vvEPA
United States
Environmental Protection
Agency
Effluent Guidelines Division
WH-552
Washington, DC 20460
EPA 440/1-79/007
Water and Waste Management
Development
Document for
Effluent Limitations
Guidelines and
Standards for the
Inorganic Chemicals
Manufacturing
Proposed
*
Point Source Category
-------
-------
TABLE OF CONTENTS
LIST
LIST
OF FIGURES
OF TABLES
ACKNOWLEDGEMENTS
1.0 CONCLUSIONS AND SUMMARY
l.L
1.2
1.3
1.4
1.5
TOXIC POLLUTANTS
CONTROL AND TREATMENT TECHNOLOGY
COSTS OF ADDITIONAL IN-PLANT TREATMENT
SUB CATEGORIZATION
RESTUDY OF REMANDED REGULATIONS
2 . 0 RECOMMENDATIONS
3.0 INTRODUCTION
3.1
3.2
3.3
AUTHORITY
3.1.1 The Federal Water Pollution
Control Act Amendments
3.1.2 Court Remand Regulations
3.1.3 The Settlement Agreement
GENERAL APPROACH AND METHODOLOGY
3.2.1 Industry Data Based Development
and Subcategorization Review
3.2.2 The Screening and Verification
Sampling Programs
3.2.3 Engineering Evaluations
3.2.4 Treatment System Cost Estimates
GENERAL CRITERIA FOR EFFLUENT LIMITATIONS
3.3.1 BPT Effluent Limitations
3.3.2 BAT Effluent Limitations
3.3.3 BCT Effluent Limitations
3.3.4 New Source Performance
Standards
3.3.5 Pretreatment Standards for
Existing Sources
3.3.6 Pretreatment Standards for
New Sources
Page
xxiii
xxxiii
li
1
1
2
2
2
3
5
23
23
23
26
28
36
37
37
37
38
38
38
39
39
42
42
42
111
-------
TABLE OF CONTENTS - Continued
4.0 SUBCATEGORIZATION REVIEW 45
4.1 BASIS FOR SUBCATEGORIZATION 45
4.1.1 Factors Considered 45
4.1.2 General Conclusions 49
4.2 SECONDARY SUBCATEGQRIZATION 49
4.2.1 Chlor-Alkali 49
4.2.2 Titanium Dioxide 49
4.2.3 Hydrogen Cyanide 51
4.3 REVIEW OF POSSIBLE INTEGRATION OF 51
SUBCATEGORIES
4.^.1 Hydrofluoric Acid and Aluminum 51
Fluoride
4.4 SUMMARY 52
5.0 SCREENING AND VERIFICATION SAMPLING PROGRAMS 53
5.1 SCOPE AND METHODOLOGY 53
5.1.1 Selecting Plants and Making 54
Preliminary Contacts
5.1.2 Screening and Verification 55
Sampling
5.1.3 Analytical Methodology for 57
Toxic Pollutants
5.1.4 Quality Assurance Provisions 64
5.2 SUMMARY OF ANALYTICAL RESULTS 65
6.0 PROCESS AND WASTE TREATMMENT INFORMATION 71
DEVELOPMENT AND EVALUATION
6.1 INDUSTRY DATA DESCRIPTION 71
6.1.1 Literature Review ' 71
6.1.2 Plant Visits 71
6.1.3 Telephone and Direct Contact 72
6.1.4 308 Questionnaire Responses 72
6.2 PROCESS WASTE SOURCES AND CURRENT 74
TREATMENT PRACTICES
6.2.1 Data Acquisition 74
6.2.2 Evaluation of Data 74
6.2.3 Model Plant and BPT Treatment 75
System Specification
6.2.4 Dissolved Solids in Waste Water 76
Effluents
-------
TABLE OF CONTENTS - Continued
Page
7.0 ASSESSMENT OF TECHNOLOGY FOR ADVANCED TREATMENT 79
AND CONTROL
7.1 INTRODUCTION 79
7.2 HYDROXIDE PRECIPITATION 80
7.3 FERRITE COPRECIPITATION 85
7.4 SULFIDE PRECIPITATION 85
7.5 THE XANTHATE PROCESS . 87
7.6 ION EXCHANGE 89
7.7 REDUCTION PROCESSES 91
7.8 OXIDATION PROCESSES 93
7.9 MEMBRANE PROCESSES 96
7.10 ADSORPTION 98
7.11 FLUORIDE REMOVAL 101
7.12 CHLORINE REMOVAL 102
8.0 TREATABILITY ESTIMATES AND LONG-T1RM DATA 103
ANALYSIS
8.1 THE DEVELOPMENT OF TREATABILITY ESTIMATES 103
8.2 THE USE OF HISTORICAL POLLUTANT DATA 117
8.2.1 Determination of Limitation 117
Guidelines Based Upon
Historical Performance
8.2.2 Assumptions Concerning Daily 118
Pollutant Level Measurements
8.2.3 Assumptions Concerning 30-Day 123
Average Pollutant Level
Observation
9.0 TREATMENT TECHNOLOGY APPLICATIONS FOR TOXIC 131
POLLUTANT REMOVAL
9.1 SELECTION OF POLLUTANTS TO BE CONTROLLED 131
9.2 APPLICATION OF ADVANCE LEVEL TREATMENT 131
AND CONTROL ALTERNATIVES
9.2.1 General Design Objectives 131
9.2.2 Pretreatment Technology 134
9.2.3 New Source Performance 134
Standards
9.3 ESTIMATED ACHIEVABLE PERFORMANCE ' 134
CHARACTERISTICS FOR ADVANCED LEVEL
APPLICATIONS
9.3.1 Advanced Level Removal of BPT 135
Pollutants
9.3.2 Advanced Level Removal of Toxic 135
Pollutants
v
-------
TABLE OP CONTENTS - Continued
9.4 POLLUTION CONTROL PARAMETERS TO BE
REGULATED
9.4.1 Conventional Pollutants
9.4.2 Nonconventional Pollutants
9.4.3 Toxic Pollutants
10.0 COST OP TREATMENT AND CONTROL SYSTEMS 139
10.1 INTRODUCTION 139
10.1.1 Purpose of Cost Data 139
10.1.2 General Approach 140
10.1.3 Cost References and Rationale 340
10.1.4 Definition of Levels of 141
Treatment and Control Cost
Development
10.1.5 Treatment and Disposal 141
Rationale Applied to Cost
Development
10.1.6 Expression of Costs 142
10.2 COST ESTIMATES FOR EACH SUBCATEGORY 149
11.0 CHLOR-ALKALI INDUSTRY 351
11.1 MERCURY CELL PROCESS INDUSTRY PROFILE 151
11.1.1 General Description 151
11.1.2 General Process Description and 151
Raw Materials
13.2 WATER USE AND WASTE WATER SOURCE 155
CHARACTERISTICS
11.2.1 Water Use 155
11.2.2 Waste Sources 155
11.3 DESCRIPTION OF SPECIFIC PLANTS 158
11.3.1 Screening Program 158
11.3.2 Verification 160
11.3.3 Descriptions of Plants Not 166
Sampled
11.3.4 Summary of the Toxic Pollutant 369
Data
11.4 POLLUTION ABATEMENT OPTIONS 174
11.4.1 Toxic Pollutants of Concern 174
11.4.2 Prevailing Control and 174
Treatment Practices
11.4.3 Process Modifications and 174
Technology Transfer Options
11.4.4 Best Management Practices 176
11.4.5 Advanced Treatment Technologies 176
-------
TABLE OF CONTENTS - Continued
11.5 SELECTION OF APPROPRIATE TECHNOLOGY AND
EQUIPMENT
11.5.1 Technologies for Different
Treatment Levels
11.5.2 Equipment for Different 177
Treatment Levels
11.6 TREATMENT COST ESTIMATES 180
11.6.1 General Discussion 180
11.6.2 Chlorine Bearing Wastes 183
11.6.3 Model Plant Treatment Costs 183
11.7 BASIS FOR REGULATIONS 182
11.7.1 Basis for BPT Limitations 382
11.7.2 Basis for Proposed BAT Effluent 3.94
Limitations
11.7.3 Basis for Proposed BCT Effluent 205
Limitations
11.7.4 Basis for New Source 205
Performance Standards
11.7.5 Basis for Proposed Pretreatment 205
Standards
11.8 DIAPHRAGM CELL PROCESS INDUSTRY PROFILE 206
11.8.1 General Description 206
11.8.2 General Process Description 206
11.9 WATER USE AND WASTE WATER SOURCES 209
11.9.1 Water Use " 209
11.9.2 Waste Sources 211
11.10 DESCRIPTIONS OF SPECIFIC PLANTS 214
11.10.1 Screening 214
11.10.2 Verification 214
11.10.3 Descriptions of Plants Not 221
Sampled
11.30.4 Toxic Pollutant Concentrations 224
11.11 POLLUTION ABATEMENT OPTIONS 231
11.11.1 Toxic Pollutants of Concern 231
11.11.2 Prevailing Control and 236
Treatment Practices
11.13.3 Process Modifications and 237
Technology Transfer Options
11.11.4 Best Management Practices 238
11.11.5 Advanced Treatment Technologies 239
11.12 SELECTION OF APPROPRIATE TECHNOLOGY AND 239
EQUIPMENT
11.12.1 Technologies for Different 239
Treatment Levels
11.12.2 Equipment for Different 244
Treatment Levels
Vll
-------
TABLE OF CONTENTS - Continued
11.13 TREATMENT COST ESTIMATES
11.13.1 General Discussion
11.13.2 Model Plant Treatment Costs
11.14 BASIS FOR REGULATIONS
11.14.1 Basis for BPT Limitations
11.14.2 Basis for BAT Effluent
Limitations
11.14.3 Basis for BCT Limitations 268
11.14.4 Basis for New Source 270
Performance Standards
11.14.5 Basis for Pretreatment 273
Standards
12.0 HYDROFLUORIC ACID INDUSTRY 275
12.1 INDUSTRY PROFILE 275
12.1.1 General Description 275
12.1.2 General Process Description and 275
Raw Materials
12.2 WATER USE AND WASTE SOURCE 279
CHARACTERISTICS
12.2.1 Water Use 279
12.2.2 Waste Sources 279
12.3 DESCRIPTION OF PLANTS VISITED AND SAMPLED 285
12.3.1 Screening 285
12.3.2 Verification 291
12.3.3 Summary of the Toxic Pollutant 291
Data
12.4 POLLUTION ABATEMENT OPTIONS 295
12.4.1 Toxic Pollutants of Concern 295
12.4.2 Process Modifications and 298
Technology Transfer Options
12.4.3 Best Management Practices 299
12.4.4 Prevailing Control and 299
Treatment Practices
12.4.5 Advanced Treatment Technologies 300
12.5 SELECTION OF APPROPRIATE TECHNOLOGY AND 30J
EQUIPMENT
12.5.1 Technologies for Different 301
Treatment Levels
12.5.2 Equipment for Different 306
Treatment Levels
12.6 TREATMENT COST ESTIMATES 308
12.6.1 General Discussion 308
12.6.2 Model Plant Control Costs for 313
Existing Sources
Vlll
-------
TABLE OF CONTENTS - Continued
12.6.3 Model Plant Control Costs for 317
New Sources
12.7 BASIS FOR REGULATIONS 324
12.7.1 Evaluation of BPT Treatment 324
Practices
12.7.2 Basis for Proposed BPT Effluent 329
Limitations
32.7.3 Basis for Proposed BCT Effluent 340
Limitations
12,7.4 Basis for Proposed BAT Effluent 341
Limitations
12.7.5 Basis for Proposed New Source 349
Performance Standards
12.7.6 Basis for Proposed Pretreatment 354
Standards
13.0 HYDROGEN PEROXIDE INDUSTRY 357
13.1 SUMMARY OF DETERMINATIONS 357
13.2 ASSESSMENT OF THE WATER POLLUTATION 357
POTENTIAL
13.2.1 Production Processes and 357
Effluents
13.2.2 Plants 358
13.2.3 Toxic Pollutants 358
13.3 STATUS OF REGULATIONS 358
J4.0 TITANIUM DIOXIDE INDUSTRY 361
(RUTILE/UPGRADED ILMENITE—CHLORIDE PROCESS)
14.1 INDUSTRY PROFILE 361
14.1.1 General Description 361
14.1.2 General Process Description and 361
Raw Materials
14.2 WATER USE AND WASTE SOURCE 365
CHARACTERISTICS
14.2.1 Water Use 365
14.2.2 Waste Sources 365
14.3 DESCRIPTION OF PLANTS VISITED AND SAMPLED 368
14.3.1 Screening 368
14.3.2 Verification 371
14.3.3 Toxic Pollutant Concentrations 371
14.4 POLLUTION ABATEMENT OPTIONS 375
14.4.1 Toxic Pollutants of Concern 375
14.4.2 Process Modification and 377
Technology Transfer Options
-------
TABLE OF CONTENTS - Continued
14.4.3 Best Management Practices 377
14.4.4 Prevailing Control and 377
Treatment Practices
14.4,5 Advanced Treatment Technologies 378
14.5 SELECTION OP APPROPRIATE TECHNOLOGY AND 379
EQUIPMENT
14.5.1 Technologies for Different 379
Treatment Levels
14.5.2 Equipment for Different 379
Treatment Levels
14.6 TREATMENT COST ESTIMATES 383
14.6.1 General Discussion 383
14.7 BASIS FOR REGULATIONS 385
14.7.1 Evaluation of BPT Treatment 385
Practices
14.7.2 Basis for Proposed BPT Effluent 392
Limitations
14.7.3 Basis for Proposed BCT Effluent 400
Limitations
14.7.4 Basis for Proposed BAT Effluent 400
Limitations
14.7.5 Basis for Proposed New Source 403
Performance Standards
14.7.6 Basis for Pretreatment 406
Standards
14.8 TITANIUM DIOXIDE - SULFATE PROCESS 408
INDUSTRY PROFILE
14.8.1 General Description 408
14.8.2 General Process Description and 408
Raw Materials
14.9 WATER USE AND WASTE SOURCE 411
CHARACTERISTICS
14.9.1 Water Use 411
14.9.2 Waste Sources 411
14.10 DESCRIPTION OF PLANTS 415
14.10.1 Screening 415
14.10.2 Verification 415
14.10.3 Other Plant Descriptions 417
14.10.4 Toxic Pollutant Concentrations 420
14.11 POLLUTION ABATEMENT OPTIONS 425
14.11.1 Toxic Pollutants of Concern 425
14.11.2 Process Modifications and 426
Technology Transfer Options
14.11.3 Best Management Practices 426
14.11.4 Prevailing Control and 426
Treatment Practices
-------
TABLE OF CONTENTS - Continued
14.11.5 Advanced Treatment Technologies 426
14.12 SELECTION OP APPROPRIATE TECHNOLOGY AND 427
EQUIPMENT
14.12.1 Technologies for Different 427
Treatment Levels
14.12.2 Equipment for Different 430
Treatment Levels
14.13 TREATMENT COST ESTIMATES 431
14.13.1 General Discussion 431
14.13.2 Model Plant Control and 432
Treatment Costs
14.14 BASIS FOR REGULATIONS 433
14.14.1 Evaluation of BPT Practices - 433
14.14.2 Basis for Proposed BPT Effluent 444
Limitations Technology Basis
14.14,3 Basis for Proposed BCT Effluent 452
Limitations
14.14.4 Basis for Proposed BAT Effluent 452
Limitations
14.14.5 Basis for Proposed N'ew Source 452
Performance Standards
14.14.6 Basis for Proposed Pretreatment 452
Standards
14.15 TITANIUM DIOXIDE - CHLORIDE ILMENITE 455
PROCESS INDUSTRY PROFILE
14.15.1 General Description 455
14.15.2 General Process Description and 455
Raw Materials
14.16 WATER USE AND WASTE SOURCE 458
CHARACTERISTICS
14.16.1 Water Use 458
14.16.2 Waste Sources 458
14.17 DESCRIPTION OF PLANTS VTSTED AND SAMPLED 461
14.17.1 Screening 461
14.17.2 Verification Program 462
14.17.3 Toxic Pollutant Concentration 462
14.18 POLLUTION ABATEMENT OPTIONS 466
14.18.1 Toxic Pollutants of Concern 466
14.18.2 Process Modifications and 466
Technology Transfer Options
14.18.3 Best Management Practices 466
14.18.4 Prevailing Control and 467
Treatment Practices
14.18.5 Advanced Treatment Technology 467
14.19 SELECTION OF APPROPRIATE TECHNOLOGY AND 468
EQUIPMENT
-------
TABLE OF CONTENTS - Continued
14.19.1 Technologies for Different
Treatment Levels
14.19.2 Equipment for Different
Treatment Levels
14.20 TREATMENT COST ESTIMATES
14.20.1 General Discussion
14.20.2 Model Plant Control and
Treatment Costs
14.21 BASIS FOR REGULATIONS 479
14.21.1 Evaluation of BPT Treatment 479
Practices
14.21.2 Basis for Proposed BPT Effluent 479
Limitation
14.21.3 Basis for Proposed BCT Effluent 485
Limitations
14.21.4 Basis for Proposed BAT Effluent 485
Limitations
14.21.5 Basis for the Proposed New 485
Source Performance Standards
14.21.6 Basis for Proposed Pretreatment 492
Standards
15.0 ALUMINUM FLUORIDE INDUSTRY . 493
15.1 INDUSTRY PROFILE 493
15.1.1 General Description 493
15.1.2 General Process Description and 493
Raw Materials
15.2 WATER USE AND WASTE SOURCE 493
CHARACTERISTICS
15.2.1 Water Use 493
15.2.2 Waste Sources 497
15.3 DESCRIPTION OF PLANTS VISITED AND SAMPLED 500
15.3.1 Screening 500
15.3.2 Verification 500
15.3.3 Summary of the Toxic Pollutant 500
Data
15.4 POLLUTION ABATEMENT OPTIONS 510
15.4.1 Toxic Pollutants of Concern 510
15.4.2 Process Modifications and 510
Technology Transfer Options
15.4.3 Best Management Practices 510
15.4.4 Prevailing Control and 511
Treatment Practices
15.4.5 Advanced Treatment Technologies 511
Xll
-------
TABUE OF CONTENTS - Continued
Page
15.5 SELECTION OF APPROPRIATE TECHNOLOGY AND 512
EQUIPMENT
15.5.1 Technologies for Different 512
Treatment Levels
15.5.2 Equipment for Different 512
Treatment Levels
15.6 TREATMENT COST ESTIMATES 518
15.6.1 General Discussion 518
15o7 BASIS FOR REGULATIONS 536
15.7.1 Evaluation of BPT Treatment 536
Practices
15.7.2 BPT Effluent Limitations 536
15.7.3 Basis for Proposed BCT Effluent 542
Limitations
15.7.4 Basis for Proposed BAT Effluent 542
Limitations
15.7.5 Basis for Proposed New Source 548
Performance Standards
15.7.6 Basis for Proposed Pretreatment 549
Standards
16.0 CHROME PIGMENTS INDUSTRY 55.1
16.1 INDUSTRY PROFILE 551
16.1.1 General Description 551
16.1.2 General Process Description and 551
Raw, Materials
16.2 WATER USE AND WASTE SOURCE 560
CHARACTERISTICS
16.2.1 Water Use 560
16.2.2 Waste Sources 560
16.3 DESCRIPTION OF PLANTS 564
16.3.1 Screening 564
16.3.2 Verification 567
16.3.3 Toxic Pollutant Concentrations 570
16.4 POLLUTION ABATEMENT OPTIONS 577
16.4.1 Toxic Pollutants of Concern 577
16.4.2 Process Modifications and 578
Technology Transfer Options
16.4.3 Best Management Practices 579
16.4.4 Prevailing Control and 579
Treatment Practices
16.4.5 Advanced Treatment Technologies 580
16.5 SELECTION OF APPROPRIATE TECHNOLOGY AND 581
EQUIPMENT
Xlll
-------
TABUS OF CONTENTS - Continued
16.5.1 Technologies for Different 581
Treatment Levels
16.5.2 Equipment for Different 581
Treatment Levels
16.6 TREATMENT COST ESTIMATES 584
16.6.1 General Discussion 584
16.6.2 Model Plant Costs 592
16.7 BASIS FOR REGULATIONS 592
16.7,1 Evaluation of BPT Treatment 592
Practices
16.7.2 Basis for Proposed BPT Effluent 595
Limitations
16.7.3 Basis for Proposed BCT 602
Limitations
16.7.4 Basis for Proposed BAT Effluent 602
Limitations
16.7.5 Basis for Proposed New Source 604
Performance Standards
16.7.6 Basis for Proposed Pretreatment 609
Standards
17.0 HYDROGEN CYANIDE INDUSTRY 611
17.1 INDUSTRY PROFILE ' 611
17.1.1 General Description 611
17.1.2 General Process Description and 611
Raw Materials
17.2 WATER USE AND WASTE SOURCE 614
CHARACTERISTICS
17.2.1 Water Use 614
17.2.2 Waste Source 614
17.3 DESCRIPTION OF PLANTS VISITED AND SAMPLED 637
17.3.1 Screening 617
17.3.2 Verification 619
17.3.3 Toxic Pollutant Concentrations 623
17.4 POLLUTION ABATEMENT OPTIONS 628
17.4.1 Toxic Pollutants of Concern 628
17.4.2 Process Modifications and 628
Technology Transfer Options
17.4.3 Best Management Practices 630
17.4.4 Prevailing Control and 630
Treatment Practices
17.4.5 Advanced Treatment Technologies 630
17.5 SELECTION OF APPROPRIATE TECHNOLOGY AND 631
" EQUIPMENT
xiv
-------
TABLE OF CONTENTS - Continued
17.5.1 Technologies for Different 631
Treatment Levels ''
17.5.2 Equipment for Different 633
Treatment Levels
17.6 TREATMENT COST ESTIMATES 635
17.6.1 General Discussion 635
17.7. BASIS FOR REGULATIONS 636
17.7.1 Evaluation of BPT Treatment 636
Practices
17.7.2 Basis for Proposed BPT 636
Limitations
17.7.3 Basis for Proposed BCT 649
Limitations
17.7.4 Basis for Proposed BAT 649
Limitations
17.7.5 Basis for Proposed New Source 652
Performance Standards
17.7.6 Basis for Proposed Pretreatment 652
Standards
18.0 SODIUM DICHROMATE INDUSTRY 655
18.1 INDUSTRY PROFILE 655
18.1.1 General Description 655
18.1.2 General Process Description and 655
Raw Materials
18.2 WATER USE AND WASTE SOURCE 658
CHARACTERISTICS
18.2.1 Water Use 658
18.2.2 Waste Sources 658
18.3 DESCRIPTION OF PLANTS VISITED AND SAMPLED 661
18.3.1 Screeninq 661
18.3.2 Verification 664
18.3.3 Toxic Pollutant Concentrations 664
and Loadings
18.4 POLLUTION ABATEMENT OPTIONS 668
18.4.1 Toxic Pollutants of Concern 668
18.4.2 Process Modifications and 671
Technology Transfer Options
18.4.3 Best Management Practices 671
18.4.4 Prevailing Control and 671
Treatment Practices
18.4.5 Advanced Treatment Technoloqies 672
18.5 SELECTION OF APPROPRIATE TECHNOLOGY AND 672
EQUIPMENT
18.5.1 Technology fo'r Different 672
Treatment Levels
xv
-------
TABLE OF CONTENTS - Continued
18.6 TREATMENT COST ESTIMATES 676
18.6.1 General Discussion 676
18.6.2 Model Plant Control Costs 677
18.7 BASIS FOR REGULATIONS 681
18.7.1 BPT Effluent Limitations 681
18.7.2 BCT Effluent Limitations 688
18.7.3 BAT Effluent Limitations 690
18.7.4 NSPS Effluent Limitations 693
18.7.5 Pretreatment Standards 694
19.0 CARBON DIOXIDE INDUSTRY 697
19.1 SUMMARY OF DETERMINATIONS 697
19.2 ASSESSMENT OF THE WATER POLLUTION 697
POTENTIAL
19.2.1 Production Processes and 697
Effluents
19.3 STATUS OF REGULATIONS 699
20.0 CARBON MONOXIDE AND BY-PRODUCT HYDROGEN INDUSTRY 701
20.1 SUMMARY OF DETERMINATIONS 701
20.2 ASSESSMENT OF THE WATER POLLUTION 701
POTENTIAL
20.2.1 Production Processes and 701
Effluents
20.3 STATUS OF REGULATIONS 703
21.0 COPPER SULFATE INDUSTRY 705
21.1 INDUSTRIAL PROFILE 705
21.1.1 General Description 705
21.1.2 General Process Description and 705
Raw Materials
21.2 WATER USE AND WASTE SOURCE 708
CHARACTERISTICS
21.2.1 Water Use 708
21.2.2 Waste Sources 708
21.3 DESCRIPTION OF PLANTS ¥ISITSD AND SAMPLED 711
21.3.1 Screening 711
21.3.2 Verification 713
21.3.3 Toxic-Pollutant Concentrations 716
21.4 POLLUTION ABATEMENT OPTIONS 719
21.4.1 Toxic Pollutants of Concern 719
21.4.2 Process Modifications and 719
Technology Transfer Options
xya.
-------
TABLE OF CONTENTS - Continued
21.4.3 Best Management Practices 719
21.4.4 Prevailing Control and 720
Treatment Practices
21.4.5 Advanced Treatment Technologies 720
21.5 SELECTION OP APPROPRIATE TECHNOLOGY AND 721
EQUIPMENT
21.5.1 Technologies for Different 721
Treatment Levels
21,5.2 Equipment for Different 721
Treatment Levels
21.6 TREATMENT COST ESTIMATES 724
21.6.1 General Discussion 724
21.6.2 Model Plant Cost Estimates 726
21.7 BASIS FOR REGULATIONS 726
21.7.1 Evaluation cf BPT Treatment 726
Practices
21.7.2 Basis for Proposed BPT Effluent 731
Limitations
21.7.3 Basis for Proposed BCT Effluent 733
Limitations
21.7.4 Basis for Proposed BAT Effluent 733
Limitations
21.7.5 Basis for Proposed New Source 738
Performance Standards
21.7.6 Basis for Proposed Pretreatment 740
Standards
22.0 NICKEL SDLFATE INDUSTRY 741
22.1 INDUSTRIAL PROFILE 741
22.1.1 General Description 741
22.1.2 General Process Description and 741
Raw Materials
22.2 WATER USE AND WASTE SOURCE 744
CHARACTERISTICS
22.2.1 Water Use 744
22.2.2 Waste Sources 744
22.3 DESCRIPTION OF PLANTS VISITED AND SAMPLED 747
22.3.1 Screening 747
22.3.2 Verification 747
22.3.3 Summary of Toxic Pollutant Data 752
22.4 POLLUTION ABATEMENT OPTIONS 755
22.4.1 Toxic Pollutants of Concern 755
22.4.2 Process Modifications and 755
Technology Transfer Options
22.4.3 Best Management Practices 756
xvn
-------
TABLE OF CONTENTS - Continued
Paqe
22.4.4 Prevailing Control and 756
Treatment Practices
22.4.5 Advanced Treatment Technologies 757
22.5 SELECTION OF APPROPRIATE TECHNOLOGY AND 757
EQUIPMENT
22.5.1 Technologies for Different 757
Treatment Levels
22.5.2 Equipment for Different 757
Treatment Levels
22.6 TREATMENT COST ESTIMATES 760
22.6.1 General Discussion 760
22.6.2 Model Plant Control Costs 761
22.7 BASIS FOR REGULATIONS 767
22.7.1 Evaluation of BPT Treatment 767
Practices
22.7.2 Basis for Proposed BPT Effluent 770
Limitations
22.7.3 Basis for Proposed BCT Effluent 770
Limitations
22.7.4 Basis for Proposed BAT Effluent 770
Limitations
22.7.5 New Source Performance 777
Standards
22.7.6 Basis for Proposed Pretreatment 777
Standards
23.0 SILVER NITRATE INDUSTRY 779
23.1 SUMMARY OF DETERMINATIONS 779
23.2 ASSESSMENT OF THE WATER POLLDTATION 779
POTENTIAL
23.2.1 Production Processes and 779
Effluents
23.3 STATUS OF REGULATIONS 781
24.0 SODIUM BISULFITE INDUSTRY 783
24.1 INDUSTRY PROFILE 783
24.1.1 General Description 783
24.1.2 General Process Description and 783
Raw Materials
24.2 WATER USE AND WASTE SOURCE 783
CHARACTERISTICS
24.2.1 Water Use 783
24.2.2 Water Sources 787
XVlll
-------
TABLE OF CONTENTS - Continued
24.3 DESCRIPTION OF PLANTS VISITED AND SAMPLED
24.3.1 Screening
24.3.2 Verification
24.3.3 Toxic Pollutant Analytical
Results
24.4 POLLUTION ABATEMENT OPTIONS 800
24.4.1 Toxic Pollutants of Concern 800
24.4.2 Prevailing Control and 800
Treatment Practices
24.4.3 Advanced Treatment Technologies 800
24.5 SELECTION OF APPROPRIATE TECHNOLOGY AND 801
EQUIPMENT
24.5.1 Technologies for Different 801
Treatment Levels
24.5.2 Equipment for Different 801
Treatment Levels
24.6 TREATMENT COST ESTIMATES 805
24.6.1 General Discussion 805
24.6.2 Cost Estimates 806
24.7 BASIS FOR REGULATIONS 813
24.7.1 Evaluation of BPT Treatment 833
Practices
24.7.2 Basis for Proposed BPT Effluent 813
Limitations
24.7.3 Basis for Proposed BCT Effluent 822
Limitations
24.7.4 Basis for Proposed BAT Effluent 822
Limitations
24.7.5 Basis for Proposed New Source 823
Performance Standards
24.7.6 Basis for Proposed Pretreatment 823
Standards
25.0 SODIUM HYDROSDLFITE INDUSTRY 827
25.1 INDUSTRY PROFILE 827
25.1.1 General Description 827
25.1.2 General Process Description and 827
Raw Materials
25.2 WATER USE AND WASTE SOURCE 830
CHARACTERISTICS
25.2.1 Water Use 830
25.2.2 Waste Sources 830
25.3 DESCRIPTION OF PLANTS VISITED AND SAMPLED 832
25.3.1 Screening and Verification 832
25.3.2 Toxic Pollutant Concentrations 834
xix
-------
TABLE OF CONTENTS - Continued
25.4 POLLUTION ABATEMENT OPTIONS 841
25.4.1 Toxic Pollutants of Concern 841
25.4.2 Prevailing Control and 841
Treatment Practices
25.4.3 Advanced Treatment Technologies 842
25.5 SELECTION OP APPROPRIATE TECHNOLOGGY AND 842
EQUIPMENT
25.5.1 Technologies for Different 842
Treatment Levels
25.5.2 Equipment for Different 842
Treatment Levels
25.6 TREATMENT COST ESTIMATES 846
25.6.1 General Discussion 846
25.6.2 Cost Estimates 846
25.7 BASIS FOR REGULATIONS 846
25.7.1 Evaluation of BPT Treatment 846
Practices
25.7.2 Basis for Proposed BPT Effluent 853
Limitations
25.7.3 Basis for Proposed BCT Effluent 855
Limitations
25.7.4 Basis for Proposed BAT Effluent 855
Limitations
25.7.,5 Basis for Proposed New Source 860
Performance Standards
25.7.6 Basis for Proposed Pretreatment 861
Standards
26.0 EXCLUDED SDBCATEGORIES 863
26.1 ALUMINUM SULFATE 863
26.2 AMMONIUM CHLORIDE 864
26.3 AMMONIUM HYDROXIDE " 866
26.4 BARIUM CARBONATE 868
26.5 BORAX 870
26.6 BORIC ACID 871
26.7 BROMINE 873
26.8 CALCIUM CARBIDE 874
26.9 CALCIUM CARBONATE 874
26.10 CALCIUM CHLORIDE 877
26.11 CALCIUM HYDROXIDE 878
26.12 CHROMIC ACID 879
26.13 CUPROUS OXIDE 879
26.14 FERRIC CHLORIDE 880
26.15 FERROUS SULFATE 881
26.16 FLUORINE 882
-------
TABLE OF CONTENTS - Continued
26.17
26.18
26.19
26.20
26.21
26.22
26.23
26.24
26.25
26.26
26.27
26.28
26.29
26.30
26.31
26.32
26.33
26,34
26.35
26.36
26.37
26.38
26.39
26.40
26.41
26.42
26.43
26.44
26.45
26.46
REFERENCES
BIBLIOGRAPHY
APPENDIX A
APPENDIX B
HYDROCHLORIC ACID
HYDROGEN
IODINE
LEAD MONOXIDE
LITHIUM CARBONATE
MANGANESE SULFATE
NITRIC ACID
OXYGEN AND NITROGEN
POTASSIUM CHLORIDE
POTASSIUM DICHROMATE
POTASSIUM IODIDE
POTASSIUM METAL
POTASSIUM PERMANGANATE
POTASSIUM SULFATE
SODIUM BICARBONATE
SODIUM CARBONATE
SODIUM CHLORIDE
SODIUM FLUORIDE
SODIUM HYDROSULFIDE
SODIUM METAL
SODIUM SILICATE
SODIUM SILICOFLUORIDE
SODIUM SULFITE
SODIUM THIOSULFATE
STANNIC OXIDE
STRONG NITRIC ACID
SULFUR DIOXIDE
SULFURIC ACID INDUSTRY
ZINC OXIDE
ZINC SULFATE
Analysis of Long-Term Effluent Monitoring
Data for the Inorganic Chemicals Industry
pH Control of Industrial Was;fce Waters in
the Inorganic Chemicals Industry
882
885
886
887
887
888
889
890
894
895
896
899
899
900
901
903
902
904
905
906
909
912
912
914
914
916
918
919
922
923
925
931
A-l
B-J
xxi
-------
-------
LIST OF FIGURES
Page
5-1 Sample flow sheet for metal analysis 59
7-1 Solubility of metal hydroxides and sulfides 82
as a function of pH.
7-2 Electrodialysis process 97
8-1 Cumulative distribution of daily concentrations 120
of mercury in treated effluent from plant
#251.
8-2 Cumulative distribution of daily concentrations 121
of cyanide in treated effluent from plant
1765.
8-3 Statistical distribution for daily pollution 124
measurements.
8-4 Cumulative distribution of 30-day averages 126
of total cyanide in treated effluent from
plant 1782.
8-5 Cumulative distribution of 30-day averages 127
of ammonia in treated effluent from plant
1782.
8-6 Statistical distributions for 30-day average 129
pollution measurements.
11-1 General process diagram for production of 156
chlorine/caustic by mercury cells
11-2 General process flow diagram at plant f299 161
showing the sampling points. Chlorine/caustic
(mercury cell) manufacture
11-3 General process flow diagram at plant f747 165
showing the sampling points. Chlorine/caustic
(mercury cell) manufacture
XXlll
-------
I.IST OF FIGURES - Continued
Page
11-4 General process flow diagram at plant fl67 165
showing the sampling points. Chlorine/caustic
(mercury cell) manufacture
11-5 General process flow diagram at Plant f317 167
showing the sampling points. Chlorine/caustic
(mercury cell) manufacture
11-6 Level 1 waste water treatment for 178
chlorine-mercury cell subcategory
11-7 Level 2 waste water treatment for 179
chlorine-mercury cell subcategory
11-8 Annual treatment cost versus production for 186
the chlorine subcategory (mercury cell process)
11-9 Annual unit treatment cost versus production 187
for the chlorine subcategory (mercury cell
process)
11-10 Annual treatment cost versus production for 189
the chlorine subcategory (mercury cell process)
11-11 Annual unit treatment cost versus production 190
for the chlorine subcategory (mercury cell
process)
11-12 Annual treatment cost versus production for 191
the chlorine subcategory (mercury cell process)
11-13 Annual unit treatment cost versus production 192
for the chlorine subcategory (mercury cell
process)
11-14 General process flow diagram for production 210
of chlorine/caustic by diaphragm cells
11-15 General process flow diagram at plant f034 215
showing the sampling points. Chlorine/caustic
(diaphragm cell) manufacture
11-16 General process flow diagram at plant £261 218
showing the sampling points. Chlorine/caustic
(diaphragm cell) manufacture
xxxv
-------
LIST OF FIGURES - Continued
Page
11-17 General process flowsheet at Plant 1738-A 219
showing the sampling points. Chlorine/caustic
(diaphragm cell) manufacture
11-28 General process flow diagram at Plant 220
f738-B showing the sampling points.
Chlorine/caustic (diaphragm cell) manufacture
11-19 General process flow diagram at Plant f736 222
showing the sampling points. Chlorine/caustic
(diaphragm cell) manufacture
11-20 General process flow diagram at" Plant $967 223
showing the sampling points. Chlorine/caustic
(diaphragm cell) manufacture
11-21 Level 1 waste water treatment for 241
chlorine-diaphragm cell subcategory
11-22 Level 2 waste water treatment for 242
chlorine-diaphragm cell subcategory
11-23 Level 3 waste water treatment for 243
chlorine-diaphragm cell subcategory
11-24 Annual treatment cost versus production for 253
the chlorine subcategory (diaphragm cell process)
11-25 Annual unit treatment cost versus production 254
for the chlorine subcategory (diaphragm cell
process)
12-1 General process flow diagram for production 280
of hydrofluoric acid
12-2 Production vs. waste flow data for HF plants 284
12-3 General process flow diagram at plant f705 289
showing the sampling points. Hydrofluoric
acid manufacture
12-4 General process flow diagram at Plant i'251 293
showing the sampling points. Hydrofluoric
acid manufacture
12-5 Level 1 waste water treatment for hydrofluoric 302
acid subcategory
xxv
-------
LIST OF FIGURES - Continued
Page
12-6 Level 2 waste water treatment for hydrofluoric 303
acid subcategory
12-7 Level 3 waste water treatment for hydrofluoric 304
acid subcategory
12-8 Level 4 waste water treatment for hydrofluoric 305
acid subcategory
12-9 Waste water treatment new source performance 307
standard for hydrofluoric acid subcategory
12-10 Annual treatment cost versus production for 314
the hydrofluoric acid subcategory
12-11 Annual unit treatment cost versus production 315
for the hydrofluoric acid subcategory
12—12 Annual treatment cost versus production for 321
the hydrofluoric acid subcategory (NSPS)
12-13 Annual unit treatment cost versus production 322
for the hydrofluoric acid subcategory (NSPS)
12-14 Fluoride loads and concentrations discharged 330
at selected hydrofluoric acid plants
14-1 General process diagram for production of 366
titanium dioxide (chloride process) from high
grade ores
14-2 General flow diagram at Plant f559 showing 370
the sampling points. (Titanium dioxide -chloride
process manufacture)
14-3 General flow diagram at Plant ^172 showing 373
the sampling points. Titanium dioxide (chloride
process) manufacture
14-4 Level 1 waste water treatment for titanium 380
dixoide - chloride process
14-5 Level 2 waste water treatment for titanium 381
dioxide - chloride process
14-6 Level 3 waste water treatment for titanium 382
dioxide - chloride process
xxvx
-------
LIST OF FIGURES - Continued
gage
14-7 Annual treatment cost versus production for 389
the titanium dioxide subcategory, chloride
process
14-8 Annual unit treatment cost versus production 390
for the titanium .dioxide subcategory, chloride
process
14-9 General process flow diagram for production 412
of titanium dioxide by sulfate process
14-10 General flow diagram at Plant |559 showing 418
the sampling points. (Titanium dioxide -sulfate
process)
14-11 Level 1 waste water treatment for titanium 428
dixoide - sulfate process
14-12 Level 2 waste water treatment for titanium 429
dioxide - sulfate process
14-13 Annual treatment cost versus production for 437
the titanium dioxide subcategory, sulfate
process
14-14 Annual unit treatment cost versus production 438
for the titanium dioxide subcategory, sulfate
process
14-15 General process flow diagram of the titanium 457
tetrachloride portion of a titanium dioxide
plant using the chloride-ilmenite process.
14-3.6 Level 1 waste water treatment for titanium 469
dioxide - chloride (ilmenite ore) process
14-17 Level 2 waste water treatment for titanium 470
dioxide - chloride {ilmenite ore) process
15-1 General process flow diagram for production 496
of aluminum fluoride
15-2 General process flow diagram at Plant £705 501
showing the sampling points. (Aluminum fluoride
manufacture)
xxvii
-------
LIST OF FIGURES - Continued
Page
15-3 General process flow diagram at Plant f251 503
showing the sampling points. Aluminum fluoride
manufacture
15-4 Level 1 waste water treatment for aluminum 513
fluoride subcategory
15-5 Level 2 waste water treatment for aluminum 514
fluoride subcategory
15-6 Level 3 waste water treatment for aluminum 515
fluoride subcategory
15-7 Level 4 waste water treatment for aluminum 516
fluoride subcategory
15-8 Annual treatment cost versus production for 523
the aluminum fluoride subcategory
15-9 Annual unit treatment cost versus production 524
for the aluminum fluoride subcategory
15-10 Effect of variation of pollutant load on 527
treatment cost at level 1 technology
15-11 Effect of variation of pollutant load on 528
treatment cost at level 4 technology
15-12 Effect of variation of hydraulic load on 532
treatment cost at level 2 technology
15-13 Effect of variation of hydraulic load on 533
treatment cost at level 3 technology
15-14 Effect of variation of hydraulic load on 534
treatment cost at level 4 technology
16—1 General process diagram for production of 555
anhydrous chrome oxide
16-2 General process diagram for production of 556
hydrated chromic oxide
16-3 General process diagram for production of 558
chrome yellow
16-4 General process diagram for production of 559
molybdenum orange
xxviii
-------
LIST OF FIGUBES - Continued
16-5 General process diagram for production of 561
chrome green
16-6 General process diagram for production of 562
zinc yellow
16-7 General process diagram for production of 565
chrome pigment complexes
16-8 General waste water treatment process flow 568
diagram at Plant §002 showing the sampling
points. (Chrome pigment manufacture)
16-9 General waste water treatment process flow 571
diagram at Plant |894 showing the sampling
points. (Chrome pigment manufacture)
16-10 Level 1 waste water treatment for chrome pigments , 582
16-11 Level 2 waste water treatment for chrome pig'ments 583
16-12 Annual -treatment cost versus production for 590
the chrome pigments subcategory
16-13 Annual unit treatment cost versus production 591
for the chrome pigments subcategory
17-1 General process flow diagram for production 615
of hydrogen cyanide by the Andrussow Process
17-2 General waste water treatment process flow 620
diagram at Plant t765 showing the sampling
points. (Hydrogen cyanide manufacture)
17-3 General waste water treatment process flow 624
diagram at Plant 1782 showing sampling points.
(Hydrogen cyanide manufacture)
17-4 Level 1 waste water treatment for hydrogen 632
cyanide subcategory
17-5 Level 2 waste water treatment for hydrogen 634
cyanide subcategory
17-6 Annual treatment cost as a function of production 640
for the hydrogen cyanide subcategory
xxix
-------
LIST OP FIGURES - Continued
17-7 Annual unit treatment cost as a function of 641
production for the hydrogen cyanide subcategory
18-1 General process diagram for production of 659
sodium dichromate
18-2 General waste water treatment process flow . 662
diagram at Plant f493 showing the sampling
points. (Sodium dichromate manufacture)
18-3 General waste water treatment process flow 665
diagram at Plant f376 showing the sampling
points. (Sodium dichromate manufacture)
18-4 Level 1 waste water treatment for sodium dichromate 674
subcategory
18-5 Level 2 waste water treatment for sodium dichromate 675
subcategory
»
18-6 Relationship of annual treatment cost to production 682
for the sodium dichromate subcategory
18-7 Relationship of annual unit treatment cost 683
to production for the sodium dichromate subcategory
21-1 General block diagram of the manufacture of 709
copper sulfate
21-2 General process flow diagram at plant |034 714
showing the sampling points. (Copper Sulfate
manufacture)
21-3 Level 1 waste water treatment for copper sulfate 722
subcategory - batch process
21-4 Level 2 waste water treatment for copper sulfate 723
subcategory - batch process
22-1 General process flow diagram for nickel sulfate 745
manufacture
22-2 General waste water treatment process flow 748
diagram showing sampling points at Plant f369.
(Nickel sulfate subcategory,)
xxx
-------
LIST OF FIGURES - Continued
Page
22-3 General process flow diagram at Plant |572 750
showing the sampling points. {Nickel sulfate
manufacture,)
22-4 General waste water treatment process flow 751
diagram at Plant £120 showing the sampling
points. (Nickel sulfate manufacture)
22-5 Level 1 waste water treatment for nickel sulfate 758
subcategory - batch process
22-6 Level 2 waste water treatment for nickel sulfate 759
subcategory - batch process
22-7 Relationship of annual treatment cost to production 765
for the nickel sulfate subcategory
22-8 Relationship of annual unit treatment cost 766
to production for the nickel sulfate subcategory
24-1 General process flow diagram at Plant f282 790
showing the sampling points. Sodium bisulfite
manufacture
24-2 General flow diagram at Plant f586 showing 791
the sampling points. Sodium bisulfite manufacture
24-3 General process flow diagram at Plant 1987 794
showing the sampling points. Sodium bisulfite
manufacture
24-4 Level 1 waste water treatment for sodium bisulfite 802
subcategory - batch process
24-5 Level 2 waste water treatment for sodium bisulfite 803
subcategory - batch process
24-6 Level 3 waste water treatment for sodium bisulfite 804
subcategory
24-7 Variation of annual treatment cost with production 810
for the sodium bisulfite subcategory
24-8 Variation of annual unit treatment cost with 811
production (sodium bisulfite subcategory)
xxxi
-------
LIST OF FIGUBES - Continued
Page
25-1 General process flow diagram at Plant f672. 831
(Sodium hydrosulfite manufacture)
25-2 General process flow diagram at Plant $672 835
showing the sampling points. (Sodium hydrosulfite
manufacture)
25-3 Level 1 waste water treatment for sodium 843
hydrosulfite subcategory
25-4 Level 2 waste water treatment for sodium 844
hydrosulfite subcategory
XX2O.X
-------
LIST OF TABLES
Page
2-1 Summary of Proposed Regulations - Best 6
Practicable Control Technology Currently
Available (BPT)
2-2 Summary of Proposed Regulations - Best 9
Available Technology (BAT)
2-3 Summary of Proposed Regulations - Pretreatment 12
Standards for Existing Sources (PSES)
2-4 Summary of Proposed Regulations - New 15
Source Performance Standards (NSPS)
2-5 Summary of Proposed Regulations - Pretreatment 19
Standards for New Sources (PSNS)
2-6 Summary of Proposed Regulations - Best 22
Conventional Pollutant Control Technology
(BCT)
3-1 Recommended List of Toxic Pollutants 28
3-2 Scope of Industry Coverage within the 34
Inorganic Chemicals Manufacturing Point
Source Category
5-1 Analytical Detection Limits for Metals 62
5-2 Pollutant Frequency Based on Sampling 66
Program Results Including Raw Waste
5-3 Distribution of Pollutants According 67
to Subcategory
6-1 308 Questionnaire Response Data 73
7-1 Solubility Products of Trace Metals 83
7-2 Comparison of Reverse Osmosis Concepts 99
XXXill
-------
LIST OF TABI^ES - Continued
Page
8-1 Waste Water Treatment Options and Performance 104
Data Summary - Antimony and Arsenic Removal
8-2 Waste Water Treatment Options and Performance 105
Data Summary - Beryllium and Cadmium Removal
8-3 Waste Water Treatment Options and Performance 106
Data Summary - Copper Removal
8-4 Waste Water Treatment Options and Performance 107
Data Summary - Chromium III and Chromium
VI Removal
8-5 Waste Water Treatment Options and Performance 108
Data Summary - Lead Removal
8-6 Waste Water Treatment Options and Performance 109
Data Summary - Mercury II Removal
8-7 Waste Water Treatment Options and Performance 110
Data Summary - Nickel Removal
8-8 Waste Water Treatment Options and Performance 111
Data Summary - Silver Removal
8-9 Waste Water Treatment Options and Performance 112
Data Summary - Selenium and Thallium Removal
8-10 Waste Water Treatment Options and Performance 113
Data Summary - Zinc Removal
8-11 Estimated Achievable Maximum 30-Day Averages 115
for the Applied Technologies
9-1 Prioritization of Toxic Metals Found in 132
Each Subcategory
11-1 Subcategory Profile Data Summary 152
11-2 Status of Regulations - Effluent Limitation 153
Guidelines
11-3 Summary of Waste Water Plow Data for Chlorine 159
Mercury Cell Plants
11-4 Pollutant Concentration and Loads at Plant 162
£299
XXXIV
-------
LIST OF TABLES - Continued
Paqe
11-5 Pollutant Concentrations and Loads at 163
Verification Plants (1)
11-6 Toxic Pollutant Raw Waste Concentrations 172
and Loads at Verification Plants mg/1
kg/kkg
11-7 Summary of Raw Waste Loadings at Verification 173
Plants
11-8 Model Plant Treatment Costs 183
11-9 Model Plant Treatment Costs 184
11-10 Model Plant Treatment Costs 185
11-11 Model Plant Dnit Treatment Costs 188
11-12 Estimated Chemical Dechlorination Costs 193
for the Chlor-Alkali Industry
11-13 Mercury Discharges from Selected 195
Chlor-Alkali Mercury Cell Plants*
11-14 Residual Chlorine Discharges at Selected 196
Chlor-Alkali Plants*
11-15 Comparison of Raw Waste Concentrations 198
of Toxic Pollutants with Treatability
11-16 Proposed Limitations, BAT 199
11-17 Effluent Concentrations of Toxic Pollutants 202
from Verification Sampling
11-18 Subcategory Profile Data Summary 207
11-19 Status of Regulations - Effluent Limitation 208
Guidelines
11-20 Waste Water Plows at Diaphragm Cell Chlorine 213
Plants
11-21 Pollutant Concentrations and Loads at 216
Screening and Verification Plants
xxxv
-------
LIST OF TABLES - Continued
Page
11-22 Results of Asbestos Sampling at Diaphragm 225
Cell Plants
11-23 Maximum Raw Waste Concentrations of Toxic 226
Metals Observed at Diaphragm Cell Chlorine
Plants (mg/1)
11-24 Toxic Metal Concentrations and Loads at 229
Screening and Verification Plants/' mg/1
\( mg/1 \
V kg/kkg /
11-25 Summary of Raw Waste Loadings at Screening 230
and Verification Metal Anode Plants
11-26 Toxic Metal Concentrations and Loads in 232
Cell Room Waste Waters at Screening and
Verification Plants/ mg/1\
\kg/kkg/
13-27 Raw Waste Toxic Metals Concentration and 233
Loads in Process Streams Other Than Cell
Room Wastes from Screening and Verification
Plants
11-28 Raw Waste Toxic Organics at a Graphite 234
Anode Plant
11-29 Raw Waste Toxic Organics by Waste Water 235
Source at a Graphite Anode Plant
11-30 Model Plant Treatment Costs 249
11-31 Model Plant Treatment Costs 250
11-32 Model Plant Treatment Costs 251
11-33 Model Plant Treatment Costs 252
11-34 Summary of Unit Flows at Diaphragm Cell 255
Plants
11-35 Comparison of Toxic Metals Treatability 257
with Screening and Verification Sampling
Data
11-36 Proposed Limitations, BPCTCA 258
XXXVI
-------
LIST OP TABLES - Continued
11-37
11-38
11-39
11-40
11-41
11-42
12-1
12-2
12-3
12-4
12-5
12-6
12-7
12-8
!
Lead and TSS Discharges from Selected Diaphragm
Cell Chlorine Plants (1)
Toxic Pollutants in Diaphragm Cell Plant
Effluents
Proposed Limitations, BAT
Proposed Limitations, BCT
i
Proposed Limitations, NSPS
Comparison of Raw Waste Ch-a-racteristics
at a New Metal Anode Plant with Treatability
of Toxic Metals
Subcategory Profile Data Summary-
Status of Regulations - Effluent Limitation
Guidelines
Water Usage in the Hydrofluoric Acid
Subcategory
Waste Water Flow and Reuse Data for the
Hydrofluoric Acid Subcategory
Waste Flow from Hydrofluoric Acid
Manufacturing Plants
Solid Waste Generated at the Hydrofluoric
Acid Plants Sampled
Gypsum Solids Production in the Hydrofluoric
Acid Subcategory
Plow and Pollutant Concentration Data of
Page
261
262
265
269
271
272
276
277
281
282
286
287
288
290
the Sampled Waste Streams for Plant f705
Producing Hydrofluoric Acid
12-9 Plow and Pollutant Concentration Data of 292
the Sampled Waste Streams for Plants f705,
$251, and f!67 Producing Hydrofluoric Acid
12-10 Toxic Pollutant Raw Waste Data 296
12-11 Summary of Raw Waste Loadings Found in 297
Screening and Verification Sampling
xxxvii
-------
LIST OF TABLES - Continued
12-12
12-13
12-14
12-15
12-16
12-17
12-18
12-19
12-20
12^21
12-22
12-23
12-24
12-25
12-26
12-27
12-28
12-29
13-1
14-1
14-2
Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Treatment Costs
Summary of Waste Water Control and Treatment
Technology Employed at Hydrofluoric Acid
Plants
Summary of Long-Term Monitoring Data from
Pour Hydrofluoric Acid Plants
Toxic Pollutant Treated Effluent Data
Development of TSS and Fluoride Limitations
Proposed Limitations, BPCTCA
Proposed Limitations, BAT
Performance of Alternative Technology,
Level 3 Treatment
Performance of Alternative Technology,
Level 4 Treatment
Toxic Pollutant Raw Waste Data Used to
Represent New Sources*
Proposed Limitations, NSPS
Subcategory Profile Data Summary
Subcategory Profile Data Summary
Status of Regulations - Effluent Limitation
Page
310
311
312
316
318
319
320
323
325
327
328
334
335
343
345
346
351
353
359
362
363
Guidelines
xxxvxxi
-------
LIST OF TABLES - Continued
Page
14-3 Water Osage in Titanium Dioxide-Chloride 367
Process/High Grade Ores Subcategory
14-4 Waste Water Flow for Titanium Dioxide-Chloride 369
Process Subcategory
1*4-5 Flow and Pollutant Concentration Data of 372
the Sampled Waste Streams of Plant f!72
Producing Titanium Dioxide by Chloride-
Rutile Process
14-6 Flow and Pollutant Concentration Data of 374
the Sampled Waste Streams for Plant f!72
Producing Titanium Dioxide (Chloride Process)
14-7 Raw Waste Pollutant Data Summary of the 376
Sampled Streams
14-8 Model Plant Treatment Costs 386
14-9 Model Plant Treatment Costs 387
14-10 Model Plant Treatment Costs " 388
14-11 Model Plant Treatment Costs 391
14-12 Historical Effluent Monitoring Data Summary 393
with Variability Factor
14-13 Historical Effluent Monitoring Data Summary 394
with Variability Factors Daily Measurements
14-14 Treatment Performance Data of Sampled Plants 395
1599 and f!72
14-15 Proposed Limitations, BPCTCA 401
14-16 Proposed Limitations, BAT . 402
14-17 Proposed Limitations, NSPS 407
14-18 Subcategory Profile Data Summary 409
14-19 Analysis of Ilmenite Ores 410
14-20 Water Usage in Titanium Dioxide - Sulfate 413
Process Subcategory
XXXLX
-------
LIST OF TABLES - Continued
Page
14-21 Raw Waste Characteristics (Industry Data) 416
for Plant 1555 (Production of Ti02 by Sulfate
Process)
14-22 Flows and Pollutant Concentrations for 419
the Waste Streams Sampled for Plant $559
Producing Titanium Dioxide
14-23 Process Waste Water Flow at Plants f555r 421
1694 and £559 Titanium Dioxide (Sulfate
Process)
14-24 Summary of Raw Waste Loadings Found in 423
Screening and Verification Sampling
14-25 Toxic Pollutants: Average Raw Waste Loads 424
and Concentrations
14-26 Model Plant Treatment Costs 434
14-27 Model Plant Treatment Costs ' 435
14-28 Model Plant Treatment Costs 436
14-29 Model Plant Treatment Costs 439
14-30 Historical Effluent Monitoring Data Summary 440
14-31 Verification Results from - Sulfate Process 443
Titanium Dioxide Plant f559
14-32 Proposed Limitations, BPTCTA 451
14-33 Proposed Limitations, BAT 453
14-34 Proposed Limitations, NSPS 454
14-35 Subategory Profile Data Summary 456
14-36 Average Water Usage for TiO~ Production 459
by the Chloride - Ilmenite process
14-37 Average Raw Waste Loads for Ti02 Production 460
by the Chloride - Ilmenite Process
14-38 Summary of Raw Waste Loadings Found in 464
Screening and Verification Sampling
-------
LIST OF TABLES - Continued
Page
14-39 Toxic Pollutant Average Raw Waste Loads 465
and Concentrations
14-40 Model Plant Treatment Costs 474
14-41 Model Plant Treatment Costs 475
14-42 Model Plant Treatment Costs 476
14-43 Model Plant Treatment Costs 477
14-44 Model Plant Treatment Costs 478
14-45 Proposed Limitations, BPCTCA 482
14-46 Proposed Limitations, NSPS 487
15-1 Subcategory Profile Data Summary 494
15-2 Status of Regulations - Effluent Limitation 495
Guidelines
15-3 Water Usage in the Aluminum Fluoride 498
Subcategory
15-4 Waste Water Plow at Plants |837, f705 and 499
1251 for Aluminum Fluoride Subcategory
15-5 Solids Generated at Plant f705 and f251 499
Producing Aluminum Fluoride
15-6 Flow and Pollutant Concentration Data of 502
the Sampled Waste Streams for Plant f705
Producing Aluminum Fluoride
15-,7 Flow and Pollutant Concentration Data of 504
the Sampled Streams for Plant f251 Producing
Aluminum Fluoride
15-8 Toxic Pollutant Average Raw Waste Loads 507
and Concentrations
15-9 Toxic Pollutant Effluent Concentrations 508
During Sampling
15-10 Summary of Raw Waste Loadings Found in 509
Screening and Verification Sampling
xli
-------
LIST OF TABLES - Continued
Page
15-11 Model Plant Treatment Costs 520
15-12 Model Plant Treatment Costs 521
15-13 Model Plant Treatment Costs 522
15-14 Model Plant Treatment Costs 525
15-15 Model Plant Treatment Costs 526
15-36 Model Plant Treatment Costs 530
15-17 Model Plant Treatment Costs 531
15-18 Model Plant Treatment Costs 535
15-19 Proposed Limitations, BPCTCA 543
15-20 Proposed Limitations, BAT 544
15-21 Performance of Alternative Technology Level 545
3 Treatment
15-22 Performance of Alternative Technology Level 546
4 Treatment
15-23 Proposed Limitations, NSPS 550
16-1 Subcategory Profile Data Summary 552
16-2 Status of Regulations - Effluent Limitation 553
Guidelines
*>
16-3 Water Usage in the Chrome Pigments Subcategory 563
16-4 Summary of Waste Water Plow 566
16-5 Plow, Pollutant, Concentration and Load 569
Data of the Sampled Waste Streams for Plant
1002
16-6 Flow, Pollutant, Concentration and Load 572
Data for the Sampled Streams at Plant £894
16-7 Toxic Pollutant Raw Waste Data 574
xlii
-------
LIST OF TABLES - Continued
Page
16-8 Summary of Raw Waste Loadings Found in 575
Screening and Verification Sampling
16-9 Toxic Pollutant Treated Waste Data 576
16-10 Model Plant Treatment Costs 586
16-11 Model Plant Treatment Costs 587
16-12 Model Plant Treatment Costs 588
16-13 Model Plant Treatment Costs 589
16-14 Model Plant Treatment Costs 593
16-15 Summary of Long Term and Verification Effluent 596
Sampling Results at Plant f894
16-16 Proposed Limitations, BPCTCA 603
16-17 Proposed Limitations, NSPS 606
17-1 Subcategory Profile Data Summary 612
17-2 Status of Regulations - Effluent Limitation 613
Guidelines
17-3 Water Usage in Hydrogen Cyanide - Andrussow 636
Process Subcategory
17-4 Waste Flow Data for HCN Production by the 618
Andrussow Process
17-5 Flow and Pollutant Data for the Raw and 621
Treated Waste Streams of Plant f765 Producing
Hydrogen Cyanide by Andrussow Process
17-6 Flow and Pollutant Concentration Data of 622
the Sampled Waste Streams for Plant |765
Producing Hydrogen Cyanide
17-7 Flow and Pollutant Concentration Data of 625
the Sampled Waste Streams for Plant #782
Producing Hydrogen Cyanide
17-8 Unit Flow and Unit Pollutant Loading for 626
Raw and Treated Waste Effluents at Plant
£782
xliii
-------
LIST OF TABLES - Continued
Page
17-9 Summary of Pollutant Raw Waste Loading 629
Pound in Screening and Verification Sampling
17-10 Model Plant Treatment Costs 637
17-11 Model Plant Treatment Costs 638
17-12 Model Plant Treatment Costs 639
17-13 Model Plant Treatment Costs 642
17-14 Statistical Analysis of the 28-Day Effluent 645
Sampling Results on Total Cyanide and Ammonia
from Plant f765
17-15 Statistical Analysis of Historical Effluent 648
Monitoring Data on Free Cyanide from Plant
f765
17-16 Proposed Limitations, BPCTCA 650
17-17 Proposed Limitations, BAT 653
17-18 Control Parameter Limitations, NSPS '654
18-1 Subcategory Profile Data Summary 656
18-2 Status of Regulations - Effluent Limitation 657
Guidelines
18-3 Water Usage in Sodium Dichromate Subcategory 660
18-4 Flow and Pollutant Concentration Data of 663
the Sampled Waste Streams for Plant f493
Producing Sodium Dichromate
18-5 Flow and Pollutant Loading Data of the 666
Sampled Waste Streams for Plant |376 Producing
Sodium Dichromate
18-6 Flow and Pollutant Loading Data of the 667
Sampled Waste Streams for Plant f398 Producing
Sodium Dichromate
18-7 Toxic Pollutant Raw Waste Data 669
xLiv
-------
LIST OF TABLES - Continued
Page
18-8 Summary of Raw Waste Loadings Pound in 670
Screening and Verification Sampling
18-9 Model Plant Treatment Costs 678
18-10 Model Plant Treatment Costs 679
18-11 Model Plant Treatment Costs 680
18-12 Model Plant Treatment Costs 684
18-13 Effluent Sampling Data from Sodium Bichromate 687
Plants
18-14 Proposed Limitations, BPCTCA 689
18-15 Proposed Limitations, BAT 695
18-16 Control Parameter Limitations, NSPS 696
19-1 Subcategory Profile Data Summary 698
20-1 Subcategory Profile Data Summary 702
21-1 Subcategory Profile Data Summary 706
21-2 Status of Regulations - Effluent Limitation 707
Guidelines
21-3 Water Usage in Copper Sulfate Subcategory 710
21-4 Waste Water Flow for the Copper Sulfate 712
Subcategory
21-5 Flow and Pollutant Concentration Data of 715
the Sampled Waste Streams for Plant f034
Producing Copper Sulfate
21-6 Raw Waste Data 718
21-7 Model Plant Treatment Costs 727
21-8 Model Plant Treatment Costs 728
21-9 Summary of Long-Term Monitoring Data from 729
Plant 1034
xlv
-------
LIST OF TABLES - Continued
Paqe
21-10 Treated Effluent Data 730
21-11 Average Pollutant Levels and Removal Efficiency 732
for Plant 1034
21-12 Proposed Limitations, BAT 736
21-13 Performance of Alternative Technology, 739
Treatment Level 2
22-1 Subcategory Profile Data Summary 742
22-2 Status of Regulations - Effluent Limitation 743
Guidelines
22-3 Water Use in the Nickel Sulfate Subcategory 746
22-4 Flow and Pollutant Concentration Data of 749
the Sampled Waste Streams for Plants Producing
Nickel Sulfate
22-5 Toxic Pollutant Raw Waste Data 754
22-6 Model Plant Treatment Costs 762
22-7 Model Plant Treatment Costs 763
22-8 Model Plant Treatment Costs 764
22-9 Model Plant Treatment Costs 768
22-10 Toxic Pollutant Treated Effluent Data 769
22-11 Proposed Effluent Limitations, BAT 773
22-12 Proposed Effluent Limitations, Treatment 776
Level 2
23-1 Subcategory Profile Data Summary 780
24-1 Subcategory Profile Data Summary 784
24-2 Status of Regulations - Effluent Limitation 785
Guidelines
24~3 Water Usage in the Sodium Bisulfate Subcategory 786
xlvi
-------
LIST OF TABLES - Continued
Page
24-4 Waste Water Flow at Plant |987 and f282 788
for Sodium Bisulfite Subcategory
24-5 Flow and Pollutant Load Data of the Sampled 789
Waste Streams for Plant f282 Producing
Sodium Bisulfite
24-6 Flow and Pollutant Load Data of the Sampled 792
Waste Streams for Plant 1586
24-7 Flow and Pollutant Load Data of the Sampled 793
Waste Streams for Plant 1987
24-8 Toxic Pollutant Raw Waste Loads 797
24-9 Summary of Raw Waste Loadings Found in 798
Screening and Verification Sampling
24-10 Toxic Pollutant Concentrations Observed 799
in Treated Effluent During Verification
Sampling
24-11 Model Plant Treatment Co'sts ' 807
24-12 Model Plant Treatment Costs 808
24-13 Model Plant Treatment Costs 809
24-14 Model Plant Treatment Costs 812
24-15 Plant Performance Evaluation Summary for 814
Conventional and Nonconventional Pollutants
24-16 Proposed Limitations, BPCTCA 818
24-17 Performance of Alterantive Technology, 824
Treatment Level 2
24-18 Performance of Alternative Technology, 825
Treatment Level 3
25-1 Subcategory Profile Data Summary 828
25-2 Status of Regulations - Effluent Limitation 829
Guidelines
25-3 Waste Source Data at Plant |672 833
xlvii
-------
LIST OF TABLES - Continued
25-4 Flowr Pollutant, Concentration, and Load 836
Data of the Sampled Waste Streams for Plant
§672 Producing Sodium Hydrosulfite
25-5 Sampling Results and Treatment System 839
Performance for Toxic Pollutants Plant
#672
25-6 Summary of Raw Waste Loadings and Concentration 840
Found at a Sodium Hydrosulfite Plant (Formate
Process)
25-7 Model Plant Treatment Costs 847
25-8 Model Plant Treatment Costs 848
\
25-9 Subcategory Performance Evaluation Summary 850
at Plant #672 for Conventional and
Nonconventional Pollutants in the Effluent
25-10 Proposed Limitations, BPCTCA 854
25-11 Proposed Limitations, BAT 858
25-12 Proposed Limitations, NSPS 862
26.2-1 Subcategory Profile Data Summary 865
26.3-1 Subcategory Profile Data Summary 867
26,4-1 Subcategory Profile Data Summary 869
26.6-1 Subcategory Profile Data Summary 872
26.9-1 Subcategory Profile Data Summary 876
26.17-1 Subcategory Profile Data Summary 884
26.23-1 Subcategory Profile Data Summary 891
26.24-1 Subcategory Profile Data Summary 893
26.27-1 Subcategory Profile Data Summary> 898
26.32-1 Subcategory Profile Data Summary 903
26.35-1 Subcategory Profile Data Summary 907
xlviii
-------
LIST OF TABLES - Continued
Page
26.36-1 Subcategory Profile Data Summary 908
26.37-1 Subcategory Profile Data Summary 911
26.38-1 Subcategory Profile Data Summary' 913
26.40-1 Subcategory Profile Data Summary 915
26.42-1 Subcategory Profile Data Summary 917
26.43-1 Subcategory Profile Data Summary ~ 920
26.44-1 Subcategory Profile Data Summary 921
xlix
-------
-------
ACKNOWLEDGEMENTS
The technical study supporting the proposed regulation was
conducted by Jacobs Environmental Division of Jacobs Engineering
Group Inc. of Pasadena, California under the direction of Mr.
Henry Cruse, Vice President and Mr. Michael Warner, Project
Manager. Major contributors were Ms. Bonnie J. Parrott, Mr.
Mazhar Mohiuddin, Mr. Mahendra L. Shah, Mr. Carl B. Johnston,
Mr. Dennis Merklin, Mr. Mark A. Jackson, Mr. Dale Newkirk, Mr.
William E. Rorie, Mr. John H. Taylor, Jr., Dr. Nelson F.
Phelan, Ms. Maureen Smith, Mr. Dale Rushneck, Dr. Ben C.
Edmondson and Dr. Martin D, Schwartz.
Cooperation and assistance of the Chemical Manufacturers
Association and the Chlorine Institute and numerous individual
corporations is appreciated for their review and comment on the
draft development document. The numerous company and plant
personnel who submitted information, opened their plants to the
program staff, and otherwise cooperated are acknowledged and
thanked for their patience and help.
The guidance and assistance of Mr. Swep Davis, Associate
Assistant Admimistrator for Water and Waste Management is
greatly appreciated.
Mr. Steve Schatzow, Deputy Assistant Administrator for
Water Planning and Standards is gratefully acknowledged for his
contributions to the project, both as an administrator and as a
member of the Office of General Counsel.
Mr. Walter J. Hunt, former Chief of the Inorganic Chemicals
and Services Industries Branch, is acknowledged for a
significant contribution in the initial stages of the project.
Ms. Susan Lepow and Staff of the Office of General Counsel
are specially acknowledged for their extensive contribution to
the drafting of the regulations and this development document.
Ms. Emily Hartnell and Mr. Richard Catz of the Office of
Analysis and Evaluation, Mr..Mark Segal, Monitoring and Data
Support Division, and Fred Talcott, Office of Planning and
Evaluation are acknowledged for their assistance.
Word processing for this project was performed by Ms.
Nancy Zrubech with assistance from Ms. Kaye Starr. Their work
is especially appreciated.
li
-------
-------
SECTION 1
CONCLUSIONS AND SUMMARY
1.1 TOXIC POLLUTANTS
For the purpose of establishing waste water effluent
limitation guidelines for existing sources and standards of
performance for new sources in this study, the following 35
inorganic chemical product subcategories were screened:
1. Chlor-Alkali
2. Hydrofluoric Acid
3. Titanium Dioxide
4. Aluminum Fluoride
5. Chrome Pigments
6. Hydrogen Cyanide
7. Sodium Bichromate
8. Copper Sulfate
9. Nickel Sulfate
10. Sodium Bisulfite
11. Sodium Hydrosulfite
12. Hydrogen Peroxide
13. Hydrochloric Acid
14. Nitric Acid
15. Sodium Carbonate
16. Sodium Metal
17. Sodium Silicate
18. Sulfuric Acid
19. Carbon Dioxide
20. Carbon Monoxide and
by-product Hydrogen
21. Silver Nitrate
22. Ammonium Chloride
23. Ammonium Hydroxide
24. Barium Carbonate
25. Boric Acid
26. Calcium Carbonate
27. Cuprous Oxide
28. Manganese Sulfate
29. Strong Nitric Acid
30. Oxygen and Nitrogen
31. Potassium Iodide
32. Sodium Hydrosulfide
33. Sodium Silicofluoride
34. Sodium Thiosulfate
35. Sulfur Dioxide
The screening studies showed that the plant process waste
waters from subcategories 1 through 11 contain the toxic metals
(see Table 3-1)f cyanide and asbestos. Very few of the organic
toxic pollutants were found in process waste streams and those
that were identified, in most cases, were present in low level
concentrations.
The screening results which indicated the presence of toxic
pollutants in significant amounts were largely confirmed by the
results of the verification program. Verification sampling
-------
accounted for 50 to 75 percent of. the current inorganic chemical
production rate in the subcategories covered.
The sources of most of the toxic pollutants found in the
raw wastes and treated effluents can be traced to specific
process-related raw materials and chemicals used in the
manufacturing operations. In the case of certain pollutants
found in widely varying amounts or with erratic frequencies of
occurrence, the precise identities of the sources remain unknown
at this timer but are suspected to be process-related.
1.2 CONTROL AND TREATMENT TECHNOLOGY
A considerable amount of toxic pollutant removal is
presently achieved in the industry by the existing control and
treatment practices. Additional removal can be accomplished by
the application of available and demonstrated technologies which
would add to or modify existing treatment systems. Recovery of
the heavy metals for value or reuse in a process does not appear
to be an attractive alternative in those industries where the
product recovery practices now in effect do not already
accomplish this.
The treatment of toxic metal-bearing waste streams results
in the production of sludges or residues which are potentially
hazardous and may require special means for handling and
disposal under the Resource Conservation and Recovery Act (RCRA)
regulations.
1.3 COSTS OF ADDITIONAL IN-PLANT TREATMENT
The estimated incremental costs of applying the candidate
BAT treatment options represent a relatively small proportion of
the investment and operating and maintenance costs already
committed to the existing BPT level treatment systems. These
costs, however, vary widely from industry to industry and are
highly dependent on site-specific factors.
1.4 SUBCATEGORIZATION
A review of the product/process basis for subcategorization
of the inorganic chemical product subcategories designated for
study revealed that certain modifications may be appropriate
in the interest of developing effective regulations. The toxic
pollutant problem per se impacts subcategorization directly only
in the Chlor-Alkali Industry where the use of graphite anodes
contributes to the generation of chlorinated hydrocarbons. In
the Titanium Dioxide Industry, major process and raw material
-------
differences justify the creation of a separate segment for the
sulfate process, the chloride process, and for the chloride
process using ilmenite ore. Consideration was given to
creating a subcategory for the combined production of
hydrofluoric acid and aluminum fluoride in view of their similar
waste characteristics and the current practice of combined
treatment at several plants. However, combining these products
into a single subcategory does not appear to offer any
regulatory advantages.
Hydrogen cyanide is produced by the Andrussow process and
as a by-product in the manufacture of acrylonitrile. By-product
hydrogen cyanide will be covered under its primary product,
acrylonitrile, in the Inorganic Chemicals Manufacturing
Category. The hydrogen cyanide subcategory includes only
manufacture by the Andrussow process.
1.5 RESTUDY OF REMANDED REGULATIONS
The Fourth Circuit, U.S. Court of Appeals remanded effluent
limitations guidelines promulgated for 11 major inorganic
chemical products. E.I. du Pont de Nemours v. Train, 541 F.2d
1018 (4th. Cir. 1976) revised in Part 430 U.S. 112 (1977). The
factors affecting the control and treatment of pollutant
discharges in those industries have been studied in response to
the remanded issues. It h'as been concluded that alternative
control and treatment technologies to those originally
considered for BAT and NSPS may be appropriate.
-------
-------
SECTION 2
RECOMMENDATIONS
•On the basis of the toxic pollutant screening and
verification results and the evaluation of applicable
technologies for discharge control and treatment, it is
recommended that effluent limitation guidelines, new source
performance standards and pfetreatment standards for new and
existing sources be proposed for the following 11 inorganic
chemical manufacturing subcategories:
Chlor-Alkali Sodium Dichromate
Hydrofluoric Acid Copper Sulfate
Titanium Dioxide Nickel Sulfate
Aluminum Fluoride Sodium Bisulfite
Chrome Pigments Sodium Hydrosulfite
Hydrogen Cyanide
Table 2-1 summarizes the proposed regulations for Best
Practicable Control Technology Currently Available (BPT).
Summaries of proposed regulations for Best Available Technology
(BAT)r Pretreatment Standards, and New Source Performance
Standards are given in Tables 2-2, 2-3, 2-4, 2-5, and 2-6.
These tables indicate that Chlor-Alkali has been divided into
two segments and Titanium Dioxide in three segments before
listing the numerical effluent limitations for the proposed
regulations.
In addition, in the following subcategories, although toxic
pollutant discharges have not been found in significant
quantities, discharge of conventional and nonconventional
pollutants should be controlled by the permitting authority.
Hydrogen Peroxide Sodium Metal
Hydrochloric Acid Sodium Silicate
. Nitric Acid Sulfuric Acid
Sodium Carbonate
-------
TABLE 2-1. SUMMARY OP PROPOSED REGULATIONS -
KRUCTICAHiE CONTROL TEOTSDLOGY
CURRENTLY AVAILABLE (BPT)
Effluent Limitations
QLUJUO, wsyu iy
Chlor-alkali,
Mercury Cells
jrcticuiRSUtiL
TSS
Mercury
pH
Max
30-day Avg
kg/kkg (or lb/1000
0.32-
0.00014
24-hr
Max
Ib. ) of product
0.64
0.00028
pH Range
6.0 to 9.0
CKLor-alteli, TSS
Diaphragm Cells Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Hydrofluoric
Acid
Sodium
Bichromate
pH
TSS
Plouride (T)
Antimony (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
PH
TSS
Hexavalent
0.51
0.00088
0.0044
0.010
0.0044
0.0044
5.3
2.9
0.044
0.0055
0.027
0.016
0.0093
0.030
0.22
1.1
0.0023
0.011
0.026
0.011
0.011
11.0
6.1
0.088
0.011
0.054
0.033
0.019
0.060
0.44
6.0 to 9.0
6.0 to 9.0
Titanium
Dioxide
(sulfate
process)
Chromium
Chromium (T)
pH
TSS
Icon (T)
Arsenic (T)
Antimony (T)
Cadmium (T)
Chromium (T)
0.0044
0.00050
30
1.2
0.24
0.38
0.070
0.070
0.0088
0.0009
6.0 to 9.1
110
4.1
0.46
0.71
0.11
0.13
(contajnued)
-------
2-1. Qontinued
Subcategory
Effluent Limitations
Parameter
Max"
30-c3ay Avg
24 -hr
Max
pH Range
..-r,. _ -1 i .. _. ni, -- -
Titanium
Dioxide
(sulfate
process)
Titanium
Dioxide
(Chloride
Process)
Titanium Diox-
ide (Chloride
Ilmenite Ero-
cess)
Aluminum
Fluoride
Copper Sulfate
kg/kkg (or lb/1000 Ib.) of product
-•,!••
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
PH
TSS
Iron (T)
Chromium (T)
pH
TSS
Iron (T)
JtatinDny (T)
Arsenic (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
pH
TSS
Fluoride (T)
Chromium (T)
Nickel (T)
pH
TSS
Copper (T)
Nickel (T)
Selenium (T)
pH
0.24
0.14
0.10
0.24
6.4
0.25
0.14
7.7
0.30
0.096
0.060
0.012
0.012
0.060
0.036
0.024
0.060
1.2
0.63
0.0012
0.0024
0.023
0.0010
0.0020
0.00050
0.46
0.21
0.18
0.50
23
0.84
0.027
28
1.0
0.18
0.11
0.019
0.023
0.11
0.054
0.046
0.013
2.4
1.3
0.0024
0.0048
0.069
0.0030
0.0060
0.0015
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
-------
2-1. Continued
Subcategory
Effluent Limitations
I&rametex
Max, 24-hr
30-day Avg Max
kg/kkg (or lb/1000 Ib.) of product
pH Range
Hydrogen
Cyanide
TSS
Jtoraonia-N
Cyanide (Free)
Cyanide (T)
2.0
4.3
0.016
0.23
5.4
12
0.043
0.65
Nickel Sulfate
Chrome Pigments
Sodium Bisul-
fite
Sodium Hydro-
sulfite
pH
TSS
Nickel (T)
pH
TSS
Antimony (T)
Cadmium (T)
Chrcanium (T)
Copper (T)
Lead {T)
Nickel (T)
Zinc (T)
PH
TSS
COD
Chrcmium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc•(T)
pH
TSS
COD
0.032
0.0020
3.9
0.051
0.020
0.12
0.042
0.15
0.018
0.12
0.033
1.2
0.00017
0.00075
0.00045
0.00030
0.00075
0.12
13
0.096
0.0060
9.4
0.12
0.048
0.29
0.10
0.36
0.043
0.29
0.12
3.6
0.00032
0.0014
0.00086
0.00057
0.0014
0.44
46
6.0 to 10.5
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
-------
TABLE 2-2, OP REOjILSTIONS -
BEST &WEL&BEE TEaWOLOGf (BRT)
Effluent Limitations
ka/kkg (or lb/1000 Ib.) of product
Chlor-alkali
Mercury Cells
Oiler-alkali
Diaphragm
Cells
Hydrofluoric
£cid
Sodium
Dichromate
Arsenic (T)
Cadmium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Silver (T)
Zinc (T)
Total Residual
Chlorine
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Total Residual
Chlorine
Fluoride (T)
Jtotimony (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Chromium (T)
Hexavalent
Chromium
Nickel (T)
Zinc (T)
0.00021
0.00011
0.00011
0.00034
0.00010
0.00021
0.00015
0.00042
0.00042
0.00044
0.0035
0.0019
0.00088
0.0035
0.0018
1.0
0.023
0.0013
0.0097
0.0020
0.0050
0.017
0.0022
"
0.00035
0.0012
0.0033
0.00046
0.00024
0.00024
0.00074
0.00022
0.00046
0.00032
0.00092
0.00071
0.00097
0.00077
0.0042
0.0019
0.0077
0.0030
2.2
0.047
0.0027
0.019
0.0040
0.010
0.035
0.0045
0.00070
0.0024
0.0066
(continued)
-------
TABLE 2-2. Continued
Subcategory
Titanium
Dioxide
StxLfate
Process
Titanim
Dioxide
Chloride
Process
Titanium
Dioxide
Chloride
Ilmenite
Process
Aluminum
Fluoride
Chrome Pigments
Parameter
Iron (T)
Arsenic (T)
Antimony (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Iron (T)
Chrcmium (T)
Iron (T)
Antimony (T)
Arsenic (T)
Cadmium (T)
Chromiurti (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Fluoride (T)
Chromium (T)
Nickel (T)
Antimony (T)
Cadmium (T)
Chrcmium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Effluent Limitations
«lfax
30-day Av§
kg/kkg (or lb/1000
1.2
0.24
0.38
0.070
0.070
0.24
0.14
0.10
0.24
0.25
0.014
0.30
0.096
0.060
0.012
0.012
0.060
0.036
0.024
0.060
0.036
0.00048
0.0020
0.051
0.020
0.12
0.042
0.15
0.018
0.12-
24 -hr
Max
Ib.) of product
4.1
0.46
0.71
0.11
0.13
0.46
0.21
0.18
0.52
0.84
0.027
1.0
0.18
0.11
0.019
0.023
0.11
0.054
0.046
0.013
0.75
0.00096
0.0040
0.12
0.048
0.29
0.10
0.36
0.043
0.29
10
-------
TABLE 2-2. Continued
Effluent Limitations
Subcategory
Parameter
Max 24 -hr
30-day Avg Max
kg/kkg (or Ib/lOQQ Ib.) of product
Copper Sulfate Antimony (T)
Arsenic (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel ,(T)
Selenium (T)
Zinc (T)
Hydrogen .Ammonia - N
Cyanide Cyanide (Free)
Cyanide (T)
Total Residual
Chlorine
Nickel Sulfate AntLomony (T)
Chromum (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Sodium COD
Bisulfite Chrcraium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Sodium COD
Hydrosulfite Zinc (T)
Nickel (T)
Lead (T)
Chromium (T)
0.00038
0.00047
0.000047
0.000047
0.00038
0.000047
O.OOOQ94
0.000094
0.00038
4.3
0.16
0.23
0.011
0.00027
0.000010
0.00027
0.000034
0.00014
0.00027
1.2
0.00017
0.00075
0.00045
0.00030
0.00075
13
, 0.0024
0.00094
0.0014
0.00047
0.00072
0.00089
0.000089
0,000089
0.00072
0.000089
0.00018
0.00018
0'. 00072
12
0.043
0.65
0.031
0.00081
0.000034
0.00081
0.00010
0.00042
0.00080
3.6
0.00032
0.0014
0.00086
0.00057
0.0014
46
0.0046
0.0018
0.0027
0.00087
11
-------
oanr,E 2-3. SUMMARY OF PROPOSED EHSJIATIONS -
PREHCKEMMENT SESMBBDS FOR EXISTING
SOURCES (PSES)
Subcategory
CKLor-alkali
Mercury Cells
CMor-alkali
Diaphragm
Cells
Hydrofluoric
Acid
Sodium Dichro-
mate
Titanium
Dioxide
Sulfate
Brocess
Parameter
Arsenic (T)
Cadmium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Silver
Zinc
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
ELuoride (T)
Antimony (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Chromium (T)
Hexavalent
Chromium
Nickel (T)
Zinc (T)
Iron (T)
Arsenic (T)
Antimony (T)
Cadmium (T)
Chromium (T)
Copper (T)
Effluent Limitations
Max
j
(mg/1)
0.10
0.050
0.050
0.16
0.048
0.10
0.070
0.20
0.050
0.40
0.22
0.10
0.40
30 •
0.70
0.040
0.29
0.060
0.15
0.52
0.32
0.050
0.17
0.47
2.5
0.50
0.80
0.15
0.14
0.50
30-day
^vg
or (kg/kkg)
0.00021
0.00011
0.00011
0.00034
0.00010
0.00021
0.00015
0.00042
0.00044
0.0035
0.0019
0.00088
0.0035
1.0
0.023
0.0013
0.0097
0.0020
•> 0.0050
0.017
0.0022
0.00035
0.0012
0.0033
1.2
0.24
0.38
0.07
0.07
0.24
24-hr
Max
(mg/1) or
0.22
0.11
0.11
0.35
0.10
0.22
0.15
0.44
0.11
0.88
0.48
0.22
0.88
66
1.4
0.080
0.58
0.12
0.30
1.0
0.64
0.10
0.34
0.94,
8.5
0.95
1.5
0.24
0.27
0.95
(kg/kkg)
0.00046
0.00024
0.00024
0.00074
0.00022
0.00046
0.00032*
0.00092
0.00097
0.0077
0.0042
0.0019
0.0077
2.2
0.047
0.0027
0.019
0.0040
0.010
0.035
0.0045
0.0070
0.0024
0.0066
"
4.1
0.46
0.71
0.11
0.13
0.46
(continued)
12
-------
TlffiLE 2-3. Continued
Subcategory
Effluent Limitations
Parameter
-Max 30-day
Avg
(mg/L) or (kg/kkg)
24-hr
Max
(mg/1) or (kg/kkg)
Titanium
Dioxide
Chloride
Process
Titanium
Dioxide ,
Chloride
Ilmenite
Process
Aluminum
Fluoride
Chrome
Pigments
Copper
Sulfate
Lead (T)
Nickel (T)
Zinc (T)
Iron (T)
Chromiun (T)
Iron (T)
Antimony (T)
Arsenic (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Fluoride (T}
dsrcmium (T)
Nickel (T)
Antimony (T)
Cadmium (T)
Ohrcsnium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Copper (T)
Nickel (T)
Arsenic (T)
Selenium (T)
Cadmium (T)
Zinc (T)
0.30
0.20
0.50
2.5
0.14
2.5
0.80
0.50
0.10
0.10
0.50
0.30
0.20
0.50
30
0.040
0.17
0.48
0.19
1.1
0.40
1.4
0.17
1.1
0.40
0.10
0.50
0.10
0.050
0.40
0.14
0.10
0.24
0.25
0.014
0.30
0.096
0.060
0.012
0.012
0.060
0.036
0.024
0.060
0.36
0.00048
0.0020
0.051
0.020
0.12
0.042
0.15
0.018
0.12
0.00038
0.000094
0.00047
0.000094
0.000047
0.00038
0.45
0.37
1.1
8.4
0.27
8.5
1.5
0.95
0.16
0.19
0.95
0.45
0.38
1.1
63
0.080
0.34
1.2
0.46
2.6
0.96
3.4
0.41
2.6
0.76
0.19
0.95
0.19
0.095
0.76
0.21
0.18
0.52
0.84
0.027
1.0
0.18
0.11
0.019
0.023
0.11
0.054
0.046
0.013
0.75
0. 00096
0.0040
0.12
0.048
0.29
0.10
0.36
0.043
0.29
0.00072
0.00018
0.00089
0.00018
0.000089
0.00072
(continued)
13
-------
3&ELE 2-3. Continued
Effluent Limitations
Subcategory
I&rameter
Max 30-day
Avg
(mg/1) or (kg/kkg)
24-hr
Max
(mg/1) or (kg/kkg)
Hydrogen
Cyanide
Nickel Sol-
fate
Sodium Bi~
sulfate
Sodium H^3ro-
sulfite
Ctoamium (T)
Lead (T)
Antimony (T)
Cyanide (Eree)
Cyanide (T)
Ararc>nia-$I "
Antimony (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
COD
Chromium (T)
Zinc (T)
Copper (T)
Lead (T)
Nickel (T)
COD
Zinc (T)
Nickel (T)
Lead ,{T)
Chromium (T)
0.050
0.05
0.40
0.27
4.0
75
0.40
0.050
0.40
0.050
0.20
0.40
680
0.11
0.50
0.50
0.30
0.20
2700
0.50
0.20
0.30
0.10
0.000047
0.000047
0.00038
0.016
0.23
4.3
0.00027
0.00010
0.00027
0.000034
0.00014
0.00027
1.2
0.00017
, 0.00075
0.00075
0.00045
0.003
13
0.0024
0.00094
0.0014
0.00047
0.095
0.095
0.76
0.74
11
210
1.2
0.15
1.2
0.15
0.60
1.2
2400
0.22
d.O
1.0
0.57
0.38
9700
0.95
0.38
0.57
0.19
0.00089
0.000089
0.00072
0.043
0.65
12
0.00081
0.000034
0.00081
0.00010
0.00042
0.00080
3.6
0.00032
0.0014
0.0014
0.00086
0. 00057
46
0.0046
0.0018
0.0027
0.00089
-------
TABLE 2-4. SUMmRY OP PROPOSED BEGUIATIONS -
NEW SOURCE PERFORMANCE STANDARDS
(NSPS)
Subcategory
Chlor-alkali
Mercury Cells
Chlor-alkali
Diaphragm Cells
Hydrofluoric
Acid
Sodium Dichro-
mate
Parameter
TSS
Arsenic (T)
Cadmium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Silver
Zinc
Total Residual
Chlorine
pH
TSS
Chromium (T)
Lead (T)
Total Residual
Chlorine
PH
'TSS
Fluoride (T)
Chromium (T)
Nickel (T)
Zinc (T)
PH
TSS
Chromium (T)
Effluent Limitations
Max
30-day Avg
kg/kkg (or lb/1000
0.32
0.00021
o.oooii
1 0.00011
0.00034
0.00010
0.00021
0.00015
0.00042
0.00042
0.10
0.00047
0.00044
0.0018
*
0.41
0.18 •
0.00024
0.0009
0.0030
0.18
0.0022
24 -hr
Max
Ib. ) of product
0.64
0.00046
0.00024
0.00024
0.00074
0.00022
0.00046
0.00032
0.00092
0.00071
0.20
0.00097
0.00097
0.0030
0.86
0.38
0.00048
0.0018
0.0060
0.35
0.0045
pH Range
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
Hexavalent Chrom- ,
ium
Nickel (T)
Zinc (T)
pa
0.00035
0.0012
0.0033
0.00070
0.0024
0.0066
6.0 to 9.0
(continued)
15
-------
2-4. Continued
Effluent Limitations
kg/Wcg (or lb/1000 Ib.) of product
Titanitm , Diox-
ide (Sulfate
toocess)
Titanitm Diox-
ide ' (Chloride
Srocess)
Titanium Diox-
ide (Chloride
Hmenite pro-
cess)
Muminum
Fluoride
TSS
Iron (T)
Arsenic (T)
Antimony (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
pH
TSS
Iron (T)
Chromium (T)
PH
*
TSS
Icon (T)
Antimony (T)
Arsenic (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
FH
TSS
ELuoride (T)
Chromium (T)
Nickel (T)
pH-
30
1.2
0.24
0.38
0.07
0.07
0.24
0.14
0.10
0.24
4.5
0.18
0.005
1.2
0.050
0.025
0.016
0.0023
0.0012
0.0090
0.0019
O.Q053
0.015
0.81
0.36
0.00050
0.0020
110
4.1
0.46
0.71
0.11
0.13
0.46
0.21
0.18
0.52
16
0.59
0.01
4.3
0.17
0.048
0.030
0.0037
0.0023
0.017
0.0029
0.010
0.032
1.7
0.75
0.0010
0,0040
6.0 to 9.0
6,0 to 9.0
6,0 to 9.0
6.0 to 9.0
16
-------
2-4. Continued
Subcategory
Effluent Limitations
Parameter
Max
30-day Avg
24 -hr
Wax
pH Range
kg/kkg (or lb/1000 Ib.) of product
Chrome Pigments
Copper Sulfate
Hydrogen Cyan-
ide
Nickel Sulfate
TSS
• Antimony (T) •
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Zinc (T)
PH ;
TSS
Copper (T)
Nickel (T)
Arsenic (T)
Selenium (T)
Cadmium (T)
Zinc (T)
Chromium (T)
Lead (T)
Antimony (T)
PH
TSS
Cyanide (Free)
Cyanide (T)
Ammonia-N
total Residual
Chlorine
pH
TSS
Antimony (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
PH
3.9
0.042
0.0011
0.0053
0.0053
0.0053'
0.0011
0.0053
0.0021
0.023
0.00038
0.000094
0.00047
0.000094
0.000047
0.00038
0.000047
0.000047
0.00038
2.0
0,016
0.23
4.3
0.011
0.032
0.00027
0.000010
0.00027
0.000034
0.00014
0.00027
9.4
0.10
0.0026
0.013
0.013
0.013
0.0026
0.013
0.0050
6.0 to 9.0
0.069
0.00072
0.00018
0.00089
0.00018
0.000089
0.00072
0.000089
0.000089
0.00072
6.0 to 9.0
5.4
0.043
0.65
12
0.031
6.0 to 10.5
0.096
0.00081
0.000034
0.00081
0.00010
0.00042
0. 00080
6.0 to 9.0
(continued)
17
-------
TftBUS 2-4. Continued
Subcategory
Parameter
Effluent Limitations
Max
30-day Avg
24-hr
Max
kg/kkg (or lb/1000 lb.) of product
pH Range
Sodium Bisul-
fite
Sodium Hydro-
sulfite
TSS
ODD
Qiranium (T)
Zinc (T)
Copper (T)
L^d (T)
Nickel (T)
PH
TSS
K30D
Chranium (T)
Lead (T)
Nickel (T)
Zinc
pH
0.033
1.2
0.00017
0.00075
0.00075
0.00045
0.00030
0.18
13
0.00047
0.0014
0.00094
0.0024
0.12
3.6
0.00032
0.0014
0.0014
0.00086
0.00057
0.44
46
0.00089
0.0027
0.0018
0.0046
6.0 to 9.0
6.0 to 9.0
18
-------
TABLE 2-5. SUM^EY OF PEDPOSED REGULATIONS -
STANDARDS FOR NEW SOURCES
(PSNS)
Effluent Limitations
buDcacegory jraiamecer • —
, 30-day Avg „ , t
(rng/L) or (Rg/kkg)
Chlor-alkali
Mercury tells
Chlor-alkali
Diaphragm
Cells
Hydrofluoric
Acid
Sodium Di-
chromate
Titanium Di-
oxide (sul-
fate pro-
cess)
Titanium
Dioxide
(chloride
process)
Arsenic (T)
Cadmium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Silver
Zinc
Chromium (T)
Lead (T)
Fluoride (T)
Chromium (T)
Nickel (T)
Zinc (T)
Chromium (T)
Hexavalent
Chromium
Nickel (T)
Zinc (T)
Iron (T)
Arsenic; (T)
Antimony (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Iron (T)
Chromium (T)
0.10
0.050
0.050
0.16
0.048
0.10
0.070
, 0.20
0.050
0.050
30
0.040
0.15
0.50
- 0.32
0.050
0.17
0.47
2.5
0.50
0.80
0.15
0.14
0.50
0.30
0.20
0.50
1.8
0.05
0.00021
0. 00011
0.00011
0.00034
0.00010
0.00021
0.00015
0.00042
• 0.00044
0.00044
0.18
0.00024
0.00090
0.0030
0.0022
0,00035
0.0012
0.0033
1.2
0.24
0.38
0.07
0.07
0.24
0-.14
0.10
0.24
0.18
0.005
24 -hr
Max
(mg/L) or
0.22
0.11
0.11
0.35
0.10
0.22
0.15
0.44
0.11
0.11
63
0.080
0.30
1.0
0.64
0.10
0.34
0.94
8.5
0.95
1.5
0.24
0.27
0.95
0.45
0.37
1.1
5.9
0.10
(kg/kkg)
0.00046
0.00024
0.00024
0.00074
0.00022
0.00046
0.00032
0.00092
0.00097
0.00097
0.38
0. 00048
0.0018
0.0060
0.0045
0.00070
0.0024
0.0066
4.1
0.46
0.71
0.11
0.13
0.46
0.21
0.18
0.52
0,59
0.01
(continued)
19
-------
ISELE 2-5. Continued
Subcategory
Titanium
Dioxide
Chloride
Ilmenite
Process
Aluminum
Fluoride
t
Chrome
Pigments
*
Copper
Sulfate
Hydrogeij
Cyanide
Parameter
Iron (T).
Antimony (T).
Arsenic Crl
Cadmium (3?)
Chromium (T)
Copper (T) •
Lead (T)
Nickel (T)
Zinc (T)
Fluoride (T)
Chromium (T)
Nickel (T)
Antimony (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Zinc (T)
Antimony (T)
Arsenic (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Selenium (T)
Zinc' (T)
Atimonia - N
Cyanide (Free)
Cyanide (T)
Effluent Limitations
30-
(mg/1)
1.6
0.80
0.50
0.075
0.040
0.029
0.060
0.17
0.47
30
0.04
0.17
0.40
0.010
0.050
0.050
0.050
0.010
0.050
0.020
0.40
0.50
0.050
0.050
0.40
0.050
0.10
0.10
0.40
75 ,
0.27
4.0
Max
•day Avg
or (kg/kkg)
0,050
Q..025
0*016
0.0023
0.0012
0.0090
0.0019
0.0053
0.015
0.36
0.00050
0.0020
0.042
0.0011
0.0053
0.0053
0.0053
0.0011
0.0053
0.0021
0.00038
0.00047
0.000047
0.000047
0.00038
0.000047
0.000094
0.000094
0.00038
4.3
0.016
0.23
24-hr
Max
(mg/L) or (kg/kkg)
5,4
1.5
0.95
0.12
0.076
0.055
0.090
0.32
0.99
63
0.08
0.34
0.96
0.024
0.12
' 0.12
0.12
0.024
0.12
0.048
0.76
0.95
0.095
0.095
0.76
0.095
0.19
0.19
0.76
210
0.74
11
tt, 017
-0.048
Q,C130
0,0037
0.0023
0.017
0.0029
0.010
0.032
0.75
0.0010
0.0040
0.10
0.0026
0.013
0.013
0.013
0.0026
0.013
0.0050
0.00072
0.00089
0.000089
0.000089
0.00072
0.000089
0.00018
0.00018
0.00072
12
0.043
0.65
(continued)
20
-------
TABLE 2-5. Continued
Effluent Limitations
Subcategory Parameter
Max
30-day Avg
(mg/1) or (kg/kkg)
24-hr
Max
(rog/1) or (kg/kkg)
Nickel
Sulfate
Sodium
Bisulfite
Sodium
Hydrosulfite
Antimony (T)
Chromium (T).
