that specifically use 1,4-
dioxane since the ESD is generic. Based on this information, EPA has concluded that the weight of the scientific evidence

Page 289 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Weight of the Scientific Evidence Conclusion in Release Estimates



for this assessment is slight to moderate and provides a plausible estimate of releases in consideration of the strengths and
limitations of reasonably available data.

Polyethylene Terephthalate
(PET) Byproduct

Air releases are assessed using reported discharges from 2019 TRI. Factors that increase the strength of evidence for this
OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium overall data
quality determination, consistency within the dataset (all reporters use the same or similar reporting forms), and
consistency with the emission data from the related life cycle analysis discussed in Appendix E.6. A strength of TRI data is
that TRI compiles the best readily available release data for all reporting facilities. Factors that decrease the strength of the
evidence for this OES include uncertainty in the accuracy of reported releases, lack of variability (only one year of data
used), and the limitations in representativeness to all sites because TRI may not capture all relevant sites. Additionally,
EPA made assumptions on the number of operating days to estimate daily releases, which introduces additional
uncertainty. Based on this information, EPA has concluded that the weight of the scientific evidence for this assessment is
moderate to robust and provides a plausible estimate of releases in consideration of the strengths and limitations of
reasonably available data.

Ethoxylation Process
Byproduct

Air releases are assessed using reported discharges from 2019 TRI. Factors that increase the strength of evidence for this
OES are that the release data are directly relevant to the OES (as opposed to surrogate), that TRI has a medium overall data
quality determination, and consistency within the dataset (all reporters use the same or similar reporting forms). A strength
of TRI data is that TRI compiles the best readily available release data for all reporting facilities. Factors that decrease the
strength of the evidence for this OES include the uncertainty in the accuracy of reported releases, lack of variability (only
one year of data used), and the limitations in representativeness to all sites because TRI may not capture all relevant sites.
Additionally, EPA made assumptions on the number of operating days to estimate daily releases, which introduces
additional uncertainty. Based on this information, EPA has concluded that the weight of the scientific evidence for this
assessment is moderate to robust and provides a plausible estimate of releases in consideration of the strengths and
limitations of reasonably available data.

Hydraulic Fracturing

Air releases are assessed using Monte Carlo modeling with information from the Draft ESD on Hydraulic Fracturing and
FracFocus 3.0. Factors that increase the strength of evidence for this OES are that the release estimates are directly relevant
to the OES (as opposed to surrogate), that the Draft ESD on Hydraulic Fracturing and FracFocus 3.0 have medium overall
data quality determinations, high number of data points (simulation runs), consistency within the dataset, and full
distributions of input parameters. The Monte Carlo modeling accounts for the entire distribution of input parameters,
calculating a distribution of potential release values that represents a larger proportion of sites than a discrete value. Factors
that decrease the strength of the evidence for this OES include that the Draft ESD has not been peer reviewed and the
uncertainties and limitations in the representativeness of the estimates for sites that specifically use 1,4-dioxane because
the default values from the Draft ESD on Hydraulic Fracturing. Another uncertainty is lack of consideration for release
controls. The ESD assumes that all activities occur without any release controls. Actual releases may be less than estimated
if facilities utilize pollution control methods, contributing to uncertainty. Based on this information, EPA has concluded
that the weight of the scientific evidence for this assessment is moderate to robust and provides a plausible estimate of
releases in consideration of the strengths and limitations of reasonably available data.

7429

Page 290 of 484


-------
7430

7431

7432

7433

7434

7435

7436

7437

7438

7439

7440

7441

7442

7443

7444

7445

7446

7447

7448

7449

7450

7451

7452

7453

7454

7455

7456

7457

7458

7459

7460

7461

7462

7463

7464

7465

7466

7467

7468

7469

7470

7471

7472

7473

7474

7475

7476

7477

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

E,6 Comparison to PET Life Cycle Analysis

EPA identified a relevant life cycle analysis (LCA) published by the National Association for PET
Container Resources (NAPCOR) that included 1,4-dioxane emission factors from PET resin production
(Franklin Associates. 2020). EPA did not use these emission factors to estimate releases in this Risk
Evaluation because there were site-specific releases reported in TRI and DMR. This LCA only provided
generic emission factors for air and surface water releases aggregated across seven unspecified sites; the
LCA did not provide 1,4-dioxane emission factors for land releases. The emission factors in the LCA
were reported by three producers (seven sites) that account for 50% of the 2015 U.S. PET production in
a survey, and the basis of the emission factors is not provided. However, the survey states that the
release data is primary data (i.e., the data were provided by directly by the surveyed PET producers). As
opposed to conventional emission factors, the report only provides the order of magnitude of the average
amount of 1,4-dioxane released per amount of PET produced. Discrete, site-specific emission factors are
not provided. As a result, the variability of 1,4-dioxane releases from site to site is unknown. EPA
prefers the use of site-specific release data as opposed to generic emission factors. Therefore, a
comparison between total annual air and water releases from the LCA and from the TRI and DMR data
used in this Risk Evaluation is provided below for context.

The LCA estimated that 4.7 million tons of PET capacity was available in 2015 in North America
(Franklin Associates. 2020). To obtain total annual air and water release estimates from the LCA, EPA
multiplied this production volume by the reported 1,4-dioxane emission factors of 0.001 kg 1,4-dioxane
emitted per 1000 kg PET for air releases, and 0.01 kg 1,4-dioxane emitted per 1000 kg PET for surface
water releases. To obtain the total annual air and water releases from the TRI and DMR used in this Risk
Evaluation, EPA summed all reported annual site-specific air emissions and surface water discharges
that were mapped to the "PET Manufacturing" OES (see Appendix E.3 and E.5 for additional
information on the use of TRI and DMR). The total annual releases from the LCA and from TRI and
DMR is compiled in Table Apx E-9. EPA did this comparison with 2019 TRI/DMR because EPA's
Risk Evaluation largely uses 2019 data, as well as 2015 TRI/DMR data because the releases estimated
with the LCA data are based on 2015 PET manufacturing data.

For air emissions, the LCA estimate and EPA's estimates from the 2019 and 2015 TRI are comparable,
being within an order of magnitude. Differences in the estimates likely arise since EPA's analysis
accounted for emissions from 13 PET manufacturing facilities compared to the seven facilities in the
LCA. Additionally, the LCA is an aggregate of releases across sites whereas EPA's analysis accounts
for variability by using data from individual sites.

For surface water discharges, the LCA estimate and EPA's estimates from the 2019 TRI and DMR show
a larger discrepancy, with EPA's estimate being two orders of magnitude larger than the LCA estimate.
However, over 2.51 million kg of the approximately 2.53 million kg (99.2%) of surface water discharges
in EPA's estimate comes from a single facility's 2019 DMR report (DAK Americas LLC, Columbia).
The other 12 facilities with non-zero surface water discharges in EPA's analysis had site-specific annual
discharges ranging between 5.37 and 8,922 kgs for 2019. EPA reviewed the DAK Americas LLC,
Columbia DMR reports from other years for comparison, which indicated approximately 14,000 kg of
1,4-dioxane were discharged in 2022, 8,800 kg in 2021, 6.8 million kg in 2020, and 2,300 kg in 2018.
DAK Americas LLC, Columbia did not include 1,4-dioxane in their DMRs in 2016 or 2017 (the two
earliest reporting years EPA looked at for this analysis). It is unclear why DAK Americas LLC,
Columbia's discharges were significantly higher in 2019 and 2020 or why these discharges were
different than other PET manufacturers in EPA's analysis. However, it is more likely that the facilities
analyzed in the LCA were more similar to the other PET manufacturing facilities in EPA's analysis,

Page 291 of 484


-------
7478

7479

7480

7481

7482

7483

7484

7485

7486

7487

7488

7489

7490

7491

7492

7493

7494

7495

7496

7497

7498

7499

7500

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

with the total surface water emissions estimated from the LCA data being within one order of magnitude
of EPA's estimate when excluding the discharges from the DAK Americas LLC, Columbia facility, as
shown in Table Apx E-9. The LCA estimate and EPA's estimates from the 2015 TRI are comparable.

As indicated above, estimates from the LCA analysis and EPA's analysis with TRI and DMR data may
differ for several reasons. First, the LCA provides a single emission factor aggregated from data across
seven sites, reducing the impact of site-specific variability in releases. Whereas EPA's analysis uses site-
specific release data from 13 sites (for air emissions) and 19 sites (for surface water discharges, six of
which reported zero surface water discharges). EPA also does not have access to site identities, or the
underlying data/methodologies used to estimate emission factors in the LCA, which limits EPA's ability
to do a direct site-to-site comparison of results between the two analyses. Additionally, the LCA study
states that some emissions are reported only by the order of magnitude of the average to protect the
confidentiality of individual companies, introducing further uncertainty in the emission factors presented
in the study. Lastly, the LCA data is from 2015 whereas EPA used data from 2019.

Table Apx E-9. Comparison of TRI/DMR Release Data to LCA Study for PET Byproduct

Data Source

Total Release for All Sites
(kg/vr)

Air emissions

EPA Estimate in this Risk Evaluation -

Based on 2019 TRI (including DAK Americas LLC, Columbia)

10,695

EPA Estimate in this Risk Evaluation -

Based on 2015 TRI (including DAK Americas LLC, Columbia)

12,407

LCA Estimate (Franklin Associates. 2020) - Based on 2015 data

4,264

Surface water discharges

EPA Estimate in this Risk Evaluation -

Based on 2019 DMR and TRI (including DAK Americas LLC,

Columbia)

2,531,730

EPA Estimate in this Risk Evaluation -

Based on 2019 DMR and TRI (excluding DAK Americas LLC,

Columbia)

19,296

EPA Estimate in this Risk Evaluation -
Based on 2015 DMR and TRI

20,511

LCA Estimate (Franklin Associates. 2020) - Based on 2015 data

42,638

E.7 Detailed Strengths, Limitations, Assumptions and Key Sources of
Uncertainties for the Environmental Release Assessment

This section includes detailed strengths, limitations, assumptions, and uncertainties associated with
EPA's approaches for estimating air, water, and land releases in this supplemental risk evaluation. This
section is intended to supplement the summary of strengths, limitations, assumptions, and uncertainties
discussed in Section 2.2.1.3 with additional details.

Page 292 of 484


-------
7501

7502

7503

7504

7505

7506

7507

7508

7509

7510

7511

7512

7513

7514

7515

7516

7517

7518

7519

7520

7521

7522

7523

7524

7525

7526

7527

7528

7529

7530

7531

7532

7533

7534

7535

7536

7537

7538

7539

7540

7541

7542

7543

7544

7545

7546

7547

7548

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Facilities Not Captured by TRI or DMR Reporting

When using TRI data to analyze chemical releases, it is important to acknowledge that TRI reporting
does not include all releases of the chemical and therefore, the number of sites for a given OES may be
underestimated. For each OES that had TRI or DMR data, the analysis of releases for those OES was
limited to the facilities that reported releases to TRI and DMR. Therefore, it is uncertain the extent to
which sites not captured in these databases have air, water, or land releases of 1,4-dioxane and what the
exact media of release for those releases would be (e.g., stack vs. fugitive air, surface water vs POTW,
RCRA or another type of landfill). To the extent additional sites are not captured, releases may be
underestimated; however, the magnitude of this underestimation is unknown. TRI data do not include

•	Releases from any facility that used the chemical in quantities below the applicable annual
chemical activity threshold (e.g., 25,000 lb manufactured or processed, or 10,000 lb otherwise
used, for most chemicals);

•	Releases from any facility that is not in a TRI covered sector; and

•	Releases from any facility that does not meet the TRI employment threshold of greater than 10
full-time employee equivalents (20,000 labor hours) for the year.

DMR Release Data

For facilities that reported having zero pollutant loads to DMR, the ECHO Pollutant Loading Tool
Advanced Search applies a hybrid method to analyze nondetects. The EZ Search Load Module uses a
combination of setting nondetects equal to zero and as one half the detection limit to calculate the annual
pollutant loadings. Specifically, if the pollutant was measured as nondetect for all monitoring periods in
a reporting year, then the EZ Search Load Module sets the annual pollutant load to zero. If the pollutant
was detected for at least one monitoring period in a reporting year, then the EZ Search Load Module
calculates the annual pollutant load by setting the nondetects equal to one half the detection limit. This
method could cause overestimation or underestimation of annual and daily pollutant loads.

TRI Release Data

EPCRA section 313 states that facilities may estimate their release quantities using "readily available
data," including monitoring data, collected for other purposes. When data are not readily available,
EPCRA section 313 states that "reasonable estimates" may be used. The facility is not required to
monitor or measure the quantities, concentration, or frequency of any toxic chemical release for TRI
reporting. TRI guidance states that not using readily available information, such as relevant monitoring
data collected for compliance with other regulations, could result in enforcement and penalties.

For each release quantity reported, TRI facilities select a "Basis of estimate" code indicating the
principal method used to determine the amount of the release. TRI provides six basis of estimate codes
to choose from: continuous monitoring, periodic monitoring, mass balance, published emissions factors,
site-specific emissions factors, or engineering calculations/best engineering judgment. In facilities where
a chemical is used in multiple operations, the facility may use a combination of methods to calculate the
release reported. In such cases, TRI instructs the facility to enter the basis of estimate code of the
method that applies to the largest portion of the release quantity. Additional details on the basis of
estimate, such as any calculations and underlying assumptions, are not reported. Depending on the
inputs and/or monitoring methods used by each facility, any of the methods used to estimate releases
may over or underestimate releases. The magnitude of this uncertainty is unknown.

For any release quantity that is less than 1,000 lb, facilities may report either the estimated quantity or a
range code. The 1,000-pound limit for range code reporting applies to each type of release reported to
TRI - fugitive air emissions, stack air emissions, water discharges, each type of land disposal, and each

Page 293 of 484


-------
7549

7550

7551

7552

7553

7554

7555

7556

7557

7558

7559

7560

7561

7562

7563

7564

7565

7566

7567

7568

7569

7570

7571

7572

7573

7574

7575

7576

7577

7578

7579

7580

7581

7582

7583

7584

7585

7586

7587

7588

7589

7590

7591

7592

7593

7594

7595

7596

7597

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

type of off-site transfer. There are three TRI range codes: 1 to 10; 11 to 499; and 500 to 999 lb. TRI data
tools display the approximate midpoint of the range {i.e., 5, 250, or 750 lb). Using this midpoint value
may be either an over or an underestimate of the true value, depending where on the range the true value
lies. Although analyses using data that was reported as a range code may add uncertainty, it is not clear
that the uncertainty associated with a range code is greater than that associated with any other estimated
release value. Range code reporting is not permitted for chemicals of special concern.

TRI guidance states that release estimates need not be reported to more than two significant figures.
However, the guidance also states that facilities should report release quantities at a level of precision
supported by the accuracy of the underlying data and the estimation techniques on which the estimate
was based. If a facility's release calculations support reporting an amount that is more precise than two
significant digits, then the facility should report that more precise amount. The facility makes the
determination of the accuracy of their estimate and the appropriate significant digits to use.

For chemicals that meet certain criteria, facilities have the option of submitting a TRI Form A
Certification Statement instead of a TRI Form R. The Form A does not include any details on the
chemical release or waste management quantities. The criteria for a Form A are that during the reporting
year, the chemical (1) did not exceed 500 lb for the total annual reportable amount (including the sum of
on- and off-site quantities released, treated, recycled, and used for energy recovery); (2) amounts
manufactured, processed, or otherwise used do not exceed 1 million lb; and (3) the chemical is not a
chemical of special concern. When conducting analyses of chemical releases and a facility has submitted
a Form A for the chemical, there is no way to discern the quantity released to each medium or even if
there were any releases. Where facilities reported to TRI with a Form A, EPA used the Form A
threshold for total releases of 500 lb/year for each release media {e.g., fugitive air, stack air, surface
water, POTW, non-POTW WWT, RCRA landfill, other landfill). EPA used the entire 500 lb/year for
each release media; however, since this threshold is for total site releases, these 500 lb/year are to only
one of these media at a time (since assessing 500 lb/year for all media at once would double count the
releases and exceed the total release threshold for Form A). Furthermore, the threshold represents an
upper limit on total releases to all environmental media from the facility; therefore, assessing releases at
the threshold value likely overestimates actual releases from the facility.

Mapping TRI and DMR Facilities to OES

EPA used a crosswalk between TRI uses/sub-uses and CDR Industrial Function Category (IFC) codes
(see Appendix E.9) along with a mapping of CDR IFC codes to OES to assign the OES for each facility
that reported to TRI. However, there are limitations to this approach. For example, this approach may
result in the mapping of multiple OES for one facility. Additionally, there are limitations to the TRI -
CDR crosswalk. For example, a TRI use/sub-use may encompass multiple uses that are not captured in
the crosswalked CDR IFC codes. In these instances, EPA determined the primary OES using the NAICS
codes that the facility reported in TRI, information on types of products that the facility manufactures
found from internet searches, and which OES is most likely to result in release. OES are related to
COUs as defined in the crosswalk in Table 2-1 and Appendix D.2.This approach requires some level of
engineering judgment to determine which OES is the most applicable to the facility, which introduces
uncertainty in the OES mapping. Additionally, this approach assumes only one OES is applicable to the
facility, which may be incorrect if the facility uses 1,4-dioxane for multiple purposes. If facilities were
categorized under a different OES, the annual releases for each site would remain unchanged; however,
average daily releases may change depending on the release days expected for the different OES.

Additional uncertainty is present in the OES mapping for TRI sites that reported using a Form A and
DMR sites because there is no reported use/sub-use information. EPA used a similar procedure as

Page 294 of 484


-------
7598

7599

7600

7601

7602

7603

7604

7605

7606

7607

7608

7609

7610

7611

7612

7613

7614

7615

7616

7617

7618

7619

7620

7621

7622

7623

7624

7625

7626

7627

7628

7629

7630

7631

7632

7633

7634

7635

7636

7637

7638

7639

7640

7641

7642

7643

7644

7645

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

described above to map these sites to an OES, involving the use of NAICS and Standard Industrial
Classification (SIC) codes reported to TRI and DMR, internet searches on the types of products made at
the facility, and which OES is most likely to result in release. Since this approach involves engineering
judgment to determine which OES is the most applicable to the facility, there is uncertainty in the OES
mapping.

There is also uncertainty in the NAICS codes and SIC codes reported in TRI and DMR. TRI facilities
enter the facility's primary NAICS code indicating the primary economic activity at the facility.
Facilities can also enter secondary NAICS codes. When using TRI chemical release data for a facility
that also reported secondary NAICS codes, there may be uncertainty as to which NAICS is associated
with the use of the chemical. Additionally, NAICS codes and SIC codes are reported for the facility as a
whole and are not chemical specific.

Estimating Daily Releases from Annual TRI and DMR Release Data

Facilities reporting to TRI and DMR only report annual releases; to assess daily air and water releases,
EPA estimated the release days and averaged the annual releases over these days. There is some
uncertainty that all facilities for a given OES operate for the assumed duration; therefore, the average
daily release may be higher if sites have fewer release days or lower if they have greater release days.
Furthermore, chemical concentrations in air emissions and wastewater streams at each facility may vary
from day to day such that on any given day the actual daily releases may be higher or lower than the
estimated average daily discharge. Thus, this approach minimizes spikes and drops in emissions and
discharges from day to day.

Pollutant Load values of "0" in the ECHO Pollutant Loading Tool Advanced Search output file are by
default reported as below the detection limit in the monitoring reports used by the Loading Calculator
Module.

EPA did not estimate daily land releases due to the high level of uncertainty in the number of release
days associated with land releases; instead, EPA estimated annual land releases.

Representativeness of TRI and DMR for an OES as a Whole

The representativeness of TRI and DMR data for an OES as a whole is dependent on (1) the extent to
which these reporting mechanisms capture all potential sites within the OES and, (2) the extent to which
the release quantities provided by reporting sites reflect releases from non-reporting sites.

For some OES, the total number of sites was determined from TRI. For these OESs, there is uncertainty
in if there may be additional sites using the chemical within the OES that did not report to TRI (e.g., due
to being below reporting thresholds). For some OES, such as manufacturing and other OES involving
larger industrial sites, TRI is more likely to capture the majority of potential sites because these sites
typically meet the reporting threshold. For other OES, such as functional fluids (open-systems), 3D
printing, and other OES that may be performed at a range of different scales, the extent to which TRI
captures all potential sites is more uncertain because not all sites may meet the reporting threshold. This
uncertainty may result in an underestimate in the total number of sites using the chemical within the
OES. To the extent additional sites are not captured, releases may be underestimated; however, the
magnitude of this underestimation is unknown for each OES. In some cases, the total number of
facilities for a given OES was estimated using data from the U.S. Census. This may result in an
overestimate of the actual number of sites that use the chemical for that OES.

Page 295 of 484


-------
7646

7647

7648

7649

7650

7651

7652

7653

7654

7655

7656

7657

7658

7659

7660

7661

7662

7663

7664

7665

7666

7667

7668

7669

7670

7671

7672

7673

7674

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Additionally, it is unknown how representative release estimates from TRI and DMR reporting sites
accurately reflect all releases from within an OES since releases from non-reporting sites cannot be
quantified. Specifically, where the number of sites was estimated from U.S. Census data, the average
daily release calculated from sites reporting to TRI or DMR was applied to the total number of sites
reported in (U.S. Census Bureau. ^ ). It is uncertain how accurate this average release is to actual
releases at these sites; therefore, releases may be higher or lower than the calculated amount.

Estimating Emissions for OES without TRI Data

For release estimates developed for an OES when directly applicable TRI or DMR data were not
available, there are uncertainties related to the use of surrogate TRI or DMR data or, in their absence,
the use of modeling.

Use of surrogate TRI or DMR data may introduce uncertainties related to the extent to which the
surrogate OES and the OES being assessed are similar. Thus, the representativeness of the surrogate
release data towards the actual releases for the OES being assessed is uncertain.

Although no new models were developed for this release assessment, the adaptations made to and uses
of these models as part of the analysis (e.g., varying input parameters, Monte Carlo simulation) may
result in release estimates higher or lower than the actual amount. EPA presented central tendency (50th
percentile) and high-end (95th percentile) modeled release values to capture a range of potential releases
and reduce the uncertainty associated with using a single release estimate. Additionally, the approaches
used for estimating releases based on modeling or literature differs from the facility-specific approach
used for OES for which TRI or DMR data were available (as described previously, the modeled
estimates are for a "generic site"). This may introduce uncertainties that differ from those of the
scenarios using TRI or DMR data, described above.

E.8 Weight of the Scientific Evidence Conclusions for Environmental
Releases

As discussed in Section 2.2.1.2, Table_Apx E-10 presents a summary of EPA's overall weight of the
scientific evidence conclusions for its release estimates for each of the assessed OES.

Page 296 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

7675 Table Apx E-10. Summary of Overall Weight of the Scientific Evidence Conclusions for Environmental Release Estimates by PES

OES

Monitoring"

Modeling

Weight of the Seientitle Evidence
Conclusion

Notes

Air

Water

Land

Data
Quality
Rating

Air

Water

Land

Data
Quality
Rating

Air

Water

Land

Manufacturing

/



•
-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Monitoring"

Modeling

Weight of the Seientitle Evidence
Conclusion

Notes

Air

Water

Land

Data
Quality
Rating

Air

Water

Land

Data
Quality
Rating

Air

Water

Land

Printing Inks
(3D)

Assumed
included
in

Industria
1 Uses
OES

/

Assumed
included
in

Industrial

Uses OES

M/H

X

X

X

N/A

Slight

Moderate to
Robust

Slight

Based on TRI and DMR which have
"medium" data quality ratings.
Information on the conditions of use of
1,4-dioxane at facilities in TRI and DMR
is limited. There is uncertainty in the
representativeness of the Industrial Uses
data for this OES.

Dry Film
Lubricant

X

X

X

N/A

/

Not
expected

/

H

Slight to
Moderate

Slight to
Moderate

Slight to
Moderate

The underlying data sources for model
parameters have a "high" data quality
rating. Modeling may not be sufficiently
representative of all the sites for this OES.

Disposal

/

/

•
-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023





Monitoring"





Modeling



Weight of the Seientitie Evidence















Conclusion





OES







Data







Data







Notes



Air

Water

Land

Quality
Rating

Air

Water

Land

Quality
Rating

Air

Water

Land



























commercial setting, and this case study

























does not represent all sites in this OES.

Dishwasher

X

X

X

N/A

Not

/

/

N/A

Indetermi

Slight

Slight

Assessed using SHEDs-HT data for the

Detergent









assess
ed







nate





Liverpool OH case study and standard
EPA/OPPT models. There is uncertainty
in the application of this modeling for a
commercial setting, and this case study
does not represent all sites in this OES.

Laundry

X

X

X

N/A



/

/

M

Moderate

Moderate

Moderate

Assessed using Laundries ESD, which has

Detergent
(Industrial)























a "medium" data quality rating. Monte
Carlo modeling allows for parameter
variation; however, the modeling may not
be sufficiently representative of all the
sites for this OES.

Laundry

X

X

X

N/A

/

/

/

M

Moderate

Moderate

Moderate

Assessed using Laundries ESD, which has

Detergent
(Institutional)























a "medium" data quality rating. Monte
Carlo modeling allows for parameter
variation; however, the modeling may not
be sufficiently representative of all the
sites for this OES.

Paints and Floor
Lacquer

X

X

X

N/A

/

Not
expected

/

M

Slight to
Moderate

Slight to
Moderate

Slight to
Moderate

Assessed using Automotive Spray
Painting ESD, which has a "medium" data
quality rating. Modeling may not be
sufficiently representative of all the sites
for this exposure scenario.

PET Byproduct

/

/

/

M/H

X

X

X

N/A

Moderate
to Robust

Moderate to
Robust

Moderate
to Robust

Based on TRI and DMR which have
"medium" data quality ratings.
Information on the conditions of use of
1,4-dioxane at facilities in TRI and DMR
is limited, but uncertainties are known.

Ethoxylation

/

/

/

M/H

X

X

X

N/A

Moderate

Moderate to

Moderate

Based on TRI and DMR which have

Process

















to Robust

Robust

to Robust

"medium" data quality ratings.

Byproduct























Information on the conditions of use of
1,4-dioxane at facilities in TRI and DMR
is limited, but uncertainties are known.

Hydraulic







M







M

Moderate

Moderate to

Moderate

Based on FracFocus 3.0 and the Hydraulic

Fracturing

















to Robust

Robust

to Robust

Fracturing ESD, which has a "medium"
data quality rating. Monte Carlo modeling
allows for parameter variation; however,

Page 299 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

OES

Monitoring"

Modeling

Weight of the Seientitle Evidence
Conclusion

Notes

Air

Water

Land

Data
Quality
Rating

Air

Water

Land

Data
Quality
Rating

Air

Water

Land

























the modeling may not be sufficiently
representative of all the sites for this OES.

7676

7677

7678

7679

7680

E.9 TRI to CDR Crosswalk

Table Apx E-l 1 presents the TRI-CDR Crosswalk used to map facilities to the OES for each chemical. Blanks in the 2016 CDR code column
indicate there is no corresponding CDR code that matches the TRI code.

Table Apx E-ll. TRI-CDR Use Code Crosswalk

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition

3.1.a

Manufacture:
Produce

N/A

N/A

N/A

N/A

N/A

3.1.b

Manufacture:
Import

N/A

N/A

N/A

N/A

N/A

3.1.c

Manufacture: For
on-site

use/processing

N/A

N/A

N/A

N/A

N/A

3.1.d

Manufacture: For
sale/distribution

N/A

N/A

N/A

N/A

N/A

3.1.e

Manufacture: As
a byproduct

N/A

N/A

N/A

N/A

N/A

3.1.f

Manufacture: As
an impurity

N/A

N/A

N/A

N/A

N/A

3.2.a

Processing: As a
reactant

N/A

N/A

PC

Processing as a
reactant

Chemical substance is used in chemical reactions for the
manufacturing of another chemical substance or product.

3.2.a

Processing: As a
reactant

P101

Feedstocks

N/A

N/A

N/A

3.2.a

Processing: As a
reactant

P102

Raw Materials

N/A

N/A

N/A

Page 300 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition

3.2.a

Processing: As a
reactant

P103

Intermediates

U015

Intermediates

Chemical substances consumed in a reaction to produce
other chemical substances for commercial advantage. A
residual of the intermediate chemical substance which has
no separate function may remain in the reaction product.

3.2.a

Processing: As a
reactant

P104

Initiators

U024

Process regulators

Chemical substances used to change the rate of a
chemical reaction, start or stop the reaction, or otherwise
influence the course of the reaction. Process regulators
may be consumed or become part of the reaction product.

3.2.a

Processing: As a
reactant

P199

Other

U016

Ion exchange agents

Chemical substances, usually in the form of a solid
matrix, that are used to selectively remove targeted ions
from a solution. Examples generally consist of an inert
hydrophobic matrix such as styrenedivinylbenzene or
phenol-formaldehyde, cross-linking polymer such as
divinylbenzene, and ionic functional groups including
sulfonic, carboxylic or phosphonic acids. This code also
includes aluminosilicate zeolites.

3.2.a

Processing: As a
reactant

P199

Other

U019

Oxidizing/
reducing agent

Chemical substances used to alter the valence state of
another substance by donating or accepting electrons or
by the addition or removal of hydrogen to a substance.
Examples of oxidizing agents include nitric acid,
perchlorates, hexavalent chromium compounds, and
peroxydisulfuric acid salts. Examples of reducing agents
include hydrazine, sodium thiosulfate, and coke produced
from coal.

3.2.a

Processing: As a
reactant

P199

Other

U999

Other (specify)

Chemical substances used in a way other than those
described by other codes.

3.2.b

Processing: As a

formulation

component

N/A

N/A

PF

Processing-
incorporation into
formulation, mixture,
or reaction product

Chemical substance is added to a product (or product
mixture) prior to further distribution of the product.

3.2.b

Processing: As a

formulation

component

P201

Additives

U007

Corrosion inhibitors
and anti-scaling
agents

Chemical substances used to prevent or retard corrosion
or the formation of scale. Examples include
phenylenediamine, chromates, nitrates, phosphates, and
hydrazine.

Page 301 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition

3.2.b

Processing: As a

formulation

component

P201

Additives

U009

Fillers

Chemical substances used to provide bulk, increase
strength, increase hardness, or improve resistance to
impact. Fillers incorporated in a matrix reduce production
costs by minimizing the amount of more expensive
substances used in the production of articles. Examples
include calcium carbonate, barium sulfate, silicates, clays,
zinc oxide and aluminum oxide.

3.2.b

Processing: As a

formulation

component

P201

Additives

U010

Finishing agents

Chemical substances used to impart such functions as
softening, static proofing, wrinkle resistance, and water
repellence. Substances may be applied to textiles, paper,
and leather. Examples include quaternary ammonium
compounds, ethoxylated amines, and silicone compounds.

3.2.b

Processing: As a

formulation

component

P201

Additives

U017

Lubricants and
lubricant additives

Chemical substances used to reduce friction, heat, or wear
between moving parts or adjacent solid surfaces, or that
enhance the lubricity of other substances. Examples of
lubricants include mineral oils, silicate and phosphate
esters, silicone oil, greases, and solid film lubricants such
as graphite and PTFE. Examples of lubricant additives
include molybdenum disulphide and tungsten disulphide.

3.2.b

Processing: As a

formulation

component

P201

Additives

U034

Paint additives and
coating additives not
described by other
codes

Chemical substances used in a paint or coating
formulation to enhance properties such as water
repellence, increased gloss, improved fade resistance,
ease of application, foam prevention, etc. Examples of
paint additives and coating additives include polyols,
amines, vinyl acetate ethylene emulsions, and aliphatic
polyisocyanates.

3.2.b

Processing: As a

formulation

component

P202

Dyes

U008

Dyes

Chemical substances used to impart color to other
materials or mixtures (i.e., substrates) by penetrating the
surface of the substrate. Example types include azo,
anthraquinone, amino azo, aniline, eosin, stilbene, acid,
basic or cationic, reactive, dispersive, and natural dyes.

3.2.b

Processing: As a

formulation

component

P202

Dyes

U021

Pigments

Chemical substances used to impart color to other
materials or mixtures (i.e., substrates) by attaching
themselves to the surface of the substrate through binding

Page 302 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition













or adhesion. This code includes fluorescent agents,
luminescent agents, whitening agents, pearlizing agents,
and opacifiers. Examples include metallic oxides of iron,
titanium, zinc, cobalt, and chromium; metal powder
suspensions; lead chromates; vegetable and animal
products; and synthetic organic pigments.

3.2.b

Processing: As a

formulation

component

P203

Reaction
Diluents

U030

Solvents (which
become part of
product formulation or
mixture)

Chemical substances used to dissolve another substance
(solute) to form a uniformly dispersed mixture (solution)
at the molecular level. Examples include diluents used to
reduce the concentration of an active material to achieve a
specified effect and low gravity materials added to reduce
cost.

3.2.b

Processing: As a

formulation

component

P203

Reaction
Diluents

U032

Viscosity adjustors

Chemical substances used to alter the viscosity of another
substance. Examples include viscosity index (VI)
improvers, pour point depressants, and thickeners.

3.2.b

Processing: As a

formulation

component

P204

Initiators

U024

Process regulators

Chemical substances used to change the rate of a
chemical reaction, start or stop the reaction, or otherwise
influence the course of the reaction. Process regulators
may be consumed or become part of the reaction product.