Copper (,Tl
Lead (T)
Nickel (T)
Zinc (T)
COD
Chromum (T)
Zinc (T)
Copper (T)
Lead (T)
Nickel (T)
COD
Zinc (T)
Nickel (T)
Lead (T)
Chromium (T)
0.40
0.05
0.40-
0.05
0.20
0.40
680
0.11
0.5
0.5
0.3
0.2
2700
0.50
0.20
0.30
0,10
Q. 00027
1,2
0,000010 6,15
Q.QQQ27
0.000034
0.00014
0.00027
, 1.2
0.00017
0.00075
0.00075
0.00045
0.00030
13
0.0024
0.00094
0.0014
0.00047
1,2
0.15
0.60
1.2
2400
0.22
1.0
1.0
0.57
0.38
9700
0.95
0.38
0.57
0.19
0.00081
0,00,0034
O..QQ081
0.00010
0.00042
0.00080
3.6
0.00032
0.0014
0,0014
0.00086
0.00057
46
0.0046
0.0018
0.0027
0.00089
21
-------
TKW..E 2-6. SDMgaKY OF PROPOSED REGDLATJDNS -
EEST OTNVENTICNMj PCUinSNT COKHTOL
TESCHNOLOGY (BCT)
Subcategory
Effluent Limitations
Parameter
Max
30-day Avg
24 -far
Max
kg/Wcg (or lb/1000 Ib.) of poxiduct
pH Range
Chlor-alkali
Diaphragm Cell TSS
pH
TSS
pH
Jteid
0.36
2.3
0.72
4.8
6.0 to 9.0
6.0 to 9.0
22
-------
SECTION 3
INTRODUCTION
3.1 AUTHORITY
3.1.1 The Federal Water Pollution Control Act Amendments
The Federal Water Pollution Control Act (the Act)
Amendments of 1972, 33 USC 1251 et seq., stated the .national
goal of attaining by July 1, 1983, a water quality which
provides for the protection and propagation of fish and
shellfish, for recreation in or on the nation's waters, and the
goal of eliminating the discharge of pollutants into navigable
waters by 1985.
Purpose and Authority
The Federal Water pollution Control Act Amendments of 1972
established a comprehensive program to "restore and maintain the
chemical, physical, and biological integrity of the Nation's
waters," Section 101(a). By July 1, 1977, existing industrial
dischargers were required to achieve "effluent limitations
requiring the application of the best practicable control
technology currently available" ("BPT"), Section 301(b) (1) (A);
and by July 1, 1983, these dischargers were,required to achieve
"effluent limitations requiring the application of the best
available technology economically achievable...which will
result in reasonable further progress toward the national goal
of eliminating the discharge of all pollutants" ("BAT"), Section
301(b)(2)(A). New industrial direct dischargers were required
to comply with Section 306 new source performance standards
("NSPS"), based on best available demonstrated technology? and
new and existing dischargers to publicly owned treatment works
("POTW") were subject to pretreatment standards under Sections
307 (b) and (c) of the Act. While the requirements for direct
dischargers were to be incorporated into National Pollutants
Discharge Elimination System (NPDES) permits issued under
Section 402 of the Act, pretreatment standards were made
enforceable directly against dischargers to POTW (indirect
dischargers).
23
-------
Although Section 402(a)(1) of the 1972 Act authorized the
setting of requirements for direct dischargers on a case-by-case
basis, Congress intended that for the most part control
requirements would be based on regulations promulgated by the
Administrator of EPA. Section 304 (b) of the Act required the
Administrator to promulgate regulations providing guidelines
for effluent limitations setting forth the degree of effluent
reduction attainable through the application of BPT and BAT.
Moreover, Sections 304 (c) and 306 of the Act required
promulgation of regulations for NSPS, and Sections 304(f),
307(b), and 307 (c) required promulgation of regulations for
pretreatment standards. In addition to these regulations for
designated industry categories, Section 307(a) of the Act
required the Administrator to promulgate effluent standards
applicable to all dischargers of toxic pollutants. Finally,
Section 501(a) of the Act authorized the Administrator to
prescribe any additional regulations "-necessary to carry out his
functions" under the Act.
The EPA was unable to promulgate many of these regulations
by the dates contained in the Act. In 1976, EPA was sued by
several environmental groups, and in settlement of this lawsuit
EPA and the plaintiffs executed a "Settlement Agreement" which
was approved by the Court. This Agreement required EPA to
develop a program and adhere to a schedule for promulgating BAT
effluent limitations guidelines, pretreatment standards, and
new source performance standards for 65 "priority" pollutants
and classes of pollutants for 21 major industries. See Natural
Resources Defense Council, Inc. v. Train, 8 ERG 2120 (D.D.C.
1976), modified March 9, 1979.
On December 27, 1977, the President signed into law the
Clean Water Act of 1977. Although this law makes several
important changes in the Federal water pollution control
program, its most significant feature is its incorporation of
several of the basic elements of the Settlement Agreement
program for toxic pollution control. Sections 301 (b) (2) (A) and
301(b) (2) (C) of the Act now require the achievement by July 1,
1984 of effluent limitations requiring application of BAT for
"toxic" pollutants, including the 65 "priority" pollutants and
classes of pollutants which Congress declared "toxic" under
Section 307(a) of the Act. Likewise, EPA's programs for new
source performance standards and pretreatment standards are now
aimed principally at toxic pollutant controls. Moreover, to
strengthen the toxics control program Section 304 (e) of the Act
authorizes the Administrator to prescribe "best management
practices" ("BMPs") to prevent the release of toxic and
hazardous pollutants from plant site runoff,spillage or leaks,
sludge or waste disposal, and drainage from raw material
storage associated with, or ancillary to, the manufacturing or
treatment process.
24
-------
In keeping with its emphasis on toxic pollutants, the Clean
Water Act of 1977 also revises the control program for non-toxic
pollutants. Instead of BAT for "conventional" pollutants
identified under Section 304(a)(4) (including biochemical
oxygen demand, suspended solids, fecal coliform and pH), the new
Section 301(b)(2)(E) requires achievement by July 1, 1984, of
"effluent limitations requiring the application of the best
conventional pollutant control technology" ("BCT"). The
factors considered in assessing BCT for an industry include the
cost of attaining a reduction in effluents and the effluent
reduction benefits derived compared to the costs and effluent
reduction benefits from the discharge of publicly owned
treatment works (Section 304(b) (4) (B). For non-toxic,
nonconventional pollutants, Sections 301(b)(2)(A) and (b) (2) (F)
require achievement of BAT effluent limitations within three
years after their establishment or July 1, 1984, whichever is
later, but not later than July lr 1987.
The purpose of these proposed regulations is to provide
effluent limitations guidelines for BPTr BAT, and BCT, and to
establish NSPS, pretreatment standards for existing sources
(PSES), and pretreatment standards for new sources (PSNS), under
Sections 301, 304, 306, 307, and 501 of the Clean Water Act.
The United States Environmental Protection Agency (the
Agency) was entrusted with the responsibility to carry out the
requirements of the Act, and initiated an intensive effort to
develop the necessary regulatory means which would achieve the
stepwise reduction and elimination of pollutant discharge
practices in all major U.S. Industries. For the Inorganic
Chemicals Manufacturing Point Source Category, the Agency
designed a comprehensive, two phase program to identify the
control parameters and establish the technological basis for
regulations development. Phase I covered 22 Major Inorganic
Chemical Products (1) , and the fjinal regulations for these
industrial subcategories were published in the Federal Register
on March 12, 1974. The regulations included specific numerical
effluent limitations and •standards of performance for both
existing and new sources. Zero-discharge requirements
specified for many of the subcategories were to be applied
either at the 1977 BPT step or later. Phase II of the Agency's
effort resulted in the promulgation of BPT based effluent
limitations for an additional group of 27 subcategories referred
to as Significant Inorganic Chemical Products (2). The interim
final regulations were published on May 22, 1975. Taken
together, the two groups of regulations cover 49 inorganic
chemical subcategories many of which include more than one
specific chemical product. Although some toxic pollutant
parameters were covered in cases where a direct relationship to
the process was obvious (e.g., mercury and/or lead in the Chlor-
Alkali Industry), the'main thrust of the regulations was the
25
-------
control of the bulk pollutant parameters which accounted, in
terms of quantity, .for most of fche pollution loading of
navigable waters attributable to the manufacture of inorganic
chemicals.
3.1,2 Court Remand of Regulations
On March 10, 1976, the United States Court of Appeals for
the Fourth Circuit decided in E.I. duPont de Nemours & Company,
et al. v. Train, 541 F.2d 1018 (4th Cir. 1976), to set aside and
remand for reconsideration a number of general definitions and
specific -discharge regulations promulgated in 1974. These
regulations are all within Title 40, Parts 401 and 415 of the
Code of Federal Regulations and are listed below:
General Provisions
401,11 (i) - Definition of effluent limitations
401.11 (g) - Definition of process waste water
401.11 (r) ~ Definition of process waste water
pollutant
Chlor-Alkali
415.63 - BATEA
Hydrochloric Acid
415.72 - BPCTCA
415.73 - BATEA
415.75 - New sources
Hydrofluoric Acid
415.82 - BPCTCA
415.83 - BATEA
415.85 - New sources
Hydrogen Peroxide
415.93 - BATEA
415.95 - New sources
Nitric Acid
415.102 - BPCTCA
415.103 - BATEA
415.105 - New sources
26
-------
Sodium Carbonate
415.152 - BPCTCA
415.153 - BATEA
415.155 - New sources
Sodium Bichromate
415.173 - BATEA
Sodium Metal
415.182 - BPCTCA
415.183 - BATEA
415.185 - New sources
Sodium Silicate
415.192 - BPCTCA
415.193 - BATEA
415.195 - New sources
Sulfuric Acid
415.210 - Applicability
415.212 - BPCTCA
415.213 - BATEA
415.215 - New sources
Titanium Dioxide
415.220 - Applicability
415.222 - BPCTCA
415.223 - BATEA
415.225 - New sources
, -,,
For the most part, the main target of the remand was the
zero discharge regulations from which the industry petitioners
sought relief on grounds of technological infeasibility. During
1975, the Agency funded a special study of the remand issues (3)
and was prepared to propose amended regulations. Where
appropriate, the results of that study are included in an
Addendum to the present report covering those remanded
regulations for subcategories which have been excluded from the
present study.
Following the court remand of the Phase I final
regulations, the Agency revoked the Phase II interim final and
proposed regulations published in May, 1975, for Aluminum
Fluoride, Chrome Pigments, Hydrogen Cyanide, and Sodium
Silicofluoride. -In €his instance, the Agency's intent was to
reconsider the specific effluent limitations established for
27
-------
these industries (1977 step) in the light of information made
available on process differences between plants and additional
data on the actual concentrations and treatability of the
regulated discharge constituents. The information was presented
to the Agency in the form of various documents prepared by
members of the industries concerned. These sources are also
cited in the appropriate sections of this report.
3.1.3 The Settlement Agreement
A consent decree was issued in a suit filed by fdur
environmental groups in Natural Resources Defense Council v.
Train, 8 EEC ,2120 (June 8, 1976) modified 12 IRC 1833 (December
15, 1978) . The consent decree contained a Settlement Agreement
wherein the Agency agreed to regulate 65 toxic pollutants under
Sections 301, 304, 306, and 307 of the Act in accordance with
the schedule and provisions stipulated. The original list of 65
chemicals and classes of chemicals attached to the Settlement
Agreement was redefined to cover 129 chemical substances,
including specific organic compounds, pesticides and their
metabolites, polychlorinated' biphenyls (PBC's), cyanide, 13
heavy metals and asbestos. Table 3-1 lists the 129 toxic
pollutants (sometimes re'ferred to in the literature as "priority
pollutants11) .
TABU! 3-1. RECOMMENDED IiIST OF TOXIC POLLUTANTS
Compound Name
1. *Acenaphthene
2. *Acrolein
3. *Acrylonitrlle
4, *Benzene
5. *Benzidine
6. *Carbon tetrachloride (tetrachloromethane)
*Chlorinated benzenes (other than dichlorobenzenes)
7. *Chlorobenzene
8. 1,2,4-Trichlorobenzene
9. Hexachlorobenzene
*Chlorinated ethanes (including 1,2-dichloroethane,
1,1,1,-trichloroethane and hexachloroethane)
10. 1.2-Dichloroethane
11. 1,1,1-Trichloroethane
28
-------
12. Hexachloroethane
13. 1,1-Diehloroethane
14. 1,1,2-Trichloroethane
15. 1,1,2,2-Tetrachloroethane
16. Chloroethane
*Chloroalkyl ethers (ehloromethylr chloroethyl and
mixed ethers)
17. Bis(chloromethyl) ether
18. Bis(2-chloroethyl) ether
19. 2-Chloroethyl vinyl ether (mixed)
*Chlorinated naphthalene
20. 2-Chloronaphthalene
*Chlorinated phenols (other than those listed
elsewhere? includes trichlorophenols
and chlorinated cresols)
21. 2,4,6-Trichloro.phenol
22. Parachlorometa cresol
23. *Chloroform (trichloromethane)
24. *2-Chlorophenol
*Dichlorobenzenes
25. Ir2-Dichlorobenzene
26. Ir3-Dichlorobenzene
27. 1,4~Dichlorobenzene
*Dichlorobenzidine
28. 3r3'-Dichlorobenzidine
29
-------
*Dichloroethylenes (lrl-<3ichloroethylene and
1,2-dichloroethylene)
29. 1,1-Dichloroethylene
30. Ir2-Trans~dichloroethylene
31. *2r4-Diehlorophenol
*Dichloropropane and dichloropropene
32. 1,2-Dichloropropane
33. 1,2-Dichloropropylene (1,3-dichloropropene)
34. *2,4-Dimethylphenol
*Dinitrotoluene
35. 2,4-Dinitrotoluene
36. 2r6-Dinitrotoluene
37. *lr2-Diphenylhydrazine
38. *Ethylbenzene
39. *Fluoranthene
*Haloethers (others than those listed elsewhere)
40. 4-Chlorophenyl phenyl ether
41. 4-Bromophenyl phenyl ether
42. Bis(2-chloroisopropyl) ether
43. Bis(2-chloroethoxy) methane
*Halomethanes (other than those listed elsewhere)
44. Methylene chloride (dichloromethane)
45. Methyl chloride (chloromethane)
46. Methyl bromide (bromomethane)
47. Bromoform (tribromomethane)
48. Dichlorobromomethane
49. Trichlorofluoromethane
50. Dichlorodifluoromethane
51. Chlorodibromomethane
52. *Hexachlorobutadiene
53. *Hexachlorocyclopentadiene
54. *Isophorone
55. *Naphthalene
56. *Nitrobenzene
30
-------
*Nitrophenols (including 2,4-dinitrophenol and
and dinitrocresol)
57. 2-Nitrophenol
58. 4-Nitrophenol
59. 2r4-Dinitrophenol
60. 4,6-Dinitro-o-cresol
*Nitrosamines
61. N-nitrosodimethylamine
62. N-nitrosodiphenylamine
63. N-nitrosodi-n-propylamine
64. *Pentachlorophenol
65. *Phenol
*Phthalate esters
66. Bis(2-ethylhexyl) phthalate
67. Butyl benzyl phthalate
68. Di-n-butyl phthalate
69. Di-n-octyl phthalate
70. Diethyl phthalate
71. Dimethyl phthalate
*P_plynu clear aromatic hydrocarbons
72. Benzo(a)anthracene (1,2-benzanthracene)
73. Benzo (a) pyrene (3,4-bejizopyrene)
74. 3,4-Benzofluoranthene
75. Benzo(k)fluoranthane (11,12-benzofluoranthene)
76. Chrysene
77. Acenaphthylene
78. Anthracene
79. Benzo(ghi)perylene (1,12-benzoperylene)
80. Fluorene
81. Phenanthrene
82. Dibenzo (a,h)anthracene (I,2r5r6-dibenzanthracene)
83. Indeno (lr2,3-cd)pyrene (2,3,-o-phenylenepyrene)
84. Pyrene
85. *Tetrachloroethylene
86. *Toluene
87. *Trichloroethylene
88. *Vinyl chloride (chlorethylene)
31
-------
*Pesticides and metabolites
89. *Aldrin
90. *Dieldrin
91. *Chlordane (technical mixture & metabolites)
DDT and metabolites
92. 4,4'-DDT
93. 4f4'-DDE (pfp'-DDX)
94. 4,4'ODD (p,p'-TDE)
*Endosulfan and metabolites
95. A-endosulfan-Alpha
96. B-endosulfan-Beta
97. Endosulfan sulfate
*Endrin and metabolites
98. Endrin
99. Endrin aldehyde
*Heptachlor and metabolites
100. Heptachlor
101. Heptachlor epoxide
*Hexac'hlorocyclohexane (all isomers)
102. A-BHC-Alpha
10'3. B-BHC-Beta
104. R-BHC (lindane)-Gamma
105. G-BHC-Delta
*Polychlorinated biphenyls (PCB's)
106. PCB-1242 (Arochlor 1242)
107. PCB-1254 (Arochlor 1254)
32
-------
108. PCB-1221 (Arochlor 1221)
109. PCB-1232 (Arochlor 1232) '
110. PCB-1248 {Arochlor 1248)
111. PCB-1260 (Arochlor 1260)
112. PCB-1016 (Arochlor 1016)
113. *Toxaphene
114. *Antimony (Total)
115. *Arsenic (Total)
116. *Asbestos (Fibrous)
117. *Beryllium (Total)
118. *Cadmium (Total)
119. *Chromium (Total)
120. *Copper (Total)
121. *Cyanide (Total)
122. *Lead (Total)
123. *Mercury (Total)
124. *Nickel (Total)
125. *Selenium (Total)
126. *Silver (Total)
127. *Thallium (total)
128. *Zinc (Total)
129. **2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
*Specific compounds and chemicals classes as listed in the
Consent Decree.
**This compound was specifically listed in the Consent Decree.
Because of the extreme toxicity of TCDD, the Agency recommended
that laboratories not acquire analytical standards for this
compound. Categories and specified the scope of application of
effluent limitations, new source performance standards, and
pretreatment standards within each category in terms of the
Standard Industrial Classification (SIC) code numbers. For the
Inorganic Chemicals Manufacturing Point Source Category, the
major industries included are:
SIC 2812 - Alkalies and Chlorine
SIC 2813 - Industrial Gases
SIC 2816 - Inorganic Pigments
SIC 2819 - Industrial Inorganic Chemicals,
Not Elsewhere Classified
Within these industries, the Agency has identified 63
subcategories listed in Table 3-2 for the initial study of the
toxic pollutant problem. Most of these subcategories, 49 in
all, had already been covered by BPT and BAT discharge
regulations promulgated in 1974 and 1975. Those regulations
established point of discharge control levels for the
conventional parameters such as pH, TSS, TOC, BOD, and oil and
grease. In many cases, specific chemical parameters were
regulated, particularly Arsenic, Chromium, Copper, Mercury,
33
-------
TABLE 3-2. SCOPE OF INDUSTRY COVERAGE WITHIN THE INORGANIC
CHEMICALS MANUFACTURING POINT SOURCE CATEGORY
Subcategories Designated for Initial Study
1. Chlor-Alkali 33.
2. Hydrofluoric Acid 34.
3. Hydrogen Peroxide 35.
4. Titanium Dioxide 36.
5. Aluminum Fluoride 37.
6. Chrome Pigments 38.
7. Hydrogen Cyanide 39.
8, Sodium Bichromate 40.
9. Carbon Dioxide 41.
10. Carbon Monoxide/Hydrogen 42.
11. Copper Sulfate 43.
12. Nickel Sulfate 44.
13. Silver Nitrate 45.
14. Sodium Bisulfite 46.
15. Sodium Hydrosulfite 47.
16. Hydrochloric Acid 48.
17. Nitric Acid 49.
18. Sodium Carbonate 50.
19. Sodium Metal 51.
20. Sodium Silicate 52.
21. Sulfuric Acid 53.
22. Ammonium Chloride 54.
23. Ammonium Hydroxide 55.
24. Barium Carbonate 56.
25. Boric Acid 57.
26. Calcium Carbonate 58.
27. Copper Oxide 59.
28. Manganese Sulfate 60.
•29. Strong Nitric Acid 61.
30. Oxygen and Nitrogen 62.
31. Potassium Iodide 63.
32. Sodium Hydrosulfide
Sodium Silicofluoride
Sodium Thiosulfate
Sulfur Dioxide
Bromine
Calcium Hydroxide
Chromic Acid
Fluorine
Hydrogen
Iodine
Potassium Chloride
Stannic Oxide
Zinc Sulfate
Calcium Carbide
Calcium Oxide
Potassium Metal
Potassium Sulfate
Sodium Bicarbonate
Borax
Ferric Chloride
Lead Monoxide
Sodium Fluoride
Aluminum Chloride
Aluminum Sulfate
Potassium Dichromate
Calcium Chloride
Sodium Chloride'
Sodium Sulfite
Pdtassium Permanganate
Zinc Oxide
Lithium Carbonate
Ferrous Sulfate
34
-------
Nickelr Lead, Selenium, Zinc, and Cyanide, which are now
included in the list of toxic pollutants. Other regulated
parameters such as Al, Bar Fe, ammonia, fluoride and sulfide are
not presently listed as toxic chemicals but are to be treated as
nonconventional pollutants under future discharge limitations
and standards of performance.1
Nearly half of the initial 63 subcategories have been
recommended for exclusion from this study on the basis of
specific provisions for such exclusion under Paragraph 8 of the
Settlement Agreement. The bases for these exclusions are as
follows:
No. 63, Ferrous Sulfate, is already covered by the Titanium
Dioxide - Sulfate Process subcategory and does not require
separate consideration.
No's. 60, 61, and 62 (Potassium Permanganate, Zinc Oxide,
and Lithium Carbonate) have only one plant each (or one
plant with a wet process discharge), and represent
nonsignificant discharges of toxic pollutants. No's. 27
and 28 (Copper Oxide and Manganese Sulfate) are also single
plants, but were covered in screening.
No's. 36 through 59 have existing BPT or BAT regulations
requiring zero discharge of process waste water to
navigable water and there are no known discharges to a
POTW. Continued enforcement of the existing regulations
will provide adequate control of toxic pollutants.
The remaining 35 nonexcluded subcategories (Table 3-2,
No's. 1 through 35) are covered in this report. This group also
includes the 11 subcategories whose final regulations were
remanded for restudy in E.I. duPont de 'Nemours and Company, et
al. v. Train, supra, and the four additional subcategories whose
interim, final or proposed regulations were revoked and reserved
by the Agency.
It was anticipated by the Agency that a substantial number
of the 35 industries to be screened would also qualify for
exclusion under Paragraph 8 on the basis of the analytical
results obtained from the process waste water toxic pollutant
screening program. A preliminary prioritization indicated that
the initial detailed study and regulation development would
focus on the first 15 subcategories.
This judgment has been substantially supported by the
analytical results of the screening programs and a number of
additional exclusions are being recommended for subcategories
in which nonsignificant •toxic pollutant discharges have
been determined. A detailed presentation of the analytical
35
-------
results is given under the individual subcategory sections of
this report. The additional recommended exclusions include the
following:
No.
Subcategory
1. Hydrogen Peroxide
2. Carbon Dioxide
3. Carbon Monoxide/Hydrogen
4. Hydrochloric Acid
17. Nitric Acid
18. Sodium Carbonate
19. Sodium Metal
21. Sulfuric Acid
22. Ammonium Chloride
23. ' Ammonium Hydroxide
24. Barium Carbonate
25. Boric Acid
26. Calcium Carbonate
27. Copper Oxide (one plant)
28. Manganese Sulfate (one plant)
29. Strong Nitric Acid
30. Oxygen and Nitrogen
31. Potassium Iodide
32. Sodium Hydrosulfide
34. Sodium Thiosulfate
35. Sulfur Dioxide
Silver Nitrate, No. 13, and Sodium Silicofluoride, No. 33, -are
being deferred for future study under Phase II of the BAT
regulation development program for Inorganic Chemicals. This
deferrment was caused by problems with plant access during the
course of the present study.
3.2 GENERAL APPROACH AND METHODOLOGY
Initiating and undertaking a comprehensive study of the
toxic pollutant problem in the Inorganic Chemicals Industry was
preceded by an intensive evaluation by the Agency of the kinds
of data and supporting information that should be assembled as a
basis for the development of regulations. All major decisions
on the identity of pollutants and the establishment of effluent
limitations and standards of performance for each subcategory
had to be supportable by documented evidence collected from
operating production facilities. Similarly, the necessary
information on production rates, processes, raw materials, water
use, waste sources, and treatment technologies in practice had
to be acquired with sufficient detail and breadth of coverage to
permit an analysis of the engineering and economic variables
that are characteristic of each subcategory. Toxic pollutant
36
-------
control regulations would be based on the application-of best
available technology for treatment and reliable performance
evaluations for the removal of specific waste substances.
The following paragraphs briefly.describe the major study
tasks and their results as they are presented in this -report.
3.2.1 Industry Data Base Development and Subcategorization
Review
Information from individual manufacturers and previous
study documents were reviewed in detail and an evaluation of the
appropriateness of subcategorization was performed. Section 4
presents a discussion of the factors considered in
subcategorization and presents the rationale for maintaining the
present scheme of subcategorization for the industries studied.
3.2.2 The Screening and Verification Sampling Programs
*
The collection of detailed analytical data on conventional,
nonconventional and toxic pollutant concentrations in raw and
treated process waste streams was completed in a two-phase
sampling program. The first phase, screening, was designed to
provide a representative, one-time 72-hour sampling of a plant
in each subcategory in order to determine the presence of toxic
pollutantp and to evaluate their potential environmental
significance. The sampling and analytical methodology is
described in Section 5, along with the basis for making a
decision on the need for verification sampling in each
subcategory.
3.2.3 Engineering Evaluations
Section 6 describes the procedures and sources used in
developing the industry productions and waste water generation
characteristics that form the basis of the model plant concept.
The sources of detailed process and waste treatment information
are also presented. Section 7 contains an evaluation of
treatment technology presently applied in BPT systems and
advanced technologies that may be recommended for BAT and NSPS
applications. Section 8 provides estimates of the treatability
of selected toxic and nonconventional pollutants to be applied
in the development of achievable performance characteristics for
specific technologies. Section 8 also presents a discussion of,
the approach taken in the statistical analysis of long-term
monitoring data. The statistically derived parameters,
including variability factors for 24-hour maxima and maximum 30-
day averages are presented in Appendix A. Section 9 lays the
groundwork for the estimation of pollutant removal performances
for each nonexcluded subcategory. The candidate toxic
pollutants to be controlled in each subcategory are identified
37
-------
on the basis of the screening and verification data and the
rationale for the application of advanced level technologies is
presented.
3.2.4 Treatment System Cost Estimates
Section 10 presents the general approach to cost
estimating, discusses the assumptions made, and gives the
detailed cost estimates for alternative levels of treatment and
control. For each subcategory verified, the total estimated
installed cost of a typical BPT treatment system is developed on
the basis of the model plant design specifications and estimated
incremental costs are given for each of the advanced level
treatment alternatives.
3.3 GENERAL CRITERIA FOR EFFLUENT LIMITATIONS
»
3.3.1 BPT Effluent Limitations
The factors considered in defining best practicable control
technology currently available (BPT) include the total cost of
applying such technology in relation to the effluent reductions
derived from such application, the age of equipment and
facilities involved, the process employed, non-water quality
environmental impacts (including energy requirements), and
other factors the Administrator considers appropriate (Section
304(b)(1)(B)). In general, the BPT technology level represents
the average of the best existing performances of plants of
various ages, sizes, processes, or other common characteristics.
Where existing performance is uniformly inadequate, BPT may be
transferred from a different subcategory or category. BPT
focuses on end-of-pipe treatment rather than process changes or
internal controls, except where such are common industry
practice. The cost/benefit inquiry for BPT is a limited
balancing, committed to EPA's discretion, which does not require
the Agency to quantify benefits in monetary terms. See, e.g.,
American Iron and Steel Institute v. EPA, 526 F.2d 1027 (3rd
Cir. 1975). -In balancing costs in relation to effluent
reduction benefits, EPA considers the volume and nature of
existing discharges, the volume and nature of discharges
expected after application of BPT, the general environmental
effects of the pollutants, and the cost and economic impacts of
the required pollution control level. The Act does not require
or permit consideration of water quality problems attributable
to particular point sources or industries, or water quality
improvements in particular water bodies. Therefore, EPA has not
considered these factors. See Weyerhaeuser Company v. Costie,
590 F.2d 1011 (B.C. Cir. 1978).
38
-------
3.3.2 BAT Effluent Limitations -
The factors considered in assessing best available
technology economically achievable (BAT) include the age^ of
equipment and facilities involved, the process employed, process
changes, non-water quality environmental impacts (including
energy requirements), (Section 304(b) (2) (B)) . At a minimum, the
BAT technology level represents the best economically achievable
performance of plants of various ages, sizes, processes, or
other shared characteristics. As with BPT, uniformly inadequate
performance may require' transfer of BAT from a different
subcategory or category. BAT may include process changes or
internal controls, even when these technologies are not common
industry practice. The statutory assessment of BAT "considers"
costs, but does not require a balancing of costs against
effluent reduction benefits (see Weyerhaeuser v. Costle, supra).
In developing the proposed BAT, however, EPA has given
substantial weight to the reasonableness of costs. The Agency
has considered the volume and nature of discharges, the volume
and nature of discharges expected after application of BAT, the
general environmental effects of the pollutants, and the costs
and economic impacts of the required pollution control levels.
Despite this expanded consideration of costs, the primary
determinant of BAT is effluent reduction capability. As a
result of the Clean Water Act of 1977, 33 OSC 1251 et seq., the
achievement of BAT has become the principal national means of
controlling water pollution due to toxic pollutants.
3.3.3 BCT Effluent Limitations
The 1977 amendments added Section 301(b)(2)(E) to the Act,
establishing "best conventional pollutant control technology"
(BCT) for discharges of conventional pollutants from existing
industrial point sources. Conventional pollutants are those
defined in Section 304(b)(4) - BOD, TSS, fecal coliform, and pH
and oil and grease, designated by the Administrator as
"conventional" on July 30, 1979, 44 PR 44501. BCT is not an
additional limitation, but replaces BAT for the control of
conventional pollutants. BCT requires that limitations for
conventional pollutants be assessed in light of a new "cost
reasonableness" test, which involves a comparison of the cost
and level of reduction of conventional pollutants from the
discharge of publicly owned treatment works to the cost and
level of reduction of such pollutants from a class or category
of industrial sources. In its review of BAT for industries not
covered by the NRDC Consent Decree, the Agency promulgated BCT
levels based on a methodology described at 44 FR 50732 (August
26, 1979). This methodology compares subcategory removal costs
(dollars per pound of pollutant, measuring from BPT to BAT) with
costs experienced by POTWs. EPA applied this methodology to the
costs of removal of conventional pollutants in the 11
39
-------
subcategories of the inorganic chemicals manufacturing industry
affected by these proposed regulations. Models were chosen to
represent the average size plant in each subcategory. The total
annualized cost of each control technology and the total pounds
per year of TSS removed were then computed for each of these
model plants. The Agency is proposing, based on this analysis,
that BCT should be equal to BPT except in the case of the
diaphragm cell portion of the Chlor-Alkali subcategory. In this
subcategory, EPA is proposing limitations based on dual media
filtration. The costs of this technology in the diaphragm cell
portion of the Chlor-Alkali subcategory are explained below in
the discussion of the treatment options for that subcategory.
For all other subcategories, EPA is proposing that BCT equal BPT
either because additional removal failed the cost test or
because EPA is proposing BAT equal to BPT.
For Aluminum Fluoride and Sodium Dichromate, the cost in
dollars per pound for removal of additional conventional
pollutants is $2.06 and $13.40, respectively. For Chlor-Alkali -
mercury cell process, Hydrogen Cyanide, and Sodium Hydrosulfite,
BCT is being proposed equal to BPT because the technology added
for BAT does not impact the removal of conventional pollutants.
EPA is proposing a distinct BCT standard for the Chlor-Alkali -
diaphragm cell process and Hydrofluoric Acid as discussed below.
In the remaining subcategories, BPT is proposed equal to BAT
which automatically makes BCT equal to BPT.
In the Chlor-Alkali - mercury cell process segment and the
Hydrogen Cyanide and Sodium Hydrosulfite Subcategories, BCT is
being proposed equal to BPT because the technology added for BAT
does not impact the removal of conventional pollutants.
In the Aluminum Fluoride subcategory, the cost for removal
of additional conventional pollutants is $2.06 per pound. Thus,
BCT is proposed equal to BPT because the cost is greater than
the $1.15 per pound cost for removal of conventional pollutants
from a publically owned treatment works (POTW). The calculation
is as follows:
$1.77 (kg/2.2 Ib.) = $2.06 per pound
(1.2 kg/kkg - 0.81 kg/kkg) of TSS removed
Where $1.77 is the' increased cost for BAT treatment over
BPT treatment cost in dollars per kkg of production from Table
15-9, 1.2 kg/kkg is proposed for the BPT suspended solids
•limitation from Table 15-19, 0.81 kg/kkg is achievable by use of
BAT technology applied to suspended solids removal developed for
Table 15-23. A conversion factor of 2.2 pounds per kilogram is
used.
40
-------
In the Sodium Bichromate Subcategory, the cost for removal
of additional conventional pollutants is $13.40 per pound.
Thusf BCT is proposed equal to BPT because the cost is greater
then the $1.15 per pound cost for removal of conventional
pollutants from a publically owned treatment works. The
calculation is as follows:
$1.18 (kg/2.2 Ib.) = $13.40 per pound
(0.22 kg/kkg - 0.18 kg/kkg) of TSS removed
Where $1.18 is the increased cost for BAT treatment over
BPT treatment cost in dollars per kkg of production from Table
18-12, 0.22 kg/kkg is proposed for the BPT suspended solids
limitation from Table 18-15, and 0.18 kg/kkg is achievable by
use of BAT technology applied to suspended solids removal for
Table 18-16.
In the Hydrofluoric Acid Subcategory, the cost for removal
of additional conventional pollutants is $0.37 per pound. Thus,
a BCT regulation is established because the cost for removal of
additional conventional pollutants is less than the $1.15 per
pound cost for removal of conventional pollutants from a
publically owned treatment works. The calculation is as
follows:
$2.42 (kg/2.2 Ib.) = $0.37 per pound
(5.3 kg/kkg - 2.3 kg/kkg) of TSS removed
Where $2.42 is the increased cost for BAT treatment over
BPT treatment cost in dollars per kkg of production from Table
12-15, where 5.3 kg/kkg is proposed for the BPT total suspended
solids limitation from Table 12-24, where 2.3 kg/kkg is
reduction of TSS achievable by application of filtration to the
waste waters. Because additional removal of conventional
pollutants passes the cost test, the regulation for BCT for TSS
is set at 2.3 kg/kkg as a 30-day maximum average and using a
variability factor ratio (VFR) of 2.1 to establish a daily
maximum of 4.8 kg/kkg.
In the diaphragm cell segment of the Chlor-Alkali
Subcategory, the cost for removal of additional conventional
pollutants is $1.09 per pound. This is less than the $1.15 per
pound cost of conventional pollutant removal in a publically
owned treatment works. This determination was made by
estimating the BAT cost (Table 11-33) of a 30 percent reduction
in the BPT maximum 30-day average TSS effluent loading (Table
11-36) as follows:
($0.36/kkg) (kg/2.2 Ib.) = $1.09/
0.51 kg/kkg - (1.00 - 0.30) (0.51 kg/kkg) Ib.
41
-------
Thus, on the basis of adding dual media filtration to the
BPT treatment, the Agency is proposing a BCT regulation of TSS.
The proposed maximum 30-day average effluent limitation is:
(1.00 - 0.30) (0.51 kg/kkg) = 0.36 kg/kkg
and the proposed daily maximum is obtained by applying the
variability factor ratio (VFR) value of 2.0 as follows:
(2.0)(0.36 kg/kkg) = 0.72 kg/kkg
In the remaining subcategories, BPT is proposed equal to
BAT which automatically makes BCT equal to BPT.
3«3-4 New Source Performance Standards
The basis for new source performance standards (NSPS) under
Section 306 of the Act is the best available demonstrated
technology. New plants have the opportunity to design the best
and most efficient inorganic chemicals manufacturing processes
and waste water treatment technologies, and Congress therefore
directed EPA to consider the best demonstrated process changes,
in-plant controls, and end-of-pipe treatment technologies which
reduce pollution to the maximum extent feasible.
3.3.5 Pretreatment Standards fear Existing Sources
Section 307(b) of the Act requires EPA to promulgate
pretreatment standards for existing sources (PSES) which must be
achieved within three years of promulgation. PSES are designed
to prevent the discharge of pollutants which pass through,
interfere with, or are otherwise incompatible with the operation
of POTWs. The Clean Water Act of 1977 adds a new dimension by
requiring pretreatment for pollutants, such as heavy metals,
that limit POTW sludge management alternatives, including the
beneficial use of sludges on agricultural lands. The
legislative history of the 1977 Act indicates that pretreatment
standards are to be technology-based, analogous to the best
available technology for removal of toxic pollutants. The
general pretreatment regulations which served as the framework
for these proposed pretreatraent regulations can be found at 40
CWR Part 403, 43 PR 27736 (June 26, 1978). In some instances
PSES regulations have been established for subcategories not
presently discharging to a POTW. This establishes regulation
for plants that may choose to change their discharge to a POTW.
3.3.6 Pretreatment Standards for New Sources
Section 307 (c) of the Act requires EPA to promulgate
pretreatment standards for new sources (PSNS) at the same time
that it promulgates NSPS. New indirect dischargers, like new
42
-------
direct dischargers, have the opportunity to incorporate the best
available demonstrated technologies including process changes,
in-plant controls, and end-of-pipe treatment technologies, and
to use plant site selection to ensure adequate treatment system
installation.
43
-------
-------
SECTION 4
SUBCATEGORIZATION REVIEW
4.1 BASIS FOR SDBCATEGORIZATION
4.1.1 Factors Considered
The inorganic chemicals industry is very large and
diversified and has been segmented into subcategories for
the purpose of establishing effluent guidelines. Factors
taken into consideration for subcategorization includes
raw materials used, product produced, manufacturing process
employed, geographical location, size and age of equipment
and facility involved, nonwater quality aspects of waste
characteristics, water pollution control technology,
treatment costs, energy requirements and solid waste
disposal. Following is a discussion of each of the general
factors considered for this industry.
Raw Materials
Different raw materials are used to manufacture a wide
variety of products, and vary from raw brines and ores to pure
reagent chemicals. Some proceses use waste or by-product
streams from other plants or from other processes within the
same plant. *
Because of this diversification, raw material
characteristics generally do not constitute a logical basis
for subcategorization. Variations in raw material quality or
purity are not normally sufficient to cause a great difference
in waste water treatment needs, except in the case "of
trace toxic materials which may occur in some sources but not
in others.
Dominant Product
Subcategorization by chemical name of the dominant
inorganic chemical produced involves the least ambiguity
in applying standards to a given point source. This is
critical because of the great variety of product mix,
45
-------
manufacturing processes, waste water constituents, and
other factors at existing plants. Subcategorization by
product becomes less useful as product mix increases in
complexity because multi-product waste water also becomes
more complex and less susceptible to simple uniform
treatment.
h subcategory established on the basis of product
manufactured might have two or more different processes but,
in the majority of cases, the characteristic of the waste
waters is similar and the same treatment technology can be
applied for different process waste waters. If two or
more dissimilar processes produce waste water of different
quality, and different treatment technologies have to be used,
then the subcategory has to be further classified or
segmented, for example, the Chlor-alkali Industry.
Manufacturing Process
Typically, inorganic chemicals are manufactured for
captive or merchant use in four or more steps starting from raw
material to final product. Two or more different products
might use the same process but then the raw materials used,
process sequence, control, recycle potential, handling, and
quality control will vary, producing wastes of
different quality. Primary subcategorization, therefore,
by process is unlikely to be useful. However, secondary
subcategorization by process has been necessary in some cases.
Geographical Location
Inorganic chemical plants exist in all parts of the
United States but subcategorization on this basis is not
appropriate. Geographical location is important in analyzing
the feasibility of various treatment alternatives.
Evaporation ponds are functional only in areas. where
evaporation exceeds rainfall. Ocean dumping and deep well
disposal are possible only *in certain areas, and must be
consistent with local, State and Federal laws.
Theppossibility of ground water contamination may preclude the
use of unlined holding and settling ponds in many locations.
In the northern regions, climatic conditions may
necessitate the inclusion of special provisions to prevent
freezing of treatment system components, particularly
biological oxidation units, clarifiers, ponds, and open
collection systems. The costs of utilizing waste heat
sources from the process or providing various types of
thermal protection, such as insulation or burial of pipes and
tanks and .building structural shelters, may add considerably
to the capital and O & M cost associated with a treatment
technology.
46
-------
Thus, the influence of geography, climate, geology,
etc. is reflected in waste treatment modifications
and is primarily manifested in the cost of treatment. This,
of itself, is not a good basis for subcategorization.
Plant Size
Plant size and production capacity were not found to
affect the characteristics of the waste produced. Although
plant size can affect treatment cost, this variability can
be expressed graphically or mathematically without the need
for further segmentation of the category.
Plant Age
Plant age can have an important bearing on waste water
volume and quality and is, therefore, a significant factor to
consider in evaluating the applicability of treatment
technologies and assessing the relative costs of treatment
for plants of widely differing age producing the same or
similar products. A particular problem with older plants is
that their present patterns of water use may have evolved over
a long period of time with little consideration for the
principles of efficient waste segregation, collection, and
treatment. To a limited degree, plant modernization can
correct or at least mitigate some of these shortcomings in
older facilities, however, only a small proportion of the
cost of revamping collection systems or of converting from
contact to noncontact cooling systems can be offset by the
resulting lower cost of treatment. In general, older plants,
even after considerable modernization, normally have a
higher volume of waste water flow and higher waste
loadings (although pollutant concentrations may be lower
due to poor segregation from noncontact sources) in
comparison to relatively new plants. The present and
forthcoming requirements for pollution control may impose a
severe treatment cost penalty on older plants due to the need
for backfitting and replumbing of outdated collection systems.
Land'' availability and land use restrictions are also
factors which may translate into higher treatment costs for
older facilities which find themselves surrounded by highly
developed industrial and residential areas.
Unfortunately, plant age does not readily lend itself
to an unambiguous definition where a series of
plant modifications has taken place. The extent of
modifications also varies greatly among plants within the
same product industry. For those plants that have been
enlarged or modified from their original status, plant age is
not unambiguously calculable and therefore not a reasonable
basis for subcategorization.
47
-------
Nonwater Quality Characteristics
Airborne emissions from manufacturing operations can be
kept within air quality control limits through the use of
cyclones, wet scrubbers and other methods. The nature of the
air pollution is related to the products(s) manufactured
and/or the raw material used. Since both of these elements
vary widely within the inorganic chemicals industry, there
is no logic in subcategorization on the basis of
nonwater quality characteristics.
Treatment Cost
From a technical viewpoint, subcategorization by common
technological requirements for treatment processes could
provide a logical basis for selecting one or more unit
processes to accomplish the same treatment "function,
regardless of the source of the waste water. For example,
residuals of dissolved heavy metals will respond to lime
precipitation and sedimentation at high pH without respect
to the specific origin of the metals. This "building block"
concept could conceivably result in selecting various
combinations of unit processes to meet the
treatment requirements. However, if the treatment cost must
be expressed in terms of dollars per unit production, this
method of subcategorization crosses product lines and
interferes with comparison of treatment costs based on the
production of a specific chemical. Even if the unit^
operation is commonly applicable for treating waste flows
of different products, the cost of treatment will fluctuate
because of variations in quality, loading and flow rates
and subcategorization on the basis of treatment cost is not
recommended.
Energy Cost
Manufacturing processes in the Inorganic Chemicals
Industry typically have large energy requirements. In
contrast, waste water treatment processes consume a small
fraction of the total energy used. There appears to be no
major energy requirements for the waste water treatment
facility and subcategorization on the basis of energy cost is
not justified.
Solid Waste
Not all inorganic manufacturing processes produce
solid wastes. Solid waste producers practice various disposal
methods, such as on-site landfills, contract 'hauling to
approved dump sites or incineration. Solid waste
disposal becomes, very site specific and exhibits a wide range
48
-------
of costs. Because of the lack of uniformity within the
industry, solid waste, generation and disposal practices are
not a staisfactory basis for subcategorization.
B
4.1.2 General Conclusions '
If effluent limitations are to be tied to units of
production, only one method of primary subcategorization is
broadly applicable to the inorganic chemicals point source
category; viz., subdivision by dominant product. However,
there are three subcategories, Chlor-Alkalir Titanium
Dioxide, and Hydrogen Cyanide which require further
subdivision based on the difference in the quantity and
quality of the waste water from the processes, and two others,
Hydrofluoric Acid and Aluminum Fluoride, have been reviewed
for possible integration (see Section 4.3).
4.2 SECONDARY SUBCATEGORIZATIOW
4.2.1 Chlor-Alkali
Mercury and .diaphragm cells are the two distinct types
of electrolytic cells that are used in the production of
chlorine and caustic soda. Major process differences
between mercury cell and diaphragm cell plants produce
corresponding differences in the volume and nature of waste
water generated. A principal difference is the presence pf
mercury as a contaminant in the waste waters from the mercury
cell process and asbestos in the diaphragm cell plant
wastes.. The TSS discharges from diaphragm cell plants are
generally larger^ than from mercury cell plants, due to the
higher volumes of contact and noncontact water used. Also, in
diaphragm cells a large amount of water is used and an
appreciable quantity of waste water is produced in the
caustic evaporation process. Such water is not produced in
mercury cell plants. The quantity of waste water generated
from the diaphragm cell plants is almost double that of the
mercury cell plants for the same chlorine production
capacity. Based on the quantity and characteristics of the
waste water, further subcategorization is justified.
4.2.2 Titanium Dioxide
Two major ores, rutile and ilmenite, are used for the
manufacture of titanium dioxide. The ilmenite ore contains 40-
70 percent titanium dioxide (Ti02), up to 35 percent ferrous
oxide (PeO), and 25 percent ferric oxide (Fe203). Rutile ore
contains more than 90 percent Ti02. Two processing techniques,
the sulfate process and the chloride process, are used to
extract titanium dioxide from the ores.
49
-------
The sulfate process uses ilmenite ore and sulfuric acid
as raw materials. The chloride process uses rutile ores
and chlorine. The high grade rutile ore is expensive and
its availability is declining. In recent years, new
technological advances have alleviated the raw material
shortage problem. By upgrading the ilmenite ore quality, the
chloride process can be used to produce titanium dioxide of
high purity. Because of the difference in quality and quantity
of waste waters generated from the sulfate and chloride
processes using the two different ores, the titanium dioxide
industry may be further subdivided into three segments as
follows:
a. Sulfate process
b. Chloride process using rutile ore
c. Chloride process using ilmenite ore (one step).
The sulfate process generates large amounts of strong and
weak sulfuric acid water-borne wastes. Application of
pollution control technology to the acid wastes generates about
five times as much gypsum as product. The chloride process
generates large amounts of dissolved metal chlorides and the
treatment technology is expensive. Solid waste from both
processes present difficult disposal problems. These solids
include ferrous sulfate (FeSO4) and a hydrated by-product
from the sulfate process and heavy metal sludges from the
chloride process. Ilmenite ore has to be upgraded before
it is used to extract titanium dioxide by the chloride
process, and this beneficiation process step generates
additional wastes.
The application of the chloride process to ilmenite ore
may proceed in either one or two steps. A patented one-step
process accomplishes both "beneficiation and chlorination of
the ore in a single fluidized bed reactor and generates
raw waste loadings which are similar to those from the
sulfate process in terms of acidity and metals, and similar
to wastes from the chloride-rutile process in terms of spent
coke solids and still residues. In the two-step process, ore
beneficiation resulting in either a synthetic rutile or an
enriched titanium oxide slag is carried out separately at
the mine or the plant. The discharge of waste water
generated by the beneficiation step would be regulated under
the Ore Mining and Dressing Point Source Category and will not
be considered in this document. The second step of the two-
step process generates wastes that are very similar in
quantity and quality to those from the chloride-rutile process
and will be governed by the discharge regulations for that
segment of the TiO2 subcategory.
50
-------
Therefore, further subclassification based on the amount
and characteristics of the waste water appears to be
justified, and the three process subdivisions indicated
above are appropriate for this purpose.
4.2.3 Hydrogen Cyanide
Hydrogen cyanide (HCN) is made from two different
processes, the Andrussow process and as a by-product of
acrylonitrile manufacture. In the Andrussow process, air,
ammonia, and natural gas are reacted to produce the dominant
product hydrogen cyanide.
Water-borne wastes from the process consist
" principally of ammonia and sulfates in addition to cyanide and
nitriles.
The primary product in the other process is
acrylonitrile (CH2 = CHCN) and the hydrogen cyanide is a by-
product. Because the hydrogen cyanide is a by-product it will
be covered in the organic chemicals manufacturing category
with the primary product.
4.3 REVIEW OP POSSIBLE INTEGRATION OP SOBCATEGORIES
4.3.1 Hydrofluoric Acid and Aluminum Fluoride
Aluminum fluoride (A1F3) usually is produced by the
reaction of hydrated alumina (A12O3.3H20) with hydrogen
fluoride (HF), although one plant produces aluminum fluoride
from fluorosilicic acid (H2SiF6), a by-product of phosphoric
acid (H3FO4). With one exception, all the aluminum
fluoride plants are integrated with hydrogen fluoride
(or hydrofluoric acid) production.
The two major uses of hydrogen fluoride are in
the fluorocarbon industry and as raw material in the
manufacture of aluminum fluoride. A ban on the fluorocarbon
propellants has curtailed the use of hydrogen fluoride in
that industry and it was completely stopped in 1978. The
selling of hydrogen fluoride in the merchant market has
declined and the primary use is limited to the
production of aluminum fluoride and fluorocarbon plastics
until some other major use is found.
For both products (HF and A1F3) , process waste
waters are generated by the various gas scrubbers and by leaks
and spills. In both cases, air pollution control scrubber
effluents contain mainly fluoride, acidity and sulfate. The
fluoride is present as the free ion as well as various complex
51
-------
fluoro anions. Calcium fluoride (CaP2), generated as a solid
waste, is a disposal problem for both the subcategories
because of its moderate toxicity. Only one additional solid
waste, gypsum (CaSO4.2H2O), is generated from the hydrogen
fluoride manufacture alone, and it can be treated and handled
independently.
Combining hydrofluoric acid and aluminum fluoride
into a single subcategory does not appear to offer any
regulatory advantages when the two products are manufactured
at the same plant location. The waste waters associated
with the two products are similar and a common treatment
facility is normally utilized. In addition, the combined
manufacture of these products does not create a unique or
unusual situation, either with regard to the waste water
treatment requirements or compliance with discharge
regulations. Although the waste gypsum produced at an HF
plant supplies enough calcium for adequate fluoride removal
from neutralized scrubber waste waters generated by both HF
and A1F3 production, the applied treatment technology is
essentially the . same as that applied by manufacturers
of either product alone. However, the effluent water
quality and the toxic pollutant loadings would not be
expected to be the same. Further, the opportunities for drip
acid recycle (or the hydrolysis of complex fluorides prior to
treatment) and scrubber water recycle are a function of plant
design and age, rather than product mix.
In view of these considerations, a recommendation
for the creation of an HF/A1F3 combined product subcategory is
not being made at this time,
4.4 SUMMARY
The recommended subcategorization with process
subdivisions include the following:
Subcategory Process Subdivisions
Chlor-Alkali Mercury Cell
Diaphragm Cell
Titanium Dioxide Sulfate
Chloride-Rutile
Chloride-Ilmenite
Hydrogen Cyanide Andrussow Process
Acrylonitrile By-Product
52
-------
SECTION 5
SCREENING AND VERIFICATION SAMPLING PROGRAMS
5.1 SCOPE AND METHODOLOGY
The specific objective of the sampling programs was to
establish the extent of the required regulation of toxic
pollutant discharges in the inorganic chemicals industry in
terms of factual information derived from the chemical analysis
and flow measurement of representative process raw waste water
streams and treated effluents. Prior to this study, most of the
information available on toxic pollutants has been concerned
with a relatively small number of known process-related
substances -contaminating a variety of direct and indirect
contact process waters discharged from a production facility.
There had been no previous requirement for a comprehensive
survey of waste water chemistry addressing the possibility that
a large number of other potentially toxic substances could be
present, albeit at extremely low concentrations.
The screening phase of the sampling program was designed to
ascertain the presence in each subcategory of any of the 129
listed toxic pollutants at raw waste * concentrations or daily
loadings which, if untreated, could be environmentally
significiant. Screening is based on the sampling of one or more
typical manufacturing operations in each subcategory. Where
significant pollutant concentrations were found, additional
plants were sampled during the verification phase for
confirmation and further quantification of data on the
particular toxic pollutants in question. A, goal was set for
screening and verification sampling of a sufficient number
of plants to account for at least 75 percent of the total-U.S.
production, in each .subcategory having significant
concentrations of priority pollutants.
A detailed description of the screening and verification
programs is presented in the'paragraphs below.
53
-------
5.1.1 Selecting giants and Making Preliminary Contacts
In each subcategory, plants were selected for screening on
the basis of the following general criteria:
1. Minimal product mix and no organic product lines which
could 'increase the potential for interprocess cross
contamination of waste waters.
2. Presence of a physical chemical treatment facility
rather than a biological one, or no treatment system.
(Biological systems are neither widely used nor
generally applicable in the inorganic chemicals
industries.)
3. Manufacture of industrial grade products in volume,
rather than low volume reagent grade products.
4. Median production capacity within the subcategory.
5. Segregated waste streams to facilitate sampling.
6. NPDES discharges rather than POTW discharges, since
treatment for a NPDES discharge is usually more
extensive.
7. Geographical clustering of selected plants to
facilitate field logistics, but only extent that other
factors are equal.
Preliminary phone contacts were made with plant
representatives of those facilities which satisfied the above
criteria. If requested,1 a letter was written to describe the
objectives of the sampling program and to cite the legal
authority of the Agency and its sampling contractor under
Section 308 of the Federal Water Pollution Control Act
Amendments of 1972. Secrecy agreements, when required, were
executed at this time for the protection of any company
proprietary information disclosed to the sampling contractor.
Prior to the actual sampling of waste streams, a lead visit
to the selected plant was made to gather background information,
confirm and update any 308 Questionaire responses, and to obtain
additional technical information regarding processes and waste
treatment practices. Sampling sites were selected and described
relative to a detailed waste source inventory and a flow
diagram of the process and waste treatment system. Arrangements
were made for the subsequent .sampling visit and the details of
the lead visit and sampling point descriptions were documented
in an interim report to the Agency.
54
-------
5.1.2 Screen!rig and Verification Sampling
Collection of Samples for Screening
In the screening phase of the sampling program, the
specific objective was the detection and quantification of
water-borne waste constituents included on the list of 129 toxic
pollutants (Table 3-1). Each sample of an individual raw waste
stream, a combined waste stream, or a treated effluent was
collected where possible by an automatic, time series,
compositor over a single 72-hour sampling period. Where
automatic compositing was not possible, grab samples were taken
at intervals during the same sampling period and composited
manually.
Each sample was divided into several portions and
preserved, as required for different types of analysis, in
accordance with the procedure established by EPA (4) for the
measurement of toxic pollutants.
" \
Samples were also taken from the composites, or as
individual grabs, for the analysis of the conventional and
nonconventional pollutants.
Collection of Samples for Verification
The objective of verification sampling was to confirm the
first observations from screening and further quantify the
concentrations and waste loadings of the toxic pollutants and
conventional and nonconventional pollutants. Where any toxic
pollutant metals were found during screening sampling of a
particular plant^, analyses were made for all toxic pollutant
metals during the verification sampling.
The established protocol for verification sampling
required the collection of three 24-hour composites at each
sampling point. Again, where composites could not be taken with
automatic samplers, grab samples were taken periodically over
the same time period and composited manually.
Sample Shipping
All samples, individually labeled, were placed in large
plastic bags, which were then placed in a waterproof insulated
shipping container. - Enough ice was included to maintain a
temperature of approximately 4 degrees C. during shipment to the
laboratory.
Containers were shipped by the best available route,
usually air freight, usually arriving at the laboratory on the
same day,* but occasionally taking overnight. Upon receipt, all
samples were immediately placed in a walk-in refrigerator
maintained at 4 degrees C.
55
-------
In order to maintain the chain of custody and to keep track
of samples, sampling personnel kept logs of samples taken in ink
in page numbered hard-bound books. The data recorded included:
date, time, plant code, number, sample type, and sampler. This
information was also included on the label of individual
samples. Prior to their arrival at the laboratory, a list of
samples shipped, including number, type of samples, and analysis
to be performed, was sent to each department supervisor to alert
him of incoming work.
A master analytical control chart was maintained which
included: date sample was received, date due, number and type of
each sample, and the analysis required.
At the time of analysis, the individual samples were
distributed to the analytical chemists along with a list which
included: I.D. number of sample, type of sample, analysis
required, date samples received, and due dates.
Upon completion of analysis, the sample was sent back to
the refrigerator and placed in identified bins. All samples
were kept in the refrigerator at 4 degrees C. when not being
analyzed. A list of completed samples was then sent to the EPA
Sample Control Center.
Verification Sampling Plant Selection
After the decision was made to verify the presence of toxic
pollutants found in the screening of a subcategory, verification
plants were selected. The basis for selection was essentially
the same as that used in selecting screening plants.
The screening program results were evaluated to identify
those toxic pollutants that were present at significant
concentration or significant daily loadings. Concentrations or
loadings which could be reduaed by the highest quality treatment
systems were considered significant. Two situations occurred:
1. A subcategory which had a significant raw waste
concentration of any toxic pollutant would be subject to
verification sampling, and BAT-based regulations would likely be
proposed by' the Agency for the treatment and control of that
toxic pollutant,
2. A subcategory which had no significant raw waste
concentration of any toxic pollutant would not be subject to
verification sampling and would likely be excluded from
regulatory coverage at this time in accordance with the
provisions for exclusion under Paragraph 8 of the Settlement
Agreement.
56
-------
In analyzing screening data, only those pollutants
attributable to process sources were considered. Pollutants
which result from cooling tower operations, corrosion or
corrosion control, control of biological growth, or any other
operation not directly tied to the production process were not
used as a basis for verification.
The number of plants selected for verification in each
subcategory was roughly proportional to the number of existing
plants in that subcategory with a maximum of five plants
selected. In small subcategories (relatively few production
facilities), an effort was made to select a sufficient number of
plants to account for the majority of the total U.S.
Production.
When the verification phase of the program was initiated,
an important decision was made with regard to metals analysis.
First, in view of the frequent presence of metal contamination
in the wastes screened, and the inability in some cases to show
a direct relationship between certain metals found and the
known process chemicals or the materials of construction, it
was decided that all 13 of the toxic metals should be determined
again during verification, regardless of whether they were found
in screening. This was intended to provide a much more complete
data base than would be obtained by running verification
analyses for only those metals found in screening to exceed the
verification criteria levels at the time of sampling.
5.1.3 Analytical Methodology for Toxic Pollutants
The analytical protocol for the screening and verification
of toxic pollutants was established in Sampling and Analysis
Procedures for Priority Pollutants by U.S. Environmental
Protection Agency, Environmental Monitoring and Support
Laboratory, Cincinnati, Ohio, April 1977.
The specified analytical methodologies were employed
without modification except where noted below in connection with
toxic metals analysis during verification.
Implementation of the methodology and quality assurance
provisions required the establishment of special sample handling
and control procedures specifically suited to each type of
analysis. These procedures, together with a discussion of the
achievable detection limits for each parameter or group of
similar parameters are presented in the following paragraphs.
57
-------
Trace Metal Analysis
Figure 5-1 shows a data flow diagram for metals analysis.
Atomic absorption methods described in 40 CFR 136 per Section
304 (h) were used. A set procedure was followed in the
laboratory to generate the analytical values and the quality
control data. The data flow diagram shows the actual sequence
employed in verification analysis and the following notes, which
are keyed to the diagram, provide additional information on the
procedures:
1. Blanks — two for each set,of analyses digested.
Duplicates — one every seventh sample.
2. Quality Control at Operator Level (Atomic Absorption):
Blanks —
Standards —
Spikes —
Duplicates —
These were run at the beginning and the
end of every set analyzed for each
metal. Also, air blanks were run on
furnace, or heated graphite atomizer,
(EGA) , after any sample with a large
positive value.
Three different concentrations were run
at the beginning and end of every set
analyzed for each metal. Standards
were also run every tenth sample during
the analysis of a set.
These were run every seventh sample,
and were made by taking a mixture of
equal parts of a sample and standard
and comparing the resulting absorbance
with individual sample and standard
absorbances.
For furnace analysis, the sample was
run twice when the absorbance was low to
identify errors. The average of the two
values was used as the determinate
value.
UTD = "Unable
interferences.
To
Determine"
due
to
matrix
Criteria Employed in Spike Selection:
a. Samples were chosen to be spiked based upon the
following criteria;
— those which were not subject to interference
effects.
58
-------
FIELD
SAMPLING
PRESERVATIVE ADDED
ICED , AND AIR SNIPED
RECEIPT . LOS IN SAMPLES
AND REFRIGERATE
U1
STANDARDS
PREPARED
UTD
O)
DETERMINATE
VALUE
t
FLAME ANALYSIS
QUALITY CONTROL BLANKS AND
DUPLICATES CR1AT1O (1)
ATOMIC ADSORPTION ANALYSIS (2)
I
NACE
b,Sb,T1 )
FLAME
- t Aa3»,Cr,Cu,NI,Zn !
~»
*•
VAPOR OENERTION
(Ho)
HYDRIDE GENERATION
( Aa,S» )
OFF-SCALE
RESULT
DETERMINATE
VALUi
Figure 5-1. Sample flow sheet for metal analysis.
-------
— those that had a measurable concentration of
the metal being determined.
— those whose concentration was in the linear
range of the instrument.
— approximately every seventh sample.
b. The level of spike chosen was controlled by the
following factors:
— it should be approximately 40-60 percent of
the determinate value.
— the determinate value absorbance 4- spike
absorbance must give total absorbance" that
was within the linear range.
c. A reagent blank was run with each set of spiked samples
prepared.
During the screening phase of the sampling program, the
standard protocol followed for metals analysis was:
1. Twelve elements were determined by AA
spectrophotometry in the furnace (HGA) mode.
2. If subject to matrix interference (UTD), they were then
determined in the flame mode.
3. Mercury was determined by the standard cold vapor
method.
Certain changes in analytical protocol were instituted
during verification analysis in order to avoid the excessive
matrix interference experienced during screening when the heated
graphite atomizer (HGA) was the primary method applied to the
analysis of 12 of the metals. The modified protocol for metals
was:
1. Six elements were determined by flame only, namely, Ag,
Be, Cu, Cr, Ni and Zn.
2. Pour elements were determined by furnace (HGA), namely,
Cd, Pb, Tl and Sb. If interference occurred, Cd, Pb,
Tl and Sb were determined by flame.
3. Hg was still analyzed by the cold vapor method.
This modification reduced the number of preparations per
sample from three to two and^achieved adequate detection limits
which were still well below the verification criteria levels.
60
-------
Additional modifications were made during the verification
program to improve the reproducibility and precision for Hg, As
and Se. These wer,e:
1. The cold vapor procedure for Hg was modified to
eliminate the pump and allow dilution and rerun from
the same sample. This saved time and increased
reproducibility.
2. Selenium and arsenic were determined by hydride
generation using sodium borohydride (NaBH4). This
greatly minimized problems associated with matrix
interference. The method is very reproducible and the
detection limits were at levels well below the
verification criteria for thes_e two elements.
After the above modifications were adopted, screening
samples which originally were unable to be analyzed, or which
were recorded as below excessively high detection limits due to
the effects of matrix interferences, were rerun. Satisfactory
results were then obtained in nearly all cases due to the
greatly improved sensitivity and reproducibility.
It should be noted that these modifications of the
analytical protocol were in the direction of improved precision
and reproducibility'and not towards lower detection limits. The
original screening procedures generally had a lower detection
limit when it was achievable. However, the methods were too
susceptible to giving no r.esult at all with complex industrial
matrices, and so the revised protocols sacrificed some
sensitivity for precision and reproducibility. The final
detection limits -were still below levels that would be regarded
as significant.
Table 5-1 presents a summary of the analytical detection
limits for each of the 13 toxic metals using the original
protocol and the two subsequent modifications which were
applied.
Organic Compound Analysis
The organic toxic pollutants were determined by the
standard protocol (40 CFR 136 proposed December 3, 1979) which
includes sample preparation, extraction, and analytical
methodologies. Extractions were carried out using methylene
chloride in the case of the acid and base/neutral organic
fractions and with hexane/methylene chloride to obtain the
pesticide-containing fractions. The acid and base/neutral
fractions were reduced in volume and analyzed by gas
chromatog,raphy-mass spectrometry (GC/MS) . The pesticides were
analyzed by electron capture gas chromatography followed by
61
-------
5-1. ANALYTICAL DETECTION LIMITS FOR METALS
(1)
Element
Antimony, Sb
Arsenic, As
Beryllium, Be
Cadmium, Cd
Chromium, Cr
Copper, Cu
Lead, Kb
Mercury, Hg
Nickel, Ni
Selenitm, Se
Silver, Ag
Thallium, Tl
Zinc, Zn
Original Screening
Protocol t2)
Method (pg/D
EGA*
EGA
EGA
EGA.
EGA
EGA.
HG&
Cold Vapor
EGA
EGA
EGA
EGA.
EGA
10
3
0.2
1
1
1
10
0.5
1
9
0.5
2
1
First Modification Second Modification
of Protocol ^ of Protocol ^
Method (jag/1) Method (pg/1)
EGA
EGA
Flame
HG&
Flame
Flame
EGA
Cold Vapor
Flame
EGA
Flame
EGA
Flame
10
3
15
1
25
20
10
0.5
25
9
15
2
25
EGA
Eydride
Flame
EGA
Flame
Flame
EGA
New Cold
Vapor
Flame
Eydride
Flame
EGA
Flame
- 10
10
15
1
25
20
10
0.5
25
10
15
2
1
Heated Graphite Atomizer
(1) Assuming no matrix interferences requiring dilution of sample.
(2) EPA Contract No. 68-01-4492 (September 29, 1977), Exhibit C,
"Protocol for the Measurement of Toxic Substances", Environmental
MDnitoring and Support Laboratory, Cincinnati, Ohio
(3) June, 1978
(4) August, 1978
62
-------
GC/MS confirmation of positive results. Volatile organics
were analyzed by the purge and trap method of introducing the
material into the GC/MS inlet system.
Cyanide Analysis
The standard methods for the wet chemical analysis of total
cyanide and cyanide amenable to chlorination'(Cyanide, A). were
utilized (40 CRF 136) Cyanide analysis is subject to several
sources of interference including:
Metals - The presence of Fer Cd, Ca, Ni, Ag, and Zn may
cause low results due to the formation of stable complexes with
cyanide. The iron complexes may form insoluble precipitates
which are particularly difficult to break up both at the time
of alkaline chlorination of the sampled waste water and during
the,chemical analysis for cyanide.
\
Oxidizing agents - The presence of free chlorine in the
waste water sample will destroy cyanide and cause low analytical
results. The addition of ascorbic acid to destroy chlorine at
the time of sampling is intended to mitigate this problem.
Other oxidizing agents such as peroxides and chromates may also
react with cyanides over a period of time and cause low results.
Sulfides - Sulfide or bisulfide will interfere in the
analysis of cyanide by reacting with the colorometric reagents.
The presence of sulfur dioxide or bisulfite in the waste
water sample should have no appreciable effect on cyanide
results. Detection limits on the order of 1-4 ug/1 can be
achieved by the analytical method employed, but the results
have to be interpreted with regard to the possible interfering
components of the sample.
Hexavalent Chromium (Cr VI) Analysis
The determination of Cr VI in waste water samples is also
subject to a number of interferences which can take effect
either during sampling and storage or during analysis.
Acids - Samples taken and held at a very low pH can
experience the conversion of other forms of chromium into Cr VI
causing a positive interference.
Reducing agents - Samples containing sulfur dioxide,
bisulfite, bisulfide, sulfide, ferrous iron, and other reducing
agents will result in low values of Cr VI by converting it to
trivalent chromium (Cr III) . Under these conditions the
chromates originally present would be included in the total
chromium determination but the analytical results for hexavalent
chromium would be proportionately low. (See Reference 52.)
63
-------
The detection limits for Cr VI using the diphenylcarbazide
colorometric method are on the order of 1-3 ug/1 in the absence
of substances which interfere with color development.
Asbestos Fiber Analysis
The analysis of selected samples for asbestos fiber
(chrysotile) was conducted by the recommended method utilizing
transmission electron microscopy with selected area electron
diffraction as described by Dr. Charles Anderson (EPA, Athens,
Georgia) at the Analytical Protocol Meeting in Denver (November,
1977) (56).
Conventional and Nonconventional Pollutants
All techniques used for the analysis of BPT control
parameters (conventional and nonconventional pollutants) were
those recommended by the Agency. The list of approved test
procedures was published in the Federal Register on October 16,
1973 (38 FR 28758) and may be also found in Title 40 of the Cod'e
of Federal Regulations (40 CFR 136).
5.1.4 Quality assurance Provisions
The Agency and the contractor's analytical laboratories
maintain consistently high standards for accuracy and quality
control. As an in-house requirement, a minium of ten percent of
all samples are routinely run in duplicate. Quantitation is
based on standards that are prepared in the same matrix as the
samples. The standards are also checked by participation in the
EPA Reference Sample Program that utilizes a double blind
technique. (EMSL, Cincinnati, Ohio, Office of Research and
Development.)
Additionally, outside laboratories are retained for checks
on quality by analyzing split samples and running submitted
standards. Accuracy is also insured by analysis of a minimum of
fifteen percent of all samples with spikes by the method of
standard additions. The spikes are added prior to sample
preparation and are carried through the entire sample analysis
procedure.
The contractor's laboratories have consistently maintained
the standards for laboratory certification which are imposed by
the State of California. Certification is dependent upon the
accurate performance of routine analyses on check samples
submitted by the State, as well as on-site inspections by the
State of California's Sanitation and Radiation Laboratory,
Department of Fish and Game, and the U. S. Environmental
Protection Agency, NEIC, Denver, Colorado.
64
-------
The quality assurance provisions outlined in the EPA
Protocol for GC/MS Analysis of Toxic Pollutants are rigorously
adhered to with one added precaution, namely, the use of
internal standards as a means of measuring recovery. Although
not required by the protocol for pesticide analysis, this
technique is utilized as an in-house quality control requirement
to insure the accuracy of results in this analysis.
The high sensitivity of instrumentation used in trace
organic chemical analysis dictates that contamination of the
samples from any possible source must be diligently guarded
against. Accordingly, only glass sample containers with Teflon-
lined lids were used and these were subjected to a three step
cleaning procedure prior to use, even though only new liners and
glass containers were used. All glassware used for sample
preparation and analysis was subjected to a dual cleaning
system.
The sample extraction and preparation rooms are dedicated
solely to toxic' pollutant analysis, and have their own
ventilation systems that are isolated from the other sample
preparation and receipt areas of the laboratories.
A documented system of existing practices, including
calibrations and operational checks is maintained to assure
uniformity of performance and to serve as a basis for alteration
of standardization intervals. A chemist is assigned full time
to maintain this system, assure strict record formating and
controls, and to direct the quality control program of the
laboratories. The primary vehicle of this system is the quality
assurance manual containing the detailed procedures used in
sample preparation and analysis, and the complete records of all
quality control standards, blanks, spikes and duplicates.
5.2 SUMMARY OF ANALYTICM. HESDLTS
f
The results obtained during the screening and verification
sampling program are summarized in Table 5-2 and Table 5-3.
These tables show the frequency and distribution of the
pollutants according to selected plant groupings, concentration
ranges, and subcategories in which the pollutants occur.
Pollutant frequencies as shown in columns 5, 6, 7, and 8
of Table 5-2 are based on the highest individual pollutant
concentration found for each plant's raw waste during the
screening and verification sampling program.
The toxic pollutant asbestos has not been included in
either of the two tables mentioned above. Asbestos
65
-------
TABLE 5-2.
POLLUTANT FBEQUENClf BASED ON SAMPLING PROGRAM RESULTS
INCLUDING RAW WASTE
Pollutants Detected
Antimony
Arsenic
Beryllium
Cadmium
Chrcrniun
Copper
Cyanide
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Zinc
Benzene
Carbon Tetrachloride
Chlorobenzene
1, 2-Dichloroethane
1/1, 1-Trichloroethane
Hexachloroe thane
1/1, 2-Trichloroethane
1,1,2, 2-Tetrachloroethane
Chloroform
1,2-Dichlorobenzene
1, 1-Dichloroethylene
1, 2-Dichlorcpropylene
2, 6-Dinitrotoluene
Ethylbenzene
Fluoranthene
Bis(2-Chloroisopropyl) ether
Hsthylene chloride
Dichlorobromore thane
Trichlorofluorcroethane
f'^riyrv'HVu tn»"4pA^fraTtf»
Naphthalene
Nitrophenol
Pentachlorophenol
' Phenol
Bi3(2-Etnylhexyl) phthalate
Butyl benzyl phthalate
Di-n-butyl phthalate
tdethyl phthalate
Dimethyl phthalate
Benzo(a) anthracene
, Benzo(a) pyrene
3, 4-Benzof luoroethane
Chrysene
Anthracene
Fluorene
Phenanthrene
Pyrene
Tetrachloroethylene
Toluene
Trichloroethylene
Nitrobenzene
2,4-Dinitrophenol
Pollutant Occurrence Based
on Plant Grouping
5 or <5
Plants
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
>5 but slO
Plants
• X
X
>10 Plants
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Pollutant Occurrence Based on
Concentration Classification (ug/1)
£50
28
38
49
45
20
21
25
46
17
46
45
41
9
6
2
1
2
4
1
2
3
15
1
3
1
7
1
1
11
5
2
2
1
1
2
2
20
3
15
5 ,
2 "
1
1
1
1
1
1
1
1
4
7
3
>50 but
<500
19
12
4
4
13
16
15
2
20
7
7
11
18
1
1
2
1
1
_
3
1
1
3
>500 but
12,500
4
3
4
9
9
7
8
1
1
14
1
1
1
1
1
1
2
2
>2,500
1
10
7
2
6
5
8
12
1
1
66
-------
TKSLE 5-3. DISTRIBUTION OF lOLHJffiNTS ACCORDING
TO SUBCMEGORY 1
Pollutants Detected Subcategory Numbers Where Pollutants Found
Antimony ' AlAut 7, 23, 27, 28, 33
Arsenic " " " " " "
Beryllium " " "
Cadmium " " " " " "
Chromium " " " " "
Copper " " " " " "
Cyanide 7
lead , Air but 7, 23,.27, 28, 33
Mercury " " " "„„ " "
Nickel " " " " " »
Selenium " " " » " »
Silver " " " » " »
Thallium " " " « "
Zinc " " " " " "
Benzene 1,' 3, 4, 10, 11, 25, 32
Carbon Tetrachloride 1, 2
Chlorobenzene 1, 35
1,2-Dichloroethane 1, 11, 13, 22, 35
1,1,1-Trichloroetfaane 1
Hexachlorcethane - 4,11
1,1,2-Trichloroethane 1, 10, 35
1,1,2,2-Tetxachloroethane 1, 3, 4, 10, 13, 15, 19, 21, 22, 25, 32, 35
Chloroform 24
1,2-Dichlorobenzene 1, 11, 13
1,1-Dichloroethylene 26
1,2-Dichloropropylene 1
2,6-Dinitrotoluene 1
Ethylbenzene 1, 3, 4, 9, 11, 21, 25, 32
Fluoranthene 8
Bis{2-€hloroisopropyl) ether 22
Methylene chloride 1, 4, 8, 9, 12, 13, 19, 21, 22, 25,26, 32, 35
Dichlorobromoroethane 1, 4, 19, 32
Trichlorofluoromethane 1, 4, 25
Clilorodibrornomethane 19, 32
Naphthalene 1, 32
4-Nitrophenol 17
Pentachlorophenol 2, 3, 4, 8, 15
Phenol 2, 15, 26, 31, 32
Bis(2-Ethylhexyl) phthalate 1, 4, 7, 8, 10, 11, 12,13,15,18,24,25,26,30,31,35
Butylbenzyl phthalate 1, 2, 12
Di-n-butyl phthalate 1, 4, 8, 11, 17, 18, 19, 21, 22, 30, 31, 34, 35
Diethyl phthalate ' 8, 10, 11, 19, 31
Dimethyl phthalate 12, 31
Benzo(a) anthracene 8
Benzo (a) pyrene 8
For name of subcategory, refer to Table 3-2. (Continued)
2 "All" means subcategory numbers 1 through. 35 of Table 3-2.
67
-------
TABLE 5-3. Continued
Pollutants 'Detected Subcategory Numbers Where Pollutants Found
3,4-Benzofluoranthane 8
Chr^sene 8
Jtatihraeene 8
Fluorene 8, 12
Phenanthrene 8
Pyrene 8
ffetradiloroetfiylene 1, 4, 10, 22
toluene 1, 3, 4, 10, 11, 15, 18, 32
Trichloroetbylene lf 4, 25
6P
-------
concentration is reported in million fibers per liter (MFL)
which is not compatible with the concentration units in which
the other pollutants have been reported. Asbestos was found
in three plants at concentration levels of 2.1E8, 2.QE7, and
9.4E4 MFL, respectively, where E is exponential on base 10.
All three plants belong to the Chlor-Alkali subcategory.
69
-------
-------
SECTION 6
PROCESS AND WASTE TREATMENT? INFORMATION DBVU.QPM1OT
AND EVALUATION
6.1 INDUSTRY DATA BASE DESCRIPTION
Information and data on the Inorganic chemicals industry
were obtained from a number of sources. These sources included
literature reviews, plant visits, telephone contacts, and
industry responses to the Section 308 Questionnaires, The type
material gathered from these sources is discussed below.
6.1.1 Literature Review
A review -of the literature has been conducted to identify
and collect information related to manufacturing processes, raw
materials, water use, waste water sources, waste water treatment
technology, raw waste characteristics, and economic data.
Relevant information from reports,, books, papers, conference
presentations and periodicals were identified by computer search
and are presented in the reference section of this report. This
information was incorporated into a broad based assessment of
process and technology practices aimed at selecting the best
available treatment technology and best demonstrated technology
for the various industry subcategories. It also provided the
background reguired for evaluating the subcategorization of the
industries.
6.1.2 Plant Visits
During the screening and verification phase of this
project, much information was gathered from individual plants
relating to production capacity, manufacturing processes, waste
flows, water reuse, waste water treatment systems and
performance, and best management practices (BMP). The lead
visits also provided an opportunity to update and clarify some
of the information given in the 308 responses.
71
-------
6.1.3 Telephone and Direct Contact
Numerous contacts were made with knowledgeable persons in
both industry and government to gather and exchange information
concerning all phases of this study. These sources are cited
in the text as personal communications.
6.1.4 308 Questionnaire Responses
The basis for much of the work in this study is the
responses from industrial inorganic chemical firms to the 308
data requests.
Data from 284 manufacturers' responses were utilized by the
project team for the development of appropriate guidelines for
the inorganic chemicals subcategory. Industrial firms, through
their compliance with the needs of the 308 Questionnaire,
provided a valuable industry-wide data base used extensively in
this analysis.
Essential data elements from each questionnaire were
extracted for the purpose of creating a working data base for
this report. Specific elements selected for this smaller, more-
manageable data base are given in Table 6-1.
These data provided the basis for the subcategory review
through a profile of each industry. After compilation of the
questionnaire data, industry totals for capacity and production
(for the respondents) were available. In addition, derivative
quantities such as percent utilization, effluent per ton of
product, and conversion to metric units were compiled.
72
-------
TABLE 6-1. 308 QUSTIONNAIRE RESPONSE DATA
DATA ELEMENTS
INORGANIC CHEMICALS GUIDELINES STUDY
Datum Reference Description Comments
Manufacturer
Product
Plant
Process
Effluent Treatment
Name
Location
EPA Region
Name
Subcategory
Number of other
Products
Capacity
Production
Age
Name
Volume of Process
Effluent
Volume of Noncontact
Effluent
Type
Permit
Major Pollutants
Confidential
Inorganic
Chemicals
Fiscal year
1976
1976
1976
73
-------
6.2 PROCESS WASTE SOURCES AND CURRENT TREATMENT PRACTICES
6.2.1 Data Acquisition
The information presented in this section was obtained from
a variety of published sources and the available industry
responses to the 308 Questionnaires as well as from plant visits
and interviews with industry personnel conducted by the Agency
and its contractor during the toxic pollutant screening and
verification program. The results of visits and interviews are
documented in field notebooks, interim plant visit reports, and
telephone communication records which are part of the rule
making record.
Plant visits were particularly useful for confirming and
updating the detailed technical information contained in the 308
Questionnaire responses. The cooperative attitude displayed by
industry greatly facilitated the acquisition of reliable
operating data and meaningful sampling results.
6.2.2 Evaluation of Data
Each of the various industrial subcategories in which
verification sampling was conducted was the subject of an
extensive evaluation to provide the technical basis for
selecting candidate advanced treatment technologies and
developing the related base and incremental cost estimations.
In the subsections which follow, individual plant descriptions
are presented according to the general format for each
subcategory:
General Process Description
Description of process reactions and unit operations.
Inventory of raw materials used.
Typical process flow diagram.
Water Use and Waste Source Inventory
Description of individual plants visited, sampled
and plant information from other sources.
Inventory of water uses for contact and noncontact
purposes.
Inventory of raw process waste water sources and
identification of sampling points.
Process waste water quality and flow data.
Solid waste generation and disposal.
74
-------
Control and Treatment Practices
Description of specific treatment technologies
and operating facilities.
Description of the total input to the treatment system
including sources attributed to other production
operations and noncontact water (e.g., cooling
water, etc.).
Evaluation of Production and Waste Flow Data
Tabular summary of plant-specific data.
Waste flows per unit of production (unit waste flows)
with the range and average values.
Solid waste quantities.
Treatment chemical requirements.
Process Modifications and Technology Transfer Options
Best Management Practices (BMP)
Plant area operations and housekeeping.
Runoff control.
Solid waste handling (e.g., fugitive dust and
leachate control, etc.).
6.2.3 Model Plant and BPT Treatment System Specification
The model plant concept plays a central role in both the
development of alternative treatment system designs for priority
pollutant removal and for estimating the related internal costs
of such treatment in each sufecategory. In order to be
representative of a subcategory, each set of model plant
specifications was composited from a profile data summary
derived from the available information on production and waste
flow.
Based on the typically achievable waste flow rate per unit
of production, the model plant was used as a starting point for
an appropriately designed and sized BPT level waste water
treatment system. Certain assumptions were made regarding
"the possible process variations and the specific raw waste
sources incorporated into each model. In most cases, ,it was
appropriate to assume that the waste flow per unit of production
did not vary over the particular range of production capacities
covered. Production rates were selected in most subcategories
to represent the small, mid-range and large size plants
presently in operation. Small subcategories were represented by
single mid-range production rates for the model plants. Cost
estimates were developed for each set of bas,e level (BPT) and
advanced level (BAT/NSPS) treatment system design
specifications.
75
-------
Beginning with Section 11, the model plant and BPT level
treatment system descriptions and specifications for each
subcategory include the following information:
Production rates and mode of operation.
Specific process type and waste sources.
Waste flow per unit of production.
Solid waste generation and handling.
Treatment chemical requirements.
If applicable, the new source model plant is also described
and the design specifications given for its waste treatment
system.
The model plants do not represent exemplary or specific
existing plants, but are typical plants of adequate design
derived from the range of plants and treatment facilities found
in the entire subcategory. For the.purpose of cost estimating,
it is necessary to specify cost rationale, define a set of
initial assumptions, and consider the variability of factors
such as waste water flows, pollutant load, unit treatment
process, plant ager etc. General assumptions have been detailed
under Section 10 of this report and are employed as the basis
for developing baseline model plant cost estimates presented in
the subsequent sections dealing with individual industries.
6.2.4 Dissolved Solids in Waste Water Effluents
Many waste treatment plants discharge final effluent into
watercourses which feed fresh water streams used as sources of
water supply by downstream agencies or industries. Groundwater
aquifers which underlie large portions of the country are tapped
to supply fresh water through wells serving public and
industrial water needs. Saline wastes discharged into streams
or into unlined lagoons can significantly alter the salt content
(total dissolved solids) of the fresh water. Although Federal
regulations seldom limit the total dissolved solids or the
various ions such as chloride, sulfate, bicarbonate, and
nitrate, these constituents can be of serious 'concern to local
water users.
To protect the mineral quality of ground and surface waters
State and local water pollution control agencies typically
establish limits on the discharge of substances which contribute
sodium, potassium, hardness, chloride, sulfate, or
conductivity, which is a measure of total solids in solution.
This restriction can affect the chemicals chosen for waste
treatment. For example, alkaline precipitation can be
accomplished by using lime, which forms an insoluble calcium
sludge, or by adding caustic soda, forming a soluble sodium
salt.
76
-------
In choosing an acid for neutralization of alkaline wastes,
it is important to weigh the overall effects of chloride (from
hydrochloric acid) and sulfate (from sulfuric acid),
particularly with respect to irrigational use of the receiving
water.
Chemicals used in the model plant processes were selected
on the basis of best performance, including consideration of
scaling problems, which can be severe when calcium and sulfate
are at saturation levels. It may be necessary to alter the
nature of chemicals used at a specific plant, in order to meet
local water quality requirements.
77
-------
-------
SECTION 7
ASSESSMENT OF TECHNOLOGY FOR ADVANCED TREATMENT AND CONTROL
7 .1 INTRODUCTION
In the inorganic chemicals industry, pollution abatement
practices vary and a wide range of treatment technologies can be
found, ranging from no treatment to the application of highly
advanced technologies for the removal of specific pollutants.
Until the NRDC Consent Decree, industry attention was
primarily directed towards general pollution problems including
removal of trace metals, but not towards treatment of over 100
individual specific organic compounds now listed as toxic
pollutants. Even with the classical (conventional and
nonconventional) pollutants, treatment technology has been
directed to removal down to the part per million level, whereas
now the thrust is towards part per billion level requirements.
For both these reasons, higher level technologies are not in
place in the inorganic chemicals industry, and it is necessary to
look into technologies that have been applied in other industries
or developed at the laboratory or pilot plant scale specifically
for the removal of these toxic substances from industrial waste
water, and determine whether they can be adopted as viable
technological options.
A list of candidate technologies was compiled from the
literature, in-house expertise, and industry contacts. These
were,evaluated with respect to:
1. Treatment effectiveness
2. Cost
3. Nonwater pollution environmental effects
4. Applications in the inorganic chemicals industry or on
other industrial wastes with similar waste water
characteristics.
79
-------
The anticipation that few of the organic toxic pollutants
would be found in inorganic chemical wastes was justified by the
results of the analytical programs. Only one industrial
subcategory, namely, Chlor-Alkali production using graphite
anodes had potentially significant levels of organic toxic
pollutants. As a result, the initial search for candidate BAT
technologies became limited to treatment technolbgies for the
thirteen metals, cyanide, and asbestos.
The technologies finally adopted were not new or untried
technologies since it was found that most treatment requirements
could be met by taking conventional techniques—for example,
chemical precipitation—and developing them to a higher degree of
engineering and design sophistication, so that optimum removal
efficiencies could be achieved.
The following pages describe the theoretical basis for
treatment systems adopted for BAT application.
7.2 HYDROXIDE PKECIPITATION
Hydroxide precipitation is the most widely used technology
for removing trace metals from waste waters, with lime or caustic
soda commonly used to supply the hydroxide ions. Under suitable
conditions the metals form insoluble metal hydroxides which can
be separated from solution.
The chemistry of the process is not simple, and must be
understood for each metal. Many metals are amphoteric, the
optimum pH for precipitation varies, and organic complexes can
interfere. A simple form of the reaction may be written as:
M-H- + 20H- = M(OH)2 (1)
Metal ion + two hydroxyl ion = insoluble metal hydroxide
If the pH is below the optimum for hydroxide precipitation
soluble complexes form:
M-H- + OH- = M(OH)+ ,(2)
Metal ion -t- hydroxyl ion = soluble metal complex
Since most metals have the capability of coordinating with
other ions or molecules, these simple equations assume that the
hydroxonium ion is the coordinated species. However, if organic
radicals are present, they can form chelates and mask the typical
precipitation reactions:
80
-------
M++ +OH- +nR = M (R)nOH+ (3)
Metal ion 4- hydroxyl ion + organic ions = soluble
metal chelate
Such complexes may require unusual treatment to hydrolyze
them, and their presence often explains why some treatment
practices yield relatively poor results.
Assuming the absence of organic complexing agents, the
treatment levels attainable by hydroxide precipitation can be
forecast from a knowledge of the pH of the system. Figure 7-1
shows the theoretical solubility of those metals which form
insoluble hydroxides, while Table 7-1 shows the solubility
product constants. For comparison, the values for sulfides are
also given.
It is .clear from the range of optimum pH's illustrated that
for waste waters containing more than one metal, no single
optimum pH exists, and problems arise at the threshold of the
alkaline range , (circa pH 10) where some metals have least
solubility, while others are at the point of redissolving as an
anionic species. For successful application as a waste water
treatment technology, careful control of pH must be practiced if
the best removals are to be achieved.
In practice the solubility of metallic hydroxides; and the
tendency for fine insolubles to remain in suspension may yield
effluents which will not meet ug/1 standards, and hydroxide
precipitation is often supplemented by the use of coagulating
agents to improve solids removal, or sulfide co-precipitation to
reduce ultimate solubilities.
In practice, ,the technology uses unit process steps which are
simple, well-established, and well-understood by the industry.
Depending on the quantity of waste flow, the treatment can
either be a batch or continuous operation, with batch treatment
being favored when waste flows are small. In batch treatment the
equipment usually consists of two tanks, each with a capacity to
treat the total waste water volume expected during the treatment
period. These systems can be economically designed for flows up
to 50,000 gallons per day (5).
The treatment tanks serve the multiple functions of
equalizing the flow, acting as a reactor and as a settler.
During operation the waste water is stirred, and a homogeneous
sample is taken and analyzed to determine the chemical dosage
requirements. The chemicals are then added, mixed and stirred
for about 10 minutes. After the reaction is complete, the solids
81
-------
Pb(OH),
10
10
-12
'0123
Figure '7-1.
45 67 8 9 10 11 12 13
Solubility of metal hydroxides and sulf ides
as a function of pH.
14
82
-------
TABLE 7-1. SOLUBILITY PRODUCTS OF TMCE METALS
Solubility Product Constant (log K^}
Ethyl Xanthate
Metal Hydroxide SoLrLde
Cadmium, Cd
Copper, Cu
4-2
Ferrous, Fe
Lead, Pb
Mercury, Hg
Nickel, Ni
Zinc, En
Omcmium (VT) ,Cr+6
13.6
18.6
15.3
16.1
25.4
14,8
15.7
8.9
26.1
35,2
16.9
26.6
52.2
25,7
25.2
_
13.6
-
7.1
16.9
37.8
11.9
8.3
-.
83
-------
are allowed to settle for a few hours. The clear liquid is then
decanted and discharged. Settled sludge is retained to serve as
a seed for crystal growth for the next batch, but must be drawn
off periodically and disposed of, usually in a chemical landfill.
For layer daily flows a typical continuous flow treatment
scheme consists of a flash mixer, flocculator, settling unit with
sludge storage tank, and, in some cases, a filtration system.
The ability to separate the solids from the waste water is
important. Metallic hydroxides tend to be gelatinous and
separ-ate poorly in gravity separators. Finely suspended solids
tend to pass out with the effluent and increase the total metal
content. Thus, improvements in precipitation applications have
been directed toward fine solids removal, and this is reflected
in the addition of various filtration systems and the use of
flocculant aids as improved levels of treatment.
Hydrated lime suspensions are more commonly used than caustic
soda as the hydroxide source because they are cheaper. However,
if there is sulfate ion present in the waste water, gypsum will
be formed:
Ca (OH) 2 + (S04)— = CaS04 + 20H- (4)
Hydrated lime -f sulfate ion = calcium sulfate (gypsum) +
hydroxyl ions
This increases the sludge produced, may cause scaling
problems in pipelines, and may clog a dual media filter. Using
caustic soda is more expensive^' but it generally feliminates the
scaling problem. Total dissolved solids in the form of sodium
salts are increased in the caustic soda tre'ated waste waters.
Although low concentrations of sodium ,are not regarded as
polluting, high levels can make drinking water unpalatable, limit
the use of water for agriculture, and promote degradation of the
structure of arable soils. Thus, where high total dissolved
solids are of concern, lime would be the preferred neutralizing
agent.
This treatment technology is wide'ly applied in treating
industrial waste waters. Industries that are using hydroxide
precipitation include:
Inorganic Chemicals
Plating and Metal Finishing
Mining
Textiles
Steel and Iron
Non-Ferrous Metal Processing and
Electronics
84
-------
Better than 99 percent removal of trace metals have been
reported in the literature with final concentrations in the
treated effluents ranging from sub ppm to low ppm (see Tables 8-1
through 8-10) .
7.3 PERRITE COPKECIPITATION
An interesting variation on the theme of hydroxide
precipitation is a process developed in Japan for the removal of
heavy metals from acidic waste water. The process, known as
ferrite coprecipitation, has the potential for producing a
marketable residual by converting the metal ions in solution into
insoluble ferromagnetic oxides or ferrites which can be removed
magnetically or by filtration (5). The treatment is applied by
adding a ferrous salt to the metal-bearing waste water, then
neutralizing and oxidizing the complex heavy metal-ferrous
hydroxide precipitate by aeration to form the stable ferrite
coprecipitate. Particle sizes are reported to be relatively
large and sludges formed can be safely disposed, of by
landfilling.
Although extensive performance data have not been developed,
the information available indicates that very hiqh removal
efficiencies can be achieved for most of the common heavy metals,
including mercury and hexavalent chromium. The method has not
been considered here as an available technology due to the lack
of sufficient information on chemical dosing requirements, energy
requirements, and performance in situations similar to those
found in the inorganic chemicals industry. In connection with
waste water treatment in the Titanium Dioxide Subcategory for the
sulfate process, the wastes contain considerable amounts of
ferrous iron from the processing of ilmenite ore and the current
practice of neutralization and aeration may involve the same
chemistry as the ferrite coprecipitation process.
7.4 SDLFIDE PRECIPITATION
The basic principle of sulfide treatment technology is
similar to that of hydroxide precipitation. Sulfide is added to
precipitate the metals as metal sulfides and the sludge formed is
separated from solution by gravity settling or filtration.
Sodium sulfide and sodium bisulfide are the two chemicals
commonly used, with the choice between these two precipitation
agents being strictly an economic consideration.
Metal sulfides form according to the following equation:
M-H- -f Na2S = MS + 2Na+ (5)
85
-------
Metal ion + sodium sulfide = insoluble metal sulfide +
sodium ions
Figur.e 7-1 shows the theoretical solubility of the metals
that form insoluble sulfides, while Table 7-1 shows the
corresponding solubility product constant.
The major problem in applying sulfide precipitation
techniques is associated with the toxicity of sulfides. This
warrants both care in application and post treatment systems to
remove excess sulfide. Pretreatment involves raising the pH of
the waste stream to minimize evolution of hydrogen sulfide gas.
A recently developed and patented process to eliminate the
potential hazard of excess sulfide in the effluent and the
formation of gaseous hydrogen sulfide uses ferrous sulfide as the
sulfide source (6) . The fresh ferrous sulfide is prepared by
adding sodium sulfide to ferrous sulfate. The ferrous sulfide
slurry formed is added to a waste water to supply sufficient
sulfide ions to precipitate metal sulfides ' which have lower
solubilities than ferrous sulfide. Typical reactions are:
FeS + Cu++ = CuS + Fe++ (6)
Ferrous sulfide + copper ion = insoluble copper sulfide 4-
iron ion
FeS + Ni (OH) 2 - Fe(OH)2 + NiS ^ (7)
Ferrous sulfide + nickel hydroxide = ferrous hydroxide +
insoluble nickel sulfide
A detention time of 10-15 minutes is sufficient to allow the
reaction to go to completion (7) . Ferrous sulfide itself is also
a relatively insoluble compound. Thus the sulfide ion
concentration is limited by the solubility of ferrous sulfide,
which amounts to about 0,02 mg/lf and the inherent problems
associated with conventional sulfide precipitation are minimized
(8).
One other advantage of this process is that if chromium (VI)
is present, it will also be reduced at the pH of normal operation
(8 to 9) and precipitate as the trivalent hydroxide (Cr III).
Treatment systems for sulfide precipitation are similar to
those used for hydroxide precipitation. A continuous treatment
scheme generally consists of a pH adjustment tank, flash mixer,
flocculator, settling units with sludge storage, and a dual media
filter..
Before the addition of sodium sulfide or bisulfide the pH of
the incoming wasteflow is adjusted to pH of 7-8 in the first
86
-------
reaction tank to reduce the formation of obnoxious hydrogen
sulfide gas. The chemicals are then added to the flash mixer
where they are thoroughly mixed with the waste water.
After the flash mix, the waste water passes through a
flocculating basin where the floe agglomerates and settles in the
settling unit. The overflow from the settling unit generally
passes through a filter to remove any fine precipitates. Any
excess sulfide will need to be removed before final discharge.
This can be achieved either by aeration or other chemical
oxidation techniques.
Sulfide precipitation is being practiced in the inorganic
chemicals industry, mining industry, textile industry, and
nonferrous metal processing industry. Most of the Chlor-Alkali
industry is applying this technology to remove lead or mercury
from its waste streams.
Literature citations on the efficiency of sulfide
precipitation (9, 10, 11) indicate that most results are in the
sub ppm range, and that sulfide treatment is superior to
hydroxide treatment for the removal of several trace metals. A
recent report concluded that, with no complexing agents in the
waste, the following effluent quality can be achieved (11).
Meta1s Coneentration
Cadmium 0.01 mg/1
Copper 0.01 mg/1
Zinc 0.01 mg/1
Nickel 0.05 mg/1
Chrome (total) 0.05 mg/1
Adding ferrous "sulfide as a polishing step to remove residual
metals appears to be a promising, economical technology.
Although there is no full-scale treatment system operating in the
inorganic chemicals industry, pilot studies on chrome pigment
waste indicate that this process is superior to sulfur dioxide
reduction followed by hydroxide precipitation (12).
7.5 THE XANTHATK PROCESS
The use of xanthates for the removal of metals from waste
streams appears to be a new, promising technology for treating
metal-bearing waste waters. Xanthates contain functional groups
capable of forming insoluble complexes with metals, and the
sludge so formed can be separated by conventional means.
87
-------
Xanthates can be generated by mixing starch or cellulose with
carbon disulfide in a caustic medium. Three types of xanthates
have been proven in bench pilot scale studies to be effective in
removing cadmium, chromium (III), copper, iron, lead, mercury,
nickel, silver and zinc from industrial waste waters (13-20).
These are:
Soluble starch xanthate with a cationic polymer,
Insoluble starch xanthate, and
Fibrous cellulose xanthate
The general removal mechanism is as follows:
2 [ROCS (=S)Na] + M-H- = [ROCS(=S)2M] + 2Na+ (8)
Xanthate + metal ion = insoluble metallic xanthate +
sodium ions
where R = starch or cellulose
Unlike hydroxide precipitation, this process is reported to
be effective in removing metals over a wide pH range of 3 to 11,
with an optimum range between 7 and 9.
Brass mill waste waters, lead battery effluent, circuit board
rinse waters, electroless copper plating rinse waters,
pyrophosphate electroplating rinse waters, and copper etching
rinse waters were studied in a pilot plant with insoluble starch
xanthate as the complexing agent (20). This pilot study
demonstrated that the xanthates can either be added to a reactor
to mix with the waste waters or be applied as a precoat on a
pressure filter (20) . Results of these pilot studies showed that
metals were reduced to below 50 jug/1 (ppb).
Another study indicated cellulose xanthate is as effective as
starch xanthate in removing trace metals. The following table
summarizes the result of the study with a cellulose xanthate
dosage of 90 mg/1 and a contact time of 30 minutes (18-19):
Concentration, mg/1
Metals Influent Effluent
Cadmium
Chromium
Copper
Iron
Lead
Nickel
Zinc
1.35
0.30
1.6
3.1
3.9
2.4
1.0
0.027
0.022
0.06-0.14
0.08-0.36
0.008-0.021
0.077
0.03-0.04
88
-------
This study also concluded that cellulose xanthate is superior
to starch xanthate in terms of sludge settling characteristics,-
filterability,'and handling.
Xanthate may also be used as a complexing agent to prevent
the formation of soluble anions from insoluble amphoteric metal
hydroxides.
The xanthate process is a relatively new technology, and the
reagent compounds are not yet available in commercial quantities.