3.2.b

Processing: As a

formulation

component

P205

Solvents

U030

Solvents (which
become part of
product formulation or
mixture)

Chemical substances used to dissolve another substance
(solute) to form a uniformly dispersed mixture (solution)
at the molecular level. Examples include diluents used to
reduce the concentration of an active material to achieve a
specified effect and low gravity materials added to reduce
cost.

3.2.b

Processing: As a

formulation

component

P206

Inhibitors

U024

Process regulators

Chemical substances used to change the rate of a
chemical reaction, start or stop the reaction, or otherwise
influence the course of the reaction. Process regulators
may be consumed or become part of the reaction product.

3.2.b

Processing: As a

formulation

component

P207

Emulsifiers

U003

Adsorbents and
absorbents

Chemical substances used to retain other substances by
accumulation on their surface or by assimilation.
Examples of adsorbents include silica gel, activated
alumina, and activated carbon. Examples of absorbents
include straw oil, alkaline solutions, and kerosene.

Page 303 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition

3.2.b

Processing: As a

formulation

component

P208

Surfactants

U002

Adhesives and sealant
chemicals

Chemical substances used to promote bonding between
other substances, promote adhesion of surfaces, or
prevent seepage of moisture or air. Examples include
epoxides, isocyanates, acrylamides, phenol, urea,
melamine, and formaldehyde.

3.2.b

Processing: As a

formulation

component

P208

Surfactants

U023

Plating agents and
surface treating agents

Chemical substances applied to metal, plastic, or other
surfaces to alter physical or chemical properties of the
surface. Examples include metal surface treating agents,
strippers, etchants, rust and tarnish removers, and
descaling agents.

3.2.b

Processing: As a

formulation

component

P208

Surfactants

U031

Surface active agents

Chemical substances used to modify surface tension when
dissolved in water or water solutions or reduce interfacial
tension between two liquids or between a liquid and a
solid or between liquid and air. Examples include
carboxylates, sulfonates, phosphates, carboxylic acid,
esters, and quaternary ammonium salts.

3.2.b

Processing: As a

formulation

component

P209

Lubricants

U017

Lubricants and
lubricant additives

Chemical substances used to reduce friction, heat, or wear
between moving parts or adjacent solid surfaces, or that
enhance the lubricity of other substances. Examples of
lubricants include mineral oils, silicate and phosphate
esters, silicone oil, greases, and solid film lubricants such
as graphite and PTFE. Examples of lubricant additives
include molybdenum disulphide and tungsten disulphide.

3.2.b

Processing: As a

formulation

component

P210

Flame
Retardants

U011

Flame retardants

Chemical substances used on the surface of or
incorporated into combustible materials to reduce or
eliminate their tendency to ignite when exposed to heat or
a flame for a short period of time. Examples include
inorganic salts, chlorinated or brominated organic
compounds, and organic phosphates/phosphonates.

3.2.b

Processing: As a

formulation

component

P211

Rheological
Modifiers

U022

Plasticizers

Chemical substances used in plastics, cement, concrete,
wallboard, clay bodies, or other materials to increase their
plasticity or fluidity. Examples include phthalates,
trimellitates, adipates, maleates, and lignosulphonates.

Page 304 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition

3.2.b

Processing: As a

formulation

component

P211

Rheological
Modifiers

U032

Viscosity adjustors

Chemical substances used to alter the viscosity of another
substance. Examples include viscosity index (VI)
improvers, pour point depressants, and thickeners.

3.2.b

Processing: As a

formulation

component

P299

Other

U003

Adsorbents and
absorbents

Chemical substances used to retain other substances by
accumulation on their surface or by assimilation.
Examples of adsorbents include silica gel, activated
alumina, and activated carbon. Examples of absorbents
include straw oil, alkaline solutions, and kerosene.

3.2.b

Processing: As a

formulation

component

P299

Other

U016

Ion exchange agents

Chemical substances, usually in the form of a solid
matrix, that are used to selectively remove targeted ions
from a solution. Examples generally consist of an inert
hydrophobic matrix such as styrenedivinylbenzene or
phenol-formaldehyde, cross-linking polymer such as
divinylbenzene, and ionic functional groups including
sulfonic, carboxylic or phosphonic acids. This code also
includes aluminosilicate zeolites.

3.2.b

Processing: As a

formulation

component

P299

Other

U018

Odor agents

Chemical substances used to control odors, remove odors,
mask odors, or impart odors. Examples include
benzenoids, terpenes and terpenoids, musk chemicals,
aliphatic aldehydes, aliphatic cyanides, and mercaptans.

3.2.b

Processing: As a

formulation

component

P299

Other

U019

Oxidizing/
reducing agent

Chemical substances used to alter the valence state of
another substance by donating or accepting electrons or
by the addition or removal of hydrogen to a substance.
Examples of oxidizing agents include nitric acid,
perchlorates, hexavalent chromium compounds, and
peroxydisulfuric acid salts. Examples of reducing agents
include hydrazine, sodium thiosulfate, and coke produced
from coal.

3.2.b

Processing: As a

formulation

component

P299

Other

U020

Photosensitive
chemicals

Chemical substances used for their ability to alter their
physical or chemical structure through absorption of light,
resulting in the emission of light, dissociation,
discoloration, or other chemical reaction. Examples
include sensitizers, fluorescents, photovoltaic agents,
ultraviolet absorbers, and ultraviolet stabilizers.

Page 305 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition

3.2.b

Processing: As a

formulation

component

P299

Other

U027

Propellants and
blowing agents

Chemical substances used to dissolve or suspend other
substances and either to expel those substances from a
container in the form of an aerosol or to impart a cellular
structure to plastics, rubber, or thermo set resins.
Examples include compressed gasses and liquids and
substances which release ammonia, carbon dioxide, or
nitrogen.

3.2.b

Processing: As a

formulation

component

P299

Other

U028

Solid separation
agents

Chemical substances used to promote the separation of
suspended solids from a liquid. Examples include
flotation aids, flocculants, coagulants, dewatering aids,
and drainage aids.

3.2.b

Processing: As a

formulation

component

P299

Other

U999

Other (specify)

Chemical substances used in a way other than those
described by other codes.

3.2.c

Processing: As
an article
component

N/A

N/A

PA

Processing-
incorporation into
article

Chemical substance becomes an integral component of an
article distributed for industrial, trade, or consumer use.

3.2.c

Processing: As
an article
component

N/A

N/A

U008

Dyes

Chemical substances used to impart color to other
materials or mixtures (i.e., substrates) by penetrating into
the surface of the substrate. Examples types include azo,
anthraquinone, amino azo, aniline, eosin, stilbene, acid,
basic or cationic, reactive, dispersive, and natural dyes.

3.2.c

Processing: As
an article
component

N/A

N/A

U009

Fillers

Chemical substances used to provide bulk, increase
strength, increase hardness, or improve resistance to
impact. Fillers incorporated in a matrix reduce production
costs by minimizing the amount of more expensive
substances used in the production of articles. Examples
include calcium carbonate, barium sulfate, silicates, clays,
zinc oxide and aluminum oxide.

3.2.c

Processing: As
an article
component

N/A

N/A

U021

Pigments

Chemical substances used to impart color to other
materials or mixtures (i.e., substrates) by attaching
themselves to the surface of the substrate through binding
or adhesion. This code includes fluorescent agents,
luminescent agents, whitening agents, pearlizing agents,

Page 306 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition













and opacifiers. Examples include metallic oxides of iron,
titanium, zinc, cobalt, and chromium; metal powder
suspensions; lead chromates; vegetable and animal
products; and synthetic organic pigments.

3.2.c

Processing: As
an article
component

N/A

N/A

U034

Paint additives and
coating additives not
described by other
codes

Chemical substances used in a paint or coating
formulation to enhance properties such as water
repellence, increased gloss, improved fade resistance,
ease of application, foam prevention, etc. Examples of
paint additives and coating additives include polyols,
amines, vinyl acetate ethylene emulsions, and aliphatic
polyisocyanates.

3.2.c

Processing: As
an article
component

N/A

N/A

U999

Other (specify)

Chemical substances used in a way other than those
described by other codes.

3.2.d

Processing:
Repackaging

N/A

N/A

PK

Processing-
repackaging

Preparation of a chemical substance for distribution in
commerce in a different form, state, or quantity. This
includes transferring the chemical substance from a bulk
container into smaller containers. This definition does not
apply to sites that only relabel or redistribute the
reportable chemical substance without removing the
chemical substance from the container in which it is
received or purchased.

3.2.e

Processing: As
an impurity

N/A

N/A

N/A

N/A

N/A

3.2.f

Processing:
Recycling

N/A

N/A

N/A

N/A

N/A

3.3.a

Otherwise Use:
As a chemical
processing aid

N/A

N/A

U

Use-non incorporative
Activities

Chemical substance is otherwise used (e.g., as a chemical
processing or manufacturing aid).

3.3.a

Otherwise Use:
As a chemical
processing aid

Z101

Process Solvents

U029

Solvents (for cleaning
or degreasing)

Chemical substances used to dissolve oils, greases, and
similar materials from textiles, glassware, metal surfaces,
and other articles. Examples include trichloroethylene,
perchloroethylene, methylene chloride, liquid carbon
dioxide, and n-propyl bromide.

Page 307 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition

3.3.a

Otherwise Use:
As a chemical
processing aid

Z102

Catalysts

U020

Photosensitive
chemicals

Chemical substances used for their ability to alter their
physical or chemical structure through absorption of light,
resulting in the emission of light, dissociation,
discoloration, or other chemical reaction. Examples
include sensitizers, fluorescents, photovoltaic agents,
ultraviolet absorbers, and ultraviolet stabilizers.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z102

Catalysts

U025

Processing aids,
specific to petroleum
production

Chemical substances added to water-, oil-, or synthetic
drilling muds or other petroleum production fluids to
control viscosity, foaming, corrosion, alkalinity and pH,
microbiological growth, hydrate formation, etc., during
the production of oil, gas, and other products from
beneath the earth's surface.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z102

Catalysts

U026

Processing aids, not
otherwise listed

Chemical substances used to improve the processing
characteristics or the operation of process equipment or to
alter or buffer the pH of the substance or mixture, when
added to a process or to a substance or mixture to be
processed. Processing agents do not become a part of the
reaction product and are not intended to affect the
function of a substance or article created. Examples
include buffers, dehumidifiers, dehydrating agents,
sequestering agents, and chelators.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z103

Inhibitors

U024

Process regulators

Chemical substances used to change the rate of a
chemical reaction, start or stop the reaction, or otherwise
influence the course of the reaction. Process regulators
may be consumed or become part of the reaction product.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z103

Inhibitors

U025

Processing aids,
specific to petroleum
production

Chemical substances added to water-, oil-, or synthetic
drilling muds or other petroleum production fluids to
control viscosity, foaming, corrosion, alkalinity and pH,
microbiological growth, hydrate formation, etc., during
the production of oil, gas, and other products from
beneath the earth's surface.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z103

Inhibitors

U026

Processing aids, not
otherwise listed

Chemical substances used to improve the processing
characteristics or the operation of process equipment or to
alter or buffer the pH of the substance or mixture, when

Page 308 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition













added to a process or to a substance or mixture to be
processed. Processing agents do not become a part of the
reaction product and are not intended to affect the
function of a substance or article created. Examples
include buffers, dehumidifiers, dehydrating agents,
sequestering agents, and chelators.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z104

Initiators

U024

Process regulators

Chemical substances used to change the rate of a
chemical reaction, start or stop the reaction, or otherwise
influence the course of the reaction. Process regulators
may be consumed or become part of the reaction product.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z104

Initiators

U025

Processing aids,
specific to petroleum
production

Chemical substances added to water-, oil-, or synthetic
drilling muds or other petroleum production fluids to
control viscosity, foaming, corrosion, alkalinity and pH,
microbiological growth, hydrate formation, etc., during
the production of oil, gas, and other products from
beneath the earth's surface.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z104

Initiators

U026

Processing aids, not
otherwise listed

Chemical substances used to improve the processing
characteristics or the operation of process equipment or to
alter or buffer the pH of the substance or mixture, when
added to a process or to a substance or mixture to be
processed. Processing agents do not become a part of the
reaction product and are not intended to affect the
function of a substance or article created. Examples
include buffers, dehumidifiers, dehydrating agents,
sequestering agents, and chelators.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z105

Reaction
Terminators

U024

Process regulators

Chemical substances used to change the rate of a
chemical reaction, start or stop the reaction, or otherwise
influence the course of the reaction. Process regulators
may be consumed or become part of the reaction product.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z105

Reaction
Terminators

U025

Processing aids,
specific to petroleum
production

Chemical substances added to water-, oil-, or synthetic
drilling muds or other petroleum production fluids to
control viscosity, foaming, corrosion, alkalinity and pH,
microbiological growth, hydrate formation, etc., during

Page 309 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition













the production of oil, gas, and other products from
beneath the earth's surface.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z105

Reaction
Terminators

U026

Processing aids, not
otherwise listed

Chemical substances used to improve the processing
characteristics or the operation of process equipment or to
alter or buffer the pH of the substance or mixture, when
added to a process or to a substance or mixture to be
processed. Processing agents do not become a part of the
reaction product and are not intended to affect the
function of a substance or article created. Examples
include buffers, dehumidifiers, dehydrating agents,
sequestering agents, and chelators.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z106

Solution Buffers

U026

Processing aids, not
otherwise listed

Chemical substances used to improve the processing
characteristics or the operation of process equipment or to
alter or buffer the pH of the substance or mixture, when
added to a process or to a substance or mixture to be
processed. Processing agents do not become a part of the
reaction product and are not intended to affect the
function of a substance or article created. Examples
include buffers, dehumidifiers, dehydrating agents,
sequestering agents, and chelators.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z199

Other

U002

Adhesives and sealant
chemicals

Chemical substances used to promote bonding between
other substances, promote adhesion of surfaces, or
prevent seepage of moisture or air. Examples include
epoxides, isocyanates, acrylamides, phenol, urea,
melamine, and formaldehyde.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z199

Other

U006

Bleaching agents

Chemical substances used to lighten or whiten a substrate
through chemical reaction, usually an oxidative process
which degrades the color system. Examples generally fall
into one of two groups: chlorine containing bleaching
agents (e.g., chlorine, hypochlorites, N-chloro compounds
and chlorine dioxide); and peroxygen bleaching agents
(e.g., hydrogen peroxide, potassium permanganate, and
sodium perborate).

Page 310 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition

3.3.a

Otherwise Use:
As a chemical
processing aid

Z199

Other

U018

Odor agents

Chemical substances used to control odors, remove odors,
mask odors, or impart odors. Examples include
benzenoids, terpenes and terpenoids, musk chemicals,
aliphatic aldehydes, aliphatic cyanides, and mercaptans.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z199

Other

U023

Plating agents and
surface treating agents

Chemical substances applied to metal, plastic, or other
surfaces to alter physical or chemical properties of the
surface. Examples include metal surface treating agents,
strippers, etchants, rust and tarnish removers, and
descaling agents.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z199

Other

U025

Processing aids,
specific to petroleum
production

Chemical substances added to water-, oil-, or synthetic
drilling muds or other petroleum production fluids to
control viscosity, foaming, corrosion, alkalinity and pH,
microbiological growth, hydrate formation, etc., during
the production of oil, gas, and other products from
beneath the earth's surface.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z199

Other

U026

Processing aids, not
otherwise listed

Chemical substances used to improve the processing
characteristics or the operation of process equipment or to
alter or buffer the pH of the substance or mixture, when
added to a process or to a substance or mixture to be
processed. Processing agents do not become a part of the
reaction product and are not intended to affect the
function of a substance or article created. Examples
include buffers, dehumidifiers, dehydrating agents,
sequestering agents, and chelators.

3.3.a

Otherwise Use:
As a chemical
processing aid

Z199

Other

U028

Solid separation
agents

Chemical substances used to promote the separation of
suspended solids from a liquid. Examples include
flotation aids, flocculants, coagulants, dewatering aids,
and drainage aids.

3.3.b

Otherwise Use:
As a

manufacturing
aid

N/A

N/A

U

Use-non

incorporative

Activities

Chemical substance is otherwise used (e.g., as a chemical
processing or manufacturing aid).

3.3.b

Otherwise Use:
As a

Z201

Process
Lubricants

U017

Lubricants and
lubricant additives

Chemical substances used to reduce friction, heat, or wear
between moving parts or adjacent solid surfaces, or that

Page 311 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition



manufacturing
aid









enhance the lubricity of other substances. Examples of
lubricants include mineral oils, silicate and phosphate
esters, silicone oil, greases, and solid film lubricants such
as graphite and PTFE. Examples of lubricant additives
include molybdenum disulphide and tungsten disulphide.

3.3.b

Otherwise Use:
As a

manufacturing
aid

Z202

Metalworking
Fluids

U007

Corrosion inhibitors
and antiscaling agents

Chemical substances used to prevent or retard corrosion
or the formation of scale. Examples include
phenylenediamine, chromates, nitrates, phosphates, and
hydrazine.

3.3.b

Otherwise Use:
As a

manufacturing
aid

Z202

Metalworking
Fluids

U014

Functional fluids
(open systems)

Liquid or gaseous chemical substances used for one or
more operational properties in an open system. Examples
include antifreezes and de-icing fluids such as ethylene
and propylene glycol, sodium formate, potassium acetate,
and sodium acetate. This code also includes substances
incorporated into metal working fluids.

3.3.b

Otherwise Use:
As a

manufacturing
aid

Z203

Coolants

U013

Functional fluids
(closed systems)

Liquid or gaseous chemical substances used for one or
more operational properties in a closed system. Examples
include: heat transfer agents (e.g., coolants and
refrigerants) such as polyalkylene glycols, silicone oils,
liquified propane, and carbon dioxide;
hydraulic/transmission fluids such as mineral oils,
organophosphate esters, silicone, and propylene glycol;
and dielectric fluids such as mineral insulating oil and
high flash point kerosene. This code does not include
fluids used as lubricants.

3.3.b

Otherwise Use:
As a

manufacturing
aid

Z204

Refrigerants

U013

Functional fluids
(closed systems)

Liquid or gaseous chemical substances used for one or
more operational properties in a closed system. Examples
include: heat transfer agents (e.g., coolants and
refrigerants) such as polyalkylene glycols, silicone oils,
liquified propane, and carbon dioxide;
hydraulic/transmission fluids such as mineral oils,
organophosphate esters, silicone, and propylene glycol;
and dielectric fluids such as mineral insulating oil and

Page 312 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition













high flash point kerosene. This code does not include
fluids used as lubricants.

3.3.b

Otherwise Use:
As a

manufacturing
aid

Z205

Hydraulic Fluids

U013

Functional fluids
(closed systems)

Liquid or gaseous chemical substances used for one or
more operational properties in a closed system. Examples
include: heat transfer agents (e.g., coolants and
refrigerants) such as polyalkylene glycols, silicone oils,
liquified propane, and carbon dioxide;
hydraulic/transmission fluids such as mineral oils,
organophosphate esters, silicone, and propylene glycol;
and dielectric fluids such as mineral insulating oil and
high flash point kerosene. This code does not include
fluids used as lubricants.

3.3.b

Otherwise Use:
As a

manufacturing
aid

Z299

Other

U013

Functional fluids
(closed systems)

Liquid or gaseous chemical substances used for one or
more operational properties in a closed system. Examples
include: heat transfer agents (e.g., coolants and
refrigerants) such as polyalkylene glycols, silicone oils,
liquified propane, and carbon dioxide;
hydraulic/transmission fluids such as mineral oils,
organophosphate esters, silicone, and propylene glycol;
and dielectric fluids such as mineral insulating oil and
high flash point kerosene. This code does not include
fluids used as lubricants.

3.3.b

Otherwise Use:
As a

manufacturing
aid

Z299

Other

U023

Plating agents and
surface treating agents

Chemical substances applied to metal, plastic, or other
surfaces to alter physical or chemical properties of the
surface. Examples include metal surface treating agents,
strippers, etchants, rust and tarnish removers, and
descaling agents.

3.3.c

Otherwise Use:
Ancillary or
other use

N/A

N/A

U

Use-non

incorporative

Activities

Chemical substance is otherwise used (e.g., as a chemical
processing or manufacturing aid).

3.3.c

Otherwise Use:
Ancillary or
other use

Z301

Cleaner

U007

Corrosion inhibitors
and antiscaling agents

Chemical substances used to prevent or retard corrosion
or the formation of scale. Examples include
phenylenediamine, chromates, nitrates, phosphates, and
hydrazine.

Page 313 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition

3.3.c

Otherwise Use:
Ancillary or
other use

Z301

Cleaner

U029

Solvents (for cleaning
or degreasing)

Chemical substances used to dissolve oils, greases, and
similar materials from textiles, glassware, metal surfaces,
and other articles. Examples include trichloroethylene,
perchloroethylene, methylene chloride, liquid carbon
dioxide, and n-propyl bromide.

3.3.c

Otherwise Use:
Ancillary or
other use

Z302

Degreaser

U003

Adsorbents and
Absorbents

Chemical substances used to retain other substances by
accumulation on their surface or by assimilation.
Examples of adsorbents include silica gel, activated
alumina, and activated carbon. Examples of absorbents
include straw oil, alkaline solutions, and kerosene.

3.3.c

Otherwise Use:
Ancillary or
other use

Z302

Degreaser

U029

Solvents (for cleaning
or degreasing)

Chemical substances used to dissolve oils, greases, and
similar materials from textiles, glassware, metal surfaces,
and other articles. Examples include trichloroethylene,
perchloroethylene, methylene chloride, liquid carbon
dioxide, and n-propyl bromide.

3.3.c

Otherwise Use:
Ancillary or
other use

Z303

Lubricant

U017

Lubricants and
lubricant additives

Chemical substances used to reduce friction, heat, or wear
between moving parts or adjacent solid surfaces, or that
enhance the lubricity of other substances. Examples of
lubricants include mineral oils, silicate and phosphate
esters, silicone oil, greases, and solid film lubricants such
as graphite and PTFE. Examples of lubricant additives
include molybdenum disulphide and tungsten disulphide.

3.3.c

Otherwise Use:
Ancillary or
other use

Z304

Fuel

U012

Fuels and fuel
additives

Chemical substances used to create mechanical or thermal
energy through chemical reactions, or which are added to
a fuel for the purpose of controlling the rate of reaction or
limiting the production of undesirable combustion
products, or which provide other benefits such as
corrosion inhibition, lubrication, or detergency. Examples
of fuels include coal, oil, gasoline, and various grades of
diesel fuel. Examples of fuel additives include
oxygenated compound such as ethers and alcohols,
antioxidants such as phenylenediamines and hindered
phenols, corrosion inhibitors such as carboxylic acids,

Page 314 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition













amines, and amine salts, and blending agents such as
ethanol.

3.3.c

Otherwise Use:
Ancillary or
other use

Z305

Flame Retardant

U011

Flame retardants

Chemical substances used on the surface of or
incorporated into combustible materials to reduce or
eliminate their tendency to ignite when exposed to heat or
a flame for a short period of time. Examples include
inorganic salts, chlorinated or brominated organic
compounds, and organic phosphates/phosphonates.

3.3.c

Otherwise Use:
Ancillary or
other use

Z306

Waste
Treatment

U006

Bleaching agents

Chemical substances used to lighten or whiten a substrate
through chemical reaction, usually an oxidative process
which degrades the color system. Examples generally fall
into one of two groups: chlorine containing bleaching
agents (e.g., chlorine, hypochlorites, N-chloro compounds
and chlorine dioxide); and peroxygen bleaching agents
(e.g., hydrogen peroxide, potassium permanganate, and
sodium perborate).

3.3.c

Otherwise Use:
Ancillary or
other use

Z306

Waste
Treatment

U018

Odor agents

Chemical substances used to control odors, remove odors,
mask odors, or impart odors. Examples include
benzenoids, terpenes and terpenoids, musk chemicals,
aliphatic aldehydes, aliphatic cyanides, and mercaptans.

3.3.c

Otherwise Use:
Ancillary or
other use

Z306

Waste
Treatment

U019

Oxidizing/reducing
agent

Chemical substances used to alter the valence state of
another substance by donating or accepting electrons or
by the addition or removal of hydrogen to a substance.
Examples of oxidizing agents include nitric acid,
perchlorates, hexavalent chromium compounds, and
peroxydisulfuric acid salts. Examples of reducing agents
include hydrazine, sodium thiosulfate, and coke produced
from coal.

3.3.c

Otherwise Use:
Ancillary or
other use

Z306

Waste
Treatment

U028

Solid separation
agents

Chemical substances used to promote the separation of
suspended solids from a liquid. Examples include
flotation aids, flocculants, coagulants, dewatering aids,
and drainage aids.

Page 315 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition

3.3.c

Otherwise Use:
Ancillary or
other use

Z307

Water Treatment

U006

Bleaching agents

Chemical substances used to lighten or whiten a substrate
through chemical reaction, usually an oxidative process
which degrades the color system. Examples generally fall
into one of two groups: chlorine containing bleaching
agents (e.g., chlorine, hypochlorites, N-chloro compounds
and chlorine dioxide); and, peroxygen bleaching agents
(e.g., hydrogen peroxide, potassium permanganate, and
sodium perborate).

3.3.c

Otherwise Use:
Ancillary or
other use

Z307

Water Treatment

U018

Odor agents

Chemical substances used to control odors, remove odors,
mask odors, or impart odors. Examples include
benzenoids, terpenes and terpenoids, musk chemicals,
aliphatic aldehydes, aliphatic cyanides, and mercaptans.

3.3.c

Otherwise Use:
Ancillary or
other use

Z307

Water Treatment

U019

Oxidizing/reducing
agent

Chemical substances used to alter the valence state of
another substance by donating or accepting electrons or
by the addition or removal of hydrogen to a substance.
Examples of oxidizing agents include nitric acid,
perchlorates, hexavalent chromium compounds, and
peroxydisulfuric acid salts. Examples of reducing agents
include hydrazine, sodium thiosulfate, and coke produced
from coal.

3.3.c

Otherwise Use:
Ancillary or
other use

Z307

Water Treatment

U028

Solid separation
agents

Chemical substances used to promote the separation of
suspended solids from a liquid. Examples include
flotation aids, flocculants, coagulants, dewatering aids,
and drainage aids.

3.3.c

Otherwise Use:
Ancillary or
other use

Z308

Construction
Materials

N/A

N/A

N/A

3.3.c

Otherwise Use:
Ancillary or
other use

Z399

Other

U001

Abrasives

Chemical substances used to wear down or polish
surfaces by rubbing against the surface. Examples include
sandstones, pumice, silex, quartz, silicates, aluminum
oxides, and glass.

3.3.c

Otherwise Use:
Ancillary or
other use

Z399

Other

U013

Functional fluids
(closed systems)

Liquid or gaseous chemical substances used for one or
more operational properties in a closed system. Examples
include: heat transfer agents (e.g., coolants and

Page 316 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

TRI

Section

TRI Description

TRI Sub-
use Code

TRI Sub-use
Code Name

2016
CDR

Code

2016 CDR Code
Name

2016 CDR Functional Use Definition













refrigerants) such as polyalkylene glycols, silicone oils,
liquified propane, and carbon dioxide;
hydraulic/transmission fluids such as mineral oils,
organophosphate esters, silicone, and propylene glycol;
and dielectric fluids such as mineral insulating oil and
high flash point kerosene. This code does not include
fluids used as lubricants.

3.3.c

Otherwise Use:
Ancillary or
other use

Z399

Other

U014

Functional fluids
(open systems)

Liquid or gaseous chemical substances used for one or
more operational properties in an open system. Examples
include antifreezes and de-icing fluids such as ethylene
and propylene glycol, sodium formate, potassium acetate,
and, sodium acetate. This code also includes substances
incorporated into metal working fluids.

3.3.c

Otherwise Use:
Ancillary or
other use

Z399

Other

U018

Odor agents

Chemical substances used to control odors, remove odors,
mask odors, or impart odors. Examples include
benzenoids, terpenes and terpenoids, musk chemicals,
aliphatic aldehydes, aliphatic cyanides, and mercaptans.

3.3.c

Otherwise Use:
Ancillary or
other use

Z399

Other

U020

Photosensitive
chemicals

Chemical substances used for their ability to alter their
physical or chemical structure through absorption of light,
resulting in the emission of light, dissociation,
discoloration, or other chemical reaction. Examples
include sensitizers, fluorescents, photovoltaic agents,
ultraviolet absorbers, and ultraviolet stabilizers.

3.3.c

Otherwise Use:
Ancillary or
other use

Z399

Other

U023

Plating agents and
surface treating agents

Chemical substances applied to metal, plastic, or other
surfaces to alter physical or chemical properties of the
surface. Examples include metal surface treating agents,
strippers, etchants, rust and tarnish removers, and
descaling agents.

7681

7682

Page 317 of 484


-------
7683

7684

7685

7686

7687

7688

7689

7690

7691

7692

7693

7694

7695

7696

7697

7698

7699

7700

7701

7702

7703

7704

7705

7706

7707

7708

7709

7710

7711

7712

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

E.10 Developing Models that Use Monte Carlo Methods

This appendix provides background information on Monte Carlo methods, including an overview of
deterministic and stochastic processes, an overview of the implementation of Monte Carlo methods, and
a discussion of EPA's approach for building models that utilized Monte Carlo methods.

This appendix is only intended to provide general background information; information related to the
specific models for which EPA implemented Monte Carlo methods is included in Appendices E. 11
through E.13 and Appendices F.5 through F.9.

E.10.1 Background on Monte Carlo Methods

A deterministic process has a single output (or set of outputs) for a given input (or set of inputs). The
process does not involve randomness and the direction of the process is known.

In contrast, stochastic processes are non-deterministic. The output is based on random trials and can
proceed via multiple, or even infinite, directions.

Monte Carlo methods fall under the umbrella of stochastic modeling. Monte Carlo methods are a
replication technique for propagating uncertainty through a model. The model is run multiple times, and
each run uses different input values and generates different output values: each run is independent of
each other. The sample of output values is used to estimate the properties of the actual probability
distribution of the outputs.

E.10.2 Implementation of Monte Carlo Methods

The implementation of Monte Carlo methods generally follows the following steps:

1.	Define probability distributions for input parameters.

2.	Generate a set of input values by randomly drawing a sample from each probability distribution.

3.	Execute the deterministic model calculations.

4.	Save the output results.

5.	Repeat steps 2 through 4 through the desired number of iterations.

6.	Aggregate the saved output results and calculate statistics.

Figure Apx E-l illustrates a flowchart of a Monte Carlo method implemented in a Microsoft Excel-
based model using a Monte Carlo add-in tool, such as the Palisade @Risk software.

Page 318 of 484


-------
7713

7714

7715

7716

7717

7718

7719

7720

7721

7722

7723

7724

7725

7726

7727

7728

7729

7730

7731

7732

7733

7734

7735

7736

7737

7738

7739

7740

7741

7742

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Results are Stored until
Desired Iterations are
Met

Iterations
Met

Results Summary
and Descriptive
Statistics

Define
Inputs

Monte Carlo
Add-In Tool

Randomly
Selected
Inputs

Deterministic Model

Outputs

FigureApx E-l. Flowchart of a Monte Carlo Method Implemented in a Microsoft Excel-Based
Model Using a Monte Carlo Add-In Tool

E.10.3 Building the Model

The steps for building a release or exposure model that incorporates Monte Carlo methods are as
follows:

1.	Build the deterministic model.

2.	Define probability distributions for input parameters.

3.	Select model outputs for aggregation of simulation results.

4.	Select simulation settings and run model.

5.	Aggregate the simulation results and calculate output statistics.

Each of these steps is discussed in the subsections below.

E. 10.3.1 Build the Deterministic Model

First, the model is built as a deterministic model. EPA uses Microsoft Excel in order to use Palisade's
@Risk software that is used for probabilistic analyses in Excel. The model parameters and equations are
programmed into the spreadsheet. Model parameters are programmed in a summary table format for
transparency and to aid in the assignment of probability di stributions. Such summary tables are included
in the model-specific write-ups in Appendices E, 11 through E. 13 and Appendices F.5 through F.9.

E. 10.3.2 Define Probability Distributions for Input Parameters

Defining a probability distribution for an input parameter generally involves three steps:

1.	Select the model input parameters for which probability distributions will be developed.

2.	Determine a probability distribution from the available data.

3.	Investigate if any parameters are statistically correlated. Define a statistical correlation among
parameters if a correlation is desired.

Step 1: Select Input Parameters for Probability Distribution Development

When selecting parameters for which probability distributions will be developed, the following factors
are considered:

•	The availability of data to inform a distribution.

•	The dependency of the input parameters on one another.

The sensitivity of the model results to each input parameter.

Page 319 of 484


-------
7743

7744

7745

7746

7747

7748

7749

7750

7751

7752

7753

7754

7755

7756

7757

7758

7759

7760

7761

7762

7763

7764

7765

7766

7767

7768

7769

7770

7771

7772

7773

7774

7775

7776

7777

7778

7779

7780

7781

7782

7783

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Availability of Data to Inform a Distribution: Data sources to investigate for available data to inform
probability distributions of model inputs include but are not limited to the following:

•	EPA Generic Scenarios,

•	OECD Emission Scenario Documents,

•	Peer reviewed literature,

•	Published chemical assessments, and

•	Other gray literature.20

Model parameters may vary greatly in their available data. There may be a single study that provides
detailed measurements or observation data. There may be multiple studies that provide limited
measurements or observations. There may be only overall statistics available for a parameter. For a
given model development, the available data goes through a systematic review process to evaluate the
data quality, integrate the data, and decide how to use the data.