More information is needed on dosage rates in continuous flow
operations. Potentially the metals can be recovered by leaching
the xanthate complex with nitric acid,-but metal recovery has not
been demonstrated yet. Sludge disposal problems may arise if the
sludge complex is unstable and, if xanthates are to be generated
on site, care will be needed in handling the hazardous carbon
bisulfide.
7.6 ION EXCHANGE
Ion exchange is a chemical reaction between the ions in
solution and the ionic sites on an exchange resin. Many natural
solids (e.g., soils, proteins, and zeolites) exhibit such
exchange characteristics. However, synthetic resins are the
predominant ones used for ion exchange applications in modern
industrial technology. These resins contain functional groups
that can react with the ions in solution*. Depending on these-
functional groups, the resins can be classified into:
Strongly acidic cation exchanger,
Weakly acidic cation exchanger,
Strongly basic anionic exchanger, and
Weakly basic anionic exchanger.
Cation exchangers are qapable of exchanging with cations in
solution. Strongly acidic cation exchangers contain functional
groups such as sulfonates, (-SO3H and -SOSNa), while weakly
acidic exchangers have functional groups derived from carboxylic
acids, (-COOH and -COONa).
Anionic exchangers are used to exchange with the anions in
solution. In general, strongly basic exchangers contain amine
functional groups {-R3NOH and -R3NC1), and weakly basic
exchangers contain ammonia functional groups (-NH3OH and -NH3C1)
When the functional groups are used up in the reaction, the
resins can usually be regenerated. Cationic resins can be
regenerated by sodium chloride, hydrochloric acid, sulfuric acid
or sodium hydroxide. Anionic resins are regenerated by sodium
89
-------
hydroxide, ammonium hydroxide, sodium carbonate, sodium
chloride, or hydrochloric acid.
The exchanger can either be added to the waste waters in
batch operations or be packed in a fixed bed or column. Fixed bed
is by far the more effective and hence more popular. The
operation generally follows a four-step cycle: exchange
(service), backwash, regeneration, and rinse.
During the exchange step, the reaction between the ions in
solution and the ionic sites in the resin takes place as the
waste water passes down the bed. The reaction is generally
regarded as a result of electrostatic attraction (20).
Therefore, the size of the hydrated ion and the charge on the ion
are the determining factors for the exchange reaction. A
trivalent ion is attracted more strongly than a bivalent ion
which is in turn attracted more strongly than a monovalent ion.
For ions with the same charge, the smaller hydrated ion is
capable of moving closer to the exchange site, and is thus
favored.
Many synthetic resins contain functional groups that are
selective to certain metals. For example, a resin manufactured by
a European company reacts preferentially with mercury (Hg+4-) and
mercuric chloride (HgCl+) ions according to the following
equations:
2RSH + Hg++ = RSHgSR + 2H+ (9)
Resin + mercury ion = insoluble resin complex +
hydrogen ions
I
RSH -f HgCl+ = RSHgCl + E+ (10)
Resin •*• mercuric chloride ion = insoluble resin complex +
hydrogen ions
The exchange reaction is governed by the law of mass action.
During the reaction, the affinity of the resin for the two ions
is so great that essentially all the mercury or mercury
chloride-resin complex formation equilibria are shifted toward
the formation of Hg++ and HgCl+ which are rapidly removed. A 5
ppb residual mercury concentration in the effluent is achieved by
this process (22).
After all the exchangeable sites in the resin are used up,
the bed is backwashed by passing clean water through to loosen up
the bed and to remove any fine particulates that are trapped
inside the bed.
After the backwash cycle the resins can be regenerated with
the appropriate regenerant.
90
-------
RSHgCl + HC1 = RSH + HgC12 (11)
Insoluble resin complex + hydrochloric acid = regenerated
resin + mercuric chloride
One attractive feature of the ion exchange process is that it
concentrates the metals in the regeneration step, and thus
provides a potential for their recovery. However, if recovery is
not feasible, this creates a secondary stream which needs to be
treated.
A recent study found that sodium alumino silicates (zeolites)
might be a low-cost exchanger that can be discarded after a one-
time use (22). This would eliminate the regeneration step. On a
batch study with a five-minute contact time, cadmium and mercury
were removed to below 10 ppb. Thermodynamic considerations show
this exchanger to have a high affinity for cadmium, copper,
mercury, nickel, silver, zinc, cesium, and barium.
Ion exchange is a proven technology that can reduce metal
concentrations down to low levels. However this technology is
used only in limited industrial pollution abatement applications
because of the high cost associated with the process.
Consequently, ion exchange has not been recommended in this
report for BAT technology.
7.7 REDOCTIOS PROCESSES
«
Many metals can exist in solution in several oxidation
states, and it may be necessary to convert from a higher valency
state to a lower one in order to apply a given chemical reaction.
The classic example is chromium, which as the trivalent chromic
ion will precipitate as the hydroxide in alkaline solution, while
the hexavalent chromate or dichromate ion will not. The latter
needs to be reduced if precipitation is to occur.
Hexavalent chromium (e.g., CrO4= and Cr207=) is toxic and
soluble. The most efficient way of removing this from solution
is a two-step process of reduction followed by precipitation.
Chromium (III) is much less toxic than chromium (VI), and
forms an insoluble hydroxide which can be removed from solution
by settling and filtration.
A number of chemicals are used for the reduction of chromium.
Most common are sodium bisulfite, sodium metabisulfite, sulfur
dioxide and ferrous salts. The reduction is accomplished readily
at low pH with these reagents. Typical reduction reactions are:
91
-------
3SO2 + Cr207= 4- 2H+ = 2Cr-t--H- 4- 3804= 4- H20 (12)
Sulfur dioxide + dichromate ion 4- hydrogen ion = trivalent
chromium ion 4- sulfates and water
3303= + Cr207= + 8H+ = 2Cr+++ + 3SO4 = +4H20 (13)
Sulfite ion + dichromate ion + hydrogen ion = trivalent
chromium ion +• water
6Fe++ 4- Cr207= + 14H+ = 2Cr+++ + 6Fe+++ + 7H20 (14)
Ferrous ion + dichromate ion 4- hydrogen ion = trivalent
chromium ion + ferric ion + water
The reduced chromium and the ferric ions produced in the
third equation will exist as the soluble sulfate at acid pH's.
If the pH is above 5, the reaction rate is drastically reduced,
and although dithionite will effect reduction at neutral pH's, it
is very costly and its use may be contraindicated.
After the reduction step, lime or caustic soda is added to
raise the pH to 8.5-9.0. Trivalent chromium will be
precipitated.
Cr-H-+ + 30H- = Cr(OH)3 (15)
Trivalent chromium ion + hydroxide ion = insoluble
chromium hydroxide
The theoretical solubility limit of chromium hydroxide is
above 0.02 mg/1 (8). It is reported that applying sulfur dioxide
to a pigment waste consistently reduces Cr.(VI) and Cr(T) to 0.5
mg/1 and 1.5 mg/1 respectively as 30-day averages (9) (10) . By
applying ferrous sulfide to a plating waste with an initial
Cr(VI) concentration of 128 mg/1 and Cr(T) concentration of 153
mg/lr an effluent quality of less than 0.05 mg/1 of either
species is achieved (12).
A one-step precipitation-reduction process using sodium
bisulfide is used in a dichromate plant to remove chromium' from
its waste water. An effluent quality 'with less than 1 mg/1
Cr(VI)r and less than 5 mg/1 Cr(T) was reported (3)..
One other common reduction process is the application of
sodium borohydride to reduce metals in waste streams. Sodium
borohydride is a mild but effective reducing agent (3), and is
currently used in some chlor-alkali plants to reduce the soluble
mercury ion to metallic mercury which is removed from solution by
carbon adsorption:
92
-------
4Hg++ + BH4- + 80H- = 4Hg +' B (OH) 4- + 4H20 (16)
Mercury ion + borohydride ion 4- hydroxyl ion = insoluble
mercury metal + borate ion + water
A mercury level of 0.01 mg/1 in the final effluent has been
reported (3).
Sodium borohydride is also reported to be effective in
removing silver, mercury, gold, lead, and cadmium (5). However,
this technology is only being applied in limited cases, the cost
of the chemical being the major drawback. The cost of sodium
borohydride was $16.00 per pound in 1978 (23).
7.8 OXIDATION PROCESSES
The oxidation of organic substances is generally carried out
by thermal processes such as wet oxidation and incineration, or
by biological processes such as the activated sludge process,
trickling filters, biodiscs, and aerated lagoons.
Incineration is actually a combination of oxidation and
pyrolysis. Both involve chemical changes resulting from heat.
Oxidation involves actual reaction with oxygen, while pyrolysis
refers to rearrangement or breakdown of molecules at high
temperatures in the absence of oxygen. There are five types of
incinerators available commercially. These are rotary kiln,
multiple hearth, liquid injection, fluidized bed, and pyrolysis
(24) . A minimum temperature of 1000 degrees C and a residence
time of two seconds is required for the reaction to proceed.
This process has been shown to be successful in reducing
pesticides to harmless molecules (25) .
Wet oxidation is a process in which an aqueous waste can be
oxidized in the liquid phase in a closed, high-temperature, high
pressure vessel. This reduces some of the problems (such as air
pollution from exhaust gas), inherent in incineration. Wet
oxidation has been used for a variety of wastes including pulping
waste and acrylonitrile liquor (26). A percent reduction in
excess of 99.8 of some of the toxic pollutants has been reported
(27).
Thermal oxidation processes are not expected to have much
application in the inorganic chemicals industry, mainly because
of the high energy cost required and the low level of organic
contamination found in the wastes.
The application of chemical oxidation to industrial wastes is
well established for cyanides, sulfite, ammonia, and other
93
-------
harmful species in dilute waste streams (phenols, mercaptans,
polysulf ides, etc.). Common chemicals used as oxidizing agents
included chlorine, hypochlorite, hydrogen peroxide, potassium
permanganate, ozone, and chlorine dioxide. Air and oxygen are
also used.
The most widely used chemical oxidation technology applicable
to the inorganic chemicals industry is the oxidation of cyanide.
The oxidation reaction between chlorine and cyanide is believed
to proceed in two steps as follows:
CN- + C12 = CNC1 + Cl- (17)
Cyanide + chlorine = cyanogen chloride + chloride ion
CNC1 + 2OH- = CNO- + Cl- + H20 (18)
Cyanogen chloride + hydroxyl ion = cyanate ion + chloride
ion + water
The formation of cyanogen chloride (CNC1) is essentially
instantaneous. The second reaction, the formation of cyanate, is
accomplished most rapidly and completely at a pH of 10 or higher
(9r 28). A detention time of 30 minutes to two hours is usually
allowed .
The cyanates can be further decomposed into nitrogen and
carbon dioxide by excess chlorination or acid hydrolysis:
2CNO- -I- 40H- + 3C12 « 6C1- + 2CO2 + N2 + 2H2O (19)
Cyanate + hydroxyl ion + chlorine = chloride ion +
carbon dioxide + nitrogen + water
CNO-
2H2O = CO2 + NH3 + OH-
(20)
Cyanate + water = carbon dioxide + ammonia + hydroxyl ion
The first reaction can be accomplished in about one hour if
the pH is adjusted to 8.0-8.5. Acid hydrolysis usually takes
place at pH 2-3 and care must be taken to avoid the liberation of
the toxic cyanogen chloride as a gas. Hydrolysis is not usually
the chosen option.
Other common chemicals used to oxidize cyanide include sodium
hypochlorite, ozone, and hydrogen peroxide. The reaction for
sodium hypochlorite is essentially the same as for chlorine. For
ozone and hydrogen peroxide, the oxidation step proceeds as
follows:
94
-------
O3 + CN- = O2 + CNO- (21)
Ozone + cyanide = oxygen + cyanate ion
H202 4- CN- = CNO- + H2O (22)
Hydrogen peroxide + cyanide = cyanate ion + water
The advantage of using these two oxidizing reagents is that
no dissolved solids are added to the waste water. In addition,
excess chlorine is not discharged.
A patented process uses hydrogen peroxide and formaldehyde to
decompose cyanide at about 120 Deg. F. This has the advantage of
precipitating cadmium and zinc simultaneously (9).
Alkaline chlorination is currently being practiced in one
hydrogen cyanide production plant. Laboratory studies in the
plant indicated that the presence of ammonia in the waste water
reduces the efficiency of cyanide removal. It is well known that
ammonia reacts with chlorine or hypochlorous acid to form
chloramines:
NH3 + HOC1 = NH2C1 + H20 (23)
Ammonia -f hypochlorous acid = monochloramine + water, etc.
NH2C1 + HOC1 = NHC12 + H20 (24)
NHC12 + HOC1 = NC13 + H2O (25)
If excess chlorine is added, chloramines can be converted
into nitrogen oxide(s):
2NH3 + 4HOC1 - N2O + 4HC1 + 3H20 . , (26)
This equation is not exact because the final form of nitrogen
oxide is. believed to be a mixture of nitrous oxide, nitrogen
dioxide and nitric oxide.
The treatment of cyanide by chemical oxidation is currently
practiced in the following industries:
Inorganic Chemicals (Hydrogen Cyanide Production)
Mining
Plating
The free cyanide level after treatment is generally below 0,1
mg/1 (9) .
95
-------
7.9 MEMBRANE PROCESSES
Membrane processes have emerged in the last decade as a
promising new technology for the treatment of saline water and
waste waters. A membrane is a semi-permeable barrier which
allows the transport of some molecules (ions) and retains others.
The driving force can either be electropotential differences
(electrodialysis) or pressure difference (reverse osmosis and
ultrafiltration). The major application of these processes has
been the desalination of brackish water and sea water. More
recently, these have also found application in a number of
industries, including;
Mining
Electroplating
Metal Finishing
Printed Circuit Board Manufacturing
Battery Manufacturing
Pulp' and Paper
Food Processing
In electrodialysis, an even number of alternating anion and
cation selective membranes are placed between two electrodes.
When current is applied the anions are attracted to the anode,
and cations are attracted to the cathode. In the process of
migration, the cations pass through the cation-permeable membrane
and are blocked by the anion-permeable membrane. Likewise, the
anions pass through the anion-permeable membrane and are blocked
by the cation membrane. This results in alternating paths of
purified water and concentrated- reject (Figure 7-2) J
The electrodialysis membranes are made very thin and are
assembled in stacks. The flow path is the active portion of the
cells. Pretreatment to.remove suspended materials is absolutely
essential. Other materials in the waste feed that may lead to
membrane fouling include high organic content, calcium sulfate,
and certain complex ions such as ZnCl- which can .partially
convert the anion membrane to the cation form, with significant
loss in system performance (28).
As ionic concentration decreases, the electroconductivity of
the water also decreases, making it less efficient to remove the
remaining salt. Most operations do not produce a product water
of less than 500 mg/1 total dissolved solids.
Reverse osmosis (RO) and ultrafiltration (UP) are similar in
basic concepts. Both are pressure-driven separation processes
that employ high-flux semi-permeable membranes operating under
96
-------
PEOOOCT
WATER
Figure 7-2. Electzodial^is process.
97
-------
dynamic flow conditions (29). In contrast to electrodialysis,
these involve the transport of solvent, not solute, across the
membrane.
Osmosis is a process in which solvent from a dilute solution
is transported spontaneously across a semi-permeable membrane-
into a concentrated solution. By applying enough pressure to
overcome this osmotic pressure, reverse osmosis, i.e., the
passage of solvent from a concentrated solution to a dilute
solution through a semi-permeable membrane, occurs. The
operating pressure of reverse osmosis units is usually between
350 and 600 psi. Ultrafiltration usually operates at a much
lower pressure (5 to 100 psi). The predominant transport
mechanism is selective sieving through pores. The membrane
retains high molecular weight dissolved solids such as synthetic
resins, colloids, and proteins. The upper and lower molecular
weight limit is generally defined as 500,000 ~and 500
respectively.
Membranes are usually fabricated in flat sheets or tubular
forms. The most common material is cellulose acetate but other
polymers such as polyamides are used. There are four basic
module designs: plate-and-frame, tubular, spiral-wound, and
hollow fiber. Table 7-2 is a comparison between the various
reverse osmosis modules. Membrane processes are effective in
removing (concentrating) inorganic and organic substances from a
wastestream. Usually extensive pretreatnient is required to
reduce the suspended solids and control pH. There are
uncertainties about" operation efficiency, membrane lifetime,
rejection specificity, and other factors. If recovery is not
feasible, the concentrated reject must be disposed or treated by
other methods. The high operation and capital cost limits the
widespread application of these technologies. For these reasons
membrane technique is not recommended as a BAT technology for
this industry.
7.10 ADSORPTION
Adsorption is a surface phenomenon in which a substance is
accumulated on the surface of another substance. Sorption of a
solute on a solid surface is widely used in pollution abatement
practices. The term "adsorbate" refers to the substance being
concentrated, and the term "adsorbent" refers to the material
that provides the surface.
Activated carbon is the prevalent adsorbent used. Both
inorganic and organic substances are known to be removed
effectively by activated carbon. Certain chlor-alkali plants are
currently using activated carbon as a polishing step to remove
mercury.
98
-------
TABLE 7-2. COMPARISON OF REVERSE OSMOSIS CONCEPTS
vo
Plate-and-Frame
Large tubes
Spiral
,
Polyamide hollow
fine fibers
Cellulose acetate
hollow fine
fibers
Packing
Density
2 3
(ftyftj)
150
50
250
5000
2500
Water Flux
at 600 psi
(gal/
day/ft2)
10
10
10
1(400 psi)
3(250 psi)
Water Output
Per Unit
Volume (gal/
day/ft2)
1500
500
2500
" 5000
7500
Parasitic Pressure
Sodium
Chloride
Rejection
Very good
Very good
Very Good
Fair
Good
Losses (psi) Useful
Peed
Channel
30
50
10
10
10
Product
Channel
30
10
50
50
50
pH
Range
2-8
2-8
2-8
0-12
3-7
Ease of
Cleaning
Fair
Very good
Good to
very good '
Fair
Fair
Source; Weber, Physicochemical Processes, 1972,
-------
Activated carbon is made by charring basic substratesr such
as wood, cokef coal, shell, husks, etc., at 600 degrees C in a
controlled atmosphere, where oxygen is kept low by adding carbon
dioxide or steam. This process drives out volatiles, leaving a
porous carbon lattice in an "activated" state.
Activated carbon can be obtained in powdered and granular
form. Powdered carbon is about 50-70 microns in diameter, and 90
percent should pass through a 300-mesh screen. Granular carbon
is about 0.1-1 mm in diameter, and because of this is three times
more expensive than powdered carbon.
The application involves the passage of the waste waters through
a contact bed. When the bed is exhausted, the carbon is either
regenerated or sent to landfill. It is economical for large
plants to regenerate the carbon. This can be done either by
thermal regeneration'in a rotary kiln or multihearth incinerator,
or by chemical regeneration by using oxidizing agents such as
hydrogen peroxide or acids and bases.
The application of carbon adsorption has been mainly in
organic waste treatment. Recently, there are studies indicating
the effectiveness of carbon adsorption in removing mercury,
cadmium, cyanide, chromium, lead, nickel, zinc, arsenic, and
copper (30, 31).
An interesting development in carbon technology is its use
after the waste water is- ozonized. This combination (known as
Bacteriologically Activated Carbon or BAG) has proved effective
in treating otherwise biologically inactive organic compounds.
The process involves chemical modification of the organics by the
ozone. Maintenance of an aerobic region on the carbon allows a
biologically activated film to develop and the modified organics
are further treated by a mixed process of biological oxidation
and carbon adsorption. The system has the advantage of being a
potential add-on to existing BPT systems, and should be cost
effective since it has been found that the carbon only needs
regeneration at infrequent intervals.
No industrial applications of this technology are known,
although research is under way (32).
Bacteriologically Activated Carbon is a very attractive
potential BAT technology for the removal of organic toxic
pollutants from waste streams, although no application to 'the
industry subcategories studied in this report was found.
100
-------
7.11 FLOORIDE REMOVAL
The conventional method of treating fluoride-bearing wastes
is to precipitate the fluoride as calcium fluoride by the
addition of lime. The reaction is:
Ca(OH)2 + 2F- = CaF2 + 20H- (27)
Hydrated lime + fluoride ion = insoluble calcium fluoride +
hydroxyl ion
Using this process alone, it is difficult to remove fluoride
to below 8 mg/1 due to the solubility of calcium fluoride (9,
33). Adding alum with the lime generally improves the removal
efficiency. Fluoride ions are removed as follows:
A1(OH)3 + P- = A1(OH)2F + OH- (28)
Aluminum hydroxide + fluoride ion =
aluminum monofluorohydroxide + hydroxyl ion, etc.
A1(OH)2F + F- = Al(OH)F2 + OH- (29)
A1(OH)F2 + F- = A1F3 + OH- . (30)
Complexed fluorides are also adsorbed to some extent on the
aluminum hydroxide surface and removed in the coagulation process
(33). Large amounts of alum (5000 mg/1) are required to reduce
the fluoride concentration to below 1 ppm.
Activated alumina has been shown to be effective in removing
fluoride and arsenic in waste water (34) and fluoride from
drinking water in municipal water treatment practice (35-38).
Typically, the fluoride content of raw water can be reduced from
about 8 to 1 ppm (38) . Application of activated alumina to high
fluoride industrial wastes shows that a low ppm effluent can be
achieved (39), although high capital and operation costs
generally limit the wide application of this process.
Certain process operations used -in the manufacture of
inorganic fluoride compounds involve the use of sulfuric acid and
starting materials which contain silicate or borate impurities.
This may lead to the formation of wastes containing
fluorosulfonate, hexafluorosilicate or tetrafluoroborate complex
ions. Although tetrafluoroborate is usually a very minor
constituent and the hexafluorosilicate is readily hydolyzed in
treatment systems, the fluorosulfonate ion is fairly stable and
presents a serious problem where low levels of total fluoride are
required. The lime precipitation method is not effective in
removing the fluorosulfonate and the effectiveness of adsorption
techniques is not known.
101
-------
7.12 CHLORINE KEMO¥AL
The removal of residual chlorine (in the form of
hypochlorite) in industrial waste water is normally accomplished
by the addition of sulfur dioxide or a related reducing agent
such as sodium bisulfite or sodium metabisulfite. Typical
reactions are shown in Equations 31 and 32.
SO2 4- OC1- 4 H20 = H2SO4 4 Cl- (31)
Sulfur dioxide 4 hypochlorite ion 4 water = sulfuric acid
+ chloride ion
Na2SO3 4- OC1- = Na2SO4 + Cl- (32)
Sodium sulfite + hypochlorite ion = sodium sulfate 4
chloride ion
Alternatively, hydrogen peroxide, although relatively
expensive may also be used for dechlorination according to
Equation 33.
H202 4 OC1- = H20 4 O2 4 Cl- (33)
Hydrogen peroxide 4 hypochlorite ion = water 4 oxygen 4
chloride ion
In the chlor-alkali industry, certain waste water streams may
have a sufficiently high loading of - chlorine to warrant recovery
of the product by air stripping, steam stripping, or extraction
by carbon tetrachloride. In some locations, a market exists for
sodium or calcium hypochlorite solutions which can be generated
by treating the tail gases with caustic soda or lime. This may
serve as a means for disposing of waste chlorine which cannot be
economically recovered. As alternatives for waste chlorine
disposal, the streams may be treated to form the hypochlorite and
then decomposed thermally or catalytically. These technologies
are discussed in Section 11 dealing with the chlor-alkali
industry. Chlorine residuals remaining after the recovery and/or
decomposition steps have been taken would be amenable to
treatment with reducing agents such as sulfur dioxide, bisulfite,
or hydrogen peroxide as described above.
102
-------
SECTION 8
TREATABILITY ESTIMATES AND LONG-TERM DATA ANALYSIS
8.1 THE DEVELOPMENT OF TREATABILITY ESTIMATES
The review of technological treatment options applicable to
the removal of toxic pollutants has led to the conclusion that
the particular contaminants found in the raw process waste
waters of the subject industries can be effectively controlled
by the proper application of fairly well-known and demonstrated
techniques. In order to proceed from a general discussion and
description of techniques to a detailed evaluation for each
subcategory of the levels of removal that can be expected, a
summary is now presented of selected treatability data for the
13 toxic metals.
The treated waste concentrations and removal efficiencies
reported in the literature are assumed to represent the best
performance characteristics that can be obtained under the
specified operating conditions. The treatment technologies
considered can thus be assigned a set of optimum conditions and
best performance estimates for removal of the particular toxic
metals that are amenable to treatment. Taking each metal in
turn, Tables 8-1 through 8-10 give the initial and final
concentrations, the removal efficiencies, and the pH conditions
for different treatment technologies. The best performance
estimates for metal removal are derived from the tabulated data
and are utilized in turn as the bases for making long-term
achievable performance estimates. The sequence of analytical
steps is:
1. Review and analyze applicable performance data.
2. Estimate best performance under optimum treatment
conditions.
3. Estimate achievable performance under expected
industrial operating conditions.
103
-------
8-1. WKEER TREKIMENr OPTIONS AND DMA. -
MTEQDNY AND ARSENIC REMOVAL
Treatment ^technology
Antdmony
Lame/Filter
Ferric cMoride/Filter
M.wiyS'il'ter
Arsenic
Lime Softening
Sulfide/Filter
Lime (260 mg/1) /Filter
Lime (600 rng/1) /Filter
Ferric sulfate
Ferric sulf ate
Ldbme/Ferric Chloride/
Filter
Activated alumina
(2 mg/1)
Activated carbon
(3 mg/1)
Ferric Chloride
Ferric Chloride
pH
11.5
6.2
6.4
—
6-7
10.0
11.5
5-7.5
6.0
10.3
6.8
3.1-3.6
_
—
Initial
Concen-
tration
(mg/D
0.6
0.5
0.6
0.2
-
5.0
5.0
0.05
5.0
3.0
0.4-10
0.4-10
0.3
0.6-0.9
Final
Concen-
tration
(mg/1)
0.4
0.2
0.2
0.03
0.05
1.0
1.4
0.005
0.5
0.05
<0.4
<4.0
1 0.05
<0.13
Removal
(%)
28
65
62
85
-
80
72
90
90
98
96-99+
63-97
98
-
References
40
40
40
9,
9,
41
41
42
41
9,
43
43
9,
9,
10
10
10
10
10
104
-------
8-2. TREMMENT OPTIONS MID DftTA -
BEEYLLIUM -KND CMMIOyi REMOVAL
Treatment Technology pH,
Beryllium
Lime/Filter - 11.5
Cadmium
Line (260 mg/1) /Filter 10.0
Lime (600 rag/1) /Filter 11.5
Lime Softening 5-6.5
LiirB/SxxLficte 8.5-11.3
Perrons Sulfide (Sulfex) 8.5-9.0
Ferrite cx>precipitation/ neutral
Filter
Initial
Concen-
tration
Cmg/1)
0.1
5.0
5.0
0.44-1.0
0.3-10
4.0
240
Final
Concen-
tration
(mg/D
0.006
0.25
0.10
0.008
0.006
<0.01
0.008
jRernoval References
(%)
99.4 40
95 41
98 41
92-98 8
98+ 44
99+ 7,8,11
99+ 5
105
-------
TASTE 8-3. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
COPPER REMOWL
Treatment Technology
lajre/Filter
Lime (260 ng/1) /Filter
Line (600 mg/1) /Filter
Ferric sulfate/Pilter
Line
Lima
Mum
Idne/Sulfide
Ferrous sulfide (Sulfex)
Ferrous sulfide (Sulfex)
Ferrite Coprecipitation/
PH
8.5-9.0
10.0
11.5
6.0
>8.5
9.5
6.5-7.0
5.0-6.5
8.5-9.0
8.5-9.0
_
Initial
Concen-
tration
(mg/1)
3.2
5.0
5.0
5.0
10-20
3.0
3.0
50-130
3.2
4.0
Final
Concen-
tration
-------
TSBLE 8-4. WftSTE WKEER TRES3WEOT OFTICKS SND EERFQi8fflHCE DKEA SWMBJSf -
CHBCMIOM III MJD GHKMIUM VI SEMCWSL
Treatment Technology
Chromium
Lime (260 rag/1) /Piltar
Line {600 mg/L) /Filter
Ksduct ion/Litre
Reduction/Liire
lame Softening
Lime/Filter
lane
lams
Ferrite aaprecipitatlon/
Filter
Iterric sulfate
Ferric sulfate/Filter
Chrcmiura VI
Activated carixsi
(pulverized, Pitts-
burgh type EC)
Same as above
Activated carbon
(granular)
Ferrite cxjprecipitation
Sulfur dioxide reduction
Bisulfite reduction
pa
10,0
11.5
7-8
7-8
10.6-11.3
7-9
9.5
9.5
6.5-9.3
3.0
2.0
6.0
. —
Initial
Concen-
tration
Omg/l)
5.0
5,0
140 (as
Cr VI3
1300 (as
Cr VI)
—
—
15
3.2
25
5.0
10
10
3
0.5
Final
Concen-
tration
(mg/1)
0.1
0.1
1.0
0.06 CrIII
0.15
0.05
0.1
<0.1
0.01
0.05
1.5
0.4
0.05
not
detectable
0.01-0.1
0.05-1.0
Rancfval Eef erenees
(%)
98 41
98 41
9,10
3,9,10
98+ 46
47
45
45
n,,,,,^, e
984- 46
99 41
85 48
96 48
98 41
5
9,10
9,10
107
-------
TOBT.E 8-5. WSSTE WATER IKEATMENT OPTIONS MID PERFOKffiNCE Dftim SOMMARY
EEM) KEMOWAL
Treatment Technology
Lime (260 mg/1)
Lime/filter
Idme (260 mg/1) /Filter
Line (600 nng/1) /Filter
Ferrous sulfate/Filter
Sodium hydroxide (1 hour
settling)
Sodium hydroxide (24 hour
settling)
Sodium hydroxide/Filter
Sodium carbonate/Filter
Sodium carbonate/Filter
Sodium carfDonate/Filter
Ferrous sulfide (Sulfex)
Ferrite coprecipitation/
PH
10.0
8.5-9.0
10.0
11.5
6.0
5.5
7.0
10.5
10.1
6.4-8.7
9.0-9.5
8.5-9.0
___
Initial
Concen-
tration
(mg/1)
5.0
189
5.0
5.0
5.0
1700
1260
10.2-70.0
5.0
189
480
Final
Concen-
tration
(mg/D
0.25
0.1
0.075
0.10
0.075
1.6
0.04
0.60
0.60
0.2-3.6
0.01-0.03
0.1
0.01-0.05
Removal
(%)
95.0
99.9
98.5
98.0
98.5
— _
99+
99+
82-99+
99+
99.9
99.9
References
41
5
41
41
41
10
10
49
49
10
9,10
8
5
Filter
108
-------
TKBLE 8-6. OPTIONS AND -
MERCURY II BEMDVKL
Treatment Technology pH
Sulfide
Sulfide 10.0
Sulf ide/Filter 5 . 5
Sulf ide/Filter 4 . 0
Sulf ide/Filter 5.8-8.0
Ferrite coprecipitation/
Filter
Activated Carbon
Activated Carbon/Mum
Activated Carbon -
Initial
Concen-
tration
fog/1)
0.3-50.0
10.0
16.0
36.0
0.3-6.0
,6.0-7.4
0.01-0.05
0.02-0.03
0.06-0.09
Final Removal
Concen- (%)
tration
fog/l)
0.01-0.12
1.8 96.4
0.04 99
0.06 99.8
0.01-0.125 87-99.2
0.001-0.005 ,99.9
<0.0005
0.009
0.006
Reference!
9,10
50
50
50
50
5
9,10
46
50
109
-------
TABLE 8-7. WASTE WATER TREATMENT OPTIONS MID PERFORMANCE DATA SUMMARY -
NICKEL REMOVAL
Treatment Technology pH Initial
Concen-
tration
(mg/1)
Line 8.5-9.0 75
Lime (260 mg/1) /Filter 10.0 5.0
Line (600 mg/1) /Filter 11.5 5.0
Caustic Soda/Filter 11.0
Iferrous sulfide (Sulfex) 8.5-9.0 75
Herrite coprecipitation - 1000
Final
Concen-
tration
(mg/1)
1.5
0.3
0.15
0.3
0.05
0.20
Rsnoval References
(%)
98 8
94 41
97 41
49
99.9 8,11
99.9 5
110
-------
TABLE 8-8. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
SILVER REMOVAL
Treatment Technology pH
Sodium hydroxide 9.0
Ferric sulfate (30 mg/1) 6-9
Lime Softening 9.0-11.5
Chloride precipitation -
Initial
Concen-
tration
(mg/1)
54
0.15
0.15
105-250
Filial
Concen-
tration
(mg/1)
15
0.03-0.04
0.01-0.03
1.0-3.5
Removal
(%)
72
72-83
80-93
97+
References
13
46
46
9,10
(alkaline chlorination
in the presence of
cyanide)
Ferric chloride/Filter
Sulfide precipitation
6.2
5-11
0.5
0.04 98.2 40
- very high 9,10
111
-------
T&BEE 8-9. WASTE WATER TREATMENT OPTIONS AM) PERPOMSNCE DATA -
SEESJIDM MED THALLIUM REMOVAL
Treatment Technology
Selenium
Ferric chloride/Filter
Ferric chloride/Filter
Mum/Filter
Ferric sulfate
Ferric sulfate
IdirMFilter
Line/Filter
Thallium
Lime/Filter
Ferric chloride/Filter
MiWFilter
pH
6.2
6.2
6.4
5.5
7.0
11.5
11.5
11.5
6.2
6.4
Initial
Concen-
tration
(mg/D
0.1
0.05
0.5
0.10
0.10
0.5
0.06
0.5
0.6
0.6
Final
Concen-
tration
(mg/1)
0.03
0.01
0.26
0.02
0.03
0.3
0.04
0.2
0.4
0.4
Removal
75
80
48
82
75
35
38
60
30
31
Beferences
40
40
40
51
51
40
40
40
40
40
112
-------
8-10. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA. SUMMARY -
ZINC EEMOVSL
Treatment Technology
Line/Filter
Lime (260 mg/1)
Lime (260 mg/1) /Filter
Lime (600 rag/1)
Lime (600 mg/1) /Filter
Lime/Filter
Sodium hydroxide
Sulfide
Ferrous sulfide (Sulfex)
Ferrite ODprecipitation
pH
8.5-9.0
10.0
10.0
11.5
11.5
-
9.0
_
8.5-9.0
_
Initial
Concen-
tration
(mj/D
3.6
5.0
5.0
5.0
5.0
16
33
42
3.6
18
Final
Concen-
tration
(rag/1)
0.25
0.85
0.80
0.35
1.2
0.02-0.23
1.0
1.2
0.02
0.02
Removal
93
83
84
93
77
-
97
97
99+
99+
References
8
41
41
41
41
5
13
5
8,11
5
113
-------
The third step involves the consideration .of treatment
system variables under full-scale operating conditions in
industrial situations where the design objective would be the
simultaneous removal of several waste load constituents. Each
industry designs for maximum removal and/or recovery of the
major process-related waste substances and utilizes an
appropriate technology which is both reliable and cost
effective. Optimum treatment conditions for the removal of a
particular pollutant, can rarely be achieved consistently and any
given set of conditions will be somewhat less than optimum for
most, if not all, of the treatable constituents. In any well-
operated production facility the normal variations in production
rates, raw material quality, the desired product mix in some
cases, and contact water use requirements may cause severe
hydraulic and pollutant load input excursions which at best can
be moderated by effective equalization in the treatment system.
This is considerably less of a problem in batch treatment than
with a continuously operating system. The latter requires
continuous feedback monitoring for pH control and chemical
dosag*e in order to maintain the effluent quality within
acceptable limits for a number of parameters. Under these
conditions, the 30-day averages derived from the actual treated
effluent monitoring data (NPDES, etc.) would equate to what has
been identified in Step 3 above as the estimated 30-day
achievable performance using the same general treatment
technology.
A statistical evaluation of long-term monitoring data is
described below and the results are presented in Appendix A
where various derivative quantities such as long-term averages
and standard deviations are tabulated and the bases for
formulating the variability factors applicable to each
subcategory are explained in detail.
For each nonexcluded subcategory, a step by step
presentation of the logic used to develop effluent limitations
is given, based on performance estimates for 30-day average
concentrations for specific pollutants. When available, these
concentrations are based on industry monitoring data. When
long-term data are not available from industry, as is the case
with most toxic pollutants, achievable concentrations are based
on the treatability of these pollutants as discussed in Section
8 and summarized in Table 8-11.
Variability factors applied to these concentrations for the
development of monthly average and daily maximum limitations
are based on statistical analysis of long-term data as presented
below and in Appendix A. In many cases, due to the limited
amount of long-term data available, variability factors observed
in one subcategory are applied in other subcategories where
similar treatment technologies are practiced.
114
-------
8-11. 30-MY FOR THE
Antimony, Sb
Arsenic V
Beryllium, Be
Cadmium, Gd
H Copper, Cu
Chrcmium III,
Cr+3
Lead, Pb
Mercury II,
Hg
Nickel, Ni
Silver, Ag
Selenium, Se
Thallium, Tl
Zinc, Zn
Lime
Settling
0.8-1.5
0.5-1.0
0.1-0.5
0.1-0.5
0.5-1.0
0.1-0.5"
0.3-1.6
Lime
Filter
0.4-0.8
0.5-1.0
0.01-0.1
0.05-0.1
0.4-0.7
0.05-0.5
0.05-0.6
Final
Sulfide
Filter
0.05-0.1
0.01-0.1
0.05-0.5
0.05-0.4
Concentrations (mg/1)
Ferrite
Coprecip- Soda Ash Soda Ash Alum
itation Settling Filter
Filter
<0.05
<0.05
0.01
0.20 0.4-0.8 0.1-0.6
0.01-0.05 <0.01
0.2-1.5
0.4-0.8
0.2-1.0
0.2-1.0
0.5-1.5
0.1-0.5
0.2-0.4
0.1-0.5
0.1-0.5
0.4-1.2
0.05-0.5
0.05-0.2
0.02-1.2
0.2-0.5
0.02-0.5
(continued)
-------
TSBIE 8-11 continued
CFs
Arsenic V, As
Chromium VI,
Cr+6
Jfercury II,
Hg
Ferric Activated
Chloride Carbon
0.05-0.5 0.3
0.1
0.01
Final Concentrations (mg/1)
S02 Bisulfite Lirte/FeCl2
EMuction Ifeduction Filter
0.02-0.1
0.01-0.1 0.05-0.5
Mkaline
Chlori-
nation
Silver, Ag
Selenium, Se
Thallium, Tl
Cyanide (Free),
0%
0.05-0.1
0.05-0.1
0.7
0.1-0.5
-------
8.2 THE USE OF HISTORICAL POLLUTANT DATA
8.2.1 Determination of Limitation Guidelines Based Upon
Historical Performance
In cases where there has been long-term monitoring of the
pollution levels in the effluent stream discharged by a plant,
it is possible to assess in-plant treatment performance through
analysis of historical data that has been collected for this
purpose. The appropriateness of standards constructed from data
collected from a single plant performance is, of course,
dependent on the plant's current performance in relation to the
performance of other plants in the manufacturing subcategory.
As economically feasible alternative waste treatment
technologies become available, pollutant discharge guidelines
need to be reviewed and revised to'reflect these advances.
Statistical analysis of historical monitoring data is
required to assess a plant's ability to discharge within set
guidelines. To perform this analysis certain assumptions must
be made regarding the nature of applicable statistical or
probabilistic models, the constancy of the operation of the
treatment facility, and the quality of the monitoring methods.
The statistical analyses contained in this development
document belong to either of two principal types: those for
daily observations of pollutant concentrations, and the others
for 30-day average pollutant levels.
Tables in Appendix A provide a summary of traditional
descriptive measures, i.e., number of observations(No),
mimima(Min), arithmetic average(Avg), maxima(Max), and
coefficient of variation(CV). In addition, a descriptive
statistic, the variability factor, pertinent to the development
of performance standards for pollution monitoring, is included.
These tables, prepared for both daily measurements as well as
30-day averages, are statistical summaries derived from data
offered by industry in response to Section 308 Questionnaires,
Data .in these tables are representative of currently achieved
pollutant discharge performance levels in the several plants
presented.
Formulation of variability factors to be used in
determination of limitation guidelines based upon historical
performance was accomplished by employing standard statistical
analysis of the data resulting from long-term monitoring of
effluent stream discharges of plants in the inorganic chemical
manufacturing subcategory. In the following paragraphs are
presented details of the theory and derivation of these
statistical procedures, and of the resulting formulae which
117
-------
relate variability factors to estimated long-term parameter
averages, standard deviations, coefficients of variation, and
"Z-values" computed from the normal probability distribution.
These details are given both for the analysis applying to daily
maxima criterion and for that applying to 30-day averages.
The term "variability factor" is used in referring to the
multiple of the long-term average which is used in formulating
performance standards. This factor allows for variation in
pollution level measurements due to sampling error, measurement
error, fluctuations in the amount of the pollutant in raw
materials, and other process variations.
In the recording of actual data, as reported by industrial
point sources in their responses to 308 Questionnaires, certain
data values were entered as "less than" detectability limits.
In these cases, the sample of monitoring data has been
"censored" in the process of data recording since only the
threshold value has been retained (i.e., if a pollutant
concentration was reported as <0.050 mg/1, the values of 0.050
mg/1 was used). In the statistical analysis of monitoring data,
censored values were included with measured values in the
sample. This practice provides a reasonable approach, both for
assessing industry's capability to perform and environmental
concerns for valid pollutant limitations.
First, since censoring was done only for "less than"
bounds, any bias from their inclusion would cause a slight
increase in the long-term average, moderately affecting (in the
direction on leniency toward industry) the estimate of long-term
average pollution levels.
On the other hand, the use of censored values combined with
measured values tends to reduce the variability slightly (or in
the direction of less leniency toward industrial point sources).
For illustration, if the sample consisted solely of censored
values, the estimated long-term average might be shightly
overstated. Nevertheless, the point source ought have no
difficulty with the threshold or detectability limit as a
performance guideline, since none of the historical data
exceeded that limit.
8.2.2 Assumptions Concerning Daily Pollutant Level Measurements
In the formulation and calculation of the following
performance standards, individual sample measurements of
pollutant levels were assumed to follow the lognormal
distribution, a well known and generally accepted statistical
probability model used in pollution analyses. Under this
assumption the logarithms of these measurements follow a normal
probability model. It was also assumed that monitoring at a
118
-------
given plant was conducted responsibly and in such a way that
resulting measurements can be considered statistically
independent and amenable to standard statistical procedures. A
final assumption was that, treatment facilities and monitoring
techniques had remained substantially constant throughout the
monitoring period.
As an indication of the appropriateness of this assumption,
the following plot of the cumulative distribution of daily
pollution concentration logarithms on normal probability paper
is illustrated in Figure 8-1.
The linearity of the cumulative plot indicates the degree
to which actual monitoring data are in agreement with the
theoretical lognormal model for their distribution.
In addition, Figure 8-2, reproduced here from ,a report
prepared by industry for consideration by EPA, also demonstrates
the validity of the lognormal assumption for daily data.
In the analysis of daily data, the inherent variability of
measured pollutant levels in the effluent stream from inorganic
chemical manufacturing processes must be incorporated in
calculating upper limits for daily pollutant discharge levels.
Even well treated and controlled plants may experience some days
when an atypically high level of pollutant discharge is present
in their waste stream. Such high variations may be due to a
variety of factors, such as short-term maladjustments in
treatment facilities, variation in flow or pollutant load, or
changes in the influent stream. To allow for this variability,
performance standards must necessarily be set above the plant's
long-term average performance and occasional, infrequent
excessive discharges permitted. Since pollutant discharge is
often expressed in terms of average level, it is convenient to
describe standards of performance and allow variability in term
of multiples of this average. Such a method of computing
standards as functions of multiples of average level performance
is explained below. The ratio of the pollutant standard level
to the estimated long-term average is commonly called the
"variability factor".
This factor is especially useful with lognormally
distributed pollutant levels because its value is independent of
the long-term average, depending only upon the .day-to-day
variability of the process and the expected number of excessive
discharge periods. For a lognormal population, the variability
factor (P/A), the performance standard P, and the long-term
average A, are related by:
sLn(P/A) = S'(Z - S'/2)
119
-------
K
o
30.0
20.0
10.0
8.0
6.0
5.0
4.0
3.0
2.0
1.0
0.01 0.1 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99
99.9
PEKCENTAS!
Figure 8-1. Cumulative distribution of daily concentrations of mercury in treated
effluent from plant #251.
-------
H
0.30
0.20
0.10
0.08
0.06
0.04
0.03
0.02
0.01
0.01 0.1 .512 5 10 20 30 40 50 60 70 80 90 95 98 99
99.9
EEK3MTAGE
Figure 8-2. Cumulative distribution of daily concentrations of cyanide in treated
effluent from plant #765.
-------
where
1. "In" represents the natural logarithm (base e) of a
numerical quantity.
2. S1 is the estimated standard deviation of the
logarithms of pollutant level measurements. In the
calculations which follow, S1 is computed by the statistical
procedure known as the "method of moments".
3. Z is a factor derived from the standard normal
distribution. Z is chosen to give performance limitations which
provide a balance between appropriate consideration of day to
day variation in a properly operating plant and the necessity to
insure that a plant is functioning properly.
The value of Z used for determining performance standards
for daily measurements of pollutant concentration is chosen as
Z=2.33. This Z-value corresponds to the 99th percentile of the
lognormal distribution meaning that only 1 percent of the
pollutant observations taken from a plant with proper operation
of treatment facilities would be greater than the performance
standard, P. This percentile is equivalent to allowing a plant
in normal operation 3 to 4 exceedances per year.
Calculation of Variability Factors
As mentioned above, development of variability factors for
daily pollution level measurements was based on the assumption
that these data, (XI,X2,...Xn), follow a lognormal distribution.
When this distribution is not a precise model, lognormally based
procedures tend to somewhat overestimate variability and produce
liberal standards which act to the benefit of permittees.
Following this assumption, if Yi=ln(Xi), where ln(Xi)
represents the natural logarithm or log base e of the pollution
measurement, then the Yi; i=l,2,...,n are each normally
distributed. If A1 and S1 are the mean,and standard deviation
of Y=ln(X) respectively, then the probability is k percent that
an individual Y will not exceed A'+ZS1, where Z is the k-th
percentile of the standard normal distribution, e.g. Z=2.33 is
the 99-th percentile of the standard normal distribution. It
follows that A'4-ZS1 is the natural logarithm of the k-th
percentile of X and that the probability is k percent that X
will not exceed a performance standard P= exp(A'+ZS') . It is
also known that the average value of X is A= exp(A'+S'(S'/2)).
The variability factor VP, is obtained by dividing P by A,
hence,
VF = P/A = exp(S' (Z - S'/2 )) , and
ln(VF) = ln(P/A) = S'(Z - S'/2)
122
-------
To estimate the W for a particular set of monitoring data,
where the method of moments- is used, S1 is calculated as the
square root of ln(l_.0 + (CV) ) , where the sample1 coefficient of
variation, CV - S/X, is the ratio of sample standard deviation
to sample average.
Example Calculation of Variability Factors From Long-Term Data
Given the following descriptive statistics for a particular
parameter, as might be found for lead fmg/1)in Appendix A.
No Min Ayg Max CV
128 0.002 0.068 0.100 0.609
Calculate the estimated standard deviation of logarithms
(S1)2 = In (1.0 + 0.6092) = 0.315
S1 = 0.56
Then
ln(P/A) = 0.56(2.33 - 0.56/2) - 1.148
The variability Factor VF is,
VF = P/A = exp{1.148) = 3.15
The performance standard P;
P = A(VP) = A (P/A) = (0.068) (3.15) = 0.2*14
The statistical distributions relevant for the analysis of
daily data are shown in Figure 8-3.
The statistical interpretation of P, the performance
standard, is that one estimates that 99 percent (for the
selected Z=2.33 value corresponding to the 99-th percentile) of
the daily pollution level measurements will not exceed P. For
large data sets, P is roughly equivalent to an upper 99 percent
confidence bound for an individual daily measurement.
8.2.3 Assumptions Concerning 30-Day Average Pollutant Level
Observation
While individual pollution level measurements should be
assumed lognormally distributed, that assumption is not
appropriate when analyzing 30-day averages. These averages
generally are not distributed as lognormal quantities. However,
for averages of daily (lognormal) measurements, a statistical
123
-------
NOKMAL DISTRIBUTION
(MODEL DENSITY OF LOGARITHMS OF POLLUTION VALUES)
ln(P) = A1 + 2.33(S')
Y = ln(X) = Logarithm (mg/1)
A1
LOGNORMAL DISTRIBUTION
(MODEL DENSITY OF
POLLUTION VALUES)
X(mg/l)
_ P(Performance Standard)
"—A (Long Term Arithmetic'Average)
SAMPLE DISTRIBUTION OF N MEASUREMENTS
(LONG TERM MONITORING DATA)
X(mg/l)
X" (Sample Average)
Note: (a) S1 is estimated as (S1)- = Cln(l + CV2)J
Figure 8-3.
t? 2
X^ ZX/N
Statistical distribution for daily pollution measurements.
124
-------
principle, the "Central Limit Theorem", provides the basis for
using the normal probability model. Therefore, the methods used
in computing historical performance characteristics for 30-day
averages differ from those used for daily samples. In this
case, the sample coefficient of variation is the primary
determinant of the variability factor, and there is no need to
resort to logarithmic transformation. Examples of the
appropriateness of this assumption is the cumulative
distribution of 30-day averages shown in Figures 8-4 and 8-5. A
straight line plot here on normal probability paper indicates
validity of this model.
Under these conditions, the 30-day average values (XI, X2,
Xm), for m months behave approximately as random data from
a normal distribution with mean A and standard deviation S"_.
Therefore, the probability is k percent that a monthly average X
will not exceed the performance standard P, where
P = A + Z(S")
The variability factor is
VF = P/A o 1.0 + Z(S"/A) and will be estimated by
VF = 1.0 + Z{CV)
Where
1. z is a factor derived form the standard normal
distribution. If one wishes a performance standard based upon
expecting 95 percent of monthly averages to be within
guidelines, then Z=1.64 should be used.
2. CV is the estimated coefficient of variation of the 30-
day averages and is computed by Sx/X, the ratio of standard
error of sample means to overall or grand average of monthly
.averages.
Calculation of Variability Factors
A sample calculation of 30-day average variability factor
is shown below. The descriptive statistical data 'is for zinc
(mg/1) from Appendix A.
No Min Avg Max CV
30 0.010 0.151 0.815 1.03
VP - 1 + Z(CV) = 1.0 -f 1.64(1.03) = 2.69
P - A(VF) * (0.151) (2.69) - 0.406
125
-------
-» 3.00
!
2.00
1.00
0.01 0.1 0.5 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 99.9
EEFCENTAGE
Figure 8-4. Cumulative distribution of 30-day averages of total cyanide in treated
effluent from plant #782.
-------
t-I
H
H
to
-J
fl
700 --FFOT
600
400
200
100
. .1
^
Ml
:•:,
1
an
:::
1
0.01 0.1 125 10 20 30 40 50 60 70 80 90 95 98 99
99.9
Figure 8-5. Cutmilative distribution of 30-day averages of ammonia in treated
effluent from plant #782.
-------
Given the previous descriptive statistics for a particular
sample, one obtains the performance standard P, by multiplying
the mean of the 30-day averages in the data set by VF, An
appropriate statistical interpretation' is that, for the
selected value of Z=1.64 corresponding to the 95th percentile of
a normal distribution, one estimates that 95 percent of the 30-
day average pollution level measurements will not exceed p.
Figure 8-6 shows the relationship between the normal
probability model and frequency distribution of set of 30-day
averages.
128
-------
NORMAL DISTRIBUTION
DENSITY OF 30-DAY AVERAGE POLLUTION MEASUREMENTS)
XCmg/1)
— P (Performance Standard)
_ A (Long Term Average)
SAMPLE DISTRIBUTION OF M MONTHLY AVERAGES
(LONG TEfiM MDNZDORING DATA)
^>* M-i
Min
I
X (rag/1)
X (Average of 30-Day Averages)
Note: (a) P/A = 1+1.64(CV)
07 = S^/X
{S-)2=(Z (
f=Z X/M ,
Figure 8-6. Statistical distributions for 30-day average pollution measurements.
129
-------
-------
SECTION 9
TREATMENT TECHNOLOGY APPLICATIONS
FOR TOXIC POLLUTANT REMOVAL
9.1 SELECTION OF POLLUTANTS TO BE CONTROLLED
In order to determine which toxic pollutants, if any, may
require effluent limitations, the pollutants observed in each
subcategory were evaluated with regard to their fereatability and
potential environmental significance on the basis of the raw
waste concentrations and mass loadings found during screening
and verification. In an attempt to prioritize the need for
regulation the toxic metals were divided into two groups:
Group 1 - Those metals which appear at concentration levels
that are readily treatable using available
technology and which have environmentally
significant mass emission rates. *
Group 2 - Potentially significant metals observed in the
subcategory. These include toxic metals which
exist at concentrations below the minimum
treatability limit and above the minimum
detection level.
Table 9-1 presents the significant toxic pollutant metals
found in each group. In general, those metals occurring in the
first group are of prime concern and may require regulation,
while those occurring in the second group are of somewhat less
concern and are not expected to require regulation.
9.2 APPLICATION OF ADVANCE LEVEL TREATMENT AND CONTROL
ALTERNATIVES
i
9.2.1 General Design Objectives
Beginning with Section 11 of this document, the selection
and application of toxic pollutant treatment and control
131
-------
TABLE 9-1. PKKDRITIZATICN CP TCKTC METALS root® IB EACH SOBCSTEGOIS
SUBCHZEGQBY
Gilorine-diaphragm cell
Chlorine-mercury cell
Hydrof luoric Sold
Group 1 (1)
Chromium
Copper
Lead
Nickel
Zinc
Arsenic
Cadmium
Copper
Xead
Mercury
Nickel
Silver
Zinc
antimony
Chromium
copper
Lead
Nickel
Zinc
Group 2 (2)
Jteitiitcny
Srsenia
Cadmium
Mercury
Selenium
Kiallium
antimany
Chranium
Biallium
Arsenic
Cadmium
tfercury
Selenium
ffliallium
Titanium Dioxide - Chromium Lead
ChlorMe Process Nickel
Zinc
Titanium Dioxide - flntimsny Selenium
Sulfate Process Arsenic Thallium
and Cadmium
Chloride llnenite Process Chromium
Copper
Lead
Nickel
Zinc
(1) Group 1 - dominant raw waste pollutants selected as control parameters
for the proposed effluent limitations.
(2). Group 2 - secondary raw waste pollutants found less frequently and at
lower concentrations. These pollutants have not been selected
as control parameters but are expected to receive adequate
treatment as a result of the proposed effluent limitations on
the Group 1 pollutants.
(continued)
132
-------
9-1 - continued.
SUBCATEGORY Group 1 Group 2
Aluminum Fluoride Copper Arsenic
Nickel Cadmium
Chromium
Jfercury
Zinc
Chrome Pigments Jtatuiwiy Cyanide
Cadmium Mercury
Chronium
Cyanide
lead
Nickel
Zinc
%drogen Cyanide Cyanide
Sodium Dichromate Chrcmium Copper
Nichel Selenium
Zinc Silver
Copper Sulfate antimony
arsenic
Cadmium
Chromium
Copper
laad
Nictel
Selenium
Zinc
Nickel Sulfate Antimony Arsenic
Chromium Cadmium
Copper Mercury
Lead Selenium
Nickel "Thallium
Zinc
Sodium Bisulfite Chromium Antimony
Copper Cadmium
Lead Mercury
Nickel
Zinc
Sodium Bydrosulfxfce Chromium Copper
formate Process lead Pentachlorophenol
Hickel Phenol
Zinc
133
-------
technology for model plant systems for each of the subcategories
proposed for regulation are described. Several levels of
treatment are proposed. Level 1 represents existing BPT
treatment systems and the advanced levels (Level 2, 3, etc.) are
the selected technologies for step-wise improvements in toxic
pollutant removal over that achieved by the BPT system. Flow
diagrams show BPT components as a starting point for advanced
level treatment additions and incremental cost estimates.
For both existing and new sources, the advanced level
technology options are selected as candidates for BAT with toxic
pollutant removal as the primary objective. Although the
advanced level systems chosen also give improved performance
over the Level 1 (BPT) systems for the removal of conventional
and nonconventional pollutants, this is regarded as a secondary
design objective.
9.2.2, Pretreatment .Technology
Since untreated heavy metal ions will usually pass through the
treatment provided in a typical POTW, or will be precipitated
with the POTW solid residue, pretreatment of wastes containing
significant amounts of heavy metals is necessary. As a general
rule, alkaline precipitation, followed by settling and removal
of the solids will suffice. In certain subcategories, such as
the chlorine industry, specific treatment will be required for
highly critical constituents (mercury, lead, chlorinated
organics and asbestos). Normally the Level 2 model treatment
processes shown in the following subsections will be appropriate
for pretreatment prior to discharge to a POTW.
9.2.3 New Source Performance Standards
New Source Performance Standards are at least equal to BAT.
In a few cases where new plants have the opportunity to design
systems for better toxic removal performance without expensive
retrofitting the higher technology systems have been used as a
basis.
9.3 ESTIMATED ACHIEVABLE PERFORMANCE CHARACTERISTICS FOR
ADVANCED LEVEL APPLICATIONS
Advanced level control and treatment alternatives for
reduction of pollutant discharges and their applicability to
each subcategory are presented in the sections dealing with
individual products. With few exceptions, these alternatives
were selected specifically for removal of priority pollutants
and were des-igned for end-of-pipe treatment.
134
-------
Treatment technologies practiced outside the industry are
recommended when appropriate and, in most cases, apply to the
removal of toxic pollutant metals. The estimated 30-day average
treatability levels (Sections 8, Table 8-11), long-term data
parameters, and the screening and verification results are all
utilized in the development of estimated performance
characteristics for the recommended treatment applications in
each subcategory.
9.3.1 Advanced Ley_el Removal of BPT Pollutants
Performance estimates for these systems, when possible, were
based on effluent quality achieved at plants currently
practicing these technologies. However, in most cases, the
advanced levels are not currently being practiced within the
specific subcategory of concern, and performance information
from other appropriate sources is necessarily utilized.
When established waste water treatment practices, such as
clarification or filtration, form a part of advanced treatment
alternatives, the specified achievable effluent quality has been
based on concentrations accepted as achievable through proper
design and control. The prime example of this is suspended
solids reduction by filtration.
9.3.2 Advanced Level Removal of Toxic Pollutants
Performance estimates for toxic pollutants were also based,
when possible, on effluent quality achieved at plants currently
practicing these technologies. However, in most subcategories,
toxic pollutant analyses are not conducted unless a specific
pollutant is regulated and requires monitoring. Where transfer
of technology is applied as a treatment alternative, performance
estimates for toxic pollutant removals were based on the
demonstrated performances in other industries while
incorporating allowances for specific differences in process
waste characteristics and operating conditions. Statistically
derived long-term monitoring data parameters were described in
Section 8 and are compiled in tabular form in Appendix A. The
screening and verification data are used to supplement the
available long-terra data applied to each subcategory. A
judgment is made whether the screening and verification data
represent a well-performing system or one which is not
performing at its technological potential. For a
well-performing system, the data are regarded as representative
of 30-day averages and are compared with the estimated
treatability ranges from Table 8-11, as well as the 30-day
averages developed from thte long-term data. In this manner, the
performance estimates for each pollutant, at each treatment
level for the nonexcluded subcategories, are developed and
135
-------
presented in tabular summaries. By starting with the estimated
achievable 30-day averages, the specific variability factor
ratio derived for each pollutant is used to estimate the daily
maximum values.
The model plant waste flow per unit of production is then
taken to calculate the estimated mass emission values of the 30-
day average and daily maximum limits for each pollutant to be
controlled.
9.4 POLLUTION CONTROL PARAMETERS TO BE REGULATED
9.4.1 Conventional Pollutants
Waste water quality parameters which are identified as
conventional pollutants include the following;
PH
Total Suspended Solids (TSS)
Biochemical Oxygen Demand, 5-Day (BOD-5)
Fecal Coliform
Oil and Grease
Only the first two parameters (pH and TSS) in this group
have been selected for regulation in the Inorganic Chemicals
Manufacturing Point Source Category. For direct dischargers,
the pH range of 6 to 9 has been established as the general
control limitation and the permissible frequency and duration of
excursions beyond this range is to be specified in individual
plant discharge permits. The limitations on TSS are specified
for both BPCTCA and BATEA-based regulations, the former being
largely a function of industry performance and the latter
stemming from treatability estimates with the appropriate
technologies.
9.4.2 Nonconventional Pollutants
The waste water quality parameters classified as
nonconventional pollutants include the nontoxic metals tsuch as
Al, B, Ba, and Fe along with chemical oxygen demand (COD) , total
residual chlorine, fluoride, ammonia, nitrate, and "phenols,"
etc. Of these, only Fe, COD, total residual chlorine, fluoride,
and ammonia are considered for regulation in the inorganic
chemicals industry. Due to its toxicity, chlorine would be
controlled in direct discharges, but would be excluded from
control in pretreatment regulations. A similar argument is made
for the control of ammonia. However, since many POTW1s are only
capable of about 20 percent ammonia removal, both direct
discharge and pretreatment regulations would specify NH3
limitations. Similarly, the type of COD found in inorganic
136
-------
chemical industry discharges is not amenable to biochemical
oxidation in a POTW. In addition, compounds which contribute to
the COD are likely to create odor and corrosion problems in
sewer systems. Therefore, its control would also be retained in
pretreatment regulations. Fluoride control is also required for
both direct and indirect discharges largely because the most
practical technology for fluoride removal (precipitation as
CaF2) must be applied to relatively concentrated waste water
sources. This treatment method achieves removal levels which at
best are still unacceptable for direct municipal or agricultural
water uses. POTWs are not effective for fluoride removal and
unless sufficient dilution occurs prior to the reuse of the
water, special techniques (e.g., adsorption on activated
alumina) would have to be applied for further fluoride removal.
9.4.3 Toxic Pollutants
The toxic pollutants found at significant levels during
screening and verification are listed by subcategory in Table
9-1. Out of these, toxic pollutant control parameters were
selected largely on the basis of treatability. Since several
toxic pollutants may be controlled by a common treatment
technology, it is possible to select one or more control
parameters which will act as a surrogate for others exhibiting
the same treatability characteristics. Treatment system
operating conditions would normally be optimized for the removal
of the specified control parameters which would be monitored on
a regular basis. The other toxic pollutants would be monitored
much less frequently as a periodic check of the effectiveness of
surrogate control.
The - following toxic pollutants have been designated as
control parameters in this point source category:
Cadmium
Chromium (Total)
- Copper
Cyanide (amenable to chlorination)
Lead
Mercury
Nickel
Selenium
Zinc
The specific control parameters selected for each
subcategory are presented in the tables entitled "Control
Parameter Limitations" in the sections of this report dealing
with the individual industries. Some general comments about
them are given here.
137
-------
The most common technology applied in industry for the
removal of chromium from waste waters involves a reduction step,
whereby Cr (VI) in solution is converted to the less toxic Cr
(III) form which can then be removed by alkaline precipitation.
The efficiency of this treatment depends upon the presence of
excess reducing agent and pH control to,drive the reduction step
to completion. When treated effluent samples are taken to
monitor residual Cr (VI) and total chromium levels, the
analytical results for Cr (VI) are subject to several factors
which adversely affect the accuracy and reproducibility of the
diphenylcarbazide (DPC) colororaetric method. The problem is not
so much one of analytical interferences with the Cr (VI) ~ DPC
color development, but rather the actual changes in Cr (VI)
concentration that can take place during sampling, sample
preservation and storage, and analysis. The major cause of such
changes is the presence of excess reducing agent in the treated
effluent. This tends to give false low readings for Cr (VI)
although in some cases the opposite may occur as a result of
sample preservation and storage under acidic oxidizing
conditions.
Thus, in view of the questionable reliability of the
presently accepted Cr (VI) monitoring procedure, total chromium,
Cr (T) , is recommended as the control parameter to be used in
the inorganic chemicals industry. The adequacy of Cr (T) as a
control parameter is predicated on its effectiveness as a
surrogate for Cr (VI) control. Since the concentration of Cr
(T) represents the summation of all forms of chromium normally
found in solution or suspension including Cr (VI), the final
concentration of Cr (T) in a treated effluent is dependent on
the effectiveness of both the reduction and the alkaline
precipitation steps. In this way, the use of Cr (T) as the
control parameter assures that adequate removal of Cr (VI) is
being achieved as a direct consequence of the treatment
technology required.
Special consideration is given to the control of copper
which may enter a POTW. At high enough concentrations, copper
may impose toxicity effects on the microorganisms in a POTW and
may accumulate in municipal sludges rendering them unusable for
certain land applications. Thus, copper may be designated as a
control parameter for pretreatment even though it may not be so
designated for direct discharges.
138
-------
SECTION 10
COST OF TREATMENT AND CONTROL SYSTEMS
10.1 INTRODUCTION
10.1.1 Purpose of Cost Data
More complex treatment methods and higher levels of
pollutant removal are reflected in increased costs of
equipment, energy, labor and chemicals. At some point, the
increasing costs of treatment will outweigh the benefits of
such treatment. Therefore, it is important that for each
subcategory the Agency know the base cost and the incremental
costs of each level of treatment which it might prescribe.
These "options" of internal costs, which are the industry's
annual costs of providing the necessary waste treatment, will
result in related increases in product costs, which are termed
external costs. Thus annual costs of waste treatment are
expressed in terms of dollars per unit of annual production of
the principal product.
Because plant visits revealed very few treatment plants
serving a single product manufacturing line, it was not feasible
to seek actual waste treatment facilities which could serve
as real models for estimating purposes. Accordingly*, the cost
data were taken from similar construction projects by tthe
contractor, and from unit process equipment costs assembled f'rom
vendors and other commercial sources. Because the model costs
apply to a wide range of climate, material sources and labor
conditions, they should be considered as preliminary estimates
within plus or minus 15 to 25 percent.
Actual costs incurred by individual plants may be more or
less than the presented model plant costs. The major causes of
variability are:
1. Waste water treatment combined with the treatment of
other product effluents.
2. Site dependent conditions, as reflected in piping
lengths, climate, land availability, water and power
139
-------
supply and the location of the points of final
discharge and solids disposal,
3. Material (reagent) costs, due to variation in
availability and distance from the source.
The construction costs are based on the Engineering News
Record Construction Index for July 1978 (ENRCI=2800), and other
costs are expressed in mid-1978 dollars.
10.1.2 general Approach
Since few single product waste treatment plants were
available for detailed study, the costs presented in this
section are based on model plants which closely resemble the
types and capacities of waste treatment facilities needed for
each separate product subcategory. The model plant selections
are based on review of Section 308 Questionnaire responses,
plant visits, development documents, contacts with the
industries to verify treatment practices and to obtain data on
size, waste water flow, and solid waste disposal systems.
Thus, each model is synthesized from actual data as a typical
plant in its subcategory with a level of waste treatment
equivalent to BPT. Variations in treatment plant capacity
are accounted for by selecting sets of models which represent
the range of existing production plant capacities in the
subcategory; large, medium and small. Thus the model plants are
not set up as exemplary plants, but as typical plants of
adequate design which represent the range of plants and
treatment facilities found in the subcategory.
10.1.3 Cost References and Rationale
Cost information contained in this report was obtained
directly from industry, engineering firms, equipment suppliers
and current experience of the contractor. Whenever possible,
costs are based on actual industrial installations or
engineering estimates for projected facilities as supplied by
industries consulted during the study. In the absence of such
information, cost estimates have been developed from either
current costs for similar waste treatment installations at
plants making other inorganic chemicals or from general cost
estimates for specific treatment technologies.
Treatment costs are based on model production plant
characteristics which determine the treatment processes
selected for each operation. Under set effluent limitations,
treatment costs are primarily functions of the pollutant load
(i.e., kg/kkg of product), waste water flow rate (i.e., cubic
meters/day). Available data indicate that both pollutant loads
140
-------
and flow rates can vary significantly among plants manufacturing
the same product.
10.1.4 Definition of Levels of Treatment and Control
Cost Development
For the purpose of establishing the base level treatment
costs, each industry is assumed to be practicing Best
Practicable Control Technology Currently Available (BPT), which
the EPA Effluent Limitations Guidelines required by 1977 for
certain pollutants (conventional and nonconventional, as well
as some of the toxic pollutants) specified for each subcategory.
The investment costs and annual costs of such BPT systems are
shown in this report as the base level or Level 1. This level of
treatment may also provide incidental removal of additional
toxic pollutants not previously specified in the regulations.
The advanced treatment levels {Level 2, Level 3, etc.) are
aimed primarily at reduction of toxic pollutants to levels
considered acceptable for July lf 1984, performance, utilizing
Best Available Technology Economically Achievable (BAT) at
incremental investment and annual costs beyond those shown for
Level 1. For example, for Level 3 treatment, the incremental
cost as given in the table is directly added to base or 1st Level
cost to obtain the total cost of the treatment system. The
addition of the Level 2 incremental cost is not required to
obtain the Level 3 total. The wa'ste water treatment flow
diagrams for Levels 2, 3, etc., as given in this report,
include the flow diagram for base or Level 1 of treatment.
10.1.5 Treatment and Disposal Rationale Applied to
Cost Development ~*~
The following assumptions are employed in the cost
development:
1. Noncontact cooling water generally is excluded from
treatment (and treatment costs) provided that no
pollutants are introduced.
•
2. Water treatment, cooling tower and boiler blowdown
discharges are not considered .process waste water
unless such flows contain significant amounts of
pollutants.
3. Sanitary sewage flow is excluded.
4. The plants are assumed to operate 24-hours per day, 350
days a year, except where otherwise noted.
5. Manufacturing plants are assumed to be single product
plants.
141
-------
6. The inorganic chemical industry extensively uses in-
plant control techniques such as in-process abatement
measures, housekeeping practices, and recycling of
process waste waters to recover, valuable materials or
use these materials as feed for other by-products.
Segregation of uncontaminated cooling and other waters
prior to treatment and/or disposal, and other similar
measures can contribute to waste load reduction. All
such costs have not been included in the cost
estimates.
7. Excluded from the estimates are any costs associated
with permits, reports or hearings required by
regulatory agencies.
10.1.6 Expression of Costs
Investment costs for Level 1 treatment systems are
expressed in mid-1978 dollars to construct base level facilities
for each single product manufacturing subcategory at various
production rates.
Similarly, operation, maintenance and amortization of the
investment are expressed as base level annual costs for Level
1 and as incremental annual costs for Level 2 and above. Where
a single product plant produces more than one waste stream
requiring treatment, the respective investment and annual costs
are the combined costs of all treatment.
Total annual costs per metric ton of product are shown in
the summaries for each product subcategory.
Direct Investment Costs for Land and Facilities
Types of direct investment costs for waste treatment
facilities and criteria for estimating major components of the
model plants are contained in the following subsections:
Construction costs - Construction costs include site
preparation, grading, enclosures, buildings, • foundations,
earthwork, roads, paving and concrete.
The costs of constructing lagoons can vary widely,
depending on local topographic and soil conditions. The
required areas of lagoons and settling ponds and their
consequent costs are developed as a function of volume
(capacity). It is assumed that reasonably level sites are
available, consisting of sandy loam with high clay content
and no large rocks or rock formations. It is assumed that
two rectangular lagoons are furnished in parallel, with one
common dike to permit alternate dewatering for sludge removal by
142
-------
the clamshell method. Using balanced cuts and fills, earth
dikes with 2:1 slopes provide liquid depths from three to
five meters. Earth moving costs are significantly affected
by site conditions and quantities. To express these
variations for a range of sizes at three depths, the cost of
clearing, excavation, dewatering, compaction, finish grading,
riprap and associated indirect expenses for earthen lagoons
were plotted against liquid volume. Piping, valving and dike
roads not included are added separately in the cost summaries.
Lagoons are unlined unless the contents are highly pollutional
or, acidic. The liner material employed for impervious lagoons
is Hypalon.. The installed cost of the liner is $11.00 per
square meter ($9.20 per square yard), which includes the
trenching and backfilling necessary for anchoring the liner.
In some subcategories, clay lining has been used in place of
Hypalon at a cost of $5.40 per square meter ($0.50 per square
foot) .
Costs of buildings may vary from $25.00 to $45.00 per
square foot. For the purpose of this study, building cost is
estimated at $377.00 per square meter ($35.00 per square foot).
Concrete construction for. miscellaneous work varies from
$260.00 to $785.00 per cubic meter ($200.00 to $600.00 per
cubic yard). For foundations and flat slabs, concrete has
been estimated at $395.00 per cubic meter ($300.00 per cubic
yard) in place. Asphalt paving which has been used on lagoon
dikes and for miscellaneous roads, is installed at a cost of
$9.70 per square meter ($0.90 per square foot), A width of
three meters is generally assumed.
Equipment costs - Depending upon the method of treatment,
equipment for waste water treatment consists of a combination
of items such as pumps, aerators, chemical feed systems,
agitators, flocculant feed systems, tanks, clarifiers,
thickeners, filters, etc. Cost tables for these items were
developed from vendors' quotations on a range of sizes,
capacities and motor horsepowers. Except for large size tanks
and chemical storage bins, the cost represents packaged,
factory-assembled units. Mechanical components are generally
skid mounted, prepiped and prewired; and include associated
pumps, meters- and instrumentation. Critical equipment is
assumed to be installed in a weatherproof structure. Chemical
storage, feeders and feedback equipment include such items as
probes, instruments, controls, transmitters, valves, dust
filters and accessories. Bulk " chemical storage bins are
designed to hold a standard bulk truck . load, plus five days
needs, between ordering and delivery. Critical pumps are
furnished in duplicate and when clarifiers are used, the flow is
split between two units, permitting one to be bypassed for
repairs. Single units are used for small flows, batch
treatment and intermittent service.
143
-------
Installation cost - Installation is defined to include
all services, activities, and miscellaneous material necessary
to implement the described waste water treatment and control
systems, including piping, fittings, and electrical work. Many
factors can impact the cost of installing equipment modules.
These include wage rates, manpower availability, whether the
job is performed by outside contractors or regular employees,
new construction versus modification of existing systems, and
site-dependent conditions (e.g., the availability of sufficient
electrical service). In these estimates, installation costs
were chosen for each application, based upon average site
conditions and considering the complexity of the* system being
installed. An appropriate cost is allowed for interconnecting
piping, power circuits and controls.
Monitoring equipment - In this report, it is assumed that
monitoring equipment will be installed at the treated effluent
discharge point. It will consist of an indicating, integrating
and recording type flow meter, pH meter with sensor and
recorder, alarms and controls and an automatic sampler.
Land - Land availability and cost of land can vary
significantly, depending upon geographical location, degree of
urbanization and the nature of adjacent development. Land for
waste treatment, and in some cases for inert solids disposal,
is assumed to be contiguous with the production plant site
and reasonably convenient to a waterway which can receive
permitted discharges of waste water. Where inert solids are
retained at the plant site, enough land is included in the base
level' model plant investment cost to accept residual solids for
a normal operating period of ten years at the same production
rate for which the plant is sized. For the purpose of this
report, land for lagoons, treatment facilities and on-site
residual waste disposal is valued at $30,000 per hectare
($12,000 per acre).
Investment costs for supporting services - Engineering
design and inspection are typical services necessary to bring a
project from a concept to an operating system. Such services
broadly include laboratory and pilot plant work to establish
design parameters, site surveys to fix elevations and plant
layout, foundation and groundwater investigations, and
operating instructions; in addition to design plans,
specifications and inspection during construction. These costs,
which vary with job conditions, are often estimated as
percentages of construction cost, with typical ranges as
follows:
Preliminary survey and construction surveying 1 to 2%
Soils and groundwater investigation 1 to 2%
144
-------
Laboratory and pilot process work 2 to 4%
Engineering design and specifications 7 to 12%
Inspection during construction 2 to 3%
Operation and maintenance manual 1 to 2%
Prom these totals of 14 percent to 25 percent, a mid-
value of 20 percent of in-place construction (installed
equipment and construction) costs has been used in this study to
represent the engineering and design costs applied to model
plant cost estimates.
The contractor's fee and contingency, usually expressed as
a percentage of in-place construction costs, includes such
general items as temporary utilities, small tools, dewatering,
field office overhead and administrative expense. The
contractor is entitled to a reasonable profit on his activities
and to the cost of interest on capital tied up during
construction. Although not all of the above costs will be
incurred on every job, an additional 20 percent of the in-place
construction costs has been used to cover related costs broadly
described as contractor's fees, incidentals, overhead and
cont i ng enc i es.
Operation and Maintenance Costs
Annual operation and maintenace costs are described and
calculated as follows:
Labor and supervis ion costs - Plant operations are assumed
to be conducted 24-hours per day 350 days per year, with
attendance for only part of each working day. For batch waste
water treatments systems adjustment are made for the number of
working days in a year. Personnel costs are based on an hourly
rate of $20.00. This includes fringe benefits and an allocated
portion of costs for management, administration and supervision.
Personnel are assigned for specific activities as required
by the complexity of the system, usually 4 to 12 hours per day.
Energy costs - Energy (electricity) costs are based on the
cost of $306.00 per horsepower operating 24 hours per day and
350 days per year. For batch processes, appropriate adjustments
are made to suit the production schedule. The cost per
horsepower year is computed as follows:
Cy= 1.1 (0.7457HP x Hr x Ckw)/('E x P) 1)
145
-------
where
Cy = Cost per year
HP * Total horsepower rating of motor (1 hp = 0,7457
kw)
E = Efficiency factor (0.9)
P = Power factor (1.00)
Hr = Annual operating hours (350 x 24 = 8400)
Ckw = Cost per kilowatt-hour of electricity ($0.040)
Note: The 1.1 factor in equation (1) represents allowance
for incidental energy used such as lighting, etc.,
It is assumed that no other forms of energy are used in the
waste treatment system.
Chemicals - Prices for the chemicals were obtained from
vendors and the Chemical Marketing Reporter. Unit costs of
common chemicals delivered to the plant site are based on
commercial grade of the strengths or active ingredient
percentages as follows;
Hydrated Lime (Calcium Hyroxide) Bulk $ 80/metric ton
Bag $ 85/metric ton
Quicklime Bulk $ 70/metric ton
Ground Limestone $ 13.20/metric ton
Soda Ash (58% Bulk) $ 85/metric ton
Caustic Soda (58% NaOH) $200/metric ton
Sodium Sulfide (60-62%) $435/metric ton
Sulfuric Acid $ 75/metric ton
Hydrochloric Acid (32%) $ 70/metric ton
Aluminum Sulfate (56% Alumina) $250/metric ton
Flocculant (Polymer) $2.00/kg
Sulfur Dioxide (Ton Containers) $335/metric ton
146
-------
Chlorine (ton Containers) $220/metric ton
Sodium Bisulfide (72-74%) $385/metric ton
Ferrous Sulfate $ 70/metric ton
Diatomaceous Earth $ 0.30/kg
Activated Carbon $ 2.00/kg
Maintenance - The annual cost of maintenance is estimated
as 10 percent of the investment cost, exluding land.
Taxes and insurance - An annual provision of three percent
of the total investment cost has been included for taxes and
insurance.
Residual waste disposal - Sludge disposal costs can vary
widely. Chief cost determinants include the amount apd type of
waste, and the choice of either on-site disposal or contract
hauling which depends on the size of the disposal operation and
transport distances. Off-site hauling and disposal costs are
taken as $13.00 per cubic meter ($10.00 per cubic yard) for
bulk hauling, with appropriate increases for small quantities in
steel containers. For on-site disposal from lagoons, a
clamshell at $600.00 and front end loader at $300.00 per
disposal day are used. For very large sludge quantities,
lower unit costs have been assumed. The computed sludge
quantities are spread on land valued at $12,000 per acre.
Monitoring, analysis and " reporting - The manpower
requirements covered by the annual labor and supervision costs
include those activities associated with the operation and
maintenance of monitoring instruments, recorders, and automatic
samplers as well as the taking of periodic grab samples.
Additional costs for analytical laboratory services have been
estimated for each subcategory assuming that sampling takes
place three times a week at the point of discharge and that
an analytical cost of $20.00 per constituent is incurred.
Approximately 10 percent of the total analytical cost has been
added for quality control and water supply samples. Unless
otherwise stated, continuous discharge is assumed and the
analytical costs associated with compliance monitoring at the
BPT level are based on the determination of four constituents.
At the advanced (BAT) levels, the determination of six
constituents is assumed. A reporting cost of $1,500 per year
is added for clerical support. Monitoring costs for periodic
batch treatments are reduced in proportion to the number of
days per year when discharges occur.
147
-------
Amor t i zation
Annual depreciation and capital costs are computed as
follows:
CA = B%r(l+r)n /M3+r)n -1 (2)
where
CA = Annual cost
B = Initial amount invested excluding cost of land
r = Annual interest rate (assumed 10%)
n - Useful life in years
The multiplier for B in equation (2) is often referred to
as the capital recovery factor, and is 0.1627 for the assumed
overall useful life of 10 years. No residual or salvage value
is assumed.
Items Hot Included in Cost Estimates
In some subcategories, a portion of the waste water is
returned to process from an intermediate treatment step. In
these cases, the costs of return piping and pumping are
considered as water development and not as waste treatment.
Costs for subsequent treatment are based on the remaining
flow after diversion of the return-to-process flows.
Although specific plants may encounter extremes of climate,
flood hazard and availability of water, the costs of model
plants have been estimated for average conditions of
temperature, drainage and natural resources. It is assumed that
any necessary site drainage, roads, water development, security,
environmental studies and permit costs are already included
in production facilities costs. Therefore, the model costs are
only for facilities, supplies and services directly related to
the treatment and disposal of waterborne wastes, including land
needed for treatment and on-site sludge disposal. Air pollution
control equipment required by the Clean Air Act is not included,
Dust collectors normally associated with package
treatment, chemical transfer and feeding systems are included.
Raw wastes from various sources are assumed to be delivered to
the treatment facility at sufficient head to fill the influent
equalization basin, and final effluent is discharged by gravity.
Costs of pumps, pipes lines etc., necessary to deliver raw
waste water to the treatment plant or to deliver the treated
effluent to the point of discharge are not included in the cost
estimates.
148
-------
Since the treatment models are designed to serve single
product manufacturing plants, no emergency holding basins or
internal bypasses are provided. Any such necessary facilities
are more appropriately furnished as part of a combined waste
treatment system serving several product lines.
10.2 COST ESTIMATES FOR EACH SUBCATEGORY
Estimated costs for the waste water treatment plants for
the different annual productions and at various levels of
treatment are calculated in terms of total annual costs. The
total annual cost is the summation of the annual amortization
of the investment costs and the annual operation and maintenance
costs.
The types of costs shown for each model plant are:
(a) Investment
(b) Annual operation and maintenance
(c) Annual amortization of investment costs (excluding
land)
The total annual costs per metric ton of product have
been calculated.
For the purpose of the cost estimate, the first level of
treatment represents the base cost of the treatment system
(BPT). The other levels (second, third, etc.) represent the
incremental cost above the base cost. The actual additional
costs a plant would incur in implementing the described
treatment processes depend on current treatment practices, and
to some extent on the availability of land.
\
In some cases, land for economical on-site sludge disposal
for a ten year period has been provided in the BPT model plant
costs. Since land cost is not amortized, its value appears in
the initial investment cost but not in the total annual costs.
Where land is a major factor in the BPT estimated costs, its
significance will be mentioned in the separate reviews of each
subcategory.
For the purpose of cost estimating, a set of generally
representative model plant specifications are given for each
nonexcluded subcategory starting with the Chlor-Alkali industry
in Section 11. These specifications, together with the basic
assumptions on co,st estimating detailed in this section, form
149
-------
the basis for the baseline cost estimates for alternative
treatment systems. These cost estimates are presented in a
tabular format in the cost development portion of each
applicable subcategory section. In order to take into account
more fully the wide range of plant specific variables,
additional cost elements which may add to the baseline costs
are then considered on a case-by-case basis. The results are
either expressed graphically as a cost envelope or are given
as an estimated percentage factor to be applied to the baseline
costs.
150
-------
SECTION 11
CHLOR-M.K&LI INDUSTRY
11.1 MERCURY CELL PROCESS INDUSTRY PROFILE
1-1.1.1 General Description
Chlorine and its co-product caustic soda (alkali) are used
in large quantities in the production of plastics, organic and
inorganic chemicals, in the pulp and paper industry, in water
and waste water treatment and in a number of other industries.
The production rate in the United States is approximately 9
million metric tons (10 million short tons) of chlorine per year
and over 95 percent of that production is by the electrolysis of
a sodium or potassium chloride solution via one of two major
processes. The two processes, mercury cell and diaphragm cell,
differ in cell design and in the quantity and quality of waste
water generated, and because of these difference they are being
addressed separately under the Chlor-Alkali Subcategory.
Other processes for chlorine production such as the
recently developed membrane process are not addressed here
because only pilot-scale production exists "or no data is
available from fully operating facilities.
Approximately 30 percent of the U.S. production of chlorine
is by mercury cell plants. Of 27 known plants, 308 data was
available for 15. Table 11-1 presents a summary profile of the
subcategory. Table 11-2 presents the current status of
discharge regulations for mercury cell chlorine plants.
11.1.2 General Process Description and Raw Materials
Brine System
The sodium chloride solution (brine or salt dissolved in
water) is treated with sodium carbonate1and sodium hydroxide to
precipitate impurities such as calcium, magnesium and iron.
The precipitated hydroxides and carbonates are then settled
151
-------
TABLE 11-1. SOBeaiEQORY ERCFUE DMCA
SUBCKDSQOOT
ODJQBINE MERCURY
Total subcategory capacity rate
Itotal subcategory production rate
Number of plants in this subcategory
308 Data on file for
With total capacity of
With total production of
Representing capacity
Representing production
Plant production range;
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
3,545,000 kkg/year
2,750,000 kkg/year
27
15
1,280,600 kkg/year
1,090,000 kkg/year
36 percent
40 percent
19,100 kkg/year
198,000 kkg/year
77,900 kkg/year
70,400 kkg/year
75 percent
2 years
26 years
4 cubic meters/day
2,100 cubic meters/day
< 1 cubic meters/kkg
11 cubic raeters/kkg
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Commerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chemical Industry,"June, 1978, and "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards for the Inorganic Chemicals Industry,"
March, 1980.
152
-------
TABLE 31-2. STATUS OF REGULATIONS - EFFLUENT LIMITATION GUIDELINES
SUBCATEGORY
SUBPART
CHLORINE MERCURY CELL
F (40 CFR 415.60, 3/12/74)
STANDARDS
Product Para-
BPCTCA
(1)
BATEA*
.(2)
Max." Avg; ' Max. Avg.
NSPS
Max. Avg.
Process meters
Mercury
Cell TSS
Process
Hg,
(kg/kkg)
0.64
0.00028
0.00028
(kg/kkg)
0.32
0.00014
0.00014
(kg/kkg) (kg/kkg)
No discharge
of pwwp3
No discharge
of pwwp
(kg/kkg) (kg/kkg)
0.64 0.32
0.00014 0.00007
* Section 415.63 was remanded and is presently reserved (41 FR 51601,
November 23, 1976).
(1) Max. = Maximum of any one day.
(2) Avg. = Average of daily values for thirty consecutive days shall not
exceed.
(3) pwwp = Process waste water pollutant.
153
-------
usually in a clarifier and the underflow, known as brine mud, is
sent to a lagoon or filtered. Brine muds from mercury cell
plants usually contain small amounts of mercury because the
spent brine from the cells is recycled. Consequently brine mud
filtrate is recycled or treated before discharge and solids are
disposed of in secure landfills.
Before it is sent to the cells, treated brine is evaporated
if necessary to remove excess water and then pH adjusted. Spent
or depleted brine from the cells is acidified and dechlorinated
using vacuum and/or air stripping before being saturated with
salt and recycled.
Mercury Cell Process
The mercury cell, in general, consists of two sections:
the electrolyzer and the decomposer or denuder. The
electrolyzer is an elongated steel trough that is inclined
slightly from the horizontal. Mercury flows in a thin layer at
the bottom forming the cathode of the cell, and the brine flows
cocurrently on top of the mercury. Parallel graphite or metal
anode plates are suspended from the cover of the cell. Electric
current flowing through the cell decomposes the brine,
liberating chlorine at the anode and sodium metal at the
cathode. The metallic sodium forms an amalgam with mercury.
2 NaCl(aq) + Hg = C12 + 2 Na(Hg)
The amalgam from the electrolyzer flows to a denuder and
the spent brine is recycled to the brine purification process.
In the denuder, the amalgam becomes an anode to a short-
circuited iron or graphite cathode. Deionized water is added to
the denuder which reacts with the amalgam to form hydrogen and
caustic soda. In modern mercury cells, the denuder or
decomposer is a horizontally or vertically laid graphite-packed
bed. The water and the amalgam flow countercurrently. Mercury
is then returned to the electrolyzer.
Product Purification
Chlorine from the cell is cooled to remove water and other
impurities. The condensate is usually steam stripped" for
chlorine recovery and returned to the brine system or discharged
After cooling, chlorine gas is dried further by scrubbing with
sulfuric acid. The diluted acid is then usually regenerated,
sold or used for pH control. When chlorine gas is compressed
and liquified, it leaves behind noncondensible gases known as
tail or sniff gas. The tail gas is usually scrubbed with
caustic or lime, generating a hypochlorite solution which is
then decomposed, used on-site, sold or discharged with or
without -treatment.
154
-------
The sodium hydroxide or caustic product formed at the
denuder has a concentration of 50 percent NaOH. Some of the
impurities present in the caustic can be removed or reduced by
the addition of certain chemicals, and the caustic is then
filtered. In most cases it is sent to storage or is evaporated
if a more concentrated product is required.
Hydrogen gas is cooled by refrigeration to remove water
vapor and mercury, and can be treated further by molecular
sieves or carbon. Condensate from hydrogen. cooling is then
discharged or recycled to the denuder after mercury recovery.
Figure 11-1 presents a general process flow diagram of
chlorine production by mercury cell.
11.2 WATER USE AND WASTE WATER SOURCE CHARACTERISTICS
11.2.1 Water Use
Water is used at mercury cell plants for noncontact
cooling, tailgas scrubbing, cell washing, equipment
maintenance, floor washings and in the decompositon of sodium-
mercury almagam in the denuder to produce sodium hydroxide.
Because most brine systems at mercury cell plants are closed
systems, water use in the brine system is minimal. The total
water usage at plants was found to range from 7.6 to 204 cubic
meters per metric ton (1800 to 49,000 gallons per short ton),
with noncontact cooling water which is not covered by this
effluent guideline comprising approximately 70 percent of the
total.
11.2.2 Waste Sources
The following waste sources are or can be contaminated with
mercury and would therefore require treatment if discharged.
Brine Hud
This is the waste produced during the purification of brine
before it is introduced into the cell for electrolysis. The
metals commonly removed during purification are magnesium,
calcium, iron and other trace metals such as titanium,
molybdenum chromium, vanadium and tungsten. Calcium and iron
are removed as hydroxides. Brine mud is the major portion of
the waste solids produced from the process. The solids content
of the stream varies from 2 to 20 percent and the volume varies
from 0.04 to 1.5 cubic meters per metric ton of chlorine
produced. The waste is either sent to a pond for settling or is
filtered. The overflow from the pond or the filtrate is
155
-------
U1
BRINE •*•
1
C<
(
*
OtlOJUHe
PURIFICATION
SYSTEM
1
BRINE
HUBS
TO WASTE
AND RECYCLE
TONCONTACT
(CONTACT*)
XH.IHG WATER
i
C
o
L
E
R
PURIFIED ^ uimmrnv
-BRIHB •>• "£££** ~ r
* * t
HOBCOHTACT
COOLING HATER
^ BACKHAS
»ACID ^CELL" "*
T
D
R
Y ^
E ^ miHUiSyoR
' WEAK SULFURIC
:ONDENSATE ACID
TO HASTE OR REUSE
— IIYDROUEH
l •
«ALGAH ^ (DECOMPOSER) "*~ DBHINERALIZED
HROnlBV 1 1 "*~ "VNO- JO ATMOSPHERE
S0» iiUUlUM HOMCOIITACT K1LTBK) TOR (JsR
HYDROXIDE SOLUTIOH COOLING _^
X HATRR
COOLER CONDENSATE
t (SCRUBBING HASTE*) ^>TO O*fiSTE
en< cnr,Ttiu 1" 1 RECYCLE COOLING
HATER — +. FILTER -5°» SODIIW — (PACKAGING 1 *• TOWER
HYDROXIDE 1 1 -- -rilr~
•t 1 SOLUTION - ""'-'"-
BACKWASH SOLUTIOH TO
I ATMOSPHERE
11 RECYCLE 1 BACKWASH A 1
^ | FILTER CAUSTIC (LIME) [ f
| "" ANU HATER | BLOW DOWN
SOIfDS TO INERTS
f~ LANDFILL
r i
\ 1 ^-J R
^ -MERCURY ' ' SOLIDS ~ -*-70 "ASTE B
- TAIL GAS f» B SODIUM (OH CALCIUM)
5 HYPOCHLORITE..^.
SOLUTIOH
LIQUID _^TQ SAtES TO USE, SALES,
. j PRODUCT
REFRIGERATION
\ SYSTEM
t t
NONCONTACT
COOLING
HATER
USED AT SOME PLANTS ONLY
Figure 11-1. General process diagram for production of chlorine/caustic by mercury cells.
-------
recycled to the process as makeup water for the brine. In the
mercury cell process, only 16 percent of the NaCl solution is
decomposed in the cell and the unconverted brine is recycled to
the purification unit after dechlorination. This recycled brine
is contaminated with mercury so the resulting brine mud contains
small amounts of mercury.
Cell Room Wastes
The major components of this stream include leaks, spills,
.area washdown and cell wash waters. The amount varies from
plant to plant and depends largely on housekeeping practices.
Data indicate a range of from 0.01 to 1.5 cubic meters per
metric ton of chlorine produced. Cell room waste constitutes
the major stream requiring treatment because of the high levels
of mercury present in these wastes. If graphite anodes are used
in the cells, the wastes may also contain lead (used as an
electrical contact at the anode) and chlorinated organics.
However most mercury cell plants have converted to metal anodes.
Chlorine Condensate
Condensation from the cell gas is contaminated with
chlorine. At some plants, the condensates are recycled to the
process after chlorine recovery. Both contact and noncontact
water is used for chlorine cooling and for removal of water
vapor. Because of this, the amount and type of waste water
varies from plant to plant. Data from one plant indicates a
waste condensate flow of approximately 0.01 cubic meter per
metric ton of chlorine produced.
Spent Sulfuric Acid
Concentrated sulfuric acid is used in the dryer to remove
the residual water from the chlorine gas after the first stage
of cooling. In most cases, the acid is used until a constant
concentration of 50-70 percent is reached. The spent acids can
be regenerated for reuse, used for pH control in a treatment
system, or sold.
Tail Gas Scrubber Liquid
The tail gas containing the uncondensed chlorine gas from
the liquefaction stage, along with some air and other gases, is
scrubbed with sodium/calcium hydroxide to form sodium/calcium
hypochlorite solution. When the equipment is purged for
maintenance, the tail gas is also absorbed in calcium or sodium
hydroxide, producing the corresponding hypochlorite solution.
The hypochlorite can be used in another process on site, sold,
discharged to treatment or decomposed 'before discharge or
157
-------
treatment. The amount of tail gas scrubber water varies from
0.04 to 0.58 cubic meter per metric ton of chlorine.
Caustic Filter Washdown
The 50 percent caustic produced at the denuder is filtered
to remove salt and other impurities. The filters are backwashed
periodically as needed, and the backwash can be discharged to
treatment or filtered with the filtrate recycled to the brine
system and the solids sent for disposal or mercury recovery.
Waste water volume from caustic filter backwashing is variable
and no flow data are available.
Hydrogen Condensate
Hydrogen produced at the denuder is cooled to remove
mercury and water carried over in the gas. The condensate is
either sent to treatment facilities or to mercury recovery after
which it can be returned to the denuder. Data on the volume of
this waste stream are not available.
Summary of Waste Water Flow
Summing the flow ranges presented above for specific waste
sources results in a maximum mercury-contamined waste flow of
2.1 cubic meters per metric ton (m3/kkg) for plants where
specific stream data were available. This does not include
brine mud flows which are reused instead of discharged, and
therefore do not affect total flow.
Data available on total discharges at 13 mercury cell
plants are presented in Table 11-3. The average discharge
volume indicated is also 2.1 m3/kkg, although flows as high as
6.3 m3/kkg do exist.
11.3 DESCRIPTION OF SPECIFIC PL&HTS
The following descriptions of specific plants includes
those that 'were sampled during the screening and verification
program. The discussion primarily covers plant practices in
waste water control and treatment.
11.3.1 Screening Progran
Plant f299 was visited in. the screening and verification
phase of the program. The mercury-contaminated waste streams
include outlet end-box wash water, spills and cleanup water,
brine mud saturator sludge, and pump seals waste water. The
combined waste water is sent to a surge pond. The effluent from
158
-------
11-3. OF PLOW FOR CHLORINE MERCURY CELL HANTS
SUBCATBGORY CHLORINE ME8CURY CELL
Plant
Number
317
907
299
167
747
343
106
131
589
898
741
553
769
Average of 13 plants
feste Water Flow
^m /kkg Chlorine)
0.51 '
0.36
1.6
5.6
0.69
1.6
0.67
1.7
5.8
0.98
0.51
1.0
6.3
2.1
159
-------
the surge pond is mixed with sodium bisulfide and sent to a
settling pond. The overflow from the pond is pH adjusted,
filtered (in a filter press) and passed through activated carbon
towers before discharge. In the sampling program waste water
influent to the surge pond and the overflow from the settling
pond were sampled. Figure 11-2 gives the general process
diagram and shows all the waste streams sampled. Table 11-4
presents major pollutant concentrations and loads for the
sampled streams.
11.3.2 Verification
Four more plants (1*747, $167, £106 and f317) producing
chlorine/caustic by mercury cells were visited and sampled in
the verification program. Table 11-5 presents pollutant
concentrations for the sampled streams—and loads for TSS and
mercury.
At Plant f747r the brine dechlorination system has been
converted from barometric condensers to a steam ejector system.
The conversion resulted in increased chlorine recovery and
reduced contact waste water. By providing settling and
secondary filter facilities, the brine filter backwash has been
eliminated. The tail gas scrubber liquid (hypochlorite
solution) is offered for sale and if not marketed, is treated
for removal of chlorine and discharged. Mercury bearing waste
waters are treated with sodium sulfide (Na2S) and filtered.
Solids are retorted for mercury recovery and the filtrate is
mixed with the other process waste waters and the pH adjusted
before discharge. A flow diagram of the manufacturing process,
including the waste water treatment facility, is given in Figure
11-3.
At Plant f!67, the waste water streams, consisting of
filter backwash, cell room wash, rain water runoff, and leaks
and spills, are combined and treated for mercury removal. The
water is sent to a holding lagoon and the overflow is reduced by
reaction with ferrous chloride, which precipitates mercury. The
reacted solution is sent to a clarifier and the underflow from
the clarifier is disposed of in a landfill. The overflow is
filtered and the filtrate is passed through activated carbon and
an ion exchange column prior to discharge to a lagoon. The
effluent from the lagoon is pH adjusted and discharged. Figure
11-4 shows the simplified process flow diagram for Plant f!67,
including the sampling locations.
At Plant f317, the brine purification mud is mixed with
spent sulfuric acid and sodium hypochlorite solution. The
treatment removes mercury from the mud and transfers it to the
solution. The solution is filtered and the solids landfilled.
The filtrate is mixed with other mercury-contaminated waste
160
-------
PMM,
DISCHARGE
-JO BWEB
Kibte stxewa saqpled.
Huntiers to ( ) were sw^ded in
soreavtng, others during verif IcatJoi),
Figure 11-2. General process flow diagram at plant #299 showing the sampling points.
dilorine/caustic (nercury cell) manufacture.