Dependency of Input Parameters on One Another: The model parameters are evaluated for any
dependency on each other. When each varied parameter is sampled according to its defined probability
distribution, they are sampled independently of each other. Therefore, the value of a sampled parameter
should be independent of the other sampled parameters. An exception is if a statistical correlation is
desired among two or more parameters. Correlating sampled parameters is discussed below in Step 3.

An example of dependency is the relationship between a facility's number of operating days, annual
production volume (PV), and daily PV. These three parameters are not all independent of each other.
The annual PV may be calculated from the daily PV and the operating days. Alternatively, the daily PV
may be calculated from the annual PV and the operating days. Additionally, operating days may be
calculated from the annual PV and daily PV. It is necessary to first understand the mathematical
relationship among these parameters before selecting parameters for which probability distributions will
be developed.

Sensitivity of the Model Results to Each Input Parameter: One consideration in selecting model
parameters for probability distribution development is the sensitivity of the model outputs to each
parameter. A sensitivity analysis can inform how sensitive each model output is to each model input
parameter. EPA may choose to prioritize probability distribution development for parameters to which
model outputs are more sensitive. Since the model outputs are more sensitive to these parameters, it
would be more important to capture variability and/or uncertainty for these parameters compared to
parameters to which model outputs are less sensitive.

A sensitivity analysis is conducted by varying each desired parameter and performing a Monte Carlo
simulation. The varied range for each parameter should be consistent with the expected range in values
for the parameter. The @Risk software (Palisade. 2022b) can perform sensitivity analyses. The statistic
of the outputs for which sensitivity is measured, such as mean, mode, or a percentile, can be selected. As
the simulation is run, the software tracks how each output changes with respect to each varied input.

20 Gray literature is defined as the broad category of data/information sources not found in standard, peer-reviewed literature
databases. Gray literature includes data/information sources such as white papers, conference proceedings, technical reports,
reference books, dissertations, information on various stakeholder websites, and various databases.

Page 320 of 484


-------
7784

7785

7786

7787

7788

7789

7790

7791

7792

7793

7794

7795

7796

7797

7798

7799

7800

7801

7802

7803

7804

7805

7806

7807

7808

7809

7810

7811

7812

7813

7814

7815

7816

7817

7818

7819

7820

7821

7822

7823

7824

7825

7826

7827

7828

7829

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Step 2: Determine a Probability Distribution

To determine a probability distribution, first, all the information known about the parameter is evaluated
(Oracle.: ). The following considerations can help guide summarizing important information about
the parameter (Analvtica. 2015):

•	Discrete or continuous

o Consider whether the parameter is discrete or continuous. Does the parameter have a
finite or countable number of possible values? Is the parameter logical or Boolean such as
having possible values of "yes or no" or "true or false"? Can the parameter be
represented by all real numbers within a domain?

•	Bounds

o Consider whether the parameter has bounds. A parameter may have a lower bound and/or
an upper bound. Alternatively, a parameter may be unbounded and can range to negative
and/or positive infinity.

•	Modes

o Consider whether the parameter has one or more modes. Does the parameter have no
mode (such as represented by a uniform distribution)? If it has a mode, is it unimodal or
multimodal? If multimodal, is the parameter a combination of two or more populations?
In which case, the parameter may be best separated into its separate components and then
develop probability distributions for the individual components.

•	Symmetric or skewed

o Consider whether the parameter is symmetric or skewed. If skewed, consider whether the
parameter is positively skewed (thicker upper tail) or negatively skewed (thicker lower
tail).

Second, review standard probability distributions and identify possible candidates that meet the
considerations identified in the first step (Oracle. 2017). The following are common probability
distributions:

•	Uniform distribution

o A uniform distribution has finite upper and lower bounds and all values between the
bounds have equal probability.

•	Triangular distribution

o A triangular distribution has finite upper and lower bounds and a modal value. The modal
value is the value that occurs most frequently. If the most frequent value is not known
another statistic, such as the mean or a percentile, could be used to define the triangular
distribution.

•	Normal distribution

o The parameters of a normal distribution are its mean and standard deviation. A normal
distribution is unbounded, and values range from negative to positive infinity. If desired,
the range of values of a normal distribution may be truncated to finite bounds to prevent
unrealistic values from being sampled.

•	Lognormal distribution

o If a variable is lognormally distributed, it means that the logarithm of that variable is
normally distributed. The parameters of a lognormal distribution are its mean and
standard deviation. A lognormal distribution is bounded from zero to positive infinity. A
lognormal distribution may be shifted and its upper bound truncated to fit the observed
data and prevent unrealistic values from being sampled.

Page 321 of 484


-------
7830

7831

7832

7833

7834

7835

7836

7837

7838

7839

7840

7841

7842

7843

7844

7845

7846

7847

7848

7849

7850

7851

7852

7853

7854

7855

7856

7857

7858

7859

7860

7861

7862

7863

7864

7865

7866

7867

7868

7869

7870

7871

7872

7873

7874

7875

7876

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Lastly, select the best suited probability distribution (Oracle. 2017). Review the available data for the
parameter to determine how to define the distribution's parameters. For example, if the only available
data are an overall range (with a minimum and a maximum), then a uniform distribution is the
appropriate distribution to use. If the only available data are an overall range and a mode, then a
triangular distribution is the appropriate distribution to use. If historical data for the parameter are
available, consider data fitting to determine the appropriate distribution and regress the distribution
parameter values.

Step 3: Check for and Define Statistical Correlations

When developing a Monte Carlo model and setting statistical distributions for parameters, EPA
evaluates possible correlations among parameters. When distributions are defined for the parameters,
each parameter is independently sampled on each iteration of the model. This may result in
combinations of parameter values that are not logical for the scenario. In the example of a model that
uses annual PV, daily PV, and operating days as parameters, there are set distributions for annual PV
and operating days, with the daily production volume calculated from the other two parameters. But
annual PV and operating days may be correlated. For example, if a site has a fixed manufacturing
capacity (as determined by the equipment size and production lines), then annual PV is a function of the
number of operating days. A facility is more likely to scale-up or scale-down their annual PV by varying
the operating days rather than varying their daily PV. Varying annual PV and operating days
independently in the model may arrive at value combinations that are not logical. For example, one
iteration may sample a high annual PV value with a low number of operating days that may result in a
high daily production rate that is not logical. In this example, a different probability distribution strategy
may be appropriate, such as defining probability distributions for daily PV and operating days since
those two parameters are likely more independent of each other than annual PV and operating days.

When developing distributions from observed data, there are statistical tests that can be performed to
indicate a statistical correlation. Two common ones are: 1) the Pearson product-moment correlation
coefficient, which measures the linear correlation between two data sets; and 2) Spearman's rank
correlation coefficient, which is a measure of rank correlation and how well a relationship between two
data sets can be described using a monotonic function. A monotonic relationship is one where the two
variables change together but not necessarily at a constant rate (Minitab. 2022). A linear correlation is
necessarily monotonic. But a monotonic correlation is not necessarily linear.

Both the Pearson and Spearman coefficients range from -1 to +1. A value close to +/-1 indicates a
strong correlation (either positive or negative). A positive correlation means as one variable increases,
the other also increases. A negative correlation means as one variable increases, the other decreases. A
value close to 0 means a weak or no correlation exists between the variables. The Pearson correlation
only measures linear relationships, and the Spearman correlation only measures monotonic
relationships. If two variables are correlated by a relationship that is neither linear nor monotonic, then
the Pearson and Spearman coefficients would not be informative of the nature of the correlation
(Minitab. 2022).

After testing for statistical correlations, statistical correlations can be defined for input parameters using
@Risk. @Risk only uses Spearman coefficients to define statistical correlations among input
parameters. Spearman coefficients to correlate two or more input parameters are defined through a
correlation matrix. The correlation matrix allows the Spearman coefficient to be defined for each pair of
correlated input parameters (Palisade. 2022a).

Page 322 of 484


-------
7877

7878

7879

7880

7881

7882

7883

7884

7885

7886

7887

7888

7889

7890

7891

7892

7893

7894

7895

7896

7897

7898

7899

7900

7901

7902

7903

7904

7905

7906

7907

7908

7909

7910

7911

7912

7913

7914

7915

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

E.10.3.3 Select Model Outputs for Aggregation of Simulation Results

The last step before running the model is to select the model outputs for which statistical results are
desired. Defining these outputs in @Risk will allow the software to save the output results from each
iteration and aggregate the simulation results over all iterations together.

E.10.3.4 Select Simulation Settings and Run Model

Simulation settings must be defined before running the model. Important simulation settings include the
number of iterations, the sampling method, and the random number generator.

•	Number of Iterations. Generally speaking, a larger number of iterations is desired to ensure
adequate sampling and representation of lower probability events. The number of iterations to
achieve a desired margin of error for a given confidence interval for an output can be calculated
using the Central Limit Theorem (OJ>n I'1 JO I \ <\tlisade. 2015a). The equation shows that the
margin of error is inversely proportional to the square root of the number of iterations. Therefore,
the greater the number of iterations, the smaller the margin of error. Calculating the number of
iterations can be difficult as the sample standard deviation is not known beforehand. EPA
typically uses 100,000 iterations to ensure convergence and have minimal cost to the simulation
time.

•	Sampling Method. The sampling method is the method used to draw random samples from the
input parameter probability distributions. @Risk uses two methods: Latin Hypercube (the
default) and Monte Carlo. Monte Carlo sampling is a purely random sampling method. This can
lead to clustering and under-representing low probability events. Latin Hypercube sampling is a
stratified sampling method. This ensures the sampled input parameter distribution matches the
assigned probability distribution closely. EPA typically uses Latin Hypercube sampling because
it is efficient and can achieve convergence with fewer iterations than Monte Carlo sampling
(Palisade. 2018).

•	Random Number Generator. The random number generator is used to generate pseudorandom
numbers that are used in an algorithm to draw random samples from the probability distributions.
The @Risk default is Mersenne Twister, which is a robust and efficient random number
generator (Palisade. 2015b).

E.10.3.5 Aggregate the Simulation Results and Produce Output Statistics

During the simulation, @Risk will save the defined model outputs for aggregation on each iteration.
After the simulation is completed, EPA can generate desired statistical results and distributions of the
defined outputs. EPA typically uses the 50th percentile and 95th percentile of the output as the central
tendency and high-end estimates, respectively.

E.ll Textile Dye Modeling Approach and Parameters for Estimating
Environmental Releases

This appendix presents the modeling approach and equations used to estimate environmental releases of
1,4-dioxane during the commercial use of textile dyes. This approach utilizes the OECD ESD on Textile
Dyes (OECD. 2017) combined with Monte Carlo simulation (a type of stochastic simulation). This ESD
includes a diagram of release and exposure points during textile dying, as shown in Figure Apx E-2.

Page 323 of 484


-------
7916

7917

7918

7919

7920

7921

7922

7923

7924

7925

7926

7927

7928

7929

7930

7931

7932

7933

7934

7935

7936

7937

7938

7939

7940

7941

7942

7943

7944

7945

7946

7947

7948

7949

7950

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Dve Formulation

© Dust Emissions
during Unloading

/Tn Disposal of Spent
Dyebath

Equipment Cleaning

©<§> Container
Residue

Cleaning and/or
Disposal

<£> Worker Exposure
During Dyeing
Operation

FigureApx E-2. Environmental Release and Occupational Exposure Points during Textile Dying

Based on Figure Apx E-2, EPA identified the following release points (note that diagram points 3 and 4
were combined for ease of analysis):

•	Release point 1 (RP1): Dust emissions of during unloading of solid powders to air, landfill,
POTW, or incineration;

•	Release point 2 (RP2): Container residual losses to POTW, landfill, or incineration; and

•	Release point 3 (RP3): Disposal of spent dye bath and equipment cleaning losses to POTW.

Environmental releases of textile dyes are a function of the chemical's physical properties, container
size, mass fractions, and other model parameters. Although physical properties are fixed, some model
parameters are expected to vary from one facility to another. An individual model input parameter could
either have a discrete value or a distribution of values. EPA assigned statistical distributions based on
available literature data or engineering judgment to address the variability in mass fraction of dye
formulation in the bath (Fdye dyebath). container size (Vcontainer), textile production rate (Vfabric), operating
days (OD), and container residue fractions (Fcontainer residue).

A Monte Carlo simulation was then conducted to capture variability in the model input parameters
described above. The simulation was conducted using the Latin hypercube sampling method in @Risk
(Palisade, Ithaca, NY). The Latin hypercube sampling method is a statistical method for generating a
sample of possible values from a multi-dimensional distribution. Latin hypercube sampling is a stratified
method, meaning it guarantees that its generated samples are representative of the probability density
function (variability) defined in the model. EPA performed 100,000 iterations of the model to capture
the range of possible input values, including values with low probability of occurrence.

From the distribution resulting from the Monte Carlo simulation, EPA selected the 95th and 50th
percentile values to represent a high-end release and central tendency release level, respectively. The
statistics were calculated directly in @Risk. The following subsections detail the model design equations
and parameters used for environmental release estimates.

E.ll.l Model Equations

Daily use rate of dye formulation is calculated using the following equation:

Equation Apx E-l

Qdye_formulation_day ~ ^fabric * ^fabric * ^dye_fabric

Page 324 of 484


-------
7951

7952

7953

7954

7955

7956

7957

7958

7959

7960

7961

7962

7963

7964

7965

7966

7967

7968

7969

7970

7971

7972

7973

7974

7975

7976

7977

7978

7979

7980

7981

7982

7983

7984

7985

7986

7987

7988

7989

7990

7991

7992

7993

7994

7995

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Where:

\ye_formulation_day

Qd,

^fabric

Pfabric
Pdye_fabric

Daily use rate of dye formulation [kg/site-day]

Textile production rate [kg/site-day]

Mass fraction of textiles treated with dye [kg/kg]

Mass fraction of dye used per mass of textile dyed [kg/kg]

Daily use rate of 1,4-dioxane formulation is calculated using the following equation:
EquationApx E-2

Qd

ioxane_site_day

Qdye_formulation_day * Pdioxane_dye * Pdye

Where:

Q dioxane_site_day
Q dye_formulation_day
Pdioxane_dye
Pdye

Daily use rate of 1,4-dioxane [kg/site-day]

Daily use rate of dye formulation [kg/site-day]

Mass fraction of 1,4-dioxane in dye formulation [kg/kg]
Fraction of dye containing 1,4-dioxane [kg/kg]

Containers emptied per facility is calculated using the following equation:
Equation Apx E-3

JV,

Qd

container _unload_site_yr

ioxane_site_day

* OD

Pdioxane_dye * ^•

container * 3.79 ^ * RHOform

Where:

Ncontainer _unload_site_yr
Qdioxane_site_day	~

OD

Pdioxane_dye	~

V,

container

RHO

form

Containers emptied per facility [containers/site-year]
Daily use rate of 1,4-dioxane [kg/site-day]

Operating days [days/year]

Mass fraction of 1,4-dioxane in dye formulation [kg/kg]
Container size [gal/container]

Dye density [kg/L]

Container residual fraction is calculated using the below equations. To make the simulation more
realistic, EPA assessed container size based on the dye formulation use rate. This avoids situations
where a small container size is associated with a large use rate, such that an unrealistic number of
containers are used each year, and vice-versa.

Equation Apx E-4

If Qdye_formulation_day ^700 kg/site-day.

Pcontainer_residue ~ Pcontainer_residue_tote

If Qdye_formulation_day 'S 200-700 kg/site-day.