-------
TABLE 11-4. POLLUTANT CONCENTRATIONS MUD LOADS £T PLANT I 299
SOBCATEGORY CHLORINE
Stream Stream
Number Description
Screening Phase: ^ '
1 Cell Waste
2 Mercury Treatment
Effluent
3 Tail Gas
Scrubber
Verification Phase: ^ '
1 Mercury Treatment
Influent
2 Mercury Treatment
Effluent
3 Cell Waste
4 Brijie Mod
5 Tail Gas Scrubber
(MERCURY CELL)
TSS
(mg/1) (kg/kkg)
12 0.016
5 0.007
NA NA
91 0.13
18 0.026
120 0.17
13,000 NA.
180 0.022
Mercury
(mg/1) (kg/kkg)
0.15
0.029
0.11
5,9
0.20
11.
0.54
0.17
0.0002
0.00004
NA
0.080
0.0003
0.015
NA
0.00002
NA = Not available.
(1) = Data based on one 72-hour composite sample of each stream.
(2) = Data based on three 24-hour composite samples of each stream.
162
-------
11-5. AND IOADS AT VMEPICATIDN
SUBC&TEGOEX CHLORINE {MERCURY CELL)
Stream Stream
Number Description
Plant
1
2
3
4
5
6
7
Plant
5
6
7
8
9
Plant
1
2
3
4
5
6
7
8
Plant
1
2
4
747
Cell Waste
Treated Waste
Acid Input
Acid Output
Dechlor System
G-2 Condensate
Tail Gas Scrubber
167
All C12 Wastes
Cell Wash
Brine Process
Treated Waste
Clarifier
Underflow 5
317
Cell Waste
Brine Mad
Filtrate
Tank Car Wash
Collection
Tank 21
Treated
Effluent
Deionizer
Effluent
N-C Cooling
Final Effluent
106
Cell Wash
Treated. Cell
Wash
Final Effluent
TSS
(mg/1)
700
60
NA
NA
9
"2
NA
560
57
4
2
,900
45
520
18
,000
110 ,
18
16
18
79 ,
20
2.0
Tkg/kkgJ
1.6 x 10~^
1.4 x 10
NA
NA
0.0037
2.7 x 10~5
NA
" —A
5.7 x 10 ^
7.1 x 10 „
1.3 x 10
4.0
NA
NA
NA
8.6
4.4 x 10~2
•— "\
5.2 x 10
2.2
2.4
Mercury
(mg/1)
18
0.10
0.023
0.003
0.035
0.27
0.039
3.8
0.72
0,005
0.32
10.4
14
34
0.033
123
0.10
0.001
0.001
0.002
3.9
0.015
<0.0005
(Kg/KKg)
4.3 x 10~f
2.3 x 10~g
3.5 x 10
7.2 x 10~i
1.5 x 10 jj
1.8 x 10~2
8.0 x 10
1.3 x iol:
6.7 x 10"t
9.0 x 10,
1.8 x 10
-5
8.7 x 10
NA
NA
NA
5.0 x 10
4.3 x 10~5
-7
2.9 x 10 J
1.4 x 10 j|
3.6 x 10
NA
NA = Not available.
(1) = Data based on three 24-hour composites,
163
-------
Had
DISCHARGE
lECfrO
SWP,.II!C POINTS,
MKW1 TO PPOCES5
Figure 11-3. teneral process flow diagram at plant #747 showing the sanpling points.
Qilorine/caustic (mercury cell) manufacture.
-------
RECYCLE
H
CTl
in
SALT
NONCONTACT H,0 „ n
TO WASTE * H-°
BBIHE PROCESS WA1£K
TO (>H ADJUST-
HEHT AND FIIIAL
DISCHARGE
•«—
LAGOON
«to
* W
18
ION
EXCHANGE
LEGENO
SAWLIN6 POINTS
ACTIVATED
CAHSQH
SANtt FHTE.IS
CLARIFIER
*
j
** Fel
i UNDERFLOW TO
UKBFILL
Figure 11-4. General process flow diagram at plant 1167 showing the sampling points.
Oilorine/catistic (mercury cell) manufacture.
-------
waters, which includes the brine purge, cell room liquid wastes
and plant area wash water. This is then reacted with sodium
hydrosulfide to precipitate the mercury as mercury sulfide and
then filtered. The solids are sent to a mercury recovery unit
and the filtrate is sent to a holding tank. The effluent from
the holding tank is mixed with de-ionizer waste and noncontact
cooling water before discharge. The process flow .diagram
showing the waste streams sampled is given in Figure 11-5.
At Plant §106, mercury-bearing wastes are segregated from
other waste waters and combined for batch treatment. Mercury-
bearing leaks, spills, and precipitation are contained and
collected by curbing around the cell room and collecting the
wastes in a common sump. From the sump the combined waste is
pumped for treatment. In the treatment system, the pH is
initially adjusted using waste sulfuric acid and 20 percent
caustic solution as required. Sodium sulfide and filter aid are
added and the waste agitated in fiberglass reaction tanks. The
effluent from the tanks is filtered and the filter cake is
retorted for mercury recovery. The residual waste, after
mercury recovery, is placed in a lined solid waste disposal
area. The filtrate is sent to the first of two lined lagoons.
Primary pH adjustment is made using waste sulfuric acid and 20
percent caustic before entry into the first lagoon; final pH
adjustment is made between the first and the second lagoons.
11.3.3 Descriptions of Plants Not Sampled
At Plant f589, the waste water going to the mercury
treatment system consists of cell room washdown, brine filter
backwash, leaks, spills, cleanup water, and hydrogen cooling
condensate. The waste waters are reacted with hydrochloric acid
and sodium bisulfide and then sent to a settling basin where
mercury sulfide precipitates. The overflow is passed through a
series of effluent filters before discharge.
At Plant £343, the cell room wash water, brine purification
sludge, and chlorine cooling condensate are combined and sent to
a pond. The suspended solids settle in the pond and are dredged
out once a year. The dredged sludge is "Chem Fixed" and
disposed of in an appropriate landfill. The overflow from the
pond is reacted with Na2S and the reacted solution is sent to a
clarifier. The clarifier underflow, consisting mainly of
mercury sulfide, is returned to the pond. The clarifier
overflow is discharged.
All contact waste water at Plant 1*907 is treated for
mercury removal in a patented process involving reduction of
mercury to the metallic state using sodium borohydride.
Previously contaminated wooden flooring in the cell room has
been removed and replaced with fiberglass gratings to reduce the
166
-------
H
a\
-j
SALT
CEF GAS
DE-IONIZED NQN3MSCT
WATER VJVSTTi COGtlHS
tEGBH
Waste streams canpled.
DISCI IAICE
Figure 11-5. General process flow diagram at plant #317 showing the sampling points.
Chlorine/caustic (mercury call) manufacture.
-------
amount of mercury in the effluent and for better waste control.
Molecular sieves have been installed on cell end boxes to reduce
the mercury content in the air vented from the cells. The
treatment not only cleans the air but is also believed to reduce
mercury in the plant area runoff.
In the treatment system, the mercury-contaminated waste
water is reacted with sodium borohydride to reduce dissolved
mercury to the metallic form. The reacted solution is filtered
prior to delivery to one of the banks of three columns packed
with anthracite coal. After passing through three absorption
columns in series, the treated waste water is delivered to large
holding tanks, from which it may be discharged or returned to
treatment, depending on its mercury content. Filter, cake,
resulting from the filtration of the waste prior to the coal
absorption step, is retorted for mercury recovery.
Waste solids at this facility, including mercury treatment
sludges and brine muds, are deposited in an on-site disposal
area. Chlorine discharges are essentially eliminated by three
significant waste management practices: the chlorine condensate
is collected and returned to the brine system, tail gas
scrubbing effluents are used in the manufacture of another
product, and spent sulfuric acid from chlorine drying is
dechlorinated in an air stripper and shipped off-site for the
manufacture of another product. Gases from the air stripper are
returned to the chlorine purification header.
At Plant #324, the barometric condenser on the brine
dechlorination was replaced with an indirect cooler, resulting
in a reduction of chlorinated waste water. The tail gas
scrubber effluent is used for the manufacture of another
product, and the brine muds are sent to a pond. Small amounts of
mercury, when detected in the brine mud, are leached with water
and treated with other mercury-contaminated waste waters which
include the cell room wash water, caustic filter backwash, and
brine leaks. The combined waste water is mixed with hydrogen
processing waste water, reacted with sulfuric acid, sodium
borohydride, and sodium sulfide, and then filtered. The
filtrate is adjusted for pH and recycled to process.
At Plant #385, the brine mud sludge is sent to a retention
pond where it accumulates. All process contact waste, water is
collected in an unlined pond where it is treated and the treated
effluent is used as the scrubber liquid for tail gases. The
spent scrubber solution is sent to an adjacent paper plant for
use.
At Plant C416, the cell room wastes are used for bleach
manufacture. The waste water streams from the chlorine/caustic
plant are sent to an adjacent paper company.
168
-------
At Plant f784, the waste water, consisting of KC1 brine
filter backwash and area washdown and spills, is sent to a
basin. The basin equalizes the flow and the overflow is treated
with sulfuric acid prior to reaction with NaHS and
clarification. The clarifier overflow passes through an
activated carbon filter and to a final tank where it undergoes
pH adjustment before discharge.
The wastes are segregated at Plant £674. The clarification
pond is used for waste streams containing suspended solids. The
streams going to the pond include brine purification muds and
spent chlorinated lime. The mercury-contaminated waste waters
are treated separately. These include the brine saturation
waste, brine filter backwash, cell room sumps, and tank car
washes. The combined mercury-laden waste water is sent to a
collection pond and the overflow from the pond is pH adjusted
before the addition of Na2S. The reacted solution is sent to a
another pond and the, pond overflow is passed through a carbon
adsorption column before final discharge. A part of the treated
effluent is re-injected into the brine well.
At Plant f012, the brine treatment area is paved to trap
all spills, leaks, and rain runoff from that area. The
recovered waste is recycled to the weak brine reservoir. The
contaminated waste waters from the plant are re-injected into
the brine wells to keep the hydraulic balance and maintain
pressure in the salt deposits.
11.3.4 Summary of the Toxic Pollutant Data
Presented below are the toxic pollutants found in the raw
wastes during screening and verification.
Because several waste streams usually contribute to the
total raw waste at mercury cell plants, a calculation was often
necessary to determine the pollutant concentrations that would
exist in the streams before they were mixed prior to treatment.
An example of this calculation is the "mixing" of the following
hypothetical streams:
Stream A: 100 gallons per minute, 15 mg/1
Stream B: 10 gallons per minute, 60 mg/1
(Flow x concentration) + (Flow x concentration)
Total Flow
= concentration of mixed streams
= (100 gpm)(15 mg/1) + (10 gpm) (60 mg/1) = 19 mg/1
110 gpm
169
-------
The maximum raw waste concentrations observed during any
single 24-hour sampling period were:
Maximum Raw Waste Concentrations Observed
(yg/i)
Pollutant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Thallium
Zinc
Screening
Plant
(#299)
< 250
< 10
< 1
8
350
1
150
< 100
< 1
140
230
Verification
Plants
(f299, f747, f!67,
?206, $317)
770
400
790
180
2,300
1,900
180,000
2,400
870
440
34,000
Section 5.1.2 of this report describes the methodology of
the screening and verification sampling program. In the
chlorine mercury cell industry, a total of 18 days of sampling
were conducted at Plants f299, t'747, f!67, £317 and f!06.
Thirty-two different sampling points were involved covering
various raw waste streams and the treated effluents at these
plants. The evaluation of toxic metal content of these process
related waste streams was based on 949 analytical data points.
The screening for toxic organic pollutants at Plants i'299 and
$167 generated an additional 490 analytical data points. The
daily raw waste loads were calculated from the waste stream flow
rates measured or estimated at the time of sampling and the
measured pollutant concentration.
The daily loading is determined by:
Daily loading (as kg of pollutant (C)(Q)
per day) = 1000
Where:
C is the concentration of the pollutant expressed in
units of mg/1 (Note: kg/m3 = 1000 mg/1), and
Q is the waste stream flow rate expressed in units of
m3/day (m3, a cubic meter, is equal to 264.2 U.S.
gallons).
170
-------
Similarly, the unit loadings were calculated from the
reported chlorine production rate, the waste stream flow rate,
and the measured pollutant concentration:
Unit loadino (as kg of pollutant per (C)(Q)
kkg of chlorine) = 1000P
Where C and Q are the same as described above, and P is the
chlorine production rate expressed in units of kkg/day (kkg is
1000 kg, a metric ton, which is equal to 2205 Ibs).
The minimum, average, and maximum values are based on data
from those plants where the particular pollutant was found at a
concentration greater than the analytical detection limits and
considered a "significant concentration". The term "significant
concentration" means an observed concentration in any 24- or 72-
hour composite raw waste sample that is above the analytical
detection limit, and treatable by an available technology
regardless of economic considerations.
In Table 11-6, the toxic pollutant raw waste data are
presented as the average daily concentrations and the unit
loadings found at the individual plants. These averages were
derived by averaging the concentrations and loads based on three
24rhour composite samples from each plant.
In Table 11-7 daily loadings (in kg/day) and unit loadings
(in kg/kkg) are presented as minimum, average and maximum values
based on the data presented in Table 11—6.
Based on the total annual production of this subcategory
and the average waste load generated per unit product, the
estimated total pollutant raw waste loads generated each year by
this subcategory are as follows:
Raw Waste load
Pollutant (kg/year)
Antimony 1,400
Arsenic 1,000
Cadmium 210
Chromium 360
Copper 960
Lead 880
Mercury 44,000
Nickel 820
Silver 850
Thallium 770
Zinc 7,200
171
-------
TABLE 11-6. TOXIC POLLUTANT RAW WASTE CONCENTRATIONS AND LOADS AT
VERIFICATION PLANTS
^kg/kkgy
SUBCATEGORY
Pollttfcant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Thallium
Zinc
CHLORINE (MERCURY CELL)
299
0.48
0.00077
0.23
0.00037
0.010
0.000016
0.063
0.00010
0.30
0.00047
0.060
0.000096
5.9
0.0081
*
*
0.18
0.00029
0.27
0.00043
Plant
747
0.11
0.000078
0.030
0.000021
0.020
0.000014
0.10
0.000071
0.38
0.00027
0.16
0.00011
18
0.0043
0.093
0.000066
0.047
0.000033
0.022
0.000016
0.69
0.00049
#
167
*
0^33
0.0011
*
0.12
0.00040
0.075
0.00025
0.072
0.00024
3.8
0.013
0.060
0.00020
*
*
0.17
0.00057
317
*
0.10
0.00005
0.46
0.00023
0.080
0.000040
1.2
0.00060
1.4
0.00070
123
0.048
1.4
0.00070
0.11
0.000055
*
20
0.010
106
0.49
0.00070
*
0.031
0.000044
0.013
0.000019
0.12
0.00017
0.33
0.00047
3.9
0.006
0.17
0.00024
0.58
0.00083
0.38
0.00054
0.96
0.0014
* - Concentration below significant level as defined in 11.3.4.
172
-------
11-7 „ OF RAW LOADINGS AT
VERIFICATION PLSNTS
SUECATEGORY
Pollutant
min.
Antimony
Arsenic
Cadmium
Chratmm
Copper
Lead
Hercury
Nickel
gilver
0
0
0
0
0
0
1
0
0
.044
.0054
.0062
.0043
.045
.036
.6
.037
,0059
UsUiuE 0, 0090
Zinc
0
.14
CHLORINE
Daily
Loadings
(kg/day)
avg.
0.17
0.11
0.013
0.037
0.10
0.070
3.1
0.056
0.082
0.086
0.41
(MERCURY C
max.
0,30
0.27
0.025
0.098
0.18
0.12
5.1
0.075
0.22
0,14
1.1
at)
Unit
Loadings
(kg/kkg)
min.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
000078
000021
000014
000019
00025
000096
0043
000066-
000033
000016
00043
,0
0
0
0
0
0
0
0
0
,0
0
avg.
.00052
.00038
.000076
.00013
.00035
.00032
.016
.00030
,00031
.00028
,0026
Number of
Plants
Averaged*
max.
0.
0.
0.
0.
0.
0.
0,
0.
0.
0,
0.
00077
0011
00023
00040
00060
00070
048
00070
00083
00054,
010
3
4
4
5
5
5
5 .
4
3
3
5
* - Only those plants where the pollutant was observed at "significant
concentrations" are included in the averaging. "Significant
concentrations" is defined in 11.3.4.
173
-------
11.4 POH.OTION ABATEMENT OPTIONS
11.4.1 Toxic Pollutants of Concern
Mercury is the major toxic pollutant of concern in the
production of chlorine by the mercury cell process. Other toxic
metals often found in significant concentrations in raw wastes
include arsenic, antimony, cadmium, chromium, copper, lead,
nickel, silver, thallium and zinc. Sources of these metals are
assumed to be impurities in the raw salt or brine and corrosion
products from the reaction between chlorine and process
equipment materials of constuction. No toxic organics were
found at significant levels.
11.4.2 Prey jailing Control and Treatment Practices
Section 11.2.3 described specific control and treatment
practices at 14 plants. All known mercury cell plants practice
treatment of mercury-bearing wastes, but control practices such
as recycling of brine mud filtrate or pond overflow, chlorine
condensates, hydrogen condensates and caustic filter backwash,
and solids handling vary from plant to plant. Although all
known treatment facilities precipitate mercury and separate the
solids formed by clarification and/or filtration, sampling data
has shown that some treatment systems including those with more
advanced technologies such as' adsorption or ion exchange, are
not operating efficiently.
11.4.3 Process Modi£ications and Technology Transfer Options
The following process modifications are being practiced at
one or more mercury cell plants and can significantly reduce
pollutant loads discharged.
Anode Material
Nearly all mercury cell plants now use metal anodes. Their
use, as opposed to graphite anodes, improves the power
efficiency of the cells and reduces the potential pollutant
load.
Liquefaction of Chlorine
Utilization of high pressure and refrigeration for chlorine
recovery will reduce the chlorine content of tail gases.
Brine Recycling •
Although practiced at many facilities not all plants are
using a closed-loop brine system which eliminates a significant
waste volume requiring mercury treatment.
174
-------
Mercury Emissions
Hydrogen gas produced in the denuder can be refrigerated
and passed through treated carbon or molecular sieves to remove
the mercury escaping with the gas.' This will reduce the mercury
emissions and reduce atmospheric fallout in the neighborhood of
the plant. This in turn will reduce mercury concentrations in
storm runoff. Two plants are practicing this control
technology.
Tail Gas Emission Control
When chlorine gas produced from the cell is compressed and
cooled, chlorine separates ' as liquid chlorine, and
noncondensable gases (tail or sniff gas containing residual
chlorine vapor) are produced at the discharge end of the
condenser. The amount of chlorine present in the tail gas is
significant and has to be removed and treated or recovered
before the tail gas is vented to the atmosphere. The common
industrial practice is to scrub the gas with caustic soda or
lime solution thus producing the corresponding hypochlorite.
The hypochlorite solution is either sold, used on-siter sent to
a waste water treatment plant, or discharged without treatment.
Treatment of this waste is a relatively recent practice.
Decomposition is a common method of treatment using catalytic,
thermal, and chemical methods as described below.
Catalytic decomposition involves the addition of small
quantities of cobalt, nickel, and iron chloride to the waste
streams, followed by retention in reaction tanks for periods up
to several days. Of the two plants employing this technology,
one reports zero discharge of chlorine, and the other reports
respective average and maximum chlorine discharge rates of 0.015
and 0.14 kg per metric ton of chlorine produced.
Thermal decomposition occurs when the temperature of the
solution containing hypochlorite reaches 175 degrees F. Lime
reacts with chlorine exothermically, producing heat and calcium
hypochlorite. If the hypochlorite solution is not cooled,
thermal decomposition occurs . One chlorine/caustic plant is
using this treatment method and another is planning to use it.
The plant using thermal decomposition reports complete
conversion of hypochlorite to chloride.
Chemical decomposition takes place by reacting the
hypochlorite solution with a chemical reactant which is usually
sodium sulfite or hydrogen peroxide. Chemical decomposition is
expensive but complete and rapid.
When chlorine is present in a dissolved form (hypochlorous
acid) in water, a stripping technique may be applied to recover
175
-------
the chlorine. Chlorine condensate streams and spent chlorine-
drying acid are most commonly treated by steam or vacuum
stripping, with the chlorine frequently returned to process for
purification and recovery as a product. The tail gas is not
generally scrubbed with water because water does not effectively
remove chlorine and the chlorine concentration in the exhaust
will reach 0.1 to 4.5 percent by volume after scrubbing with
water. One effective method of chlorine recovery from the tail
gas is by the passage of the gas through an absorbing material
such as carbon tetrachloride and subsequent recovery of the
chlorine. The process is proprietary and little information is
available on its design or performance.
11.4.4 Best Management Practices
Area Runoff
Provisions can be made to divert and contain storm runoff
from plant areas. Collected runoff can then be sent to the
waste water treatment system.
Leaks and Spills
The brine treatment area and the cell room areas can be
paved with fiberglass gratings, and provision should be made to
collect the leaks and spills from the operation.
Mercury Contaminated Solids
The precipitated mercury waste should be stored in a lined
pond, disposed of in a secured landfill or sent to mercury
recovery operations. Brine mud should be discharged to a lined
pond or a secure landfill after filtration. The brine mud
contains small amounts of mercury which can leach into the
ground water if proper safety precaution are not taken.
Transportation, Handling and Abnormal Operations
'Provisions should be made to remove chlorine from air
emissions resulting from abnormal operating conditions such as
start up and shut down, or from vents on returned tank cars,
cylinders, storage tanks, and process transfer tanks during
handling and loading of liquid chlorine.
11.4.5 Advanced Treatment Technologies
Methods available for the removal of elemental mercury or
mercuric salts from plant waste waters include precipitation
with sodium sulfide to form insoluble mercuric sulfide,
adsorption by activated carbon, adsorption by ion-exchange and
other resins, reduction by borohydrate, hydrazine, sulfite,
176
-------
hypophosphite or iron, and biological reduction (57). All of
these methods are patented? many of these methods have been
proven on a pilot scale only. Sulfide precipitation and
adsorption techniques will also provide for the removal of other
toxic metals.
11.5 SELECTION OP APPROPRIATE TECHNOLOGY AND EQUIPMENT
11.5.1 Technologies for Different Treatment: Levels
Following the evaluation of significant toxic pollutants
found in raw waste waters, current industry treatment practices
and applicable treatment alternatives, two levels of end-of-pipe
treatment were selected as alternatives for application in the
mercury cell chlorine subcategory.
Level 1
This treatment consists of sulfide precipitation of
mercury-bearing waste water followed by pressure filtration.
This level of treatment, which will also reduce other heavy
metals, includes recycle of the brine waste stream back to
process, and the settling and storage of brine muds. Mercury-
bearing solids can be sent to mercury recovery or disposal. The
flow diagram for this treatment level is shown in Figure 11-6.
Level 2
The filtered Level 1 effluent is passed through a granular
activated carbon bed where residual metal sulfides and any
metallic mercury will be removed. The flow diagram for this
treatment level is shown in Figure 11-7.
11.5.2 Equipment for Different Treatment Levels
Equipment Functions
In Level 1, typical of existing treatment facilities,
mercury-bearing wastes are equalized in a surge tank, and
following chemical mixing, sulfide precipitates are removed in a
conventional plate and frame filter press followed by final pH
adjustment of the filtrate before discharge. In Level 2 a
conventional granular activated carbon filter is added for
further removal of residual metals before pH adjustment.
Chemical Handling
Sodium bisulfide is used with filter aid after pH
adjustment to pH 5-7. Care is needed to prevent escape of toxic
177
-------
]• ^
BRIM:
HUD STREW
1 f
« *^V MGOOH X »*
. RECYCLE TO
00
sutftmicftdD
FEWER
AID
HERCUOT
CXWBWJKMH)-
VJASTESWEPH
SODIUM
BISULFIDE
n
FH/THt
HOLDING WNK MUttHS
MIXING
i—-g-
SOLIBS TO
MERCURY
RBOOVHW
ORU11WIU.
BETCHCV BBUJHH UM!
1
Includes pH monitoring, flow monitoring and sampler
Figtire 11-6. Level 1 waste water treatment for chlorine - mercury cell subcategory.
-------
H
-4
vo
i' -' ^
BRIM: MUG
smsm
mcuig
aOBBlMtNKTEB
S&S1E S&T13R
fflSCENCir W.TUPN LINE
SDUBS TO
Harare
KBCWSW i
OR IMtnU. I
-a^-EFPUiwr
Inclndeii pH monitoring, flow monitoring and sampler
Figure 11-7. -Level 2 waste water treatment for chlorine - mercury cell subcategory.
-------
and obnoxious H2S fumes at neutral and acid pH levels. At Level
2 no additional chemicals are used since the activated carbon
bed is not regenerated but is periodically removed and replaced.
The. handling of granular carbon may cause temporary dust
problems but it causes no special hazards.
Separation and Removal of Solids
Conventional settling and filtration methods are used, but
because of the toxicity of mercury, precipitated sludges should
be disposed of in a safe chemical waste area.
Monitoring Requirements
Both levels of treatment include provisions for sampling
and monitoring of the waste water discharge. Monitoring of
heavy metals is done by atomic absorption methods at a qualified
commerical laboratory: Simple field tests for heavy metals as a
group are available for routine process control.
11.6 TREATMENT COST ESTIMATES
11.6.1 General Discussion
To prepare treatment cost estimates, a model plant concept
was developed. The proposed model plant characteristics are:
Waste Water Plow
Data presented in Table 11-3 indicate an average waste
water flow of 2.1 m3/kkg for 13 plants, while the average of the
five plants surveyed during this study averaged ].7 m3/kkg. The
latter value was used for developing the detailed cost estimates
presented in the cost tables because the technology base for the
model plants was that observed in the field.
For effluent limitation calculations (see 11.7.2) the more
conservative unit flow from the larger data base and 2.1 m3/kkg
has been used. Cost estimates will be adjusted to reflect the
larger unit flow before promulgation.
Chlorine Production
Approximately 50 percent of the production data for all the
chlorine/caustic plants using mercury cells is available on
file. Production ranges from 19,000 to 198,000 kkg of
chlorine/year. Three model plants with productions of 19,100
kkg/yrr 95,500 kkg/yr and 191,000 kkg/yr were selected to
180
-------
represent the subcategory production range. The flow per unit
of production is assumed to be the same for each size of model
plant. Seventy-seven percent of the plants for which flow data
was available have flows per unit of production equal to or less
than the average unit flow (Table 11-3).
Solid Waste Produced
Brine mud constitutes the major source of solid waste
generated at chlorine plants. Although flows and solids content
varies considerably from plant to plant, an average flow of 0.42
m3/kkg at 10 percent suspended solids gave an estimated solids
load of 42 kg/kkg to be used for cost estimating purposes.
11.6.2 Chlorine Bearing Wastes
In the selection of model plants, the following assumptions
have been made for the chlorine contaminated waste streams. The
chlorine condensate waste stream has not been included in the
waste streams going to the treatment facility. In the majority
of the chlorine/caustic plants, this stream is stripped of
chlorine by steam or vacuum and the chlorine is recycled to the
purification operation. The waste water is then returned to the
process and introduced to the brine purification unit or sent to
the treatment unit. The quantity of waste water generated by
this operation is* small and does not significantly affect the
flow determination. In some cases the chlorine gas from the
cells is contact cooled with water and the scrubbed liquid,
after steam stripping, is reused. The stripping operation in
the recovery of chlorine is part of the process and, therefore,
its cost is not included in the treatment cost. The spent tail
gas scrubber solution, which contains mainly calcium or sodium
•hypochlorite, is assumed to be used or decomposed before it is
discharged or sent to treatment. Thermal decomposition can be
practiced at no additional cost at some facilities, while
another efficient treatment method is catalytic decomposition.
The cost estimates for decomposition are not included here
because at many plants the hypochlorite stream is sold, used on-
site or only infrequently discharged depending on market demand.
However, because of the environmental effects of high
levels of chlorine in waste water discharges, the cost for the
dechlorination of total plant discharges using sulfur dioxide
has been included because this is the treatment method on which
control of total residual chlorine is based.
11.6.3 Model Plant Treatment Costs
On the basis of the model plant specifications and design
concepts presented earlier, the estimated.costs of treatment for
three models having different production levels are shown in
181
-------
Table 11-8r 11-9 and 11-10. The costs of Level 2 treatment are
incremental over Level 1 costs. Annual treatment costs as a
function of production are shown graphically in Figure 11-8.
Similarly, treatment cost per metric ton of product is shown in
Figure 11-9. Table 11-11 presents a summary of the unit cost
distribution between amortization and operation and maintenance
components.
Variability in specific plant water use practices and
treatment applications may be responsible for treatment costs
that are higher than those developed for €he model plant. These
variations have been considered. Using the model plant annual
cost curve as a baseline, consideration of the additional plant
specific cost factors results in the cost analysis as shown in
Figures 11-10 and 11-11 for Level 1 treatment and Figures 11-12
and 11-13 for the Level 2 treatment. The cost envelopes reflect
the impact of higher flows (2.4 m3/kkg) which are required at
some plant locations and the consequent increase in costs due to
additional chemical requirements and the variability in the
costs associated with solid waste disposal. A combination of
these and other specific plant factors may result in additional
costs ranging from 30 to 125 percent of the baseline costs.
Cost estimates are presented in Table 11-12 for plants
requiring dechlorination of waste waters by sulfur dioxide. For
the range of model plant productions, the annual cost of sulfur
dioxide treatment varies from $1.72 to $0.40 per metric ton of
product.
11.7 BASIS FOR REGULATIONS
11,7.1 Basis for BPT Limitations
Technology Basis
Existing mercury cell chlorine plants are controlling
mercury in their waste waters in accordance with existing BPT
regulations which require a discharge of less than 0.00014
kg/kkg of product as a 30-day average. These BPT regulations,
40 CFR.415.62 (a) presently in effect will not be revised.
Pollutants regulated include TSS and mercury. The technology
basis of sulfide precipitation and filtration of mercury bearing
streams (Level 1) is currently being applied at 24 plants in
this subcategory. Other plants in the industry use mercury
control methods that are different in detail but with the same
objective.
The existing regulations, presented in Table 11-2, are
sustained by the fact that plants having properly operated BPT
technology have demonstrated the achievability of the effluent
182
-------
TABLE 11-8 MODEL PLANT TREATMENT COSTS
Subcategory CHLORINE Mercury cell
Production 19,100 metric tons per year (21,057 tcais pet year)
54 metric tons per day (60 tons per day)
Waste water flow 91 cubic meters" per day.
LEVEL OF 7REMMENT*
FIRST SECOND
A. INVESTMENT COST
Construction $49,100 $500
Equipment in place,
including piping,
fittings, electrical
work and controls 68,100 15,000
Monitoring equipment
in place... 9,000
Engineering design
and inspection... 25,240 3,100
Incidentals, overhead,
fees, contingencies... 25,240 3,100
Land....... 21,000
TOTAL INVESTMENT COST $197,680 $21,700
B. OPERATION AND
MAINTENANCE COST
Labor and supervision. $112,000 _ $14,000
Energy 1,250
Chemicals 500 1,400
Maintenance 17,668 2,170
Taxes and insurance... 5,930 651
Residual -waste
disposal...... „.„ 4,400
Monitoring, analysis
and reporting...,,..... 15,000 7,500
TOTAL OPERATION AND
'MAINTENANCE COST $156,748 $25,721
C. AMORTIZATION OF
INVESTMENT COST $28,745 $3,530
TOTAL ANNUAL COST $185,493 $29,251
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
183
-------
TABLE .^11-9 MODEL PLANT TREATMENT COSTS
Subcategory CHLORINE Mercury cell
Production 95,500 metric tons per year (105,288 tons per year)
272 metric tons per day (300 tons per day)
Waste water flow 455 cubic meters per day.
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT COST
Construction
Equipment in place,
including piping,
fittings, electrical
work and controls
Monitoring equipment
in place
Engineering design
and inspection
Incidentals, overhead,
fees, contingencies...
Land
TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Energy
Chemicals
Maintenance
Taxes and insurance...
Residual waste
disposal
Monitoring, analysis
and reporting
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL AtttJUAL COST
$134,500
141,300
9,000
56,960
56,960
63,000
$461,720
$112,000
3,700
2,500
39,872
13,851
21,400
15,000
$208,323
$64,871
$273,194
$1,000
61,000
12,400
12,400
$86,800
$14,000
7,000
8,680
2,604
7,500
$39,784
$14,122
$53,906
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
184
-------
TABLE 11-10 MODEL PLANT TREATMENT COSTS
Subcategory CHLORINE Mercury cell
Production 191,000 metric tons per year (210,577 tons per year)
545 metric tons per day (601 tons per day)
Waste water flow 910 cubic meters per day.
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT COST
Construction $257,700 $2,'000
Equipment in place,
including piping,
fittings, electrical
work and controls 213,200 115,000
Monitoring equipment
in place 9,000
Engineering design
and inspection 95,980 23,400
Incidentals, overhead,
fees, contingencies... 95,980 23,400
Land 123,000
TOTAL INVESTMENT COST $794,860 $163,800
B. OPERATION AND
MAINTENANCE COST
Labor and supervision, $112,000 $14,000
Energy 6,400
Chemicals............. 5fOOO 14,000
Maintenance 67,186 16,380
Taxes and insurance... 23,845 4,914
Residual vaste
disposal.... 42,600
Monitoring, analysis
and reporting 15,000 7,500
TOTAL OPERATION AND
MAINTENANCE COST $272,031 $56,794
C. AMORTIZATION OF
INVESTMENT COST $109,311 $26,650
TOTAL ANNUAL COST - $381,342 $83,444
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
185
-------
500
o 400
o
•w-
1 30°
200
inn
100 500 " 1000 DAILY
1 1 1 (m3
!
1 !
1
! i
.
-------
u
u
10
LEVEL i'
\
1 !
I I
! I
50 100 150 200
PRODDCTICH (METRIC TONS/SEAR X 1000)
Figure H-9. Annual unit txeatment cxjst vs. production for the Chlorine
Sxibcatecpry (Mercury Cell Process)
187
-------
TABLE 11-11. MODEL ELMO? UNIT TR1MMENT COSTS
Subcategory CHLORINE Mercury cell
Annual Treatment Costs ($/kkg)
LEVEL OF TREATMENT
COST HEM PRODUCTION FLOW FIRST SECOND* THIRD FOURTH
(kkg/yr) (m3/<3ay)
Annual Operation
and Maintenance
Annual
Amortization
Total Cost
19,100
95,500
191,000
19,100
95,500
191,000
19,100
95,500
191,000
91
455
910
91
455
910
91
455
910
8.21
2.18
1.42
1.50
0.68
0.57
9.71
2.86
2.00
1.35
0.42
0.30
0.18
0.15
0.14
1.53
0.56
0.44
JN&t Applicable
* — These costs are incremental to first level costs.
188
-------
500
•400
300
200
f 1 COST
100
I
I
J_
I
I
50
300
100 150 200 250
PKSOCTION (METRIC ims/mR x 1000)
Figure 11-10. Annual treatment cost vs. production for tihe Chlorine
Subcategory (Mercury Cell Process).
189
-------
14
12
10
to-
IEVEL fl COST
I
I
I
50 100 150
PRXDCTION (METRIC
200
x 1000)
Figure 11-11. Annual unit treatment cost vs. production for :the
Chlorine Subcategory (Mercury Cell Process).
190
-------
600
500
©400
300
IEVEL #2 COST RSNGE
200
100
I
J_
50 100 150 200
PIO3UCTIQN (METRIC TCSNS/XE&R x 1000)
Figure 11-12.
annual treatment cost vs. production for the Chlorine
Subeategory (Msrcury Cell Process).
191
-------
16
14
12
10
RANGE
I
J_
I
50 100 150 200
PTODUCTION (METRIC TONS/YEAR x 1000)
Figure 11-13. Annual unit treatment cost vs. production for the
Chlorine Subcategory (Marcury Cell Process),
-------
TABLE 11-12. ESTIMATED CHEMICftL DECHLORINATION COSES FOR THE CHDORrALKALI
SUBCATEGORY CHLORINE
Chlorine Production (kkg/yr)
A.
B.
C.
INVESTMENT COST
Construction. ..........
Equipment in place,
including piping,
fittings, electrical
Monitoring equipment
Engineering design
Incidentals, overhead,
fees, contingencies. . . .
Land
TOTAL INVESTMENT COST
OPERATING AND
MA3MTENANCE COST
Labor and supervision. .
Enerqv
Chemicals (S00)..
Maintenance. ...........
Taxes and insurance. . . .
Residual waste
disposal. ..............
Monitoring, analysis,
and reporting. .........
TOTAL OPERATING AND
MAINTENANCE COST
AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
(MERCURY CELL).
19,100
$3,000
•j'1— ' f W W V
20,100
^.^
4,600
4,600
— —
$32,200
14,000
500
1,500
3,220
966
,„„.„
7,500
$27,686
$5,239
$32,925
31, 850
.
$5,000
35,000
mmm—
8,000
8,000
—
$56,000.
26,000
659
2,000
5,600
1,680
ox... ..........
7,500
$43,439
$ 9,111
$52,550
191,000
$10,000
«!»**«. %* f \s w w
50,000
W.M.
12,000
12,000
w****.
$84,000
,
28,000
1,220
15,000
8,400 *
2,520
— .
7,500
$62,640
$13,660
$76,300
COST PER KKG OF
PHDDUCT (Dollars)
1.72
1.65
0.40
193
-------
limitations based on available long-term monitoring data. Table
11-13 presents data from eleven mercury cell plants, seven of
which are meeting the 30-day average limitations. The other
four plants have mercury control technology installed but are
not meeting BPT limits.
Flow Basis
The existing regulations contained only load limitations,
kg/kkg, and no flow basis or concentration limit was provided.
But the regulations did consider the inclusion of noncontact
cooling water in determining discharge load limitations.
11.7.2 Basis for Proposed BAT Effluent Limitations
The original BAT limitations for this subcategory required
zero discharge of process waste water pollutants. These
regulations were remanded and are not in effect. The proposed
regulations allow for the discharge of process waste water
following treatment.
Technology Basis
Utilizing the cost estimates presented in this report, the
Agency has analyzed the cost effectiveness of Level 1 and Level
2 treatment options for pollutant removal. The economic impact
on the mercury cell chlorine subcategory has been evaluated in
considering the technology basis for proposed BAT limitations.
For BAT, the Agency is proposing limitations based on BPT
technology (Level 1) with the addition of dechlorination.
Dechlorination is being included in BAT because the toxicity of
chlorine to aquatic life is well documented (59) and it is a
pollutant of concern to the Agency. Dechlorination, currently
practiced at two plants, may be required only at fewer than half
of the plants in the subcategory because hypochlorite produced
in tail gas scrubbers is often sold or used in other operations
while residual chlorine in condensates is usually stripped or
recovered. Table 11-14 presents residual chlorine discharges at
plants that have reported the use, sale or treatment of
chlorine-bearing waste waters. This data indicates that
dechlorination technology has not been successfully
implemented.
The Agency considered the addition of carbon adsorption for
additional mercury removal but rejected its use because of high
cost and questionable performance in this industry.
194
-------
TABLE 11-13. MERCURY DISCHARGES FROM SELECTED CHLOR-ALKKLI MERCURY
CELL PISHES*
SUBCATEGORY
Plant
#343
#907
#898
1195
#106
#589
#299
#747**
#317**
#195**
#324**
Average
0.000025
0.000020
0.000060
0.000040
0.000065
0.000055
0.000040
0.000055
0.000006
0.000022
0.00086
CHLORINE (?>IERCURY CELL)
Mercury
Daily Maximum
0.00094
0.00026
0.0025
0.00073
0.00022
0.00086
0.00019
0.000083
0.000048
0.00066
0.0022
laste Load (kg/kkg)
Maximum 30 -day Average
0.00029
0.000030
0.00043
0.00015
0.000096
0.00049
0.000056
0.000065
0.000010
0.00010
0.0018
* See Reference 3
t
** Ercm Plant Long Term Monitoring Data presented in Appendix A.
195
-------
TlfflEB 11-14. RESHTOL CHLORINE DISCHARGES AT SELECTED
CHDOR-JffiKKLI
Plant
1207
f 014
* 819
# 747
# 106
* 589
# 747**
* 324**
Average
0.33
0.04
ND
0.002
0.001
0.003
. 0.0025
3.72
Chlorine Waste Load (kg/kkg)
Range
1.4 maximum
0 to 1.29
0.016 to 0
0 to 0.006
0 to 0.14
0.001 to 0
ND
0.38 to 12
.14
.011
.2
*See Reference 3
**Frcro Plant long Teim Monitoring Data
196
-------
Flow Basis
The flow basis for BAT limitations is 2.1 m3/kkg based on
the average of discharge data of 13 plants presented in Table
11-3 The order of magnitude of this unit flow volume was
supported by data obtained during sampling visits to five plants
at which flows ranged from 0.5 m3/kkg to 5.6 m3/kkg with an
average of 1.7 m3/kk"g.
Selection of Toxic Pollutants to be Regulated
The selection of pollutants for which specific effluent
limitations are proposed was based on the evaluation of raw
waste concentrations found during the sampling program and on
t'he treatability of toxic pollutants using BAT technology.
*
Table 11-15 presents the achievable concentrations of toxic
pollutants using the BAT technology of sulfide precipitation
followed by filtration. The concentrations, based on literature
treatability data presented in Section 8,1 and summarized in
Table 8-11, reflect the lowest level achievable by this
technology for arsenic, cadmium, copper, lead, nickel, silver
and zinc. For antimony, chromium and thallium, literature
treatability data are not available for this technology. Also
presented in Table 11-15 are the maximum and average raw waste
concentrations of toxic pollutants found during the sampling
program with an indication of the number of plants where the
treatability concentration was exceeded.
Based on the occurrence of treatable levels of specific
toxic metals in raw wastes and the fact that the sulfide
precipitation technology is already utilized as BPT in the
chlorine mercury cell subcategory, arsenic, cadmium, copper,
lead, nickel, silver and zinc were selected as additional toxic
pollutants proposed for BAT regulations. Antimony, chromium and
thallium were included for guidance but no limits are proposed
because concentrations found in the raw waste load were below
treatable levels.
Basis of Pollutant Limitations
Limitations are presented as both concentrations (mg/1) and
loads (kg/kkg) for each pollutant. The relationship between the
two is based on the unit flow rate. Although actual unit flow
rates at plants vary by an order of magnitude due to such
factors as raw materials and plant control practices, the Agency
has determined that the load limitations can be met by well-
operated treatment facilities. The concentration or quality
limits are included below.
BAT proposed limitations are presented in Table 11-16.
197
-------
TABLE 11-15. COMPARISON OF RAW WASTE CONCENTRATIONS OP TOXIC POLLUTANTS
WITH TREATAKLLITY
SUBCATEQORY
Pollutant
Antimony
Arsenic
Cadmium
Chroroium
Copper
Lead
Nickel
Silver
Thallium
Zinc
CHLORINE
Treatability * '
CtngA)
_ (2)
t
0.05
0.01
__ (2)
0.05
0.10
0.05
0,05
__ (2)
0.20
(MERCURY
Kfeximum
Plant
Average
tragA)
0.49
0.33
0.46
0.12
1.2
1.4
1.4
0.58
0.38
20
CELL)
Average of
5 Plants
(mg/1)
< 0.28
0.14
0.11
0.075
0.41
0.40
0.35
0.15
0.17
4.4
Number Plants out
of Five
Exceeding
Treatability
Level
__C2)
3
3
_J2)
5
3
2
2
__(2)
4
(1) Literature-based txeatabJJLity estimates from Section 8.1. Table 8-11,
given as the lower limit of txeatability expressed as a 30-day average.
(2) No data available on treatability with sulfide/f ilter.
198
-------
TABLE 11-16. PBDBOSED UMIXKE210NS
Chlorine - Mercury Cell
Best Available Technology
Wbste Eater flows 2.1 -n -
SOBCATEGORY
CHLORINE MRBTTTRY fTT.T.
Subcategory
Performance „,
Pollutant fog/1} VER* ;
Bfcaiconventional
Total Besidual
Chlorine(6)
Concentration Basis Effluent Limit
30-day 24-hour SSday 24-hour
avg. max. avg. max.
Pollutants:
0.2
1.7
0.20
0.34
0.00042
0.00071
Toxic Ksllutants
Antimony ^
Arsenic^53
Cadmium^53
Chrcmium '"
Copper^ 5)
Lead (5)
Mercury ^
Nickel153
Silver^55
1tellium(5)
Zinc {55
0.23 (33
0.10(3>
o.oso{3>
0.040(33
o.oso<2)
0.16(3)
0.020(2)
0.10 <23
0.070(33
0.17 (3)
0.15(3)
2.2
2.2
2.2
' 2.2
2.2
2.2
2.2'
2.2
2.2
2.2
2.2
0.23
0.10
0.050
0.040
0.050
0.16
0.048
0.10
0.070
0.17
0,15
0.51
0.22
0.11
0.088
0.11
0.35
0.10
0.22
0.15
0.37
0.33
_ W)
0.00021
0.00011
_ (4)
0.00011
0.00034
0.00010
0.00021
0.00015
— (4)
0.00032
-W)
0.00046
0.00024
-(4)
0.00024
0.00075
0.00022
0.00046
0.00032
__(4)
0.00070
(1) V5^, t±e variability factor ratio, is the ratio of the -variability factor
for daily measurements to the variability factor for 30-day average.
(2) Lower limit of treatability fior sulfide/filter technology according to
literature treatability data (Table 8-11}.
(3) Average effluent concentration fron verification sampling,
(4) tfo load limits proposed; concentration limits are provided for guidance
purposes.
(5) Limits are also applicable to PSES and PSNS and HSPS.
(6) Limits are also applicable to NSPS.
199
-------
C hlor ine - Total residual chlorine limits are based on data
transfer from the utility industry (58) and the detectable
concentration of chlorine (0.2 mg/1) because treatment should
remove essentially all chlorine. Thus the maximum 30-day
average concentration limit was set at 0.20 mg/1.
The daily maximum limit for total residual chlorine was set
at 0.34 mg/1 based on an evaluation of long-term monitoring data
for total residual chlorine as presented in Appendix A (Table A-
la and c). The ratio of 24-hour maximum variability factors to
30-day average variability factors for two plants was 1.7, thus:
VFR = 2.28 = 1.7
1.38
( 0.20 mg/1 \ / 1.7 j\ = 0.34 mg/1
\30~day averagelimit/ \24-hour maximum limit/
The determination of load limitations for total residual
chlorine (kg/kkg) was calculated based on the unit flow rate of
2.1 m3/kkg, thus:
(0.20 mg/1) (2.1 m3/kkg) (kg/m) = -0.00042 kg/kkg
3000 mg/1
for the 30-day average limit. The 24-hour maximum limit was
calculated similarly, i.e.,
(0.34 mg/1)(2.1 m3/kkg)/_kg/m3_V 0.00071 kg/kkg
Mercury - The proposed BAT limitations for mercury,
although based on the same technology, are more stringent than
BPT limitations. Dechlorination does not affect mercury
removal. The Agency considered the following data in
establishing the BAT limits of 0.00010 kg/kkg for a maximum 30-
day average.
«
o Half of the 'plants with monitoring data presented in
Table 11-13'are meeting the limits.
o Three of five plants were meeting the limits during
sampling of their wastes.
o Three of four plants with long-term monitoring, data
presented in Appendix A are meeting the limits.
The daily maximum limit of 0.00022 kg/kkg for mercury was
based on an evaluation of long term monitoring data from four
200
-------
plants presented in Appendix A. The average variability factor
ratio for the four plants was 2.2. Thus:
(0.00010 kg/kkg) (2.2) = 0.00022 kg/kkg
The concentration limitations for mercury were then
calculated based on the unit flow rate of 2.1 m3/kkg. That is:
(0.00010 kg/kkg) f WOO mg/lN (2.1 m3/kkg) = 0.048 mg/1
—- j
V
and (2. 2) (0.048 mg/1) » 0.11 mg/1
respectively for the maximum 30-day average and 24-hour maximum.
Additional Toxic Pollutants - The effluent limitations
proposed for the selected additional toxic pollutants were
derived from two sources of information? sampling data and
literature-based treatability estimates. Dechlorination does
not affect toxic metals removal.
The results of analysis of treated effluent represents
plant performance observed during three days of sampling. The
effluent data for toxic pollutants found above treatable
concentations in raw wastes are summarized in Table 11-17. Data
are presented from four plants practicing BPT technology
(sulf ide precipitation followed by filtration) . Sampling data
for the fifth plant, £299, reflect effluent quality prior to
filtration.
It is apparent from the sampling data that the BAT
technology systems are generally achieving higher quality
effluents than treatability literature indicates. This could be
a reflection of low influent concentrations and incidental
removal of metals, which indicate that applying effluent
limitations to a dominant metal pollutant (mercury) assures
effective control of other metals.
The concentration bases for the proposed limitations are
derived from average effluent sampling unless the observed
concentration was below the literature treatability level. In
such cases the lowest applicable treatability level from Table
8-11 was used. Because long-term monitoring data from mercury
cell chlorine plant effluents was not available for these
metals, the variability factor ratio established for mercury
limits (2.2) was also applied to these metals. The VFR used in
the existing regulations (2) agrees witli this.
A. Arsenic: Because the sampling data from five plants
(Table 11-17) indicated an achievable . average arsenic
201
-------
11-17. EPELUMT CDNCENTRaTIONS OF TOXIC POIIOTMiTS
FEDM ITEKEFICmTION SAMPLING
SUBCA3DGORY
Pollutant
Jtofcunony
Arsenic
Caflmiuni
Geranium
Copper
Lead
Mercury
Nickel
Silver
Thallium
Zinc
CHLORINE (MERCURY
CHuL)
Plant Effluent Concentrations
(mg/1)
Plant
§299
0.15
' 0.063
0.073
<0.06
0.038
<0.050
0.029
<0.050
<0.015
0.20
0.100
#747
<0.2-5
-------
concentration of less than 0.096 mg/1, this value, rounded to
0.10 mg/1, is supported by the estimated range of treatability
from Table 8-11 and was selected in the concentration basis for
the proposed maximum 30-day average limitation. This effluent
limitation is:
{0.10 mg/1) (2.1 m3/kkg)/_kg/m3 \ = 0.00021 kg/kkg
/ kg/m3 \
\1000 mg/1/
and the proposed daily maximum arsenic limitation is obtained by
applying the VFR value of 2.2 that is:
(2.2)(0.00021 kg/kkg) = 0.00046 kg/kkg
B, Cadmium: For cadmium, the plant sampling data
indicated an achievable average concentration of less than 0.050
mg/1 in the sulfide/filter treated effluent. This falls well
within the range of published treatability values (Table 8-11)
and was used as the concentration basis for the proposed 30-day
average effluent limitations. Thus:
(0.050 mg/1)(2.1 m3/kkg) / kg/m3 \ = 0.00011 kg/kkg
VlQOO mg/1/
for the maximum 30-day average and using the VFR value of 2.2,
the proposed daily maximum is:
(2.2)(0.00011 kg/kkg) = 0.00024 kg/kkg
C. Copper: In the case of copper, the average plant
performance derived from sampling data (Table 11-17) showed an
effluent concentration of less than 0.033 mg/1 which is slightly
below the accepted lower limit of treatability based on
literature data. The latter is approximately 0.050 mg/1 and was
selected as the concentration basis for the proposed maximum 30-
day average limitation on copper. Thus:
(0.050 mg/1)(2.1 m3/kkg) / kg/m3 N = 0.00011 kg/kkg
\iO00 mg/1/
and the proposed daily maximum limitation is obtained by
applying the VFR value of 2.2, that is:
(2.2)(0.00011 kg/kkg) = 0.00024 kg/kkg
D, Lead: The proposed maximum 30-day average limitation
for lead is based on sampling data shown in Table 11-17 which
indicate and achievalbe effluent concentration of less than 0.16
mg/1. Thus:
203
-------
(0.16 mg/1) (2.1 m3/kkg) / kg/m3 *\ = 0.00034 kg/kkg
\1000 mg/1/
and, applying the VPR value of 2.2, the proposed daily maximum
limitation is:
(2.2) (0.00034 kg/kkg) = 0.00075 kg/kkg
E. Nickel; The average plant effluent concentration of
less than 0.074 mg/1 of nickel is slightly less than the
accepted lower limit of treatability (0.10 mg/1) based on
literature data. This lower limit of 0.10 mg/1 was Selected as
the concentration basis for the proposed maximum 30-day average
limitation for nickel. Thus:
(0.10 mg/1) (2.1 m3/kko) / kg/m3 \ « 0.0021 kg/kkg
VLOOO rag/1/
and the proposed daily maximum limitation is obtained by
applying the VFR value of 2.2, that is:
(2.2) (0.00021 kg/kkg) » 0.00046 kg/kkg
F. Silver: For silver, the average effluent data
indicated an achievable concentration of less than 0.067 mg/1.
This is within the range of published treatability values (Table
8-11) and is used as the concentration basis for .the proposed
30-day average effluent limitation. Thus:
(0.067 mg/1) (2.1 m3/kkg) Skg/m3\ = 0.00014 kg/kkg
/ kg/m3 N
VjLOOO mg/1/
and the proposed daily maximum limitation derived from the VFR
value is 2.2 is:
(2.2)(0.00014 kg/kkg) = 0.00031 kg/kkg
G. Zinc: The average plant effluent for zinc is less than
0.15 mg/1. This is greater than the accepted lower limit of
treatability which is approximately 0.02 mg/1. The observed
performance level of 0.15 mg/1 is used as the concentration
basis for the proposed maximum 30-day average limitation of
zinc. Thus;
(0.15 mg/1)(2.1 m3/kkg)f kg/m3 \ = 0.00032 kg/kkg
/ kg/m3 \
\1000 mg/1 /
and the proposed daily maximum limitation is obtained by
applying the VFR value of 2.2, that is:
204
-------
(2.2)(0.00032) = 0.00070 kg/kkg
H. Antimony and Thallium: The sampling data indicate an
average effluent concentration of less than 0,23 mg/1 antimony
and less than 0.17 mg/1 thallium. These relatively high
concentrations are the result of analytical difficulties with
some samples which gave high "less than", results.* Because of
this and the fact that no data are available for the
treatability of antimony or thallium with sulfide/fliter, these
concentrations are being offered as 30-day average maximum
limitations for guidance purposes only.
I. Chromium: The sampling data indicate that plants are
achieving effluent concentrations of less than 0./044 mg/1
chromium. Because no data is available for the treatability of
chromium with sulfide/filter, this concentration is used as the
basis for the proposed maximum 30-day average limitation. Since
there is no treatability data, the limitation is being offered
as guidance and no load limitations (kg/kkg) are presented.
11.7,3 Basis for Proposed BCT Effluent Limitations
The BCT limitation (applicable only to TSS) was set equal
to BPT because the treatment technology for BAT is the same as
for BPT except for dechlorination. Dechlorination does not
affect conventional pollutants.
11•7-4 Basis for Mew Source Performance Standards
For NSPSr the Agency is proposing limitations equal to BPT
for TSS and BAT for other.pollutants because of the prohibitive
cost of additional technology. Pollutants to be limited are pHr
TSS, mercury, arsenic, cadmium, copper, lead, nickel, silver,
zinc and total residual chlorine.
11 -7 *5 Basis f_or Proposed Pr e t reataaent Standards
For pretreatment standards for new and existing sources,
the Agency is proposing limitations based on BAT technology
excluding dechlorination. Dechlorination is unnecessary for
discharges to POTWs because chlorination of influent to
treatment works is common. Pollutants to be limited are pH,
mercury, arsenic, cadmium, copper, lead, nickel, silver and
zinc.
205
-------
11.8 DIAPHRAGM CSLL PROCESS INDUSTRY PROFILE
11.8.1 General Description
Approximately 65 percent of the U.S. production of chlorine
is by diaphragm cell plants. Of 40 known plants, 308 data are
available for 19. Table 11-18 presents a summary profile of the
subcategory. Table 11-19 presents the current status of
discharge regulations for diaphragm cell chlorine plants.
11.8.2 General Process Description
Brine System
As in the mercury cell process, the sodium chloride
solution (brine or salt dissolved in water) is purified before
it is sent to the electrolytic cells. Precipitation of major
impurities with sodium carbonate and sodium hydroxide followed
by clarification generates a brine mud waste which is then sent
to a lagoon or filtered. The settled brine is saturated further
by the addition of salt from caustic evaporators and then is
sent to the cells.
The fundamental difference between diaphragm and mercury
cell brine systems is that unconverted sodium chloride in
diaphragm cell processes is carried with the sodium hydroxide
(caustic) from the cell and is then removed as a solid in
caustic evaporators. In mercury cells the unconverted sodium
chloride is discharged as a spent brine from the cell and
recycled directly through the brine system.
Diaphragm Cell
The treated brine solution is electrolyzed in the diaphragm
cell to form chlorine, hydrogen, and sodium hydroxide according
to the reaction:
2NaCl + 2H20 = C12 + 2NaOH + H2
The diaphragm cell contains a porous asbestos diaphragm
separating the anode from the cathode. Chlorine is liberated at
the anode and hydrogen and hydroxyl ions (caustic) are produced
at the cathode. In the past, the predominant material used for
anodes was graphite with lead used to provide an electrical
contact and support. The lead was joined to the graphite anode
by an organic binder. In recent years, many graphite anodes
have been replaced by stabilized metal anodes made of titanium
with a platinum or ruthenium oxide coating. (An industry
association, estimate is that approximately 49 percent of U.S.
diaphragm cell capacity still involves graphite anodes.) The
206
-------
2W3IE 11-18. SUBCATEGORY PROFILE DMCA SUMMftRY
CHLORENE (DIAPHR&Q4 CELL)
Total subcategory capacity rate
Total subcategory production rate
Number of plants In this subcategory
308 Data on file for
With total capacity of
With total production of
Representing capacity
Representing production
Plant production ranges
Minimum
Maximum
Average production
k
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste •water flew range:
Minimum
Maximum
Volume per unit products
Minimum
Maximum
8,272,600 kkg/year
6,427,000 kkg/year
40
19
6,397,000 kkg/year
4,200,000 kkg/year
77 percent
66 percent
14,700 kkg/year
1,500,000 kkg/year
221,000 kkg/year
103,000 kkg/year
67 percent
4 years
74 years
1,100 cubic meters/day
7,100 cubic meters/day
1 cubic meters/kkg
23 cubic inebers/kkg
Sources of data are Stanford Research Institute, Directory of CJhemical
Producers, U.S.A., 1977, U.S. Department of Commerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessment of Effluent limitations in the
Inorganic Chemical Industry*" June, 1978 and "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards for the Inorganic Chemicals Industry,"
March, 1980.
207
-------
11-19.
STATUS OF FEGULffiTICNS - EFFLUENT LIMITATION GUIDELINES
GHODKENE
SUBP&KE P (40
(DIAPHRAGM
CFR 415.60
CKLTi)
, 3/12/74)
STANDARDS
Product
Process
BPCTC&
Max.1 Avg.2
Para- kg/kkg kg/kkg
meters (mg/1) (rag/1)
BATEA
Max. Avg.
kg/kkg kg/kkg
'(mg/1) (mg/1)
NSPS
Max. Avg.
kg/Meg kg/kkg
(mg/L) (mg/1)
Diaphragm TSS 0.64
tell
Process
Pb 0.005
0.32 No discharge
of pwwp
0.0025 No discharge
of
0.64
0.32
0.00008 0.00004
Section 415.63 was remanded and is presently reserved (41 FR 51601,
November 23, 1976).
"wax. = Maximum of any one day.
Avg.
imum. of daily values for thirty consecutive days.
= process waste water pollutant.
208
-------
advantages of using metal anodes compared to graphite anodes are
increased power efficiency of the cells, longer anode life and a
reduction in potential pollutant loads of lead and chlorinated
organics.
Product Purification
As with mercury cell plants, chlorine liberated at the
anode must be cooled and dried to remove moisture and other
impurities. The cooling generates a chlorine condensate stream
which can be stripped to recover chlorine then returned to the
brine system or discharged. Drying the chlorine gas is
accomplished by scrubbing with sulfuric acid. The resulting
diluted acid can subsequently be regenerated, sold or used for
pH control. When the chlorine gas is compressed and liquified,
noncondensible gases known as tail or sniff gases remain. These
are usually scrubbed with caustic soda or lime generating a
hypochlorite solution which can be sold, used on-site or
discharged, with or without decomposition or treatment.
The sodium hydroxide or caustic from,the diaphragm cell has
a concentration of about 14 percent NaOH and a sodium chloride
content as high as 17 percent. The caustic is usually filtered
to remove some of the impurities and then evaporated to 50
percent NaOH by multiple effect evaporators. Sodium chloride
remains as a solid salt which is then returned to the brine
system. Further purification of the caustic is necessary for
some applications (such as rayon production) and extraction or
adsorption techniques have been used to remove small amounts of
impurities. The caustic can be evaporated further if more
concentrated products are required. The vapor evolved from the
last of multiple effect evaporators is condensed in barometric
condensers generating contact cooling water, or in surface
condensers using noncontact cooling water.
The hydrogen gas generated in the process can be vented or
cooled by refrigeration to remove water vapor before sale or use
as a fuel.
Figure 11-14 is a general flow diagram for the manufacture
of chlorine by the diaphragm cell process.
11.9 WATER OSS AND WASTE WATER SOURCES
11.9.1 Water Use
Water use at diaphragm cell plants is similar to that at
mercury cell plants with one exception. Common uses include
noncontact cooling, tail gas scrubbers, cell wash, equipment
209
-------
LI HE
i
SU1FATB gmi,
PURGE
HO
I
t t
COOMHS
TOHBR"
to
l_l *
O I
SLOWDOWN
TO WASTE
WATER
t
LEAKS,
SPILLS
WASIIDOWH
ETC.
t
TO WASTE
I DR1KB
HCOOTACT •
COOLING -W
WATER ^
11*
ECTC1B
f
PtMIFieWMOH
SKSTEH
t
DIAFHRAGH
CELL
1
ro 121 soDim
HYDROXIDB
SOtUTIOH
t
EWkPORMOR
ran-rur- ijjm.t ^.*m wyTTR
DRIIW BUUS »• ^gnjuj HOHCOWACT COOMMG HATBB
t 4
COOIBB OR usg
WATER
— BAROHBIRIC
^ COHDEHSER
t
COOMNG
HAfEE
,_£*
COOLER I ^ CHUJRlHATBn HATER COHDBNSATB
* f *
SALT
REMOVAL
!
SODMIH
MSDROXIDE
SOLUTION
1
FIWER
SUIHJMC-^
«nn
f •MMB) f 4
— — «""* iwwijK iX)!)lUH
r*"W4PMKKK
-------
maintenance, floor washings and filter backwashing. The
exception at diaphragm cell plants is the use of wate'r for
barometric condensers in the evaporation of caustic.
11.9.2 Waste Sources
Brine Mud
As with mercury cells, this is the waste produced during
purification of brine before it is introduced into the cells for
electrolysis. It consists of precipitated hydroxides and
carbonates of calcium, magnesium, iron, and other metals. The
mud can be a major source of solid waste depending on, the purity
of the raw salt used. At diaphragm cell plants brine muds are
filtered or settled in lagoons. The solids are landfilled and
the filtrate or overflow is discharged or recycled to the brine
system.
Brine mud is the major source of solid waste at chlorine
plants, and discharges range from 0.04 to 1.5 cubic meters per
metric ton (m3/kkg), with a solids content of from two to 20
percent.
Cell Room Wastes
These wastes include leaks, spills, area washdown and cell
wash waters. At diaphragm cell plants cell wash waters are
heavily laden with asbestos and are therefore settled and/or
filtered before chemical treatment or discharge. At plants
using graphite anodes in the cells, the cell room wastes also
contain lead. Data from diaphragm cell plants indicate a waste
flow from 0.02 to 1.2 m3/kkg from cell room operations.
Chlorine Cooling Condensate
Condensation from the indirect cooling of cell gas is
contaminated with chlorine. The chlorine is removed (stripped)
or recovered from the stream before discharge or recycle.
Condensate flows from three plants range from 0.16 to 0.9
m3/kkg.
Spent Sulfuric Acici
Concentrated sulfuric acid is used to dry chlorine gas
after the first stage of cooling. Once diluted to 50 to 70
percent, the spent acid can be regenerated, sold or used for pH
control.
211
-------
Tail Gas Scrubber Liquid
The uncondensed chlorine gas from the liquefaction stage is
scrubbed with sodium or calcium hydroxide producing the
corresponding hypochlorite. The hypochlorite can be used in
other processes, sold, decomposed, or discharged. The amount of
tail gas scrubber water generated at diaphragm cell plants
ranges from 0.1 to 0.29 m3/kkg.
Filter Backwashes
Backwashing of filters used to treat brine before it is
sent to the cells at one graphite anode diaphragm cell plant
generated a waste water flow of 0.45 m3/kkg. Backwashing of
filters used to clarify caustic product at the same plant
resulted in an average flow of 5,4 m3/kkg. At some diaphragm
cell plants these waste waters are partially recycled to
process.
The relatively high flow of caustic filter backwash is due
to the need to remove sodium sulfate, an impurity in the
caustic. Sulfate ions, if allowed to accumulate in the brine
system at graphite anode plants will interfere with cell
performance.
Hydrogen Condensate
Cooling of hydrogen gas for use or sale produces a
condensate stream which can be discharged. Although no data are
available on the volume of this flow, it is small.
Barometric Condenser Waste Water
When vapors from caustic evaporators are contact-cooled, a
significant amount of waste water can be generated. Plows of
from 90 to 300 m3/kkg have been reported at facilities where
barometric condenser water is "once through" and not
recirculated. Recirculation of barometric condenser water
requires a cooling step and a blowdown discharge. A number of
facilities are accomplishing this with a corresponding reduction
in water use. The necessary blowdown of recirculating
barometric condenser waste water at two plants ranges from a
flow of 0.82 m3/kkg to Ol89 m3/kkg.
Summary of Waste Water Flows
Table 11-20 summarizes unit waste water flow data available
by specific sources. A separate list of flows at one graphite
anode plant is presented to compare waste water generation
between metal anode and graphite anode plants.
212
-------
TABIE 11-20. WaSTE WATER FLOWS AT DIAPHRAGM CELL CHDDRINE PLANTS
Stream Description
Flow (nP/kkg)
Plants with Plant with
jfetal Anodes Graphite Anode
min. avg. max.
room wastes
and cell wash
Chlorine Condensate
Spent Sulfuric Acid
Tail Gas Scrubber
Caustic Filter Wash
Brine Filter Backwash
Caustic Cooling Slowdown
Brine Mud
0.02 0.38 0.67
0.16 0.49 0.90
0.01
0.10 0,17 0.29
NA.
0.82 0.86 0.89
0.04 0.42 1.5
1.2
0.78
NA
0.11
5.4
0.45
NA.
NA
NA: Not Available
213
-------
11.10 DESCRIPTIONS OF SPECIFIC PLANTS
The following descriptions of plants includes those plants
that were sampled during the screening and verification program.
The discussion primarily covers plant practices in waste water
control and treatment. Plants were selected for screening and
verification sampling because they were representative of the
industry in that they included a wide range of sizes and
variation in process detail.
11.10.1 Screening
At Plant f014r visited during the screening program, the
chlorine condensate is stripped with steam to remove and recover
chlorine. Brine precipitates (muds) are land disposed, while
the spent sulfuric acid and scrubber solutions are used at an
adjacent plant. The condensate from the hydrogen cooler is used
as makeup water for a cooling tower system, and the condensate
from the evaporative concentration of sodium hydronide is used
to dissolve salt reclaimed from the concentration process. The
cell washings are sent to a collection pond where asbestos and
other suspended solids are removed. In Figure 11-15 the general
process flow sheet is presented. The waste streams sampled and
their waste loadings are presented in Table 11-21.
11.10.2 Verification
Four plants were visited and their waste streams sampled
during the verification program. The results of analysis of the
waste waters are presented in Table 11-21.
At Plant f261, the cathode wash water is passed through a
filter and the asbestos drummed and disposed of in an off-site
landfill, while the filtrate goes to the sewer. Brine
purification muds at this facility are utilized for their
alkalinity on-site and then they are settled prior to discharge
of the supernatant. Spent sulfuric acid is used for
neutralization of waste waters. Dechlorination of the drying
acid by reaction with sodium bisulfite is planned in the near
future. Figure 11-16 shows the process flow diagram and
sampling points.
Plant |738 has two production lines, 738A and 738B, that
are almost identical. At the new plant (738B) the NaOH is not
concentrated nor is the waste from the chlorine disposal system
scrubbed. In addition, the inert gases from the liquefaction
step are put through the chlorine disposal system. The process
flow sheets are shown in Figures 11-17 and 11-18.
214
-------
ro
H
tn
BW3
•»«
1 1
BRINE MUD
VENT GAS
1
1,00000001 *"«2 •Bffl.GBSIg.
98% H^SO, ** SCMUMHiH
4, A0 i3^
T
-------
TABLE 11-21. POEiOTSNT CmCENTRftTIONS AND LOADS AT
AND VppIPICRTIOP PLANTS
stBCASEoany CELCRINE DIAPHRAGM CEO
Plant &
Stream _ Stream
No. Description (mg/1)
#014
3
4
5
6
§261
1
2
3
4
5
#738A1
2
3
4
5
I738B6
7
8
9
10
11
12
13
14
0-2 condensate
Cell.wash
Brine mud
Bar. condenser
Brine mud
Cell wash
Asbestos filtrate
Filter cake
Bar. condenser
Cell room waste
Asbestos wash
Hypo scrubber
0-2 cooling water
Caustic cooling
tower
Cell room waste
Asbestos wash
Hypo scrubber
€3-2 cooling water
Caustic cooling
wtower
Chlorate sump
Plant effluent(B)
Final effluent
(Total)
Brine mud
.2
1600
NA
7
NA
4800
9
NA
6
27
57
290
35
48
95
72
160
20
4.7
32,
63
58
270
«r
TSS
(kg/kkg)
1.
2.
1.
1.
7.
2.
2.
4.
4.
8.
1.
1.
3.
7.
5.
4 x
4 x
NA
3.6
NA
8 x
NA
NA
NA
4 x
0 x
7 x
2 x
3 x
5 x
3 x
4 x
7 x
8 x
0 x
7 x
NA
NA
10
10
10
10
10
10
10
10
10
10
10
-3
—2
-1
-3
-3
-2
-1
-2
-3
-3
-2
10-2
10
10
10
-3
-3
—1
Lead
(mg/1) (kg/kkg)
0.
0.
0.
0.
0.
2.
0.
42
< 0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0055
26
72
005
36
0
075
010
077
031
18
28
51
067
13
20
20
< 0.010
< 0.
0.
0.
0.
010
12
078
10
5.
3.
1.
0x10
9x10
3x10
1.5x10
3.
7.
3.
3.
1.
-6
-6
-5
-3
OxlO"4
6x10
NA
NA
NA
9x10
8X10
7x10
1.3x10
4.
3.
1.
1.
1.
< 8.
< 2.
1.
5x10
2x10
5x10
7x10
7x10
2x10
3x10
1K10
NA
NA
-5
-6
-6
~5
-4
-4
-6
-5
-5
-5
—6
-6
-3
(Continued)
216
-------
TABLE 11-21 (continual)
Plant & Stream
Stream No. Description
#736
1
2
3
4
5
6
7
#9671
2
3
4
5
6
7
Cell wash
Cell room drain
Brine mud
50% Bar, condenser
70% Bar. condenser
95% Bar. condenser
* Chlorine condensate
Cell bldg wastes
Lead pond effluent
Caustic backwash
Brine backwash
Cell wash
Condensate and 112804
Scrubber waste
TSS
(mg/1) (kg/kkg)
934
283.5
20,000
32
21
90.33
2.4
1000
54
« 160
13,000
310
1100
270
6.0 x 10~2
4.6 x 10~3
33
NA
NA
NA
3.9 x 10""4
1.8 x 10"1
3.0 x 10"2
8.6 x 10-1
5.8
5.6 x 10~2
8.7 x 10-1
1.2 x 10~2
Le
(rag/1)
0.014
0.17
0.019
0.010
0.010
0.010
0.010
680
29
0.32
0.52
48
0.92
0.67
ad
(kg/kkg)
9,lxlO~6
2.8xlO~6
S.lxlO""5
NA
NA
NA
1.6xlO"6
1.2X10"1
1.6xlO"2
1.7xlO~3
2.3xlO~4
8.6xlO~3
7.3xlO~4
2.9xlO~5
NA: Not Available
217
-------
SODIUM
CAR90KATE H»OH SALT
1 i '
HAW ... Hirrn<: ^ CLARI- ^ SAND KlS», grrriiUTfit -^^X—
IMKI "*" "' " FLOCCULATORS ^ FILT£RS ~^j^ SATURATOR -J^—
I BACK
4*^f| ,rWASH
£ -{^ BRINE RECYCLED
f SETTLER TO PROCESS
MINE HUBS 1
ADJACENT PLANT w
HjSO^ SLUDGE TO LANDFILL
f COOLING WATER
i — I — j C\
-, - NfX
CI2 ff
1 ' S
" — -— • UQ f(jH
__ CELL
*1 ROOK
GD'
CELL
WASH ^~"
) ^.HYDROGEN TO
^
^COOLING WATER
BEHSATE
n
•"""" """** FIL'IER
STEAM
1 1»> •»
MULTIPLE EFFECT
EVAPORATOR
V CI2 CONDENSATE CAUSTIC j
SPENT HjSOfc UATER Tfl AOjACEHT f (/-
tO PLANT FOR USE fc
^ UVPOCHLORIT
TOWER
"*• PURIFICATION PURIFICATION », LIOUrFlc.
^ BOTTOM TOWER *" TOWER ^ LIQUEFIER _ ^
_ 1 ' f Cl, STORAGE TANK
^ ' "" "' "~ " IHLUHINk AND RAILROAO CAR
, , TO STORAGE WASHOOWN
PURGE FOR DISPOSAL
BY .CONTRACT
LE6EMP
^•^ SAHPLIN6 POIHTS
E -*-T°tfMT COOL.C
UATER
HVPOCHLORITE STEAM
SOLUTION TO 1
ADJACENT PLANT V 1
m e«
EVAPOF
£
(
U5TIC i,,,,™ £
ATOR
NaOH
BOILER
«
T l^\ ^ FILTRATE TO
'" '\±l ^"PROCESS SEWER
/•-^ ASBESTOS TO
Ij^-r-"**' LANDFILL
|li ^ TO
| ^"EJECTOR
BAROMETRIC COOLING
CONOENSEII wut
TO BOILER
<^N _ I FEED WAT£R
^K —tot T0 5L*K£ LIH£
* ^"(rOR LIHI PLANT}
^-PROCESS SEWER
„ . c rf_ COOLING
' WATER
^^ PROCESS
^ SEWER
Figure 11-16. General process flow diagram at Plant #261 showing the sampling points,
Chlorine/Caustic (Diaphragm Cell) manufacture
-------
tvJ
BRINK
73% NaOH •<
LEEESP
•« 8aif>Ung points.
(TYiint {12 - waste aim> (ccnfaination o£ all wastes) . )
Figure 11-17. General process flowsheet at plant I738-A- stowing the sampling points.
Chlorine/caustic (diaphragm cell) manufacture.
-------
VEW
BRINE
to
w
o
Figure 11-18. General process flow diagram at plant #738-B showing the sampling points.
Oilorine/caustic (diaphragm cell) manufacture.
-------
Plant §736 has installed demisters to control the vapors
evolved from the last stage of the evaporator during the
concentration of caustic soda. In this treatment, the steam
evolved from the concentration of cell liquors passes through
metal wool filters to reduce entrained solids. The cell room
washings are sent to a settling chamber and the settled asbestos
is sent to a landfill. The other waste waters, consisting of
caustic evaporator washings and wastes from salt separation,
brine purification operations, and caustic filtration backwash
waters, are combined and sent to one of two settling ponds.
Skimming devices on the settling ponds remove any oil that
separates, while the settled solids in the ponds are dredged and
disposed of in an abandoned brine well. Figure 11-19 shows the
process flow diagram and sampling points.
Plant 1967 uses graphite anodes in its diaphragm cells.
The cell washings at this plant are sent to an asbestos pond
that has a continuous cover of water. Periodically, the settled
solids are removed, sealed in drums and disposed of in a
landfill. The overflow from the pond is treated with soda ash
to precipitate lead, and then filtered. Sulfuric acid is used
to bring the pH down to the 6 to 9 range. Figure 11-20 is a
general process flow diagram for Plant |967.
11.10.3 Descriptions of Plants Hot Sampled
At Plant 1999 brine mud and other streams with high
suspended solids are collected and filtered with leaf filters.
The cake is disposed of in a landfill and the filtrate returned
to the brine system.
At Plant |326, waste water from the diaphragm cell process
is combined with other process waste waters. The combined waste
water is sent to two settling tanks in series. In one of the
settling tanks, skimmers have been installed to remove oil and
the overflow from the second is filtered before discharge.
At Plant 1589, the brine mud from the clarifier underflow
is sent to a brine mud settling pond. The overflow, which is
mostly brine, is returned to the process. The cell room
washings are sent to a settling pit and the settled asbestos
fibers are removed by the use of a vacuum truck, and disposed of
in a landfill. The chlorine from the cells is contact-cooled
with the tail gas scrubber water. The resulting waste water is
steam stripped for chlorine recovery before discharge.
At Plant 1741, chlorine, caustic soda, and potassium
hydroxide are produced using both mercury and diaphragm cells.
Mercury-bearing effluent at this facility is treated by sulfide
precipitation. Tail gas absorption wastes are treated by
catalytic decomposition by a process which consists of scrubbing
221
-------
MH08.
N)
to
cait/ct ASBESTOS
wrex 10 EOUD
TORIVEH W3IE
DISPOSAL
Figure 11-19. General process floi'j diagram at Plant #736 showing the sanpling points.
CMorine/Caustic (Diaphragm Cell) manufacture
-------
TO PChtR HOUSE
NJ
Figure 11-20. Ger^ral process flow diagram at Plant 1967 showing the saropling points.
CMorine/Caustic (Diaphragm Cell) manufacture
-------
with caustic soda solution and treating the resulting
hydrochlorite solution with nickel chloride and iron chloride.
Consumption of iron and nickel chloride is approximately equal
and consists of 0.01 kilogram per metric ton of chlorine
produced. The catalytic decomposition proceeds relatively
slowly, and wastes are retained in the treatment tanks for
approximately three days, after which time no residual chlorine
is reported to be present (3).
11.10.4 Toxic Pollutant Concent r at ions
Analytical Data Base
Section 5.1.2 of this report describes the methodology of
the screening and verification sampling program. In the
chlorine diaphragm cell industry, a total ,of 15 days of sampling
were conducted at Plants #014, #261, f738, f967, and #736.
Thirty-seven different sampling points were involved covering
various raw waste streams and the treated effluents at these
plants. The evaluation of the toxic metal content of these
process related waste streams was based on 975 analytical data
points. The sampling for toxic organic pollutants at Plants
1014 and |967 generated 2300 analytical data points. Analysis
of waste for asbestos generated -an additional 13 data points,
Asbestos
Asbestos, used as a diaphragm separating the cell anode and
cathode, is the major toxic pollutant consistently found in
process waste water from diaphragm cell plants. It occurs
primarily in wastes resulting from activities such as cell room
washdown and cell repair and cleaning.
Table 11-22 presents the results of asbestos determinations
of supply water and waste waters at three diaphragm, cell plants.
Results are expressed as total fibers per liter (in millions) as
well as crisotile and amphibole fibers per liter.
There ar.e no standardized analytical techniques and no
definition of asbestos in water. Because of this, EPA is
excluding limitations for asbestos from these proposed
regulations and deferring regulation to a later date.
Toxic Metals
Table 11-23 presents maximum daily concentrations of toxic
metals found in raw waste samples during the screening and
verification of diaphragm cell chlorine plants. Maximum
concentrations observed at one graphite anode plant are
presented separately. It is clear that except for lead, toxic
metals concentrations at the graphite anode plant are
224
-------
-TABLE 11-22. RESULTS OF ASBESTOS SAMPLING AT DIAPHRAGM CELL PLANTS
Plant
#261
#736
#967
Stream
Supply
Cell Wash
Filtered Discharge
Barometric
Condenser
Supply
Cell Wash
Cell Room Waste
Barometric
Condenser
Barometric
Condenser
Barometric
Condenser
Supply
Cell Ifeste
Pond Effluent
Caustic Wash
Brine Filter
Backwash
Cathode Wash Waste
Condensate & Spent
Acid
Neutralizer Waste
Total Asbestos
Fibers (MFL)*
8.0
2.1 X 108
1.6 X 103
0.4
• 0.7
2.0 X 107
2.9 X 102
1.8
5.3
1.4 X 102
9.7 X 102
2.4 X 104
2.4 X 103
7.8 X 103
8.0 X 102
3.2 X 105
2.7 X 102
2.1 X 103
Chrisotile
MFL
7.5
2.1 X 108
1.6 X 103
0.4
0.7
2.0 X 107
2.8 X 102
0
5.3
1.4 X 102
9.7 X 102
2.4 X 104
2.4 X 103
7.8 X 103
6.2 X 102
3.2 X 105
1.8 X 102
2.1 X 103
Amphibole
MFL
0.4
0
0
0
0
0
8
1.8
0
0
0
8 X 102
0
0
1.8 X 102
0
8.9 X 10
0
*Million fibers per liter
225
-------
TABLE 11-23. R?W OF TOXIC AT
DIAPHR&Qd CELT. CHLORINE PLSNTS (mg/1)
SUBCMEGOBX'
Toxic
Metal
jtotimony
Arsenic
Beryllium
Cadmium
Qiromium
Copper
Leac1
Mercury
Nickel
Selenium
Silver
OSiallium
Zinc
CHLORINE DTKPHBMM CELL
Plants with
Metal Jtaodes
<0.25
0.17
<0.014
0.037
7.4
17
2.0
<0.003
22
<0.020
0.018
<0.25
3.0
Plant with
Graphite Anode
<0.065
0.59
<0.001
0.017
<0.048
0.27
44
0.004
0.070
O.Q30
<0.016
<0.050
0.25
226
-------
essentially no higher than at the metal anode plants. Because
several waste streams usually contribute to the total raw waste
at chlorine plants, a calculation was often necessary to
determine the pollutant concentrations that would exist when the
streams were mixed prior to treatment. An example of this
calculation is the "mixing" of the following hypothetical
streams:
Stream A: 100 gallons per minute, 15 mg/1 of pollutant
Stream B: 10 gallons per minute, 60 mg/1 of pollutant
The weighted average for the mixed streams is given by:
Concentration of mixed stream =
(Flow A x Concentration A) +(Flow B x Concentration B)
(Plow A + Flow B)
Substituting numerical values gives:
(100 gpm) (15 mg/1) + (10 gpm) (60 mg/1) « 19 mg/1
110 gpm
This method was used to calculate raw waste concentrations
of pollutants as presented in Table 11-23. Barometric condenser
waste water when "once through" was not included because of the
high dilution effect of these large flows. Brine mud flows were
also not included.
The daily raw waste loads were calculated from the waste
stream flow rates measured or estimated at the time of sampling
and the measured pollutant concentration.
That is,
Daily loading (as kg of pollutant per day) = (C) (Q)
1000
Where:
C is the concentration of the pollutant expressed in unit
of mg/1 (Note: kg/m3 = 1000 mg/1) and
Q is the waste stream flow rate expressed in units of
m3/day (m3, a cubic meter, is equal to 264.2 U.S. gallons).
Similarly, the unit loadings were calculated from the
reported chlorine production rate, the waste stream flow rate,
and the measured pollutant concentration.
221
-------
Unit loading (as kg of pollutant per kkg (C) fQ)
of chlorine = 1000 P
Where C and Q are as described above, and P is the chlorine
production rate expressed in units of kkg/day (kkg is 1000 kg, a
metric ton, which is equal to 2205 Ibs).
The minimum, average and maximum values were calculated
based on data from those plants where the particular pollutant
was found at a detectable concentration.
In Table 11-24, the toxic pollutant raw waste data are
presented as the average daily concentrations (based on three
24-hour samples) and the unit loadings found at the individual
plants. Beryllium, selenium, and thallium are not included in
the table because average concentrations were below detectable
limits.
In'Table 11-25 plant average daily and unit loadings are
presented as minimum, average, and maximum values based on data
presented in Table 11-24 for metal anode plants only. (The
graphite anode plant is considered separately due to its
particular waste source characteristics.)
Based on the average waste loads generated per unit of
product at metal anode plants and one graphite anode plant, and
the estimated total subcategory production, the estimated total
pollutant raw waste loads generated each year by this
subcategory are as follows:
Raw Waste Load
Pollutant kg/year
Antimony 483
Arsenic 6,300
Cadmium 41
Chromium 3,100
Copper 4,400
Lead 470,000
Mercury 48
Nickel 3,600
Silver 5
Zinc 5,100
Because cell room wastes including cell or cathode wash
wastes, leaks, spills and washdown are usually treated
separately at diaphragm cell plants and because other process
wastes such as filter backwashes, condensates and caustic
evaporation wastes are usually discharged after the settling,
228
-------
11-24.
TOXIC AND LOADS AT AND
VERIFICATION PLANTS
(mg/1)
(kg/kkg)
SlBCMEGOKf
Pollutant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Zinc
CHLORINE DIAPHRAGM nKT,T,
014
*
*
0.002
0.0000018
0.019
0.000017
0.015
0.000014
0.006
0.0000045
' 0.002
0.0000018
0.90
0.00081
*
*
Plant!
261 738A
*
0.17
0.0000064
0.037
0.0000014
1.9
0.000071
17
0.00064
2.0
0.000075
*
22
0.00081
0.018
0.0000007
'1.5
0.000054
*
*
*
*
0.52
0.0046
0.045
0.00039
0.082
0.00060
*
0.21
0.0018
*
0.29
0.0021
738B
*
0.011
0.000021
*
0.066
0.0012
0.12
0.00023
0.11
0.000021
*
0.067
0.00013
*
0.093
0.00018
736
0.010
0.0000033
0.057
0.000014
0.025
0.0000061
0.18
0.000044
0.43
0.00011
0.016
0.0000039
0.003
0.0000007
0.22
0.000054
*
3.0
0.00074
967**
0.011
0.00015
0.30
0.0021
*
0.004
0.000032
0.16
0.0011
21
0.015
0.002
0.000014
0.068
0.00049
*
0.19
0.0014
* Below measurable concentrations
** Graphite Anode plant
229
-------
11-25.
OP
TOSK LQMXDKS KT SCREENING ABD VBRIFICKHCK MEEM. MJQEE EMNTS
SOBaflEQQfS
Pollutant
£* Antimony
o
Arsenic
Cadmium
Chromium
Copper
lead
Jfercury
Hickel
Silver
Zinc
min.
0.00077
O.OOlfl
0.00041
0.0042
0.0035
0.00090
0.00016
0.0066
0.00021
0.016
Loading
(kg/kkg)
avg.
0.00077
0.0084
0.00076
0.59
0.12
0.094
0.00030
0.31
0.00021
2.1
CHJJ08INE DI
AHnasicEii
Unit Loading
(kg/kkg)
HBX,
0.00077
0.020
0.0014
2.8
0.19
0.37
0.00044
1 1.1
0.00021
8.0
min.
0.0000033
0.0000064
0.0000014
0.000017
0.000014
0.0000039
0.0000007
0.000010
0.0000017
0,000054
aag.
0.0000033
0.000017
0.0000032
0.00096
0.00020
0.00016
0.0000012
0.00057
0.0000007
0.00078
mix.
0.0000033
0.000030
o.oooooei
0.0046
0.00064
0.00060
0.0000018
0.0018
0.0000007
0.0021
*Nuntoer of
Plants
Averaged
(out of 5)
1
3
3
5
5
5
2
5
1
4
Cnly those plants where the pollutant was observed at measurable concentrations.
-------
these two waste mixes were evaluated separately. Table 31-26
presents average raw waste concentrations and loads of toxic
metals found in cell room wastes at the six diaphragm cell
plants sampled. Table 11-27 presents the similar data from the
sampling of other process wastes at these plants.
Toxic Organic Pollutants
The use of graphite anodes at chlorine plants results in
the generation of a variety of simple chlorinated hydrocarbon
compounds as a result of the attack of chlorine on the anodes.
These compounds are carried out of the cell with the chlorine
and find their way into the various waste streams which
originate from the chlorine cooling, drying, compression, and
liquefaction steps.
Table 11-28 presents the toxic organics that were observed
in measurable concentrations in the raw wastes at Plant |967.
The concentrations presented in the table were calculated as a
mixture of all raw waste streams weighted on a flow basis as
previously described.
Table 11-29 presents the concentrations of toxic organics
by individual raw waste stream at Plant $967. It is clear from
the table that the highest concentrations of organics occur in
wastes from chlorine treatment (condensate, drying acid and tail
gas scrubber water) and they account for 83 percent of the total
organic waste load.
11.11 POLLOTION ABATEMENT OPTIONS
11.11.1 Toxic Pollutants of Concern
Lead occurs in high concentrations in the cell room waste
waters of chlorine plants using graphite anodes. Other toxic
metals often found in ,significant concentration at diaphragm
cell plants include arsenic, cadmium, chromium, copper, nickel,
and zinc. Antimony, mercury, and silver were also detected but
at concentrations that are not treatable. These metals are not
considered further. The sources of these metals may be raw
material impurities or corrosion products from the reaction
between chlorine or acid and the process equipment materials of
construction.
Toxic organic compounds also occur in waste waters from
graphite anode plants because of the attack of chlorine on the
anode material. They appear primarily in waste streams
associated with the purification of chlorine.
231
-------
to
Ul
to
TflBIE 11-26. TOME MEEM, CCNCEOTRfimJOE MD LORDS IN CELL EDOM HRSEB HMERS KS SCREaflNG flND
ELMHS/ ngA \
\kgAkg;
Pollutant
Mtimony
Arsenic
Cadmium
Qircniuu
flapper
Lead
ffercury
Hictel
Silver
Zinc
014
*
0.010
0.0000001
*
0.94
0.000014
0.53
0.0000075
0.26
0.0000039
*
54
0.00081
*
*
261
*
0.17
0.0000064
0.037
0.0000014
1.9
0.000071
17
0.00064
2.0
0.000075
*
22
0.00081
0.018
0.0000007
1.5
0.000054
Plant f
73B&
0.050
0.0000081
*
*
*
0.24
0.000042
0.044
0.0000077
0.003
O.OOOOOOS
*
*
0.046
0.0000080
738B
*
*
*
0.075
0.000012
0.38
0.000061
0.11
0.000018
*
0.061
0.0000098
*
0.46
0.000074
736
0.038
0.0000031
0.17
0.000014
*
0.54
0.000044
1.1
0.000090
0.047
0.0000038
0.002
0.0000002
0.67
0.000055
*
0.58
0.000048
967**
0.41
0.00015
0.45
0.00017
0.016
0.0000059
0.086
0.000032
2.4
0.00089
370
0.14
0.001
0.0000004
0.36
0.00013
*
0.92
0.00034
Below detection limits
** Graphite anode plant
-------
TSffiE 11-27.
RAH WASTE TOXIC METALS CCKCEKTEATIOH AND LOADS IN PROCESS STREWS OTHER THAN CELL BOOM WASTES
EBOM SCKffiKJHG SW> VERHTCKCICH PILOTS
Pollutant
Antimony
Arsenic
Cadniun
Chromium
een»r
lead
Mercury
Nickel
Silver
Zinc
|014
*
*
0.002
0.0000018
*
0,004
0.0000036
*
0.002
0.0000018
0.003
0.0000027
*
*
I738A
*
*
*
0.53
0.0046
0.041
0.00035
0.083
0.00060
*
0.21
0.0018
*
0.29
0.0021
Plant
I738B
*
0.011
0.000019
*
0.065
0.00011
0.094
0.00016
0.11
0.00019
*
0.067
0.00012
*
0,058
0.00010
§736
*
*
0.038
0.0000062
*
0.090
0.000015
*
0.003
0.0000005
*
4.3
0.00070
1967
*
0.29
0.0020
*
#
0.030
0.00020
0.40
0.0027
0.002
0.000014
0.052
0.00035
*
0.15
0.0010
Avg
*
0.15
0.0010
0.020
0.0000040
0.29
0.00014
0.043
0.00014
0.002
0,0000054
0.088
0.00072
*
1.5
0.0037
Below detection limits
-------
TABLE H-28.
RAW mSTE TOXIC ORGANICS AT A GRAPHITE ANODE PLANT
CELL
Pollutant
benzene
carbon tetrachloride
1, 2-dichloroethane
1, 1, l-trichlorcjethane
hexachloroethane
1 , 1 , 2-trichloroethane
1,1,2, 2-tetrachloroethane
cdiloroform
1, 1-dichlorcethylene
2 , 6-dinitrotoluene
methylene cWLoride
brcmoform
dicMorcbrcnicinetnane
ctilorcxiibrcsnamethane
hexachlordbutadiene
bis (2-ethylheHyl) phthalate
di~n-butyl phthalate
tetxadiloroethlene
toluene
trichloroethylene
Concentration*
(mg/1)
0.00040
0.023
0.079
0.00014
0.010
0.00040
0.000044
0.085
0.000026
0.000026
0.00056
0.000063
0.035
0.002
0.004
0.00075
0.00078
0.036
0.0030
0.020
Load
(kg/day)
0.0011
0.066
0.23
0.00040
0.029
0.0011
0.00013
0.24
0.000074
0.000074
0,0016
0.00018
0.10
0.0057
0.011
0.0022
0.0022
0.10
0.0086
0.0057
Flow-proportioned concentration
234
-------
TKBIM 11-29.
RIW WaSTE TOXIC OBOTICS BY W5STE WKTER SOURCE M A
MODE PLffiNT
SUBCMEQORZ
CfflJORINE DIM»HRfiGM ORT.T,
Stream
Cell building wastes
Caustic filter backwash
Brine filter backwash
Cell wash
Chlorine condensate and
Spent H2S04
Scrubber waste
Totals
Total Toxic
Qrganics
(fflg/1)
0.126
0.057
0.003
0.20
2.2
0.81
0.30*
Total Toxic
Organics
(kg/day)
0.0093
0.12
0.00050
0.014
0.70
. 0.015
0.86
Percent of
Total Toxic
Organics
1.1
14
0.06
1.6
81.5
1.7
100
Flow-proportioned concentration
235
-------
Asbestos occurs in all waste waters from diaphragm cell
plants, and in large quantities in cell room waste waters when
cells are cleaned and repaired.
11.11.2 Prevailing Control and Treatment Practices
Section 11.10 described specific control and treatment
practices at eleven plants. The prevailing practices at
diaphragm cell plants are to control asbestos wastes by settling
or filtering cell wash, waste waters and to neutralize and settle
all waste waters before discharge. The recycle or reuse of
waste streams is practiced to varying degrees in the industry
depending on plant-specific factors such as raw material quality
and type of anodes used.
Plants using graphite anodes are treating lead-bearing
wastes by chemical precipitation and settling and/or filtration
before discharge.
The control of toxic organic compounds in the waste streams
at graphite anode plants also varies in'the industry. At Plant
1967 where the end use of the chlorine is captive, involving its
direct application to the manufacture of a chlorinated organic
product, the bulk of chlorinated organic impurities are not
removed.
At Plant f!95r where a more purified product is required,
the organics are accumulated in the reboiler of the chlorine
scrubber. The residues are treated batchwise for separation and
recovery of the organic phase materials which are then sold as
feedstock for the manufacture of related products. Prior to
discharge the aqueous phase is vacuum stripped to remove
additional organics and chlorine for recycle. Normally, one
batch of organics is treated per week. After separating each
batch of organics and stripping the residual aqueous phase, the
quantity of waste water discharged is approximately 5.7 m3/week
or 0.8 m3/day. The organic loading in this waste is not known,
however, if the assumption is made that the discharge is
saturated with carbon tetrachloride (CC14) (800 mg/1 8 20
degrees. C) , the waste load would be 0.5 kg/day.
A-lthough the daily mass emissions from the two plants are
likely to be similar and both would require additional treatment
to achieve acceptable discharge levels, the wide difference in
concentrations of the chlorinated organics as well as the manner
in which they are handled would necessitate the application of
an advanced treatment technology specifically suited to each
case.
Where the flow is large and the concentrations are low, the
application of activated carbon adsorption to the collected
236
-------
organic-bearing waste stream at Plant f967 would be capable of
reducing a CC14 mass emission from 0.066 kg/day to approximately
0.03 kg/day, assuming an achievable treatability level of 0.10
mg/1.
In the case of Plant |195, where the volume of waste water
is small but the concentrations of residual chlorinated organics
can be on the order of several hundred parts per million, a more
appropriate removal technology would be steam stripping with an
overhead return to the process. Assuming a treatability level
of 10 mg/1 for CC14 using this technology, its mass emission
could be reduced to approximately 0.001 kg/day.
11.11.3 Process Modifications and Technology Transfer Options
Anode Material
The use of metal anodes rather than graphite anodes
increases cell power efficiency and greatly reduces the
pollutant loads of lead and toxic organics in plant waste
waters. Approximately half of the diaphragm cell production of
chlorine is now by metal anodes.
\
Caustic Evaporation Water
The vapors from the evaporative concentration of caustic
soda are either contact-cooled or cooled in surface condensers.
Plants practicing contact cooling through barometric condensers
generate large amounts of waste water contaminated with caustic
soda and salt. By changing from contact cooling of the vapors
to noncontactf cooling, or by recirculating barometric condenser
water, the amount of waste water generated can be reduced
considerably. If the change is considered too expensive or is
not feasible, demisters_ or similar control devices can be
installed to reduce the salt and caustic carryover in the
vapors.
Diaphragm Material
Although not in full scale use at any U.S. chlorine plants,
modified diaphragms have been developed which can reduce power
consumption and minimize or eliminate asbestos discharges. The
modified diaphragms include polymer membrane and ion exchange
membrane diaphragms.
Polymer Modified Asbestos Membranes - These consist of a
polymer treated asbestos diaphragm baked into place on the
cathode. , Its usage results in power savings and has an
environmental benefit, since, -at the time of rebuilding the
237
-------
cathodes, the discarded material is produced in stablized pieces
instead of loose asbestos fibers. Final disposal is thus safer
and easier.
Polymer Membranes - These consist of a microporous Teflon®
type polymer"andtheir operation has been demonstrated
successfully in laboratory and pilot plant scale cells. In
addition to the benefits of cost savings through energy use
reduction and longer lifer their use eliminates the handling and
disposal problems associated with asbestos.
Ion Exchange Membranes - These membranes allow the
production of a concentrated caustic similar to that produced by
mercury cells. The production of salt-free concentrated caustic
will reduce the waste water associated with the caustic
evaporation process. Like the polymer membranes, the problems
associated with the handling and disposal of asbestos are
eliminated.
Itiquefaction of Chlorine
Utilization of high pressure and refrigeration for chlorine
recovery will reduce the chlorine content in tail gases.
Tail Gas Emission Control
As with mercury cell plants, chlorine in tail gases has to
be removed and treated or recovered before venting to the
atmosphere. The common practice is to scrub the gas with
caustic soda producing a hypochlorite solution. This
hypochlorite can then be sold, used on-site or, discharged.
Decomposition is a common method of removing the chlorine in
this stream prior to discharge. Catalytic, thermal and chemical
methods of decomposition, described in Section 11.4.3, are
effective.
11.11.4 Best Management Practices
The following Best Management Practices are common industry
practices and are provided for guidance purposes although they
may not meet the requirements of the Resource Conservation and
Recovery Act -(as amended, 42 USC 6901 et. seq.).
Area Runoff
Provisions can be made to divert and contain storm runoff
from areas where lead or asbestos contamination could occur.
Collected runoff can then be treated with other wastes.
238
-------
Leaks and Spills
Provisions can be made in cell room areas to control and
collect the leaks or spills contaminated with lead or asbestos.
Contaminated Solids
Asbestos waste and precipitated metals wastes should be
stored in a lined pond or disposed of in a secure landfill.