Pcontainerjresidue ~ Pcontainer_residue_drum

if Qd

\ye_formulation_day

< 200 kg/site-day

container _residue

= P,

container _residue_pail

Page 325 of 484


-------
7996

7997

7998

7999

8000

8001

8002

8003

8004

8005

8006

8007

8008

8009

8010

8011

8012

8013

8014

8015

8016

8017

8018

8019

8020

8021

8022

8023

8024

8025

8026

8027

8028

8029

8030

8031

8032

8033

8034

8035

8036

8037

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Where:

Qdye_formulation_day ~
Fcontainer residue ~
Fcontainer _residue_tote ~
Fcontainer _residue_drum~
Fcontainer_residue_pail ~

Daily use rate of dye formulation [kg/site-day]
Container residual fraction [kg/kg]

Container residual fraction for totes [kg/kg]
Container residual fraction for drums [kg/kg]
Container residual fraction for pails [kg/kg]

Mass fraction of 1,4-dioxane in dye bath is calculated using the following equation:
EquationApx E-5

Where:

Fdioxane_dyebath

Fdioxane_dye

Fdye_dyebath

Fdioxane_dyebath ~ Fdioxane ciye * Fdye_dyebath

Mass fraction of 1,4-dioxane in dye bath [kg/kg]

Mass fraction of 1,4-dioxane in dye formulation [kg/kg]
Mass fraction of the dye formulation in the dyebath [kg/kg]

Release point 2 (container residual) release per day is calculated using the following equation:
Equation Apx E-6

Release_perDayRP2 = Qd

Where:

Release_perDayRP2 =

Qdioxane_site_day	~

Fcontainer _residue ~

ioxane_site_day * ^container_reSidiie

Container residual release from release point 2 [kg/site-day]
Daily use rate of 1,4-dioxane [kg/site-day]

Container residual fraction [kg/kg]

Release point 3 (spent dye bath and equipment cleaning) release per release day is calculated using the
following equation:

Equation Apx E-7

Release _perD ayRP3 = Qd

ioxane_site_day

(1 ^fixation)

Where:

Release_perDayRP 3 =

Qd

ioxane_site_day

Ffixation

Dye bath and equipment cleaning release from point 3 [kg/site-day]

Daily use rate of 1,4-dioxane [kg/site-day]

Fraction of dye affixed to textile during dye process [kg/kg]

E.11.2 Model Input Parameters

Table Apx E-12 summarizes the model parameters and their values for the Monte Carlo simulation.
Additional explanations of EPA's selection of the distributions for each parameter are provided after this
table. High-end and central tendency releases are estimated by selecting the 50th and 95th percentile
values from the output distribution.

Page 326 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

8038 Table Apx E-12. Summary of Parameter Values and Distributions Used in the Textile Release Model

Input Parameter

Symbol

Unit

Constant Model
Parameter Values

Variable Model Parameter Values

Rational/Basis

Value

Lower
Bound

Upper
Bound

Mode

Distribution

Type

Number of Sites

Ns

sites

783

-

-

-

-

See Section E.11.3

Fraction of dye
containing 1,4-dioxane

Fdye

kg/kg

1

-

-

-

-

See Section E.11.4

1,4-dioxane vapor
pressure

VP

Torr

40

-

-

-

-

Physical property

1,4-dioxane molecular
weight

MW

g/mol

88.1

-

-

-

-

Physical property

Operating days

OD

days/year

157

30

296

157

Triangular

See Section E.11.5

Mass fraction of 1,4-
dioxane in dye
formulation

Fdioxane dye

kg/kg

0.0000047









See Section E.11.6

Textile production rate

V fabric

kg/day

9,100

3,250

50,000

9,100

Triangular

See Section E.11.7

Mass fraction of textiles
treated with dye

F fabric

kg/kg

0.3

-

-

-

-

See Section E.11.8

Mass fraction of dye used
per mass of textile dyed

Fdye fabric

kg/kg

0.1

-

-

-

-

See Section E.11.9

Mass fraction of the dye
formulation in the
dyebath

Fdye dyebath

kg/kg

0.06

0.002

0.06

0.02

Triangular

See Section E.11.10

Container size for dye
formulation

V container

gal

35

7

264

35

Triangular

See Section E.ll.ll

Dye density

RHOform

kg/L

1









ESD assumes a
density equal to that
of water

Container residual
fraction for totes

Fcontainer residue totes

kg/kg

0.002

0.0007

0.002

0.0007

Triangular

See Section E.11.12

Container residual
fraction for drums

Fcontainer residue drums

kg/kg

0.03

0.0003

0.03

0.025

Triangular

See Section E.11.13

Page 327 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Input Parameter

Symbol

Unit

Constant Model
Parameter Values

Variable Model Parameter Values

Rational/Basis

Value

Lower
Bound

Upper
Bound

Mode

Distribution

Type

Container residual
fraction for pails

Fcontainer residue_pails

kg/kg

0.006

0.0003

0.006

0.003

Triangular

See Section E.11.14

Fraction of dye product
affixed to textile during
dyeing process substrate

F fixation

kg/kg

Multiple Triangular Distributions

See Section E.11.15

8039

Page 328 of 484


-------
8040

8041

8042

8043

8044

8045

8046

8047

8048

8049

8050

8051

8052

8053

8054

8055

8056

8057

8058

8059

8060

8061

8062

8063

8064

8065

8066

8067

8068

8069

8070

8071

8072

8073

8074

8075

8076

8077

8078

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

E.11.3 Number of Sites

EPA did not find data on the number of textile dye sites that specifically use dyes containing 1,4-
dioxane. As a bounding estimate, EPA used U.S. Census and BLS data for the NAICS code 313310,
Textiles and Fabric Finishing Mills, to estimate a total of 783 sites within the industry (U.S. BLS. 2016).

E.11.4 Mass Fraction of Dye Containing 1,4-Dioxane

The ESD on the Use of Textile Dyes provided a single value for the mass fraction of dyes containing the
chemical of interest, which is 1,4-dioxane. The ESD assumes that 100 percent of dyes contain the
chemical of interest (OECD. 2017). Therefore, EPA could not develop a distribution of values for this
parameter and used the single value of 1 kg dye with 1,4-dioxane/kg dye used from the ESD.

E.11.5 Operating Days	

EPA modeled the number of operating days per year using a triangular distribution with a lower bound
of 31 days per year, and upper bound of 295 days per year, and a mode of 157 days per year. This is
based on the ESD on the Use of Textile Dyes (OECD. 2017). The ESD cites the basis of these data as
past new chemical submissions that were submitted to EPA from 2006 through 2014. EPA used the 5th
percentile, average, and 95th percentile in the ESD as the lower bound, mode, and upper bound of this
distribution, respectively.

E.11.6 Mass Fraction of 1,4-Dioxane in Dye Formulation

The December 2020 Final Risk Evaluation for 1,4-Dioxane (U.S. EPA. 2020c) indicates that 1,4-
dioxane is a byproduct in dye formulations and provided a single value of 0.0000047 kg 1,4-dioxane/kg
dye. Therefore, EPA could not develop a distribution of values for this parameter and used the single
value of 0.0000047 kg 1,4-dioxane/kg dye from the risk evaluation.

E.11.7 Textile Production Rate

EPA modeled textile production rate using a triangular distribution with a lower bound of 3,520 kg/site-
day, and upper bound of 50,000 kg/site-day, and a mode of 9,100 kg/site-day. This is based on the ESD
on the Use of Textile Dyes (OECD. 2017). The ESD cites data provided in fabric finishing new
chemical submissions during 1993 and 1994. Note that the ESD uses a "typical" value as default and
does not say what the typical is based on (e.g., average, median). EPA used the range of textile
production rates and the default typical value provided in the ESD as the lower and upper bounds and
mode of the triangular distribution for this parameter, respectively.

E.11.8 Mass Fraction of Textiles Treated with Dye

The ESD on the Use of Textile Dyes provided a single value for the mass fraction of all textiles treated
with dyes. The ESD states that the median share of textiles processed per day using the primary dyestuff
is 30 percent (OECD. 2017). Therefore, EPA could not develop a distribution of values for this
parameter and used the single value of 30 percent from the ESD.

E.11.9 Mass Fraction of Dye Used per Mass of Textile Dyed

The ESD on the Use of Textile Dyes provided a single value for the mass fraction of dye used per mass
of textile dyed. The ESD states that as a "realistic worst case," liquid dye formulations are used in an
amount of 10 percent (	). Therefore, EPA could not develop a distribution of values for this

parameter and used the single value of 0.10 kg dye/kg textiles from the ESD.

Page 329 of 484


-------
8079

8080

8081

8082

8083

8084

8085

8086

8087

8088

8089

8090

8091

8092

8093

8094

8095

8096

8097

8098

8099

8100

8101

8102

8103

8104

8105

8106

8107

8108

8109

8110

8111

8112

8113

8114

8115

8116

8117

8118

8119

8120

8121

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

E.11.10 Mass Fraction of the Dye Formulation in the Dyebath

EPA modeled mass fraction of dye formulation in the dyebath using a triangular distribution with a
lower bound of 0.002 kg dye/kg bath, an upper bound of 0.06 kg dye/kg bath, and a mode of 0.02 kg
dye/kg bath. This is based on the ESD on the Use of Textile Dyes. The ESD states that typical dye
concentrations may range from 1.5 to 2.5 percent, lighter shades may be as low as 0.2 to 0.3 percent, and
heavier shades may be between 4 to 6 percent (OECD. ^ ). EPA used the overall range of dye
concentrations (0.2 to 6 percent) and the mid-range of the typical concentration (2 percent) provided in
the ESD as the lower and upper bounds and mode of the triangular distribution for this parameter,
respectively.

E.ll.ll Container Size for Dye Formulation

EPA modeled container size using a triangular distribution with a lower bound of 7 gallons, an upper
bound of 264 gallons, and a mode of 35 gallons. This is based on the ESD on the Use of Textile Dyes.
The ESD states that dyes can be transported in containers ranging from 25 kg through 1,000 kg, but
most are shipped in 35-galIon drums (OE	). EPA converted this range from kilograms to

gallons using an assumed dye density of 1 kg/L and a conversion factor of 3.785 L/gal.

E.11.12	Container Residual Fraction for Totes	

EPA modeled container residual fraction for totes using a triangular distribution with a lower bound of
0.0007 kg residual/kg dye, and upper bound of 0.002 kg residual/kg dye, and a mode of 0.0007 kg
residual/kg dye. The lower and upper bounds of this distribution are based on the central tendency and
high-end values listed in the EPA/OPPT Bulk Transport Residual Model from the ChemSTEER User
Guide (	). EPA used the central tendency value as the mode of the triangular

distribution.

E.11.13 Container Residual Fraction for Drums

EPA modeled container residual fraction for drums using a triangular distribution with a lower bound of
0.0003 kg residual/kg dye, an upper bound of 0.03 kg residual/kg dye, and a mode of 0.025 kg
residual/kg dye. The lower bound is based on the minimum value for pouring and the upper bound is
based on the default high-end value in the EPA/OPPT Drum Residual Model from the ChemSTEER
User Guide (	). EPA used the central tendency value for pumping as the mode of the

triangular distribution.

E.11.14 Container Residual Fraction for Pails

EPA modeled container residual fraction for small containers using a triangular distribution with a lower
bound of 0.0003 kg residual/kg dye, an upper bound of 0.006 kg residual/kg dye, and a mode of 0.003
kg residual/kg dye. The lower bound is based on the minimum value for pouring and the upper bound is
based on the default high-end value listed in the EPA/OPPT Small Container Residual Model from the
ChemSTEER User Guide (	). EPA used the central tendency value for pouring as the

mode of the triangular distribution.

E.11.15 Fraction of Dye Product Affixed to Textile during Dyeing Process Substrate

EPA modeled the fraction of dye product affixed to textiles during dyeing process substrate using
multiple triangular distributions. The ESD on the Use of Textile Dyes provides a table containing ranges
and averages for dye fixation percentages based on the nine different classes of dyes (OECD. 2017).
EPA used the low-end of the range of dye fixation as the lower bound, the high-end of the range of dye
fixation as the upper bound, and the average dye fixation as the mode for each of the nine triangular
distributions. In the Monte Carlo simulation, each of the nine triangular distributions from the ESD has

Page 330 of 484


-------
8122

8123

8124

8125

8126

8127

8128

8129

8130

8131

8132

8133

8134

8135

8136

8137

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

an equal probability of being selected and used for the parameter's output. The distribution selection
probabilities and values are shown in Table Apx E-13.

Table Apx E-13. Triangular Distributions Ffixation

Dye Type

Dye Fixation (%) Triangular Distribution

Lower Bound

Upper Bound

Mode

Acid

85

98

93

Azoic

76

95

84

Basic

95

100

99

Direct

64

96

88

Disperse

80

100

96

Metal-
Containing

82

98

94

Reactive

50

97

85

Sulfur

60

95

70

Vat

70

95

85

Source: ("OECD. 2017)

E.12 Laundry Detergent Modeling Approach and Parameters for
Estimating Environmental Releases

This appendix presents the modeling approach and equations used to estimate environmental releases of
1,4-dioxane during the industrial and institutional use of laundry detergents. This approach utilizes the
OECD ESD on the Chemicals Used in Water Based Washing Operations at Industrial and Institutional
Laundries (OECD. i ) combined with Monte Carlo simulation (a type of stochastic simulation).

This ESD categorizes laundry facilities into either industrial or institutional facilities and includes
different process parameters for each. Therefore, EPA modeled the two types of laundry facilities
separately. In addition, laundry detergents can be in liquid or powder physical forms. Because the
difference in physical form results in different parameter distributions, EPA modeled liquid and powder
detergents separately. This ESD includes a diagram of release and exposure points during textile dying,
as shown in Figure Apx E-3.

Page 331 of 484


-------
8138

8139

8140

8141

8142

8143

8144

8145

8146

8147

8148

8149

8150

8151

8152

8153

8154

8155

8156

8157

8158

8159

8160

8161

8162

8163

8164

8165

8166

8167

8168

8169

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

©©©Container Residue and Cleaning ©©©Container Residue and Cleaning
©@ Fugitive Air Release and Dust	©@ Fugitive Air Release and Dust

Emissions During Transfers	Emissions During Transfers

©@© Releases During
Operations

FigureApx E-3. Environmental Release and Occupational Exposure Points during
Industrial/Institutional Laundering Operations

Based on Figure Apx E-3, EPA identified the following release points:

•	Release point 1 (RP1): Container residual losses to POTW, landfill, or incineration;

•	Release point 2 (RP2): Fugitive air releases during container cleaning;

•	Release point 3 (RP3): Fugitive air releases during container unloading;

•	Release point 4 (RP4): Dust releases during container unloading;

o 4a: Uncaptured dust releases;
o 4b: Captured, uncontrolled dust releases;
o 4c: Captured and controlled dust releases;

•	Release point 5 (RP5): Fugitive air releases during washing; and

•	Release point 6 (RP6): Release from washing and drying operations to fugitive air, stack air, or
POTW.

Environmental releases of laundry detergent are a function of the chemical's physical properties,
container size, mass fractions, and other model parameters. Although physical properties are fixed, some
model parameters are expected to vary from one facility to another. An individual model input
parameter could either have a discrete value or a distribution of values. EPA assigned statistical
distributions based on available literature data or engineering judgment to address the variability in mass
fraction of 1,4-dioxane in the detergent (Fdioxane laundry), container size (Vcontainer), daily use rate of
detergent (Qfaciiity day), air speed (RATEair speed), duration of release (OHCOnt unload), operating days (OD),
container residue fractions (Fcontainer residue), and dust capture/control efficiency (Fdust).

A Monte Carlo simulation was conducted to capture variability in the model input parameters described
above. The simulation was conducted using the Latin hypercube sampling method in @Risk (Palisade,
Ithaca, New York). The Latin hypercube sampling method is a statistical method for generating a
sample of possible values from a multi-dimensional distribution. Latin hypercube sampling is a stratified
method, meaning it guarantees that its generated samples are representative of the probability density
function (variability) defined in the model. EPA performed 100,000 iterations of the model to capture
the range of possible input values, including values with low probability of occurrence.

Page 332 of 484


-------
8170

8171

8172

8173

8174

8175

8176

8177

8178

8179

8180

8181

8182

8183

8184

8185

8186

8187

8188

8189

8190

8191

8192

8193

8194

8195

8196

8197

8198

8199

8200

8201

8202

8203

8204

8205

8206

8207

8208

8209

8210

8211

8212

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

From the distribution resulting from the Monte Carlo simulation, EPA selected the 95th and 50th
percentile values to represent a high-end release and central tendency release level respectively. The
statistics were calculated directly in @Risk. The following subsections detail the model design equations
and parameters used for environmental release estimates.

E.12.1 Model Equations	

Daily use rate selection based on physical form of detergent is based on the following two equations, the
first being for liquid detergent and the second being for powder detergent:

EquationApx E-8

Qfacility_day ~ Qfacility_day_liquid

or

Qfacility_day ~ Qfacility_day_powder

Where:

Qf acility_day	= Daily use rate based on physical form of detergent [kg/site-day]

Qf acility dayUquid = Daily use for liquid form detergent [kg/site-day]
Qf acility_day_powder = Daily use for powder form detergent [kg/site-day]

Daily use rate of laundry detergents containing 1,4-dioxane is calculated using the equation below:

Equation Apx E-9

Q facility _day_adjusted ~ Qf acility_day * Fformulations_dioxane

Where:

Qfacility_day_adjusted = Daily use rate of detergent containing 1,4-dioxane selected based

on the physical form of the detergent [kg/site-day]

Qf acility _day	= Daily use rate based on physical form of detergent [kg/site-day]

Pformuiations_dioxane = Fraction of laundry detergents containing 1,4-dioxane [kg/kg]

Daily use rate of 1,4-dioxane is calculated using the equation below:

Equation Apx E-10

Qdioxane_day ~ Qfacility_day_adjusted * Pdioxanejaundry

Where:

Qdioxane_day	= Daily usage rate of 1,4-dioxane [kg/site-day]

Qf acility _day_ad justed = Daily use rate of detergent with 1,4-dioxane [kg/site-day]

Fdioxanejaundry = Mass fraction of 1,4-dioxane in laundry detergent [kg/kg]

Number of containers used per year is calculated using the equation below:

Page 333 of 484


-------
8213

8214

8215

8216

8217

8218

8219

8220

8221

8222

8223

8224

8225

8226

8227

8228

8229

8230

8231

8232

8233

8234

8235

8236

8237

8238

8239

8240

8241

8242

8243

8244

8245

8246

8247

8248

8249

8250

8251

8252

8253

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

EquationApx E-ll

N,

Q facility j

day_ad justed

* OD

cont_site_yr

V,

container

* gal * ^^^detergent

Where:

Ncont_site_yr
Q facility _day _ad justed

OD

^container

RHO

detergent

Number of containers used per site per year [containers/site-year]
Daily use rate of detergent with 1,4-dioxane [kg/site-day]
Operating days [days/year]

Container volume [gal/container]

Detergent density [kg/L]

Vapor pressure correction factor for release points 2 and 3 is calculated using the equation below:
Equation Apx E-12

Fdioxane

X,

e laundry/
/,

MW

cleanjinload

Fdioxane Jaundry , 1 FdioxaneJaundry

MW +	18

Where:

X,

clean unload

Fd ioxane Jaundry

MW

Vapor pressure correction factor for release points 2 and 3
[mol 1,4-dioxane/mol water]

Mass fraction of 1,4-dioxane in detergent [kg/kg]
1,4-dioxane molecular weight [g/mol]

Fraction of 1,4-dioxane in wash water is calculated using the equation below:
Equation Apx E-13

Fni

dioxanejwash

~ Fdilution * Fdioxanejaunciry

Where:

Fd ioxane _w ash

Fdilution

Fd ioxane Jaundry

Fraction of 1,4-dioxane in wash water [kg 1,4-dioxane/kg water]
Dilution factor for detergent in the wash [unitless]

Mass fraction of 1,4-dioxane in detergent [kg/kg]

Vapor pressure correction factor for release point 5 is calculated using the equation below:
Equation Apx E-14

Fdioxanejwash /

'MW

X,

washing

Fdioxanejwash _j_ ^	 ^df'o:mr!e_was/i

MW

18

Where:

Xwashing
Fd ioxane_wash

MW

Vapor pressure correction factor for release point 5
[mol 1,4-dioxane/mol water]

Fraction of 1,4-dioxane in wash water [kg 1,4-dioxane/kg water]
1,4-dioxane molecular weight [g/mol]

Page 334 of 484


-------
8254

8255

8256

8257

8258

8259

8260

8261

8262

8263

8264

8265

8266

8267

8268

8269

8270

8271

8272

8273

8274

8275

8276

8277

8278

8279

8280

8281

8282

8283

8284

8285

8286

8287

8288

8289

8290

8291

8292

8293

8294

8295

8296

8297

8298

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Container residual fraction is calculated using the following equations. To make the simulation more
realistic, EPA assessed container size based on the detergent use rate. This avoids situations where a
small container size is associated with a large use rate, such that an unrealistic number of containers are
used each year, and vice-versa:

EquationApx E-15

If Qfacility_day > 600 ~ '

site—day

Pcontainer residue ~ Pcontainer residue tote

KQ facility .day = 200 - 600^^'

site—day

Pcontainer residue ~ Pcontainer residue drum

If <3facility_day < 200 sitJday-

Pcontainer_residue ~ Pcontainer_residue_pail

If physical form of detergent is powder:

Pcontainer_residue ~ Pcontainer_residue_powder

Where:

Qfacility_day	=	Daily use rate based on physical form of detergent [kg/site-day]

pcontainer residue =	Container residual fraction [kg/kg]

p'container residuejote =	Container residual fraction for totes [kg/kg]

pcontainer residue_drum=	Container residual fraction for drums [kg/kg]

pcontainer residue_Vaii =	Container residual fraction for pails [kg/kg]

p'container residue_Powder	= Container residual fraction for solid detergents [kg/kg]

Release Point 1 site release per day is calculated using the equation below:

Equation Apx E-16

Release_perDayRp1 — Qdioxane_day * Pcontainerjresidue

Where:

Release_perDayRP1 = Daily 1,4-dioxane release at release point 1 [kg/site-day]
Qdioxane_day	= Daily usage rate of 1,4-dioxane [kg/site-day]

pcontainer residue = Container residual fraction [kg/kg]

Release Point 2 fugitive emissions from container cleaning for pails and drums per day is calculated using
the Penetration Model equation below (air speed <100 ft/min):

Equation Apx E-17

Release jper DayRP2 =

s ka x ) * (MW ) * Xcleanun[oad * VP *	* (0-257lDcontainer_opening)

3600— * 0.001— *	 y		

fir	a	t0-05 * ff)	.[p

a	i	<\j ^container_o~pening *1

Where:

Page 335 of 484


-------
8299

8300

8301

8302

8303

8304

8305

8306

8307

8308

8309

8310

8311

8312

8313

8314

8315

8316

8317

8318

8319

8320

8321

8322

8323

8324

8325

8326

8327

8328

8329

8330

8331

8332

8333

8334

8335

8336

8337

8338

8339

8340

8341

8342

8343

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Release_perDayRP2 =
MW

X clean_unload	~

VP
T

R(Xt6airspeed	~

Dcontainer _opening ~

Release point 2 fugitive emissions from pail/drum cleaning

per day [kg/site-day]

1,4-dioxane molecular weight [g/mol]

Vapor pressure correction factor release point 2

[mol 1,4-dioxane/mol water]

1,4-dioxane vapor pressure [torr]

Ambient temperature [K]

Air speed [cm/s]

Diameter of container opening [cm]

Atmospheric pressure [atm]

Release Point 2 fugitive emissions from container cleaning per day for totes is calculated using the Mass
Transfer Coefficient Model equation below (air speed >100 ft/min):

EquationApx E-18

Release jper DayRP2 =

3600—*0.001
hr	g

kg (1-93 x 10 7) * (MVK0'78) * Xcleanun[oad * VP * RO-tSair_speed *	containerjopenlng) J29 ^ MW

Where:

Release_perDayRP2 =

X clean_unload	~

MW

VP

T

R(Xt6airspeed	~

Dcontainer _opening ~

j10.4 n 0.11

container _opening

(VT - 5.87)2/3

Release point 2 fugitive emissions from tote cleaning per day
[kg/site-day]

Vapor pressure correction factor release point 5
[mol 1,4-dioxane/mol water]

1,4-dioxane molecular weight [g/mol]

Vapor Pressure [torr]

Ambient Temperature [K]

Air speed [cm/s]

Diameter of container opening [cm]

Atmospheric pressure [atm]

Release Point 3 fugitive emissions from unloading of pails and drums during the day is calculated using
the Penetration Model equation below (air speed <100 ft/min):

Equation Apx E-19

Release jper DayRP3 =

OHcont_unload * 3600 — * 0.001— *

kg (8.24 x lO"8) * (WJS) * Xcleanunload * VP * jRateairspeea * (0.25*#,

)1- + —

container _opemng J ..I 29 MW

'VA

container_opening

VP

Where:

Release_perDayRP3

X clean_unload

MW

VP

T

Point 3 fugitive emissions from unloading during the day
[kg/site-day]

Vapor pressure correction factor release point 5
[mol 1,4-dioxane/mol water]

1,4-dioxane molecular weight [g/mol]

Vapor pressure [torr]

Ambient temperature [K]

Page 336 of 484


-------
8344

8345

8346

8347

8348

8349

8350

8351

8352

8353

8354

8355

8356

8357

8358

8359

8360

8361

8362

8363

8364

8365

8366

8367

8368

8369

8370

8371

8372

8373

8374

8375

8376

8377

8378

8379

8380

8381

8382

8383

8384

8385

8386

8387

8388

8389

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Rateair_Speed	= Air speed from EPA model [cm/s]

Dcontainer opening = Diameter of the opening for containers [cm]
P	= Atmospheric pressure [atm]

OHcont unload	= Duration of container unloading [hrs/day]

Release Point 3 fugitive emissions from unloading totes during the day is calculated using the Mass
Transfer Coefficient Model equation below (air speed >100 ft/min):

EquationApx E-20

Release jper DayRP3 =

s kq (193 x 10 7) * (MW0,78) * Xcieanunload * VP * Rate°Jr8speed * (0.25?lD2container opening)\js +
OHcont_unload3600—* O.OOl^*	3	

hr ' g	7-o.40o.il	(^/T-5.87)2^

1 ^container_opemng\v 1	j

Where:

Release_perDayRP3 = Point 3 fugitive emissions from unloading during the day

[kg/site-day]

Xdeanunioad	= Vapor pressure correction factor release point 5

[mol 1,4-dioxane/mol water]
MW	= 1,4-dioxane molecular weight [g/mol]

VP	= Vapor pressure [torr]

T	= Ambient temperature [K]

Rateair Speed	= Air speed from EPA model [cm/s]

Dcontainer opening = Diameter of the opening for containers [cm]
P	= Atmospheric pressure [atm]

OHcont unload	= Duration of container unloading [hours/day]

Release Point 4a dust not captured to fugitive air, water, incineration, or landfill is calculated using the
following equation:

Equation Apx E-21

Release_perDayRPAa — Qdioxane_day * Fdustgenerati0n * Fdustcapture)

Where:

Release_perDayRP4a= Dust not captured to fugitive air, water, incineration, or landfill

[kg/site-day]

Fdustgeneration	= Fraction of chemical lost during transfer of solid powders [kg/kg]

Fdustcapture	= Capture efficiency for dust capture methods [kg/kg]

Release Point 4b dust captured but not controlled to stack air is calculated using the following equation:
Equation Apx E-22

Release_perDayRP4b = Qdioxane_day * Fdustgeneration * Fdustcapture * (1 — Fdustcontrol)

Where:

Release_perDayRP4b= Dust captured but not controlled to stack air [kg/site-day]
Fdustgeneration	= Fraction of chemical lost during transfer of solid powders [kg/kg]

Fdustcavture	= Capture efficiency for dust capture methods [kg/kg]

Page 337 of 484


-------
8390

8391

8392

8393

8394

8395

8396

8397

8398

8399

8400

8401

8402

8403

8404

8405

8406

8407

8408

8409

8410

8411

8412

8413

8414

8415

8416

8417

8418

8419

8420

8421

8422

8423

8424

8425

8426

8427

8428

8429

8430

8431

8432

8433

8434

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Fdustcontroi	= Control efficiency for dust control methods [kg/kg]

Release Point 4c dust captured and controlled to incineration of landfill is calculated using the following
equation:

EquationApx E-23

Release_perDayRP4b = Qdioxane_day * Fdustgeneration * Fdustcavture * Fdustcontroi

Where:

Release_perDayRP4b= Dust captured but not controlled to stack air [kg/site-day]
Fdustgeneration	= Fraction of chemical lost during transfer of solid powders [kg/kg]

Fdustcapture	= Capture efficiency for dust capture methods [kg/kg]

Fdustcontroi	= Control efficiency for dust control methods [kg/kg]

Release Point 5 fugitive emissions during washing per day is calculated when air speed <100 ft/min using
the Penetration Model in the equation shown below:

Equation Apx E-24

Release jper DayRPS =

(8.24 x 10~8) * (Miy0,835) * Xcleanunload *VP* lRateairspeed * (025uDtontainer_opening)% + ^

OH * 3600— * 0.001— *	 N		

hr	q	T005 * fn	¦ \fP

u	V container_opemng v 1

Where:

Release_perDayRP5	=	Point 5 fugitive emissions from washing [kg/site-day]

Xdean unload	=	Vapor pressure correction factor release point 5

[mol 1,4-dioxane/mol water]

MW	=	1,4-dioxane molecular weight [g/mol]

VP	=	Vapor pressure [torr]

T	=	Ambient temperature [K]

Rateair speed	=	Air speed [cm/s]

Dcontainer opening	=	Diameter of the opening for containers [cm]

P	=	Atmospheric pressure [atm]

OH	=	Operating hours [hours/day]

Release Point 5 fugitive emissions during washing per day is calculated when air speed >100 ft/min using
the Mass Transfer Coefficient Model shown below:

Equation Apx E-25

Release jper DayRP5 =

s kq (193 x 10 7) * (MW0,78) * Xcieanunload * VP * Rate0Jr8speed * (0.25?lD2container opening)\js +
OHcont_unload3600—* 0.001^*	3	

hr ' g	ro.40O.ii	(Vf- 5.87)2/3

1 ^container_opemng\v 1	j

Where:

Release_perDayRP5 = Point 5 fugitive emissions from washing [kg/site-day]
Xdeanunioad	= Vapor pressure correction factor release point 5

[mol 1,4-dioxane/mol water]
MW	= 1,4-dioxane molecular weight [g/mol]

Page 338 of 484


-------
8435

8436

8437

8438

8439

8440

8441

8442

8443

8444

8445

8446

8447

8448

8449

8450

8451

8452

8453

8454

8455

8456

8457

8458

8459

8460

8461

8462

8463

8464

8465

8466

8467

8468

8469

8470

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

VP	=	Vapor pressure [torr]

T	=	Ambient temperature [K]

Rateair speed	=	Air speed [cm/s]

Dcontainer opening	=	Diameter of the opening for containers [cm]

P	=	Atmospheric pressure [atm]

OH	=	Operating hours [hours/day]

Release Point 6 site release per day (washing and drying) is calculated using the equations and criteria
below:

EquationApx E-26

If £f=i Release_perDayRPi < Qdioxane_day:

5

Release _per Day RP6 Qdioxane_day I Release_perDayRPi

i=1

If £f=i Release_perDayRPi > Qdioxane_day:

Liquid detergent:

Release.jperDayRP6 = Qdioxane_day ~ Release jperD ayRP1

Powder detergent:

Release _perD ay RPb Qdioxane_day ReleasepgrDay RP^ Release _per Day RPA

Point 1 container residual releases [kg/site-day]

Point 2 fugitive emissions from container cleaning [kg/site-day]
Point 3 fugitive emissions from unloading [kg/site-day]

Point 4 fugitive dust emissions [kg/site-day]

Point 5 fugitive emissions from washing [kg/site-day]

Point 6 daily site releases (washing and drying) [kg/site-day]
Daily usage rate of 1,4-dioxane [kg/site-day]

The sum of release points 1-5 emissions [kg/site-day]

Where:

Release_perDayRP1 =
Release_perDayRP2 =
Release_perDayRP3 =
Release_perDayRP4 =
Release_perDayRPS =
Release_perDayRP6 =
Qdioxane_day	~

£f=1 Release _perD ay RPi =

E.12.2 Model Input Parameters	

Table Apx E-14 summarizes the model parameters and their values for the Monte Carlo simulation.
Additional explanations of EPA's selection of the distributions for each parameter are provided after this
table. High-end and central tendency releases are estimated by selecting the 50th and 95th percentile
values from the output distribution.

Page 339 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Table Apx E-14. Summary of Paramel

ter Values and Distributions Used in the Industrial and Institutional Laundry E

please Model

Input Parameter

Symbol

Unit

Constant Model
Parameter Values

Variable Model Parameter Values

Rational/ Basis

Value

Lower
Bound

Upper
Bound

Mode

Distribution
Type

Operating Days

OD

days/year

260

Industrial:

19

Institutional:
249

366

260

Triangular

See Section
E.12.3

Mass Fraction of
1,4-dioxane in
Laundry Detergent

Fdioxane laundry

kg/kg

0.000014

5.00E-08

0.000014



Uniform

See Section
E.12.4.