* 2
11.11.5 Advanced Treatment Technologies
The methods available and currently used in the industry
for the removal of lead and other toxic metals from plant waste
waters include hydroxide or carbonate precipitation followed by
settling or filtration. Further removal of metals can be
effected using sulfide precipitation, adsorption and ion
exchange.
Removal of asbestos from cell wastes is improved with the
addition of coagulating agents prior to filtration of these
wastes,
11.12 SELECTION OF APPROPRIATE TECHNOLOGY AND EQUIPMENT
11.12.1 Technologies for Different Treatment Levels
Level 1 (BPT)
Level 1 treatment addresses the waste characteristics
associated with diaphragm cell plants using graphite anodes.
The data from graphite anode plants were used because the
pollutant load is greater than for metal anode plants. Existing
plants that have changed from graphite anode to metal anode will
have residual effects that increase their loads for an extended
time after the change - possibly as long as two years. Waste
streams from the cell rooms and the cathode wash station
(asbestos pond overflow) are collected in a holding tank where
they are combined with any other process waste sources
containing treatable levels of lead and other toxic metals.
Alkaline precipitation of the toxic metals is accomplished by
the addition of soda ash. The solids are removed by filtration
and the filtrate may be combined with other process waste
streams such as chlorine condensate, tail gas scrubber water,
caustic filter backwash and barometric condenser waters found to
be contaminated with toxic metals at levels usually below the
limits of treatability by alkaline precipitation. Because the
other process water sources are normally alkaline the pH is
relatively unchanged and clarification for suspended solids
239
-------
removal also achieves some additional removal of traces of toxic
metal hydroxides. Thus, the combined flow is clarified and
discharged directly or, in some cases, it may be combined with
noncontact waste waters and passed through a polishing pond
system prior to final discharge. At all levels of treatment,
the brine mud is collected in lagoons and the effluent recycled
to process. The flow diagram for Level 1 treatment is shown in
Figure 11-21.
Level 1 treatment was ultimately selected as the basis for
BPT because it represents a typical and viable industry practice
for the control of asbestos fiber, lead, and other toxic metals
in waste waters associated with diaphragm cell plants using
graphite anodes. Plants utilizing metal anodes are expected to
have lower levels of toxic metal emissions and may not require
alkaline precipitation to meet the proposed BPT limitations.
All 39 plants in the industry presently have BPT or equivalent
treatment technology installed.
Level 2 (BAT)
The objective of Level 2 treatment technology is to
achieve, at a reasonable cost, a greater degree of asbestos
fiber and toxic metals removal than provided by Level 1. Thus,
Level 2 adds dual-media filtration to the combined effluent from
Level 1 treatment excluding noncontact waste streams.
Dechlorination of the final plant effluent is also included in
Level 2 (BAT) treatment. This assumes treatment by sulfur
dioxide or bisulfite to remove total residual chlorine to the
detection limit of approximately 0.2 mg/1. This is a reasonable
value for a waste ^ater sample, since the Iodine Method for
determining total residual chlorine is affected by the color of
the sample.
Level 2 was finally selected as the basis for BAT
regulations on the strength of technology transfer options
within the inorganic chemicals industry and because four out of
five plants sampled were meeting limits derived from published
treatability data. In addition, two plants are known to
practice dechlorination of the final effluent. The flow diagram
for Level 2 is shown in Figure 11-22.
Level 3
The practice of sulfide precipitation of mercury in the
mercury cell segment of the chlor-alkali industry suggested the
application of this technology for achieving greater removal of
toxic metals in diaphragm cell plants. Level 3 adds sulfide
precipitation to Level 2 as shown in Figure 11-23. This option
was not selected due to its relatively high cost per pound of
additional metal removal obtained.
240
-------
r-\
ro
*>.
H
Ds
incite WHK
fcfcstff
OEUittW) .
HDUJINGIM*
rnswMD
SMEGE
moans
1' "»Q »• groan1
n»
IWEFIU,
Includes pH monllorltif?, flaw tnonltoring
and atamptf t-
Figure 11-21. Level 1 wste -water treatmert:--fer chlorine - diaphragm cell subcategory.
-------
p
I
•i
n
< *<
1
i-.
I
J
• I
1
II
II
8
CM
CV
A
242
-------
10
*»
U)
(tew,
r-\
/—n
I SCOllK
r
f
WaMSH
'
Jj WfDUKi 3SHK
i
1
i
i
1
1^
lUXniUTIMB
«>w
1
Icaj.
PO^H • — .^
. MSIES "
1
1
SMSMS
3MUWIB))
,
IBM
IKLC
W
1
'C *
\ /
\ /
\
1
Ml
IMS
NK
**"
i
|r
r*—
-------
11.12.2 Equipment for Different Treatment Levels
Equipment Functions
Conventional sludge dewatering by filter press is used for
asbestos sludge before disposal and dual-media filter backwash
is returned to the influent surge tank. Level three requires
the addition of a reagent mixing tank and chemical solution
feeder. Level 3 treatment requires the addition of a reagent
mixing tank and chemical solution feeder to introduce ferrous
sulfide ahead of the Level 2 dual-media filter. All equipment
is conventional and readily available.
Chemical Handling
Nonhazardous solutions of aluminum sulfate and sodium
carbonate are the only solutions used at Levels 1 and 2. Inert
filter aid used in the alum sludge filter process presents no
unusual hazard. At Level 3 the potential hazard of handling
sodium sulfide is nullified by reacting it with ferrous sulfate
to form ferrous sulfide. Any excess ferrous sulfide will
oxidize to a ferric sulfide precipitate. At the point where
sodium sulfide is reacted with ferrous sulfate, good ventilation
is essential to avoid the hazards associated with hydrogen
sulfide gas.
Solids Handling
For all three levels of treatment, brine mud solids are
accumulated in lined lagoons on-site. Asbestos solids and
precipitated metals wastes are to be sent to suitable chemical
landfills.
11.13 TKE&TMENT COST ESTIMATES
11.13.1 General Discussion
To prepare treatment cost estimates, a model plant concept
was developed. Because higher pollutant loads and larger unit
flows exist at graphite anode plants as opposed to metal anode
plants the characteristics associated with these plants were
used when possible for the model plant characteristics as
discussed below.
The preliminary cost estimates presented, in this report
were based on incomplete industry data on waste source
characteristics and flow rates. The cost estimates assumed a
flow rate of approximately 1.2 m3/kkg for the waste stream from
the cell room^ asbestos pond, and other sources (Table 11-20}.
244
-------
Later revisions of the flow rates and the inclusion of other
waste sources in the total volume of process related waste
streams to be handled brought the total model plant flow rate up
to the currently estimated 8.8 m3/kkg {Table 11-34). This is
the flow used for the development of proposed regulations for
the diaphragm cell segment of the industry. Since cost
estimates were based on the preliminary flow estimate of 1.2
m3/kkg, the Agency is assessing the need for making appropriate
adjustments in the cost estimates. Such adjustments will be
made prior to final promulgation of the regulations. The model
plant specifications given below are those used for regulation
development purposes.
Chlorine Production
Approximately 60 percent of the production data for all
chlorine plants using diaphragm cells is available on file.
Production ranges from 15,000 to 1,500,000 kkg of chlorine per
year. Three model plants with production rates of 19,100,
95,500, and 191,000 kkg per year were selected to represent the
subcategory production range.
Waste Water Flow
Based on industry flow data (Table 11-20), waste streams in
the model plants are segregated into brine mud, cell wash and
cell room wastes, and other process wastes such as filter
backwashes, condensates and tail gas scrubber wastes. For
treatment cost estimates at all levels of treatment the
following flow basis was used.
A. A brine mud flow of 0.42 m3/k*kg is sent to lagoons for
solids removal. Solids are disposed of on-site and other
overflow is recirculated to process.
B. Cathode or cell wash waters, heavily laden with
asbestos are sent to asbestos removal at a flow rate of 0.07
m3/kkg.
C. Cell room wastes consisting of leaks, spills, and area
washdown contaminated with lead and other metals are combined
with treated cell wash waters for a total flow of 1.2. m3/kkg to
be treated for"metals removal.
D. Other process waste water sources account for 'an
additional 7.6 m3/kkg which is combined with effluent from the
treatment of wastes from the cell room and cathode wash areas.
This brings the mod'el plant total flow rate to an estimated 8.8
m3/kkg. The final, combined process waste flow is either
clarified and discharged as in Level 1 treatment or clarified,
245
-------
passed through dual-media filtration, and dechlorinated prior to
discharge as in Level 2 treatment.
Solid Waste Produced
Brine mud constitutes the major source of solid waste
generated at chlorine plants. Although solids content varies
from plant to plant, an average of ten percent solids was used
for the model resulting in a solids load of 42 kg/kkg. Asbestos
from cell wash operations and precipitated solids from metal
treatment generate a solid waste of 0.83 kg/kkg.
Chlorine Bearing Wastes
In the selection of model plants, the following assumptions
have been made for the chlorine contaminated waste streams.
The chlorine condensate waste stream has not been included
in the waste streams going to the treatment facility. In the
majority of the chlorine/caustic plants, this stream is stripped
of chlorine by steam or vacuum and the chlorine is recycled to
the purification operation. The waste water is then returned to
the process and introduced to the brine purification unit or
sent to the treatment unit. The quantity of waste water
generated by this operation is small. In some cases the
chlorine gas from the cells is contact-cooled with water and the
scrubbed liquid, after steam stripping, is reused. The
stripping operation in the recovery of chlorine is part of the
process and, therefore,' its cost is not included in the
treatment cost.
The spent taii gas scrubber solution, which is mainly
calcium/sodium hypochlorite, is assumed to be used or decomposed
before it is discharged. Thermal decomposition can be practiced
at no additional cost at some facilities, while another
efficient method is catalytic decomposition. The cost estimates
for decomposition are not included here because at many plants
the hypochlorite stream is sold, used on-site or only
infrequently discharged, depending on market demand.
However, because of the environmental effects of high
levels of chlorine in waste water discharges, a separate set of
cost estimates have been prepared for the dechlorination of
total plant discharges using sulfur dioxide.
Chlorinated organic wastes - The chlorine-bearing waste
streams at graphite anode plants are also those streams carrying
the highest concentrations of toxic organics as indicated in
Table 11-29. Section 11.11.2 discussed the techniques used to
recover and remove organics from waste streams at Plant f!95 and
the fact that organic contaminated streams can exist as either
246
-------
high volume-low concentration or low volume-high concentration
depending on plant specific factors. Costs for removing
organics are not included in the model plant cost estimates
because organics are not limited in the regulation. Orqanics
occur at low levels at most of the plants and when they are
present the appropriate treatment method is site specific.
Althdugh the costs are not included the following information is
provided as guidance. The additional costs for steam stripping
in a plant (such as Plant |195) which already has a vacuum
vaporizer, would be under $10,000 for modification of the
existing equipment. Steam costs could vary from $1,000 to
$5,000 per year. If a vaporizer is not in place, a steam
stripper to process 5 to 30 m3/week would cost roughly $50,000
to $100,000, depending on the input concentrations to be
handled. The corresponding steam costs would range from $2,000
to $10,000 per year.
The capital costs of an activated carbon adsorption unit
for handling the relatively high volume wastes with a low
influent organic loading (as found ~at Plant 1967) cannot be
reliably estimated in the absence of specific treatability data
on the waste streams in question.
Alternatively, incineration of the chlorinated organic
residuals is an effective means of destroying and disposing of
this material provided that adequate measures are taken to
control the release of HC1 to the atmosphere. •
A process evaluation should be made to determine the most
efficient means for isolating and collecting the organic bearing
waste streams prior to treatment.
Incidental removal of chlorinated organics will occur with
the application of model plant treatment levels previously
presented. Such removal, however, is expected to be erratic and
therefore cannot be predicted. Because raw waste concentrations
of these organics vary considerably depending on plant practices
and are marginally treatable at times, applicable control and
treatment technologies will need to be assessed on a
case-by-case basis.
For these reasons, the Agency is not providing specific
numerical discharge limitations for organic pollutants, but is
providing guidance for evaluating control options that could be
applied in the industry.
11.13.2 Model Plant Treatment Costs
On the basis of the model plant specifications and design
concepts presented earlier, the estimated costs of treatment for
three models having different production levels are shown in
247
-------
Tables 11-30, 11-31, and 11-32. The cost of Levels 2 and 3 are
incremental over Level 1 costs and provide for higher effluent
quality with respect to toxic pollutants.
Table 11-33 presents a summary of the unit cost
distribution between amortization and operation and maintenance
components. Annual treatment costs as a function of production
rate is shown graphically in Figure 11-24. Similarly presented
in Figure 11-25 is the relationship of unit cost (cost per
metric ton) to production rate.
For plants requiring dechlorination of waste waters, cost
estimates for dechlorination of plant effluents using sulfur
dioxide are discussed in Section 11.6.3.
11.14 BASIS FOR REGULATIONS
11.14.1 Basis for BPT Limitations
BPT regulations are currently in effect for the diaphragm
cell chlorine subcategory, 40 CFR 815.62(b) . The Agency is
proposing to revise the limitations, however, based on an
increased unit flow rate.
Technology Basis
For BPT, the Agency is proposing limitations based on
equalization, alkaline precipitation and settling of lead and
asbestos-bearing wastes and neutralization and settling of all
waste waters before discharge. All diaphragm cell chlorine
plants are known to be using this technology (Level 1) or its
equivalent.
Flow Basis
As described in Section 11.13.1, waste water streams at
diaphragm cell plants1 are separated into two types, those that
require treatment for asbestos-and metals removal and those that
do n6t require such treatment. From data presented in Table
11-20, the unit flow rate of 1.2 m3/kkg of cell room and cell
wash wastes from one graphite anode plant was selected as the
flow basis for wastes to be treated. Graphite anode plant data
were used in this instance because the flows were higher than
those of other plants and thus represent a conservative estimate
of flow for other plants in the industry. Using available flow
data the remaining waste streams total 7.6 m3/kkg as shown in
Table 11-34. Thus the total unit flow discharge 'used in the
development of effluent limitations is 8.8 m.3/kkg.
248
-------
TABLE 11-30. MODEL PIANT TREATMENT COSTS
Subcategory CHLORINE Diaphragm cell
Production 19,100 metric tons per year
54 metric tons par day
Waste water flow 68 cubic meters per day
A.
B.
C.
INVESTMENT COST
Equipment in place,
including piping,
fittings, electrical
Monitoring equipment
in place ...«*.,.......
Engineering design
Incidentals, overhead,
fees, contingencies...
TOTAL INVESTMENT COST
OPERATION AND
MAINTENANCE COST
Labor and supervision.
Taxes and insurance...
Residual vaste
Monitoring, analysis
TOTAL OPERATION AND
MAINTENANCE COST
AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
FIRST
$57,100
106,850
9,000
34, 590
34,590
21,000
$263,130
«
$112,000
2,200
1,500
24,213 '
7,893
5,800
15,000
$168,606
$39, 394
$208,000
( ,21,057 tons per
( 60 tons per
*
year)
day )
LEVEL OF TREATMENT*
SECOND THIRD
$1,800 $2,250
17,900 20,400
3,940 4,530
3,940 4,530
$27,580
$14, 000
300
2,758
827
7,500
$25,385
$4,487
$29,872
$31,710
$14,000
300
100
3,171
951
7,500
$26,022
$5,159
$31,181
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
249
-------
TABLE 11-31. MODEL PLANT TREATMENT COSTS
Subcategory CHLORINE Diaphragm cell
Production 95,500 metric tons per year
272 metric tons per day
Waste water flow 340 cubic meters per day.
( 105,288 tons per
( 300 tons per
year)
day )
LEVEL OF TREATMENT*
A. INVESTMENT COST
Equipment in place,
including piping,
fittings, electrical
work and controls.....
Monitoring equipment
Engineering design
Incidentals, overhead,
fees, contingencies. . .
TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Maintenance. ..........
Taxes and insurance ...
Residual waste
Monitoring, analysis
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
FIRST
, $148,100
219,700
9,000
75,360
75,360
63,000
$590,520
$112,000
4,900
7,500
52,752
17,715
29,000
15,000
$238,867
$85,827
$324,694
SECOND
$2,900
27,000
5,980
5,980
$41,860
$14,000
600
4,186
1,255
7,500
$27,541
$6,810
$34,351
THIRD
$3,350
29,500
9
6,571
6,571
$46,002
$14,000
600
500
4,600
1,380
7,500
$28,580
$7,484
$36,064
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost
250
-------
TABLE 11-32 MODEL PLANT TREATMENT COSTS
Subcategory CHLORINE Diaphragm cell
Production 191,000 metric tons per year ( 210,577 tons per year)
545 metric tons per day ( 601 tons per day )
Waste water flow 680 cubic meters per day.
A. INVESTMENT COST
Equipment in place,
including piping,
fittings, electrical
Monitoring equipment
Engineering design
and inspection ........
Incidentals, overhead,
fees , contingencies ...
TOTAL INVES1MENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Energy. ...*..»........
Taxes and insurance...
Residual waste
disposal ..............
Monitoring, analysis
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
DIVESTMENT COST
TOTAL ANNUAL COST
FIRST
$271,900
295,500
9,000
115,280
115,280
123,000
$929,960
$112,000
8,000
15,000
80,696
27,898
58,000
15,000
$316, 594
$131,292
$447,886
LEVEL OF TREATMENT *
SECOND
$4,800
43,500
9,660
9,660
$67,620
$14,000
600
6,762
2,028
7,500
$30,890
$11,001
$41,891
THIRD
$5,250
46,000
10,250
10,250
$71,750
$14,000
^600
1,000
7,175
2,152
7,500
$32,427
$11,673
$44,100
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
251
-------
TABLE 11-33. MODEL PLANT TREATMENT COSTS
Subcategory CHLORINE Diaphragm cell
Annual Treatment Costs ($/kkg)
COST ITEM
PRODUCTION FLOW
(kkg/yr) (m3/day)
LEVEL OF TREATMENT
FIRST SECOND THIRD FOURTH
Annual Operation
and Maintenance
Annual
Amortization
Total Cost
19,100 68 8.83 1.33 1.36 Not
95,500 340 2.50 0.29 0.30 ApplieabL
191,000 680 1.66 0.16 0.17
19,100 68 2.06 0.23 0.27
95,500 340 0.90 0.07 0.08
191,000 680 0.69 0.06 0.06
19,100 68 10.89 1.56 1.63
95,500 340 3.40 0.36 0.38
191,000 680 2.34 0.22 0.23
252
-------
500
400
SOO
u
D
<
200
100
TT
i t I/
f
W I/I "
i
, . I I
Z
I
z
TTT
JZ
/TT/T
z
i r
! i '
Z
/LEVEL, #1
I I !
J I
1
i i
i ' i
L.L
I I
1
J i
I I
i I
I I
50 100 150 200
PRODUCTION (METJUC TONS/YEAR x 1000)
Figure 11-24, Annual treatment cost vs. production for the Chlorine
Subcategory (Diaphragm Cell Process)
253
-------
1 I
I i
TT
J I
I i
12
10
I i
S i
IT
I i
LEV! LS i»2 EC #3
I I
i I
I !
\
TTTv
i i
v\
T\
\r
I !
I TV
i i
i :
: LEA 'EL #11 ;
I I
! i
TT
i i
i !
i i i
I
i !
I
! I
I i
I I I
50 100 150 200
PRODUCTION (METRIC .TONS/XEAR X 1000)
Figure 11-2 5. Annual unit treatment cost vs. production for the
Chlorine Subcategory (Diaphragm Cell Process)
254
-------
11-34
OF UNIT FECWS Jffi DiaPHRSSM CFT,T,
SUBCATEGOKSf
DlftPHRAQf
Stream Description
Onit Plow
Data
Source
Cell room and cell
wash wastes
Chlorine eondensate
Tail gas scrubber waste
Caustic filter wash
Brine filter wash
Caustic cooling blowdown
Spent sulfuric acid
1.2
0.78
0.11
"s.4
0.45
0.86
0.01
Graphite anode plant
Graphite anode plant
Graphite anode plant
Graphite anode plant
Graphite anode plant
Ifetal anode plants
average
Mstal anode plants
average
Total Unit Plow Discharge 8.8
255
-------
Selection of Pollutants to be Regulated
The selection of pollutants for which specific effluent
limitations are being proposed was based on an evaluation of raw
waste data from screening and verification sampling and on the
treatability of toxic pollutants.
Table 11-35 presents the achievable concentrations of toxic
metal pollutants (found at detectable levels in raw waste
streams) using the available treatment technology options.
Based on literature treatability data presented in Section 8.1
and summarized in Table 8-11, the concentrations reflect the
lowest level achievable by these technologies. Also presented
in the table are the maximum three-day average raw waste
concentrations observed during the sampling program with an
indication of the .number of plants where treatable
concentrations were exceeded.
Based on the occurrence of treatable levels of specific
toxic metals, arsenic, cadmium, chromium, copper, lead, nickel
and zinc were selected as candidate toxic pollutants proposed
for- BPT regulations. Antimony, mercury, and silver were
detected but at less than treatable levels.
Basis of BPT Pollutant Limitations
Limitations are presented as both concentrations (mg/1) and
loads (kg/kkg), and the relationship between the two is based on
the unit flow rate (8.8 m3/kkg). The concentration basis
therefore represents the concentration of the total plant
discharge including both treated and untreated waste waters.
BPT proposed limitations are presented in Table 11-36.
Convent ional Pollutants -
A. pH: The treated effluent is to be controlled within the
range of 6.0 to 9.0. This limitation is based on the data
presented in Appendix B of this report^and the JRB Study (52).
B. TSS:-The proposed BPT limitations for TSS are based on a
summary of monitoring data from Plant 1207 (3) . The average
discharge load of 0.30 kg/kkg is used to develop discharge
limitations. Because variability factors for TSS were not
available for this plant, factors obtained from the hydrofluoric
acid subcategory were used. In that subcategory, where the same
technology of alkaline precipitation and settling is used, the
average variability factor for daily measurements of TSS is 3.5
.and the average factor for 30-day averages is 1.7. Thus,
utilizing the long-term average discharge load of 0.30 kg/kkg
one obtains a maximum 30-day average load limit ofs
256
-------
TABLE 11-35. COMPARISON OP TOXIC METALS TREATABILITY WITH SCREENING
AND VERIFICATION SAMPLING DATA
Arsenic
Antimony
Cadmium
Chrcmium
Copper
Lead
Mercury
Nickel
Silver
Zinc
Tree
(i
Level
1
0.5
0.8
0.1
0.1
0.5
0.3
_J3>
0.2
0.4
0.5
ng/D
Level
2
0.5
0.4
0.05
0.05
0.4
0.05
_»)
0.1
0.2
0.4
Plant Raw
Level Waste Average
3 (rag/1)
0.05
NA
0.01
NA
0.05
0.05
0.01
0.05
0.05
0.02
0.30
0.011
0.037
1.9
17
21
0.003
22
0.018
3.0
Number of Plants
Exceeding
Treatability
3
0
2
4
4
4
0
6
0
3
(1) Literature-based treatability estimates from Table 8-11.
(2) Of 6 plants, number exceeding treatability by sulfide/filter.
(Level 3)
(3) Treatability with this technology not available.
NA Not Applicable
257
-------
11-36. PROPOSED LTMITOTIONS
Chlorine - Diaphragm Cell
Best Practicable Control Technology Currently Available
Waste Water Plow: 8.8m3Akg
Pollutant
Subcategory
Performance
(mg/1)
Concentration Basis
24-hr
max.
30-day
avg.
Effluent Limit
(kg/kkg)
Max
30-day 24-hr
avg. max.
Conventional Pollutants
TSS
Toxic
57
2,1
57
120
0.51
1,1
Arsenic
Cadmium
Chromium
Copper
Lead
Nickel
Zinc
0.50^ ' 2.6
0,1Q(3) 2.6
G.10t3) 2.6
0.50(3) 2.6
1.1^) 2.6
OCT f\ V*""/ *"i f~
,50 2.6
0.50{3) 2.6
0.50
0.10
0.10
0.50
1.1
0.50
0.50
1.3
0.26
0.26
1.3
2.9
1.3
1.3
(5)
(5)
0.00088
0.0044
0.010
0.0044
0.0044
(5)
(5)
0.0023
0.011
0.026
0.011
0.011
(1) - VFR: ratio of the 24 hour variability factor to the 30-day variability
factor
(2) - Verification sampling
(3) - Lower limit of literature treatability (Table 8-11): used when observed
sampling data
(4) - Based on long-term monitoring data
(5) - No effluent limitation proposed
258
-------
(0.30 kg/kkg) (1.7) = 0.51 kg/kkg
and a maximum daily limit of:
(0.30 kg/kkg) (3.5) - 1.1 kg/kkg
The concentration bases then are derived by applying the
model plant flow rate of 8.8 m3/kkg to obtain a 30-day average
concentration of 57 mg/1 derived as follows:
(
0.51 kg/kkgN/lOOO ing/A = 57/mg/l
8.8 m3/kkg/\kg/mi
and a daily maximum concentration of 120 mg/1 derived from the
variability factor ratio (VFR: 3.5/1.7 = 2.1) as follows:
(2.1) (57 mg/1) = 120 mg/1
Toxic Pollutants -
A. Lead: The proposed BPT limitations for lead are based
on long-term monitoring data from one graphite anode plant as
presented in Appendix A. The plant is achieving a long-term
average lead discharge of 0.0064 kg/kkg.
Statistical analysis of monitoring data from the plant
established a 30-day average variability factor of 1.6 and a 24-
hour variability factor of 4.1. The ratio of the two
variability factors, VFR, is 2.6. The proposed 30-day average
limitation for lead was then obtained by multiplying the
variability factor for 30-day averages by the long-term average
load; i.e., 1.6 x 0.0064 kg/kkg = 0.010 kg/kkg. Similarly the
daily maximum limitation was obtained by multiplying the daily
maximum variability factor by the long-term average load; i.e.,
4,1 x 0.0064 kg/kkg = 0.026 kg/kkg.
The concentration basis for lead is derived from the
relationship between concentration (C), unit lead (L), and unit
flow (Q) .
C (mg/1) * L (kg/kkg) /1000 mg/1
,Q (m3/kkg)\ kg/m3
Thus the concentration basis for the maximum 30-day average
for lead is:
/
V
0.010 kg/kkg\ /1000 mg/I\ - 1.1 mg/1
8.8 m3/kkg ) \ kg/m3 ~)
The concentration basis for the daily maximum limitation is
obtained -similarly or by applying the variability factor ratio
(VFR) of 2.6 to the maximum 30-day average concentration:
259
-------
(1.1 mg/1) (2.6) = 2.9 mg/1
Monitoring data from six diaphragm cell plants presented in
Table 11-37 indicates that plants using metal anodes are meeting
the BPT lead limitations. One of two graphite anode plants is
meeting the limitations.
The limitations proposed for additional toxic pollutants
are derived from two sources - sampling data and
literature-based treatability estimates. The concentration
bases are derived from effluent sampling unless the observed
concentrations are below treatability estimates. In such cases
the lower limit of the applicable treatability level was used
(Table 8-1).
B. Chromium; Raw waste concentrations of chromium were
observed as high as 1.9 mg/1 (Table 11-26). Table 11-38
presents effluent data from the sampling of two diaphragm cell
plants which indicate an achievable final discharge
concentration of 0.05 mg/1 chromium.' Because this is below the
treatability estimate of 0.10 mg/1 with BPT technology (Table
11-35), the treatability concentration has been used as the 30-
day average basis for deriving BPT limitations for chromium.
Because no long-term monitoring data is available for
chromium in this industry, the same variability factor ratio
(VPR) obtained from monitoring lead in the discharge at one
plant is used to obtain daily concentration limits.
(0.10 mg/1) (2.6) = 0.26 mg/1
To obtain effluent lead limitations for chromium, the 30-
day average concentration is multiplied By the unit flow:
(0.10 mg/1) (8.8 mS/kkg)/ kg/m3 N = 0.00088 kg/kkg
VlOOO mg/1/
and the daily maximum effluent limit is
(0.26 mg/1) (3.8 m3/kkg)/ kg/m3 N = 0.0023 kg/kkg
V1000 mg/1/
C. Copper, Nickel, and Zinc: Raw waste concentrations of
these metals were observed as high as 17 mg/1 copper, 22 mg/1
nickel, and 3.0 mg/1 zinc. Table 11-35 indicates an achievable
final discharge concentration of less than 0.10 mg/1 for these
metals. Because this is below the literature-based treatability
estimate of 0.50 mg/1 using BPT technology (Table 11-35), the
treatability concentration has been used as the 30-day average
basis for deriving BPT limitations for these metals.
260
-------
TABLE 11-37.
LEAD AND TSS DISCHARGES ERCM SELECTED DIAPHRAGM CELL
CHLORINE PLANTS ^
SUBCATEGORY
CHLORINE - DIAPHRAQ1 CRT.T,
Plant
Lead Discharge
kg/kkg
.Average
Maximum
#589*
#738*
§261*
#014*
1967 (3)
#207
Plant
#014*
1207
0.0020
0.0010
0.0025
0.0060
0.0064
0.021
TSS Discharge
kg/kkg
Average
2.8^
0.30
0.030
0.015
0.019
NA
0.026
0.054
Maximum
NA
0.57
(1) Fran Reference 3
(2) Plant has "once-through" barometric condenser water
(3) Long Term Data Appendix A
* Plants with metal anodes
NA: Not Available
261
-------
TSELE 11-38. TOXIC POH1HSNTS IN DIAPHRAGM OKTiT. PLSNT EFFLUENTS
Effluent Concentration
(mg/1)
Pollutant
Arsenic
Cadmium
Chroitiium
Copper
Nidcel
Zinc
Metal .anode Plant
#261 (1)
influent effluent
0.17
0.037
1.9
17
22
1.5
0.12
0.004
< 0.050
< 0.025
< 0.050
< 0.025
Graphite Jtaode Plant
#967 , }
Lead Treatment Plant ^
influent effluent discharge
0.28
< 0.023
0.10
1.6
0.070
0.93
0.36
< 0.015
< 0.050
0.030
< 0.050
< 0.10
0.30
<'0.015
< 0.050
0.031
< 0.050
0.15
(1)
Cell wash waste filtered with coagulant to remove asbestos
(2) Flow-proportioned average discharge, consisting of lead treatment
discharge and untreated filter backwashes, condensates and scrubber
wastes
262
-------
Osing the same VPR of 2.6 the daily concentration limits
are thus,
(0.50 mg/1) (2.6) » 1.3 mg/1
To obtain effluent load limitations for these metals, the
30-day average concentration is multiplied by the unit flow:
(0.50 mg/1) (8.8 m3/kkg) f kg/m3 \ = 0.0044 kg/kkg
V.1000 mg/1/
and the daily maximum effluent limit is:
(1.3 mg/1) (8.8 m3/kkg) / kg/m3 \ = 0.011 kg/kkg
\1000 mg/1/
D. Arsenic and Cadmium: The maximum raw waste
concentrations observed for arsenic and cadmium were below the
literature-based treatability concentration for these metals
(Table 11-35). For this reason only the concentration bases are
presented in Table 11-36. The concentrations represent
treatability for these metals and are meant to serve only as
guidance should these pollutants be of concern.
11.14.2 Basis for BAT Effluent It imitations
Previous BAT regulations called for no discharge of process
waste water pollutants. The regulations were remanded. The
proposed BAT regulations provide for the discharge of pollutants
following appropriate treatment of process wastes.
Technology Basis
Utilizing the cost estimates presented in this report the
Agency has analyzed the cost effectiveness of the base level
system (BPT) and the advanced level options (Levels 2 and 3) for
conventional and toxic pollutant removal. The economic impact
on the diaphragm cell chlorine subcategory has been evaluated in
consideration of the technology basis for proposed BAT
limitations. The need for a reevaluation of cost-effectiveness
based on new cost data will be assessed by the Agency before
promulgation.
For BAT the Agency is uproposing limitations based on BPT
technology with the addition of dual-media filtration (Level 2)
and dechlorination of all process waste waters. Filtration will
remove additional toxic metals and has been used successfully in
the mercury-cell chlorine subcategory. Dechlorination is being
included in BAT because the toxicity of chlorine to aquatic life
is well documented and it is a pollutant of concern to the
Agency (59). Two chlorine plants are known to be practicing
263
-------
dechlorination. The Agency considered the addition of sulfide
precipitation (Level 3) to the treatment of cell room wastes but
rejected it because further reduction of toxic pollutants in
this stream only would not substantially improve total discharge
quality.
Flow Basis
The flow basis for BAT limitations is the model plant total
discharge of 8.8 m3/kkg. This flow reflects that expected at
chlorine plants using graphite anodes.
Selection of Pollutants to be Regulated
For BAT regulations, the Agency has selected the same seven
toxic metals identified in the proposed BPT regulations, and
total residual chlorine.
Basis of Pollutant Limitations
For BAT regulations, the Agency is proposing more stringent
controls on the discharge of the seven toxic metals of concern
on the basis of physical removal by filtration. Alkaline
precipitation converts most dissolved metals into less toxic
insoluble forms and excess alkalinity exists in most of the
process wastes generated in this subcategory. Proposed BAT
limitations are presented in Table 11-39.
Nonconventional Pollutant -
Chlorine: Total residual chlorine limits are based on the
detectable concentration of chlorine (0.2 mg/1) and on
performance of dechlorination in the electric utility industry
(58) because treatment should remove essentially all chlorine.
Thus the 30-day average limit was set at 0.20 mg/1.
The daily maximum limit for total residual chlorine was set
at 0.34 mg/1 based on an evaluation of long-term monitoring data
and determination of variability factors for total residual
chlorine as presented in Appendix A. The ratio of 24-hour
maximum variability factors to 30-day average variability
factors for two plants was 1.7, thus the maximum 30-day average
is given.by:
(0.20 mg/1) (1.7) = 0.34 mg/1
The determination of load limitations for total residual
chlorine (kg/kkg) was calculated based on. the unit flow rate of
8.8 m3/kkgr thus the maximum 30-day average is given by:
264
-------
TABLE 11-39. PROPOSED LZMHATIOKS
Chlorine Diaphragm Cell
Best Available Technology
Waste Water Flew: 8.8 m^/kkg
Pollutant
Treatability VF
(mg/D
Concentration Basis
R. Max
30-day 24-hr
avg. max.
Effluent Limit
f\gTf /L*1b"#fc? i
* * \"^"J/ J\ISJ*-1 /
30-day 24-hr
avg. max.
Noneonventional Pollutant
0.2 1.7
Total Residual
" Chlorine
0.20
0.34
0.0018 0.0030 :
Toxic Pollutants
Arsenic
Cadmium
Chromium^)
O)
Copper w
Lead<2)
19\
Nickel u'
Z-c(2)
0.50^
0.05<3>
0.05' '
0.40(3)
0 . 22
._.
0.101 '
0.40<3>
*
2.
2.
2.
2.
2.
2.
2.
2
2
2
2
2
2
2
0
0
0
0
0
0
0
.50
.05
.05
.40
.22
.10
.40
1.1
0
0
0
0
0
0
.11
.11
.88
.48
.22
.88
0
0
0
0
0
(5)
(5)
.00044
.0035
.0019
.00088
.0035
(5)
(5)
0
0
0
0
0
.00097
.0077
.0042
.0019
.0077
(1) - VPR: ratio of the 24 hour variability factor to the 30-day average
variability factor
(2) - Also applicable for PSES limitations
(3) - Literature - based treatability estimate
(4) - Based on filtration for BPT subcategory performance
(5) - No effluent limitation proposed
265
-------
(0.20 mg/1) (8.8 m3/kkg)/ kg/m3 \ « 0.0018 kg/kkg
VlOOO mg/1 /
The 24-hour maximum limit was calculated similarly,
(0.34 mg/1) (8.8 m3/kkg) / kg/m3 \ = 0.0030 kg/kkg
VlOOO mg/1/
Toxic Pollutants -
Dual media filtration of BPT effluent will significantly
reduce suspended metal precipitates, BAT limitations for the
toxic metals of concern are based on literature-treatability
studies as presented in Section 8.1 and summarized in Table
8-11.
A. Lead: Filtration of the BPT effluent is estimated to
fesult in an 80 percent reduction of lead (41) giving a final
concentration of 0.22 mg/1. This value is used as the
concentration basis for the maximum 30-day average limitation of
0.0019 kg/kkg. Application of the model plant discharge' rate
results in a loading of 0.0019 kg/kkg. That is,
(0.22 mg/1) (8.8 m3/kkg) / kg/m3 \ = 0.0019 kg/kkg
VlOOO mg/1/
The variability factor ratio (VPR) of 2.2, used for BAT
limitations, is from the analysis of mercury monitoring data in
the mercury cell chlorine subcategory (Section 11.7.2). Mercury
cell chlorine plants typically practice filtration of waste
water and the value of 2.2 represents the average VFR of four
plants.
The daily maximum limitation is then,
(2.2) (0.0019 kg/kkg) - 0.0042 kg/kkg
and the daily maximum concentration basis is:
(2.2) (0.22 mg/1) = 0.48 mg/1
B. Chromium: Filtration of the BPT effluent is estimated
to reduce the chromium concentration by approximately 60 percent
(41) to give a final concentration of 0.050 mg/1. This value is
used as the concentration basis for the maximum 30-day average
effluent limitation. Application of the model plant discharge
rate results in a corresponding loading limitation of 0.00044
kg/kkg. That is,
(0.050 mg/1) (8.8 m3/kkg)/ kg/m3« \ = , 0.00044 kg/kkq
VlOOO mg/1 /
266
-------
and, for the daily maximum limitation using the VFR value of
2.2, one obtains;
(2.2) (0.00044 kg/kkg) = 0.00097 kg/kkg
The corresponding concentration basis is:
(2.2) (0.050 mg/1) =0.11 mg/1
C. Copper and Zinc: Filtration of the BPT effluent is
estimated to reduce the copper and zinc concentrations by 20
percent (41) to give a final concentration of 0.40 mg/1. This
value is used as the concentration basis for the maximum 30-day
average effluent limitation. Application of the model plant
discharge rate results in a loading limitation of 0.0035 kg/kkg.
That is,
(0.40 mg/1) (8.8 m3/kkg)/ kg/m3 A = 0.0035 kg/kkg
V1000 mg/1 )
and for the daily maximum limitation using the VFR value of 2.2,
one obtains:
(2.2) (0.0035 kg/kkg) = 0.0077 kg/kkg
and the daily maximum concentration basis is:
(2.2) (0.40 mg/1) = 0.88 mg/1
D. Nickel: The addition of filtration to the BPT effluent
is estimated to achieve a 50 percent reduction of the nickel
concentration. The basis of the proposed BAT limitation is
therefore 0.10 mg/1 and results in a maximum 30-day average
loading limitation of 0.00088 kg/kkg. That is,
(0.10 mg/1) (8.8 m3/kkg) / kg/m3 N = 0.00088 kg/kkg
V1000 mg/1/
and the daily maximum is,
(2.2) (0.00088 kg/kkg) - 0.0019 kg/kkg
with a corresponding concentration basis of:
(2.2) (0.10 mg/1) = 0.22 mg/1
E. Arsenic and Cadmium: Filtration of BPT effluent will
reduce the cadmium concentration to 0.050 mg/1 but will not
significantly reduce the arsenic concentration of 0.50 mg/1.
Because maximum plant raw wastes were below these
concentrations, no effluent limitations are being proposed.
267
-------
Concentration values presented in Table 11-37 are intended for
guidance only.
11.14.3 Basis for BCT Limitations
Technology Basis
Utilizing the cost estimates presented in this report, the
Agency has analyzed the costs of BAT technology in removing
conventional pollutants. This technology of dual-media
filtration of all process waste water was found by the Agency to
be cost effective in removing TSS. Proposed BCT limits for TSS
are given in Table 11-40. This calculation is shown in Section
3.3.3.
Flow Basis
The flow basis for BCT limitations is the same 8.8 m3/kkg
used for both BPT and BAT limitations.
Selection of Pollutants to be Regulated
BCT regulations only apply to total suspended solids {TSS)
and pH.
Basis of Pollutant Limitations
TSS limitations are based on technology performance of
filtration of waste water flow from the mercury cell segment of
the chlorine industry. Appendix Table A-l presents long-term
TSS monitoring data from a chlorine plant practicing filtration
of process wastes. The maximum 30-day average limitations are
based on a 30-day average concentration of 12 mg/1 using the
diaphragm cell model plant discharge flow rate of 8.8 m3/kkgr
namely:
(12 mg/1) (8.8 m3/kkg) f kg/m3 \ = 0.10 kg/kkg
\1000 mg/1/
The variability factor ratio for this plant was 1.9.
Applying this value to the 30-day limit, one obtains a 24-hour
maximum limit of:
(0.10 kg/kkg) (1.9) - 0.19 kg/kkg
^
with a corresponding concentration basis of:
(12 mg/1) (1.9) = 23 mg/1
268
-------
TABLE 11-40. PBQPOSED LIMITATIONS
Chlorine Diaphragm Cell
Best Conventional Technology
Waste Water Flow: 8.8 m^/kkg
Pollutant
Treatability
Concentration Basis
VFR(1) Max (mg/1)
30-day 24-hr
avg. max.
Effluent Limit
30-day 24-hr
avg. max.
Total Suspended
Solids
(2)
12
1.9 12
23
0.10
0.20
(1) - WR: ratio of the 24 hour variability factor to the 30-day variability
factor
(2) - Limitations based on technology transfer from mercury-cell chlorine
subcategory; long-term monitoring data from .Appendix A-l
269
-------
11.14.4 Basis for New Source Performance Standards
Technology Basis
The Agency is basing NSPS limitations on the BAT technology
of alkaline precipitation filtration and dechlorination and on
the performance achieved at plants using metal anodes. The
conversion to metal anodes has largely eliminated the source of
lead in waste waters, but residual lead contamination at a
converted plant may exist for as long as a year or more. New
metal anode plants should have relatively low lead
concentrations in their waste waters. Proposed NSPS limits are
presented in Table 11-41.
Flow Basis
The flow basis of 8.8 m3/kkg used for BPT and BAT
limitations is conservatively being used for new sources.
Selection of Pollutants to be Regulated
For NSPS regulations, the Agency initially considered the
same BAT pollutants (seven toxic metals and total residual
chlorine), pH and TSS. However, following an evaluation of raw
waste characteristics at a new metal anode facility (shown in
Table 11-42) where residual metals contamination from previous
graphite anode use does not exist, only two toxic metals were
selected for regulation. A discussion of the selection of these
metals is presented below.
Basis of Pollutant Limitations
For NSPS regulations the Agency is proposing more stringent
controls on the discharge of toxic metals of concern on the
basis of lower raw waste loads generated at plants using metal
anodes. NSPS proposed regulations ar_e^shown in Table 11-39.
Conventional and Nonconventional Parameters -
A. pH: The treated effluent is to be controlled within
the range of 6.0 to 9.0. This limitation is based on the data
presented in Appendix B of this report and the JRB Study (52).
B. TSS: Limitations for TSS are the same as in BAT
regulations.
C. Total Residual Chlorine: Limitations for total
residual chlorine are the same as in BAT regulations.
270
-------
TSBU3 11-41. PROPOSED LIMITAIIONS
Chlorine Diaphragm Cell
New Source Performance Standards
Waste Water Flow: 8.8 m3/tckg
,.^ C3oncentration Basis Effluent Limit
Pollutant Treatability VERUJ yigx (mg/1) Max k9>/KfcEr
(mg/1) 30-day 24-hr 30-day 24-hr
avg, max. avg. max.
Conventional and
Non-Conventional
TSS 12
Total Kesidual
Chlorine 0.2
Toxic Pollutants
jy:senic 0.50
Cadmium 0.050
Chromium^2^ 0.050
Copper 0.40
Lead(2) 0.050
Nickel 0.10
Zinc 0.40
1.9 12 23
1.7 0.2 0.34
2.2 0.50 1.1
2.2 0.050 0.11
2.2 0.050 0.11
2,2 0.40 0.88
2,2 0.050 0.11
2.2 0.10 0.22
2.2 0.40 0.88
(1) - VER: ratio of the 24 hour variability factor to
, ^variability factor
(2) - Also applicable to
PSNS limitations
0.10
0.0018
(3)
(3)
0.00044
(3)
0.00044
(3)
(3)
the 30-day
0.20
0.0030
(3)
(3)
0.00097
(3)
0.00097
(3)
(3)
(3) - No effluent limitation proposed
271
-------
TABLE 11-42. COMPARISON OF RSW WSJIE CHSRACTEBISTICS AT A NEW METAL
PLAM? WITH TEEATABILITY OF TOXIC METALS
SUBCATEGOiy
CHEDIIKE DISPHRA^I CKTJ,
Pollutant
Arsenic
Cadmiuoa
Qaomittn
Copper
Lead
Nicskel
Zinc
Treatability (1)
0.50
0.050
0.050
0.40
0.050
0.10
0.40
Cancentxation(mg/D
Plant |738B(2)
Raw Waste
0.011
<0.025
0.066
0.12
0.11
0.067
0.093
(1)- Literature based treatability estimates using BIT technology
of dual media filtration following alkaline precipitation of
metals (Table 8-11)
(2)- Verification sampling at new metal anode facility
272
-------
Toxic Pollutants -
Table 11-42 presents the results of verification sampling
of raw wastes at a new chlorine plant using metal anodes. The
total raw waste concentrations of toxic metals are, with the
exceptions of chromium and lead, substantially below the
estimated treatability of these metals using BAT technology, as
shown in the table. For this reason only chromium and lead
effluent limitations are proposed. Other metals are presented
on a concentration basis for guidance purposes only.
Lead and Chromium: The treatability of both lead and
chromium using the BAT technology of alkaline precipitation
followed by dual-media filtration is estimated at 0.05 mg/1
(Table 8-11) . This value was used as the concentration basis
for the proposed maximum 30-day average NSPS effluent
limitations. Application of the model plant discharge rate
results in a corresponding loading limit of 0,00044 kg/kkg.
That is,
(0.050 mg/1) (8.8 m3/kkg)/ kg/m3 \ = 0.00044 kg/kkg
\1000 mg/1 )
and for the proposed daily maximum limitation using the VPR
value of 2.2, one obtains:
(2.2) (0.00044 kg/kkg) = 0.00097 kg/kkg
The concentration basis for the daily maximum is,
i
(2.2) (0.050 mg/1) = 0.11 mg/1
11.14.5 Basis for Pretreatment: Standards
Existing Sources
For Pretreatment Standards for Existing Sources (PSES), the
Agency is proposing the same limitations as for BAT based on the
identical treatment technology without dechlorination being
used for indirect dischargers (see Table 11-39). Dechlorination
is unnecessary because chlorination of publicly-owned treatment
works influent is fairly common. The pollutants to be limited
are chromium, copper, lead, nickel, and zinc.
New Sources
For Pretreatment Standards for New Sources (PSNS), the
Agency is proposing the same limitations as for NSPS based on
the identical treatment technology without dechlorination being
used for indirect dischargers (see Table 11-41). Dechlorination
is unnecessary because chlorination of publicly-owned treatment
273
-------
works influent is fairly common. The pollutants to be limited
are chromium and lead. The pollutants (arsenic, cadmium,
copper, nickel, and zinc) are not being limited based on the
assumption that all new plants will use metal anodes. As shown
in Table 11-42, these pollutants are below treatability levels
at such a plant.
274
-------
SECTION 12
HYDROFLUORIC ACID INDUSTRY
12.1 INDUSTRY PROFILE
12.1.1 General Description
Hydrofluoric acid (Hydrogen fluoride-HF) is produced both
as anhydrous and aqueous products. It is used in the
manufacture of fluorocarbons which are used as refrigerating
fluids, and plastics, for pressurized packing and as
dispersants in aerosol sprays. It is used in the production
of aluminum, in the refining and enriching of uranium fuel,
pickling of stainless steel, in petroleum alkylation, and for
the manufacture of fluoride salts. The industry data profile
is given in Table 12-1. The status of regulations is given in,
Table 12-2.
12.1.2 General Process Description and Raw Materials
HP is the most important manufactured compound of the
fluorine family in volume of production. Fluorspar (mainly
CaF2) 'and sulfuric acid are the raw materials used for its
manufacture. Fluorspar and sulfuric acid react
endothermically at 200-250 degrees C and the reaction time is
20-60 minutes. The reaction is given as:
CaF2 + H2S04 .+ heat = CaSO4 + 2HF (1)
The reaction kinetics and the yield of product depends
on the purity and fineness of the fluorspar. The concentration
of sulfuric acid, the temperature of the reaction,- and the ratio
of sulfuric acid to fluorspar are among important variables.
Crude fluorspar, as mined, varies in CaF2 content from 35
to 90 percent. The ore is upgraded by flotation which results
in 98 percent CaF2 being available for the production of HE.
The analysis of a typical upgraded fluorspar is given as:
CaF2 Minimum 97.5-98%
SiO2 Maximum 1.0%
S " 0.05%
275
-------
12-1 - SUBCATEGORY PROFILE DATA SUMMARY
SUBCATEGORY HYDROFLUORIC ACID
Total subcategory capacity rate 363,000 kkg/year
Total subcategory production rate 261,800 kkg/year
Number of plants in this subcategory 9
308 Data on file for 8
With total capacity of *
With total production of 177,§00 kkg/year
Representing capacity *•
Representing production 68 percent
Plant production range:
Minimum 7,300 kkg/year
Maximum 62,000 kkg/year
Average production 22,100 kkg/year
Median production 15,800 kkg/year
Average capacity utilization 83 percent
Plant age range:
Minimum 7 years
Maximum 58 years
Waste water flow range;
Minimum 0 cubic meters/day
Maximum 4,700 cubic meters/day
Volume per unit product:
Minimum _, 0 cubic meters/kkg
Maximum 86 cubic meters/kkg
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Commerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.j and Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chemical Industry, "June, 1978 and "Economic Analysis of Proposed Revised
Effluent Guidelines and Standards for the Inorganic Chemicals Industry," March, 1980.
* Data incomplete because certain plants did not respond to this question.
276
-------
TflBLE 12-2 -
OP BEGOLKTICNS
GOIEEEilNES
SUBCAIEQOPY
SUBPAOT
Hvdrof luoric Acid
H (40 CER 415.80, 3/12/74)
STANDARDS
Product
Process
Hydro-
fluoric
Acid
BPCTCA* ' BKE2A*
1 2 i •>
Max, Avg. Max.-1 Avg.''
Para- kg/kkg kg/kkg kg/kkg kg/kkg
rosters (rog/1) (mg/1) tog/l) Pig/i)
Fluoride (30) (15) » £^f^
NSPS*
Max. ! Avg.2
kg/kkd kg/kkg
ung/1) (rag/1)
, No discharge
of pwwp
TSS (50)
(25)
Ho discharge
of pwwp
No discharge,
of pwwp
Sections 415.82, 415.83, and 415.85 were rananded and are presently
reserved (41 FR 51601, November 23, 1976).
"Max. =» Maximum of any one day.
2
Avg. — Average of daily values for thirty consecutive days shall not exceed.
pwwp a* Process vastewater pollutants.
277
-------
H20 " 0.1%
CaCOS Principal remainder
Silica is a highly objectionable contaminant, since each
pound consumes 2.0 pounds of fluorspar and 3.3 pounds of
sulfuric acid by the following reaction:
SiO2 + 2CaP2 + 2H2SO4 = SiF4 + 2CaSO4 + 2H2O (2)
Sulfuric acid having a concentration as low as 93 percent
or as high as 99 percent is generally used. Dilute sulfuric
acid enhances better mixing and liberation of fluoride but
has two disadvantages; viz,, the dilute acid is very
corrosive and the water present in the acid evaporates and
distills off with the HP gas, thus reducing product
concentration. Concentrated sulfuric acid (greater than
98 percent) offsets these disadvantages but creates new
problems. The vapor pressure of concentrated sulfuric acid is
sufficiently high to cause large amounts of sulfuric acid to
be carried away by the HF. Excess sulfuric acid, when used,
will leave with the gypsum as part of the residue.
HP generators are, in the majority of cases, externally
fired rotary kilns in which acid and fluorspar are fed
continuously through a screw conveyor at the forward end and
gypsum is removed from the other end through an air lock. The
product HP may discharge from either end. The theoretical
amount of gypsum produced is 3.4 kg/kg of HP produced, but
because of the impurities in the fluorspar the actual amount
°f gyPsum produced is higher and varies from 3.6 to 4.8 kg/kg of
HP.
One manufacturer uses a patented process to supply
internal heat to the reactor. The heat is supplied by
introducing sulfur trioxide (SOS) and water (as steam).
The exothermic heat liberated by the reaction of S03 and water
is used for the heat required for HF generation. Thus a part of
the sulfuric acid is supplied as 803.
The HF gas leaving the reactor is cooled in a precooler to
condense high boiling compounds. The condensables are known
as drip acid and largely consist of fluorosulfonic acid (HS03F)
and unreacted sulfuric acid. In 1978, nine plants out of a
total of eleven returned the drip acid to the reactor,
while the remaining two sent the drip acid to the waste
treatment plant. The HP gas from the precooler is cooled
further and condensed in a cooler/refrigeration unit. The
uncondensed gas containing the HF is scrubbed with sulfuric
acid and refrigerated to recover the product. The scrubbed
acid liquor is returned to the kiln, and residual vent gases
are scrubbed further with water to remove HF and other
fluoride compounds before they are vented to the atmosphere.
The scrubber water is sent to the waste water treatment
278
-------
plant. Figure 12-1 is a block flow diagram of the
manufacturing process.
The crude HF is then distilled to remove the residual
impurities, and "the condensate, which is anhydrous HF, is
stored in tanks. If aqueous HF is desired, the crude product is
then diluted with water to form a 70 percent HF solution as the
final product.
12.2 WATER USE AND WASTE SOURCE CHARACTERISTICS
12.2.1 Water Use
Water is used in HF production in noncontact cooling, air
pollution control, product dilution, seals on pumps and
kilns, and for equipment and area washdown. Although
noncontact cooling constitutes the major use of water, water
is also used, in a majority of cases, in the transport of gypsum
as a slurry to the waste water treatment facility. The
water for gypsum transport is provided by either reusing the
water from the treatment facility or by using once-
through cooling water. Table 12-3 summarizes the water usage
found in this study.
12.2.2 Waste Sources
Gypsum Solids
Gypsum solids are generated as a by-product. The
amount produced is in the range of 3.6 4.8 kg/kg of HF
produced. The gypsum also contains small amounts of sulfuric
acid, HF, and calcium fluoride. Minor amounts of other
impurities present in fluorspar are also removed with the
gypsum. In five out of eleven plants producing HF, gypsum
is slurried with treated waste water, neutralized with lime
or soda ash, and pumped to a gypsum storage pond. In one plant
the gypsum slurry is pumped to the storage pond without
treatment and in another plant partial neutralization is
employed. Three plants transport the gypsum as a dry solid and
dispose of it as a solid waste after mixing with lime for
neutralization. The disposal method of one plant is not
known. It should be noted that two of the eleven plants have
recently discontinued HF production, one of which is in the
group of five.
When gypsum solids from the kiln are slurried with water
for treatment, the resulting stream constitutes the major
source of waste water. When kiln residue is disposed of as a
solid waste, scrubber waste water is the major source of
waste. Table 12-4 gives the data for the direct and indirect
279
-------
SLOBBY WATER
TO WASTE
1OXER'
SULFURIC
ACID
FLUORSPAR
SOLFATE (GYPSUM)
SOLIDS
SULFONIC ACID
NONCONTACT
COOLING OR
REFRIGERATION
SYSTEM
CRUDE HYDROGEN
SOLFURIC ACID
AQUEOUS HYDROGEN
HYDROGEN FLUORIDE
FLUORIDE PRODUCT
LEGEND
COMMON PRACTICE
INIERMICTSHT
PROCESS (OR PRO
CESS KS DULY
SOME PLANTS)
EJECTOR
WSTSHMER
to
TREATMENT
TO SALES
Figure 12-1. General process flow diagram for production
of hydrofluoric acid.
280
-------
TABLE 12-3. WATER USAGE IN THE HYDROFLUORIC ACID SUBCATEGORY
Water Usage at Plants
{m3/kkg of HP)
(1) (11 (1)
Source §987 #251 #753 #426 #120 #722 #167 #705
ton-contact 154 NA 63.5 110 NA 13.6 116 30.0
Cooling
Gypsum Slurry NA 64.0 NA * NA 22.5 41.6 30.0
Transport
Maintenance, NA 2.40 2.11 NA 0.1 12.2 5.00 16.9
Equipment and
Area Washdown
Air Pollution 7.90 14.4 4.23 NA 0.586 14.5 40.0 11,3
Control
NA = Not Available
* = Not Applicable
(1) Discontinued HP production.-
281
-------
TABLE 12-4. WATER FLOW AND FOR THE
HYDROFLUORIC ACID SU8CATEGORY
I
Plant
1120
1426
(3)
1987
$837
1967
(3)
1251
(3)
1705
1167
1753
1928
1664
1722
Averages:
(1) D * DE
Ciln Residue
(D
Handling
D
D
D
5
S
S
S
S
S
S
S
S
(S only)
•y disposal
Reuse for Influent to
Kiln Residue Treatment
(2)
Slurry Facility
(Percent) (m3/kkg)
HF
(4)
— 9.10
(4)
— o
(4)
"**•*""' JL%MP* O
0 120
0 125
0 84.7
30.0-35.0 58.2
47.0 166
65.0 31.4
83.0 55.5
94.0 96.6
92.0-100 120
42.8 percent 95.4 m3Akg
S = Slurried to treatment
Treated
Effluent
Discharged
(m3/kkg)
HP
9.10
Not available
13.6
120
125
84.7
39.3
88.2
11,1
9.40
5.80
7.20
54.6 m3/kkg
(2> Percent of waste water flow reused for residue slurry after
treatment.
(3) Dicontinued HF production.
(4) Not Applicable.
282
-------
process contact waste water going to treatment facilities.
Noncontact cooling water has not been included in the figures
given in Table 12-4. Figure 12-2 is a graphical
representation of production versus waste water flow to in-
plant treatment facilities for plants whose waste water
includes the gypsum slurry and for those "practicing disposal
of kiln residue as a solid waste.
Drip Acid
This is formed in the first stage of the cooling {i.e.,
in the precooler) of the gases emitted from the kiln. Drip
acid mostly contains high boiling compounds consisting of
complex fluorides, especially fluorosulfonic acid, and small
amounts of hydrofluoric acid, sulfuric acid, and water.
Pluorosulfonic acid is formed by reaction between
hydrofluoric acid and sulfuric acid in the absence of water.
The quantity of drip acid produced is relatively small. In the
plants which recycle the drip acid back to the reactor, it is
mixed with the sulfuric acid feed stream before it enters the
kiln where it is hydrolyzed to form sulfuric acid and
hydrofluoric acid. The critical factors for hydrolysis are
temperature and retention time and enough water is normally
present in the kiln for the reaction.
Noncontact Cooling Water
Noncontact cooling water is used for precooling the
product gases emitted from the kiln. The possibility of product
or other process compounds leaking into the cooling water is
very small? however, in the event that the cooling water does
become contaminated, the proposed limitations for fluoride may
be exceeded. Depending on the merits of the situation, the
upset and bypass provisions may apply. In some plants, the
cooling water is used to transport the waste gypsum.
Scrubber Haste Water
Scrubber water is another waste water source, and in
plants whieh practice dry disposal of gypsum, scrubber water
constitutes the predominant and major source of waste water. It
contains fluoride, sulfate, and acidity. The fluoride is
present as HP, silicon tetrafluoride (SiF4), and hexafluosilicic
acid (H2SiF6). Silica present in the ore as an impurity reacts
with HP forming silicon tetrafluoride as shown in Equation 3.
Si02 + 4HP , » SiF4 + 2H2O (3)
In the scrubber, the tetrafluoride is converted to
hexafluosilicic acid according to the following equations:
283
-------
15,000 +
12,500 4
10,000 -f
i
g 7,500
1
CO
5,000 -j-
2,500 --
2,000 -•
1,000 --
0
0
/
Slurrying Kiln Waste ~-'
+
75 100 150
HP Broduction, kkg/day
Figure 12-2. Production versus waste flow data for HF plants,
284
200
-------
SiP4 + 2HF = H2S1F6 (4A)
3S1F4 + 2H2O » 2H2S1F6 + S1O2 (4B)
Distillation Wastes
The distillation waste generally contains HF and water.
In some cases the vent gases from the distillation column are
scrubbed before they are emitted to the atmosphere, and the
resulting scrubber water requires treatment.
The range of waste water quality of the different
streams generated from the production of HF is summarized in
Table 12-5. The data are taken from the prior development
documents, 308 Questionnaire responses, and industry visits.
Other Solid Wastes
The total solids generated from the process and the
treatment system consist of gypsum and the fluoride
precipitated as calcium fluoride. Table 12-6 gives the amount
of suspended solids generated from the process and the
quantity of total suspended solids generated at the waste
water treatment plant for the HF plants visited in screening
and verification. The data- indicate that the gypsum waste
constitutes more than 95 percent of the total solids
produced. Table 12-7 gives the amount of gypsum solids
produced at different HF manufacturing facilities. The data
shows that 3.8 to 4.7 kg gypsum solids are produced per kg of
product.
12.3 DESCRIPTION OF PLANTS VISITED AND SAMPLED
12.3.1 Screening
Plant $705 was visited and process waste water samples
were collected and analyzed for conventional, nonconventional
and tonic pollutants. The process used at this site is
similar to the conventional HF manufacturing process
described earlier. The drip acid is sent to the waste water
treatment facility and the gypsum produced from the reactor
is slurried with water and also sent to the treatment
facility. The waste waters from the HF production facility
are combined with the aluminum fluoride plant waste waters.
The combined raw waste water is treated with lime and sent
to settling ponds before discharge. Figure 12-3 shows the
general process and the locations of the sampling points.
Table 12-8 gives the flow data and the total suspended
solids (TSS) and fluoride emissions.
285
-------
12-5. ffiCW HXDK3EKXBIC ACID MMfOEmCTORING ELSNTS
to
00
01
3
Flow in m /Kkg of %drofluoric Acid
Plants
Source of
Waste Water
Gypsum Slurry
Drip Acid
Scrubber
Waste Water
Other
#251 (1) #987 (1>
64.0 Dry
disposal
0.0490 0
14.4 8.30
0.530 0.530
#753 #426 §120
NA Dry Dry
disposal disposal
00 0
2.30 N& 0.624
8.40 NA 5.55
#722
(Total
Recycle)
0
(total
Recycle)
M
1167 #705 (1)
122 (Total
Recycle)
NA 0.0180
40.0 11.3
5.20 22.5
#837
6.50
0
1.12
NA
(1) Discontinued HP production
m = Not Available
* Other does not include wasteflows from storm water runoff.
-------
TABLE 12-6. SOLID WASTE GENERATED AT THE HH3RQFLUORIC ACID FLAM'S SAMPLED
Plant Gypsum Solids Going To Total Solids Produced
Treatment Facility (kg/kg of HF)
" - (kg/kg of HP)
1705 ^ 4.73 4.78
#251(1) 3.81 NA
#167 3.94 NA
(1) Discontinued HF production.
NA = Not Available
287
-------
TKEIE '12-7. GXPSUM SOLIDS PRODUCTION IN TEE HYDROFHJORIC ACID SDBCMESORY
Plant
1837
f705 <1)
*167
#722
#120
§426
%87 (D
1251 ^
*753
1967
f928
Kiln Residue Produced
(kgAg of HP)
3.86
4.73
3.94
NA
NA
4.00
4.13
3.81
m
NA
NA
Kiln Residue
Disposal/^Ireatroent Method
S
s
s
s
D
D
D
S
S
s
• s
S" * Slurried with •water and sent to wastewater treatment facility.
D » Dry disposal.
N& = Not available.
(1) Discontinued HP production. 288
-------
vats
ro
oo
Waste streams sarpled.
Figure 12-3. General process flow diagram at plant #705 showing the sampling points.
Hydrofluoric acid manufacture. •
-------
TABLE 12-8. FLOW AND POLLUTANT CONCENTRATION DATA OF THE SAT/PLED WASTE
STREAMS OF PLANT #705 PRODUCING HYDROFLUORIC
Stream
No.
1
2
3
4
Sanpled
Stream
Description
Kiln Slurry
Scrubber Waste
Water
Surface Drains
Cooling Tower
Slowdown
Treated' Effluent
Screening
Flow
(m3/kkg of HP)
26.6
10.0
20.0
23.3(3>
DataC2>
Fluoride
(kg/kkg of HP)
15
9.6
6.9
1.6
Total
Suspended
Solids
(kg/kkg of HP)
4700
0.070
3.9
•
1.9
(1) This plant has discontinued the production of HF since the time of
sartpling.
(2) One 72-hour coirposite sample of each waste water stream.-
(3) The discharged effluent consists of the treated waste waters from
hydrofluoric acid and aluminum fluoride plants.
290
-------
12.3.2 Verification
The same streams at Plant |705 were sampled again in the
verification phase. The variation in the flow of the streams
in the two sampling phases was negligible. Table 12-9 gives
the TSS and fluoride load summary of the sampled streams.
Two more HF plants (Plant |251 and f!67) were sampled in
the verification phase. The drip acid at both facilities is
also sent to the waste treatment plant and the hydrofluoric
acid waste waters are combined with aluminum fluoride
plant waste for treatment. In addition to drip acid, Plant
1251 waste water consists of scrubber water, gypsum slurry,
and plant area hose down. The treatment consists of gypsum
ponds where the suspended solids are removed. The overflow
from the last gypsum pond is neutralized and the pH adjusted
with wastes from other product lines. Figure 12-4 is a block
diagram of the process showing the sampling locations at Plant
1251.
At Plant #167, the major raw waste sources are the kiln
waste slurry, the absorber tails from the condensate {drip
acid) recycle system, and the ejector water which is used to
quench the off-gases from the absorber. All three of these
waste streams are collected in a common neutralization pit where
lime slurry is added. The waste then flows into a series of
three lagoons for solids removal and final pH adjustment prior
to discharge. Most of the gypsum settles out in the first
lagoon and the overflow enters the second lagoon where
commingling with wastes from other processes takes place.
Verification sampling data from this plant were obtained
from four sampling points. These are: 1) the kiln waste
slurry, 2) the absorber tails, 3) the ejector water, and 4) the
effluent from the first lagoon. The fourth sampling point is
the last point at which all waste water originating in the HF
plant can be intercepted.
12.3.3 Summary of the Toxic Pollutant Data
Eleven toxic pollutants were found in the raw waste
samples from HF Plant §705. They were also verified at three
other typical HF plants practicing BPT treatment. No organic
toxic pollutants were found at detectable levels. The results
were:
291
-------
TAKEE 12-9. FlOW AND PCMG3TANT (XftKEIOTR&TION DATA OF THE SAMPLED WASTE
FOR PIANTS #705, 1251, AND 1167 PRODUCING HYDBDFLUORIC ACID
Plant
*705(2)
1251 (2)
#167
Stream
Kb.
1
2
4
5
5
6
2
3
1
2
3
4
Verification Data
Sampled 3 Flow
Stream Cm /kkg of HP)
Description
Kiln Slurry
Scrubber Waste
Water
Surface Drains
Cooling Tower
Slowdown
Treated Effluent
AHF Plant
Hbsedown
SO2 Scrubber
Waste
Gypsum Pond Inlet
Gypsum Pond
Outlet
Kiln Slurry
Ejector & Absorber
Unit Wastes from
Kilns #1,#2, and
14
Ejector & Absorber
'Unit Wastes from
Kilns 15 and §6
Effluent from
First lagoon
26.6
10.0
20.0
23.3
1.20
14.4
84.7
84.7
122
25.0
14.6
162
(1)
Fluoride
(kg/kkg of HF)
3.8
1.5
3.4
0.54
1.9
0.31
58
27
4.9
14
20
11
Total
Suspended
Solids
(kg/kkg of HF)
4700
0.019
4.0
0.040
0.26
0.10
3800
0.80
170
0.36
0.41
22
(1) Three 24-hour composite samples of each waste water stream.
(2) These plants have now discontinued their HF production.
(3) Consists of the combined flow from hydrofluoric acid and aluminum
fluoride plants.
292
-------
K3
VENT
DUST
COLLECTION
WET
SPAR"
SPAR
DRYING
HOSE DOWN
WATER '
AIR
i/iiu
N
RESIDUE
HANDLING
LOSSES
DRIP
ACID
AUO
WATER
A1F3 PRODUCT
LIQUEFACTION
•*-
AHF
PURIFICATION
DILUTION WATER
O
LEGEND
SAMPLING POINTS.
15
I
44
^^
HOSE DOWN WATER
AHF PLANT
#2
NEUTRALIZATION
SYSTEM
ALKALINE STREAMS AND ««-
ACID FROM OTHER PLANTS
WATER
EFFLUENT
Figure 12-4. General process flow diagram at Plant #251 showing the sampling points.
Hydrofluoric acid manufacture.
-------
Maximum Raw Waste Concentrations Observed
(pg/1)
Pollutant Screening Verification
Plant f705 Plants 1705,
*167
Copper
Lead
Selenium
line
Antimony
Arsenic
Cadmium
Chromium
Mercury
Nickel
Thallium
770
5200
25
8100
70
10
2
73
2
150
5
.0
.0
.5
600
200
230
13000
2800
160
60
1200
43
2000
63
Section 5.1.2 of this report describes the methodology of
the screening and verification sampling program. In the
Hydrofluoric Acid industry, a total of 12 days of sampling
were conducted at Plants £705, £251, and f!67. Sixteen
different sampling points were involved covering the raw waste
source, the various raw waste streams, and the treated
effluents at these plants. The evaluation of toxic metal
content of these process related waste streams was based on
572 analytical data points. The screening for toxic organic
pollutants at Plant f'705 generated an additional 635
analytical data points. The daily raw waste loads were
calculated from the waste stream flow rates measured or
estimated at the time of sampling and the measured pollutant
concentration.
That is,
Daily loading (as kg of pollutant (C)(Q)
per day) = 1000
Where:
C is the concentration of the pollutant expressed in units
of mg/1 (Note: kg/m3 = 1000 mg/1), and
F is the waste stream flow rate expressed in units of
m3/day. (m3, a cubic meter, is equal to 264.2 U.S.
gallons)
Similarly, the unit loadings were calculated from the
reported hydrofluoric acid production rate, the waste stream
flow rate, and the measured pollutant concentration.
Unit loading (as kg of pollutant (C) (Q)
per kkg of hydrofluoric acid) = 1000 fP)
294
-------
Where C and F are the same as described above, and P is the
hydrofluoric acid production rate expressed in units of
kkg/day. (kkg is 1000 kg, a metric ton, which is equal to
2205 Ibs.)
The minimum, average, and maximum values are based on
data from those plants where the particular pollutant was
found at concentrations greater than the analytical detection
limits and significant in that it could conceivably be
treated by an available treatment technology regardless
of economic considerations.
In Table 12-10, the toxic pollutant raw waste data are
presented as the average daily concentrations and the unit
loading found at the individual plants. The overall averages
are also shown and were subsequently used in the calculations
of the average daily loadings and the average unit loadings
shown in Table 12-11 along with the corresponding minimum and
maximum values.
Based on the total annual production of this subcategory
and the average waste load generated per unit product, the
estimated total toxic pollutant raw waste loads generated each
year for this subcategory are as follows:
Pollutant Waste Load (kg/year)
Copper 6600
Lead 10000
Selenium 260
Zinc 110000
Antimony 8900
Arsenic 1400
Cadmium . 79
Chromium 4700
Mercury , 130
Nickel 10000
Thallium 840
12.4 POLLUTION ABATEMENT OPTIONS
12.4.1 Toxic Pollutants of Concern
i
Toxic pollutants in raw waste waters and slurries typical
of the HF industry include the heavy metals often found
as impurities in fluorspar. These metals are zinc, lead,
295
-------
TABLE 12-10. TOXIC POLLUTANT RAW WASTE DATA
SUBCATEGORY: HYDROFLUORIC ACID
(1)
Average Daily Pollutant Concentrations and Loadings at Plants Sampled
(mg/1)
(kg/kkg of Anhydrous HP)
#705 (S) #705 (V) 1251 (V)
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Thallium
Zinc
0.018
0.0010
0.051
0. 0029
0.0014
0.000080
0.062
0.0035
0.41
0.023
2.47
0.14
0.00090
0.000050
0.062
0.0035
0.0070
0. 00040
*
*
4.0
0.23
0.010
0.00057
*
*
0.0060
0.00034
0.26
0.015
0.26
0.015
0.044
0.0025
0.0053
0.00030
0.48
0.027
*
*
*
*
0.21
0.012
0.12
0.010
0.11
0.0091
*
*
0.47
0.040
0.12
0.010
0.059
0.0050
0.018
0.0015
1.18
0.10
. 0.017
0.0014
0.039
0.0033
0.28
0.024
I167CV)
0.74
0.12
0.028
0.0046
0.0030
0.00047
0.074
0,012
0.32
0.051
0.062
0.010
0.0010
0.00016
0.15
0.025
0. 0074
0.0012
0.019
0.0030
8.2
1.3
Overall
Average
0.22
0.033
0.062
0.0055
0.0035
0.00030
0.22
0.018
0.28
0.025
0.66
0.039
0.0060
0.00050
0.47
0.039
0.011
0.0010
0.029
0.0032
3.2
0.41
S - Screening data from one 72-hour composite sample of
individual or combined raw waste streams.
V - Verification data from three 24-hour composite samples, averaged,
from each raw waste sampling point.
* - Concentration below significant level.
(1) The methodology of the sampling program is described in Section
5.1.2, and Section 12.3.3 presents the scope of sampling in the
Hydrofluoric Acid industry.
296
-------
TABLE 12-11. SUMMARY OF RAW WASTE LOADINGS FOUND IN
SCREENING AND VERIFICATION SAMPLING
SU8CATEGORY: HYDROFLUORIC ACID
Daily " Unit
Loadings Loadings No. Of
(kg/day) (kg/kkg) Plants
Pollutant Minimum Average Maximum Minimum Average Maximum Averaged*
Toxic
Antimony
Arsenic
Cadmium
Chromium
Copper
lead
Mercury
Nickel
Selenium
Thallium
Zinc
Conventional
0.023
0.01-2
0.0031
0.15
0.60
0.10
0.0021
0.14
0.016
0.16
0.49
2.0
0.50
0.014
1.7
1.4
1.8
0.057
4.1
0.093
0.31
21
6
1
0
5
2
5
0
14
0.
0.
72
.4
.2
.025
.4
.80
.4
.21
20
45
0.00057
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0"0030
000077
0035
0096
0025
000050
00035
00040
0030
012
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
034
0055
00030
018
025
039
00050
039
0010
0032
41
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
1.
12
0090
00047
040
051
14
0015
10
0014
0033
3
4
3
. 3
' 4
4
4
4
4
3
2
4
& Nonconventional
TSS , 190000
Fluoride
13
310000
2900
520000
7900
3800
8.
8
4200
34
4800
58
3
4
* Only those plants where the pollutant was observed at significant
levels were included.
297
-------
nickel, antimony, chromium, arsenic, copper, and selenium. Raw
waste waters from plants practicing dry disposal of kiln
wastes may include some of the same heavy metals in
scrubber and area washdown wastes, but in considerably smaller
amounts, since the spent ore is hauled as a solid waste and
bypasses the waste water treatment facilities. Although the
fluorosulfonate anion is found in HP wastes containing drip
acid, organic compounds are not anticipated in waste waters
from this industry. No toxic organic pollutants were found at
significant levels.
12.4.2 Process Mod if i cat i ons and Technology Transfer Options
1. Gypsum produced in the kiln can be disposed of as a
solid waste instead of being slurried with water and sent to
the waste water treatment facility. The -solids in this case
are mixed with lime and stored in piles on the land surface
until alternative disposal methods are found or the
site abandoned. Although the dry disposal method is
labor intensive (involving transporation and landfill operating
cost) , it has been found to be less expensive due to the
reduced initial capital requirement and operating costs
relative to the wet slurry method which requires a more
extensive system of pipes, pumps, and on-site impoundments.
2. The use of soda ash in place of lime for
neutralization has some advantages. It eliminates or reduces
the problem of scale formation in the pipelines and scrubbers
when the treated waste water is recycled. It offers a faster
reaction time and better control of pH than lime. Even though
the cost of soda ash is higher than lime, soda ash has been found
overall to be a less expensive alternative at some plants. One
plant reported that a combination of brine and soda ash has
been found to present the best alternative for operation of
the recycle system at minimum cost. After the use of soda
ash, the treated effluent water can be totally recycled,
either to the scrubber or to the kiln for transportation water
for the gypsum.
As the pH approaches 6, sodium in soda ash replaces
calcium present in the gypsum waste. This frees enough
calcium ion to precipitate fluoride as calcium fluoride. Where
the scrubber water is the predominant source of waste water',
the water has to be treated first with enough lime to
precipitate fluoride as calcium fluoride. Soda ash can
then be added to the supernatant to precipitate calcium
followed by neutralization with HC1 to reduce scaling
problems.
3. Two out of a total of 11 plants manufacturing
hydrofluoric acid send the drip acid to the waste water
treatment facility. The rest of the plants recycle it to the
298
-------
reactor. When discharged to the waste treatment system, the
fluorosulfonic acid does not hydrolyze and leaves with the
treated effluent as a complex fluoride in soluble form. The ,
total fluoride concentration of the effluent will be higher
for the plants discharging drig acid compared to those which
do notr after the same neutralization treatment. The two
plants discharging drip acid to waste looked into the
feasibility of returning it to the kiln, but because of the
unique design of the kilns, they found it to be economically
unattractive. Bench scale studies have shown that the drip
acid can be hydrolyzed to free the HP.
HSO3F + H2O + heat = H2S04 + HF (5)
The two plants not returning the drip acid to the kiln
should be able to hydrolyze the material in a separate unit
before commingling it with other wastes, thus avoiding the
treatability problem associated with complex fluorides.
12.4.3 Best Management"Practices
1. Runoff can be collected from raw material and product
storage, process, and impoundment areas. It should be treated
with other process waste at the waste water treatment
facility. Leachate and permeate control needs to be practiced
on the solid waste stored in many plant premises as gypsum
piles. There is a risk that uncontrolled stockpiling _may
contaminate the local ground, water.
2. Ponds designed for solids removal must be deep enough
to have a minimum of disturbance from wind and rain. In those
areas where the rainfall rate exceeds yearly evaporation,
the collection of runoff from raw material, product storage,
process, and impoundment areas may lead to serious water balance
problems. Recycle ponds would have to be designed to handle
this excess loading.
12.4.4 Prevailing Control and Treatment Practices
Plant §705 combines the hydrofluoric acid wastes,
including the gypsum slurry, with aluminum fluoride waste.
The combined waste water, after neutralization, is sent to
settling lagoons before discharge. This plant was visited in
both the screening and verification phases of the project and
a fuller description of waste treatment practice is given below.
Plant f837 combines the gypsum slurry and plant area
hosedown waste water with the equipment washings, leaks, and
spills etc. from the aluminum fluoride plant and neutralizes
them with lime. The solids are removed in settling ponds
before discharge. The waste water from scrubbers of both
299
-------
hydrofluoric acid and aluminum fluoride plants ' is sent to
an adjoining facility for use.
Plant |251 also combines the hydrofluoric acid and aluminum
fluoride waste water. The suspended solids in the combined
waste water are removed in the gypsum ponds. The overflow
from the gypsum ponds is neutralized and the pH adjusted with
the waste water from other products which are manufactured on
the site. The plant is in the process of installing a new
proprietary treatment process to further reduce the fluoride in
its waste waters.
Two plants, $120 and f987f dispose of the kiln residue as a
solid waste after lime addition. The waste water in both cases
is treated with lime and the solids are separated? in one case
in a clarifier followed by a filtration, and in the other
by lagooning.
At Plant §167, the combined waste water (including the
gypsum) is neutralized with lime and then settled in lagoons
before discharge.
Plant f722 practices complete recycle. The gypsum slurry,
scrubber water, and other waste waters are combined and
treated with soda ash for neutralization. The neutralized
solution is settled in lagoons and then is recycled to the
scrubbers and to the kiln to slurry the gypsum.
Plant f426 disposes of the gypsum solids from the kiln as a
solid waste after lime addition. The scrubber water is used to
make another product. The noncontact cooling water is
neutralized when required with caustic soda and settled
before discharge.
12.4,5 Advanced Treatment Technologies
Although alkaline precipitation, sulfide precipitation,
the xanthate process, and ion exchange might be applied to
clarified solutions for control of metal ions, only alkaline
precipitation can be readily used for slurri.ed kiln wastes from
HF production. Sulfide precipitation from cleared solutions
could be used to provide additional removal of zinc, lead,
nickel, and copper and to a lesser extent, antimony.
300
-------
12.5 SELECTION OF APPROPRIATE TECHNOLOGY AND EQUIPMENT
12.5.1 Technologies for Different Treatment Levels
Level 1 (BPT)
Neutralization with lime, used widely in the HP industry,
is shown as the BPT treatment, principally to control pH
and the nonconventional pollutant fluoride - which is
precipitated as calcium fluoride. Sufficient settled effluent
may be reused to transport kiln waste to 'the treatment
facility as a slurry, and the remainder is adjusted to a pH
between 6 and 9 before discharge. The flow diagram is shown
in Figure 12-5,
Level 2 (BAT)
Treatment is alkaline precipitation, using additional
lime and close control of pH in second-stage
neutralization, followed by lagoon settling. Sufficient
lagoon effluent is reused to transport kiln waste to the
treatment facility as a slurry and the remainder JLS filtered
to remove finely divided metal hydroxides. The flow diagram is
shown in Figure 12-6.
Level 3
It is assumed that 65 percent of the Level 2 effluent is
reused for transporting spent kiln waste to the treatment
facility. For the remaining 35 percent, pH adjustment and
sulfide precipitation are used ahead of the Level 2 dual
media filter, to react with residual lead, copper, nickel,
zinc, and antimony which may not have reached their optimum pH
levels for alkaline precipitation. The flow diagram for this
treatment is shown in Figure 12-7.
Level 4
As an alternative to Level 2, Level 4 employs soda ash
instead of lime for neutralization, depending on the spent
ore to contain enough ' calcium to precipitate calcium
fluoride. Use of soda ash permits increased-effluent
recycling without scaling problems associated with calcium
sulfate. To control salinity and sodium alkalinity, a final
effluent blowdown of at least 10 percent of the influent rate
is maintained. The common heavy metals will be precipitated
as carbonates and hydroxides with varying degrees of
effectiveness at pH levels attainable with soda ash. The
effluent is filtered and adjusted to a pH between 6 and 9 before
discharge or process recycling. (Figure 12-8.)
301
-------
UUS
I—-fin
.RAW Ti
WASTE WATER
EQUALIZATION
RECYCLE FOR SLURRY TRANSPORT •*-
MIXING
LAGOON
LAGOON
pH ADJUSTMENT
-»•»
EFFLUENT
»•
Includes flow monitoring, pH monitoring and sampler
Figure 12-5. Level 1 waste water treatment for hydrofluoric acid sx±)category.
-------
BACKWASH
o
U)
r — nuw;
l^J— fi»-
HAW
WASTE WATER
r--@
i
i
i
f3
f
k i /,
KOUAUZATION I
l.-.torf-n a.
" : -~1
i
I
1 '
J u
,— L. -1* »X
MIXING ,
RECYCLE fOB
SUJRRY TBAHSPOBr*""
—Ji, LAOOON r-~\
•
-\ UODON /I*.
. l'» I
ADJUSTMENT,
4_h
SUMP Lrt_
FILTBO *
1JSOENU
i Uuw niDRllitiriiig«
Figxire 12-6. Level 2 waste water •treatment for hydrofluoric acid subcategory.
-------
*»
WSIBtKtEK
r
—fun
0
L
SUKW TRANSPORT
"1
M
Include* flow monltoriisg, pH manllaring aad
Figure 12-7. Level 3 v?aste water treatment for hydrofluoric acid subcategory.
-------
r
en
SMI
_r**-
•
E
a
f dfi)
/
— , /
SQKHSKfSCH J
f
HKKK
I
I
li
**
,-A UOOCH /»,
n ™
r '
1
1
V ,
L— £^ LKCX30N £m^i*
atmur n«i3H
,t,
SUMP
3RT
•rfU—
r**
1
i
X
t
racsfi
&^U£
i
,
.1. i.Wt,.l
Includes flow mojtitoHiig, pH mQUitarmg a&d *
Figure 12-8. Level 4 waste water treatment for bycfrofluoric acid subcategory.
-------
Level 5 (NSPS)
The proposed NSPS treatment is dry handling and off-site
chemical landfill for the kiln waste and two-stage
alkaline precipitation with clarification and filtration for
the liquid process wastes. Heavy metal precipitation with
soda ash permits partial recycling for uses other than slurry
transport. (Figure 12-9.)
12.5.2 Equipment for Different Treatment Levels
Equipment Functions
Level lr typical of existing treatment facilities,
utilizes very little equipment, but depends on lime
neutralization in settling lagoons, with final pH adjustment.
In Level 2, conventional dual media filtration is added to
the Level 1 system. In Level 3, standard reagent mixing and
solution feeding units are added to the Level 2 system. In
Level 4, which is an alternate to Level 2, the same type of
chemical feed equipment is used for soda ash as was used for
lime in Level 2. Conventional lagoons and dual media filters
are used in Level 4, but special attention to selection of
materials is required because of the high salinity of
recycled effluent. In the NSPS model, dry kiln waste disposal
is recommended with conventional dry solids handling
equipment. Lagoons, clarifiers, and filters are used for
scrubber, noncontact cooling, and other miscellaneous
liquid wastes. In this case, equipment for storing and
handling the dry kiln waste is not considered to be waste
water treatment, and the cost is not included in the cost
estimates.
Chemical Handling
Lime (as CaO) is the major chemical used in Levels 1 and
2, along with minor amounts of hydrochloric acid for final pH
adjustment. With normal precautions, these chemicals pose
no special hazards. In Level 3, ferrous sulfide is prepared
on-site by mixing sodium bisulfide and ferrous sulfate.
Although sodium bisulfide can release toxic H2S at pH levels
below 7, the hazard can be mitigated by avoiding acid
conditions and by providing adequate ventilation. After mixing
its components, the ferrous sulfide solution, is stable at the
pH levels employed in the process. In Level 4, only sodium
carbonate and hydrochloric acid are used, without unusual
safety hazards or special handling problems. In the NSPS system
only lime, soda ash and hydrochloric acid are used,
introducing no special problems of safety or handling.
306
-------
ll ADJUSTMENT
HECYCI.B *O
SCRtfUBER
{3-iUO OfASTEJ
Include* flow munltorlng, pH motiltartog aiul B
Figure 12-9. Waste water treatment new source performance standard for
hydrofluoric acid subcategory.
-------
Separation and Removal of Solids
Solids are accumulated in unlined settling lagoons. In
Level 4, calcium fluoride will still precipitate in the
lagoons but the total sludge quantities will be less than in
fievels If 2, and 3 where lime is used. Solids from Level 4
treatment will be alkaline, very, saline, and difficult to
consolidate. Dry solids from the Level 5 (NSPS) model are
not subjected to treatment, except for nominal application,
of lime before hauling in dry form to an approved chemical
landfill.
12.6 TBEATMEHT COST ESTIMATES
12.6.1 General Discussion
To prepare treatment cost estimates, a model plant
concept was developed. The proposed BPT model treatment
consists of:
A. Slurry transportation of kiln solids to an equalization
basin.
B. Application of lime to precipitate fluoride and toxic
metals, followed by lagoon settling,
C. pH adjustment before final discharge.
D. Scrubber, cooling, and distillation wastes enter the
equalization basin.
It is assumed that drip acid is recycled to the process
reactor and does not appear directly in the waste stream.
For new or remodeled production facilities, the NSPS model
treatment system is based on hauling dry kiln residue directly
to a landfill. Miscellaneous liquid wastes in the NSPS model
are subjected to two stage lime-soda ash
neutralization/precipitation, followed by filtration and
partial return of effluent for use in scrubbers.
Haste Water Flow
The data in Table 12-4 for plants sending the gypsum
solids to the treatment facility indicate that the unit flow
varies from approximately 31.0 m3/kkg of HP to 166 m3/kkg of
HF. For the model plants, a constant unit flow of 95.4 m3/kkg
of HF was assumed.
308
-------
HP Production
In the HP subcategory, production ranges from a minimum
of 7,300 kkg/year to a maximum of 62,000 kkg/year with a mean of
22,100 kkg/year and a median of 15,800 kkg/year. For waste
water treatment cost estimates, three production levels were
selected as model plants. These are 19,100 kkg/year, 38,200
kkg/year, and 57,300 kkg/year.
Waste Water Pollutant Load
The amount of kiln residue varies from 3.8 to 4.1 kg/kg of
HP produced. The waste water going to treatment model plants
is assumed to contain 3.8 kg of solid kiln residue per kg of
HF. Fluoride emissions in waste water have been shown to vary
as indicated below:
Source of Data Fluoride, (kg/kkg)
Reference 3 20
Reference 3 37
Screening and Verification 3.8 to 58
Phase Sampling
(Tables 12-8 and 12-9)
For the model plants, the average fluoride loading from
kiln wastes of 31 kg/kkg HF produced was used to establish
treatment requirements and related costs.
The costs shown at each level of treatment correspond to
the model plant BPT system (Level 1} and one or more
alternative BAT systems (Level 2, 3', and 4) which may add
to or modify the existing BPT system to meet more stringent
priority pollutant removal requirements. The BAT system also
provides a higher effluent water quality with respect to
the conventional and nonconventional parameters.
At each level of treatment, the cost elements
associated with the typical rates of effluent reuse have also
been included. However, the hydraulic loading on the
treatment system is unaffected by reuse, and, therefore the
total costs (including reuse) are independent of the particular
rate of reuse that may be practiced.
The estimated costs ,for three models having different
production levels are given in Tables 12-12, 12-13, and 12-
14. For these models, both the hydraulic and "the pollution
loads per unit of production are held constant over the
entire range of production. Annual treatment cost as a
309
-------
TABLE 12-12. MODEL PLANT TREATMENT COSTS
Subcategory HYDROFLUORIC
Production 19,100 n
54 n
Waste water flow 5220 <
A. INVESTMENT COST
Construction .........
Equipment in place,
including piping,
fittings, electrical
Monitoring equipment..
Engineering design
Incidentals, overhead,
fees, contingencies...
TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Maintenance. ..........
Taxes and insurance...
Residual waste disposal
Monitoring, analysis
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
ACID
letric tons per
netric tons per
ruble meters pe
FIRST
$877,500
356,000
30,000
9,000
254,500
254,500
1,020,000
$2,801,500
$56,000
14,000
534,800
172,650
6,500
84,045
350,000
15,000
1,232,995
$289,850
$1,522,845**
(1)
year (21
day (60
r day.
LEVEL OF
SECOND
$24,500
89,500
22,800
22,800
$159,600
$14,000
1,500
15,960
4,788
7,500
$43,748
$25,966
$69,714
,057 tons pe
tons per da
TREATMENT*
THIRD
»
$25,000
92,000
23,400
23,400
$163,800
$14,000
1,800
3,400
16,380
4,914
7,500
$47,994
$26,650
$74,644
(1)
r year)
y)
FOURTH
$24,500
89,500
22,800
22,800
$159,600
$14,000 '
1,500
367,700
15,960
4,788
7,500
$411,448
$25,966
$437,414
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
**!neluding $11,100 for the reuse of treated effluent to slurry
kiln residues, etc.
(1) Production year is 350 days.
310
-------
TABLE 12-13. MODEL PLANT TREATMENT COSTS
Subcategory HYDROFLUORIC ACID
(1) (1)
Production 38,200 metric tons per year (42,115 tons per year)
109 metric tons per day (120 tons per day)
Waste water flow 10450 cubic meters per day.
A. INVESTMENT COST
Equipment in place,
including piping,
fittings, electrical
Reuse facilities......
Monitoring equipment..
Engineering design
Incidentals, overhead,
fees, contingencies...
TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Taxes and insurance...
Residual waste disposal.
Monitoring, analysis
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
LEVEL OF TREATMENT*
FIRST SECOND THIRD FOURTH
$1,354,500 $35,000 $35,500 $35,000
493,500 131,000 137,500 131,000
40,000
9,000
379,400 33,200 34,600 33,200
379,400 33,200 34,600 33,200
1,944,000
$4,599,800 $232,400
$56,000 $14,000
19,500 3,100
1,069,600
257,580 23,240
10,000
137,994 6,972
700,000
15,000 7,500
2,265,674 $54,812
$432,098 $37,811
$2,697,772** $92,623
$242,200 $232,400
$14,000 $14,000
3,400 3,100
6,700 735,350
24,220 23,240
7,266 6,972
7,500 7,500
$63,086 $790,162
$39,405 $37,811
$102,491 $827', 973
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
**Includes $16,500 for the reuse of treated effluent to slurry
kiln residues, etc.
(1) Production year is 350 days.
311
-------
TABLE 12-14. MODEL PLANT TREATMENT COSTS
Subcategory HXDRQffLUORIC AGED
(1) (1)
Production 57,300 metric tons per year (63,173 tons per year)
163 metric tons per day (180 tons per day)
Waste water flow 15700 cubic meters per day.
A. INVESTMENT COST
Construction
Equipment in place,
including piping,
fittings, electrical
work and controls.....
Reuse facilities
Monitoring equipment..
Engineering design
and inspection.
Incidentals, overhead,
fees, contingencies...
Land
LEVEL OF TREATMENT*
FIRST SECOND THIRD
$1,755,500 $49,000 $50,000
848,000
50,000
9,000
203,500 215,500
FOURTH
$49,000
203,500
532,500
532,500
2,880,000
50,500
50,500
53,100
53,100
50,500
50,500
B.
TOTAL INVESTMENT COST
OPERATION AM)
MAINTENANCE COST
$6,607,500 $353,500 $371,700 $353,500
Labor and supervision.
Taxes and insurance...
Residual waste disposal.
Monitoring, analysis
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION Of
INVESTMENT COST
TOTAL ANNUAL COST
$56,000
28,000
1,604,400
362,350
13,000
198,225
1,050,000
15,000
3,326,975
$606,464
$3,933,439**
$14,000
4,600
35,350
10,605
7,500
$72,055
$37,514
$129,569
$14,000
4,900
10,070
37,170
11,151
7,500
$84,791
$60,475
$145,266
$14,000
4,600
1,103,025
35,350
10', 605
7,50i
$1,175,080
$57,514
$1,232,594
*FIrst level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
**Includes $21,200 for the reuse of treated effluent to slurry
kiln residues, etc.
(1) Production year is 350 days.
312
-------
function of production is shown graphically in Figure 12-10.
Similarly, treatment cost per metric ton of product is given
in Figure 12-11.
To indicate the effect on costs of an increased pollution
load per unit of production for the 38,200 kkg/year model plant,
the pollution load was increased by 100 percent and the
hydraulic load was held constant. The cost estimate indicated
that the annual unit cost per metric ton of product at first" and
fourth (incremental) levels of treatment increased
approximately 40 percent and 90 percent respectively over the
original model unit cost. The increased cost is due mainly to
the additional cost of chemicals. Increase of pollutant loading
had no effect on the unit cost of treatment at other levels of
treatment.
Similarly, for the same model plant, the hydraulic load
was increased by 100 percent and the pollutant load was
held constant. The cost estimate indicated that the annual
unit cost per metric ton of product at the second and fourth
levels of treatment increased approximatley 70 percent and 10
percent respectively over the original model unit cost. There
was no significant impact on the unit cost at other levels of
treatment.
Table 12-15 presents a summary of the unit cost
distribution between amortization and operation and maintenance
cost components at various production rates and levels of
treatment.
At the second, third and fourth levels of treatment, the
cost estimates are based on part of the waste water flow being
recirculated and the remaining flow being treated, thus the
subsequent treatment units are sized and estimated for lower
flows than if recycling were not practiced,
12.6.2 Model Plant Control Costs for Existing Sources
For the model plant control costs for existing sources at
the first -level of treatment, the disposal of the sludge is
on-site and hence the land requirements are fairly large.
Chemicals, sludge hauling, and .disposal costs have a
significant impact on the total annual costs. At the second
and third levels of treatment however, amortization, labor and
supervision costs constitute a major portion of the additional
annual costs.
The fourth level of treatment is designed for recirculation
of the major portion of the treated effluent and therefore, soda
ash is used for neutralization in place of lime. Due to this
change, chemic.al cost has a significant impact on the additional
annu al cos ts.
313
-------
a
o
o
^
o
o
o
•9
<4
Cd2
14
x
X
•BV:SL
Figure 12-10.
10 20 30 40 50 60
PRODUCTION (METRIC TONS/YESR X 1000)
Jtonual treatinent cost vs. production for the Hydrofluoric
Acid Subcatecpry
314
-------
110
100
90
80
70
60
10 20 30 40 50
EEODDCTICN (METRIC 1OTS/YEAR X 1000)
Figure 12-11. jtonual unit treatment cost vs. production for the
Hydrofluoric Acid Subcategory
315
-------
TABLE 12-15. PLANT COSTS
Subcategory HYDROFLUORIC ACID
Annual Treatment Costs ($/kkg) of HP
COST ITEMS
PRODUCTION FtOW
(kkg/yr) (m3/day)
LEVEL OF TREA1MENT
FIRST SECOND THIRD FOURTH
Annual Operation
and Maintenance
Annual
Amortization
Total Cost
19,100 5,220 64.55 2.29 2.51
38,200 10,450 59.31 1.43 1.65
57,300 15,700 58.06 1.26 1.48
19,100 5,220 15.18 1.36 1.40
38,200 10,450 11.31 0.99 1.03
57,300 15,700 10.58 1.00 1.06
19,100 5,220 79.73 3.65 3.91
38,200 10,450 70.62 2.42 2.68
57,300 15,700 68.65 2.26 2.54
21.54
20.68
20.51
1.36
0.99
1.00
22.90
21.67
21.51
316
-------
12.6.3 Model Plant Control Costs for New Sources
The basis of the selection of the model plant representing
a new source is described earlier in this section. The
estimated costs for three different models, having three
different production levels are given in Tables 12-16, 12-17,
and 12-18. Both the hydraulic and pollutant loads are directly
proportional to the production, i.e., -the waste flow per unit
of production and the pollutant loading per unit of production
are held constant.
Annual treatment cost as a function of production is shown
graphically in Figure 12-12. Treatment cost per metric ton of
product is given in Figure 12-13.
Table 12-19 presents a summary of the unit cost
distribution between amortization and operation and maintenance
components.
For the model plant, the dry solids generated in the kiln
are hauled to approved chemical dump sites, eliminating kiln
waste slurry. The waste water sources are air pollution control
(scrubbers), leak, spills, and washdowns.
The cost of transporting dry kiln waste 'sludge to the
approved chemical dump site has been included in the cost
estimates. The cost of conveying the dry solids from the kiln
operation to the trucks (for transporting to the dump site) is
not included in the cost estimate. Such costs, which can vary
widely with site conditions, are considered to be process
costs and not part of treatment. However, if such costs are
to be considered as part of the treatment costs, then the
estimated total annual costs per metric ton of product for the
three model plants would ,be as follows:
Production Flow Total Annual Cost
(kkg/year) (m3/day) ($/kkg)
19,100 680 14.81
38,200 1,370 9.68
57,300 2,030 8.03
Since the sludge disposal is not on site, the land cost
has negligible impact on total annual cost. However, the cost
of transporting the dry solids to the dump site constitutes
about 75 percent of the annual costs.
317
-------
TABLE 12-16.
MODEL PLANT TREATMENT COSTS
Subcategory HYDROFLUORIC ACID
Production
Waste water flow
(1)
19,100 metric tons per year
54 metric tons per day
680 cubic meters per day.
(1)
(21,057 tons per year)
(60 tons per day)
A. INVESTMENT COST
Construction
Equipment in place,
including piping,
fittings, electrical
work and controls
Monitoring equipment
in place
Engineering design
and inspection
Incidentals, overhead,
fees, contingencies...
Land
TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Energy
Chemicals.
Maintenance
Taxes and insurance...
Residual waste
disposal
Monitoring, analysis
and reporting
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
LEVEL OF TREATMENT*
FIRST
$64,000
327,000
9,000
80,000
80,000
30,000
$590,000
$56,000
6,100
44,000
56,000
17,700
742,000
15,000
$936,800
$91,112
$1,027,912
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
(1) Production year is 350 days.
318
-------
TABLE 12-17. MODEL PLANT TREATMENT COSTS
Subcategory HYDROFLUORIC ACID
(1) CD
Production 38,200 metric tons per year (42,115 tons per year)
109 metric tons per day (120 tons per day)
Waste water flow 1370 cubic meters per day.
LEVEL OF TREATMENT*
FIRST
INVESTMENT COST
Construction $94,500
Equipment in place,
including piping,
fittings, electrical
work and controls 468,500
Monitoring equipment
in place.... 9,000
Engineering design
and inspection... 114,400
Incidentals, overhead,
fees, contingencies... 114,400
Land 60,000
TOTAL INVESTMENT COST $860,800
B. OPERATION AND
MAINTENANCE COST
Labor and supervision. $56,000
Energy 8,300
Chemicals 88,000
Maintenance • 80,080
Taxes and insurance... 25,824
Residual waste
disposal.. 1,480,000
Monitoring, analysis
and reporting 15,000
TOTAL OPERATION AND
MAINTENANCE COST " $1,753,204
C. AMORTIZATION OF
INVESTMENT COST $130,290
TOTAL ANNUAL COST $1,883,494
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
(1) Production year is 350 days.
319
-------
TABLE 12-18. MODEL PLANT TREATMENT COSTS
Subcategory HYDROFLUORIC ACID
(1) (1)
Production 57,300 metric tons per year (63,173 tons per year)
163 metric tons per day (180 tons per day)
Waste water flow 2030 cubic meters per day.
LEVEL OF TREATMENT*
FIRST
INVESTMENT COST
Construction $120,700
Equipment in place,
including piping,
fittings, electrical
work and controls 601,000
Monitoring equipment
in place 9,000
Engineering design
and inspection 146,140
Incidentals, overhead,
fees, contingencies... 146,140
Land 84,000
TOTAL INVESTMENT COST $1,106,980
B. OPERATION AND
MAINTENANCE COST
Labor and supervision. $56,000
Energy 12,250
Chemicals 132,000
Maintenance 102,298
Taxes and insurance... 33,209
Residual waste
disposal 2,226,000
Monitoring, analysis
and reporting 15,000
TOTAL OPERATION AND
MAINTENANCE COST $2,576,757
C. AMORTIZATION OF
INVESTMENT COST $166,438
TOTAL ANNUAL COST $2,743,195
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
(1) Production year is 350 days.
320
-------
o
« 2
O
O
o
w-
10 20 30 40 50 60
PIODUCTICN CMETOIC TCNS/YEAR x 1000)
Figure 12-12. Annual treatment cost vs. production for the
Hydrofluoric Acid Subcategory
321
-------
60
50
LIE''
'EL
40
30
J_
10 20 30 40 50
PKCOJCTION CMETKCC TOTS/YEAR X 1000)
60
Figure 12-13. Annual unit treatment cost vs. production for
the Hydrofluoric Acid Subcategory (NSPS)
322
-------
TABLE 12-19. MODEL PLANT TREATMENT COSTS
Subcategory HYDROFLUORIC ACID
Annual Treatment Costs ($/kkg) of HF
COST ITEM
LEVEL OF TREATMENT*
PRODUCTION FLOW FIRST SECOND THIRD FOURTH
(kkg/yr) (m3/day)
Annual Operation
and Maintenance
Annual
Amortization
Total Cost
19,100
38,200
57,300
19,100
38,200
57,300
19,100
38,200
57,300
680
1,370
2,030
680
1,370
2,030
680
1,370
2,030
49.05
45.90
44.97
4.77
3.41
2.90
53.82
49.31
47.87
Not Applicable
* Only applies to first level,
323
-------
12.7 BASIS FOR REGULATIONS
12.7.1 Evaluation of BPT Treatment Practices
Control and treatment practices for eleven plants
producing HF are presented in Table 12-20. Also indicated
are other product-related waste water sources and pollutant
loads discharged.
It is clear from the table that a wide variation in
effluent quality exists within this subcategory. The factors
believed to cause these variations are the following:
Dry Residue Handling
The disposal of kiln waste by dry handling rather than
slurrying is practiced currently at three plants. This process
eliminates the major source of waste water generated at most
plants, greatly reducing the raw waste loads to be treated.
The only sources of waste water remaining are from air pollution
control and washdown,
Effluent Reuse
Reuse of treated waste water for slurry transport of kiln
wastes is commonly practiced to varying degrees and clearly has
a major effect on pollutant loads discharged. Although four
plants do not practice reuse, it has been demonstrated
sufficiently that this practice is both technologically and
economically feasible.
Recycle of Condensables
Recycling of drip acid or condensable cooler bottoms
reduces the loading of fluoride in the treated effluent
since the fluoride species (fluorosulfonic acid) in this
material is not removed by conventional lime treatment. Only
two plants do not recycle drip acid.
Other Related Products
Most hydrofluoric acid plants also discharge wastes from
related products such as aluminum fluoride, fluorocarbons,
hexafluorosilicic and tetrafluoroboric acids to treatment.
These other product wastes can account for higher raw waste
loadings and increase the potential for complex fluorides
formation and can also impact treatment efficiency by diluting
the raw waste. In addition, commingling of other product
wastes will limit the percentage of reuse of the total plant
treated effluent.
324
-------
, TKBCB 12-20. SUMMARY OS WASfE SATES QUaStUi HO} TREATMENT radWOKXBf EMPLOYED AT HYDROFLUORIC ACID PUNTS(1)
Plant
* 426
i 664
* 167
1 120
Product-Related
Haste Water Sources
Hydrofluoric acid
fluosilicic acids
production
Hydrofluoric acid
production
Hydrofluoric acid,
fluorocarbon,
Color tn*/sodiu»
hydroxide, and
hydrochloric acid
production
Hydrofluoric acid
production
Control and Treatment
Technology Employed
Dry residua hauling
and dumping; neutra-
lization with caustic
of noncontact cooling
water and floor
drainage
Residue slurry, neutra-
lization with sodium
carbonate, settling,
recycle
Residue slurry, lime
treatment , settling ,
recycle
planned dry residua
handling, lime
Amount of Cooler Bottoms Effluent volume
Treated (Condenaables) in m3 /metric ton
Waste Water Recycled? (gal/short ton) of
Reused Actual Production
0 Tea 465 (111,397)
includes noncon-
tact cooling
water
94% Yes 5.78 (1,360)
47% Ye» 103 (24,200)
0 Yes m
Long Term
Average Pollutant
wasteload Discharged
(kg/metric ton)
(lb/1000 Ibj
Fluoride »SS
1.2 HD
0.10 0.27
18 0.4S (Net)
ND NO
<3
t 967 Hydrofluoric acid,
fluorocarbon, and
sulfuric acid
production
treatment, clarification
Residue slurry, settling Present: 0
(Recycle and pH 'planned: 70%
polishing facilities to 75%
under construction.)
* 928 Hydrofluoric acid
and aluminum
fluoride production
I 837 All hydrofluoric
acid genorated as
used captivoly for
aluminum fluoride
production
f 753 Hydrofluoric acid
production
Residue slurry, settling, 83%
recycle (Flocculation,
lime treatment, and
clarification facilities
under construction.)
Residue slurry, line 0
treatment, settling
Residue slurry, lime 65%
treatment, settling,
recycle, pH polishing
Yes
Yes
Ye»
Yes
125 (30,000) Present: 24 16
Expected
with 1.8 2.1
additional
facilities
9.44 (2,260) Present! 1 1.7
Sxpected
with 0.65 0.75
additional
facilities
134 (32,200)
11.0 (2,650)
1.8 3.1
0.64 0.38
t 251 (2!
* 70S<2»
* 722
* 987 (2>
HF, AlFt, chlorine/ Residue slurry, settling, 0
sodium hydroxide, neutralization
aluminum oxide, and
fluorocarbon
production
Hydrofluoric acid Residue slurry, lime 30% to 35%
and aluminum treatment, settling,
fluoride production recycle, pH polishing
Hydrofluoric and, Residue slurry, lime 92% to 100%
in recent past, treatment, settling,
fluoboric, recycle, pH polishing
acid production
Hydrofluoric! Acid Dry residue hauling 0
1 Kiln: Yes 22,2 x 10* 46 530
3 Ktlnss Ho (553 x 10*5
So 25.9 (8,204) 3.2 0.64
Yes 0-10.3 (0-2,460) 0-0.81 0 to 0.54
Yes 8.8 ND ND
(1) Adapted from Calapan (Reference 3).
(2) Hydrofluoric Acid production has been discontinued at these plants since the time of sampling.
(3! affluent loading less the influent loading.
HD » Hot determined,
-------
In addition to the above factors, the design and operation
of the treatment facilities affect effluent quality. Solids
removal depends on retention time and surge capacity.
Precipitation of fluoride requires careful pH control and in
areas of heavy rainfall or winds, adequate freeboard or multiple
ponds are necessary to limit the discharge of high pollutant
loads due to unfavorable climatic conditions.
Pollutant Removal with BPT Treatment
Treatment level 1 is equivalent to the proposed BPT in the
Hydrofluoric Acid industry. Table 12-21 presents a summary of
long term effluent monitoring data on total suspended solids
(TSS) and fluoride from four plants. Means, standard
deviations, and variability factors are given where sufficient
data are available. These performance characteristics are later
utilized for the development of the proposed regulations on TSS
and fluoride.
The ability of BPT treatment to remove toxic pollutants
can be estimated by comparing the raw waste data presented in
Table 12-10 with the corresponding treated effluent data
presented in Table 12-22. The latter expresses the removal
efficiency as the calculated average percent removal observed
at these plants. The BPT removal efficiency for some of the
toxic metals is undoubtedly augmented to some degree by the
fact that the raw waste may carry insoluble forms of the metals
that were never completely leached out of the ore. Removal of
these forms would take place simply by settling out? however,
the effluent concentrations of some metals such as chromium,
nickel, and zinc remain at concentrations higher than should be
achievable by alkaline precipitation. This suggests that these
metals are largely in solution coming into the treatment system
and that the optimum conditions for metal hydroxide formation
were not being attained at the time of sampling.
\
The original BPCTCA limitations for this subcategory shown
in Table 12-2 required zero pollutant discharge except during
periods of excess rainfall. Objections to the-zero-discharge
limitations concerned the feasibility of using gypsum-saturated
water for reuse in the air pollution control scrubbers.
The proposed BPT waste water control and treatment
technology allows for the discharge of process waste water after
appropriate treatment. This technology is practiced widely in
the industry and should pose no technical problems.
Implementation of BPT at all sites in the industry will achieve
the indicated pollutant discharge levels.
The nine plants presently producing hydrofluoric acid all
have installed BPT treatment or the equivalent. At the time *bf
326
-------
TABLE 12-21. SUMMARY .OF LONG TERM MONITORING DATA FROM FOUR
(1)
HYDROFLUORIC ACID PLANTS
Treated Waste Load (kg/kkg) or (Ib/lOOOlb)
Daily Data 30-Day Average Data
Long Term (2) Long Term (2)
Plant Avejrage St.Dav. W Ave£age St.Dev. VF
No. Parameter (X) (S) (Sg) (X) (S)
#664
#753
#722
Fluoride
TSS
Fluoride
TSS
Fluoride
TSS
0.10
0.29
0.72
0.38
0.81
0.54
0
0
0
0
0.090 0.77 4.5 0.10 0.040 1.7
0.27 0.36 2.2 0.64 0.15 1.4
0.52 ' 0.59 3.3 -- — —
0.37 0.62 3.5 — — —
(3)
#705 Fluoride — — — — 0.49 0.22 1.7
TSS — _•____ o.84 0.37 1.7
(1)
Based on Reference'3 data.
(2)
In the case of daily measurements, the variability factor, W,
for a lognormal distribution is found by the expression ln(W) -
S1(Z - 0.5S1), where S" is the estimated standard deviation of
the logarithm derived from the arithmetic mean, X, and the
arithmetic standard dev:
(S')2 = in
[1.0 +f_Sj\2'
L ^x/ _
ation, S, according to the relationship,
.When the value of Z is 2.33, the
variability factor for the 99 percentile is obtained.
For 30-day average measurements, a normal distribution is
obtained and the variability factor is found by the expression,
W = 1.0 + Z f S\. When the value of Z is 1.64, the >
VxV
variability factor is for the 95 percentile. Please refer to
Section 8.2 for a more detailed discussion of the statistical analysis
of long term data.
(3)
Although Plant #705 does not recycle the drip acid, the TSS
data is not adversely affected and is used as the basis for
the 30-day average W.
— Not Available.
-------
TABLE 12-22. TOXIC POLLUTANT TREATED EFFLUENT OATA
SUBCWEGQRY: HXDROFLUORIC &CID
(U
Average Dally fellutant Concentrations and loadings at Plants Sampled
(1B3/1)
(kgAkg of anhydrous HP)
(2>
Overall Average
f705(8) 1705 (V) 1251 (V) H67(V) Average S Removal
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
ffercury
Nickel
Selenium
Thallium
Zinc
<0.010
<0. 00021
<0.0030
<0. 000063
0.00030
0.0000060
0.014
0.00029
0.10
0.0021
0.0060
0.00012
<0. 00040
<0. 0000080
0.050
0.0010
0.033
0.00069
0.0070
0.00015
0.071
0.0015
<0.0020
<0. 000042
<0.010
<0. 00021
<0.0017
<0. 000035
<0.046
<0. 00096
<0.020
<0. 00042
<0.022
<0. 00046
<0. 00050
<0. 000010
<0.010
<0. 00021
<0.0050
<0. 00010
<0.0012
<0. 000025
0.053
O.OOU
<0.17
<0.017
<0.020
<0.0020
<0.0020
<0. 00020
0.22
0.022
0.070
0.0069
<0.031
<0,0031
<0.0010
<0. 00010
0.52
0.052
<0.071
<0.0070
<0.0070
<0. 00069
0.16
0.01S
0.047
0.012
0.016
0.0040
0.0087
0.0022
0.050
0.013
0.060
0.015
0.010
0.0026
0.0065
0.0017
0.090
0.023
0.010
0.0026
0.0030
0.00069
1.9
0.49
<0.057
<0.0073
<0.012
<0.0016
<0.0032
<0. 00060
<0.083
<0.0091
<0.063
<0,0061
<0.017
<0.0015
<0.0020
<0. 00044
<0.17
<0.019
<0.030
<0.0025
<0.0045
<0. 00039
0.55
0.13
74
81
9
9
62
77
97
67
<54
Effluent
>Influent
85
83
(S) Screening data from one 72-hour composite sample of treated
effluent.
(V) Verification data from three 24-hour composite samples.
(1) The effluent data presented here corresponds to the raw waste
data shown in Table 12-10. The methodology of the sampling
program is described in Section 5.1.2, and the scope of
sampling in the Hydrofluoric Scid industry is described in
Section 12.3.3,
(2) Wien averaging values indicated as "less than" «}, the
absolute value MBS used and the resulting average was indicated
as a "less than" value.
328
-------
sampling, seven of the 12 plants operating were meeting the
proposed fluoride limitations and eight were meeting the
proposed TSS limitations according to the data available.
Although there is practically no long term monitoring data
available to support the additional proposed limitations on
toxic metals, the screening and verification data indicate
that all three plants sampled are meeting the proposed
limitations on antimony, copper, and lead, while two of the
plants are meeting the proposed zinc limitation and one plant
is meeting the proposed chromium and nickel limitations.
With the limited amount of toxic metal data, it is not possible
to estimate compliance or noncompliance on a statistical
basis. The Agency is conducting additional treatability
studies.
12.7.2 Basis for Proposed BPT Effluent Limitations
Technology Basis
For BPT, the Agency is proposing limitations for which the
technology basis is, or is equivalent to, equalization, lime
neutralization/alkaline precipitation, solids removal by
settling or thickening, final pH adjustment, and discharge
of clarified effluent. The in-house process recycling of
the reactor condensables (drip acid) is required for meeting
the proposed fluoride limitations.
Flow Basis
The reuse of treated waste water to slurry kiln residues
to the treatment system is not required for meeting the BPT
limitations. BPT or its equivalent is practiced by all plants
in this industry including six which reuse, for slurrying
residues, proportions of their treated waste water ranging from
30 to 100 percent of the plant flow as shown in Tables 12-4 and
12-20.
The practice of reusing waste water in this manner has two
opposing effects on the plant effluent:
A. A decrease in the net discharge unit flow rate
(m3/kkg), and
B. An increase in the fluoride concentrations (mg/1).
As a result, the fluoride unit loading (kg/kkg) in the
effluent does not decrease as a direct proportion to the
decrease in the flow rate, but is partially offset due to the
increase in fluoride concentration as a function of percent
reuse. The relationship of percent water reuse to fluoride
concentrations and unit loadings is shown in Figure 12-14. The
apparent reason for the increase in fluoride concentration
with reuse is a calcium deficiency which may result from the
329
-------
2'5 *
Percent Reuse
I'OO
SO
Percent Reuse
75
100
I12GEND
• Long-term data
O Expected with treatment system upgrading
D Screening and verification sampling results
Figure 12-4.4
FluorMe loads and concentrations discharged at
selected hydrofluoric acid plants.
330
-------
buildup of sulfate or bicarbonate concentrations in the
treatment systems at plants where reuse is practiced. Other
pollutants such as TSS and metals would not be expected to
exhibit similar concentration offset effects in these systems.
It should be noted that while the practice of reusing
waste water for kiln residue slurrying may be advantageous in
some locations with respect to alternative water supply costs,
there is no associated reduction in the hydraulic load, size,
or cost of the BPT treatment system itself.
The net result of water reuse is a moderate decrease in the
effluent fluoride loadings which is achieved at a small
additional annual cost of less than one percent of the estimated
BPT treatment systems cost (Tables 12-12r 12-13, and 12-14).
The model plant BPT treatment system is based on an inflow
rate of 95.4 m3/kkg derived from the average of nine plants
which handle the kiln residues in a slurry system as shown in
Table 12-4. The treated effluent flow rate is 54.6 m3/kkg
which is the average effluent flow rate for the same nine
plants and corresponds to the reuse of about 43 percent of the
flow for residue slurrying and other uses.
Selection of Pollutants to be Regulated
The selection of pollutants for which specific numerical
effluent limitations are proposed was based on an evaluation of
raw waste data from the screening and verification sampling
program. The following two major factors were considered:
Raw waste pollutant concentrations - A tabular summary of
maximum raw waste concentrations is presented in Section
12.3.3. Data from the one plant sampled for screening were
used to determine the need for verification sampling. The
maximum concentrations found during verification are also shown
for comparison. For each pollutant, the maximum concentration
observed gave a preliminary indication of its potential
significance in the subcategory. On this basis, the preliminary
selection of candidates for regulation included zinc, lead,
antimony, nickel, chromium, and copper in decreasing order of
their apparent pollution potential. These pollutants were
observed at least once during the sampling program at
concentrations considered treatable in this industry using one
of the available treatment technology options. The other
metals, cadmium, thallium, and mercury exhibited maximum
concentrations that were considerably lower.
Total subcategory _raw waste pollutant loadings -Pollutant
raw waste loading data were used to evaluate the overall
magnitude of the pollution potential for the subcategory. Data
from the plants sampled are presented in Table 12-10 and the
331
-------
daily and unit loadings are summarized in Table .12-11. This
information, coupled with the estimated total hydrofluoric
acid production rate of 261,800 kkg/year, yielded the
approximate total annual pollutant loading rates for the
subcategory shown in Section 12.3.3. This method of ranking
the pollution potential of the observed toxic metals confirmed
the maximum concentration based ranking and indicated that
zinc, nickel, lead, antimony, copper, and chromium were the
six dominant toxic metals in terms of both total mass
loadings and treatable raw waste concentrations.
In view of the treatment technology already implemented in
this industry, the added BPT regulation of any one of these
pollutants may provide assurance that all of the observed
toxic metals would receive adequate treatment and control.
This includes taking credit for incidental removal of metals
which are either below practical treatability limits or
are not particularly amenable to removal by alkaline
precipitation methods. The latter includes cadmium, selenium,
thallium, and mercury. Thus, because zinc, nickel, lead,
antimony, copper, and chromium were observed most frequently at
treatable concentrations and may serve as reliable indicators of
overall treatment system performance, these metals were selected
as the additional parameters proposed for BPT regulations.
Basis of Pollutant Limitations
Conventional and nonconventional parameters -
A. pH: The treated effluent is to be controlled within
the range of 6.0 to 9.0. This limitation is based on the
data presented in Appendix B of this report and the JRB Study
(52).
B. TSS and Fluoride: The data presented in Tables 12-
20 and 12-21 were used for the development of TSS and
fluoride limitations. However, because of the wide range of
product mixes, significant differences in residue handling,
waste water treatment, reuse practices, and dilution with other
product waste -streams, it was necessary to select only those
plants where the effect of BPT technology could be clearly
observed. The plants excluded are:
|426 and f!20 because kiln residues are handled as a dry
solid,
1167, 1967, and £251 because the combined treatment of HF
wastes along with the waste waters from other major
products generated high fluoride loadings in the large
volume discharges with fluoride at its minimum treatability
concentration,
33-2
-------
#705 because cooler bottom condensables (drip acids) are
not recycled back to the process but are added to the raw
waste contributing complex fluorides which tend to remain
in solution after lime treatment, TSS data are not
affected.
Data from the remaining five plants are presented in Table
12-23 which summarizes the development of the proposed
regulations for total suspended solids and fluoride. Since the
BPT level of treatment does not require the reuse of treated
waste water for slurrying kiln residues, the performance of
Plant f837 was used as the long term average unit loading basis
for the TSS and fluoride limitations. The variability factors
used for fluoride are based on the long term data from Plants
1664 and #753 and those used for TSS are derived from Plant #722
for daily measurements and Plant §705 for 30-day average
measurements as indicated in Table 12-23.
The proposed maximum 30-day average TSS limitation was
obtained by multiplying the variability factor for 30-day
averages from Table 12-23 by the long term average waste load;
i.e., 1.7 x 3.1 kg/kkg = 5.3 kg/kkg. Similarly, the daily
maximum TSS limitation was obtained by multiplying the
variability factor for daily measurements by the long term
average? i.e., 3.5 x 3.1 kg/kkg = 11 kg/kkg. The same approach
was taken to obtain the proposed fluoride limitations; i.e.,
1.6 x 1.8 kg/kkg = 2.9 kg/kkg for the maximum 30-day average,
and 3,4 x 1.8 kg/kkg = 6.1 kg/kkg for the daily maximum
limitation. These computations are shown on Table 12-23 and
the proposed BPT limitations are presented in Table 12-24.
The concentration basis (C) for each effluent is
derived from the relationship between concentration (C) ,
flow (Q), and unit loading,
C (as mg/1) = 1000 (L)
(Q)
Where L is the effluent limitation expressed as a unit
loading in kg of pollutant per kkg of product (kg/kkg) , and
Q is the flow rate expressed as cubic meters per kkg of
product (m3/kkg). (Note: kg/m3 = 1000 mg/1.)
Thus, the concentration basis for the maximum 30-day
average TSS limitation is:
(5.3 kg/kkg) (WOO mg/l\ = 97 mg/1
(54.6 m3/kkg) \ kg/m3 )
and the concentration basis for the daily maximum limitation is
obtained by a similar calculation or simply by applying the
333
-------
TABLE 12-23. DEVELOPMENT OF TSS AND FLUORIDE LIMITATIONS
Long Term Average
Waste Load Discharged
Plant
#837
#753
#928
#722
#664
Average of
practicing
(excluding
Reuse
(percent)
0
65
83
92
94
four plants
effluent reuse
#837)
Fluoride
(kg/kkg of HF)
1.8
0.72
1.0
0.81
0.10
0.'66
TSS
(kgAkg of HF)
3.1
0.38
1.7
0.54
0.29
0.73
Variability Factor for 3.4^ 3.5(5)
Daily Measurements
Variability Factor for 1.6(1) 1.7(6)
30-Day Averages
(?} (7}
Variability Factor Ratio (VFR) 3.4/1.6 = 2.r ; 3.5/1.7 = 2.r ;
Effluent Limitations for BPT
(from Plant #837) ,_. ,,.
a. Daily Max 3.4 X 1.8 kg/kkg = 6.1JJ 3.5 X 3.1 kg/kkg = 11 *~>
b. Max 30-Day Avg 1.6 X 1.8 kgAkg = 2.9^ ' 1.7 X 3.1 kg/kkg = 5.3W
Effluent Limitations for BAT
(from average of four plants) /.,%
a. Daily Max 3.4 X 0.64 kgAkg = 2.2JJ NA
b. Max 30-Day Avg 1'.6 X 0.64 kg/kkg = 1.01 ; NA
NA - Not Applicable
(1) Variability factor average of Plants #664,1722 and #753 from Table 12-21.
(2) Ratio of the daily (24-hr) variability factor to the 30-day
average variability factor. This value appears on the Proposed
Limitations tables.
(3) The long term average loading in kgAkg multiplied by the
variability factor for daily measurements as shown.
(4) The long term average loading in kgAkg multiplied by the
variability factor for 30-day measurements as shown.
(5) Variability factor from Plant #722, Table 12-21.
(6) Variability factor from Plant #705, Table 12-21.
334
-------
TABLE 12-24. PROPOSED LIMITATIONS
Hydrofluoric Acid
Best Practicable Control Technology Currently Available
Waste Water Flow: 54.6 m3/kkg of HP (43% Reuse) *
Subcategory
Pollutant Performance
(rag/1)
Conventional and
Nonconventional
Pollutants:
Total Suspended
Solids
Fluoride
Toxic
Pollutants:
Antimony
Arsenic
Chromium
Copper
Lead
Nickel
Selenium
Zinc
(2)
57
(2)
33
(3)
0.80
(3)
0.50
(3)
0.10
(3)
0.50
(3)
0.30
(4)
0.17
(3)
0.20
(4)
0.55
Concentration Basis
(1) (mq/1)
VFR
2.1
2.1
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
Max
30 -day
Avg
97
53
0.80
0.50
0.10
0.50
0.30
0.17
0.20
0.55
24-hr
Max
200
110
1.6
1.0
0.20
1.0
0.60
0.34
0.40
1.1
Effluent Limit
(ko/kkct) of HP
Max
30-day
Avg
5.3
2.9
0.044
_ » '
0.0055
0.027
0.016
0.0093
— (5)
0.030
24-hr
Max
11
6.1
0.088
— **'
0.011
0.054
0.033
0.019
_'5)
0.060
(1) - VFR: ratio of the 24 hour variability factor to the
30 day variability factor.
(2) - Long term average based on loading data and
variability factors selected from Table 12-21.
.(3) - The lower limit of the literature treatability estimate
(Table 8-11) is used as the basis for the 30-day average
limitation when the observed average of the sampling data
is below this level.
(4) - Average effluent concentration from screening and verification
sampling data.
(5) - Tito effluent limitation proposed.
* From Table 12-4.
335
-------
variability factor ratio, VFR, from Table 12-24 to the maximum
30-day average concentration; that is,
(VFR) (max. 30-day average concentration or loading)
= daily maximum concentration or loading In this
case, the daily maximum TSS concentration is 2.1 X 97 mg/1 =
201 mg/1.
In the same manner, the concentration basis for the
maximum 30-day average fluoride limitation is,
(2.9 kg/kkg) / 1000 mg/l'N = 53 mg/1
(54.6 m3/kkg) V
and the daily maximum fluoride concentration is 2.1 X 53 mg/1 =
111 mg/1. (Note: due to rounding off, this value differs just
slightly from the value that appears in Table 12-24 which was
obtained by calculating the concentration directly from the
daily maximum limitation; i.e.,
(6.1 kg/kkg) /1000 mg/A = 112 mg/1.
(54.6 m3/kkg) \ kg/m3/
In either case, only two significant figures should be taken.)
Performance evaluation and review of discharge quality has
been complicated by problems associated with chemical analysis."
Prior to July 1976, the methods generally used for the analysis
of fluoride in industry were specific ion electrode or
colorimetry. These methods did not detect the soluble complex
fluoride species present in the waste water. The best method
of total fluoride detection (free as well as complex) is
distillation followed by analysis using the specific ion
electrode. Using the distillation method, the complex fluorides
are hydrolyzed and the resulting HF is carried over with the
distillate along with any free HF in the sample. Thus, the
method of total fluoride analysis used for effluent monitoring
is capable of measuring free fluoride and the fluoride present
in the form of complex ions which are not removed by lime
treatment. Monitoring data on effluent fluoride levels using
the revised method are likely to be higher than the levels
previously reported under the same treatment conditions.
Toxic pollutants - The effluent limitations proposed for
the selected toxic pollutant control parameters are derived
from three sources of information. These are 1) screening
and verification sampling data, 2) literature based
treatability estimates (Section 8.1), , and 3) a limited
amount of long term monitoring data from Plant &251.
336
-------
The sampling results represent plant performance observed
during three days of sampling at each of the plants. The
treated effluent data on the toxic pollutants found at
significant levels are summarized in Table 12-22. The average
values shown for each pollutant are interpreted as being
approximately equal to a maximum 30-day average unless
there is some reason to believe that abnormal conditions
existed either in the process operation or in the treatment
system at the time of sampling. Abnormal conditions would
dictate that high values should either be excluded or
regarded as daily maxima rather than monthly averages.
For this subcategory, the screening and verification data are
believed to represent normal influent and effluent values at
the plants sampled.
For a number of the metal pollutants, the sampling data
demonstrate that the effluent quality and percent removal with
full scale BPT systems are considerably better than the
literature treatability data in Section 8.1 would indicate for
that particular technology. For example, even though arsenic,
cadmium, mercury, and thallium average influent concentrations
are well below the accepted treatability limits for
lime/settling shown in Table 8-11, greater than 60 percent
removals were observed for all but cadmium as is shown in
Table 12-22. This high degree of incidental removal supports
the contention that by applying effluent limitations just to
the dominant metal pollutant(s), an effective control of the
other metals may also be assured.
In Table 12-24, the concentration bases for the proposed
BPT limitations are derived from the averaged effluent
sampling data unless the observed pollutant concentration is
actually below the literature treatability level. In such
cases, the lowest applicable treatability level from Table
8-11 is used. This approach results in the setting of
achievable effluent limitations for all of the pollutants of
concern and provides for the possibility of wider variations
in the influent quality. Such variations may be associated
with different fluorspar impurity levels or other process
variables not fully taken into account by the limited data
obtained.
The basis for the proposed BPT limitations on each of the
selected metals is given below.
A. Zinc: The raw waste concentrations of zinc ranged as
high as 11.3 mg/1 (Section 12.3.3, Table of Maximum
Concentrations Observed) and averaged about 3.2 mg/1 (Table
12-10) for the plants sampled. BPT treatment achieved an
average removal of better than 80 percent with an average
performance concentration of about 0.55 mg/1 in the treated
effluent shown in Table 12-22. This level of performance
337
-------
approximately equals that obtained from the literature
treatability data in Table 8-11. The average performance
value is used as the concentration basis for the proposed
maximum 30-day average effluent limitation of 0.030 kg/kkg
using the model plant flow of 54.6 m3/kkg (Table 12-4). This
limitation was achieved by all but one of the plants sampled.
Using the model plant flow of 54.6 m3/kkg from Table 12-14, the
limitation was calculated as follows:
(0.55 mg/1)(54.6 m3/kkg) / kg/m3 \ = 0.030 kg/kkg
\1000 mg/1/
Since long term monitoring data on zinc are not
available from the HF industry, the variability factor ratio
(VFR) of 2.0 was selected on the basis of lead monitoring data
from Plant #251 presented in Tables A-lOa and A-lOc. This is
justified by the similarity in the chemistry of lead, zinc,
and the other metals of concern under BPT treatment conditions.
Thus,
VFR » VF of da ily measu rements
VF of 30-day averages
= 2.0
and the daily maximum limitation for zinc is,
,(2.0) (0.030 kg/kkg) = 0.060 kg/kkg.
The proposed effluent limitations on zinc and the other metals
of concern are given in Table 12-24.
B. Nickel: The sampling data indicate better than 60
percent BPT removal of nickel resulting in an average
effluent quality of about 0.17 mg/l.A The literature
treatability data in Table 8-11 show an effluent level
approximately equal to this value. Thus, 0.17 mg/1 is used as
the concentration basis for the proposed maximum 30-day
average effluent limitation of 0.0093 kg/kkg. A VFR of 2.0 was
used following the same rationale described for zinc. Thus,
the proposed maximum 30-day average limitation is,
(0.17 mg/1)(54.6 m3/kkg) / kg/m3 % » 0.0093 kg/kkg,
V1000
and the daily maximum limitation is,
'(2.0) (0.0093 kg/kkg) = 0.019 kg/kkg.
C. Lead: Because the observed average raw waste
concentration of lead (0.66- mg/1) was very close to the 0.30
338
-------
mg/1 lower limit of its estimated treatability according to
literature data, the latter was selected as the concentration
basis for the proposed maximum 30-day average effluent
limitation rather than using the observed performance average
of less than 0.02 mg/1. This results in setting the
limitation at 0.016 kg/kkg, a level which would be achievable
with BPT treatment even when higher influent levels occur.
It also avoids taking credit for incidental removal and a
higher removal efficiency than can be justified by the use of
this technology. A WR of 2.0 was used for lead on the basis
of long term data from Plant £251. The proposed maximum 30-
day average limitation is,
(0.30 mg/1) (54.6 m3/kkg) / kg/m3 \ = 0.016 kg/kkg,
V1000 mg/V
and the daily maximum limitation is,
(2.0) (0.016 kg/kkg) = 0.032 kg/kkg.
D. Antimony: In a manner similar to that described for
establishing the lead regulation, the concentration basis for
the proposed maximum 30-day average effluent limitation on
antimony was set at 0.80 mg/1 in accordance with literature
treatability data. The resulting limitation of 0.044 kg/kkg
was met in two of the four sampling data sets. A VFR of 2.0
was also used for antimony although a wider range of variation
may be observed when more operating data are collected. The
proposed maximum 30-day average limitation isr
(0.80 mg/1)(54.6 m3/kkg) ( kg/m3 ^ = 0.044 kg/kkg,
\1000 mg/1/
and the daily maximum is,
(2.0) (0.044 kg/kkg) = 0.088 kg/kkg.
E. Copper : The concentration basis for the proposed
maximum 30-day average effluent limitation on copper was set at
0.50 mg/1 in accordance with the literature treatability data.
All of the plants sampled had average loadings below the
proposed 0.027 kg/kkg limitation. A VPR of 2.0 was used
following the same rationale described for zinc. Thus, for
copper, the proposed maximum 30-day average limitation is,
(0.50 mg/1)(54.6 m3/kkg) / kg/m3 \ = 0.027 kg/kkg,
\1000 mg/1/
and the daily maximum is,
(2.0) (0.027 kg/kkg) = 0.054 kg/kkg.
339
-------
F. Chromium: The concentration basis for the proposed
maximum 30-day average limitation on chromium was set at 0.10
mg/1 in accordance with the literature treatability data. Two
of the plants sampled exceeded the proposed limitation of
0.0055 kg/kkg. A VFR of 2.0 was used following the same
rationale described for zinc. The proposed maximum 30-day
average BPT effluent limitation is,
(0.10 mg/1)(54.6 m3/kkg) / kg/m3 \ = 0.0055 kg/kkg
/ kg/m3 \
V1000 mg/1/
andf the daily maximum is,
(2.0)(0.0055 kg/kkg) = 0.011 kg/kkg.
G. Other metals: The concentration bases for arsenic
and selenium are also presented in Table 12-24. These are
intended to serve as guidance in cases where these pollutants
are found to be of serious concern.
12.7.3 Basis for Proposed BCT Effluent Limitations
For the Hydrofluoric Acid Subcategory, the Agency is
proposing BCT limitations applicable to total suspended solids
(TSS) based on the estimated performance of Level 2 (BAT)
treatment. Assuming that the addition of dual media filtration
to "the BPT system removes approximately 30 percent more
suspended solids, the maximum 30-day average TSS loading of 5,3
kg/kkg (Table 12-24) would be decreased to 3.7 kg/kkg as
follows:
(1.00 - 0.30)(5.3 kg/kkg) = 3.7 kg/kkg
By adjusting the loading to account for the decrease in
effluent flow rate from BPT (54.6 m3/kkg) to BAT (33.4 m3/kkg),
the proposed BCT maximum 30-day average effluent limitation
becomes,
(3.7 kg/kkg) /33.4 m3/kkg\
V54.6 m3/kkg/
* 2.3 • kg/kkg
The corresponding daily maximum limitation is then obtained
by applying the VFR value of 2.1 (Table 12-24). That is,
(2.1)(2.3 kg/kkg) = 4.8 kg/kkg
340
-------
12.7.4 Basis for Proposed BAT Effluent Limitations
The Application of Advanced Level Treatment
Utilizing the cost estimates presented in this report, the
Agency has analyzed the cost/ effectiveness of the base level
systems (BPT) and the various advanced level options for
conventional, nonconventional and toxic pollutant removal. The
economic impacts on the Hydrofluoric Acid industry have been
evaluated in detail (53) and taken into consideration in the
selection of the technology basis for the proposed BAT
regulations.
For BAT, the Agency is proposing limitations based on
treatment consisting of Level 2 technology. It has been
estimated that this will remove 11,100 pounds per year of toxic
metals and 104,000 pounds per year of fluorides in addition to
the pollutant removals already being achieved by BPT treatment.
The Agency considered the use of treatment Level 3
(addition of sulfide precipitation) but rejected it due to
lack of performance data. EPA also considered Level 4, a
variation of Level 2, that would substitute soda ash in the lime
precipitation step and allow 90 percent recycle of effluent.
This option was rejected due to being prohibitively expensive.
Pollutants limited by the proposed BAT are fluoride, antimony,
chromium, copper, lead, nickel, and zinc.
Technology Basis
For BAT, the Agency is proposing more stringent effluent
limitations on fluoride and the toxic metals based on the
addition of dual media filtration or its equivalent to the BPT
treatment system, coupled with the requirement of at least 65
percent effluent reuse for kiln residue slurrying. This
technology aims at both the reduction of suspended solids
containing fluorides and metal precipitates and the reduction
of the dissolved component loadings of these substances in the
final effluent. The minimum reuse rate of 65 percent was
selected because it is typical of the five plants (Plants f!67,
#753, 1928, ^664, and f-722) which presently practice reuse as
is shown in Table 12-4.
Flow Basis
With the model plant inflow rate of 95.4 m3/kkg and the
reuse of 65 percent of the treated effluent, the quantity
discharged is 33.4 m3/kkg; i.e., (1.00 -0.65) (95.4 m3/kkg) =
33.4 m3/kkg.
341
-------
Selection of Pollutants to be Regulated
For the BAT regulations, the Agency has selected fluoride
and the same six toxic metals identified in the proposed BPT
regulations. The rationale for their selection is discussed in
Section 12.7.2.
Basis of Pollutant Limitations
Nonconventional pollutants - The only nonconventional
pollutant is fluoride. The limitation proposed for BAT is based
on the performance of the four plants shown in Table 12-23 that
presently reuse at least 65 percent of their treated effluent.
The long term average effluent loading taken from Table 12-23
is 0.66 kg/kkg for the four plants and this is equal to the
performance of Plant f753 which reuses 65 percent, the lowest
reuse rate of the four plants. Although these plants do not
employ the filtration technology which is the basis for the
BAT regulation, the use of this performance in conjunction with
the 30-day average variability factor of 1.6 and the model
plant net discharge rate of '33.4 m3/kkg results in a
calculated maximum 30-day average concentration of 30 mg/1
total fluoride. Thus, the maximum 30-day average limitation
is,
(1.6) (0.66 kg/kkg) = 1.1 kg/kkg
and its concentration basis is,
(1.1 kg/kkg? /1000 mg/lN = 33 mg/1
(33.4 inS/kkg) \kg/m3 )
This represents a 58 percent reduction in fluoride
concentration in going from BPT (43 percent reuse) to BAT (65
percent reuse plus filtration). The use of a fixed loading
limitation allows the permissible concentration to increase
as a function of percent reuse. The proposed daily maximum
limitation on fluoride is obtained by utilizing the long term
average and variability factor for daily measurements,
(3.4) (0.66 kg/kkg) = 2.2 kg/kkg
and the concentration basis is,
(2.2 kg/kkg) /lOOOmg/l\ = 66 mg/1
(33.4 m3/kkg) ykg/mS )
The variability factors used for the BAT limitations on
fluoride are the same as for BPT shown in. Table 12-23. The
proposed BAT limitations for the Hydrofluoric Acid
Subcategory are presented in Table 12-25.
342
-------
TABLE 12-25. PROPOSED LBHTATIONS
Hydrofluoric Acid
Best Available Technology
Waste Water Plow: 33.4 m3/kkg of HF (65% Beuse)*
Concentration Basis Effluent Limit
(1) (mg/1) (kg/kkg of HF)
rua.j.uucuiu J
Nonconventional
(2)
Fluoride, F
Toxic
Pollutants:
(2)
Antimony1 '
Arsenic
Chromium^
Copper'2'
Lead ~(2)
Nickel (2*
Selenium
Zinc (2)
', L cauauii jLi_y
(mg/1)
Pollutants:
33
0.70
0.50
0.040
0.29
0.060
0.15
0.18
0.52
vrn — — — — •—
30-day
2.1 33
2.0 0.70
2.0 0.50
2.0 0.040
2.0 0.29
2.0 0.060
2.0 0.15
2.0 0.18
2.0 0.52
24-hr
Max
66
1.4
1.0
0.080
0.58
0.12
0.30
0.36
1.0
30-day
1.1
0.023
_J4)
0.0013
0.0097
0.0020
0.0050
_J4>
0.017
24-hr
Max
2.2
0.047
_J4)
. 0.0027
0.019
0.0040
0.010
_J4>
0.035
(1) - VFR: ratio of the 24 hour variability factor to the
30 day variability factor.
(2) - Also applicable for PSES limitations.
(3) - 30-Oay average calculated for the model plant based on
data In Table 12-2J..
(4) - No effluent limitation proposed..
* !he effluent flow rate is 35 percent of the average influent
shown in Table 12-4 (i.e., 0.35 X 95.4 is3/kkg = 33.4 m3/kkg).
343
-------
The estimated performance of Level 3 and Level 4
alternative technologies are presented in Tables 12-26
and 12-27, respectively. For these options, the expected
performance is expressed in the achievable maximum 30-day
average and daily (24-hour) maximum concentrations. These
are presented for comparison purposes and are not the
bases for any proposed regulations.
The Agency is currently conducting treatability studies on
dual media filter performance with HF industry wastes. The
results will be available prior to promulgation.
Toxic pollutants - For BAT regulations, the EPA is
proposing more stringent controls on the discharge of the six
toxic metals of concern on the basis of a reduced volume of
dicharge and physical removal by filtration. Alkaline
precipitation converts most of the dissolved metals into less
toxic, insoluble forms such as hydroxides and hydrated oxides.
Other mechanisms of removal including coprecipitation and
flocculation are undoubtedly involved during the treatment
process and probably account for a substantial portion of the
removal of certain toxic metals. Because there is no directly
applicable data on filter performance in the HF industry,
literature treatability studies (40, 41) have been evaluated in
order to estimate the probable efficiency of filtration for
the removal of residual suspended metal precipitates. The
following information was derived from pilot scale tests on
raw municipal, waste water samples spiked with toxic metals,
treated with lime, and settled, followed by dual media
filtration of the clarified effluent:
Removal of Suspended Metal Precipitates by Filtration
(Percent)
Antimony 7
Arsenic 0
Chromium (III) 60
Copper 42
Lead 80
Nickel 14
Selenium 12
Zinc 6
The filter efficiency values have been used in setting
the proposed BAT limitations. The basis for the limitation on
each metal is given below.
A. Zinc: Filtration of the BPT effluent is estimated to
reduce the zinc concentration by 6 percent to approximately
0.52 mg/1. This value is used as the concentration basis
for the ma'ximum 30-day average effluent limitation. Application
344
-------
12-26. OF
Hydrofluoric Acid
Level of Treatment: 3
Waste Water Plow: 33.4 m3/kkg of HP (65% Reuse)
Pollutant Treatability WR
(rag/1)
Nonconventional
Fluoride, F
Toxic
Pollutants:
Antimony
Arsenic
Chromium
Copper
Lead
Nickel
Selenium
Zinc
Pollutants:
33 2.1
0.70 2.0
0.50 2.0
0.010 2.0
0. §50 2.0
0.060 2.0
0.10 2.0
0.18 2.0
0.20 2.0
Achievable Concentration
Max
30-day 24-hr
Avg Max
33 66
0.70 1.4
0.50 0.10
0.010 0.020
0.050 0.10
0.060 0.12
0.10 0.20
0.18 0.36
0.20 0.20
(1) - WR: ratio of the 24-hour variability factor to the 30-day
variability factor.
345
-------
ISBLE 12-27. PERFORMANCE OF ALTERNATIVE TECHNOLOGY
Hydrofluoric Acid
Level of Treatment: 4
Waste Water Plow: 9.5 m3/kkg of HF (90% Reuse)
Pollutant Treatability WR(15
(ng/i)
Nbnconventional
Fluoride, F
TOXIC
Pollutants;
Antimony
Arsenic
Chrcniium.
Copper
I^ead
Nickel
Selenium
Zinc
Pollutants:
33 2.1
0.70 2.0
0.50 2.0
0.040 2.0
0.29 2.0
0.060 2.0
0.15 2.0
0.18 2.0
0.52 2.0
Achievable Concentration
(irer/1)
Max
30-day
Avg
33
0.70
0.50
0.040
0.29
„ 0.060
0.15
0.18
0.52
24-hr
Max
66
1.4
1.0
0.080
0.58
0.12
0.30
0.36
1.0
(1) - WE: ratio of the 24-hour variability factor to the 30-day
variability factor.
346
-------
of the BAT model plant discharge rate results in a corresponding
loading limitation of 0.017 kg/kkg. That is,
(0.52 mg/1) (33.4 m3/kkg) / kg/m3 \ = 0.017 kg/kkg
VLOOO mg/y
and, for the daily maximum limitation using the VFR value of
2.0, one obtains,
(2.0) (0.017 kg/kkg) = 0.034 kg/kkg.
This represents an overall reduction of 43 percent from the
BPT loading limitation. The VFR value of 2.0 used for BPT was
also used for BAT because the variability of the filtrate
quality is expected to be no greater than the observed
variability of the unfiltered effluent at Plant 1251 (Tables
A-lOa and A-lOc). Treatability studies .are being conducted
by the EPA and the results on filter performance will be
available prior to promulgation. The proposed BAT limitations
on the toxic metals are included in Table 12-25.
B. Nickel: The addition of filtration to the BPT
effluent is estimated to achieve a 14 percent reduction in.
the nickel concentration. The concentration basis for the
proposed BAT limitation is therefore set at 0.15 mg/1 and
results in a maximum 30-day average loading limitation of
0.0050 kg/kkg. That is,
(0.15 mg/1)(33.4 m3/kkg) / kg/m3 \ = 0.0050 kg/kkg
\1000 mg/1/
and the daily maximum is, .
(2.0) (0.0050 kg/kkg) = 0.010 kg/kkg.
This represents an overall 46 percent decrease from the
corresponding BPT level. A VFR value of 2.0 was used following
the same rationale as applied to zinc.
C. Lead: With the addition of filtration, providing an
estimated 80 percent removal of the residual lead, the
concentration basis for the proposed BAT limitation is set at
0.060 mg/1. This is in close agreement with the literature
treatability data summarized in Table 8-11. O'n this basis,
the maximum 30-day average effluent limitation .for lead is
0.0020 kg/kkg. That is,
(0.060 mg/1) (33.4 m3/kkg) / kg/m3 N = 0.0020 kg/kkg
VLOOO mg/1/
347
-------
and, the proposed daily maximum is,
(2.0) (0.0020 kg/kkg) = 0.0040 kg/kkg.
This represents nearly an 88 percent reduction from the
BPT level. A VPR of 2.0 is used following the same rationale
as applied to zinc.
D. Antimony: The addition of filtration is expected to
decrease the antimony concentration by approximately 7 percent.
Thus, the maximum ' 30-day average concentration basis is
estimated as 0.70 mg/1. This establishes the corresponding
loading limitation at 0.023 kg/kkg which is about 48 percent
lower than the BPT limitation. The proposed maximum 30-day
average limitation is,
i
(0.70 mg/1) (33*.4 m3/kkg) /' kg/m3 "\ = 0.023 kg/kkg
VLOOO mg/y
and the daily maximum is,
(2.0)(0.023 kg/kkg) = 0.046 kg/kkg.
The VFE is estimated to be 2.0 for the reason that this
value was used for the BPT limitation on antimony.
E. Copper: Filtration of the BPT effluent is estimated to
achieve approximately a 42 percent reduction in the average
copper concentration. Thus, a value of 0.29 mg/1 is used as the
concentration basis for the proposed 30-day average effluent
limitation of 0.0097 kg/kkg. That is,
(0.29 mg/1) (33.4 m3/kkg) f kg/m3 \ = 0.0097 kg/kkg
\1000 mg/1/
and, the daily maximum limitation is,
(2.0)(0.0097 kg/kkg) = 0.019 kg/kkg.
This represents an overall reduction of 64 percent below
the BPT level. A VFR value of 2.0 was used for the BAT
limitations for the same reason described for zinc.
F. Chromium: For chromium, an average additional removal
of approximately 60 percent is expected with the use of
filtration. For this reason, the concentration basis for the
proposed 30-day average BAT limitation is set at 0.040 mg/1.
This results in a corresponding loading limitation of 0.0013
kg/kkg. That is,
(0.04!0 mg/1) (33.4 m3/kkg) / kg/m3 \ = 0.0013 kg/kkg
\1000 mg/1/
348
-------
and, the daily maximum limitation is,
(2.0){0.0013 kg/kkg) = 0.0026 kg/kkg.
This represents an overall reduction of approximately 76
percent from the BPT level. A VFR value of 2.0 was also
used for chromium for the same reasons presented for zinc.
G. Other metals: The concentration basis for arsenic
and selenium are also given assuming 0 and 12 percent removals,
respectively, by the addition of filtration to the BPT system.
The values presented in Table 12-25 for these toxic pollutants
are intended to be used as guidance in cases where they are
found to be of serious concern.
12.7.5 Basis for Proposed Hew Source Performance Standards
Application of Advanced Level Treatment
Examination of raw waste loads indicates that the prime
source of pollutants at HF plants is the kiln waste.
Currently, two plants handle their kiln waste as a solid
greatly reducing the total raw waste load and subsequent
effluent. Based on this and an examination of control and
treatment alternatives available to this industry, it has
been determined that new HF facilities should achieve the
effluent quality attainable by NSPS, Level 5, technology.
The control parameters for NSPS are pH, TSS, fluoride,
nickel, and chromium. The recommended treatment technology
for new sources as described is dry handling of kiln wastes and
chemical treatment, filtration and reuse of other treated
wastes. The use of soda ash for precipitation of fluorides
will allow approximately 60 percent reuse for air pollution
control scrubbers, the second major source of waste water.
Raw waste toxic pollutant metal loadings from sources
other than kiln wastes were minimal and only occasionally
observed at potentially significant levels. It is assumed
that following chemical precipitation for fluoride removal,
the effluent loads discharged will be insignificant with regard
to these metals.
Technology Basis
For new plants in the hydrofluoric acid industry, the
specified waste treatment technology is the reduction of waste
flow and pollutant loadings by the dry handling of kiln wastes
and the treatment of other wastes by alkaline precipitation
followed by settling of solids and filtration of the
effluent. The technology also incorporates the reuse of 60
percent or more of the treated effluent for the air pollution
control scrubbers. Two plants now practice dry handling of
349
-------
kiln residues. This technology greatly reduces the waterborne
raw waste loads and is available to new plants because they
have the opportunity to design the most efficient systems
without retrofitting. The dry solids generated would have to
be handled and disposed of in a manner consistent with any
applicable requirements of the Resource Conservation and
Recovery Act (RCRA), 42 USC 6901 et. seq.
Pollutants limited by the proposed NSPS regulations are
pEr TSSr fluoride, chromium, nickel, and zinc. Metals from
scrubber water and other plant waste streams facilitates
the reuse of 60 percent or more of the treated effluent
for scrubber operation. Effluent reuse for this purpose is
presently practiced in the hydrofluoric acid industry. Plant
f722 reuses 92 to 100 percent of its soda ash treated waste
water for both scrubber operation and kiln residue transport.
Plant 1664 reuses approximately 94 percent of similarly
treated waste water for the same purposes. Information on
these plants is summarized in Tables 12-5 and 12-20,
Plow Basis
The basis for the model plant total treatment system
influent is the flow data on scrubber and other waste water
sources (excluding gypsum slurry water) for five of the nine
plants presented in Table 12-5. Plants f426, 1722, and
1837 were excluded because of incomplete data for scrubber
effluent and Plant f!67 because of an unusually high flow
rate for the scrubbers. The average raw waste flow rate
for the five remaining plants is 14.9 m3/kkg and with a
reuse rate of 60 percent the net effluent is 6.0 m3/kkg.
Selection of Pollutants to be Regulated
For NSPS, the two major waste water4 sources of concern
are the air pollution control scrubbers and "other" process
wastes. The latter includes the indirect contact wastes from
surface drains but excludes storm water runoff. The pollutant
parameters of concern are pH, TSS, fluoride, and the
toxic metals. Screening and verification sampling data on
the raw scrubber and other sources confirm the need for
limitations on pH, TSS, and fluoride. The four sets of
sampling data from these sources indicate that the relative
importance of the toxic metals is considerably less than was
found for the BPT and BAT sources which included the kiln
residue slurries. The observed maximum and average raw waste
concentrations of the toxic metals are shown in Table 12-28.
In the NSPS raw waste sources, nickel, zinc, chromium, and
selenium were the only toxic metals which showed maximum
concentrations that would be treatable by alkaline precipitation
and filtration. Of these, selenium is marginal even at its
maximum concentration and has an average concentration
350
-------
TABLE 12-28. TOXIC POLLUTANT RAW WASTE DATA USED TO
NEW SOURCES*
SUBCATEGORY: HYDROFLUORIC
Pollutant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Thallium
Zinc
ACID
Concentration
(I)
MaKimum
(tng/1)
0.030
0.014
0.021
0.41
0.12
0.029
0. 0020
0.81
0,24
0.0040
0.45
(2)
Average
(rag/1)
0.014
0.0090
0.0080
0.11
0.049
0.011
0.0010
0.18
0.068
0.0020
0.15
* Based on four sets of screening and verification sampling
data from Plants |705, #251, and #167 taking only the
scrubber and "other" waste sources.
(1) Maximum value observed from screening and verification
sampling data.
(2) Average value derived from screening and verification
sampling data.
351
-------
that is far below the minimum level treatable. For this
reason, only nickel, chromium, and zinc have been selected as
toxic pollutant control parameters for NSPS regulations.
Basis of Pollutant Limitations
Conventional Parameters -
A. pH: For NSPS, the BPT limitation is retained.
Control of the final effluent within the range of pH 6.0 to
9.0 is required on the basis of data presented in Appendix B of
this report and the JIB Study (52).
B. TSS: Although there is no applicable performance
data available on the filtration of treated and clarified NSPS
waste water, taking the proposed BPT maximum 30-day
average concentration of 97 mg/1 of TSS and assuming 30
percent additional removal by filtration, an estimated
performance level of 68 mg/1 is obtained. Pilot scale
studies (41) have demonstrated an average removal by
filtration of approximately 30 percent from waste water
containing suspended metal hydroxides after lime treatment.
A WR of 2.1 is used on the basis of long term data summarized
in Table 12-21 and described in the BPT section. Thus, the
proposed maximum 30-day limitation on TSS is,
(68 mg/1)(6.0 m3/kkg) / kg/m3 \ » 0.41 kg/kkg
V1000 mg/1/
and, using the VFR value of 2.1,
(2.1)(0.41 kg/kkg) = 0.86 kg/kkg is the proposed daily
maximum.
The proposed NSPS limitations are presented in Table 12-29.
Nonconventional pollutants - The only nonconventional
pollutant of concern is fluoride. The concentration basis for
the proposed maximum NSPS 30-day average limitation is set
equal to the 30 mg/1 BAT model plant performance level (Table
12-25), because'the treatment technology is the same. A VFR of
2.1 is used on the same basis given for the use of this ratio
in the BPT and BAT limitations. Thus, the proposed 30-day
average is,
(30 mg/1)(6.0 m3/kkg) / kg/m3 \ = 0.18 kg/kkg
V1000 mg/1/
and, using the VFR value of 2.1, the daily maximum is,
(2.1)(0.18 kg/kkg) = 0.38 kg/kkg.
352
-------
TABLE 12-29. PROPOSED LIMITATIONS
Hydrofluoric Acid
New Source Performance Standards
Waste Water Flow: 6.0 m3Akg (60% Reuse)
Pollutant Treatability
(mg/1)
Conventional and
.Nonconventional Pollutants:
Total Suspended 68
.Solids, TSS
(2)
Fluoride, F 30
Toxic
Pollutants :
Antimony 0.70
Arsenic 0 . 5
(2)
Chromium 0.040
Copper 0.29
Lead 0.060
(2)
Nickel 0.15
Selenium 0.18
(2)
Zinc 0.52
Concentration Basis Effluent Limit
(1) (rog/1) (kg/kkg of HP)
*je»r* —-.——,—,— ——«__.. —«.««,„— ,.__™™__
30-day 24-hr 30-day 24-hr
Avg Max Avg Max
2.1 68 143 0.41 0.86
2.1 3H 63 0.18 0.38
(3) (3)
2.0 0.70 1.4 — —
(3) (3)
2.0 0.5 1.0 — —
2.0 0.040 0.080 0.00024 0.00048
(3) (3)
2.0 0.29 0.58 — -—
(3) (3)
2.0 0.060 0.23 -- —
2.0 0.15 0.30 0.00090 0.0018
(3) (3)
2.0 0.18 0.36 — —
2.0 0.52 1.0 0.0031 0.0062
(1) - WR: ratio of ±he 24 hour variability factor to the
30 day variability factor.
(2) - Also applicable for PSNS limitations.
(3) - No effluent limitations proposed.
353
-------
Toxic pollu tants -
A. Nickel: For the proposed NSPS limitation on nickel,
the BAT concentration basis and VFR value are used because
the treatment technology remains essentially the same. The
proposed maximum 30-day average limitation is,
(0.15 mg/1)(6.0 m3/kkg) / kg/m3 \ = 0.00090 kg/kkg
V1000 mg/lj
and, with a VFR of 2,0, the daily maximum is,
(2.0)(0.00090 kg/kkg) = 0.0018 kg/kkg
The toxic pollutant limitations for NSPS are presented in
Table 12-29.
B. Chromium: Similarly for chromium, the BAT
concentration basis and VFR value are again used for the
proposed NSPS limitations. Thus, for chromium, the proposed
maximum 30-day average limitation is,
(0.040 mg/1)(6.0 m3/kkg) / kg/m3 \ = 0.00024 kg/kkg
\1000 mg/1/
and the daily maximum is,
(2.0)(0.00024 kg/kkg) = 0.00048 kg/kkg
C. Zinc: In the case of zinc, the concentration basis
for the proposed maximum 30-day average is the same as BAT
(0.52 mg/1). Thus, the maximum 30-day average limitation is,
(0.52 mg/1)(6.0 m3/kkg) / kg/m3 \ = 0.0031 kg/kkg
\1000 mg/1/
and, the daily maximum is,
(2.0)(0.0031 kg/kkg) = 0.0062 kg/kkg
D. Other metals: The concentration bases for antimony,
arsenic, copper, lead, and selenium are also provided in Table
12-29 to be used as guidance in cases where one or more of these
toxic metals may be of more serious concern.
12.7.6 Basis for Proposed Pretreatment Standards
Existing Sources
For Pretreatment Standards for Existing Sources (PSES), the
Agency is proposing limitations based on BAT. The pollutants
to be limited are fluoride, antimony, chromium, copper, lead,
354
-------
nickel, and zinc as indicated in Table 12-22. However, at this
time, there are no indirect dischargers in the HF industry.
New Sources
For Pretreatment Standards for New Sources (PSNS), the
Agency is proposing limitations based on NSPS. The pollutants
to be regulated are fluoride, nickel, chromium, and zinc
as indicated in Table 12-29.
355
-------
-------
SECTION 13
HYDROGEN PEROXIDE INDUSTRY
13.1 SUMMARY OP DETERMINATIONS
It has been determined that no further effort be given to
developing or revising BAT, NSPS, or pretreatment regulations
for the Hydrogen Peroxide Subcategory using either the
electrolytic process or the organic process.
The bases for this recommendation are: 1) only one plant
exists that manufactures hydrogen peroxide using the
electrolytic process and 2) no toxic pollutants were found in
the wastes using the organic process. Therefore this
subcategory is excluded under Paragraph 8 of the Consent Decree.
13.2 ASSESSMENT OF THE WATER POLLUTION POTENTIAL
13.2.1 Production Processes and Effluents
In the electrolytic process, ammonium (or other) bisulfate
solution is electrolyzed, yielding ammonium persulfate at the
anode and hydrogen gas at the cathode. The presulfate is then
reacted with water to yield hydrogen peroxide and original
bisulfate. Hydrogen peroxide is separated from bisulfate by
fractionation, after which it is concentrated and filtered. The
only waste is a stream of condensate from the fractionation
condenser.
The organic process involves the reduction of
alkylanthraquinone by hydrogen over a supported metal catalyst
to produce the corresponding alkylhydroanthraquinone. The
reacted mixture is oxidized to form hydrogen peroxide and
original alkylanthraquinone. The peroxide is extracted with
water and the organic material in the solvent is recycled to the
process. Since hydrogen peroxide manufactured by the organic
process consists of a series of exothermic chemical reactions,
the bulk of the water usage is for process cooling (contact and
noncontact). Noncontact cooling accounts for over 90 percent of
the total water usage in this subcategory. The waste water
sources include contact cooling (barometric-condenser) water,
357
-------
purification washing of the organic working solutions,
regeneration waste from the deionizers, and leaks and spills.
13.2.2 Plants
Only one plant exists in the United States that
manufactures hydrogen peroxide using the electrolytic process.
The hydrogen peroxide subcategory profile data received in
response to 308 letters is given in Table 13-1.
Three plants produce hydrogen peroxide by the organic
process.
13.2.3 . Toxic Pollutants
Data has been received on 100 percent of the industry as a
result of section 308 letters. A"" sampling survey for toxic
pollutants was made for three plants. At one plant,
pentachlorophenol was found in significant concentrations.
However, it was determined that is presence was due to its use
as a weed killer at the plant site and this use was
discontinued. Two more plants were sampled in the verification
phase, and the survey indicated that no toxic pollutants were
being discharged in significant quantities.
Toxic pollutants found during sampling were as follows:
Maximum Concentration
Pollutant Observed (pg/1)
Zinc 256
Pentachlorophenol 4850
Bis(2-ethylhexyl)phthalate 20
Chloroform 11
Naphthalene 11
13.3 STATUS OF REGULATIONS
Since no toxic pollutants were found in significant
concentrations, the subcategory is excluded under Paragraph 8.
358
-------
13-1
SIBCKCEQORY PROFILE DATA
SUBCATEQOKf
ffiTDBQGEN PEROXIDE
Total subcategory capacity rate
Total subcategory production rate
Number of plants in this subcategory
308 Data on file for
With total capacity of
With total production of
Representijig capacity
Representing production
Plant production range:
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range;
Minimum
Maximum
Waste water flow'range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
85,700 kkg/year
4
4
102,200 kkg/year
57,000 kkg/year
66 percent
5,560 kkg/year
28,730 kkg/year
NA
NA
NA
15 years
27 years
NA
NA
NA
NA
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Commerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.? Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chemical Industry, " June, 1978 and "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards for the Inorganic Chemicals Industry,"
March, 1980.
359
NA = Not Available.
-------
-------
SECTION 14
TITANIUM DIOXIDE INDUSTRY
(RUTILE/UPGRADED ILMENITE — CHLORIDE PROCESS)
14.1 INDUSTRY PROFILE
14.1.1 General Description
Titanium dioxide is manufactured by a chloride process, a
sulfate process, and a chloride-ilmenite process. This
subcategory is subdivided into three segments, one for each
processs because of the difference in raw materials used, waste
water flows, and raw waste characteristics. Ti02 is a high
volume chemical, ranking within the first fifty of all U.S.
chemicals production. Over fifty percent of the titanium
dioxide produced is used in paints, varnishes and lacquers.
About one third is used in the paper and plastics industries.
Other uses are found in ceramics, ink and rubber manufacturing.
The industrial profile data for the chloride segment are
presented in Table 14-1, while the status of regulations is
given in Table 14-2.
14.1.2 General Process Description and Raw Materials
In the chloride process, the raw materials used are rutile
or upgraded ilmenite ore, which are relatively pure materials
with a high titanium and a low iron content. For upgrading
ilmenite (FeTiOS), a beneficiation process removes a part or all
of the iron. Several patented processes exist for the
beneficiation step and two or three are in current operation on
a commercial scale. The wastes from the chloride process
using beneficiation of ilmenite in titanium dioxide production
are different from those produced using high grade titanium ore
(rutile or upgraded ilmenite). The Titanium Dioxide Subcategory
has been classified further into three separate categories:
sulfate process using ilmenite ore, chloride process using
rutile or upgraded titanium ore, and chloride process using
ilmenite ore. This section is restricted to the chloride
process using rutile ore.
361
-------
IftBEE 14-1
SDBCATBQOEY ERCFIIE DKCA SUMMARY
DIOXIDE )
Total subcategory capacity rate
Stotal subcategory production rate
Number of plants in this subcategory
308 Data on file for
With total capacity of
With total production of
Representing capacity
Representing production
Plant production ranges
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flowirange:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
610,000 kkg/year
389,000 kkg/year
5
5
184,600 kkg/year
142,000 kkg/year
30 percent
37 percent
16,900 kkg/year
45,200 kkg/year
28,400 kkg/year
25,600 kkg/year
77 percent
6 years
15 years
1,140 cubic meters/day
4,770 cubic meters/day
29.3 cubic meters/kkg
110 cubic meters/kkg
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Commerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary iksonomic Assessment of Effluent Limitations in the
Inorganic Chemical Industry," June, 1978, and "Economic Analysis of
Proposed Revised Effluent Guidelines and Standards for the Inorganic Chemicals
Industry," March, 1980.
362
-------
TKBXE 14-2 - STftTOS OP HEGOIATICNS - EEET£ENT IJMEEKEECN GUIDELINES
SCffiCXCEGQRX Titanium Dioxide
SUBPAKT -V (40 CFR 415. 220, 3/12/74)
STANDARDS
Product Para-
Process meters
Chloride ^g
Process
Iron
SUlfate _gg
Process
YWM.
BPCTCA* BKCEA*
1 2
Max. Avgr. Max. Avg.
kg/kkg hg/kkg ka/Mcf kg/kkg
(nig/1) ftngA) fcig/1) ftng/1)
4.6
0.72
21.0 ^
(100.0)
1.7
2.3 2.6 1.3
0.36 0.36 0.18
10.5 10.6 5.3
(50.0)
0.84 0.84 0.42
NSPS*
Max. Avg.
kg/kkg kg/kkg
(mg/1) (rag/1)
2.6 1.3
0.36 0.18
10.6 5.3
0.84 0.42
(8.1) (4.0)
Sections 415.220, 415.222, 415.223, and 415.225 were remanded and are
presently reserved (41 FR 516CEL, Ifeventoer 23, 19761.
Ttox, at Maximum of any one day.
Avg. - Maximum average of daily values for thictv mr>s<=«utive dairs.
**£low basis 210,000 1/kkg.
363
-------
In the chloride process, the ore and coke are dried and then
reacted with chlorine to form titanium tetrachloride. The
chemical reaction taking place in the reactor is given as:
3C + 2TiO2 + 4C12 = 2TiC14 + C02 -I- 2CO (1)
The reaction takes place at a temperature of 800-1000
degrees C and a fluidized bed reactor is generally used. The
product gases leaving the reactor consist of titanium
tetrachloride, unreacted chlorine, carbon dioxide, carbon
monoxide and minor amounts of heavy metal chlorides. The gases
are initially cooled to 250 degrees C to remove the impurities,
although in some cases purification is accomplished by washing
the gases with liquefied titanium tetrachloride. Iron chloride
and small amounts of vanadium, zirconium, and other trace metal
chlorides are removed by centrifugation and the liquid recycled
to the absorber. Titanium tetrachloride is liquefied from the
gases after the first stage of cooling by further cooling to
ambient temperature. Copper, hydrogen sulfide and, in some
cases, proprietary organic complexing agents are added for
purification to the condensed solution. Copper acts as a
catalyst to decompose the phosgene formed in the TiC14 stream.
Organic complexing agents aid in separation of the T1C14 from
other chlorides such as cupric chloride and silicon
tetrachloride.
The residual uncondensed gases generally consist of
hydrochloric acid, chlorine, carbon monoxide, carbon dioxide,
nitrogen, and some titanium tetrachloride. They are treated to
remove acidic materials before being vented to the atmosphere.
The liquefied titanium tetrachloride contains impurities
such as aluminum chloride, silicon tetrachloride, etc., which
are removed by distillation. The distillate is the purified
titanium tetrachloride and the impurities remain as a residual
which becomes waste. The tail gases from the distillation
column are scrubbed to remove acidic materials. The titanium
tetrachloride product is then reacted with oxygen, as air,
forming titanium dioxide and chlorine:
TiC14 + 02 = Ti02 + 2C12 (2)
The rate of reaction is negligible below 600 degrees C but
increases rapidly above this temperature, and is generally
maintained between 1200-1400 degrees C for efficient reaction
and conversion. The needed heat is supplied by passing the
reactants through heat exchangers, by electric dischargers, or
by use of fluidized beds. After the oxidation reaction, the
titanium dioxide forms a solid and is separated from the gases
either in cyclones, baghouse filters, or Cottrell precipitators.
The residual chlorine is refrigerated and liquefied. The tail
364
-------
gases are scrubbed with caustic soda to remove chlorine before
being vented to the atmosphere. When air is used for oxidation,
chlorine recovery is achieved by absorption in trichlorethylene,
followed by distillation to remove chlorine. The titanium
dioxide is then sent to the finishing operation where it is
vacuum degassed and then treated with alkali, using a minimum
amount of water to remove traces of absorbed chlorine and
hydrochloric acid. The pigment is then milled, surface treated
for end-use application, dried, and packaged for sale. A
generalized process flow diagram, including the waste streams,
is shown in Figure 14-1.
14.2 WATER USE AND WASTE SOURCE CHARACTERISTICS
14.2.1 Water Use
Water Use
Water is used in noncontact cooling, for scrubbing the tail
gases from the purification and oxidation reactor to remove
contaminants, and in the finishing operation of the product.
The total amount of water usage varies from 45.3 to 555 m3/kkg
of Ti02 produced, as shown in Table 14-3. The table also shows
that cooling water constitutes the major use of water and varies
from 10.7 to 426 m3/kkg of Ti02 produced*
14,2.2 Waste Sources
Wastes from Cooling Chlorinator Gas
The waste consists of solid particles of unreacted ore,
coke, iron, and small amounts of vanadium, zirconium, chromium,
and other heavy metal chlorides. They are either dissolved in
water and sent to the waste water treatment facility or disposed
of in landfills as a solid waste.
Chlorinator Process Tail Gas Scrubber Waste
The uncondensed gases, after the liquefaction of titanium
tetrachloride, are initially wet scrubbed to remove hydrogen
chloride, chlorine, phosgene, and titanium tetrachloride and
chlorine. In a second stage, they are scrubbed with caustic
soda to remove chlorine as hypochlorite.
Distillation Bottom Wastes
These contain copper, sulfide, and organic complexing
agents added during purification in addition to aluminum,
silicon, and zirconium chlorides. These are removed as
waterborne wastes and reaction with water converts silicon and
anhydrous aluminum chlorides to their respective oxides.
365
-------
OR
awns one
ATHOSPHEHS
HSSBS
HATER
Figure 14-;!. • General process diagram for production of titanium dioxide
(chloride process) from high grade ores.
366
-------
14-3. USAGE IN DIOXIDEHSHLQRIDE PROCESS/HIGH
ORES SUBC&TEGORY
Water usage at plants
Water Use (m3/kkg of TiO_)
«£
Water Use
Plant #102
Noneoataat cooling
Direct process contact
Indirect process contact
Maintenance, equipment
cleaning and wark area
wasbdown
Mr pollution control
Istoncontaofc ancillary uses
Sanitary & potable water
fbtal
182
10.5
NA
6.65
0.25
11.60
0.23
211.23
Plant 1172
10.66
15.53
0.72
0.52
7.14
10.4
0.31
45.28
Plant #199
426
73.2
26.5
2.80
11.3
9,5
5.6
554.9
N& = Not available
367
-------
Oxidation Tail Gas Scrubber Wastes
The gases from the oxidation Unit are cooled by
refrigeration to liquefy and recover chlorine. The uncondensed
off-gases are scrubbed with water or caustic soda to remove
residual chlorine. When caustic soda is used as the scrubbing
solution, the resulting solution of sodium hypochlorite is
either sold, decomposed, sent ,to the waste water treatment
facility, or discharged without treatment. The scrubber waste
stream also contains titanium dioxide particulates.
Finishing Operations Waste
The liquid wastes from the finishing operation contains
titanium dioxide as a suspended solid and dissolved sodium
chloride formed by the neutralization of residual HC1 with
caustic soda.
The range of waste water flows requiring treatment is
summarized in Table 14-4. The wide range of flow occurs because
some plants use additional water to wash solid process residues
to the waste treatment system.
14.3 DESCRIPTION OF PLANTS VISITED AND SAMPLED
14.3.1 Screening
Plant =P559 was visited and the waste effluents sampled in
the screening phase of the program. Plant #559 makes titanium
dioxide using both the sulfate and the chloride processes. The
waste waters from both processes are mixed and undergo combined
treatment.
The solids from the chloride process, called pit solids,
(mainly unreacted ore, coke, iron, and trace metal chlorides
including TiC14) are separated from the first stage cooling of
the chlorinated gases and are slurried with water. The slurried
pit solids and the distillation column bottom residue effluents
from the chloride process are sent to a large settling pond
(called the weak acid pond) where they are mixed with the weak
acid from the sulfate process. The overflow from the settling
pond is neutralized with ground calcium carbonate in a reactor.
The reactor effluent is filtered, aerated to remove iron and
combined with neutralized strong-acid waste effluent (from the
sulfate process). The combined scrubber and contact cooling
waste water from both sulfate and chloride processes is also
combined at this point. The combined waste water is neutralized
and solids settled out in a pond prior to final discharge, A
flow diagram of the treatment facility including the sampling
locations is shown in Figure 14-2.
368
-------
ISfflEE 14-4. mSEE FLOW K)R DIC3XID1-CHIJORIDE
SUBC&TEGORY
SUBCATEQORY TITANIUM DIOXIDE CChlQrjx3e Process)
Plant Unit Ifeste Water Plow Going to Treatment Plant
. (m3/Mcq of TiD2)
102 29.3(1)
172 34.7(1)
559 91.0(2)
199 110.0(2)
(1) Offsite disposal of process solid residues.
(2) Process solid residues are slurried to «aste treatment. The average
flow of Plants #559 and #199 »s used as the model plant flow for
cost estimating and regulation development.
369
-------
KlVtll
SWPU MATE*'
MUNICIPAL
SUPPLY WATER
uai
SUPPLY WATER
e.
vo
U)
<1
o
OTHER PRODUCT
(WEAK ACID) '
WASTE VATEft
SAI1ITAKY
AMD TIO- FIHISHIHG
AMA HASTE VATEP.
OTHEK PRODUCT
(STRONG ACID)
VAST! WATER
SETTLING
URRIEO —p.
T SOLIDS
PCN
|9
i
«
^.ri-i Dl
.„„„., ft:
\> £-*.
I
STIILATIOM BOTTOM
WASTE WATER
?°
FINAL EFHUiMT
e
LEGEND
SAMPLING POINTS
CHLORIDE PROCESS SCRUBSEH
WASTE WATER
OTHER PRODUCT
WASTE HATiR
Fimire 14-2
*
General &cw diagram at Plant #559 showing the sanpling points.
(Titanium dioxide — chloride process manufacture.)
-------
Problems were encountered during the sampling of the pit
solids and the distillation bottoms. The pipes carrying the
wastes from the process discharged at the bottom of the settling
pond and it was not possible to take the samples right at the
outlet of the pipe. The combined sample of the two streams was
taken at the surface of the discharge. It is probable that some
solids settled before the stream reached the surface. Table 14-
5 gives the waste flows and pollutant loadings for the streams
sampled at Plant 1559. Because of the intermixing of the waste
effluents from both chloride and sulfate processes, the
pollutant loadings in Table 14-5 were calculated by
proportioning according to the relative hydraulic loadings.
14.3.2 Verification
Plant f!72 was sampled in the verif-ication phase. Titanium
dioxide is made at this facility by the chloride process only.
The strong acid wastes and the spent coke and ore residues are
hauled to a secure chemical landfill for disposal. The waste
water from the process, mainly the scrubber water, is collected
in trenches and sent to a central reactor basin. Other
discharges, including a part of the total rain runoff, are also
collected in ditches and sent to the reactor basin. In the
reactor basin, sodium hydroxide is used for neutralization, and
the resulting effluent is mixed with the remaining rain water
runoff and sent to the first of two retention basins arranged in
series. The overflow from the second retention is pH adjusted
with sulfuric acid before discharge. A simplified diagram of
the treatment system, including the sampling points, is shown in
Figure 14-3. Table 14-6 gives the waste flow and pollutant
loadings for the streams sampled.
14.3,3 Toxic Pollutant Concentrations
Five toxic pollutants were found above the treatability
levels in the raw waste of plant $559. It is possible that some
of the pollutants might be from the sulfate process waste water
as the two raw waste effluents are intermixed before treatment.
One pollutant was found above the treatability level in the raw
waste of Plant fl72. No organic toxic pollutants were found
above treatment levels in the raw wastes of either plant. The
maximum concentration of the toxic pollutants found in the raw
waste in significant concentration in the screening and
verification program weres
371
-------
OBOE 14-5. FICM SND KEIW»W CSXCSOimnm D?3& OP THE SfiMELED
STEAMS OF PWOTfl72 BKTOCING IHMmM XttCBOBE W CHMSCDDe-KOTIIE HCCESS
STREAM |2 STREAM f 5
(li
CAICULOTED ESSJKKEB SERESM f6
Bit a>lidsand Scrufcter and • IWal Raw WSste Treated Effluent
Distillation Bottoms Contact Cooling Witet *«»«« ^.ij.ucin.
A B" C D 1
RAlutant (ftxBsKf3)
Ohit Flow OXKJ. , Unit Load (Unit Flow) Gone.
(nfl/kkg) (tng/I) (kg/Hcg) {m3/Kkg) (mg/1)
10.9 80.1
TSS 6903 75.2 314
Iron 1348 14.7 143
w chromium 112 1.2 0.11
to
Lead 3.53 0.04 0.009
Nickel 3.46 0.04 0.016
Zinc 2.12 0.02 0.13
F G H I J K
(DxExlO"3) (AH)) (CW) (KdO~3/G)
ttoit toad ttiit Flow Unit Load Cone. Unit Flow Oonc.
(kg/Beg) (ra3/lckg) (kg/Meg) tog/1) (nQ/kkg) 6ngA)
91 91
25.2 100.4 1103 23
11.5 26'2 mB 4.4
0.01 1.21 13.3 0.03
0,001 0.041 0.5 0.002
0.001 0.041 0.5 0.005
0.01 0.03 0.3 0.06
I.
Unit toad
(kg/kkg)
2.1
0.4
0.003
0.0002
0.0004
0.005
(1) See Figure 14-2 for location of sampling points
-------
PROCESS
WASTE WATER
HOLDING POND
FOR
RETREATMENT
#1
.RETENTION
BASIN
RETENTION
BASIN -
^,
#3
DISCHARGE
NaOH
MIXING
BASIN
NEUTRALIZE
RAIN RUNOFF
RAIN RUNOFF
pH ADJUSTMENT
LEGEND
ft SAMPLING POINTS
^
* THE TOTAL RETENTION TIME
OF WATER IN THE TWO- PONDS
IS 5 DAYS.
Figure
General flow diagram at Plant #172 showing the sampling points.
Titanium dioxide (chloride process) manufacture.
373
-------
TABLE 14-6. FLOW AND KHXUTAMF CONGEKmATE)N mm OP THE SftMELED
FOR §172 DIOXIDE (CHLORIDE
SAMPLED STREAM #1
SAMPLED STREAM #3
JEtollutant Saw Wfe.sfce Influent
Treated Effluent
A B C-D E F ,
(A+BxlO ) (D+ExlO )
Unit Plow Avg'. Cone. Unit Load Unit Plow Avg, Cone. Unit Load
(kg/kkg) (roVkkg) (rag/1) (kg/kkg)
34.7
TSS
171
34.7
5.93
6.7
0.23
Iron
Chromium
Lead
Nickel
2.9
0.72
0.005
0.08
0.10
0.03
0.0002
0.003
0.33
0.02
0.002
0.01
0.01
0.0007
0.00007
0.0003
Zinc
0.3
0.01
0.09
0,003
374
-------
Maximum Raw Waste Concentrations Observed (jig/1)
Pollutant
Plant 1559
Chromium
Lead
Nickel
Zinc
Screening
Plant f!72
152,000
5,150
6,320
3,300
Verification
1800
NS*
NS
NS
*NS - Concentration was found below the lower limit of
treatability; i.e., not significant.
The screening and verification sampling program and the
methodology used have been described in Section 5.1.2 of this
report. A total of six days of sampling was conducted at Plants
1559 and f!72. Five effluent waste streams were sampled at
Plant |559 and three streams were sampled at Plant #172. At
each sampling point, three 24-hour composite samples were
collected for analysis. The evaluation of toxic metal content
of these process related waste streams was based on 550
analytical data points. The average unit loadings and
concentrations for conventional, nonconventional, and toxic
pollutants found in the raw waste effluents for Plants f559 and
1172 are given in Table 14-7.
The total quantities of toxic pollutants generated each
year for this subcategory (calculated as total subcategory
production times average unit toxic pollutant load from Table
14-7) are as follows:
Pollutant Waste Load (kg/year)
Chromium 241,000
Lead 8,200
Nickel 8,500
Zinc 7,800
14.4 POLLUTION ABATEMENT OPTIONS
14.4.1 Toxic Pollutants of Concern
The dominant toxic pollutant in untreated effluents in the
Titanium Dioxide (chloride process) Subcategory is chromium.
Chromium was found in treatabl.e concentrations at both plants
sampled in the screening and verification phase. Lead, nickel,
and zinc were'found in the raw waste of Plant 1559 at treatable
levels, but- were not present in the Plant $172 raw waste. At
Plant 1559, the chloride process waste effluents are mixed with
375
-------
14-7. RAW msiE
mtm OF THE SAMPLED
SUBCMBGORY:
DIOXIDE (CHDORIDE PROCESS)
Average Daily
Pollutant
Toxic •
Iron
Cteomium
Lead
Nictel
Zinc
Conventional!
TSS
Pollutant Concentration and
(kg/kkg of Ti02)
(mg/1)
Plant
1559
26.2
(288)
1.21
(13.3)
0.041
(0.5)
0.041
(0.-5)
0.03
(0,3)
100.4
(1103)
Loadings at Plants
Plant
1172
0.10
(2.9)
0.03
(0.72)
0.0002
(0.005)
0.003
(0.08)
0.01
(0.3)
5.93
(171)
Sampled
Overall
Average
13.15
0.62
0.021
0.022
0.02
53.17
376
-------
the sulfate process waste effluents before treatment. It is
likely that the three major toxic pollutants found were
contributed by the sulfate process wastes, as it uses a low
purity ore (ilmenite), At Plant 1172, the solids generated from
the chloride process (which consist of solid particles of
unreacted ore, coke, iron, and other heavy metals) are hauled to
a landfill for disposal. It seems probable that the three
pollutants are present in this solid waste and hence do not
appear in the waste waters.
14.4.2 Process Modi£i cation and Technology Transfer Options
1. Research to develop economical techniques to recover
the vanadium and other metal values from the solid wastes
generated from the process waste treatment system would appear
to be a fruitful area of investment.
2. New plants can utilize refrigeration and high pressures
for chlorine liquefaction. This would reduce or eliminate the
chlorine residual problem in the tail gases. The capital cost
to modernize old plants is high, but these plants should have a
caustic soda or lime scrubber instead of a water scrubber to
remove residual chlorine from the tail gases. . Caustic or lime
scrubbing removes a significant portion of the chlorine from the
tail gases as seen from the analagous data for the chlorine
subcategory given in Section 11.
14.4.3 Best Management Practices
Provision should be made at all plants to collect storm
water runoff from the plant site and send it to the treatment
facility. Three out of a total of five existing plants are
presently treating storm water runoff.
14.4.4 Prevailing Control and Treatment Practices
At Plant §172, the solid wastes consisting of spent ore and
coke are hauled to an off-site landfill. Process waste waters
consisting of scrubber and contact-cooling effluents and a part
of the surface runoff are sent to a mixing basin where they are
neutralized with caustic soda. The effluent from the basin is
then sent to two retention ponds in series. Additional or
residual rain water runoff is added to the ponds for
clarification. The overflow from the last pond is monitored and
discharged to a surface stream. At Plant ^559, the waste waters
from both chloride and sulfate processes are mixed and treated
together. The distillation bottoms and the unreacted ore and
coke from the chloride process are combined with the weak acid
effluent from the sulfate process in a pond. The overflow from
the pond is neutralized with limestone and oxidized with air for
the removal of iron. The waste water is then mixed with the
377
-------
neutralized strong acid waste (from the sulfate process) and
scrubber waters (from both the chloride and sulfate processes)
and neutralized with lime in a reactor and sent to a final
settling pond. The overflow from the pond is the final
discharge.
At Plant ^199, all the process waste waters are combined,
including storm water and sanitary waste water. The combined
was.te water is sent to a four-stage neutralization system, and
the effluent from each of the four stages of neutralization is
sent to a thickener. The thickener overflow is transferred to
the first of three settling ponds, also in series. The
underflow from the thickener is heated to improve its filtration
characteristics and filtered in four rotary drum filters. The
thickened solids from the filters are disposed of in a landfill
and the filtrate is combined with wash water, and vacuum pump
seal water prior to being recycled to the fourth stage of the
neutralization train. The overflow from the last settling pond
is discharged.
The process waste water streams at Plant #102 are received
in two tanks, neutralized with lime, and then sent to a settling
basin. The settled solids are retained in the settling lagoons.
The plant has future plans for treating boiler blowdown, and
cooling tower blowdown, leaks and spills with the process waste
water.
At Plant 1605, the unreacted ore and coke is disposed of as
a solid waste in the pit. The waste water from the process is
passed to two tanks for flow equalization, and the water is then
reacted with ground limestone slurried in water. The treated
solution is centrifugally treated to remove coarse solids which
are separated and landfilled. A flocculating agent is added to
the centrate and the solution is sent to a clarifier. The
clarifier overflow is degassed and the pH adjusted with caustic
soda (if required) before discharge.
14.4.5 Advanced Treatment Technologies
Neutralization and settling are practiced for the treatment
of chloride process raw waste effluents at all the five plants
for which 308 data are available. Air oxidation,- sulfide
precipitation, xanthate precipitation, and ion exchange might be
applied to the clarified solutions for control of metals.
Sulfide precipitation or the xanthate process could be used to
provide additional removal of zinc, lead, and nickel*
378
-------
14.5 SELECTION OF APPROPRIATE TECHNOLOGY AND EQUIPMENT
14.5.1 Technologies for Different Treatment Levels
Level 1 (BPT)
The chloride process wastes are equalized, neutralized with
lime to a pH range of 6 to 9, and settled in lagoons before
discharge. Level 1 treatment is typical of industry practice
and for this reason was selected as the technology basis for BPT
regulations.
Level 2
Alkaline precipitation as a second-stage lime treatment to
an optimum pH (9 to 10) is added to Level 1 to precipitate
metallic hydroxides, which are then filtered before discharge.
Filtration removes traces of metallic hydroxides which do not
separate in a gravity system. Level 2 technology was
utilimately selected as the basis for the proposed BAT
limitations because it provides an economical method for the
removal of additional toxic metals.
Level 3
Ferrous sulfide treatment is added ahead of the Level 2
filter to precipitate the heavy metals more effectively.
Alkaline precipitation was chosen as Level 2 because it
readily supplements existing lime neutralization by the simple
addition of filtration and increasing the Level 1 lime dosage.
Sulfide precipitation was chosen at Level 3 because it provides
a polishing treatment for most residual heavy metals beyond
Level 2 treatment.
Figures 14-4, 14-5, and 14-6 show the model treatment
systems adopted for the chloride process wastes.
14.5.2 Equipment for Different Treatment Levels
Equipment Functions
BPT treatment is essentially lagooning with lime
neutralization, using no special equipment except a lime feeder
and mixer.
In Level 2, second stage lime treatment is followed by
gravity clarification and multi-media filtration, with
necessary pH controls.
379
-------
LIME
RAW
WASTE WATER
— \ " JLAGOON JL — .
LAGOON / —
*9
W
MDttNG
LAGOON
*o
•*»
EFFLUENT
LAGOON
>J
w
00
o
Includes flow monitoring, pH monitoring and sampler.
Figure 14-4. Level, 1 waste water treatment for titanium dioxide — chloride
process.
-------
BACKWASH
RAW
WASTE WATBi
U)
00
pH ADJUSTMENT
SUMP
TO LANDFILL
'ncludes flow monitoring, pll monitoring and sampler.
Figure 14-5. Level 2 waste water treatment for titanium dioxide — chloride
process.
-------
-A LAGOON ym»-
RAW
WASTE WATER
LAGOQN
oo
r
LIME ±
!V
LAGOON
MIXING
JLAGOON
FEilROUS SODIUM
SULFATE BISULFIDE
BACKWASH
4
t-
0
FILTER PRESS
SUMP
TO LANDFILL
Includes flow monitoring, pH monitoring and sampler
r
'h^l
SUMP DUAL
MEDIA
FILTER
CLARIFIER -,
* EFF.
Figure 14-6. Level 3 waste water treatment for titanium dioxide
process.
chloride
-------
In Level 3, ferrous sulfide is added ahead of the Level 2
filter, to react with residual heavy metals more completely than
in the alkaline precipitation step at Level 2.
Chemicals and Handling
Lime and hydrochloric acid are fed with conventional
equipment at all levels, and ferrous sulfide is prepared on-site
by mixing ferrous sulfate with sodium bisulfide. When normal
dust control and good ventilation are used, there should be no
adverse effects from handling these chemicals, although care
should be taken that hydrogen sulfide gas is not generated.
Separation and Removal of Solids
Inert ore fractions and precipitated solids are accumulated
in clay-lined lagoons, which are alternately drained. Solids
are mechanically removed to self-draining 18 ft. high storage
piles on land provided at the site for a 10-year operating
period. At Levels 2 and 3, small amounts of heavy metal
precipitates in the clarifier underflow are filter pressed and
hauled to a secure landfill.
14.6 TREATMENT COST ESTIMATES
14.6.1 General Discussion
To determine the treatment cost, a model plant concept was
developed. A raw waste unit flow was selected and pollutants to
be treated were selected, based on the treatment system data
available for the five TiO2 plants and the screening and
verification sampling program. Three production levels were
then selected to cover the entire subcategory range. Treatment
costs for Levels 1, 2, and 3 were calculated for each of the
model plant production ranges using the unit flow and unit
pollutant loads. The preliminary cost data given in the cost
tables and figures were generated using a low unit flow of 31
m3/kkg of Ti02 based on incomplete industry data. The new unit
flow of 100 m3/kkg used for the model plant in regulation
development has been selected to be more representative of the
subcategory and it is assumed that the unreacted ore and coke
are slurried and sent to the treatment system, instead of being
disposed of in a landfill as a solid waste. The need for
revising the preliminary cost estimates is being evaluated by
the Agency and any appropriate changes will be made prior to
promulgation. The model plant specifications presented here
were used in regulation development.
383
-------
Waste Water Flow
The unit waste effluent flow varies from 29.3 to 110.0
m3/kkg of Ti02 for the four plants as shown in Table 14-4. The
primary reason for the variation in the flow is that some plants
slurry the spent ore and coke (solid waste from chloride
process) and send it to the treatment system, and others haul
the dry solids to a landfill. The flow variation is also
dependent on the difference in the chlorine recovery process
from the tail gas and the amount of scrubbing liquid used.
Small variations in flow also result from the finishing
operation which is dependent on the type of titanium dioxide end
product desired. Plants #559 (unit flow of 91 m3/kkg) and #199
(unit flow of 110 m3/kkg) sent the solid waste from the
manufacturing process to the treatment facility. It is.assumed
for treatment system cost estimation that the solids are
included in the raw waste flow to the treatment system. A
constant unit flow of 100 m3/kkg of Ti02 has been used for the
model plants, which is an average of the unit flows of Plants
f559 and #199.
Pollutant Load
The primary pollutants occurring in the waste water are
suspended solids, acidity, and the chlorides of ferric iron,
chromium and other trace metals. The suspended solids (TSS)
loading values for Plants #559 and #172 are 100.4 and 5.93
kg/kkg of TiO2 (Table 14-7). The low value represents a plant
that hauls ore and coke off-site, while the high value is
believed to be due to nonrepresentative sampling. The amount of
solids produced are higher than the values indicated for, the
sampled plants. Consequently, a higher suspended solids loading
of 500 kg/kkg of TiO2 ("reported in the 308 data from Plant #199)
is assumed for the model plants. To establish treatment
chemical requirements and related costs, the toxic pollutant
loadings for the model plant are taken as the average values of
the unit pollutant loadings of the plants sampled in the
screening and verification program (Table 14-6) and the selected
pollutant values are:
Pollutant Unit Loading (kg/kkg of Ti02)
Chromium 0.62
Lead 0.021
Zinc 0.020
Iron 13.15
Nickel 0.022
Production Rates
Five plants produce titanium dioxide from rutile ore or
ilmenite ore, using the chloride process at a total production
384
-------
rate of 142,000 metric tons per year. Production ranges from a
minimum of 16,900 kkg/year to a maximum of 45,200 kkg/year with
a mean of 28,400 kkg/year and a median of 25,600 kkg/year. For
waste water treatment cost estimates, three production levels
were selected as model plants. These are 16,900 kkg/year, 25,500
kkg/year, and 45,200 kkg/year. This range of production
includes all United States plants.
The estimated costs for the three models having different
production levels are given in Tables 14-8, 14-9, and 14-10.
Annual treatment costs as a function of production are shown
graphically in Figure 14-7. Similarly, treatment costs per
metric ton of product are given in Figure 14-8. Table 14-11
presents a summary of the unit cost distribution between
amortization, and the operation and maintenance cost components
at various production rates and levels of treatment. The costs
shown at each level of treatment correspond to the model plant
BPT (Level 1} system and higher level (2 or 3) systems which may
add to or modify the existing BPT system to meet more stringent
toxic pollutant removal requirements. The higher levels (2 and
3) also furnish a better effluent quality with respect to the
conventional and nonconventional parameters. For model plants
at the base level of treatment, amortization, chemicals and the
residual waste disposal costs have a significant impact on the
total annual costs. At treatment levels 2 and 3, amortization,
chemicals and labor constitute a major portion of the additional
annual costs.
14.7 BASIS FOR REGULATIONS
14.7.1 Evaluations of BPT Treatment Practices
All the plants producing titanium dioxide by the chloride
process using rutile ore or upgraded ilmenite ore practice
neutralization and settling for control and treatment of the
waste effluents. A variation in the effluent quality is
expected because of the method of handling the unreacted ore and
coke (generated as solid residue from the chloride process).
Two of the five plants haul the residue to a secure landfill for
disposal while the remainder slurry the residue with water and
send it to the treatment system. No information is available
about recycling the treated waste water at any of the plants.
Pollutant Removal with BPT Treatment
Treatment Level 1 is equivalent to the proposed BPT in the
TiO2 subcategory (chloride process).
Plants |559 and f!72 practice neutralization and settling
of the raw waste. At Plant 1559, the chloride process raw waste
385
-------
TABLE 14-8. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE-Chloride Process
Production 16,900 metric tons per year (18,632 tons per year)
48 metric tons per day (53 tons per day)
Waste water flow 1485 cubic meters per day.
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT COST
Construction- $368,500 $49,000
Equipment in place,
including piping,
fittings, electrical
work and controls 209,000 389,000
Monitoring equipment
in place 9,000
Engineering design
and inspection 117,300 87,600
Incidentals, overhead,
fees, contingencies... 117,300 87,600
Land 192,000 6,000
TOTAL INVESTMENT COST $1,013,100 $619,200
B. OPERATION AND
MAINTENANCE COST
Labor and supervision. $56,000 $84,000
Energy 3,700 4,300
Chemicals 140,000 34,100
Maintenance 82,110 61,320
Taxes and insurance... 30,393 18,576
Residual waste
disposal 108,000 9,000
Monitoring, analysis
and reporting 15,000 7,500
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
$435,203
$133,592
$568,795
$218,796
$99,767
$318,563
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
386
-------
TABLE 14^9. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE-Chloride Process
Production 25,500 metric tons per year (28,113 tons per year)
72 metric tons per day (80 tons per day)
Waste water flow 2240 cubic meters per day.
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT COST
Construction $525,000 $50,800
Equipment in place,
including piping,
fittings, electrical
work and controls 228,000 450,000
Monitoring equipment
in place 9,000
Engineering design
and inspection 152,400 100,160
Incidentals, overhead,
fees, contingencies... 152,400 100,160
Land 276,000 6,000
i?
TOTAL INVESTMENT COST $1,342,800 $707,120
B. OPERATION AND
MAINTENANCE COST
Labor and supervision. $56,000 $84,000
Energy 4,000 5,500
Chemicals „ 211,000 51,000
Maintenance 106,680 70,112
Taxes and insurance... 40,284 21,213
Residual waste
disposal 164,000 11,000
Monitoring, analysis
and reporting... 15,000 7,500
TOTAL OPERATION AND
MAINTENANCE COST $596,964 $250,325
C. AMORTIZATION OF
INVESTMENT COST $173,568 $114,072
TOTAL ANNUAL COST $770,532 $364,397
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
387
-------
TABLE 14-10. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE-Chloride Process
Production 45,200 metric tons per year (49,833 tons per year)
129 metric tons per day (142 tons per day)
Waste water flow 3980 cubic meters per day.
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT COST
Construction
Equipment in place,
including piping,
fittings, electrical
work and controls
Monitoring equipment
in place
Engineering design
and inspection
Incidentals, overhead,
fees, contingencies...
Land
TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Energy
Chemicals
Maintenance
Taxes and insurance...
Residual waste
disposal
Monitoring, analysis
and reporting
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
t
TOTAL ANNUAL COST
$815,500
283,000
9,000
221,500
221,500
504,000
$2,054,500
$56,000
4,600
374,000
155,050
61,635
294,000
15,000
$960,285
$252,266
$1,212,551
$76,800
590,000
133,360
133,360
6,000
$939,520
$84,000
7,650
95,000
93,352
28,185
20,000
7,500
$335,687
$151,883
$487,570
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
388
-------
2n
,_,
o
o
o
o
0 1 5
rH
X
>
EH
O
o
EH
8
§ 1.0
3
H
g
S
0.5
•4
<3
/
/
y
^
/
.