Daily Use Rate of
Liquid Laundry
Detergents

Qfacility day liquid

kg/day

Industrial:
35.7

Institutional:
16

Industrial:
0.116

Institutional:
0.124

Industrial:
814

Institutional:
513



Discrete

See Section
E.12.5

Daily Use Rate of
Powder Laundry
Detergents

Qfacility day_powder

kg/day

Industrial:
110.45
Institutional:
8.63

Industrial:
1.33

Institutional:
3.71

Industrial:
1,917.44
Institutional:
15



Discrete

See Section
E.12.5

Container Size

V container

gal

55

5

550

55

Triangular

See Section
E.12.6

Air Speed

RATEair speed

cm/s

10

1.3

202.2

—

Lognormal

See Section
E.12.7

Container residual
fraction for Totes

Fcontainer residue totes

kg/kg

0.002

0.0002

0.002

0.0007

Triangular

See Section
E.12.8

Container residual
fraction for Drums

Fcontainer residue drums

kg/kg

0.03

0.017

0.03

0.025

Triangular

See Section
E.12.9

Container residual
fraction for Pails

Fcontainer residue_pails

kg/kg

0.006

0.0003

0.006

0.003

Triangular

See Section
E.12.10

Container residual
fraction for Powders

Fcontainer residue_powders

kg/kg

0.01

—

—

—

—

See Section
E.12.11

Fraction of Laundry
Detergents
Containing 1,4-
dioxane

Fformulations dioxane

unitless

0.5

0.111

1



Industrial:
Discrete

Institutional:

See Section
E.12.12

Page 340 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Input Parameter

Symbol

Unit

Constant Model
Parameter Values

Variable Model Parameter Values

Rational/ Basis

Value

Lower
Bound

Upper
Bound

Mode

Distribution
Type















Uniform



Duration of Release
for Container
Unloading

OUcont unload

h/day

Industrial:
0.0043
Institutional:
0.0114

Industrial:
0.0043
Institutional:
0.0114

Industrial:

12

Institutional:
8



Uniform

See Section
E. 12.13

Fraction of chemical
lost during transfer
of solid powders

F dust_generation

kg/kg

0.0050

0.0010

0.03

0.005

Triangular

See Section
E. 12.14

Control efficiency
for dust control
methods

Fdust control

kg/kg

0.7900

0.0000

1

0.79

Triangular

See Section
E. 12.15

Capture efficiency
for dust capture
methods

Fdust capture

kg/kg

0.9633

0.9310

1

0.9633

Triangular

See Section
E. 12.16

Number of Sites

Ns

sites

Industrial:

2,453

Institutional:

95,533









See Section
E. 12.17

Vapor Pressure of
1,4-dioxane

VP

Torr

40

—

—

—

—

Physical
property

Molecular Weight of
1,4-dioxane

MW

g/mol

88.1

—

—

—

—

Physical
property

Diameter of
Container Opening

Dcontainer opening

cm

5.08

—

—

—

—

See Section
E. 12.18

Diameter of Wash
Opening

Dwash opening

cm

73

—

—

—

—

See Section
E. 12.19

Ambient
Temperature

T

K

298

—

—

—

—

Process
parameter

Ambient Pressure

P

atm

1

—

—

—

—

Process
parameter

Page 341 of 484


-------
PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

Input Parameter

Symbol

Unit

Constant Model
Parameter Values

Variable Model Parameter Values

Rational/ Basis

Value

Lower
Bound

Upper
Bound

Mode

Distribution
Type

Dilution Factor

F dilution

unitless

0.016

—

—

—

—

See Section
E. 12.20

Density of Laundry
Detergent

RHOform

kg/L

1









ESD assumes a
density equal to
that of water

Container Fill Rate

RATEfin

containers
/ hour

20

—

—

—

—

See Section
E.12.21

8472

Page 342 of 484


-------
8473

8474

8475

8476

8477

8478

8479

8480

8481

8482

8483

8484

8485

8486

8487

8488

8489

8490

8491

8492

8493

8494

8495

8496

8497

8498

8499

8500

8501

8502

8503

8504

8505

8506

8507

8508

8509

8510

8511

8512

8513

8514

8515

8516

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

E.12.3 Operating Days

EPA modeled the operating days per year using a triangular distribution with a lower bound of 20 days
per year, an upper bound of 365 days per year, and a mode of 260 days per year for industrial laundries.
EPA used a triangular distribution with a lower bound of 250 days per, an upper bound of 365 days per
year, and a mode of 260 days per year for institutional laundries. This is based on the ESD on the
Chemicals Used in Water Based Washing Operations at Industrial and Institutional Laundries (
2i ). The ESD provides the range and average of operating days for six separate years, which EPA
took the minimum, maximum, and average of the 6 years to form the distributions.

E.12.4 Mass Fraction of 1,4-Dioxane in Laundry Detergent

EPA modeled the mass fraction of 1,4-dioxane in laundry detergent using a uniform distribution with a
lower bound of 5.00x 10~8 kg of 1,4-dioxane/kg detergent and an upper bound of 1.4/ 10 5 kg of 1,4-
dioxane/kg detergent for both industrial and institutional laundries. This is based on the December 2020
Final Risk Evaluation for 1,4-Dioxane (	,020c). The risk evaluation indicates that 1,4-dioxane

is a byproduct in the laundry detergents and provides a concentration range of 0.05 to 14 ppm of 1,4-
dioxane in the detergent.

E.12.5 Daily Use Rate of Detergent	

EPA modeled the daily use rate of detergent using a discrete distribution. For industrial laundries, the
distribution ranged from 0.116 kg/day to 814 kg/day for liquid detergents and 1.33 kg/day to 1917.44
kg/day for powder detergents. For institutional laundries, the distribution ranged from 0.124 kg/day to
513 kg/day for liquid detergents and 3.71 kg/day to 15 kg/day for powder detergents. This discrete data
was pulled from survey data from laundries sites used in the ESD on the Chemicals Used in Water
Based Washing Operations at Industrial and Institutional Laundries (OECD. ^ ). Equal probability
was given to each discrete survey value.

E.12.6 Container Size

EPA modeled container size using a triangular distribution with a lower bound of 5 gallons, an upper
bound of 550 gallons, and a mode of 55 gallons for industrial laundries. Because EPA expects industrial
laundries to have variation in the sizes of containers, EPA used values of 5, 55, and 550 gallons for the
triangular distribution based on the default values from the EPA/OPPT Small Container Residual Model,
Drum Residual Model, and Bulk Transport Residual Model, respectively.

EPA used a single value of 5 gallons for institutional laundries. This is based on the ESD on the default
value for institutional laundries from the Chemicals Used in Water Based Washing Operations at
Industrial and Institutional Laundries ESD (OEC	).

E.12.7 Indoor Air Speed

Baldwin (1998) measured indoor air speeds across a variety of occupational settings in the United
Kingdom. Fifty-five work areas were surveyed across a variety of workplaces. EPA analyzed the air
speed data from Baldwin (1998) and categorized the air speed surveys into settings representative of
industrial facilities and representative of commercial facilities. EPA fit separate distributions for these
industrial and commercial settings and used the industrial distribution for laundry facilities.

EPA fit a lognormal distribution for both data sets as consistent with the authors observations that the air
speed measurements within a surveyed location were lognormally distributed and the population of the
mean air speeds among all surveys were lognormally distributed. Since lognormal distributions are
bound by zero and positive infinity, EPA truncated the distribution at the largest observed value among

Page 343 of 484


-------
8517

8518

8519

8520

8521

8522

8523

8524

8525

8526

8527

8528

8529

8530

8531

8532

8533

8534

8535

8536

8537

8538

8539

8540

8541

8542

8543

8544

8545

8546

8547

8548

8549

8550

8551

8552

8553

8554

8555

8556

8557

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

all of the survey mean air speeds from Baldwin (1998). EPA fit the air speed surveys representative of
industrial facilities to a lognormal distribution with the following parameter values: mean of 22.414
cm/s and standard deviation of 19.958 cm/s. In the model, the lognormal distribution is truncated at a
maximum allowed value of 202.2 cm/s (largest surveyed mean air speed observed in Baldwin (1998)) to
prevent the model from sampling values that approach infinity or are otherwise unrealistically large.

Baldwin (1998) only presented the mean air speed of each survey. The authors did not present the
individual measurements within each survey. Therefore, these distributions represent a distribution of
mean air speeds and not a distribution of spatially variable air speeds within a single workplace setting.

E.12.8 Container Residual Fraction for Totes

EPA modeled container residual fraction for totes using a triangular distribution with a lower bound of
0.0007 kg residual/kg detergent, and upper bound of 0.002 kg residual/kg detergent, and a mode of
0.0007 kg residual/kg detergent. The lower and upper bounds of this distribution are based on the central
tendency and high-end values listed in the EPA/OPPT Bulk Transport Residual Model from the
ChemSTEER User Guide (	). EPA used the central tendency value as the mode of the

triangular distribution.

E.12.9 Container Residual Fraction for Drums

EPA modeled container residual fraction for drums using a triangular distribution with a lower bound of
0.0003 kg residual/kg detergent, an upper bound of 0.03 kg residual/kg detergent, and a mode of 0.025
kg residual/kg detergent. The lower bound is based on the minimum value for pouring and the upper
bound is based on the default high-end value in the EPA/OPPT Drum Residual Model from the
ChemSTEER User Guide (	15a). EPA used the central tendency value for pumping as the

mode of the triangular distribution.

E.12.10	Container Residual Fraction for Pails	

EPA modeled container residual fraction for small containers using a triangular distribution with a lower
bound of 0.0003 kg residual/kg detergent, an upper bound of 0.006 kg residual/kg detergent, and a mode
of 0.003 kg residual/kg detergent. The lower bound is based on the minimum value for pouring and the
upper bound is based on the default high-end value listed in the EPA/OPPT Small Container Residual
Model from the ChemSTEER User Guide (	) EPA used the central tendency value for

pouring as the mode of the triangular distribution.

E.12.11 Container Residual Fraction for Powders

The EPA/OPPT Solid Residuals in Transport Containers Model provides a loss fraction 0.01 kg of solid
chemicals remaining in a container per kg transported. Therefore, EPA could not develop a distribution
of values for this parameter and used the single value 0.01 kg/kg from the model (	|).

E.12.12 Fraction of Laundry Detergents Containing 1,4-Dioxane

EPA modeled the fraction of laundry detergents containing 1,4-dioxane using a discrete distribution. For
industrial and institutional laundries, the distribution ranged from 0.111 to 1 kg detergents containing
1,4-dioxane/kg all detergents. This discrete data was pulled from survey data from laundries sites used
in the ESD on the Chemicals Used in Water Based Washing Operations at Industrial and Institutional
Laundries (OEC	). Equal probability was given to each discrete survey value.

Page 344 of 484


-------
8558

8559

8560

8561

8562

8563

8564

8565

8566

8567

8568

8569

8570

8571

8572

8573

8574

8575

8576

8577

8578

8579

8580

8581

8582

8583

8584

8585

8586

8587

8588

8589

8590

8591

8592

8593

8594

8595

8596

8597

8598

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

E.12.13 Duration of Release for Container Unloading

EPA modeled the duration of release for container unloading using a uniform distribution. For industrial
and institutional laundries, EPA assumed the distribution had a maximum of 12 hours/day. The lower
bound was based on the length of time to unload detergent containers each day, calculated using the
number of containers used per day and the container fill rate (see Section E. 12.21). This means that each
iteration of the simulation would calculate a new lower bound based on the parameters for that iteration.

E.12.14 Fraction of Chemical Lost during Transfer of Solid Powders

EPA modeled the fraction of chemical lost during transfer of solid powders using a triangular
distribution with a lower bound of 0.001 kg dust lost/kg transferred, an upper bound of 0.03 kg dust
lost/kg transferred, and a mode of 0.005 kg dust lost/kg transferred for both industrial and institutional
laundries. These values were taken from the EPA/OPPT Dust Emissions from Transferring Solids Model
from the ChemSTEER User Guide (	015a).

E.12.15 Control Efficiency for Dust Control Methods

EPA modeled the control efficiency for dust control methods using a triangular distribution with a lower
bound of 0 kg controlled/kg transferred, an upper bound of 1 kg controlled/kg transferred, and a mode of
0.79 kg controlled/kg transferred for both industrial and institutional laundries. These values were taken
from the EPA/OPPT Dust Emissions from Transferring Solids Model from the ChemSTEER User Guide
(I	>015a).

E.12.16 Capture Efficiency for Dust Capture Methods

EPA modeled the capture efficiency for dust capture methods using a triangular distribution with a
lower bound of 0.9310 kg captured/kg transferred, an upper bound of 1 kg captured/kg transferred, and a
mode of 0.9633 kg captured/kg transferred for both industrial and institutional laundries. These values
were taken from the EPA/OPPT Dust Emissions from Transferring Solids Model from the ChemSTEER
User Guide (	).

E.12.17	Number of Sites	

EPA did not find data on the number of laundry sites that specifically use detergents containing 1,4-
dioxane. As a bounding estimate for the number of industrial laundries, EPA used U.S. Census and BLS
data for the NAICS code 812330, Linen and Uniform Supply, to estimate a total of 2,453 industrial
laundry sites within the industry (\' S HI S 101 * -). As a bounding estimate for the number of
institutional sites, EPA used industry information as described in the ESD to estimate a total of 95,533
institutional laundries (OECD. 201 lb).

E.12.18 Diameter of Container Opening

The ChemSTEER User Guide (U.S. EPA. 2015a) provides a typical diameter of container openings as
5.08 cm. Therefore, EPA could not develop a distribution of values for this parameter and used the
single value 5.08 cm from the ChemSTEER User Guide.

E.12.19	Diameter of Wash Opening	

The ESD on the Chemicals Used in Water Based Washing Operations at Industrial and Institutional
Laundries provided a single value for the diameter of washer openings to estimate air releases during
operation (	). The ESD states that the wash opening is 73 cm (OECD. 2017). Therefore,

EPA could not develop a distribution of values for this parameter and used the single value of 73 cm
from the ESD.

Page 345 of 484


-------
8599

8600

8601

8602

8603

8604

8605

8606

8607

8608

8609

8610

8611

8612

8613

8614

8615

8616

8617

8618

8619

8620

8621

8622

8623

8624

8625

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

E.12.20 Dilution Factor

The December 2020 Final Risk Evaluation for 1,4-Dioxane (U.S. EPA. 2020c) provided a single value
for the dilution factor of 1,4-dioxane in laundry detergents. The risk evaluation states that a dilution
factor of 0.016 was estimated assuming a high-end mass of product used (60g) in one gallon of water
(U.S. EPA. 2020c). Therefore, EPA could not develop a distribution of values for this parameter and
used the single value of 0.016 from the 2020 RE.

E.12.21 Container Fill Rate

The ChemSTEER User Guide (U.S. EPA. 2015a) provides a typical fill rate of 20 containers per hour
for containers with 20 to 100 gallons of liquid. Therefore, EPA could not develop a distribution of
values for this parameter and used the single value 20 containers/hour from the ChemSTEER User
Guide.

E.13 Hydraulic Fracturing Modeling Approach and Parameters for

Estimating Environmental Releases	

This appendix presents the modeling approach and equations used to estimate environmental releases of
1,4-dioxane during hydraulic fracturing. This approach utilizes the Draft ESD on Chemicals Used in
Hydraulic Fracturing (U.S. EPA. 2022d) combined with Monte Carlo simulation (a type of stochastic
simulation). This ESD includes a diagram of release and exposure points during textile dying, as shown
in FigureApx E-4.

FigureApx E-4. Environmental Release and Occupational Exposure Points during Hydraulic
Fracturing

Based on Figure Apx E-4, EPA identified the following release points:

•	Release point 1 (RP1): Fugitive air releases during container unloading;

•	Release point 2 (RP2): Container residue losses to surface water, incineration, or landfill;

•	Release point 3 (RP3): Fugitive air releases during container cleaning;

Page 346 of 484


-------
8626

8627

8628

8629

8630

8631

8632

8633

8634

8635

8636

8637

8638

8639

8640

8641

8642

8643

8644

8645

8646

8647

8648

8649

8650

8651

8652

8653

8654

8655

8656

8657

8658

8659

8660

8661

8662

8663

8664

8665

8666

8667

8668

8669

8670

PUBLIC RELEASE DRAFT - DO NOT CITE OR QUOTE

July 2023

•	Release point 4 (RP4): Equipment and storage tank cleaning losses to surface water, incineration,
or landfill;

•	Release point 5 (RP5): Fugitive air releases during equipment and storage tank cleaning;

•	Release point 6 (RP6): Release of hydraulic fracturing fluid that remains underground to deep
well injection; and

•	Release point 7 (RP7): Hydraulic fracturing fluid flowback and produced wastewater to
recycle/reuse, deep well injection, surface water, or land.

Environmental releases of hydraulic fracturing are a function of the chemical's physical properties,
container size, mass fractions, and other model parameters. Although physical properties are fixed for a
chemical, some model parameters are expected to vary from one facility to another. An individual model
input parameter could either have a discrete value or a distribution of values. EPA assigned statistical
distributions based on available literature data or engineering judgment to address the variability in
operating days (OD), mass fraction of 1,4-dioxane in fracturing fluid (Fdioxane fracturing fluid), mass fraction
of 1,4-dioxane in additive (Fdioxane additive), container container size (Vcont), annual use rate of fracturing
fluids (Qsite yr), saturation factor (Fsaturation), container cleaning losses (FCOnt cleaning), and fraction of
injected fracturing fluid that returns to the surface (Frecovered).

A Monte Carlo simulation (a type of stochastic simulation) was conducted to capture variability in the
model input parameters. The simulation was conducted using the Latin hypercube sampling method in
@Risk (Palisade, Ithaca, NY). The Latin hypercube sampling method is a statistical method for
generating a sample of possible values from a multi-dimensional distribution. Latin hypercube sampling
is a stratified method, meanin