/
s
/
Ci
'
(•'
/
/
f
/
{
f
/
/
/
/
/
/
t
/
/
/
/
/
/
/
/
/
'
/
<£
-------
60
50
40
I
30
ffl
I El
rat.
il
20
10 20 30 40 50
ERODOCTICN .(METRIC TONS/YEAR X 1000 )
Figure 14-8. Annual unit treatment cost vs. production for the
Titanium Dioxide Subcategory, Chloride Process
390
-------
TABLE 14-11. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE-Chloride Process
Annual Treatment Costs {$/kkg)
COST HEM
PRODUCTION FLOW
(kkg/yr) (m3/day)
LEVEL OF TREATMENT
FIRST SECOND THIRD FOURTH
Annual Operation
and Maintenance
Annual
Amorti zation
Total' Cost
16,900
25,500
45,200
1,485
2,240
3,980
16,900
25,500
45,200
16,900
25, 500
45,200
1,485
2,240
3,980
1,485
2,240
3,980
25.75
23.41
21.25
7.90
6.81
5.58
33.66
30.22
26.83
12.95
9.82
7.43
5.90
4.47
3.36
18.85
14.29
10.79
13.27
10.09
7.65
6.07
4.60
3.47
19.33
14.68
11.12
Ntot
Applicable
391
-------
water is mixed with the sulfate process waste water for
treatment. Also at Plant 1559, the spent ore and coke (solid
residues from the chloride process) are slurried with water and
sent to the treatment facility whereas at Plant ^172, the solid
residues are hauled to a chemical landfill. Long-term treated
effluent data have been submitted by both Plants 1559 and 1172.
The derivation of the variability factors for daily and 30-day
averages for both plants are given in Tables 14-12 and 14-13.
The concentration of the raw waste and treated effluent
along with the percent removal of the pollutants by the
treatment system for Plants 1559 and 1172 sampled in the
screening and verification program are given in Table 14-14.
14.7.2 Basis for Proposed BPT Effluent Limitations
Technology Basis
For BPT, the Agency is proposing limitations based on
equalization, neutralization, and settling or clarification.
All plants in this segment of the industry have BPT technology
installed.
Plow Basis
The flow going to the treatment system at different plants
varies and is dependent on the method of disposal of the spent
ore and coke (pit solids) and on the finishing operation. The
spent ore and coke are either hauled to a landfill as solid
residue or sent to the treatment system. For the purpose of the
model plant treatment system, the solid residues from the
manufacturing process are assumed to be slurried with water
and sent to the treatment system. Plants f559 and f!99 do, in
fact, send the solid residues to the treatment system. The
model plant treatment system is based on an inflow rate of 100
m3/kkg of Ti02 which is an average value of the effluent flow of
Plants §559 and |199. The treated effluent flow is assumed to
be the same as the influent flow. The water added or
removed in the treatment system through chemical addition,
precipitation, and evaporation have been neglected, as it
varies from plant to plant and is dependent on the selection
of treatment chemicals as well as climatic conditions and is
insignificant in comparison to the total flow.
Selection of Pollutants to be Regulated
The selection of pollutants for which regulations are
being proposed was based on an evaluation of the waste data
from the screening and verification sampling program. The
two major factors considered were the individual plant raw
waste concentrations and the total subcategory pollutant
loadings.
392
-------
14-12. HISTORICAL
DATA SUMMARY WHS mRIABILirY FACTOR
Daily Measurements
Subcategory: Titanium Dioxide
Chloride Process (Rutile Ore)
Plant #559
April 76 through September 78
Pollutant
Daily Data
TSS
(1)
No. of Points 889
Average x,
ppn 21
Cadmium Chromium Iron Lead
109 128 854 128
0.058 0.072 0.620 0.068
Nickel
128
0.08
Zinc
128
0.151
Standard
Deviation, S
Standard
Deviation, S1
65.93 0.044 0.054 3.46 0.041 0.07 0.20
1.54 0.68 0.67 1.86 0.56 0.76 1.02
13.5 3.2 4.4 6.4
30-day (1)
Averages
No. of Points 30 26 30
Sfa^d 21.84 0.042 0.038
Deviation
Variability •* 04 24 2 04
Factor
Variability
Factor Ratio
WR 3.6 1.6 1.9
28 30 30 30
0.94 0.04 0.05 0.155
4.0 2,1 4.4 3.1
3.4 1.5 1 2.1
(1) Section 8.2 presents a discussion of the approach and methodology employed
in the statistical evaluation of data.
(2) VER is the ratio of the variability factor for daily measurements to the
variability factor for 30-day averages.
393
-------
14-13. HISTORICAL EEFHJHST MONZDORJNG niaca. SUMMARY WITH VARIABILITY
FACTORS
EMLY MEASUREMENTS
SOBCATEGORY: TITANHM DIOXIDE-ChlorMe Process
(RutxLe/Opgraded Ilmenite Ore)
Plant 1172
Daily Data{1)
No. of Points
Average ic, ppm
Standard deviation, S
Standard deviation, S1
Variability factor
30-Day Averages
No. of Points
Standard deviation, S
Variability factor
/ 1*
Variability Factor Ratio ^
VFR
TSS
454
5.39
9.13
1.16
7.6
15
6.31
2.92
0
2.6
Pollutant
Chromium
454
0.008
0.016
1.27
8.6
15
0.012
3.46
2.5
Copper
454
0.02
0.03
1.08
6.9
15
0.028
3.29
2.1
Einc
454
0.02
0.027
1.02
6.4
15
0.026
3.13
2.1
(1) Section 8.2. presents a discussion of the approach and methodology employed
in the statistical evaluation of data.
(2) VFR is the ratio of the variability factor for daily measurements to the
variability factor for 30-day averages.
394
-------
TABLE 14-14. TREATMENT PERFORMANCE DATA OF SAMPLED PLANTS #559 MID #172
SUBC&TEGOKY: THANIUM DIOXIDE-ChlorMe Process
Pollutant
Plant #559 Plant #172
Pollutant
Concentration
(rag/1)
A
Raw
Waste
B
Treated
Effluent
Percent
Removal
cK^oo
Pollutant
Concentration
(rag/1)
D
Raw
Waste
E
Treated
Effluent
Percent
Removal
-f^oo
TSS
Iron
Chromium
Lead
Nickel
Zinc
1103
288
13.3
0.5
0.5
0.3
23
4.4
0.03
0.002
0.005
0.06
97.9
98.5
99.8
99.6
99.0
80.0
171
2.9
0.72
0.005
0.08
0.3
6.7 .
0.33
0.02
0.002
0.01
,0.09
96.1
88.6
97.2
60 -
87.5
70.0
395
-------
Raw waste concentration - Plant #559 was visited in the
screening phase for sampling of the raw and treated waste
water. For each pollutant, the maximum concentration observed
gave a preliminary indication of its pollution potential.
Five pollutants were found above treatabilitv levels in the raw
waste of Plant f559 and they were: chromium, iron, nickel,
lead, and zinc (Section 14.3.3). A second plant, f.172, was
sampled in the verification phase and chromium was the only
pollutant found above treatability levels in the raw waste. At
Plant f559, the waste water from the chloride process is mixed
with the sulfate process waste water and the chloride process
effluents were sampled at the point of mixing. It is highly
probable that the sampled waste included the sulfate process
effluent impurities. The sulfate process for the manufacture of
TiO2 uses an ore of lesser purity. For this reason the nickel,
lead, and zinc found are attributed to the sulfate process and
are not further considered in this segment. The nonconventional
and toxic pollutants of concern include chromium and iron.
Iron, a nonconventional pollutant is significant because it is
present as a major impurity in the rutile or upgraded ilmenite
ore and was found at treatable levels in the Plant |559 raw
waste.
Total subcategory raw waste pollutant loading - Chromium
was the only toxic pollutant found in significant concentrations
in the raw waste of both plants sampled in the screening and
verification phase. The average unit raw waste chromium loading
(Table 14-7) obtained from the plants sampled was multiplied by
the total TiO2 subcategory production by the chloride process to
evaluate the overall magnitude of the pollutant potential for
the subcategory. The value of 241,000 kg/year of chromium
discharged by the subcategory in the effluent indicated the
necessity of control of this pollutant.
The treatment technology practiced by the industry removes
the chromium and iron to low levels as seen from the effluent
quality of the plants sampled and shown in Table 14-14.
Basis of Pollutant Limitations
Conventional and nonconconventional parameters -
A. pH: The treated effluent is to be controlled within
the range of 6.0 to 9.0. This limitation is based on the data
presented in Appendix B of this report and the JRB Study (52).
B. TSS: Long-term effluent data is available for TSS for
Plants f559 and #172. At Plant §172, the solid residues from
the manufacturing process are sent to a landfill. Although the
amount of solids sent to the treatment system at Plant ^559 is
high compared with that selected for the model plant (because of
396
-------
intermixing of sulfate waste), the long-term average
concentration of 21 mg/1 (Table 14-12) calculated from the
monitored data submitted by Plant $559 is selected as the
treatment performance basis for the subcategory. The daily and
the 30-day average variability factors (11 and 3.04) derived
from the long-term data of Plant #559 and given in Table 14-12
are used to calculate the concentration basis. The proposed
unit effluent limitations are calculated using the model plant
unit flow of 100 m3/kkg. The calculations are given below:
Proposed 30-day average concentration
= (21 mg/1)(3.04) = 64 mg/1
Proposed 24-hour maximum concentration
= (21 mg/1)(11) = 230 mg/1
Proposed 30-day average effluent limit
= (64 mg/1) (100 mS/kkg)/' kg/m3 \
VlOOO mg/1/
= 6.4 kg of TSS
kkg of Ti02
Proposed 24-hour maximum effluent limit
= (230 mg/1) (100 m3/kkg) f kg/m3 >\
V1000 mg/V
= 23 kg of TSS
kkg of TiO2
C. Iron: The subcategory performance standard of 0.62
mg/1 selected for iron is based on the long-term average of the
effluent data submitted by Plant f559 (Table 14-12).
For the model plant, it is assumed that iron is present in
the ferric state in the raw waste from the chlorination process.
Using the daily variability factor of 4.0 and the 30-day average
variability factor of 13.5 estimated from the long-term
monitored effluent data of Plant $559 for iron (Table 14-12),
and the model plant unit flow of 100 m3/kkg, the proposed
concentration basis and effluent limitations are determined as
shown below.
Proposed 30-day average concentration basis is:.
(0.62 mg/1)(4.0) = 2.5 mg/1
397
-------
Proposed 24-hour maximum concentration basis is:
{0.62 mg/l}{13.5) = 8.4 mg/1
Proposed 30-day average effluent limit is:
(2.5 mg/1) (100 m3/kkg)/ m3/kkg N
\1000 mg/lj
= 0.25 kg of iron
kkg of Ti02
The proposed daily maximum effluent limit is:
(8.4 mg/1)(100 m3/kkg) / m3/kkg \ = 0.84 kg of iron
V1000 mg/lj kkg of Ti02
The proposed maximum 30-day average concentration of 2.5
mg/1 is higher than the achievable effluent concentration of 1.6
mg/1 reported in the literature (10). The latter concentration
is based on the performance of lime neutralization followed by
settling of acid mine drainage waste, and may not be appropriate
for this subcategory.
Toxic pollutants - Chromium is the only regulated toxic
pollutant because of its presence in the raw waste of the plants
surveyed at treatability levels.
A. Chromium: The proposed chromium limitations are based
on the long-term data of the treated effluent of Plant 1559.
The influent to the treatment system at Plant fS59 contains
significant amounts of chromium because of the intermixing of
sulfate process waste. The long-term average of 0.07 mg/1
derived from the monitored data of the treated effluent of
Plant 1*559 (Table 14-12) is taken as the proposed subcategory
performance concentration. The daily variability factor of 2,0
and 30-day variability factor of 3.8 estimated from the long-
term data of Plant |559 (Table 14-12) and the model plant
effluent flow of 100 m3/kkg of Ti02 are used to derive the
proposed concentration basis and effluent limitations. The
calculations are shown below.
The proposed 30-day average concentration basis is:
(0.07 mg/1)(2.0) = 0.14 mg/1
The proposed daily maximum concentration basis is:
(0.07 mg/1)(3.8) = 0.27 mg/1
398
-------
The proposed 30-day average effluent limit is:
(0.14 mg/1)(100 m3/kkg) / kg/m3 N
V1000 mg/lj
= 0.014kg of chromium
kkg of Ti02
The proposed daily maximum effluent limit is;
(0.27 mg/1}(100 m3/kkg)/kg/m3 \ = 0.027 kg of chromium
\1000 mg/1/ kkg of Ti02
B. Other metals: Lead, nickel, and zinc were found in the
raw waste of Plant |559 in the treatability range. They were
not found in the raw waste of Plant |172. Plant 1559 intermixes
the chloride and sulfate process waste before treatment. The
presence of these pollutants in the raw waste at Plant £559
might result from the sulfate process waste. The limitations
for the three pollutants are given and are intended to serve as
guidelines in cases where the pollutants are found to be of
serious concern.
The selected 30-day concentration basis for lead, nickel,
and zinc are based on the lower treatability limits achieved by
the lime precipitation and settling of metal contaminated waste
(Table 8-11) and the values are:
Lead = 0.3 mg/1
Nickel = 0.2 mg/1
Zinc = 0.5 mg/1 ,
The variability factor ratio of 1.9 for chromium estimated
from the long-term data of Plant |559 (Table 14-12) is used to
obtain the daily maximum proposed limits for lead, nickel, and
zinc. This variability factor ratio was used because
precipitation of chromium is similar in performance to
precipitation of other metals. Calculations are as follows:
The proposed 24-hour maximum concentration basis for lead
is:
(0.30 mg/1) (1.9) = 0.60 mg/1
The proposed 24-hour maximum concentration basis for nickel
is:
(0.20 mg/1)(1.9) = 0.40 mg/1
The proposed 24-hour maximum concentration basis for zinc
is:
(0.50 mg/1)(1.9) = 1.0 mg/1
399
-------
The proposed limitations for BPT are given in Table 14-15.
14.7.3 Basis for Proposed BCT Effluent Limitations
The BCT limitation (applicable only to TSS) was set equal
to BPT because BAT is equal to BPT.
14.7.4 Basis for Proposed BAT Effluent Limitations
The Application of Advanced Level Treatment
The advanced level technologies, viz,, the use of sulfide
and xanthate as a polishing step to the base level treatment
system (BPT) f were considered for BAT and NSPS but were rejected
on the basis of cost (Level 3 Table 14-11). Level 1, used for
BPTr is selected for BAT treatment technology.
Technology Basis
Alkaline precipitation followed by settling used for BPT
(Level 1) is proposed for BAT.
Flew Basis
A unit waste water flow rate of 100 m3/kkg of TiO2 used for
the BPT model plants has been selected for BAT.
Selection of Pollutants to be Regulated
Chromium and iron are the two pollutants identified for
regulation.
Nonconventional pollutants -
The proposed iron limitations are the same as those
selected for BPT.
Toxic pollu t an t s -
A. Chromium: The limitations proposed for BAT are
selected for BPT.
B. ' Other metals: Concentration limits for lead, nickel,
and zinc are not proposed as limitations. However, they are
contained in this document for use if these pollutants are found
to be of concern. The values are the same as those selected for
BPT.
Table 14-16 gives the proposed limitations for BAT.
400
-------
TABLE 14-15- PROPOSED LMECATIONS
TITANIUM DIOXIDE - CHLORIDE PROCESS (HJTILE OR UPGRADED nMENTTE ORE)
Best Practicable Control Technology*Currently Available
Waste Hfeter Flow: 100 m3/kkg of Ti02
Concentration Effluent Limit
Subcategory Basis (mg/1) (kg/kkg of Ti°2)
Pollutant Performance VFR(1) ffex ^^
(mg/1) 30-day 24-hr 30-day 24-hr
Avg Max Avg Max
Conventional and
Non Conventional
Pollutants:
(2)
Total Suspended Solids 21 v ' 3.6
(2) 3 4
Iron 0.62^' J'
Toxic Pollutants:
Chromium 0.070^ 1.9
Lead 0.30(3) 1.9
Nickel 0.20(3) 1.9
Zinc 0.50(3) 1.9
64 230 6.4 23
2.5 8.4 0.25 0.84
0.14 0.27 0.014 0.027
0.30 0.60 _ (4) _ (4)
0.20 0.40 _ (4) - (4)
0.50 1.0 _<4> _(4)
(1) VFR: Ratio of the 24-hour variability factor to the 30-day variability
factor.
(2) Long term average based on loading data and variability factors of plant
#559 selected from Table 14-11.
(3) The lower limit of the literature treatability estimate (Table 8-11) is
used as the basis of or the 30-day average limitation.
(4) No- effluent limitation proposed.
401
-------
TABLE 14-16. PROPOSED LMTATIONS
THENIOM DIOXIDE - CHLORIDE PROCESS (RUTILE/UPGRADED ILMENITE ORE)
Best Available Technology
Waste Water Flow; 100 m3/kfcg of Ti02
Pollutant
Concentration Effluent Limit
Basis (mg/1) (kg/kkg of TiO )
30-day 24-hour 30-day 24-hour
Avg Max Avg Ma,x
Nonconventional
Pollutants ;
W4>
Toxic
PoXtatants:
Chromium^ '
Nickel
Zinc
0.62 3.4 2.5 . 8.4 0.25 0.84
0.070 1.9 0.14 0.27 0.014 0.027
0.30 1.9 0.30 0.60 (3) {3)
0.20 1.9 0.20 0.40 (3) (3)
0.50 1.9 0.50 1.0 (3) (3)
(1) See Table 14-14 for details.
(2) WE: Ratio of the 24-hour variability factor to the 30-day variability
factor.
(3) No effluent limitation proposed.
(4) Limitations are applicable for PSES.
402
-------
14.7.5 Basis for Proposed New Source Performance Standards
Application of Advanced Level Treatment
The concentration of conventional, nonconventional, and
toxic pollutants can be reduced by filtering the clarified
effluent from BPT in a dual media filter.
Technology Basis
For new plants, the recommended waste water treatment
technology is lime neutralization and precipitation, settling,
and dual media filtration (equivalent to Level 2) . All the
existing chloride process plants using rutile/upgraded ilmenite
ore currently practice lime neutralization and settling, but
only published treatability data is available on the performance
of dual media filters.
Flow Basis
The raw effluent flow rate is the same as that used for
BAT, namely 100 m3/kkg of Ti02. It is assumed that the
unreacted ore and coke are slurried with water and sent to the
treatment system. The selected flow value is an average of the
unit effluent flow rate of two plants (|559 and f!99) practicing
this method of solids disposal.
Selection of Pollutants to be Regulated
It is proposed that the pollutants regulated for BPT are
also regulated for NSPS. The pollutant parameters of concern
are pH, TSS, iron, and chromium. Concentration limits are
provided for lead, nickel, and zinc in cases where these
pollutants become of concern.
C onv e n tiona1 parameters ~
A. pH: For NSPS, the BPT limitation is retained. Control
of the final effluent within the range of pH 6.0 to 9.0 is
required on the basis of data presented in Appendix B of this
report and the JRB Study (52).
B. TSS: There are no in-plant performance data available
on the filtration of treated and clarified NSPS waste water, so
a 30 percent additional removal is assumed. This assumption is
based on treatability studies (41) using filtration. This
reduction is applied to the selected BAT (or BPT) maximum 30-day
average of 64 mg/1. The proposed maximum 30-day average
concentration basis is then given by: 64 mg/1 (1.00 - 0.30) = 45
mg/1. Likewise, the proposed 24-hour maximum concentrations
and unit effluent limitations are obtained from the BAT
limitations (Table 14-15) as shown below.
403
-------
The proposed 24-hour maximum concentration is:
(230 mg/1)(0.70) - 160 mg/1
The proposed 30-day average effluent limit is;
(6.4 kg/kkg)(0.70) = 4.5 kg of TSS
kkg of Ti02
The proposed 24-hour maximum effluent limit is:
(23 kg/kkg)(0.70) = 16 kg of TSS
kkg of TiO2
Konconventional pollutants -
A. Iron: No in-plant performance data is available on the
effect of dual media filtration on the removal of iron from the
lime treated and clarified waste water. The removal efficiency
of 30 percent obtained for TSS from the treatability studies
(41) is assumed also to apply to iron, since the iron is present
as a floe. The proposed concentration basis and effluent
limitation for NSPS are obtained by multiplying the selected BAT
(or BPT) limitations (Table 14-15) by 0.70 as follows:
The proposed 30-day average concentration basis is:
(2.5 mg/1)(0.70) » 1.8 mg/1
The proposed 24-hour maximum concentration basis is:
(8.4 mg/1)(0.70) - 5.9 mg/1
The proposed 30-day average effluent limit is:
(0.25 kg/kkg)(0.70) = 0.18 kg of iron
kkg of T1O2
The proposed 24-hour maximum effluent limit is:
(0.84 kg/kkg)(0.70) = 0.59 kg of iron
kkg of TiO2
Toxic pollutants -
A. Chromium: For NSPS, the Agency is proposing more
stringent controls on the discharge of chromium. There is no
directly applicable data on filter performance in the TiO2
(chloride process) industry. Therefore, the proposed
limitations are based on literature treatability studies
(40,41). In pilot scale treatability tests, raw municipal waste
404
-------
water samples were spiked with toxic metals, treated with lime
and settled. This was followed by dual media filtration of the
clarified effluent. For chromium, an additional removal of 60
percent was obtained by filtration. This reduction factor is
assumed to be applicable. The proposed limitations for NSPS are
obtained by multiplying the respective BAT (or BPT) limitations
(Table 14-16) by 0.40 as follows:
The proposed maximum 30-day concentration basis is:
(0.14 mg/l)(0.40) = 0.060 mg/1
The proposed daily maximum concentration basis is;
(0.30 mg/1) (0.40) = 0.10 mg/1
The proposed 30-day average effluent limit is:
(0.014 kg/kkg) (0.40) = 0.0060 kg of chromium
kkg of TiO2
The proposed maximum daily effluent limit is:
(0.030 kg/kkg) (0.40) = 0.010 kg of chromium
kkg of TiO2
B. Other metals: Treatability studies have indicated
that the following increased removals of lead, nickel, and zinc
can be achieved by filtration (40,41).
Additional Removal by
Filtration Using Settled Effluent
Lead 60
Nickel 14
Zinc 6
The additional levels of removal are applied to the
corresponding BAT (or BPT) concentration for the above metals to
get the NSPS concentrations.
The proposed 30-day average lead concentration basis is:
(0.30 mg/1) (0.40) = 0.12 mg/1
The proposed 24-hour lead concentration basis is:
(0'.60 mg/1) (0.40) = 0.24 mg/1 ^
405
-------
The proposed 30-day average nickel concentration basis is:
(0.20 mg/l)(0.86) = 0.17 mg/1
The proposed 24-hour maximum concentration basis for nickel
is:
(0.40 mg/1) (0.86) = 0.34 mg/1
The proposed 30-day average concentration basis for zinc
is:
(0.50 mg/1)(0.94) = 0.47 mg/1
The proposed 24-hour maximum concentration basis for zinc
is:
(1.0 mg/1) (0.94) = 0.94 mg/1
The proposed conventional, nonconventional, and toxic pollutant
limitations for NSPS are given in Table 14-17.
14.7.6 Basis for Pretreatment Standards
Existing Sources
For Pretreatment Standards for Existing Sources (PSES), the
Agency is proposing limitations based on BAT. The
pollutants to be limited are iron and chromium (see Table 14-
15).
New Sources
For Pretreatment Standards for New Sources (PSNS), the
Agency is proposing limitations based on NSPS. The pollutants
to be regulated are iron and chromium (see Table 14-17).
406
-------
Pollutant
TABUS 14-17. PROPOSED LIIOTIATIONS
TITANIUM DIOXIDE - CHLORIDE PROCESS
New Source Performance Standards
Waste Water Flow: 100 m3/kkg of TiO2
Treatability VFR
(1)
Concentration Effluent Limit
Basis, (ragA) (kg/kkg of TioJ
Max Max
30--day 24-hour 30-day 24-hour
Avg Max Avg Max
Conventional and
Nonconventional-
Pollutants :
Total Suspended
Solids
lron<2>
Toxic Pollutants;
(2)
Chromium
Lead
Nickel
Zinc
15
0.40
0.030
0.060
0.17
0.47
3.6
3.4
1.9
1.9
1.9
1.9
45 160 4.5 16
1.8 5.9 0.18 0.59
0.060 0.12 0.0060 0.012
0.12
0.17
o.47
0.24
0.34
0.94
(3) (3)
(3) (3)
(1) VFR: Ratio'of 24-hour variability factor to> the 30-^day variability
factor.
(2) Also applicable for PSNS limitations.
(3) No effluent limitations proposed.
407
-------
14.8 TITANIUM DIOXIDE - SDLFATE PROCESS
INDUSTRY PROFILE
14.8.1 General Description
The industrial profile for the Sulfate Process Segment of
the Titanium Dioxide Subcategory is presented in Table 14-18 and
the status of regulations is shown in Table 14-2.
14.8.2 General Process Description and Raw Materials
Sulfate Process - General Description
Among the various titanium ores, ilmenite is available in
abundance. Ilmenite is a low-grade titanium ore with a Ti02
content varying from 45 to 60 percent. Ilmenite ore and slag
from iron production generally comprise the raw materials used
for the preparation of titanium dioxide by the sulfate process.
Large amounts of water and sulfuric acid are used in this
process, and the majority of the plants are co-located with
sulfuric acid plants. Table 14-19 gives the analysis of various
ilmenite ores. The preparation of T102 by the sulfate process
utilizes three important steps:
1. Digestion: FeO.Ti02 + 2H2S04 = PeS04 + TiO.S04
+ 2H20
2. Precipitation: TiO.S04 + 2H20 • Ti02.H20 4- H2S04
3. Calcination: Ti02.H20 = Ti02 + H20
The ore is dried, ground, and then reacted with sulfuric
acid* The reaction takes place at 160 degrees C and the reacted
mixture consists of titanyl, ferrous, and ferric sulfates. The
product is dissolved in water. The total iron in the reacted
product is kept in the ferrous state by the addition of scrap
iron. After the reduction, the product is dissolved in water
and clarified with the aid of flocculating agents to remove
insoluble impurities such as silicon, zirconium, and unreacted
ore. The iron is removed from the clear solution by cooling the
solution to 10 degrees -C when FeS04.7H20 crystallizes. The
ferrous sulfate crystals, commercial copperas, are mechanically
separated from the solution by filtration or centrifugation.
The concentrated titanyl sulfate solution is diluted with water
and heated to form titanium dioxide hydrate, which is known as
strong acid, is separated and either discharged or recycled.
The Ti02.H20 filter residue is slurried with water and
conditioning agents are added to control particle size, color,
dispersibility, and photochemical stability. The conditioning
agents include potassium, zinc, antimony, and calcium compounds,
"and phosphate salts. The solution is filtered and the filtrate
408
-------
TABLE 14-18. - SUBGffiEGQRY PROFILE DHSA SUMMARY
SIBCMEGORY
TITANIUM DIOXIDE
SULEATE PROCESS
Total subcategory capacity rate
Total subcateogry production rate
Number of plants in this subcategory
308 Data on file for
With total capacity of
With total production of
Representing -.capacity
Representing production
Plant production range:
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volume per unit product
Minimum
Maximum
401,000 kkg/year
259,000 kkg/year
4
5
320,000 kkg/year
246,000 kkg/year
80 percent
95 percent
31,000 kkg/year
74,500 kkg/year
49,000 kkg/year
43,000 kkg/year
76 percent
23 years
54 years
35,000 cubic meters/day
125,000 cubic meters/day
300 cubic meters/kkg
780 cubic meters/kkg
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Cotanetce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chemical Industry," June, 1978 and "Economic Analysis of Proposed Revised
Effluent Guidelines and Standards for the Inorganic Chemicals Industry "
March, 1980.
409
-------
5MBIJ3 14-19. SHMXSIS OF HMEKCHE OSES1
H
O
Chemical
Constituent
TiO,
PeO
Fe203
si°2
A1203
P2°5
ZrO2
MgO
MnO
CaO
^2%
Cr203
Virginia
Piney
River Eoseland
44.3
35.9
13.8
2.0
1.21
1.01
0.55
0.07
0.52
0.15
0.16
0.27
51.4
37.9
1.6
4.6
0.55
0.17
NA
2.35
0.70
0.59
0.07
NA
UNITED SMCES
Hew York Florida
44.4
36.7
4.4
3.2
0.19
0.07
0.006
0.80
0.35
1.0
0.24
0.001
64.1
4.7
25.6
0.3
1.5
0.21
NA
0^35
1.35
0.13
0.13
0.1
California
48.2
39.1
10.4
1.4
0.2
NA
0.05
0.6
0.1
0.1
0.05
0.03
Ivry
42.5
39.1
20.7
0.88
1.05
NA
NA
2.0
0.04
0.1
0.36
0.15
CANADA
Bourget
22.4
36.9
31.2
1.0
6.01
0.93
NA
1,50
NA
0.55
NA
NA
Allard
37.3
26,3
30.0
NA
NA
0.004
NA
NA
0.10
NA
0.39
NA
Constituents expressed as weight percent.
NA! Not Available
-------
is known as weak acid. Residual acids and iron originally
present in the precipitate are removed with the water of
hydration by calcination. The resulting Ti02 pigment is sent to
finishing operations, which vary according to the end product
requirement and application. The wet finishing operations may
include some, or all, of the following steps: repulping,
milling, surface treatment with proprietary agents in solution,
washing, and drying. The alternative dry finishing operations
may include one or more milling steps followed by packaging. A
simplified block diagram of the sulfate process is shown in
Figure 14-9.
14.9 WATER USE AND WASTE SOURCE CHARACTERISTICS
14.9.1 Water Use
Water is used in the preparation of titanium dioxide by the
sulfate process for noncontact cooling, air emission control,
and for process reactions. In the process, water is used to
leach the soluble sulfate salts from the reaction mass and to
convert the titanyl sulfate to titanium dioxide hydrate. Water
is also used to wash the titanium dioxide hydrate precipitate
free from residual acid and iron. Water is used for air
emission control during the drying of ore, on digester units,
and for the cleaning of the kiln gases before they are vented to
the atmosphere. In the digester unit, water seals are used to
maintain a vacuum on the digester units. Large amounts of water
are also used in the finishing operations. Table 14-20 is a
summary of water usage in the titanium dioxide subcategory using
the sulfate process.
14.9.2 Waste Sources
Digester Sludge
After the digestion of the ore in sulfuric acid, the
result-ing sulfates are dissolved in water and the insoluble
impurities are removed in a clarifier or filter. These include
silica, alumina, sulfuric acid, and unreacted iron. The quality
of this waste varies and depends on the type and quality of ore
used. Data on the quantity of this waste indicates that
approximately 210 kg/kkg is produced.
Copperas
\
The recovered ferrous sulfate is marketed or disposed of as
a solid waste. The amount of copperas generated is about 950
kg/kkg of Ti02. The copperas generally contain small amounts of
adsorbed "sulfuric acid.
411
-------
HH5JUTB
STROHG-ACID {
WAfSR — •
DIGESTER
HATER ... , fc,
WATER"""*
STEAM — »,
WEAK-ACID
RECfCLE ^
WATER— »
CLARIFIER
*
EVAPORATOR
*
PRECIPITATION
1
FLASH
COOLER
— EHI8SIOHS — »"
FPRAY COK01HSEBS uvurjieu'r fci
WATER — to ADO VEOTORI SCRUBBERS «rrl««ia. ^
Mnn qt.flTIBV itmn.it sfTnV «-
_• EMISSIONS — », ' ' -
„.___ CONDENSERS — EFFLUENT —*•
WATER 1>
CONDEHSERS — EFFLUENT ~*
tjumoo ^
J
1 FIRST MOORE
PTr.TBn
HATER — *»
SECOND MOORE
FILTER
STEftli *^ •
1 RECXCLE »
HATER — •>
STEAM »
— *
DZ
„„ , „, HEAR M"TP m
! '
CALCINER
HEX MILL
— KMIB5IOMS » COOLIMG SPRAYS AHD ELECJ-._..fc ..,„,., J,CID_
HATER UROSTATIC PRECIPITATORS | ~~
1
EMISSIONS
*
HATER * MIST ELIMINATORS 1 EFFLDEHT — »•
|»BI.nBHT ^
" ""'•"• ff , i ....
1
JET MILLS
i- - •» i H5» i un^ fc
IfATEH •— -9
I
TITANIUM
OXIDE PIGMENT
4 '•
PACKA6IHG '
HATER — - * Jgl MILL SCRUBBERS ^.EFFLUENT——
t
TO SALES
r.m
Figure 14-9. General process flow diagram for production of titanium dioxide by
sulfate process.
£TE DISPOSAL •
-------
14-20. IN DIOXIDE - PROCESS SOBCKEBGOEZ
Uses
Water Usage per Unit of Production
s( m3/kkg of TiO,}
Noncontact cooling
Direct process contact
Indirect process contact
(pumps, seals, leaks,
spills, etc.)
Maintenance, equipment
cleaning and work area
washdown
Mr pollution control
Nbncontact ancillary
uses (boilers, utilities,
etc.)
Plant #555
47.8
390
6
258
36
Plant #694
408
588
1.6
1.8
78
33
Plant #696
149
297
4
81
NA
NA: Not Available
413
-------
Strong Acid Waste
When water is added to titanyl sulfate solution after the
removal of copperas, sulfuric acid and the hydrate of titanium
dixoide are formed. The acid contained in solution is removed
by filtration and the filtrate is known as strong acid solution.
The concentration of sulfuric acid varies from 15 to 30 percent
as H2S04. In addition to sulfuric acid, the waste stream
contains ferrous sulfate, titania, antimony, and other heavy
metal salts. A part of the acid is returned to the process and
the rest sent to the treatment facility.
Weak Acid Waste Stream
The waste generated from washing the titanium dioxide
hydrate precipitate is known as weak acid. The concentration of
sulfuric acid in this waste varies from two to four percent as
H2S04 and contains various impurities, including iron sulfate,
titania, antimony, and other heavy metal salts. It also
includes, in some cases, the conditioning agents added to the
precipitate prior to washing, to control and improve the quality
of the final product. The weak acid may also include the kiln
exhaust scrubber waste.
Scrubber Wastes
Scrubber waste water results from the scrubbing of vapors
emitted during the drying of the ore, during digestion, and
during kiln drying. The amount of waste water generated depends
on the amount of water used and type of emission controls
practiced. The scrubber water contains titanium dioxide
particulate, acid mist, sulfur trioxide and sulfur dioxide. Of
all the waste produced from titanium dioxide-sulfate process
manufacture subcategory, the scrubber waste water constitutes
the major portion.
Wet Milling Waste
These wastes are generated during wet finishing of the
titanium dioxide pigment. Wet milling is used to produce
pigment particles of the desired size and surface character and
requires steam and water for repulping the pigment. Caustic
soda is also used to remove any residual acidity from the
titanium dioxide pigment during the finishing operation. The
waste water from wet finishing operation, therefore, contains
titania, sodium sulfate, and other agents added to improve or
achieve desired properties in the final product.
414
-------
14.10 DESCRIPTION OP PLANTS
14.10.1 Screening
Plant |555 was visited and its waste streams sampled in the
screening phase by an EPA Region II team. The pigment
manufacturing operation utilizes a titania slag for the
production of Ti02 by the sulfate process. After digestion of
the slag in sulfuric acid the residual gangue material is
filtered out and the clear liquor is concentrated by
evaporation. The crude pigment is formed by hydrolysis with
water and steam and processed to form both anatase and rutile
type pigment products. Table 14-21 presents raw waste flows and
pollutant characteristics for Plant f555.
Waste water samples were collected at five points and
analyzed for the conventional, nonconventional, and toxic
pollutants. These sampling points were designated as 1) the
digestion suppression flume containing waste water from direct
contact air scrubbers on the digesters, 2) the black end flume
containing wastes from major cuttings, filter sludges, acid
filtrates, and evaporator and condenser waters, 3) the white end
flume carrying finishing process filtrates, noncontact cooling
water, and sanitary wastes, 4) northside jet air scrubbers, and
5) southside jet air scrubbers.
At present , all of the process waste streams are collected
in a settling basin which is open to tidal fluctuations that
provide diurnal flushing of the effluent into the receiving
waters.
14.10.2 Verification
Plant f559 was surveyed in both the screening phase and
verification phase of the study. At this plant the strong acid
is sent to a lined holding pond for equalization. Effluent from
the pond is neutralized with ground calcium carbonate limestone
in a reactor. A minimum amount is added to raise the pH to a
level such that calcium sulfate, but not ferrous hydroxide, is
precipitated. The CO2 formed during the reaction is vented to
the atmosphere and the , calcium sulfate slurry goes to a
clarifier. The underflow from the clarifier is filtered to
produce pure gypsum crystals at a concentration of 70 to 80
percent.
The weak acid is sent to a settling, pond, where it is
combined with a small quantity of other wastes. The effluent
from the weak acid pond is mixed with the calcium sulfate
clarifier overflow and neutralized with ground calcium carbonate
in a three-stage reactor. Pebble and slaked lime are also added
415
-------
14-21.
BSW C3BWM3ERISTICS (INDUSTRY DATA)
(PRODUCTION OP Ti02 BY
PMNT #555
Waste Source
Digestion
Clarification
Evaporation
Cooling
Unit
(nvVkkg pH*
of TiC>2)
115
3.58
113
20
Strong Acid from 8.49
first Moore Filtration
Weak Acid from
12.2
3.0
2.5
4.0
6.1
< 0.5
2.0
Pollutant Waste Loads, (kg/kkg of TiC^)
Acidity NEU Fe TSS TDS
(asH2S04> (asN)
20.8
26.7
18.7
2.49
2.360
88.3
NA
NA
NA
NA
NA
NA.
0.042
8.42
1.14
0.099
139
3.8
9.3 35.7
175 40 .'8
3.2 20.2
0.46 3.09
0.959 2.815
0.23 98.8
first Moore Filtration
Weak Acid from
10.4
1.7
148
NA
0.29
0.13 151
second Moore Filtration
Weak Acid from
12.0
2.0
20.8
NA
0.22
2.0 7.50
first stage
Calcination
Weak Acid from
second stage
Calcination
40.0
2.2
19.2
NA 0.64
NA: Not A-yailable
* Value in pH units
(1) - Response to 308 Questionnaire, 1976
4.92 33.1
Calcination Mist
Bliminators
Wet Milling Washing
and Drying
Jet-Mill Condenser
Jet-Mill Scrubbers
Boiler and Water
Plants
38.
11.
27.
18.
16.
7
1
0
0
6
3.
8.
6.
7.
9.
0
0
5
4
0
7.50
NA
NA
NA.
NA
NA
8.6
NA
NA
NA
0
0
0
0
0
.02
.01
.01
.13
.66
0.
2.
1.
1.
5.
21
13
1
7
25
27.9
11.0
2.7
3.58
8.92
416
-------
to raise the pH and precipitate more calcium sulfate. Air is
also introduced to convert the ferrous iron to ferric. The
effluent from the reactor goes to another clarifier, and the
clarifier underflow is filtered to concentrate the solids to 70
percent. The overflow from the second clarifier is mixed with
the other process waste waters. These include the scrubber,
finishing, and cooling waste waters. The combined water is
neutralized with slaked lime before it is sent to a final
settling pond, the effluent from which is discharged. Figure
14-10 gives the flow diagram of the treatment process and shows
the sampling locations for both screening and verification.
Table 14-22 gives the flow data for the waste streams and
conventional and nonconventional pollutant emissions,
14.10.3 Other giant Descriptions
At Plant #694, the clarification sludge which contains the
unreacted ore is sent to waste disposal. The weak acid
effluent from the plant is neutralized with slaked lime and the
grit is settled out for landfill disposal. After the separation
of grit, the aqueous stream is discharged to a municipal
treatment system. The other wastes, together with runoff from
the plant site, are collected, and sent to a lagoon for solids
removal, and the overflow discharges to a river.
At Plant 1696, the raw wastes are sent to thickeners to
remove the suspended solids and the overflow is discharged.
Depending on the titanium content, the underflow from the
thickeners is either recycled or disposed of in a landfill.
This plant has discontinued operations.
At Plant |605f the process raw waste streams are combined
and sent to a reactor for neutralization with a water slurry of
finely ground calcium carbonate. The effluent from the reactor
is hydrocycloned into three fractions. The first fraction,
which is the coarse gypsum slurry, is separated from the reactor
effluent at a concentration of 85 to 90 percent, and placed in a
self-draining dewatering system. The "dry" solids are finally
trucked" to a landfill. The second fraction separated in, the
hydrocyclone is a fine gypsum slurry which is recycled to the
neutralization reactor. The residual gel slurry forms the third
fraction, and this is sent to a thickener after C02 degassing.
A flocculating agent is added to the flow to the thickener to
promote solids separation and thickening. The underflow from
the thickener is centrifuged and the solids landfilled. The
filtrate from the centrifuge is recycled to the thickener, and
the thickener overflow is discharged.
The volume and characteristics of waste water streams from
different sulfate process titanium dioxide plants do not differ
greatly. Some variations, however, are noted as a result of
417
-------
WtAK ACID .__,,»,
WASTE STREAK
Itl
SI
H
OTHiR MOPUCT
WASTE VATEA
k
WEAK ACIO rONO
VER
PPL* WATER "
mteiMi
SUPPLY WATER
\
H
00
STRONG AC IB .^^
WASTE STMAH
tt\\ r
iUPPUf WATER
STROKG ACID
MHO
n
/o
i\
^
ft
REACTORS
FILTER
SOLIDS
Tft
,
REACTORS
FILTER
1
»
•" ' HHAL f~+*
i
SOLIDS TO
STORAGE/ ,,
LAHOF1LL k
pnmj ^_7
IS
Vs
^ OTHER PRODUCT
WASTE WATER
TIO. {SULfATI PROCESS)
Z SCRUBBER
WASTE WATER
STORAGE/LANDFILL
LEGEND
mm SAMPLING POINTS
Figure 14-10. Gaieral flow diagram at plant 1559 shewing the sanp-ling points.
(Titanium dioxide - sulfate process.)
-------
14-22. AND FOR THE
FOR
Stream
No. W>
Sampled Unit TSS Iron
Stream Plow Load Load
Description (irtVkkg) (kg/kkg of Ti02) (kg/kkg of Ti02)
Weak Acid
Pond Overflow
68
.I1"2' 1.23
1.23
Strong Acid
Pond Overflow
Scrubber and
Contact Cooling
Water
Final Treatment
Effluent
6.1 205.85
(1) (2)
361.4 113.5
(1) (2) (3)
436 10.0
106.34
51.68
1.92
(1) - The flow is contributed by the sulfate process stream.
(2) - The pollutant load was calculated by multiplying the flow contributed
by the sulfate process stream tines the concentration of pollutant.
Pollutant Load = (total stream flow) x (fraction contributed by sulfate
process waste) x stream pollutant concentrated.
(3) - "While calculating the "unit flow the contributions to the treatment
process from precipitation, the water in the treatment chemicals,.
losses from evaporation and from solids leaving the process have
not been considered.
(4) - See Figure 14-10 for sampling point location
419
-------
differences in ore qualities, in location and in process
details. The majority of the dissolved pollutants in waste
water from this segment of the Ti02 industry consist of acidity
and iron. Segregation of the waste water is important for
control and treatment practices and aids in developing
economically feasible treatment systems. Generally, weak and
strong acid streams are segregated from each other as well as
from the less contaminated waste waters which include contact
cooling, scrubbing, and some finishing operation wastes. The
unit flows for the segregated raw waste streams at different
facilities are shown in Table 14-23.
The average total effluent flow rate is 475 m3/kkg (Table
14-23) for Plants #555, £694, and #559. Complete flow data is
not available for Plants #696 and $605.
14.10.4 Toxic Pollutant Concentrations
Section 5.1.2 of this report describes the scope and
methodology of the sampling program. In the Sulfate Process
segment of the Titanium Dioxide Subcategory, 18 different
sampling points were selected for studying the toxic pollutant
characteristics of the water supplies, the raw process waste
waters, and the plant effluent at two major manufacturing
facilities. For the inorganic constituents 575 analytical data
points were generated and an additional 1,824 data points were
obtained for the organic toxic pollutants excluding blanks and
duplicates for quality control.
The only organic toxic pollutant found during the screening
program was phenol which was observed at only one of the two
plants sampled. The maximum raw waste concentration of phenol
was 0.020 mg/1, however the raw water source for the plant
contained as much as 0.007 mg/1. This is well below the
treatability level for phenol, therefore, phenol is not
considered a significant or process related pollutant.
Daily raw waste loads were calculated from the flow rates
measured or estimated at the time of sampling and the measured
pollutant concentrations., That is,
Daily loading (as kg of pollutant per day) = (C)(Q)
1000
*
Where the concentration (C) of the pollutant is expressed
in units of mg/1 (Note: 1 kg/m3 = 1000 mg/1), and the flow
rate (Q) is expressed in units of m3/day (m3, a cubic
meter, is equal to 264 U.S. gallons).
Similarly, the unit loadings were calculated from the
reported Ti02 productions rate (P), the waste stream flow rate
(Q)/ and the measured pollutant concentration (C).
420
-------
TABLE 14-23. PROCESS WASTE WATER FLOW AT PLANTS # 555, #694 and #559
TITANIUM DIOXIDE (SULFATE PROCESS)
Plant
#555
#694
#559
Average
A
Strong
8.49
16
6.10
10
Flow in (m3/kkg
B
acid Weak acid
78.2 '
67
69
72
of Ti02)
C
Scrubber and
contact cooling
water
362
457
361
393
D = A
Total
449
540
436
475
+ B 4- C
Effluent
421
-------
Onit loading (as kg of pollutant
per day kkg of Ti02)
(C) (Q)
1000 (P)
Where C and Q are expressed in the same units described
above, and the production (P) is expressed in units of
kkg/day (kkg is 1000 kg, a metric ton, which is equal to
2205 Ib).
The maximum concentration of toxic pollutants found in the
raw waste at concentrations above the treatability level in the
screening and verification program were:•
Maximum Concentration Observed (ug/1)
Pollutant
Screening
(Plants f555 & 1559)
Verification
(Plant H559)
Cadmium
Chromium
Copper
Lead
Nickel
Zinc
Antimony
Arsenic
Thallium
Selenium
124
1
3
6
3
340
,000
,500
,700
,400
,800
20
11
19
360
31,
1,
5,
1,
17,
1,
12
000
000
200
300
000
400
340
41 ,
Below detection
limit
A summary of daily and unit (per unit of production) raw
waste loads for all plants sampled can be found in Table 14-24.
Individual plant raw waste loads and concentrations found in
sampling are given in Table 14-25.
Based on the total annual production of this industry and
the average waste load generated per unit product, the estimated
total toxic pollutant raw waste loads generated each year for
this subcategory are as follows:
422
-------
TABLE 14-24. SlMffiRY OF BAH WASTE LOADINGS FOUND IN SCREENING AND VERIFICATION SAMPLING
SIBCKIEQOKX THSWUM DIOXIDE - SUIESE PROCESS
Pollutant: Loading Range,
(kg/day)
Minimum Maximum
Toxic
it1- antimony 5.0 28
to
U)
flrsenic 1.9 4.0
Cadmium .068 7.2
Chromium 140 530
Copper 8.2 19
Lead 3.0 65
Nickel 3.7 23
Selenium 7 9.5
Thallium .47 1.2
Zinc 1.8 85
Conventional and Nonconventional
IBS
Icon, Pe
Minimum
.032
.012
.00044
1.1
.065
.024
.029
.0020
.0030
.014
Unit Loading,
(kg/Meg)
Average Maximum
0.11 0.22
0.19 .032
0.19 .057
2.0 3.4
.085 .12
.18 .42
.080 0.15
.031 .060
.0055 .0080
.34 .55
320
600
Ito. Of
Plants'1*
3
3
3
3
3
3
3
2
2
3
1
1
(1) - Data are taken only from those plants where pollutants were found above detection limits, or, in the
case of TSS and Iron, where data are available.
-------
14-25. "rescte wuumnsi AVERSCE raw K»BS AND
K)
SOBCffilEGOHX
t
TEEBflltM DIOXIDE - SULEKTE PROCESS
Screening
.antimony
.arsenic
Cactnium
Chroniun
Copper
Lead
Nickel
Selenium
Thalliun
Zinc
Cmj/1)
0.77
0.11
0.29
3.8
0.20
0.075
0.091
MA
m
0.088
Plant S555
(kg/kkg)
0.22
0.032
0.057
1.1
0.065
0.024
0.029
< 0.06
MA
0.014
Plant
0.16
0.029
0.0020
7.0
0.25
0.20
0.31
NA
0.020
1.1
1559
(kg/kkg)
0.080
0.014
0.0009
3.4
0.12
0.10
0.15
HA
0.0080
0.55
Verification
Plant
(rnq/1)
0.074
0.028
0.0010
3.1
11
0.96
0.14
0.0050
0.0070
1.04
1559
(kg/kkg)
0.032
0.012
0.00044
1.4
0.070
0.42
0.061
0.0020
0.0030
0.45
KB. = Not Available
-------
Pollutant Total Annual Raw
Waste Load (kg/year)
Cadmium 5,000
Chromium 510r000
Copper 22,000
Lead 47,000
Nickel 21,000
Zinc 88,000
Antimony 29,000
Arsenic 49,000
Selenium 8,000
Thallium 1,400
14.11 POLLUTION ABATEMENT OPTIONS
14.11.1 Toxic Pollutants of Concern
The toxic pollutants found above treatability levels in
this industry were evaluated on the basis of the maximum
concentration observed in the process raw waste waters. These
values are shown in Section 14.10.3. Using cadmium as an
example of a borderline case, its maximum observed concentration
of 0.34 mg/1 is considered significant because removal
efficiencies ranging from 70 to 97 percent could possibly be
achieved on the basis of the lower limits of treatability shown
in Table 8-11 for lime/settling, lime/filter, and sulfide/filter
technologies. The BAT utlimately selected as a basis for
regulations may not be as effective as the most advanced
technology considered at this stage of the evaluation of
alternatives.
The sampling data from this industry indicate that the
toxic pollutants of concern are chromium, zinc, nickel, lead,
copper, antimony, arsenic, and cadmium in decreasing order of
the amounts found. Selenium and thallium were detected at
levels too low to be treated effectively. The relative
pollutant concentrations and loadings in the raw waste largely
reflect the amounts of impurities in the ilmenite ore or titania
slag being processed. The major impurity found in the various
grades of raw material is ferrous iron as indicated in Table 14-
19. The toxic metal impurities would also be expected to occur
in a wide range of concentrations in the raw materials.
The advanced treatment technology options evaluated for
sulfate process segment of the industry were selected for their
ability to remove toxic metals of concern with greater
efficiency than the prevailing (BPT) practice which also removes
TSS, iron, and sulfate from the waste waters.
425
-------
14.11.2 Process Modifications and Technology Transfer Options
Specific process modification recommendations are not
made. However, several areas for further investigation suggest
themselves. They are:
1. One of the water borne wastes, the strong sulfuric acid
produced from the Ti02 sulfate process, has a sulfuric acid
concentration that varies from 15 to 30 percent as H2S04.
Currently, only a small portion of it is recycled. Research is
needed to find cost-effective ways to concentrate the acid to 90
percent and to eliminate the impurities (especially iron) so
that it can be reused in the digester. This will eliminate much
of the alkali requirements for neutralization and relieve
disposal problems associated with solid waste gypsum.
2. Economical methods need to be developed for the
recovery of iron oxide, aluminum, and vanadium from the waste to
the extent that markets are available for these materials.
3. If markets could be developed for the sale of ferrous
sulfate (copperas), solid waste disposal problems would be
reduced. Currently, a portion is sold and the rest disposed of
as a solid waste.
14.11.3 Best Management Practices
Storm water runoff from the plant site should be collected
and sent to the treatment facility for the removal of suspended
solids.
14.11.4 Prevailing Control and Treatment Practices
The treatment practices of the plants producing Ti02 by the
sulfate process is given in Sections 14.10.1 to 14.10.3.
14.11.5 Advanced Treatment Technologies
Although sulfide precipitation, the xanthate process, and
ion-exchange might be applied to the clarified solution obtained
by alkaline precipitation, oxidation and settling the cost
incurred are high because of the large quantity of water (more
than 400 m3/kkg of Ti02) that must be treated. The sulfate
process is one of two subcategories (the other being Soda Ash
Solvay Process) in the Inorganic Chemicals Industry studied in
this report that generates, the largest quantities of waste
effluent.
426
-------
14.12 SELECTION OF APPROPRIATE TECHNOLOGY AND EQUIPMENT
14.12.1 Technologies for Different Treatment Levels
Level 1 (BPT)
In the Level 1 treatment the blended strong and weak acid
streams are neutralized with calcium carbonate. The toxic
pollutants are precipitated and separated along with gypsum in
first stage thickeners. Aeration then oxidizes any ferrous iron
present and removes C02 before mixing with miscellaneous plant
waste containing minor amounts of heavy metal priority
pollutants. The combined stream is then given lime treatment of
pH 9 and settled in polishing lagoons before discharge. This
three-step system is patterned after existing systems which
separate the acid streams from miscellaneous wastes in order to
make possible the recovery of pure and impure gypsum from the
relatively consistent acid streams. Alkaline precipitation of
heavy metals, and significant removal of arsenic occur during
the last two-stages of lime neutralization, and settling of
precipitated toxic pollutants occurs in the final polishing
lagoons. Because waste flow rates are unusually high in the
sulfate process, long-term lagoon settling is more cost
effective than dual media filtration. The mechanical aeration
step used for oxidizing ferrous iron may contribute an important
mechanism for the simultaneous removal of other heavy metals
present very similar to the ferrite coprecipitation method
described in the, Treatment Technology Assessment section. The
flow diagram of the treatment system is shown in Figure 14-11.
Although the Model Plant does not include equipment for
gypsum recovery, it is based on separation of waste streams,
making pure or impure gypsum recovery possible by intercepting
thickener underflow(s). Recovery of gypsum as a saleable by-
product is not a viable option since no market appears to exist
at this time.
Level 2
Level 2 for the sulfate process employs the described BPT
treatment for strong acid, weak acid, and 55% of the "other
wastes". The remaining other wastes receive soda ash treatment
and settling, to permit recycling a nonscaling effluent for
scrubbers and miscellanous uses. Heavy metal pollutants in the
separated recycle stream are settled as carbonates and
periodically removed to a secure l;andfill. The flow diagram of
this treatment is shown in Figure 14-12.
427
-------
I
I
I
I
a
CO
rH
3
CD
428
-------
V£>
GROUND
LIMESTONE
CiCOj
C.C
-•"•"
"^
X.
S
,1
fin ri son*
Y LJ ASII
1
' u
1
»BC1tCl« %i I, 1|
(Misc. west V | jyr | y
/
1
L
«
45%
OTHER
WASTES
HAP1O MIX AND SETTUNC1
1
^p"V- —
X
nr
^Q-1
Oj SLUHHT
TANK
>»\ WEAK ACin/*~1pP1f
BilSTgOMfi AraliA— f\J^
t*
k
t f
JQ
IUI.T1PL
RECYCLED EFFLUENT
-fr. C08*0
ATMOSPIIEHE
__ --"
*
E
REACTORS
if
^—
-V
tIME
"^
*^-^
X_X" ^"*^i
I' I «l — - | FOMTMEU
1 -J—WI ^ | *
I***" §
MIXING I
| f _» A -»
1
n
i
Jy
1 AERATION * MIXING
X^ ^X^ OOZ REMOVAL
lit STAGE
TIBCKENER
f
k,/
.
'C")
1
i¥
i
roy
7
MIXING
TANK
I 2ND STAGE
I THICKENER .
i \
«*
ffi*J
I r
/ " EFFIAJEHT
FOUSIBHa
IAGOON
RECYCLE , SOUU.SW5TOSAI-
flow iRoaitorlag, pH tnonitdirlnf
Figure 14-12. Level 2 waste water treatment for titanium dioxide - sulfate process.
-------
14.12.2 Equipment for Different Treatment levels
Equipment Functions
Treatment of waste water from the sulfate process involves
the mechanized handling of large quantities of chemicals and
reaction products, primarily gypsum.
The BPT model includes rail car deliveries of ground
limestone and lime, bucket elevators, storage bins, multiple
feeders, mechanical feeders, mechanical aerators, and two-stage
thickening for removal of pure and iron-bearing gypsum for the
treated acid waste streams. Calcium saturated thickener
overflow and miscellaneous other waters are subjected to
alkaline precipitation and settled in a one-day polishing pond.
In Level 2, to reduce the mass discharge of heavy metals, only
55% of the BPT "other waste" flow joins the treated acid waste
stream, for BPT treatment as described above. However, the
remaining 45% of "other wastes" is given separate treatment with
soda ash settled in a lagoon, for recycle to miscellaneous
scrubber and noncontact cooling purposes. Treatment of the
strong and weak acid streams, including oxidation and settling
of ferrous iron, remains the same as in the BPT model.
Chemicals and Handling
First stage neutralization employs ground limestone, while
lime is used for second stage and final alkaline precipitation.
Oxygen is supplied from atmospheric air, and polymer is added to
assist in the second stage settling of iron hydroxide. Aside
from the bulk handling of large amounts of these common
chemicals, there are no special hazards involved in their use.
Separation and Removal of Solids
Large quantities of thickener underflow are pumped to
spreading areas for consolidation of the solids, which are later
pushed into 18 foot high piles on land provided for 10 years of
operation. Solids from occasional draining of the polishing
lagoon and the Level 2 recycling lagoon are returned to the
aeration step of the waste acid streams, after which they will
be settled out in the second stage thickener, being handled as
part of the thickener underflow. Although no dewatering
equipment is'provided, the first and second stage thickeners can
be sources of pure and' impure gypsum for future byproduct
recovery.
430
-------
14.13 TREATMENT COST ESTIMATES
14.13.1 General Discussion
To prepare treatment cost estimates, a model plant concept
was developed. For conceptual design a representative unit
waste flow was selected, together with three different Ti02
production rates. The latter were chosen to cover most of the
rates typical for the Ti02 subcategory (Sulfate Process). The
selected daily Ti02 production for the model plant was
multiplied by the selected unit flow to obtain the volume of
influent to the treatment system. The selected unit raw waste
pollutant loads were also multiplied by the model plant
production rate to determine the pollutant load on the treatment
system. Capital and equipment costs were then calculated based
on developed conceptual design parameters for each model plant
production rate. The rationale used for the model plant
selection is given below.
Waste Water Flew
Waste effluent data is available for three plants and is
given in Table 14-23. For the model the average value of the
three plant data has been used. The unit flow data for strong
acid ranges from 6.10 to 16 m3/kkg of Ti02. (Table 14-23). For
the model plant the average value of 10 m3/kkg has been used.
Unit flows for the weak acid stream range from 67 to 78 m3/kkg.
For the model plants, a unit flow of 72 m3/kkg of Ti02 is used,
The third segregated stream includes contact cooling water,
scrubber water, and finishing operation waste water. The unit
flow for this stream varies from plant to plant and depends
largely on the type and quality of the Ti02 pigment end product
desired. For model plants, a unit flow of 393 m3/kkg of Ti02 was
used. For model plants a total effluent flow which consists of
the strong acid, weak acid, and scrubber effluent, etcetra, of
475 m3/kkg of Ti02 was used.
Production
Five plants produce titanium dioxide by the sulfate process
at a total production rate of 259,000 metric tons per year.
Production ranges from a minimum of 31,000 kkg/yr to a maximum
of 74,500 kkg/yr with an average of 49,000 kkg/yr and a median
of 43,000 kkg/yr. For treatment cost estimates, three
production levels were selected. These were 31,800 kkg/yr;
47,700 kkg/yr, and 74,500 kkg/yr.
Waste Water Pollutant Load
As stated before, the principal pollutants occur in the
strong and weak acid streams and include high acidity (sulfuric
431
-------
acid)r suspended solids, iron and other heavy metal sulfates.
The other waste waters contain titanium dioxide and small
amounts of other heavy metals as suspended solids. Iron
concentrations vary depending on the grade of ilmenite ore used,
14.13.2 Model Plant Control and Treatment Costs
The average raw waste pollutant loadings given in Table 14-
23 were used for the model plant. For the model plants, a total
iron loading of 600 kg/kkg was used with the assumption that
two-thirds was suspended ferric hydroxide and one-third (200
kg/kkg of Ti02) was soluble ferrous iron. The unit sulfate and
suspended solid loadings for the different waste water streams
for the model plant were:
Sulfate Loading TSS Loading
Stream (kg/kkg of Ti02) (kg/kkg of Ti02)
Weak Acid 2,300 300
Strong Acid 1,800 200
Other Waste Water Negligible 113
Chemical Useage
In the model BPT system, powdered limestone is used for
first stage neutralization of mixed strong and weak acids, at
the unit rate of 3,000 kg/kkg of Ti02. Pebble lime (CaO) is used
for second stage neutralization of the mixed acid streams and
for the final neutralization of the total combined flow,,
including the other miscellaneous wastes. The unit application
of CaO for all purposes is 0.235 kg/kkg of Ti02. In Level 2
(which is not used as a regulation basis), soda ash is added to
45% of the "other waste" flow at an approximate dosage of 130
yg/1, to permit partial recycle for miscellaneous purposes.
Solids Produced
Although some existing plants have attempted to produce two
grades of saleable gypsum from the strong and weak acid streams,
at present there is not a sufficient market for gypsum to
justify byproduct gypsum recovery in the model plants. The
solids produced from the treatment facility consist of gypsum,
iron oxide, and the original suspended solids introduced in the
influent. The total solids produced in the model plant are
assumed to be 5,500 kg/kkg of Ti02.
Additional solids generated in the soda ash treatment of
Bother wastes" at Level 2 are only a few hundred pounds per day,
432
-------
and are considered a negligible increase in total solids
production. These additional solids are periodically
transferred from the recycle polishing ponds to the main
treatment system just ahead of the aeration step. In this way,
the additional quantity of toxic metals will be subjected to the
ferric iron flocculation, lime treatment, and settling sequence
in the BPT system.
The estimated costs for three models having different
production levels are given in Table 14-26, 14-27, and 14-28.
Annual treatment costs as a function of production are shown
graphically in Figure 14-13. Similarly, treatment cost per
metric ton of product is given in Figure 14-14.
Table 14-29 presents a summary of the unit cost distibution
between amortization and operation and maintenance cost
components at different productions and at the BPT and the Level
2 treatment.
For existing sources at the first level of treatment, the
disposal of sludge is on-site, hence land requirements are
fairly large. Amortization, chemicals, labor, and residual
waste disposal costs have significant impact on the annual
costs. The treatment Level 2 amortization, chemicals, and labor
constitute a major portion of the additional costs.
14.14 BASIS FOR REGULATIONS
14.14.1 Evaluation of BPT Practices
Out of a total of four Ti02 plants (sulfate process) that
are currently in operation, only one plant (|559) has a BPT
treatment system. The other 3 plants practice partial
neutralization and settling. The proposed BPT limitations are
based on available long-term data from plant |559.
Pollutant Removal with BPT Treatment
Treatment Level 1 is equivalent to the proposed BPT in the
Titanium Dioxide (sulfate process) industry. Means, standard
deviations, and variability factors were calculated from data
submitted by Plant §559 for final effluent quality, and the
results are given in Table 14-30. The performance
characteristics are utlized for the development of the proposed
BPT regulations.
The ability of the"treatment system to remove conventional,
nonconventional, and toxic pollutants was estimated by comparing
the treated effluent qualities with the raw waste qualities of
the sampled waste streams. The data for Plant #559 are given in
Table 14-31.
433
-------
TABLE 13-26. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE Sulfate
Production 31,800 metric tons per year (35,059 tons per year)
90 metric tons per day (100 tons per day)
Waste water flow 42750 cubic meters per day.
LEVEL OF TREATMENT*
FIRST SECOND
INVESTMENT COST
Construction $701,200 $117,500
Equipment in place,
including piping,
fittings, electrical
work and controls 2,328,400 233,000
Monitoring equipment
in place 9,000
Engineering design
and inspection 607,720 70,100
Incidentals, overhead,
fees, contingencies... 607,720 70,100
Land 1,272,000 12,000
TOTAL INVESTMENT COST $5,526,040 $502,700
B. OPERATION AND
MAINTENANCE COST
Labor and supervision. $504,000 $56,000
Energy 96,000 9,000
Chemicals 1,589,000 176,000
Maintenance 425,404 49,070
Taxes and insurance... 165,781 15,081
Residual waste
disposal 210,000
Monitoring, analysis
and reporting 15,000 7,500
TOTAL OPERATION AND
MAINTENANCE COST $3,005,185 $312,651
C. AMORTIZATION OF
INVESTMENT COST $692,132 $79,836
TOTAL ANNUAL COST $3,697,317 $392,487
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
434
-------
TABLE 14-27. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE Sulfate
Production 47,700 metric tons per year (52,589 tons per year)
136 metric tons per day (150 tons per day)
Waste water flow 64600 cubic meters per day.
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT COST
Construction
Equipment in place,
including piping,
fittings, electrical
work and controls
Monitoring equipment
in place
Engineering design
and inspection
Incidentals, overhead,
fees, contingencies...
Land
TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Energy
Chemicals
Maintenance
Taxes and insurance...
Residual waste
disposal
Monitoring, analysis
and reporting
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
$958,700
2,980,200
9,000
789,580
789,580
1,920,000
$7,447,060
$672,000
138,000
2,384,000
552,706
223,411
315,000
15,000
$4,300,117
$899,252
$5,199,369
$161,000
278,000
87,800
87,800
18,000
$632,600
$56,000
12,000
265,000
61,460
18,978
7,500
$420,938
$99,995
$520,933
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
435
-------
TABLE 14-28. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE Sulfate
Production 74,500 metric tons per year
212 metric tons per day
Waste water flow 100700. cubic meters per day.
(82,136 tons per year)
(234 tons per day)
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT COST
Construction
Equipment in place,
including piping,
fittings, electrical
wark and controls
Monitoring equipment
in place
Engineering design
and inspection.
Incidentals, overhead,
fees, conting enc ies...
Land
TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Energy.
Chemicals
Maintenance
Taxes and insurance...
Residual waste
disposal
Monitoring, analysis
and reporting
TOTAL OPERATION AND
MAD1TENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
$1,293,500
3,914,500
9,000
1,043,400
1,043,400
2,940,000
$10,243,800
$672,000
199,000
3,719,000
730,380
307,314
420,000
15,000
$6,062,694
$1,188,328
$7,251,022
$208,000
322,000
106,000
106,000
24,000
$766,000
$56,000
18<000
412,000
74,200
22,980
7,500
$590,680
$120,723
$711,403
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
436
-------
j L
1 i
; L.EVEL, f2 j
XT i i x
i i
i ! I
I I
\ \
' yf \,
II
I i I
20
30 40 50
EJOTJCTICH (MESOK:
60 70
x 1000)
80
Figure 14-13. Jtonual txeatanent cost vs. production for the titanium dioxide
subcategory, sulfate process.
437
-------
130
1ZO
no
100
90
z
( 1
1 1 '
' '
'
i ! •
• l I f
t
t
. * '
: • ; :
t ' ,
I . j
i ;
' ! ' !
i ' '
• 1
j
l
;
i
t
1
i
i
i
*
i , , >
;
® i ' •
•\
\ •
\
\
:
1
• ' i
fi
1 i
^ •
X '
x
N
.
1 •
• i . :
, ;
• ; l i
X' ' '
X • •
X
; ; \_
? '^X
, i . i N
i
;
i
!
X
X
X^i
"X
t
!
; i
!
i ' ,
i
1 !
, ;
l
• i
i i
\i !
j\
iXi
X
\
X
V
IX
1
X
^
r
t
i
L> JU 4U SU b
,
1 ,
' , ' ,
! '
1 1
I !
; | ,
! i i
!
! i !
!
! ,
i 1 ,
;
I
! i !
! ! !
;
i
*s
i
X
X
X
I
i >
X!
x
^
-
.
S^
i
1 i
1
, i t
!
• 1
: i •
' I
i i
i ', .
'
i
1
1
J
1
I i
| , '
|
1
1
!
1
1
1
!
i
. l
j i
SJ 1LEVEJ
N -
i _®i i
Sy
,
IwEVEJ
\fflkl
!
i
4
u 70 a
•
1
' '
!
1 j 1 <
i i :
1 I
1 ' ' '.
1 • I
.
i !
1 1
'
j
1
111.
i : i
i
I
i
j
i
!
1
i i
!
,
;
!
• i
i
1
1
•
^ rfi
1
!
!
1
1
0
(METRIC TCNS/YERR x 1000)
Figure 14-14. Amual unit treatment cost vs. production for the titanium dioxide
subcategory, sulfate prooass.
438
-------
TABLE 14-29. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE Sulfate
Annual Treatment Costs ($/kkg)
PRODUCTION FLOW
(kkg/yr) (m3/day)
LEVEL OF TREATMENT
FIRST SECOND THIRD FOURTH
Annual Operation
and Maintenance
Annual
Amortization
Total Cost
31,800 61,600
47,700 92,600
74,500 144,000
31,800 61,600
47,700 92,600
74,500 144,000
31,800 61,600
47,700 92,600
74,500 144,000
94.50
90.15
81.38
21.77
18.85
15.95
116.27
109.00
97.33
9.83
8.82
7.93
2.51
2.10
1.62
12.34
10.92
9.55
Not Applicable
439
-------
TABLE 14-30. KESTORICMi EEFKM3T I-OCCTORING
SUBCMBGOKr - T33SNIUM DIOXIDE
snr.TWFE BRQCESS W3NT 4559
Pollutant
TSS Cadmium Chromium
Daily Data
No. of Boints 899 109 128
Average, x 21.0 0.060 0.070
nof-f^ cd) 65.93 0.044 0.054
§ Standard (2) 1.54 0.68 0.67
Deviation, S1
XSfity<3> *•» . 3-85 , 3.a
30-Day Average
No. of Boints 30 26. . ,30
Standard 21 84 0 042 0 038
Deviation , 1*B4 O'°*z °'038
Variability , „ . 2 ._ _ fl.
Factor
Variability
Bactxar Ratio
VFR(5) 3.62 1.58 1.87
Iron .Lead . Nickel Zinc
854 128 128 128
0.62 0.068 0.08 0.151
3.46 0.041 0.071 0.204
1.86 0.56 0.76 1.02
13.65 3.16 4.39 • 6.41
28 30 30 30
0.94 0.04 0.048 0.16
4.00 2.14 4.39 3.05
3.38 1.48 1.00 2.10
(Continued)
-------
TSBLE 14-30. Continued
(1) S is the arithmetic standard deviation and is given by
»-,/-?
n-1
vtere xi is the data value for point i
x is the mean value
n is the number of data points
(2) S1 is the estimated standard deviation
S1 =\/ ln(l 4
iwhere S is .the arithmetic standard deviation
x~ is the mean value
(3) Ite variability factor (W) of daily measurements for lognormal distribution
is found by the expression,
- In (W) = S1 (2- 0.5 S')
•where S1 is the estimated standard deviation
Z- 2.33 for 99th percentile
(Continued)
-------
TfiBIE 14-30. Continued
(4) The variability factor (W) for 30-day average raeasuretnents is found by the
expression
W = 1.0 + Z
8
TOiere x is the mean value
S is the arithmetic standard deviation
Z = 1.64.for 95th pereentile
(5) WR: Ratio of the 24-hour variability factor to the 30-day variability factor
-------
14-31. VEEOMEATION -RESULTS - SULFATE
DIOXIDE #559
•"
Bollutant
Total Suspended
Solids
Iron
Cadmium
Chromium
Copper
Lead
Nickel
Zinc
Arsenic
Antimony
Selenium
Thallium
Raw
Waste Treated Effluent
A BCD
Unit Load Concentration Unit Load Concentration
(kg/kkg) (mg/1) (kg/kkg^ fag/1)
116
364
0.00045
1.3
0.070
0.040
0.060
0.45
0.012
0.030
0.0020
0.0030
266 10.0 - 23
835 1.92 4.4
0.0010 0.000040 0.00010
3.1 0.011 0.025
0.16 0.002 0.0050
0.96 0.00090 0.0020
0.14 0.0020 0.0050
1.0 0.030 0.062
0.028 0.0040 0.010
0.074 0.0060 0.015
0.0050 0.0020 0.005-0
0.0070 0.0010 0.0030
E
Rstioval
Efficiency
(%)
91
99
90
99
97
99
96
94
64
80 "
' 0
60
443
-------
14.14.2 Basis for Proposed BPT Effluent Limitations
Technology Basis
For BPT, the Agency is proposing limitations based on
equalization, limestone neutralization, clarification,
aeration, alkaline precipitation and settling followed by pH
adjustment before final discharge of the effluent. This
technology is chosen because it has been installed and operated
successfully by a plant in the industry.
Flow Basis
Waste flow data is available for 3 plants and the average
value of 475 m3/kkg of Ti02 (Table 14-23) is taken as the inflow
for the model plant treatment system. The treatment plant
effluent is taken to be the same as the influent and the loss or
addition of water through chemicals, evaporation,
precipitation, and through solid removal have been neglected.
Selection of Pollutants to be Regulated
The selection of pollutants for which limitations are
proposed is based on the evalutation of raw waste data from the
screening and verification program. The following two major
factors were considered:
Raw waste pollutant concentration - Initially one plant was
visted and the waste effluent sampled in the screening phase.
The discovery of any toxic pollutants in the raw waste above the
detection level and in the treatability range was the basis for
proceeding with verification sampling. The presence of the same
pollutants in the verification phase confirmed the significance
of the screening program observation. The pollutants found in
signifcant levels in the raw waste of the plants sampled in
screening and verification are given in Section 14.10.3.
On the basis of maximum concentration of pollutants in the
raw waste the preliminary selection of candidates for regulation
included chromium, cadmium, copper, lead, nickel, zinc, arsenic,
and antimony.
Total subcategory raw waste pollutant loading - The average
unit toxic pollutant loadings of the raw waste of plants sampled
in the screening and verification program (Table 14-23) was
multiplied by the total Ti02 production rate of 259,000 kkg/year
to yield an estimate of the total annual pollutant loading for
the subcategory (see Section 14.10.3). The data give an
indication of the overall magnitude of the pollution potential
for the subcategory.
444
-------
Basis of Pollutant Limitations
Conventional and nonconventional parameters
A. The treated effluent is to be controlled within the
range of pH 6.0 to 9.0. This limitation is based on the data
presented in Appendix B of this report and the JRB Study (52).
B. TSS and Iron: The long-term average values of 21 mg/1
for TSS and 0.62 mg/1 for iron derived from the monitoring data
of Plant £559 (Table 14-30) was used as the proposed subcategory
performance values. The variability factors for daily and 30-
day average estimated from Plant 4'559 long-term data (Table 14-
30) were used in calculating the concentration basis and
effluent limitations as shown below.
Total Suspended Solids
The proposed TSS maximum 30-day average concentration is
given by:
(21 mg/1) (3.04) = 64 ing/1
The proposed TSS 24-hour maximum by
(21 mg/1)(11.0) « 230 mg/1
The proposed TSS maximum 30-day average effluent limit was
obtained by using the model plant unit flow of 475 m3/kkg,
namely
(64 mg/1)(475 m3/kkg) / kg/m3 \ = 30 kg of TSS
UOOO mg/1/ kkg of Ti02
from:
The proposed iron 24-hour maximum effluent limit:
(230 mg/1)(475 m3/kkg) / kg/m3 \ » 110 kg of TSS
\1000 mg/1/ kkg of Ti02
The proposed iron maximum 30-day average concentration:
= (0.62 mg/1) (4.0) = 2.5 mg/1
The proposed iron 24-hour maximum concentration:
= (0.62 mg/1) (13.65) = 8.5 mg/1
The proposed iron maximum 30-day average effluent limit?
(2.5 mg/1)(475 m3/kkg) / kg/m3 \ - 1.2 kg of iron
\1000 mg/1/ kkg of Ti02
445
-------
The proposed iron 24-hour maximum effluent limit
= (8.5 mg/1)(475 m3/kkg) / kg/m3 N - 4.1 kg of iron
\1000 mg/1/ kkg of Ti02
Toxic Pollutants - The effluent limitations proposed for
the selected toxic pollutant parameters are derived from two
sources of information. These are 1} long-term monitoring data
for Plant |559, 2) literature-based treatability estimates.
If the long-term data of a certain pollutant was not
available or the 30-day average obtained from the long-term data
was less than the lower level of treatability values, then the
lower limit of treatability was used as the concentration basis
for the maximum 30-day average limitation. The long-term data
of most of the toxic pollutants for Plant #559 are given in
Table 14-30.
A. Antimony: The maximum concentration of antimony
observed in the raw waste during the screening and verification
program was 1.4 mg/1 (shown as 1400 yg/1 in Section 14.10.4).
At Plant 1-559, 80 percent of the antimony is removed during
treatment (Table 14-31). The proposed 30-day average
concentration of 0.8 mg/1 is based on the lower limit of
treatability as determined by literature studies (Table 8-11).
A variability factor ratio of 1.9 (ratio of 24-hour variability
factor to the 30-day variability factor) determined for chromium
(Table 14-30) from the long-term data for Plant $559 was used to
obtain the 24-hour maximum concentration. Thus:
The proposed antimony 24-hour maximum concentration is
given by;
(0.80 mg/1)(1.9) = 1.5 mg/1
The proposed antimony 30-day average effluent limit is
given by:
(0.80 mg/1) (475 m3/kkg) / k'g/m3 \ = 0.38 kg of antimony
UOOO mg/1 / kkg of Ti02
The proposed antimony 24-hour maximum effluent limit is
given by:
(1.5 mg/1)(475 m3/kkg) / kg/m3 \ = 0.71 kg of antimony
VLOQO mg/1/ kkg of Ti02
B. Cadmium: The maximum concentration of cadmium found in
the raw wastes during the screening and verification program was
0.340 mg/1 (shown as 340 ng/1 in Section 14.10.4). The data for
Plant f559 indicated a removal efficiency of 90.0 percent (Table
446
-------
14-31). Thus, the long-term average value of 0.060 mg/1 and the
variability factor of 3.85 for daily maximum and 2.43 for 30-day
average estimated from the long-term monitoring data of Plant
1559 (Table 14-30) were used in calculating the proposed
concentrations and effluent limitations as shown below:
The proposed cadmium 30-day average concentration is given
by:
(0.060 mg/1) (2.43) = 0.15 mg/1
The proposed cadmium 24-hour maximum concentration is given
by:
(0.060 mg/1) (3.85) = 0,24 mg/1
The proposed cadmium 30-day effluent limit is given by:
= (0.15 mg/1)(475 m3/kkg) / kg/m3 \= 0.070 kg of cadmium
\1000 mg/1/ kkg of Ti02
The proposed cadmium 24-hour maximum effluent limit is
given by:
(0.24 mg/1)(475 m3/kkg)/ kg/m3 \ = 0.11 kg of cadmium
\1000 mg/1/ kkg of Ti02~
C. Chromium: The proposed subcategory limitation of 0.070
mg/1 is based on the average of the long-term monitoring data
for Plant f559 given in Table 14-30. The variability factor of
3.81 for the daily data and the variability factor of 2.04 for
30-day averages were estimated from the same data for Plant
If559r and an established model plant unit flow of 475 m3/kkg was
used in setting up the proposed limitations.
The proposed chromium maximum 30-day average concentration
is given by:
(0.070 mg/1)(2.04) = 0.14 mg/1
The proposed chromium 24-hour maximum concentration is
given by:
(0.070 mg/1) (3.81) = 0.27 mg/1
The proposed chromium 30-day average effluent limit is
given by:
(0.14 mg/1)(475 m3/kkg) / kg/m3 \ = 0.070 kg of chromium
VlOOO mg/1/ kkg of Ti02
447
-------
The proposed chromium 24-hour maximum effluent limit is
given by:
(0.27 mg/1)(475 m3/kkg) / kg/m3 \ = 0.13 kg of chromium
V1000 mg/1/ kkg of Ti02
D. Copper: The value of 0.5 mg/lf which is the lower
limit achieved from the lime-settling of copper contaminated
waste water from the treatability studies (Table 8-11) was
selected as•the proposed maximum 30-day average concentration
because no long-term data for copper is available. The
variability factor ratio of 1.87 developed from the long-term
data for Plant f559 for chromium (Table 14-30) was used to
estimate the 24-hour maximum concentration because performance
of the treatment system is expected to be the same for copper as
for chromium . The calculations for the proposed concentrations
and effluent limits are given below:
The proposed copper 24-hour maximum concentration is given
by
(0.50 mg/1)(1,87) = 0.95 mg/1
The proposed copper 30-day average effluent limit is given
by:
(0.50 mg/1)(475 m3/kkg) f kg/m3 ^ = 0.24 kg of copper
\1000 mg/1/ kkg of Ti02
The proposed copper 24-hour maximum effluent limit is given
by:
(0.95 mg/1)(475 m3/kkg) / kg/m3 N = 0.46 kg of copper
VLOOO mg/1/ .kkg of Ti02
E. Lead: The lowest concentration of lead achievable by
treatment as determined by treatability studies (value of 0.30
mg/1 from Table 8-11 for lime-settling) was selected as the
proposed maximum 30-day average concentration. The higher value
was selected because if lead is present in large quantity this
represents the achievable level. The selected value is higher
than the value obtained by multiplying the long-term average of
.070 mg/1 by the 30-day variability factor of 2.14 estimated
from the monitoring data for Plant f559 (Table 14-30) . The
variability factor ratio of 1.48 obtained from the long-term
monitoring of lead for Plant f-559 (Table 14-30) was used in
calculating the 24-hour maximum concentration. The calculations
used to establish the proposed concentrations and effluent
limitations are shown below:
448
-------
The proposed lead 24-hour maximum concentration is given
by:
(0.30 mg/1) (1.5) = 0.45 mg/1
The proposed lead 30-day average effluent limit is given
by:
(0.30 mg/1) (475 m3/kkg) / kg/m3 \ = 0.14 kg of lead
\1000 mg/1/ kkg of T5.02
The proposed 24-hour maximum effluent limit is given by:
(0.44 mg/1) (475 m3/kkg) / kg/m3 \ = 0.21 kg of lead
\1000 mg/1/ kkg of Ti02
F. Nickel: The proposed 30-day average concentration of
0.20 mg/1 is based on the lower limit established by
treatability studies and achieved using lime treatment and
settling (Table 8-11). The' proposed daily maximum concentration
was estimated by multiplying the 30-day average concentration by
the variability factor ratio of 1.87 developed for chromium from
the long-term data for Plant 1559 (Table 14-30) . The variability
factor for chromium was used because the treatment system is
expected to perform similarly for nickel and chromium. The
calculations for the proposed concentrations and effluent limits
are given below:
The proposed nickel 24-hour maximum concentration is given
by:
(0.20 mg/H (1.87) = 0.37 mg/1
The proposed nickel maximum 30-day average effluent limit
is given by:
(0.20 mg/1) (475 m3/kkg) / kg/m3 \ = 0.10 kg of nickel
\1000 mg/1/ kkg of Ti02
The proposed nickel 24-hour maximum effluent limit is given
by:
(0.37 mg/1) (475 m3/kkg) / kg/m3 \ = 0.18 kg of nickel
UOOO mg/1/ kkg of Ti02
G. Zinc: The lower limit established by treatability
studies, namely 0.5 mg/1, (Table 8-11) was used as the basis' for
the proposed 30-day average concentration limit because the
observed average effluent concentration (Table 14-31) was
considerably less. The variability factor ratio of 2.1 developed
from the long-term data for Plant 1559 for lead (Table
449
-------
14-30) was used to estimate the proposed daily maximum
concentration since similar performance with this treatment
technology is expected. The calculations used to establish the
concentration basis and effluent limitations are shown below:
The proposed zinc daily maximum concentration is given by:
(0.50 mg/1)(2.1) = 1.1 mg/1
The proposed zinc 30-day average effluent limit is given
by:
(0.50 mg/1)(475 m3/kkg) / kg/m3 \ = 0.24 kg of zinc
\1000 mg/1/ kkg of Ti02
The proposed zinc 24-hour maximum effluent limit is given
by:
(1.1 mg/1)(475 m3/kkg) / kg/m3 N = 0.52 kg of zinc
\1000 mg/1/ kkg of Ti02
H. Arsenic: The proposed 30-day average concentration of
0.5 mg/1 is based on the lower limit established by treatability
studies for lime precipitation and settling (Table 8-11) because
no long-term data for arsenic treatment is available. The
proposed daily maximum concentration was estimated by
multiplying the 30-day average concentration with a variability
factor ratio of 1.9 developed for chromium from the long-term
monitoring data for Plant f559. The calculations for the
proposed concentrations and effluent limits are given below:
The proposed arsenic 24-hour maximum concentration is given
by:
(0.50 mg/1) (1.9) = 0.95 mg/1
The proposed arsenic 30-day average effluent limit is given
by:
(0.50 mg/)(475 m3/kkg) / kg/m3 N = 0.24 kg of arsenic
\1000 mg/1/ kkg of Ti02
The proposed arsenic 24-hour maximum effluent limit is
given by:
(0.95 mg/1) (475 m3/kkg) / kg/m3 N = 0.46 kg of arsenic
UOOO mg/1/ kkg of Ti02
Summary - A summary of the proposed conventional,
nonconventional, and toxic pollutant limitations for BPT are
given in Table 14-32.
450
-------
TABLE 14-32. BROWSED LIMITATIONS
THANHM DIOXIDE SULFATE HffiCESS
Best Practical Control Technology Currently Available
Waste ite-iter Flow. 475 m3/kkg of TiO2
Pollutant
Conventional and
Nonconventional
Pollutants
Total Suspended
Solids
Iron
Toxic Pollutants
Antimony
Cadmium
Chrcroium
Copper
Lead
Nickel
Zinc
Arsenic
Sufccategory «,
Performance WR
(rag/1)
21 (2) 3'6
0.62(2) 3.4
0.80(3) 1.9(4)
o.06(2) 1.6
0.07(2) 1.9
0.50<3) 1.9(4)
0.30(3> 1.5<5>
0.20(3) 1.9(4)
0.50(3> 2.1(5)
0.50(3) 1.9
Consentration
Basis
I lax
30-day 24-hr.
Avg Max
64
2.5
0.80
0.15
0.14
0.50
0.30
0.20
0.50
0.50
230
8.5
1.5
0.24
0.27
0.95
0.45
0.37
1.1
0.95
Effluent
Limit
(kg/kka of TiO,)
Max
30-day 24-hr,
Avg Max
30
1.2
0.38
0.070
0.070
0.24
0.14
0.10
0.24
0.24
no
4.1
0.71
0.11
0.13
0.46
0.21
0.18
0.50
0.46
(1) WR: Ratio of the 24-hour •variability factor to the 30-day variability
factor.
(2) Long-term average based on loading data and variability factors of
plant #559 selected from Table 14-30.
(3) The lower limit of the literature treatability estimate (Table 8-11)
is used as the basis for the 30-day average limitation.
(4) Variability factor ratio of chromium developed from the long-term data
of plant #559 has been used (Table 14-30).
(5) Variability factor ratio estimated for this pollutant from long-term
data of plant 1559 has been used.
451
-------
14.14.3 Basis for Proposed BCT Effluent Limitations
The BCT limitation (applicable only to TSS) was set equal
to BPT because BAT is equal to BPT.
14.14.4 Basis for Proposed BAT Effluent Limitations
For BAT, the Agency is proposing limitations based on
treatment consisting of Level 1 technology, and are the same as
BPT. A treatment system requiring 55 percent recycle through
use of soda ash precipitation was considered but rejected
because its performance has not been demonstrated. The
limitations proposed for BAT are given in Table 14-33.
14.14.5 Basis for Proposed Hew Source Performance Standards
Level 1 treatment technology (also proposed for BPT and
BAT) is selected as the basis for NSPS limitations. A treatment
system requiring 55 percent recycle through use of soda ash
precipitation was considered but rejected because its
performance has not been demonstrated. Compared to BAT, NSPS
additionally limits pH, TSS and iron. The proposed NSPS
limitations are given in Table 14-34.
14.14.6 Basis for Proposed Pretreatment Standards
Existing Sources
For pretreatment standards for Existing Sources (PSES), the
Agency is proposing limitations based on BAT. The pollutants to
be limited are iron, antimony, cadmium, chromium, copper, lead,
nickel and zinc as indicated in Table 14-34.
New Sources
For pretreatment standards for New Sources (PSNS), the
Agency is proposing limitations based on NSPS. The pollutants
to be regulated are iron, antimony, cadmium, chromium, copper,
lead, nickel and zinc as indicated in Table 14-34.
452
-------
14-33. PEDPOSHD LMITATtCNS
' DICKIEE
Best Available Technology
Waste Water Plow: 475 m3/kkg of TiO0
(2)
Subcategory ,,» WET Concentration
Pollutant Perfcomnce^ Basis (mg/1)
Nonconventional
Pollutants
iron (3)
Toxic Pollutants
Antimony
Cadmium
Chromium
Copper (3)
Lead*3)
Nickel (3)
ZJno(3)
Arsenic
0.62
0.80
0.060
0.070
0.50'
0.30
0.20
0.50
0.50
3.4
1.9
1.6
1.9
1.9
1.5
1.9
2.1
1.9
3u-3ay
Avg
2.5
0.80
0.15
0.14
0.5Q
0.30
0.20
0.50
0.50
24- hour
Max
8.5
1.5'
0.24
0.27
0.95
0.45
0.37
1.1
0.95
Effluent Limit
(kg/kkg of TiC>2)
30-ctay
Avg
1.2
0.38
0.070
0.070
0.24
0.14
0.10
0.24
0.24
24 -hour
Max
4.1
0.71
0.11
0.13
0.46
0.21
0.18
0.52
0.46
(1) Proposed Lainitations for BPT Table 14-32
(2) VtR: Ratio of the 24-tour -variability factor to the 30-day variability
factor.
(3) Also applicable for PSES and PSJS limitations.
453
-------
Dollutant
Conventional and
TABLE 14-34. PROPOSED LIMITATIONS
TITANIUM DIOXIDE SULFATE PROCESS
New Source Performance Standards
ftkste Tflater Flow: 475 m3/kkg of TiO.
Subcategory
Performance
(1) VFR
(2) Concentration Effluent Limit
Basis, (mg/1) (kg/kkg of TiO,)
Max - Max ^
30-day 24-hour 30-day 24-hour
Avg Max Avg Max
Nonconventional
Pollutants
Total Suspended
Solids
Iron
Toxic Pollutants
Antimony
CadmivHn
Chromium
Copper
Lead
Nickel
Zinc
Arsenic
21
0.62
0.80
0.060
0.070
0.50
0.30
0.20
0.50
0.50
3.6
3.4
1.9
1.6
1.9
1.9
1.5
1.9
2.1
1.9
64
2.5
0.80
0.15
0.14
0.50
0.30
0.20
0.50
0.50
230
8.5
1.5
0.24
0.27
0.95
0.45
0.37
1.1
0.95
30
1.2
0.38
0.070
0.070
0.24
0.14
0.10
0.24
0.24
110
4.1
0.71
0.11
0.13
0.45
0.21
0.18
0.52
0.46
(1) For basis, see proposed limitation for BPT T-able-32,
(2) VFR: Ratio of the 24-hour variability factor to the 30-day variability
factor.
454
-------
14.15 TITANIUM DIOXIDE - CHLORIDE ILMENITE PROCESS INDUSTRY
PROFUSE
14.15.1 General Description
Total subcategory production capacity is given in Table
14-35 Profile Data Summary. The 308 data available for the
Ti02 Subcategory does not adequately cover the one-step
chloride ilmenite process; however, supplementary information
has been submitted by industry (55) . The status of
regulations is presented in Table 14-2. Additional
information on the chloride process industry is given in
Section 14.1.
14.15.2 General. Process Description and Raw Materials
For the manufacture of titanium dioxide by the combined ore
benefication-chloride process, a generalized process flow
diagram including the waste streams is shown in Figure 14-15.
The direct use of ilmenite ore for the manufacture of
titanium dioxide pigments requires the application of either
the sulfate process or the one-step ore
beneficiation/chlorination process which is referred to in
this report as the chloride-ilmenite process. Processes
which involve a separate ore beneficiation step (either at
the plant or at the ore source) resulting in an upgraded or a
synthetic rutile product to be used as feed material for a
chloride process would not be classified as a chloride-
ilmenite process. A separate ore beneficiation process would
fall within the Ore Mining and Dressing Category for
regulatory purposes, and the manufacture of TiO? from an
upgraded ilmenite or synthetic rutile would be in the same
classification as a chloride process using natural rutile ore.
The central feature of the chloride-ilmenite process is
a fluidized bed reactor, referred to as the chlorinator,
which receives the ore, coke, and chlorine. For any given
ilmenite ore composition, the differential rates of the
various metal chlori'nation reactions taking place
simultaneously in the chlorinator make impossible any clear
distinction between ore beneficiation and titanium
tetrachloride formation steps. The reaction mixture
composition is further complicated by recycling recovered ore
from the quench tower back to the chlorinator. Thus, the
wastes generated by the process are not separable into
beneficiation wastes and chlorination wastes. The chlorinator
acts as the primary source of concentrated acidic wastes
which are collected for treatment and disposal from the ore
recovery and gas scrubber units.
455
-------
EfflLE 14-35
.SOBCKCEGQRY PROFILE DATA SUMMARY
SOBCATEGORY
TZEKNIDM DIOXIDE Chloride Process CLlmenite Ore)
Total subcategory capacity rate
Total subcategory production rate
Stariber of plants in this subcategory
308 Data on file for
With total capacity of
With total production of
Representing capacity
Representing production
Plant production range;
Minimum
maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
m.
522,775 kkg/year
4
(1)
495,500 kkg/year
NA
95 percent
Unknown; wide variation
in production
75 kkg/year
228 kkg/year
151.50 kkg/year
NA.
NA
NA
8400 cubic meters/day
42,000 cubic meters/day
29 cubic meters/kkg
140 cubic meters/kkg
(1) Capacity included in Table 14-1.
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Commerce, Current Industrial
Reports, December 1977? Energy and Environmental Analysis, Inc., Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chemical Industry," June, 1978, and "Economic Analysis of
Proposed Revised Effluent Guidelines and Standards for the Inorganic
Chemicals Industry ," March, 1980.
- not available
456
-------
Figure 14-15. General process flow diagram of the titanium tetrachloride portion
of a titanium dioxide plant using the chloride-ilrnenitje'process.
45?
-------
Once the T1C14 intermediate has been isolated and
purified, the production of Ti02 is basically the same as
described in Section 14.1.2 for the ordinary chloride process.
14.16 WATER USE AND WASTE SOURCE CHARACTERISTICS
14.16.1 Water Use
Water is used in the manufacture of titanium dioxide by
the chloride-ilmenite process for noncontact cooling,
process reactions, air emission control, product treatment,
washing and transport operations. Table 14-36 presents a
summary of water use data for three plants.
14.16.2 Waste Sources
The concentrated process waste stream generated by the
chloride-ilmenite process contains the HC1 generated in the
chlorination process along with iron and other metal chlorides
in solution. The waste stream also carries the spent coke
and unreacted ore solids in suspension (TSS) .
The other major sources of process contact waste water
are combined in the dilute process waste stream. These wastes
are generated in the product finishing operations which
include the application of surface coatings (usually alumina
or silica) to the titanium dioxide pigment particles, and the
final dewatering, washing, drying, and sizing of the product.
The application of surface coatings requires the use of acid
and alkali to maintain the proper pH range for chemical
treatment of the Ti02 slurry. The resulting salts of
neutralization are washed from the product. These dilute
acid wastes are high in total dissolved solids and contain
suspended Ti02 from the finishing operations.
Table 14-37 summarizes the average raw waste loads
carried by the concentrated and dilute process waste streams
at three plants.
In Table 14-36, considerable differences in water usage
are revealed among the three plants. These differences are
largely a reflection of plant, age in the sense that the
feasibility and economics of effective contact/noncontact
waste water segregation and recycling are highly dependent on
the original plant design and facilities layout. Obviously,
Plant $713 is a new plant which incorporates modern concepts
of water use and waste handling practices and is therefore
used as the basis for the chloride-ilmenite NSPS. The high
flow plant, Plant #237, • is an older, existing facility in
458
-------
TABLE 14-36. AVERAGE "HATER USAGE FOR TiO~ PRODDCTION
BY THE CHLORIDE -HMENTTE PfiOCESS
*
Use Plant #237
*
Nbncontact 73-140
Cooling
Process Contact 100-140
and Clearup
Noncontact 9- 11
Jtacillary Uses
(Boilers,
Sanitary, etc.)
Plant #550
om /kkg of TiOj
330-390
47- 59
6- 7
Plant #713
15-16
29-33 (2)
5- 6
Source of data, (55).
(1) The average total flow of 120 m /kkg is used as the basis for BPT.
3
(2) The average flow of 31m /kkg is used as the basis for NSPS.
459
-------
TKELE 14-37. AVERAGE MW MSCE LORDS FCR Ti02 ERODUCTION
BY THE -
ot
o
Plant
#237
Cone. Process Dil. Process
Stream Stream
(kg/kkg of Ti02)
TSS
HC1
Fed3
Other metal
chlorides
100-150
200-230
900-1150
140-155
20-35
8-10
1-3
Negl.
Plant #550
Gone. Process
Stream
(kg/kkg
150-200
250-300
1000-1200
190-210
Dil. Process
Stream
of Ti02)
15-20
0,5-0.8
2-3
Negl.
Plant
#173*
Cone. Process Dil. Process
Stream Stream
.(kg/kkg of TiO2)
200-240
120-240
1000-1200
120-150
5-20
Negl.
Negl.
Negl.
* "Ehese mLues are estimated for a new plant prior to start-up.
Negl. -Negligible (< 0.5)
-------
which process contact water usage is by far the highest of the
three plants. Although some reductions in the volume and the
relative proportion of contact water usage may be feasible, the
economic incentives are lacking and it is unlikely that
older plants will be extensively modified to improve water
use patterns alone. This segment was not further subdivided
because the basic process is the same.
14.17 DESCRIPTION OF PLANTS VISITED AND SAMPLED
14.17.1 Screening
Plant f'55Q was visited during the screening phase of the
sampling program. This plant is ._ capable of producing
titanium dioxide from ilmenite ore by means of a one-step,
integrated beneficiation/ chlorination process. However, at
the time of the sampling visit, the plant was not using
ilmenite ore, but rather an upgraded ore which was similar in
quality to rutile. For this reason, the sampling results
cannot be considered representative of a chloride-ilmenite
process and are not presented in this report.
Plant $550 disposes of its concentrated acid waste by
deep well injection. These ferric chloride laden acidic
wastes are collected first in a system of four settling ponds
where the bulk of the solids are removed. Dredging of the ponds
is a continuous operation and the sludges are landfilled in an
adjacent, on-site area. Unlike the wastes from the Ti02-
Sulfate Process, the iron content of the concentrated wastes
from the Chloride-Ilmenite Process is largely in the ferric
state after chlorination and probably would not require
aeration if these wastes were treated in a conventional BPT
system utilizing neutralization and settling. After
settling and clarification, the acidic wastes at Plant |550
are deep well injected without prior neutralization. The
other process waste waters from this plant include the dilute
acid wastes from the scrubbers and white water from the
finishing operations. The dilute acid wastes are equalized,
neutralized with caustic and sent to a primary settling pond,
a polishing pond, and finally a clear pool prior to
discharge. The white water- from product finishing, first
goes to a slip pond for pigment recovery before mixing with
the neutralized dilute acid wastes in the primary settling
pond. Noncontact cooling water and sanitary wastes are
handled separately.
461
-------
14.17.2 Verification Program
No verification program was carried out in the
subcategory since the only other nominally chloride-
ilmenite plant in operation during the sampling program was
Plant |237. However, this plant was not using ilmenite ore
during the period of study and therefore was not visited or
sampled.
14.17.3 Toxic Pollutant Concentration
Because neither of the two operating Ti02 chloride-
ilmenite process plants were actually using ilmenite ore at
the time when the sampling program was being conducted, the
toxic pollutant characteristics of this process have been
estimated on the basis of sampling results from the Ti02-
sulfate process at Plant #559 where a typical ilmenite was
being used. The process waste water characteristics of the two
processes are expected to be similar because the sources of
iron and toxic metal pollutants are related to the use of the
same type of ore material. This segment was not combined with
the sulfate process segment because the manufacturing process is
different. The basic difference between the two processes is
the chemical agent used in the reaction with the ore and this
has a significant impact on the conventional and
nonconventional pollutant parameters, such as acidity,
suspended and dissolved solids, sulfate, chloride, and iron
(ferrous vs. ferric).
Thus, the toxic pollutants found at potentially significant
levels in the raw waste during sampling of Titanium Dioxide
Sulfate process plants (Section 14.10.3) are also presented
here to be used as the basis for evaluating the
pollutant characteristics of the Titanium Dioxide-Chloride
Ilmenite process.
Maximum Concentrations Observed
(ug/1)
Screening Verification
Pollutant (Plants £555 & f559) (Plant f559)
Chromium
Nickel
line
Lead
Copper
Cadmium
Selenium
Antimony
Thallium
Arsenic
124,000
6,400
3,800
3,700
1,500
340
340
20
19
11
31,000
1,300
17,000
5,200
1,000
12
< 20
1,400
41
340
462
-------
The only organic toxic pollutant found during the screening
program was phenol which was observed at only one of the two
plants sampled. The maximum raw waste concentration of
phenol was 0.020 mg/1, however, the raw water source for the
plant contained as much as 0.007 mg/1. Therefore, phenol is
not considered a significant or process-related pollutant
because it is well below the concentration which is treatable.
Daily raw waste loads were calculated from the flow rates
measured or estimated at the time of sampling and the measured
pollutant concentrations. That is,
Daily loading (as kg of pollutant per day) = (C)(Q)
1000
Where the concentration (C) of the pollutant is expressed
in units of mg/1. (Note: 1 kg/m3 = 1000 mg/1), and the flow
rate (Q) is expressed in units of m3/day (m3, a cubic meter, is
equal to 264 U.S. gallons).
Similarly, the unit loadings were calculated from the
reported Ti02 production rate (P) , the waste stream flow rate
(Q), and the measured pollutant concentration (C).
Unit loading (as kg of pollutant per kkg of Ti02) = (C)(Q)
• 1000(P)
where C and Q are expressed in the same units described above,
and the production (P) is expressed in units of kkg/day (kkg is
1000 kg, a metric ton, which is equal to 2205 Ibs).
A summary of daily and unit per unit of production raw
waste loads for the plants sampled is presented in Table 14-38
and the individual plant averages are given in Table 14-39.
The estimated total annual raw waste water load of toxic
pollutants generated by the Chloride-Ilmenite Process is given
below.
Total Annual Raw Waste Water Load
Pollutant (kg/year)
Chromium 1,050,000
Nickel 42,000
Zinc , 178,000
Lead 94,000
Copper 44,000
Cadmium 9,900
Antimony 58,000
Thallium 2,900
Arsenic 99,000
Selenium 16,000
463
-------
TABLE 14-38. SU-MflBY OF RAW HASTE LOADINGS FOUND IN SCREENING AND VERIFICATION SAMPLING
SUBCATEGORY TITANItH DIOXIDE - SULFATE PROCESS (Applied to Chloride Ilroenite Process)
Pollutant Loading Range
(kg/day)
Minimum Maximum
Priority
Antimony 5.0 28
Arsenic 1.9 4.0
Cadmium 0.068 7.2
Chromium 140 530
Copper 8.2 19
Lead 3.0 65
Nickel 3.7 23
Selenium 7.6 9.5
fliallium 0.47 1.3
Zinc 1.8 85
Conventional
OSS
Iron
Minimum
0.032
0.012
0.00044
1.1
0.065
0.024
0.029
0.0020
0.0030
,0.014
Unit Loading
(kg/kkg)
Average
0.11
0.19
0.019
2.0
0.085
0.18
0.080
0.031
0.0055
0.34
320
600
Maximum
0.22
0.032
0.057
3.4
0.12
0.42
0.15
0.06
0 .0080
0.55
No. of
Plants (1)
3
3
3
3
3
3
3
2
2
3
1
1
(1) - Data are taken only from those plants where pollutants were found above detection limits, or
in the case of TSS and Iron, where data are available.
-------
TABLE 14-39. TOXIC POLUUTANT AVERAGE RAW WftSffi LOADS WTO
cn
SUBCATEGORY TITANIUM DIOXIDE — Sulfate Process
(Applied to Chloride Ilmenite
Screening
Plant 1555
(mg/1) (kg/kkg)
antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Nickel
Selenium
Thallium
Zinc
0.77
0.11
0.29
3.8
0.20
0.075
0.091
MA
NA
0.088
0.22
0.032
0.057
1.1
0.065
0.024
0.029
< 0.06
NA
0.014
Plant 1559
(ma/1) (kg/kkg)
0.16
0.029
0.002
7.0
0.25
0.20
0.31
NA
0.02
1.1
0.080
0.014
0.0009
3.4
0.12
0.10
0.15
NA
0.008
0.55
Process)
Verification
Plant 1559
(mg/1) (kg/kkg)
0.074
0.028
0.0010
3.1
U.
0.96
0.14
0.005
0.007
1.04
0.032
0.012
0.00044
1.4
0.070
0.42
0.061
0.002
0.003
0.45
NR = Not Available
-------
Section 5.1.2 of this report described the scope and
methodology of the sampling program and Section 14.10.3
indicates the size of the analytical data base on toxic
pollutants for the sulfate process segment. This is the
basis for selecting pollutants of concern in the chloride-
ilmenite segment of the Ti02 subcategory.
14.18 POLLUTION ABATEMENT OPTIONS
14.18.1 Toxic Pollutants of Concern
Rationale for selection of the toxic pollutants of
concern is presented in Section 14.11.1 for the sulfate process
industry. The sampling data evaluations resulted in the
selection of chromium, zinc, nickel, lead, copper, antimony,
arsenic, and cadmium on the basis of raw waste maximum
concentrations and total annual industry loads.
The major impurity found in the various grades of raw
material is ferrous iron as shown in Table 14-19. In the
sulfate process the unwanted iron remains largely in the
ferrous state and may be crystallized out of the acid waste
streams and sold as coppers (ferrous sulfate). In the
chloride-ilmenite process, the same ore impurity is largely
oxidized to the ferric state during the chlorination step.
This appears in the acid waste streams as ferric chloride
(FeC13) in the amounts indicated in Table 14-38.
Iron, in either the ferrous or ferric state, is classified
as a nonconventional pollutant. However, when present in large
amounts, such as it is in the Ti02 industry, it can be a
considerable aid to toxic metal removal in treatment systems
designed to take advantage of coprecipitation processes.
14.18.2 Process Modifications and Technology Transfer Options
The comments made in regard to the Titanium Dioxide-
Chloride Process for rutile and upgraded ores in Section
14.4.2 are generally applicable to the Chloride-Ilmenite
Process.
14.18.3 Best Management Practices
Storm water runoff from the plant site should be
collected and sent to the treatment facility for the removal
of suspended solids.
466
-------
14.18.4 Prevailing Control and TreatJnent Practices
The three chloride-ilmenite plants from which the water
use and waste source information was obtained all handle the
disposal of the concentrated process waste stream separately
by either ocean dumping or deep well injection. The
availability of either of these methods of disposal for a
particular plant is a matter handled on a case-by- case basis
by the appropriate regulatory agencies from which various
approvals and permits are required under the Marine
Protection, Research and Sanctuaries Act of 1972 for ocean
disposal or by state and local authorities for deep well
injection. For the purpose of developing the model plant
concept and specifying a generally applicable- waste treatment
technology for the chloride-ilmenite industry, for the purpose
of this study the assumption has been made that neither the
ocean dumping nor the deep well injection disposal options
are generally available, and that the concentrated process
waste stream is, therefore, included in the raw waste influent
to the model plant waste water treatment system.
In practice, one plant disposes of the entire metal
chloride, HC1, and TSS waste by ocean dumping. The remainder of
the plants dispose of the concentrated waste by deep well
injection after use of surface lagoons for removal of settleable
solids.
The dilute process waste streams are segregated to the
extent possible from noncontact sources and treated in
conventional in-plant systems utilizing equalization and spill
diversion facilities followed by lime
neutralization/coagulation, solid separation in a settling
pond, and final discharge of the treated effluent. Chemical
coagulating agents such as ferric chloride and alum may be
used either before or after pH control as an aid in the removal
of metal hydroxides and other suspended solids.
14.18.5 Advanced Treatment Technology
Advanced treatment technology options for in-plant
treatment of process wastes have been evaluated as possible
polishing step additions to a conventional system for
equalization, neutralization, and clarification in ponds
prior to discharge. Such options include:
1. Aeration for a) deearbonization if limestone is -used
for neutralization, and b) ferrite coprecipitation, assuming
that sufficient ferrous iron is aleady present or is added to
the system as needed (the latter may also be accomplished by
adding scrap iron to the acid wastes).
467
-------
2. An alkaline precipitation step under optimum
conditions for metal hydroxide precipitation, i.e. pH 9-10.
3. Dual-media filtration for additional removal of
suspended solids including toxic metal hydroxides.
4. Sulfide precipitation for additional toxic metal removal
followed by filtration.
5. Other metal removal technologies including xanthate
precipitation, ion exchanger and membrane applications, all of
which were regarded as categorically inappropriate from a
practical and economic point of view.
14.19 SELECTION OF APPROPRIATE TEC1NOLOGY AND EQUIPMENT
14.19.1 Technologies for Different Treatment Levels
Level 1 (BPT)
Figure 14-16 shows the model treatment system chosen for
this • subcategory. Calcium carbonate (limestone) is used
to neutralize the concentrated acid waste stream. The
priority pollutants are precipitated in the first stage
thickeners. Aeration then oxidizes any ferrous iron present
and removes C02 before mixing with dilute and
miscellaneous plant wastes. The combined stream is then given
lime treatment to pH 9-10 for additional toxic metals removal
and settled in polishing lagoons before discharge. This
treatment system is patterned after the model plant BPT
waste water treatment technology for the sulfate process as
presented in this report.
This technology was used as the treatment model for BPT
regulations because of the similarity of wastes to those in the
Ti02-Sulfate Process industry. This technology is available
and, to some degree, already employed in the Ti02-Chloride-
Ilmenite industry. The proposed BPT treatment would remove
greater than 95 percent of the major pollutants of concern
including toxic metals according to preliminary treatability
estimates.
Level 2 (NSPS)
Level 2 treatment adds dual-media filtration to the
Level 1 technology for additional removal of suspended solids
and toxic metal hydroxides following the alkaline
precipitation and settling steps. The flow diagram for Level
2 is shown in Figure 14-17. This level of treatment was
468
-------
•a
469
-------
c!
WlSTE
siJjrtw
pl MHUSTHBJT
EFFUJENT
DUO,
(EOIA
FZUTGR
THICXQER
Figure 14-17.
Level 2 waste water treatment for titanium dioxide - chloride
(ilmenite ore) process.
-------
selected as the basis for NSPS because it provides a
relatively economical method for removing additional toxic
metals.
14.19.2 Equipment for Different Treatment Levels
Equipment Functions
Unlike treatment of the waste waters from the Ti02
Sulfate Process, limestone neutralization of the Chloride-
Ilmenite Process waste waters does not generate large
quantities of solids (e.g., 'gypsum) which require mechanized
separation and transfer to sizable on-site or off-site
disposal areas. The solids that are generated from TSS and
metal precipitate separation can be collected in moderate
sized lagoons and periodically transferred to appropriate
chemical landfill disposal sites in accordance with the
Resource Conservation and Recovery Act (RCRA) (as amended,
42 USC 6901, et seq.) . The Level" 1 treatment model
includes rail car deliveries of ground limestone amd line,
bucket elevators, storage bins, multiple reactors and chemical
feeders, mechanical aerators and thickeners for solids
removal. The clarified overflow is treated with lime for
additional toxic metals removal and settled in a one-day
polishing pond prior to final pH adjustment, monitoring and
discharge.
Chemicals and Chemical Handling
First stage neutralization utilizes ground limestone
while lime is used for second stage neutralization and final
alkaline precipitation. Oxygen is supplied as air and
treatment chemicals may be added as required for removal of
precipitated metals and other suspended solids. Aside from
the large scale bulk chemical handling requirements for
limestone and lime, there are no particular hazards
involved.
Disposal, of Solids
Periodic removal of solids from settling impoundments
will require compliance with RCRA regulations as applicable to
on-site or off-site chemical disposal site operation.
471
-------
14.20 TREATMENT COST ESTIMATES
14.20.1 General Discussion
Preliminary cost estimates hae been prepared for Level 1
(BPT) treatment only. Final cost estimates, including those
for Level 2 (NSPS) will be prepared prior to final
promulgation of the regulations. The model plant
specifications given below were utilized for preliminary cost
estimating and for development of, the proposed regulations.
Production
There are three plants at different locations producing
(or capable of producing) titanium dioxide by the combined
ilmenite ore beneficiation - chlorination process. Annual
capacity of these plants varies from 136,000 metric tons to
207,000 metric tons. For treatment cost estimates, four
production levels were selected. These were 35,000 kkg/year,
70,000 kkg/year, 113,750 kkg/year, and 157,500 kkg/year.
Waste Water Flows
Waste water is typically segregated into two streams;
strong acidic waste water flow from beneficiation -chlorination
of ilmenite ore and air emission scrubbing facilities, and
the other waste water from process reactions, washings,
product transport, cooling tower blowdown, water treatment
blowdown, and other operations. For the model plants, a unit
flow of 6 m3/kkg of product for the concentrated acidic waste
water and 114 m3/kkg of product for the dilute wastes is used.
The treatment system is designed to handle a total flow of 120
m3/kkg of product (Table 14-36).
For the NSPS model plant, a unit flow of 6 m3/kkg of
product for the concentrated acidic waste water is used.
Because of improved design which allows for recycle
systems and more efficient process water utilization,
dilute waste water is considerably reduced. The total
combined waste water flow of 31 m3/kkg of product is used
(Table 14-36) . The treatment system is Level 2 which is BPT
plus dual, media filtration.
Pollutant Load
The principal pollutants occurring in the waste waters are
TSS, iron, chromium, zinc, and hydrochloric acid. For the
model plants, the following unit pollutant loads have been
considered:
472
-------
TSS 175 kg/kkg of Ti02
HC1 230 kg/kkg of Ti02
Iron 375 kg/kkg of Ti02
Chromium 1.4 kg/kkg of Ti02
Zinc 0.5 kg/kkg of Ti02
The loading values for TSS, HC1, and iron are based on
data submitted by industry on the chloride-ilmenite
process. The chromium loading is an estimated average
derived from a wide range of ilmenite ore qualities and the
zinc loading is taken from the screening and verification
data on the Ti02 sulfate process.
Chemical Usage
In the model BPT system, powdered limestone is used for
first stage .neutralization of strong acidic waste flow at the
unit rate of 302 kg/kkg of Ti02. Pebble lime (CaO) is used for
second stage neutralization of the mixed acidic and other waste
waters and final neutralization of the total combined flows.
Lime is used at the unit rate of 42 kg/kkg of Ti02.
Solid Waste
The solids produced in the treatment facility consist of
iron hydroxides, the original suspended solids introduced in
the influent and solids derived from the treatment chemicals
added for neutralization. The total solids produced in the
model plant are assumed to be 990 kg/kkg of Ti02.
14.20.2 Model Plant Control and Treatment Costs
The estimated costs for four models having different
production levels are given in Tables 14-40, 14-41, 14-42,
and .14-43.
Table .14-44 presents a summary of the unit cost
distribution between amortization and operation and
maintenance cost components at different productions at the
BPT level of treatment.
For existing sources at the first level of treatment,
the disposal of sludges is on-site, hence land requirements
are fairly large. Amortization, chemicals, labor, and residual
waste disposal costs have significant impact on the annual
costs.
The unit waste flow of 6 m3/kkg of product for the
concentrated acidic waste water stream is the same for BPT and
NSPS systems. The NSPS treatment technology is the same as
BPT, but the total combined acidic and dilute waste water
473
-------
TABLE 14-40. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE Chloride-Il. Ore
Production 35,000 metric tons per year (38,587 tons per year)
100 metric tons per day (110 tons per day)
Waste water flow 12,000 cubic meters per day.
LEVEL OF TREATMENT*
A. INVESTMENT COST FIRST
Construction $300,500
Equipment in place,
including piping,
fittings, electrical
work and controls 696,500
Monitoring equipment
in place ' 9,000
Engineering design
and inspection 201,200
Incidentals, overhead,
fees, contingencies... 201,200
Land 252,_000
TOTAL INVESTMENT COST $1,660,400
B. OPERATION AND
MAINTENANCE COST
Labor and supervision. $336,000
Energy , 31,000
Chemicals 260,000
Maintenance 140,840
Taxes and insurance... 49,812
Residual waste
disposal 105,000
Monitoring, analysis
and reporting i5i222
TOTAL OPERATION AND
MAINTENANCE COST $937,652
C. AMORTIZATION OF
INVESTMENT COST $229^146
TOTAL ANNUAL COST . $1,166,798
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
474
-------
TABLE 14-41. PLANT COSTS
Subcategory TITANIUM DIOXH3E Chloride-Il. Ore
Production 70,000 metric tons per year (77,175 tons per year)
200 metric tons per day (220 tons per day)
Waste water flow 24,000 cubic meters per day.
LEVEL OF TREATMENT *
A. INVESTMENT COST FIRST
Construction,.......... $387,500
Equipment in place,
including piping,
fittings, electrical
work and controls...... 865,000
Monitoring equipment
in place. 9,000
Engineering design
and inspection 252,300
Incidentals, overhead,
fees, contingencies.... 252,300
Land...... 492,000
TOTAL INVESTMENT COST $2,258,100
B. OPERATION AND
MAINTENANCE COST
Labor and supervision. $504,000
Energy 43,000
Chemicals 510,000
Maintenance. 176,610
Taxes and insurance... 67,743
Residual waste
disposal 105,000
Monitoring, analysis
and reporting......... 15,000
TOTAL AND
COST $1,421,353
C. AMORTIZATION OF
INVESTMENT COST $287,344
TOTAL ANNUAL COST $1,708,697
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
475
-------
TABLE 14-42. PLANT COSTS
Subcategory T33ANIOM DIOXIDE Chloride-Il. Ore
Production 113,750 metric tons per year (125,409 tons pec year)
325 metric tons per day (358 tons per day)
Waste water flow 39,000 cubic maters per day.
LEVEL OP TREATMENT*
A. INVESTMENT COST FIRST
Construction $508,000
Equipment in place,
including piping,
fittings, electrical
work and controls 1,179,500
Monitoring equipment
in place. 9,000
Engineering design
and inspection........ 339,300
Incidentals, overhead,
fees, contingencies... 339,300
Land. 2§0£QOO
TOTAL HSraBSTMENT COST $3,155,100
B. OPERATION AND
MAINTENANCE COST
Labor and supervision. ,$588,000
Energy 62,000
Chemicals...... 823,000
Maintenance 237,510
Taxes and insurance... 94,653
Residual waste
disposal 210,000
Monitoring, analysis
and reporting :L§iP_P_P_
TOTAL OPERATION AND
MAINTENANCE COST $2,030,163
C. AbDRTIZATION OF
INVESTMENT COST !2§§ii3§
TOTAL ANNUAL COST $2,416,591
* First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
476
-------
TABLE 14-43. MODEL PLANT COSTS
Subcategory TITANIUM DIOXIDE Chloride-Il. Ore
Production, 157,500 metric tons per year (173,643 tons per year)
450 metric tons per day (496 tons per day)
Waste water flow 54,000 cubic meters per day.
LEVEL OP TREATMENT*
A. INVESTMENT COST FIRST
Construction $638,000
Equipment in place,
including piping,
fittings, electrical
work and controls 1,356,000
Monitoring equipment
in place................ 9,000
Engineering design
and inspection. 400,600
Incidentals, overhead
fees, contingencies..... 400,600
Land 1,080,000
TOTAL INVESTMENT COST $3,884,200
B. OPERATION MID
MAIMENANCE COST
Labor and supervision. $588,000
Energy. 71,000
Chemicals 1,141,000
Maintenance. 280,420
Taxes and.insurance... 116,526
Residual waste
disposal 210,000
Monitoring, analysis
and reporting i§x.2§2
TOTAL AND
COST $2,421,946
C. AMORTIZATION OP
INVESTMENT COST li§§£.243
TOTAL ANNUAL COST $2,878,189
*F±rst level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
477
-------
TABLE 14-44. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE Chloride-Il. Ore
LEVEL OF TREfflMENT
PRODUCTION FLOW
(kkg/yr) (m3/day)
FIRST
SECOND
THIRD
Annual Operation
and Maintenance
Annual
Amortization
Total Cost
35,000
70,000
113,750
157,500
35,000
70,000
113,750
157,500
35,000
70,000
113,750
157,500
12,000
24,000
39,000
54,000
12,000
24,000
39,000
54,000
12,000
24,000
39,000
54,000
26.79
20.31
17.85
15.38
6.55
4.10
Not Applicable
.40
.90
33.34
24.41
21.24
18.27
478
-------
flow for NSPS system is much less than BPT model; however,
the reduced flow has negligible impact on costs because the
unit waste loads are the same. There is insignificant
difference in the estimated total annual costs per kkg of
product between BPT and NSPS levels of treatment for the model
plant designs.
14.21 BASIS FOR REGULATIONS
14.21.1 Evaluation of BPT Treatment Practices
The prevailing control and treatment practices in the Ti02
Chloride-Ilmenite industry have been reviewed in Section
14.18.4. For the purpose of regulations development, it has
been assumed that neither ocean dumping nor deep well
injection methods are generally available as disposal options
for all or any portion of the process-related wastes. Thus,
treatment technology used in the Ti02-Sulfate Process Segment
of the industry has been used as the basis for the proposed
regulations.
14.21.2 Basis for Proposed BPT Effluent Limitation
Technology Basis
The Agency is proposing BPT limitations based on
technology used in the Ti02-Sulfate Process industry (Section
14.14.2) involving equalization, limestone neutralization,
clarification, aeration, alkaline precipitation, and settling
followed by final pH adjustment and discharge. The rationale
for the selection of Level 1 technology is given in Section
14.19.1.
Flow Basis
The BPT model plant flow rate is based on the reported
average process contact and clean up waste water flow at Plant
£237 of 120 m3/kkg as indicated in Table 14-36.
Selection of Pollutants to be Regulated
The selection of pollutants to be regulated in the Ti02
Chloride-Ilmenite industry is based on the analysis of raw waste
maximum concentrations and total industry loadings as presented
in Section 14.3.3. The significant toxic pollutants include:
479
-------
Chromium
Zinc
Nickel
Iiead
Copper
Antimony
Arsenic
Cadmium
Each of these pollutants was present at least once in raw
waste water at a maximum concentration level ' regarded as
treatable in accordance with the appropriate treatability
estimates presented in Table 8-11. Although arsenic is a
borderline case with an observed maximum of 0.34 mg/1, it is
regarded as a candidate for regulation based on alkaline
precipitation technology. All of the other pollutants
identified here as significant are definately treatable by this
technology.
This selection follows the same logic presented in
Section 14.14.2 for the Ti02-Sulfite Process. At the BPT
level, the Agency is also proposing limitations on Total
Suspended Solids (TSS) and iron which are classified as
conventional and nonconventional pollutants, respectively.
Conventional and nonconventional parameters -
A. pH: The treated effluent is to be controlled within
the range of pH 6.0 to 9.0. This limitation is based on the
data presented in Appendix B of this report and the JRB study
(52).
B, TSS and iron; The analysis of long term monitoring
data from Plant §559 (Table 14-30) indicates an achievable
long-term average of 21 mg/1 for TSS and 0.62 mg/1 for iron
(total).
For TSS, the proposed maximum 30-day average limitation
is derived from the long term average, the variability factor
of 3.0 for 30-day averages (rounded off from 3.04 in Table
14-30), and the BPT model plant flow rate of 120 m3/kkg.
The proposed TSS maximum 30-day average concentration basis is
given by:
(3.0)(21 mg/1) = 63 mg/1
and the proposed TSS maximum 30-day average by:
(63 mg/1) (120 m3/kkg)/* kg/m3 \
\1QQO mg/1/
=7.6 kg/kkg
480
-------
With a variability factor ratio (VFR) of 3.6, the
corresponding proposed TSS daily maximum limitation is given by;
(3.6)(7.6 kg/kkg) = 27 kg/kkg
and the TSS daily maximum concentration basis is:
(63 mg/1) (3.6) = 230 mg/1
Similarly for iron, the concentration basis for the
proposed maximum 30-day average limitation is derived from the
long term average and a variability factor of 4.0. The
proposed iron maximum 30-day average (Table 14-31) concentration
are given as:
(4.0) (0.62 mg/1) = 2.5 mg/1
and the proposed iron limitation is:
(2.5 mg/1)(120 m3/kkg)/ kg/m3 \
• V 1000 mg/1/
=0.30 kg/kkg
The corresponding proposed daily maximum limitation for
iron is determined by applying the variability factor ratio
(VFR) of 3.4 as follows:
(3.4) (0.30 kg/kkg) = 1.0 kg/kkg
and the iron daily maximum concentration basis is:
(3.4) (2.5 mg/1) .= 8.5 mg/1
The proposed BPT limitations are presented in Table 14-45.
Toxic Pollutants - For the Ti02 Chloride-Ilmenite process,
the proposedlimitations on the toxic metals found at
significant concentrations are based on estimates of
achievable 30-day „ average concentrations as presented in
Table 8-11 because no directly applicable industry treatment
performance data are available. The lower• limits of
treatability shown for lime/setting are taken as the
concentration bases for the proposed maximum 30-day average
limitations on the various toxic metals. The variability
factor ratio (VFR) used for each pollutant is identical to
the value derived from long-term monitoring data on Plant f559
presented in Table 14-31.
A. Antimony: The concentration basis for the proposed
maximum 30-day average limitation is the lower limit of
481
-------
TABLE 14-45. PROPOSED LIMITATIONS
Titanium Dioxide - Chloride Process Using Ilmenite
Best Practicable Control Technology Currently Available
Waste Water Flew: 120 m3/kkg
Pollutant
Estimated
Treatability
(mg/1)
Concentration Basis
WRW (rag/1)
Max
30-day
Avg
24-hr
max.
Effluent Limit
(kg/kkg)
Max 24-hr
30-day max.
Avg
Conventional and
Nbnconventional
Pollutants
Total Suspended
Solids
Iron
Toxic Pollutants
Antimony (6)
Arsenic (6)
Cadmium (6)
Chromium (6)
Copper (6)
Lead (6)
Nickel^)
Zinc (6)
2l(2) 3.6 63
0 fi? ' ) 34 75
VS * \J> <-« •w/*T± £,* * +J
0.80(3) 1.9(4) 0<80
0.50(3) l.gte) o,50
o.io (3) i.e(4) o.io
0.10(3) 1.9 (4) 0.10
0.50(3) 1.9(4) 0.50
0.30 (3) 1.5 (4) 0.30
0.20(3) 1.9(4) Qm2Q
0.50(3) 2.1(4) 0.50
230
8.5
1.5
0.95
0.16
0.19
0.95
0.45
0.38
1.1
7.6
0.30
0.096
0.060
0.012
0.012
0.060
0.036
0.024
0.060
27
1.0
0.18
0.11
0.019
0.023
0.11
0.054
0.046
0.013
(1) - WRs ratio of the 24-hr (daily) variability factor to the 30-day
average variability factor
(2) - Long term average from Plant t559 monitoring data (Table 14-31)
(3) - Estimated lower limit of treatability as* a 30-day average (Table 8-11)
(4) - Based on long term data from Plant #559(Table 14-30)
(5) - Set equal to the YFR for antimony
(6) - Applicable to proposed BAT and PSES limitations.
482
-------
treatability, 0.8 mg/1 , from Table 8-11. Applying the BPT
model plant flow rate of 120 m3/kkg.
The proposed antimony limitation is given by:
kg/m3
1000 mg/1.
(0.80 mg/1) (120 mS/kkg)/* kg/m3 \
.)
= 0.096 kg/kkg
and, by applying the VFR value of 1.9, the proposed antimony
daily maximum limitation is,
(1.9) (0.096 kg/kkg) = 0.18 kg/kkg.
for which the corresponding concentration is:
(1.9) (0.80 mg/1) =1.5 mg/1
B. Arsenic: For arsenic, the lower limit of treatability
is 0.50 mg/1. Although long-term monitoring data on arsenic are
not available, arsenic is expected to behave in a manner
similar to antimony during lime treatment and for this
reason, the same VFR value of 1.9 is utilized.
Thus, the proposed arsenic maximum 30-day average
limitation is given by:
(0.50 mg/1) (120 m3/kkg)/ kg/m3 \
\1000 mg/1/
= 0.060 kg/kkg
and the proposed antimony daily limitation by:
(1.9) (0.060 kg/kkg) = 0.11 kg/kkg
C. Cadmiums The lower limit of treatability for cadmium
is estimated at 0.10 mg/1 as a 30-day average (Table 8-11).
Using this value as the concentration basis, the proposed
cadmium maximum 30-day limitation is given by:
(0.10 mg/1) (120 m3/kkg)/" kg/m3 \
VLOOO mg/V
= 0.012 kg/kkg
and the proposed cadmium daily maximum limitation is
obtained by applying the VFR value of 1.6 from Table 14-31.
That is:
(1.6) (0.012 kg/kkg) = 0.019 kg/kkg
483
-------
D, Chromium: The proposed chromium limitations are
based on an estimated 30-day average treatability limit of
0.10 mg/1 using lime/settling treatment. The achievability
of this concentration level is predicated on the assumption
that chromium is in the trivalent state and no significant
amount of the hexavalent form is present.
Thus, the proposed chromium maximum 30-day average
limitation is given by;
(0.10 mg/1) (120 m3/kkg)/ kg/m3 \
V1000 mg/1/
= 0.012 kg/kkg
and application of the WR value of 1.9 gives the proposed
chromium daily maximum limitation as,
(1.9) (0.012 kg/kkg) = 0.023 kg/kkg.
E. Copper: Using an estimated lower limit of
treatability for copper of 0.50 mg/1 and a VFR value of
1.9, the proposed limitations are identical to those given
above for arsenic. The proposed maximum 30-day limitation
is 0.060 kg/kkg and the proposed daily maximum limitation is
0.11 kg/kkg.
P. Lead: The lower limit of treatability for lead is
estimated at 0.30 mg/1 as a 30-day average (Table 8-11) . Osing
this value as the concentration basis, the proposed lead
maximum 30-day average limitation is given by:
(0.30 mg/1) (120 m3/kkg)/" kg/m3 \
UOOO mg/1/
= 0.036 kg/kkg
and, applying the VFR of 1.5 from Table 14-31, the proposed lead
daily maximum limitation is:
(1.5) (0.036 kg/kkg) = 0.054 kg/kkg
G. Nickel: In a similar manner for nickel, the proposed
limitations -are based on an estimated treatabiliy limit of
0.20 mg/1 and a VFR value of 1.9. The proposed nickel maximum
30-day average is given by:
(0.20 mg/1) (120 m3/kkg)/ kg/m3 \
VLOOO mg/1/
= 0.024 kg/kkg
and the proposed nickel daily maximum is given by:
484
-------
(1.9) (0.024 kg/kkg) - 0.046 kg/kkg
H. Zinc: The estimated treatability limit for zinc is
the same as arsenic and copper, i.e., 0.50 mg/1, however, a VFR
value of 2.1 is applied instead of 1.9.
Thus, the proposed zinc maximum 30-day average limitation
is given by:
(0.50 mg/1) (120 m3/kkg)/ kg/m3 \
VOOO mg/1/
= 0.060 kg/kkg
and the proposed zinc daily maximum is given by:
(2.1) (0.060 kg/kkg) = 0.13 kg/kfcg
The proposed BPT limitations are presented in Table 14-45.
14.21.3 Basis for Proposed BCT Effluent Limitations
For BCT, the Agency is proposing limitations for TSS
equal to the BPT limitations because BAT is equal to BPT.
14.21.4 Basis for Proposed BAT Effluent Limitations
For BAT, the Agency is proposing limitations on iron and
the toxic pollutants based on the application of Level 1
technology which is equivalent to BPT. The model plant flow
basis of 120 m3/kkg used for BPT is also used for BAT. The
proposed BAT limitations are presented in Table 14-45. A more
advanced technology using soda ash precipitation and recycle of
waste water was considered for the similar sulfate process but
was rejected because its performance has not been demonstrated.
14.21.5 Basis for the Proposed Mew Source Performance Standards
Technology Basis
For NSPS the Agency is proposing limitations based on the
application of Level 2 treatment technology which adds dual
media filtration to the BPT system for greater efficiency
in the removal of suspended solids including iron and
toxic metal precipitates.
Flow Basis
The reported data on process contact and clean-up waste
water flow at Plant $713 is selected as the basis of a model
plant for new sources. Process modifications resulting in a
485
-------
greatly increased efficiency of water use reduce the average
flow rate to 31 m3/kkg as shown in Table 14-36.
Basis for Pollutant Limitations
Conventional parameters -
A. pH: The treated effluent is to be controlled within
the range of pH 6.0 to 9.0. This limitation is based on
data presented in Appendix B of this report and the JRB Study
(52).
B. TSS: The concentration basis for the proposed NSPS
maximum 30-day average limitation is obtained by applying an
average filtration efficiency of 38 percent removal (41) to the
corresponding BPT concentration of 64 mg/1 (Table 14-45). That
is:
(1.00-0.38) (64 mg/1) = 40 mg/1:
Then, the proposed maximum 30-day average limitation is
obtained by applying the NSPS model plant flow rate of 31
m3/kkg:
(40 mg/1) (31 m3/kkg)/ kg/m3 \
VLOOO mg/1/
=1.2 kg/kkg
The proposed TSS daily maximum limitation is determined
by multiplying this value by the VPR of 3.6 (Table 14-31) ,
namely:
(3.6) (1.2 kg/kkg) =4.3 kg/kkg
and the daily maximum concentration basis is:
(3.6) (40 mg/1) = 140 mg/1
The same VFR that was used in developing the proposed BPT
limitations is also used for NSPS because the actual
variability • of the BPT system with added filtration is
expected to be somewhat less than the statistically derived
VPR for BPT.
The proposed NSPS limitations are presented in Table 14-46.
Nonconventional pollutants - The only nonconventional
pollutantolconcern is iron. For NSPS, the Agency is
proposing a maximum 30-day average limitation based on an
average filtration efficiency of 38 percent removal (41).
486
-------
TABLE 14-46. PROPOSED LIMITATIONS
Titanium Dioxide - Chloride Process Using llmenite
New Source Performance Standards*
Waste Water Flow: 32 m3/kkg
Pollutant
Treatability (2)
(mg/1)
Concentration Basis
(1) (mg/1)
Max
30-day
Ava
24-hr
max.
Effluent Limit
(kg/kkg)
30*12 24~hr
ou— oay max.
Avq
(jonventional and
Nonconventional
Total suspended 40
Solids
Iron 1.6
3.6 40
3.4 1.6
140
5.4
1.2
0.050
4.3
0.17
Toxic Pollutants
Antimony ")
Arsenic ^ '
Cadmium' 3)
Chromium (3)
Copper '^)
Lead (35
Nickel (3)
Zd*c<3>
0.80
0.50
0.075
0.040
0.29
0.060
0.17
0.47
1.9
1.9
1.6
1.9
1.9
1.5
1.9
2.1
0.80
0.50
0.075
0.040
0.29
0.060
0.17
0.47
1.5
0.95
0.12
0.076
0.55
0.090
0.32
0.99
0.025
0.016
0.0023
0.0012
0.0090
0.0019
0.0053
0.015
0.048
0.030
0.0037
0.0023
0.017
0.0029
0.010
0".032
(1) - VFR: ratio of the 24-hour daily variability factor to the 30-day
average variability factor.
(2) - Based on the application of pollutant - specific removal efficiencies
for dual-media filtration (41) to adjust the BPT performance on
treatability estimates shown in 30-day average concentrations in
Table 14-45.
* Including pretreatment standards for new sources (PSNS) covering iron
and toxic metals which are expressed as concentrations.
(3) - Applicable to proposed PSNS limitations.
487
-------
Thus, the appropriate concentration basis is derived from the
corresponding concentration basis of 2.5 mg/1 of iron used for
the BPT maximum 30-day average (Table 14-45) . That is:
(1.00-0.38) (2.5 mg/1) = 1.6 mg/1
and the limitation proposed for NSPS is:
(1.6 mg/1) (31 m3/kkg)/ Kg/m3 \
V 1000 mg/1/
= 0.050 kg/kkg
Again, applying the same VFR value of 3.4 that was used for the
BPT limitations, the proposed NSPS daily maximum limitation for
iron is:
(3.4) (0.050 kg/kkg) = 0.17 kg/kkg
and the daily maximum concentration basis is:
(3.4) (1.6 mg/1) = 5.4 mg/1
Toxic pollutants - The Agency is proposing new source
performance standards for the eight toxic metals identified at
significant concentrations during the screening and
verification sampling program. To the extent possible, a
specific filtration removal efficiency derived from published
literature data (41) is applied for each toxic pollutant
parameter. The filtration removal efficiency (percent
removal) is applied to the estimated lower limit of
treatability (Table 8-11) for lime/settling treatment (BPT
basis) to arrive at the concentration basis for each proposed
NSPS maximum 30-day average limitations.
A« Antimony and arsenic: No credit for additional
removal by filtration is taken for either antimony or
arsenic because removal data is not available. The Agency is
proposing NSPS limitations for which the concentration bases
are identical to those used for the development of BPT
limitations. Thus, for antimony, the proposed NSPS maximum
30-day average is given by:
(0.80 mg/1) (31 m3/kkg)/ kg/m3 \
\1000 mg/1/
= 0.025 kg/kkg.
and the corresponding daily maximum limitation is obtained by
applying the VFR of 1.9, that is:
(1.9) (0.025 kg/kkg) = 0.048 kg/kkg.
488
-------
Similarly, for arsenic the proposed NSPS maximum 30-
day average limitation is:
(0.50 mg/1) (31 m3/kkg) / kg/m3 \
VJLOOO mg/1/
= 0.016 kg/kkg
and the proposed daily maximum is:
(1.9) (0.016 kg/kkg) = 0.030 kg/kkg
B. Cadmium: Employing a filtration removal efficiency of
25 percent for cadmium (41) results in the following
concentration basis for the proposed NSPS maximum 30-day average
limitation:
(1.00-0.25) (0.10 mg/1) = 0.075 mg/1
Therefore, the proposed limitation is:
(0.075 mg/1) (31 m3/kkg)/ kg/m3 \
VlOOO mg/l}
» 0.0023 kg/kkg
The corresponding proposed daily maximum limitation:
(1.6) (0.0023 kg/kkg) = 0.0037 kg/kkg.
and the daily maximum concentration basis is:
(1.6) (.075 mg/1) =0.12 mg/1
C. Chromium: For chromium the filtration removal
efficiency is reported to be approximately 60 percent (4.1) .
Thus, for the proposed NSPS maximum 30-day average
limitations, the concentration basis is given by:
(1.4)0-0.60) (0.10 mg/1) = 0.040 mg/1
and the proposed NSPS limitation is:
(0.040 mg/1) (31 m3/kkg)/ kg/m3 \
\1000 mg/1/
= 0.0012 kg/kkg.
The proposed NSPS daily maximum limitation for chromium
is then obtained by applying the VFR value of 1.9, that is:
(1.9) (0.0012 kg/kkg) = 0.0023 kg/kkg.
489
-------
and the daily maximum concentration basis is:
(1.9) (0.040 mg/1) = 0.076 mg/1
D, Copper: The estimated filtration efficiency for copper
removal is approximately 42 percent (41). Therefore,
the concentration basis for the proposed NSPS maximum 30-day
average effluent limitation is given by:
(1.00-0.42) (0.50 mg/1) = 0.29 mg/lr
and the proposed limitation is:
(0.29 mg/1) (31 m3/kkg)/ kg/m3 \ = 0.0090 kg/kkg
V1000 mg/1/
The proposed NSPS daily maximum is then obtained by
multiplying the maximum 30-day average by the VPR value of
1.9. That is:
(1.9) (0.0090 kg/kkg) = 0.017 kg/kkg
and the daily maximum concentration basis is:
(1.9) (.24 mg/1) = 0.55 mg/1
E. Lead: Starting with the estimated BPT treatability
level of 0.30 mg/1 for lead and applying a filtration
removal efficiency of 80 percent (41), one obtains:
(1.00-0.80) (0.30 mg/1) = 0.060 mg/1
This is the concentration basis for the proposed NSPS maximum
30-day effluent limitation which is:
(0.060 mg/1) (31 mS/kkg)/ kg/m3 >>
\1000 mg/1/
= 0.0019 kg/kkg
The proposed NSPS daily maximum limitation for lead is
then calculated by multiplying the 30-day average limitation
by the VFR value of 1.5 as follows:
(1.5) (0.0019 kg/kkg) = 0.0029 kg/kkg
and the daily maximum concentration basis is:
(1.5) (0.060 mg/1) = 0.090 mg/1
490
-------
One can determine the concentration basis for the
proposed daily maximum limitation by applying the VFR to the
concentration basis for the maximum 30-day average:
(1.5) (0.060 mg/1) = 0.090 mg/1
F. Nickel: For nickel the estimated efficiency of
removal by dual media filtration is approximately 14 percent
(43). Thus, the maximum 30-day average concentration is:
(1.00-0.14) (0.20 mg/1) - 0.17 mg/1
and the proposed NSPS maximum 30-day average effluent
limitation for nickel is:
(0.17 mg/1) (31 m3/kkg)/ kg/m3 \
\1000 mg/1-/
= 0.0053 kg/kkg
The corresponding proposed daily maximum effluent
limitation is obtained by applying the VFR value of 1.9. That
is:
(1.9) (0.0053 kg/kkg) = 0.010 kg/kkg
and the daily maximum concentration basis is:
(1.9) (0.17 mg/1) = 0.32 mg/1
G. Zinc: For zinc, the removal efficiency using dual
media filtration is estimated at approximately 6 percent (41).
This value is applied to the BPT concentration basis to
obtain the NSPS concentration basis as follows:
(1.00-0.06) (0.50 mg/1) = 0.47 mg/1
Thus, the proposed NSPS maximum 30-day average effluent
limitations for zinc is:
(0.47 mg/1) (31 m3/kkg)/ kg/m3 \
V.1000 mg/1/
= 0.015 kg/kkg
and the corresponding proposed daily maximum limitation is
obtained by multiplying this limitation by the VFR value of
2.1. That is:
(2.1) (0.015 kg/kkg) = 0.032 kg/kkg
and the daily maximum concentration basis is:
491
-------
(2.1) (0.47 mg/1) = 0.99 mg/1
The proposed NSPS limitations are presented in Table 14-46.
14.21.6 Basis for Proposed Pretreatment Standards
Existing Sources
The Agency is proposing Pretreatment Standards for
Existing Sources (PSES) based on Level 1 (BPT/BAT)
treatment. The pollutants to be limited are iron and the
toxic metals as indicated in Table 34-45.
New Sources
Pretreatment Standards for New Sources (PSNS) are being
proposed by the Agency on the basis of NSPS treatment
technology for the Tl02-Chloride~Ilmenite industry. The
pollutants to be limited are iron and the toxic metals as
indicated in Table 14-46.
492
-------
SECTION 15
ALUMINUM FLUORIDE INDUSTRY
*t
15.1 INDUSTRY PROFIIJi
15.1.1 General Description
Aluminum fluoride is used as a raw material in the
production of cryolite (sodium fluoroaluminate) , which in turn
is used in the production of aluminum. Aluminum fluoride is
used also as a metallurgical flux (for welding rod coatings), as
a ceramic flux (for glazes and enamels), and as a brazing flux
(for aluminum fabrication).
The industry profile data for this subcategory are given in
Table 15-1, while the status of regulations is given in Table
15-2.
15.1.2 General Process Description and Raw Materials
In the dry process for the manufacture of aluminum
fluoride, partially dehydrated alumina hydrate is reacted with
hydrofluoric acid gas. The reactions is given as:
A1203 + 6HF = 2A1F3 + 3H20 (1)
The product, aluminum fluoride, is formed as a solid, and
is cooled with noncontact cooling water before being milled and
shipped. The gases from the reactor are scrubbed with water to
remove unreacted hydrofluoric acid before being vented to the
atmosphere. A simplified flow diagram of the process is shown
in Figure 15-1.
15.2 HATER USE AND WASTE SOURCE CHARACTERISTICS
15.2.1 Water Use
Water is used in noncontact cooling of the product, for
seals on vacuum pumps and for scrubbing the reacted gases before
493
-------
TKEOZ 15-1.
SOBCKIBQDSf PBCFXGE
SUBCATEGOS^
ALOMINtM FtOORIDE
Total subeatespry capacity rate
Total subcategory production rate
Number of plants in this sufacategory
308 Data on file for
With total capacity of
With total production of
Representing capacity
Representing production
Plant production range:
Minirraam
Mastitaaa
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volune per unit product:
Minimum
Maximum
134,700 kkg/year
5*
6
204,800 kkg/year
120,000 kkg/year
UR
m
38 kkg/year
45,600 kkg/year
24,300 kkg/year
35/500 kkg/year
59 percent
5 years
21 years
539 cubic meters/day
2,200 cubic meters/day
5 cubic maters/kkg
12 cubic meters/fckg
Sources of data are Stanford Research Institute., Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Ccnroerce, Current Industrial
Reports, Decsnber 1977? Energy and Bnviromental analysis, Inc.; Draft
Report, "Preliminary Economic Assessnent of Effluent Limitations in the
Inorganic Chanical industry." June, 1978- and "Eronomic Analysis of Proposed
Bevxsed Effluent Guidelines and Standards for the Inorganic Chemicals Industry,"
March, 1980.
NA =* Not Available
* Seven plants were operating at the beginning of this study, but two closed down
production after 1978.
494
-------
15-2 .
SUBPART
OF - LIMZmTION GUIDELINES
Aluminum Fluoride
W (40 CFR 415.230, 5/22/75)
STANDARDS
BPCTCA* BA3EA*
Product
Process
A1F3
Para-
meters
Fluoride
TSS
Aluminum
Max.1
kg/kkg
tog/l)
0.68
(40) 3
0.86
(51)
0.34
(20)
2
Avg. Max. Avg.
kg/kkg kg/kkcf kg/kkg
0.34
(20)
0.43
(25)
0.17
(10)
NSPS *
Max. Avg.
kg/kkg kg/kkg
(mg/1) (mg/1)
Sections 415.230, 415.231, and 415.232 were revoked by the Agency
(41 FR 51601, November 23, 1976)-.
Tflax. = Maximum of any one day.
2.
'Avg. =Max±raann average-of daily values for thirty consecutive days.
basis 17,000 1/Kkg. ,„_
495
-------
IO
HYDRATED
ALUMINA
HYDROGEN
FLUORIDE
WATER
I
VENT
SCRUBBER
CYCLONE
REACTOR
1
WSIG WATER
NONCONTACT
COOLING WATER
COOLER
PRODUCT
COLLECTION
AND STORAGE
ALUMINUM
•FLUORIDE
PRODUCT
Figure 15-1. General process flow diagram for production
of aluminum fluoride.
-------
being vented to the atmosphere. Water is also used for leak and
spill cleanup and equipment washdown. Table 15-3 summarizes
water usage in the aluminum fluoride industry.
15.2.2 Waste Sources
Noncontact Cooling Water
Noncontact cooling water is used to cool the product coming
out of the reactor. In some cases it is recirculated and the
blowdown treated separately from other process contact waste
water or it is discharged without treatment. The water can be
monitored for fluoride and if process contamination occurs, it
can be diverted to the waste water treatment facility for
fluoride removal.
Floor and Equipment Washings N
The quantity and quality of waste water generated from
these operations varies and depend largely on the housekeeping
practices at the individual plants.
Scrubber Waste Water
This is the major source of waste water requiring treatment
before discharge or recycle to the scrubber. It is contaminated
with hydrofluoric acidr aluminum fluoride and aluminum oxide,
and, in some cases, sulfuric acid and silicontetrafluoride have
been detected. These originate as impurities in the
hydrofluoric acid used in the process. Table 15-4 presents the
waste water flows at different facilities in the subcategory.
Noncontact cooling water is excluded from consideration since it
normally does not contain pollutants.
Solid Wastes
In aluminum fluoride production, hydrofluoric gas and
solids, such as aluminum trihydrate and aluminum fluoride,
escape with the vent gases. During scrubbing, the solids are
suspended in the scrubber water, while hydrofluoric acid gas is
dissolved. In, the treatment facility, the waste water is
neutralized with lime and calcium fluoride precipitates out and
settles with other suspended solids. In the majority of cases,
the solids are retained in the lagoon for periods up to ten
years. Table 15-5 gives a summary of the amounts of solids
generated at two aluminum fluoride plants.
Different wastes from the aluminum fluoride process are
intermixed before treatment. As mentioned earlier, scrubber
water constitutes the major source of waste water in the
aluminum fluoride subcategory. If the production of aluminum
497
-------
TABLE 15-3. WATER USAGE IN THE ALUMINUM FLUORIDE SUBOYIEGORY
Source Water use per unit of production
3
(in /kkg of A1F3)
Non-contact cooling
Indirect process
Plant
# 837
14.5
12.2
Plant /9.
t 705 (Z}
NA(1)
1.15
Plant
# 188
6.95
NA
Plant
# 251 (2)
NA
NA
contact (puiqps, seals,
leakst spills)
Maintenance, e.g. 1.13 2.39 NA 1.02
cleaning and work area
washdown
ScruKber 3.45 8.92 3.46 18.7
(1) NA. = Not Available
(2) Currently not manufacturing aluminum fluoride.
498
-------
TABLE 15-4. WASTE WATER FLOW AT PLANTS #837, #705 AND #251
FOR ALUMINUM FLUORIDE SUBCATEGORY
Source Flow rate per unit of production
( m /kkg of AlFo)
Plant #837 Plant #705(4) Plant #251(4)
Scrubber water 3.45 8.92(2) 18.7(3)
Maintenance equipment 1.13 2.39 1.02-
cleaning and work area
washdown
Total raw waste flow 4.58 11.3 19.7
Average of above 11.9
three flows
(1) All flow information is from 308 Questionnaires and plant visits. Unit
flow is calculated by dividing waste water flow in m-Vday by production
in kkg/day.
(2) From Table 15-6 (see footnotes which describe basis of information).
(3) From Table 15-7 (see footnotes which describe basis of information).
(4) Currently not manufacturing aluminum fluoride.
TABLE 15-5. SOLIDS GENERATED AT PLANT #705 AND #251 PRODUCING
ALUMINUM FLUORIDE
Plant Total Solids Generated(kg/kkg of A1F-.)
< o
#705(1) , 54
#251(1) • 69
(1) Currently not manufacturing aluminum fluoride.
499
-------
fluoride is integrated with hydrofluoric acid, then the waste
waters from both plants are combined and treated.
15.3 DESCRIPTION OF PLANTS VISITED AND SAMPLED
15.3.1 Screening
Plant f705 was visited in the screening phase of the
program. Both hydrofluoric acid and aluminum fluoride are
produced at this facility by the general processes described
earlier. The waste water from the hydrofluoric acid and
aluminum fluoride plants is mixed and sent to the treatment
facility. At the treatment facility the combined waste water is
neutralized with lime and sent to a series of settling ponds.
The effluent from the last pond is given a final pH adjustment
before a portion is discharged and the rest recycled to the
process. Figure 15-2 shows a simplified block diagram of the
process including the waste water treatment facility and
sampling locations. Table 15-6 presents a summary of flow data
of the sampled streams, and the data for important classical
pollutant parameters.
15.3.2 Verification
Plant f705 was visited again and the same streams sampled
in the screening phase also were sampled and analyzed in the
verification phase. The variations in individual stream flows
were small during the two phases of sampling. Table 15-6
summarizes the flow data and important conventional and
nonconventional pollutant emissions. A second plant (Plant
$251) was visited and sampled in the verification phase. Figure
15-3 is a simplified flow diagram of the aluminum- fluoride
manufacturing plant and the waste water treatment facility
showing the sampling locations. Table 15-7 presents the flow
and pollution concentration data for the plant. The aluminum
fluoride and hydrofluoric acid waste streams are combined and
sent to a gypsum pond for suspended solids removal. The
overflow from the pond is mixed with alkaline and acid streams
from other plants for neutralization and pH adjustment before
final discharge.
15.3.3 Summary of the Toxic Pollutant Data
Following is a list of toxic pollutants which identifies
their maximum concentration levels as found in the raw process
waste streams sampled during screening and verification.
500
-------
varan
VEHf
o
H
'AQUEOUS
IIP
10
TIERE
\
'»
SCHU8BEK
VTER
1
REM2TOR
EFFLUENT
IS DISCHARGE
IBGBP
Haste atreans sanpled.
Figure 15-2. General prowess flew diagram at Plant 1705 showing the sampling points
(aluminum fluoride manufacture).
-------
15-6. BDOW AND OF TOE
STRERMS FOR PMBT §705 PBODUCENG MXMEMM FLUORIDE
Saitpling
Phase
Screening
Sanpled
Stream
No.
3
4l)
3&4
5
Verifica- 3
tion Sampling ...
3&4
5
Sanpled
Stream
Description
MF-s scrubber
Surface drains,
cooling tower,
blowdown, etc.
Total raw waste
load
(2)
Treated waste x '
AlFg scrubber
Surface drains,
cooling tower,
blowdown, etc.
Total load
Treated waste ' '
Unit
Flow
(m3/kkg)
8.92
2.39
11.3
24
8.92
2.39
11.3
24
Total
Suspended
Solids
(mg/1) (3) (kg/kkg) W
13,000 120
200 0.48
11,000(5) 120
80 2.0
1,400 13
200 0.48
1,200(5) 13
2.0 0.048
Fluoride
(3) (4)
(mg/1) (kg/kkg)
530 4.7
350 0.82
490 5.5
70 1.6
1400 12
170 0.40
1100 13
20 0.55
Muminum
(3) (4)
(mg/1) (kg/kkg)
780
40
620
10
460
27
370
1.0
7.0
0.10
7.1
0.17
4.1
0.060
4.1
0.012
en
o
to
(1) Consists of waste water from HF and A1F-, process. Flow indicated is estimated portion of total
flow contributed by MF3 nmintenance and washdown waste water from 308 Questionnaire,. Total flow
is 17.8 nr/kkg of product for both process wastes combined,
(2) Consists of waste water from HF and A1F3 process. Plant currently not manufacturing A1F3.
(3) Average of three daily composite samples during verification and single value obtained during
screening.
(4) kg/kkg of AlF^. (5) Weighted average based on unit flows.
-------
VENT
U1
o
VENT
DUST
COLLECTOR
H2S04
WET
SPAR"
SPAR DRYING
HANDLING
LOSSES
HOSE DOWN
WATER
r
HF KILN
WASTE
AIR
OR|J>
ACID
WATER
LEGEND
^ SAMPLING POINTS.
AHF
PURIFICATION
DILUTION
WATER ;
A!F3 PLANT
HOSE DOWN
t
HOSE DOWN WATER
AHF PLANT
NEUTRALIZATION
SYSTEM
WATER
_J
EFFLUENT
TO RIVER
ALKALINE STREAMS
AND ACID FROM OTHER PLANTS
Figure 15-3, General process flow diagram at Plant #251 showing the sampling points,
(aluminum fluoride manufacture).
-------
OSffiEE 15-7. HOT MD POELOEftNT OCMENTRATION DAEA. OP THE SAMPLED STOEM6
FOR PLANT 1251 PEODOCTOG fflUMMM EHX3RIDE
Stream
No.
Sanpled
Stream
Description
Obit
Plow
Cm3/kkg
of A1P3)
Verification
Sampling
4
6
45j6
2
3
AlP- fecrufcfcer
water
S02 scnabber
water '^
•Dotal raw waste
load
GypstM pond
influent^2'
Q/psum pond
effluent(25
12.6
6.10
18.7
25.1
25.1
Total
Suspended
Solids
(ng/1) (kg/kkg)
1200 16
0.0 0.0
1200 16
19,000 470
9.0 0.23
Fluoride
(mg/1) (kg/kkg)
470 5.90
20 0.14
320 6.0
660 17
320 8.0
Aluminum
(rag/1) (kg/kkg)
50 0.60
0.20 0.0010
50 0.60
26 0.65
22 0.55
8
(1) One half flow of SC>2 scrubber water is assumed to contribute to the AlP^ process since the
total flow is oomnon to the AU^ and HP process.
(2) Consists of hydrofluoric acid and aluminum fluoride waste water. Plant currently not
manufacturing A1F-,.
-------
Maximum Haw Waste Concentrations Observed
Pollutant Screening Verification
Plant |705 Plant #705 and f251
Arsenic
Selenium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Cadmium
Antimony
Beryllium
200
68
70
120
25
1.6
150
450
0.70
0
0.80
480
97
1100
250
91
11
290
450
33
3
0
.0
.80
Section 5.1.2 of this report describes the methodology of
the screening and verification sampling program. In the
aluminum fluoride industry, seven days of sampling were
conducted at Plants ^705 and f251. Seven sampling points were
identified and studied for the subcategory. The evaluation of
toxic pollutant content of these process-related waste streams
was based on 637 analytical data points. The screening for
toxic organic pollutants at Plant #705 generated an additional
645 analytical data points. The daily raw waste loads were
calculated ' from the waste stream flow rates measured or
estimated at the time of sampling and the measured pollutant
concentration.
That is,
Daily loading (as kg of pollutant per day) =
1000
Where:
C is the concentration of the pollutant expressed as
mg/1 (Note: kg/m3 = 1000 mg/1) , and
Q is the Aluminum Fluoride process - waste stream flow
rate expressed as m3/day. (m3, a cubic meter, is equal
to 264.2 U.S. gallons)
505
-------
Similarly, the unit loadings were calculated from the
reported aluminum fluoride production rate, the waste stream
flow rate, and the measured pollutant concentration.
Unit loading (as kg of pollutant //-,%
per kkg of aluminum fluoride) = -i — '
1000P
Where C and Q are the same as described above, and P is the
aluminum fluoride production rate expressed as kkg/day.
(kkg is 1000 kg, a metric ton, which is equal to 2205 Ibs.)
The P and Q factors are for the Aluminum Fluoride Process
and thereby the Agency has segregated that portion of the
effluent attributable only to the Aluminum Fluoride
Process.
Table 15-8 and 15-9 are a tabulation of the raw waste and
treated toxic pollutant concentrations and loads determined
during the three plant visits. The loads and concentrations are
based on the average of three composite samples during
verification and one composite sample during screening. These
unit loads were used to determine the minimum, average, and
maximum unit loading valves presented in Table 15-10.
Based on the total annual production of 134,700 kkg/year in
this subcategory and the average waste load generated per unit
product in Table 15-10, the estimated total toxic pollutant raw
waste loads generated each year for this subcategory are as
follows:
_ Pollutant _ Waste Load (kg/year)
Arsenic 180
Selenium 140
Chromium 400
Copper 94
Lead 20
Mercury 3.0
Nickel 180
Zinc 140
Cadmium 11
Antimony 0.70
Beryllium 0.30
506
-------
15-8. TOXIC POLLOTRMT wm imns AND
StBCATEGORy
AHMHM HTOKEDE
Pollutant
Arsenic
Selenium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Cadmium
Jtotiitony
Beryllium
Screening
Plant
fag/i) (1)
0.18
0.050
0.030
0.10
0.0050
0.00040
0.11
0.16
0.00020
_J3)
0.00020
#705
(kg/kkg) (2)
0.0020
0.0010
0.00030
0.0010
0.00010
0.0000040
0.0010
0.0020
0.0000020
__<3)
0.0000020
Verification
Plant
(mg/D
0.18
_J3)
0.44
0.070
0.020
0.00040
0.22
0.080
0.010
0.00040
"""*
#705
(kg/kkg)
0.0020
— <3)
0.0050
0.0010
0.00020
0.0000050
0.0030
0.0010
0.00020
0.0000050
— ™
Plant
(mg/1)
0.020
0.050
__(3)
0.010
0.010
0.0030
0.010
0.020
""""•
#251
(kg/kkg)
0.00030
0.0010
__(3)
0.00010
0.00010
0.000050
0.00020
0.00030
..(3)
__^
Average
Concentration
fag/1)
0.13
0.050
0.24
0.060
0.012
0.0013
0.11
0.090
0.0050
0.00040
0.00020
U1
o
(1) Concentrations based on average raw waste loads shown and total process production and waste
flows.
(2) kg/kkg of product.
(3) — below analytical detection limit. *
-------
TABLE 15-9. TOXIC POLLUTANT EFFLUENT CONCENTRATIONS DURING SAMPLING
SUBCATEGORY
ALUMINUM FLUORIDE
Pollutant
Arsenic
Selenium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Cadmium
Antimony
Beryllium
Plant and Sampling Phase
#705
Screening
fag/D
HD(1)
ND
0.0070
0.10
0.0020
ND
0.050
0.0020
0.0020
ND
0.0020
#705
Verification
fag/D
ND
ND
0.040
0.0010
0.020
ND
ND
0.0010
0.0010
ND
ND
#251
Verification
fag/D
0.0050
0.070
0.22
0.070
0.030
ND
0.45
ND
ND
ND
ND
Average
fag/D
< 0.0050
< 0.070
0.090
0.060
0.020
ND
< 0.25
0.0020
< 0.0020
ND
< 0.0020
(1) ND — Not Detected.
508
-------
TABLE 15-10.
OF RAW WRSEB LOADINGS POUND IN SCHEMING AND VERIFICATION
SDBCATEGOBy
Pollutant
Ttoxic
Arsenic
Selenium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Cadmium
Antimony
Beryllium
Conventional and
NDnconventional
TSS
Fluorine
Aluminum
Loading Bange,
kg/day
Minimum Maximum
0.050
0.030
.0.020
0.020
0.0030
0.026
0.026
0.040
0.00010
NA(2)
NA
600
250
100
0.080
0.16
0.22
0.050
0.020
0.0080
0.12
0.080
0.0070
0.00020
0.00010
5400
980
320
Minimum
0.00030
0.0010
0.00030
0.00010
0.00010
0.0000040
0.00020
- 0.00030
0.0000020
NA
NA
13
5.5
0.60
Unit Loading,
Average
0.0013
- O.OOJ.O
0.0030
0.00070
0.00015
0.000020
0.0013
0.0010
0.000080
0.0000050
0.0000020
50
8.1
3.9
kg/kkg
Maximum
0.0020
0.0010
0.0050
0.0010
0.00020
0.000050
0.0030
0.0020
0.00020
NA.
NA
119.0
13.0
7.0
No. of
Plants
Averaged
3
2
2
3
3
3
3
3
2
1
1
3
3
3
Ul
o
(1) Average unit loadings from Table 15-8.
(2) Not Applicable
-------
15.4 POLLUTION ABATEMENT OPTIONS
15.4.1 Toxic Pollutants of Concern
The toxic pollutants found in actual plant waste waters are
lead, mercury, cadmium, antimony, beryllium, copper, arsenic,
chromium, nickel, zinc, and selenium, in the case of selenium,
it is apparent that the source was the raw water supply and is
therefore not regarded as a process-related pollutant, but
control of selenium in the treated effluent may be required.
The ones of most concern are chromium and nickel.
Copper and chromium may be present as trace impurities in
the hydrofluoric acid used to react with bauxite to form
aluminum fluoride. Arsenic, zinc, and nickel may originate as
impurities in the bauxite ore. Waste treatment processes should
be designed to control TSS, fluoride, and the significant toxic
metals. Lead, mercury, cadmium, antimony, and beryllium are
eliminated as toxic pollutants of concern because levels
observed are too low to be considered treatable.
15.4.2 Process Modifications and Technology Transfer options
1. Total recycle of waste water to the scrubbers is
feasible if final neutralization is with soda ash. The calcium
in the waste is precipitated as calcium carbonate and scaling
problems in pipes and scrubbers are reduced.
2. passage of the vent gases from the reactor through a
cyclone prior to scrubbing with water will remove the aluminum
oxide and aluminum fluoride particulates. The collected
material in the cyclone can be recycled to the reactor. The
installation of a cyclone will result in material recovery and
will also reduce the suspended solids load going to the waste
water treatment facility.
15.4.3 Best Management Practices
1. Rainfall runoff in plant areas, treatment facilities
and other places susceptible to fluoride contamination can be
collected and sent to the waste water treatment facility.
2. If solid wastes containing fluoride are stored on land,
studies should be conducted to ascertain the risk of
contaminating ground water. Where necessary, provisions can be
made for collection and treatment of leachate, permeate, and
runoff.
3. Settling ponds in the waste water treatment facility
should be deep enough (or provided with baffles) to eliminate or
reduce turbulence caused by wind and rainfall. This will reduce
the incidence of weather-related plant upsets, and suspended
solids limitations will be met more consistently.
510
-------
15.4.4 Prevailing Control and Treatment Practices
Plant |705 practices lime neutralization and settling of
the waste waters. Since aluminum fluoride production is
integrated with hydrofluoric acid production, the waste waters
from the two processes are combined before treatment. The plant
does not treat noncontact cooling water.
At Plant §837 the tail gases are scrubbed with soda ash
solution, and the resulting solution is sent to an adjacent
facility for use. The water from the wet scrubbers on the
hydrated alumina dryers are also sent to an adjacent facility
for use. The waste waters from area washdown are combined with
other product waste water, treated with hydrated lime and sent
to a settling lagoon before discharge.
Plant |188 produces aluminum fluoride in small quantities
and in batches. The waste water from the batch operation is
first sent to a collection pond. It then goes to a second pond
where lime and alum are added and it finally enters a third pond
where the pH is adjusted by recarbonation.
Plant f251 mixes the aluminum fluoride waste with
hydrofluoric acid plant waste. The combined waste water is sent
t° gypsum ponds for suspended solids removal. The supernatant
is treated with an effluent from another plant for pH control
and neutralization. Because of the presence of complex
fluorides (from the HF process) in the waste waters, the plant
is planning to use a new proprietary process in the near future
to further reduce fluoride levels in the final effluent-.
15.4.5 Advanced Treatment Technologies
Metal ions can be precipitated as hydroxides at alkaline pH
levels, and in clarified solutions they may be exchanged for
hydrogen or sodium ions by ion exchange. Metal ions at low
levels may also be controlled by xanthate precipitation,
although the process is not widely used.' Sulfide precipitation
will reduce copper, nickel, and zinc to low levels but will not
control chromium or arsenic. Although the mechanism is not
clear, arsenic levels appear to be reduced -in the lime
neutralization process followed at most plants, perhaps by
entrapment or adsorption of the oxide during the precipitation
of calcium fluoride. A combination of lime and ferric sulfate
coagulation is probably the most effective and practical method
for reducing arsenic concentrations.
511
-------
15.5 SELECTION OF APPROPRIATE TECHNOLOGY AND EQUIPMENT
15.5.1 Technologies for Different Treatment Levels
Level 1 (BPT)
Neutralization with lime is widely used in the industry to
remove the primary nonconventional pollutant as calcium
fluoride. Because lime neutralization to pH 10 results in
significant incidental removal of toxic pollutants, alkaline
precipitation was chosen as BPT (Level 1) technology. The flow
diagram is shown in Figure 15-4.
Level 2 (BAT and NSPS)
A higher removal of suspended metal hydroxides, TSS, and
CaP2 can be achieved by adding dual media filtration to the
Level I system. The flow diagram is shown in Figure 15-5.
Level 3
Sulfide precipitation is added to the proposed BAT level of
treatment to attain a higher level of heavy metal removal.
Chromium and selenium levels are not appreciably reduced
although other toxic pollutant levels are. The flow diagram is
shown in Figure 15-6.
Level 4
The technology is similar to Level 2, except that soda ash
is substituted for part of the lime treatment, permitting
partial recycling of effluent. Eighty percent recycle has been
demonstrated and is used in the development of plant performance
estimates. The flow diagram is shown in Figure 15-7.
15.5.2 Equipment for Different Treatment Levels
Equipment Functions
Level 1 consists of flow equalization with first stage lime
application followed by second stage lime application and lagoon
settling. The final pH is adjusted with hydrochloric acid to
the 6-9 range before discharge through an effluent monitoring
system.
In Level 2, dual media filtration is added to provide
better control of suspended solids, including heavy metal
hydroxides, which are returned to the .lagoons as filter
backwash.
512
-------
in
H
UMS
O
-••Oh
BW
HASTE WATER
i WXJUSTOENT
MIXING
XA90CH
I fc^H-•»!
'Includes flow Monitoring, pH monitoring and sampler.
Figure 15- 4. Level 1 waste water treatment for the aluminum fluoride subcategory.
-------
r
BSOKKRSH
in
*»
^
RBH
WASTE
1
HATER
r--@-
i
i
i
"
/
^_ /^
— »* /*-"*•»
ECWALIZATICM 1
L»^>T_fc
• ««_
«~ — •
|
MDCIN;
1
1
•jl
i!
-L§*
; f
r
• i ,j —
^Includes flow ncnitoring, jtt nonltxirlng an! aanpler.
Figure 15-5. Level 2 waste water treatment for the aluminum fluoride subcategory.
-------
FEtVUIS SUtHOE
Ul
H
tn
MM
msrz-
*A i Z
r-
-ii
i'A
ll1^-
daw manttorlng, pff monitoring and lumpier
X
Figiure 15-6. Level 3 waste water treatment for the aluminum fluoride subcategory.
-------
!
1
r
i
1
i
i
X
1
W
I
I
H
ref
I
&
r-
in
H
516
-------
In Level 3, ferrous sulfide is prepared on site from
ferrous sulfate and sodium bisulfide and is added ahead of the
dual media filter shown in Level 2, to reduce heavy metals
(except chromium) to lower levels by sulfide precipitation.'
Level 4 is a modification of Level 2 which allows partial
recycling of final effluent by substituting soda ash for part of
the lime treatment, and settling the resulting calcium carbonate
in a clarifier before filtration. This step reduces the calcium
saturation and permits recycling of effluent without serious
scaling problems. Although a small blowdown of effluent is
maintained for control of salinity the total mass discharge of
toxic pollutants is less than that achieved in Level 2 due to
the lower effluent flow rate.
Chemicals and Handling
In BPT (Level 1) and in Level 2, two-stage neutralization
is accomplished with lime alone, using conventional handling
equipment to deliver milk of lime to two points of application.
In Level 3r a mixture of ferrous sulfate and sodium bisulfide is
prepared in a well-ventilated space and applied with a
conventional solution feeder to the inlet of the Level 2 dual
media filter. With adequate ventilation and proper pH control
in this chemical preparation, there are no unusual problems in
chemical handling. In Level 4, soda ash is used to furnish part
of the alkalinity, employing conventional dry chemical feeding
equipment for this nonhazardous chemical.
Separation and Removal of Solids
At all levels of treatment the precipitated solids are
removed mechanically from the lagoons at regular intervals and
are piled in self-draining areas near the lagoons, on land
provided for a ten-year operating period. Fluoride and toxic
pollutants are in the insoluble or adsorbed form and do not
constitute a hazard to the local environment when left at the
plant site under controlled conditions, i.e., with leachate and
permeate control.
Monitoring Requirements
Control of fluoride and toxic pollutants in the treatment
process can be reasonably assured by pH and fluoride ion field
testing equipment. At advanced"levels very low values of toxic
metals are detected best by atomic absorption methods, normally
performed in commercial laboratories on carefully collected and
composited samples.
517
-------
15.6 TREATMENT COST ESTIMATES
15.6.1 General Discussion
A model plant concept was developed for the subcategory for
treatment cost estimation purposes. The proposed BPT treatment
system specifications are outlined subsequently.
Waste Mater Flow
The range of waste water data on file shows flow variations
from 4.58 m3/kkg of A1F3 to 19.7 m3/kkg of A1F3 (see Table
15-4). Based on these values, a unit flow of 11.9 m3/kkg of A1P3
was taken as the average for the waste water treatment model
plant for cost estimating purposes.
Production
Six plants manufacture aluminum fluoride at a total
production rate of 120,000 kkg/yr. Individual plant production
rates range from a minimum of 38 kkg/yr to a maximum of 45,600
kkg/yr with an average of 24,300 and a median of 35,500 kkg/yr.
For waste water treatment cost estimates, three production
levels were selected as model plants. These three models
reflect the production levels of the plants for which data is on
file (excluding a small batch operation plant) and are 17,500
kkg/yr, 39,200 kkg/yr and 50,400 kkg/yr.
Pollutant Loadings
Observed pollutant loadings varied from 14 to 27 kg/kkg of
A1F3 for suspended solids and from 5.4 to 39.5 kg/kkg of A1F3
for fluoride. The data sources are as follows:
Source of Data TSS (kg/kkg-AlF3) F (kg/kkg-AlF3)
EPA Document 1974 Ref- 16-20 15-20
Screening and
Ver ification
Phase - Plant Data 14-27 5.4-40
For model plants, pollutant loadings of 20 kg of total
suspended solids and 18 kg of fluoride per kkg of A1F3 were used
to establish treatment requirements.
Treatment Chemicals
Lime (CaO powder form) is added to precipitate fluoride and
to raise the pH to a six to nine range. For each of the model
plants, lime is added at 25 percent above the stoichiometric
518
-------
requirements for fluoride precipitation. For advanced
treatment, ferrous sulfide is added to give a concentration of
10 ppm. This acts as a polishing step to remove additional
trace metals from the effluent. For a more advanced level of
treatment, soda ash is added in addition to lime (CaO) . The
soda ash dosage was assumed to be 770 kg/kkg.
Variation in Flow and Pollution Loading
To indicate the effect on costs of higher and lower
pollutant loadings, cost estimates were developed for one model
plant (35r600 kkg-A!F3/yr) at 27 kg of TSS/kkg-AlF3 and 30 kg
fluoride/kkg-AlF3 and 14 kg fluoride/kkg-AlF3. The waste water
flow for these additional estimates was held constant as in the
original mode (i.e., 15 m3/kkg-AlF3). Unit flows were also
varied to monitor the sensitivity of cost to plant size. In
this case, the pollutant loadings were assumed to be the same as
in the original model. The range of waste water flows used were
10.1 m3/kkg to 22.8 m3/kkg.
Generation of Solids
From the pollutant loadings and treatment chemi.-.als above,
the waste treatment residue consists of 20 kg/kkg of suspended
solids plus 46.2 kg/kkg from added chemicals. Thus,' the total
solids generated are 66.2 kg/kkg of product. After mechanical
removal to self-draining piles, the combined fluoride As(CaF2)
is reasonably stable at the reaction pH reached during lime
treatment.
Cost Estimates
The estimated costs for models having three different
production and four levels of treatment are given in Tables
15-11, 15-12 and 15-13. For these models, both the hydraulic
and pollution loadings per unit of production are held constant
over the entire range of production. Annual treatment cost as a
function, of production is shown graphically in Figure 15-8.
Similarly, treatment cost per metric ton of product is given in
Figure 15-9.
To indicate the effects on cost of varying the pollutant
load per unit of product, cost estimates were developed for one
medium-size production model plant at higher solids and
pollutant (fluoride) loadings. For these models the hydraulic
load per unit of production was held constant. The cost
estimates for these models are given in Tables 15-14 and 15-15.
The effects on costs of varying the unit pollutant load are
shown graphically in Figures 15-10 and 15-11 at Levels 1 and 4.
Variation of pollutant loads has a significant impact on Level
1, but had no effect on the incremental costs of treatment at
Levels 2 and 3.
519
-------
TADLE 15-11. MODEL PLANT TREATMENT COSTS
Subcategory ALUMINUM FLUORIDE
(1)
Production 15,900 metric tons par year (17
45 metric tons per day (50
Waste water flow 540 cubic meters per day.
A. INVESTMENT COST
Equipment in place,
including piping,
fittings, electrical
work and controls.....
Monitoring equipment
Engineering design
and inspection.
Incidentals, overhead,
fees, contingencies...
Land
TOTAL INVESTMENT COST
8. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Enerdv. ...............
Maintenance. ..........
Taxes and insurance...
Residual waste
disposal.
Monitoring, analysis
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OP
INVESTMENT COST
v TOTAL ANNUAL COST
FIRST
$39,800
192,000
9,000
48,160
48,160
24,000
$361,120
$56,000
3,400
35,000
33,712
10,833
5,400
15,000
$159,345
$54,849
$214,194
LEVEL OF
SECOND
$10,000
63,000
15,600
15,600
$109,200
$14,000
600
10,920
3,276
7,500
$36,296
$17,766
$54,062
(1)
,529 tons per year)
tons per day)
(2)
TREATMENT
THIRD
$14,000
74,000
17,600
17,600
$123,200
$14,000
900
800
12,320
3,696
7,500
$39,216
$20,044
$59,260
FOURTH
$20,500
172,000
38,500
38,500
$269,500
$14,000
2,500
9,800
26,950
8,085
"
7,500
$63,835
$43,847
$112,682
(1) Production year is 350
(2) First level represents
Other levels represent
days.
the base cost of treatment system.
the incremental cost above base cost.
520
-------
TABLE 15-12. sKJDSL PLANT miAWENT COSTS
Subcategory ALUMMlM FLUORIDE
1
(I) (I)
Production 35,600 metric tons per year (39,249 tons per year)
101 metric tons per day (112 tons per day)
Waste water Clow 1200 cubic meters per day.
A. INVESTMENT COST
Equipment in place,
Including piping.
fittings, electrical
Monitoring equipment
Engineering design
Incidentals, overhead,
Sees, contingencies...
tand.
TOTAL INVES1MSST COST
a. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Taxes and insurance...
Residual waste
Monitoring, analysis
and reporting.........
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTISATION OF
INVESTMENT COST
TOTAL ANNUAL COST
FIRST
$63,600
238,000
9,000
62, 120
62,120
42,000
1476,840
$56,000
5,500
80,000
43,484
14,305
12,500
15,000
$226,789
$70,748
$297,537
LEVEL OP
SECOND
$15,000
84,000
19,800
19,800
$138,600
$14,000
900
13,860
4,158
7,50i
$40,418
$22,550
$62,968
(2)
TREATMENT
THIRD
$19,000
90,508
21,900
21,900
$153,300
$14,000
1,300
1,800
15,330
4,599
7,500
$44,529
$24,941
$69,470
P0URTH
$34,000
259,000
58,600
58,600
$410,200
$14,000
3,100
18,800
41,020
12,306
7,500
$96,72
-------
TABLE 15-13. MODEL PLANT TREATMENT COSTS
Subcategory ALUMINUM FLUORIDE
(1) (U
Production 45,800 metric tons per year (50,494 tons per year)
130 metric tons per day (144 tons per day)
Waste water flow 1550 cubic meters per day.
A. INVESTMENT COST
Equipment in place,
including piping,
fittings, electrical
Monitoring equipment
Engineering design
Incidentals, overhead.
fees, contingencies...
Land
TOTAL INVESTMENT COST
8. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Enerav. ...............
Taxes and insurance. . .
Residual waste
Monitoring, analysis
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
' INVESTMENT COST
TOTAL ANNUAL COST
FIRST
$76 , 500
281,000
9,000
73,300
73,300
60,000
$573,100
$56,000
7,400
100 , 000
51,310
17,193
16,000
15,000
$262,903
$83,481
$346,384
LEVEL OF
SECOND
$20,500
110,000
, *.«.»„« — -
26,100
26,100
$182,700
$14,000
1,500
18,270
5,481
7,500
$46,751
$29,725
$76,476
(2)
TREATMENT
THIRD
$24,500
116,500
28,200
28,200
$197,400
$14,000
1,900
2,400
19,740
5,922
7,500
$51,462
$32,116
$83,578
FOURTH
$43,000
317,000
72,000
72,000
$504,000
$14,000
4,300
26,400
50,400
15,120
7,500
$117,720
$82,000
$199,720
(1) Production year is 350
(2) First level represents
Other levels represent
days.
the base cost of'treatment system.
the incremental cost above base cost.
522
-------
500
o
o
H
X
400
300
200
100
OIVEI.
10 20 30 40 so
HO3UCTIGN (METRIC TONS/HEAR X 1000)
Figure' 15-8. Annual treatment cost vs. production for the
Fluoride Subcategory
523
-------
20
15
\
•to-
\\
|V
\
\i\
\\
10
13
10 20 30 40 §0
HODOCTICN (METRIC TON/YESR X 1000)
Figure 15- 9. Jtanual unit treatment cost vs> pioduction for the
Aluminum Fluoride Subcategory
524
-------
TABLE 15-14. MODEL KANT TREATMENT COSTS
(3)
Subeategory AUMMtM FLUORIDE
(I) (1)
Production 35,600 metric tons per year (39,249 tons par year)
101 metric tons per day (112 tons per day)
Waste water flow 1200 cubic meters per day.
A. INVESTMENT COST
Equipment >in place,
including piping,
fittings, electrical
Monitoring equipment
in place. .............
Engineering design
and inspection* .......
Incidentals , overhead,
fees, contingencies...
Land.........
TOTAL INVESTMENT COST
8. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Snarctv. ...............
Chemicals. ............
Taxes and insurance...
Residual waste
disposal. .............
Monitoring, analysis
and reporting. ........
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
DIVESTMENT COST
TOTAL ANNUAL COST
FIRST
$82,000
241,000
9,000
66,400
66,400
66,000
$530,800
$56,000
5,500
130,000
46,480
15,924
19,000
15,000
$287,904
$75,622
$363,526
LEVEL OF
SECOND
$15,000
84,000
19,800
19,800
$138, S0i
$14,000
900
13,860
4,158
7,500
$40,418
$22,550
$62,968
(2)
TREATMENT
THIRD
$19,000
90,500
21,900
21,900
$153,300
$14,000
1,300
1,800
15,330
4,599
7,500
$44,529
$24,941
$69,470
FOURTH
$34,500
270,000
60,900
60,900
$426,300
$14,000
3,100
31,500
42,630
12,789
7,500
$111,519
$69,359
$180,878
(1) Production year is 350
(2) First level represents
Other levels represent
(3) Sensitivity Analysis -
days.
the base cost of treatment system.
the incremental cost above base cost.
increased pollutant load.
525
-------
TABLE 15-15. MODEL PLANT TREATMENT COSTS
(3)
Subcategocy ALlMBilM FLUORIDE
(1) (I)
Production 35,600 metric tons per year (39,249 tons per year)
101 metric tons per day (112 tons per day)
Waste water flow 1200 cubic maters per day.
A. INVESTMENT COST
Equipment in place.
including piping.
fittings, electrical
Monitoring equipment
Engineering design
Incidentals, overhead,
Eees, contingencies...
Land
TOTAL INVESTMENT COST
8. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Taxes and insurance...
Residual waste
disposal. .............
Monitoring, analysis
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OP
INVESTMENT COST
TOTAL ANNUAL COST
FIRST
$56,900
221,000
9,000
57, 380
57,380
30,000
$431,660
$56,003
5,508
60,000
40,166
12,949
9,000
15,000
$198,615
$65,350
$263,965
LEVEL 0?
SECOND
$15,000
34,300
19,800
19,800
$138,600
$14,000
900
13,860
4,158
7,500
$40,413
$22,550
$62,968
(2)
TREATMaiT
THIRD
$19,000
90,500
21,900
21,900
$153,300
$14,000
1,300
1,800
15,330
4,599
7,500
$44,529
$24,941
$69,470
FOURTH
$34,000
259,000
58,600
58,600
$410,200
$14,000
3,100
14,610
41,020
12,306
7,500
$92,536
$66,739
$159,275
(I) Production year is 350 days,
(2) fist level represents the base cost of treatment system.
Other levels represent the incremental cost above basa cost,
(3) Sensitivity Analysis - decreased pollutant load.
526
-------
20
15
•to-
10
•
.
•
,
L
/
/
DT
^
s
/
/
w
\
\
\
/
»llc
\
s
s
/
RT1
»
*^
\
s
> 1
s
s,
'O
\
\
X
r.T
(
X
s
V
m
'
^
w
tf
X,
•vj
—
IT
/
*&
f>
•t
/-
v^
-»v
"^*
i
rJrw
TT
v.
^.
^
D
i
ITT
*s»
"^
^
?R
•^*.
""^
"•••I
^.S
«.
(7y
Fir
T>
r.ii
m
w
IT
T<
,
^A
h
A1F3 PRODUCTION (METRIC TONS/YEAR X 100Q)
Figure 15-10. Effect of variation of pollutant load on treatment
cost at level 1 technology
527
-------
20
k
ji
S. Ell
I
D?D
fc
15
\
\
10
** D'JlC^EMiElb
NT
IQ?D
10 20 30 40 SO 60
A1P3 PW1DDCTICN (MEaKCC TCNSAEaR X 1000)
Figure 15-11.- Effect of variation of pollutant load on treatment
cost at level 4 technology
528
-------
To judge the effects on cost of varying the hydraulic load
per unit of production, cost estimates were developed for one
medium-size production model plant at a higher and a lower
hydraulic loadings. The pollutant load per unit of production
was held constant for these models. Tables 15-16 and 15-17 show
the cost estimates. At treatment Levels 2, 3, and 4 the effects
on costs of varying the per unit hydraulic load are shown
graphically in Figures 15-12, 15-13, and 15-14. Hydraulic load
variation had no significant effect on the costs of treatment at
Level 1. Table 15-18 presents a summary of the unit cost
distribution between amortization and the operation and
maintenance cost components at various production and levels of
treatment. The effects on cost due to variations in unit
pollutant and hydraulic loads are also shown in Table 15-18.
At the first level of treatment, chemicals, labor, and
amortization have significant impact on the annual costs. At
the second, third and fourth levels of treatment, the operation
and maintenance cost comprises approximately two-thirds of the
additional annual costs, and the remaining one-third is due to
amortization.
Effects on annual costs arising from higher and lower
pollutant loads per unit of product for a medium level of
production model plant were studied. At high pollutant loading,
the annual cost at the first and fourth levels of treatment
increased approximately by 25 and 35 percent, respectively, over
the base case cost. At the second and third levels of
treatment, annual costs per unit of product are the same as for
the original model.
At lower pollutant loadings, annual cost at the first level
of treatment decreased by 15 percent below the base case cost.
At other levels, annual costs per unit of product are the same
as for the original model.
The annual costs arising from higher and lower hydraulic
load per unit of product for a medium level of production model
indicated that at the first level of treatment, variation of
hydraulic loads had an insignificant impact on annual cost
compared to the original model annual cost.
In the second, third, and fourth levels of treatment, at a
higher hydraulic load, additional annual costs per unit of
production increased by 24, 21, and 18 percent respectively over
the original model costs.
At a lower hydraulic load, additional annual costs per unit
of production decreased by 10 percent at second and third
levels, and by 16 percent at the fourth level, compared to the
original model cost.
529
-------
TABLE 15-16. TOSL PLANT TREATMENT COSTS(3)
Subcategocy ALIMINIM FLUORIDE
(I) (I)
Production 35,603 metric tons per year (39,249 tons per year)
101 metric tons per day (112 tons per'day)
Waste water flow 2300 cubic meters per day.
A. INVESTMENT COST
Equipment in place,
including piping.
fittings, electrical
Monitoring equipment
Engineering design
Incidentals, overhead,
fees, contingencies...
Land.............
TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Enerov. ...............
Taxes and insurance...
Residual waste
Monitoring, analysis
TOTAL OPERATION AMD
MAINTENANCE: COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
TOST
$66,100
256,000
9,000
66,220
66,220
42,000
$505,540
$56,000
7,400
80,000
46,334
15,156
12,500
15,000
$232,420
$75,417
$307,837
LEVEL OP
SECOND
$21,000
117,600
27,720
27,720
$194,040
$14,300
1,500
19,404
5,821
7,500
$48,225
$31,570
$79,795
(2)
TREATMENT
THIRD
$25,000
124,000
29,800
29,800
$208,600
$14,000
1,900
1,800
20,860
6,258
7,500
$52,318
$33,939
$86,257
FOURTH
$43,500
321,000
72,903
72,900
$510,300
$14,000
4,700
18,800
51,030
15,309
7,500
$111,339
$83,025
$194,364
(!) Production year is 350
(2) First level represents
Other levels represent
(3) Sensitivity Analysis -
days.
the base cost of treatment system.
the incremental cost above base cost.
increased hydraulic load.
530
-------
TABLE 15-17. MODEL PLANT TREATMENT COSTS ^
Subcategory ALIMDKM fLUORIDE
(I) (1)
Production 35,600 metric tons per year (39,249 tons per year)
101 metric tons per day (112 tons per day)
Waste water flow 1020 cubic meters per day.
(2)
LEVEL OF TREATMENT
A. INVESTMENT COST
Squiptent In place,
including piping,
fittings, electrical
"tonitorinq equipment
Engineering design
Incidentals, overhead,
fees, contingencies...
Land.
TOTAL INVESTMENT COST
3. OPERATION AND
MAINTENANCE COST
tabor and supervision.
Taxes and insurance...
Residual waste
Monitoring, analysis
TOTAL OPSRAf ION AND
MAMTENAICE COST
C. AMORTIZATION 0?
INVESTMENT COST
TOTAL ANNUAL COST
flRST
$63,600
237,000
9,000
61,920
61,920
42,000
$475,440
$56,000
5,500
80,000
43,344
14,263
12,500
15,000
$226,607
$70,520
$297,127
SECOND
$14,500
73,300
16,960
16,960
$118,720
$14,000
600
11,872
3,561
7,500
$37,533
$19,315
$56,848
THIRD
$18,500
76,000
18,900
18,900
$132,300
$14,000
900
1,800
13,230
3,969
7,500
$41,399
$21,525
$62,924
FOURTH
$30, 000
206,000
47,2i0
47,200
$330,400
$14,000
2,500
18,800
33,040
9,912
7,500
$85,752
$53,756
$139,508
(1) Production year is 350
(2) First level represents
Other levels represent
(3) Sensitivity analysis -
days.
the base cost of treatment system.
the incremental cost above base cost.
decreased hydraulic load.
531
-------
zo
IS
I?
10
!
f
4.
/
/
a
/
x:
/
SE
k
f
H
\
\
,
>
\
\
\ N
\
/
El
\
^
s
^
/
) 1
\
^
\
s
IY
s
(\
>
DE
•
\
s\
M
\.
s^
L:
+
\
>s
N
:c
N
^x
N.
L
s
V
^
Ob
s
Is
im
TS
D
S-
1 ^
s^
-I
*«.
•*>
^«
X.
•»
*^
**•
1
SE
^-**
""""•J
"%i
i&
•«.
~~(i
~~*
!E
h
p
En
D^
yy
E.
EC
I
s
d
^D
10 20 30 40 SO 60
MP3 PBGDUCTTCN (B5ETRIC TClK/IiaR X 1000)
Figure 15-12. Effect of variation of hyojcaulic load on treatment
cost at level 2 technology
532
-------
ANNOMi TREATMENT COST ($/METRIC TON)
t- t- fJ
01 O W> O
Zl
/*
m
-
/
T"
/
?F
/
A?
I
*
n
1
^
(
*
FT
n?
.
.
\
l
\
i
\
j
^
) l
D
\
\
s
rv
\
^
\
N k
DR
\
S
V
^
PiT
\
\
rrr
\
v,
s
rr
\
V
>
Vr
y
vj x^-
^
X
XI
«S
Vkg
/
'
k.
•x.
"•»
/
^
^
^
»
">•
"*
•^*
•*^
• —
"M
j!
•rp
7B
P|
T
T>
D7
6-A
fl
,rv
UJ
,
iTi
pp
TTf
1
iTf
T
10 20 30 40 50 60
MP3 PRODUCTION (METRIC TONS/YEaR X 1000)
Fipare 15-13. Effect of variation, of bydtaulie load on treatment
cost at level 3 technology
533
-------
N
O
C
•to
W
I
»-<
Ul
•-»
O
c-rniY
±
X
\
\
jK3Eaii;
y
£=££
KDIMfll
i
IQIJ3
10 20 30 40 SO 60
A1F3 HRODDCTICN (METRIC TCNS/XEAR X 1000)
Figure 15-14. Effect of variation of hydraulic load on treatment
cost at level 4 technology
534
-------
TABLE 15-18. MODEL PLANT TREATMENT COSTS
Subcategory ALUMINUM FLUORIDE
Annual Treatment Costs/Metric Ton of Product
<~OST ITEMS
Annual Operation
and Maintenance
Annual
Amortization
Total Cost
PRODUCTION FLOW FIRST
(kkg/yr) (m3/day)
LEVEL OF TREATMENT
SECOND THIRD FOURTH
tion
nee 15,900
35,600
45,800
a 35,600
b 35,600
c 35,600
d 35,600
15,900
35,600
45,800
a 35,600
b 35,600
c 35,600
d 35,600
15,900
35,600
45,800
a 35,600
b 35,600
c 35,600
d 35,600
690
1,550
1,990
1,550
1,550
2,203
1,064
690
1,550
1,990
1,550
1,550
2,203
1,064
690
1,550
1,990
1,550
1,550
2,203
1,064
10i02
6.37
5.74
8.09
5.58
6.53
6.37
3.45
1.99
1.82
2.12
1.84
2.12
1.98
13.47
8.36
7.56
10.21
7.41
8.65
8.35
2.28
1.14
1.02
1.14
1.14
1.35
1.05
1.12
0.63
0.65
0.63
0.63
0.89
0.54
3.40
1.77
1.67
1.77
1.77
2.24
1.60
2.47
1.25
1.12
1.25
1.25
1.47
1.16
1.26
0.70
0.70
0.70
0.70
0.95
0.60
3.73
1.95
1.82
1.95
1.95
2.42
1.77
4.33
2.72
2.57
3.13
2.60
3.13
2.41
2.76
1.87
1.79
1.95
1.87
2.33
1.51
7.09
4.59
4.36
5.08
4.47
5.46
3.92
a Increased pollutant load
b Decreased pollutant load
c Increased hydraulic load
d Decreased hydraulic load
535
-------
15.7 BASIS FOR REGULATIONS
15.7.1 Evaluation of BPT Treatment Practices
EPA is proposing BPT limitations based on Level 1
treatment. All plants in this subcategory have installed BPT
technology. Pollutants limited by the proposed BPT regulations
are TSS, fluoride, chromium, nickel, and pH. The major
pollutants previously regulated are TSS, fluoride, and aluminum.
Aluminum is no longer considered a pollutant of concern due to
its relatively nontoxic nature. The treatment proposed as the
basis for BPT regulations will actually benefit from the
presence of aluminum which will precipitate under mildly
alkaline conditions and act as a coagulant to aid the removal of
toxic metals and suspended solids.
15.7.2 BPT Effluent Limitations
Technology Basis
The Agency proposes BPT limitations for which the
technology basis is, or is equivalent to, equalization, lime
neutralization/alkaline precipitation, solids removal by
settling or thickening, final pH adjustment, and discharge of
the clarified effluent. This technology represents current
practice in the Aluminum Fluoride industry and was therefore
selected as the basis for the proposed BPT effluent limitations.
Flow Basis
The basis of flow for BPT limitations is estimated from
data provided in the 308 questionnaires for three of the four
complete plant responses received, including Plant 1837, f251,
and |705. Plant f!88 was omitted in view of the batch process
utilized for the manufacture of aluminum fluoride. The other
three plants are continuous manufacturing processes.
The two major raw process waste water sources contributing
to the total plant flow estimates include scrubber and work area
washdown. These waste water sources are summarized in Table
15-4 for the three plants considered. The model plant flow for
the A1F3 industry is estimated as the average total raw waste
water flow for the three plants, and is used to estimate
pollutant discharge loadings for the purpose of regulation.
Exact measures of treated effluent from the aluminum fluoride
industry are not available, since aluminum fluoride plants
normally integrate process waste streams with those generated by
the hydrofluoric acid process prior to treatment and discharge.
The unit flow rates varied widely for the three plants in the
range between 4.58 to 19,7 m3/kkg of product which is largely
536
-------
dependant on the scrubber design and water utilization. The
A1F3 process in Plant §251 shares an S02 scrubber with the
anhydrous hydrofluoric acid process. Waste water generation
from this combined use scrubber was estimated on the basis of
hydrofluoric acid utilization in the two processes.
The cleaning and work area washdown flow is similar for the
three plants considered, ranging between 1.02 and 2.39 m3/kkg of
product.
The average total flow for the three plants is 11.9 m3/kkg
of product. This flow is used for the model plant in the
aluminum fluoride subcategory.
Selection of pollutants to be Regulated
The selection of pollutants for which specific numerical
effluent limitations are proposed was based on an evaluation of
raw waste data from the screening and verification sampling
program. Pollutant data from the plant sampled during screening
was used to determine the need for verification sampling.
Verification sampling of Plant f705 and |251 provided additional
pollutant raw waste concentration data needed to assess the
magnitude of the pollution potential.
For conventional pollutants, the Agency has selected pH and
total suspended solids for specific treatment and control.
Fluoride was selected as the only nonconventional pollutant
parameter because it is a major constituent in the process raw
waste and is a pollutant of concern to the Agency. A limitation
on aluminum is not proposed because this constituent of the
process wastes will be effectively controlled by treatment
required for removal of toxic metals.
Results of the screening and verification sampling are
tabulated in Section 15.3.3 for the raw process waste stream.
The pollutant concentration listed under verification is the
highest" value observed during sampling at the two plants
visited. Toxic pollutants are listed based on their presence,
during sampling, at detectable concentration levels. Pollutants
from this list were considered candidates for regulation if
their concentrations appeared at least once at approximately the
lowest level estimated as treatable using any available
technology appropriate for their removal. The only two metals
which passed this test were chromium and nickel. The only two
metals which passed this test were chromium and nickel. The
metals arsenic, copper, selenium, and zinc were never observed
in the raw waste at concentrations equal to or above the lowest
level estimated as treatable as presented in Table 8-11 and
therefore are not proposed for regulation.
537
-------
Specific numerical effluent loading limitations are
proposed for chromium and nickel for which the average
concentration levels (Table 15-8) are considered treatable for
at least one plant visited during sampling.
No limitation is being proposed for aluminum because of its
relatively low toxicity and its beneficial effects in removing
toxic metals by coprecipitation. In addition, control of the
major toxic metal ions should provide adequate control of the
aluminum concentration, since the treatment pH for BPT is in the
region considered optimal for alkaline precipitation of most
metal hydroxides.
Basis of Pollutant Limitations
Conventional and nonconventional parameters -
A. pH: The treated effluent is to be controlled within the
range of 6.0 to 9.0. This limitation is based on the data
presented in Appendix B of this report and the JRB Study (52).
B. TSS and Fluoride: Pollutant limitations for TSS and
fluoride were based on the evaluation of data for the
hydrofluoric acid subcategory. This evaluation is described in
Section 12.7.2 under "Basis of Pollutant Limitations." There
are no plants where the BPT treatment performance can be
evaluated for the treatment of raw aluminum fluoride process
waste water alone. Aluminum fluoride plants integrate raw
process waste water with waste waters generated from the
hydrofluoric acid process.
In view of the similar waste water characteristics, the
effluent concentration from a common treatment system would be
the same for TSS and fluoride whether it originates from the
A1F3 industry or the HF industry. Therefore, a maximum 30-day
average concentration for the A1F3 industry of 97 mg/1 and 53
mg/1 from the HF subcategory (Table 12-24) are proposed for TSS
and fluoride, respectively. These are relatively high values
that are unique to this industry. The variability factor ratio
of 2.1 was selected based on the evaluation in the HF
subcategory (Table 12-23). The unit effluent load limitation is
determined as follows:
L (as kg/kkg) = (Q) (C)
1000
Where C is the maximum 30-day average concentration in
mg/1, Q is the unit flow in m3/kkg, and 1000 is the conversion
factor for kg to grams. (Note: kg/m3 = 1000 mg/1.)
538
-------
The 24-hour maximum is determined by the following
relationship:
Maximum 30-day average X VFR = 24-hour maximum
(concentration or unit (concentration or
loading) unit loading)
In this case, the daily maximum TSS concentration is 2.1 X
97 mg/1 = 200 mg/1. The unit loading is then
97 mg/1 (11.9 m3/kkg)/ kg/m3 \ = 1.2 kg/kkg
V 1000 mg/1 )
In the same manner the concentration basis for fluorides is
2.1 X 53 mg/1 = 110 mg/1. The unit loading is then
53 mg/1 (11.9 m3/kkg)/ kg/m3 \ = 0.63 kg/kkg
V 1000 mg/1 )
The 24-hour maximum unit loading is determined by
multiplying 2.1 times the 30-day average unit loading determined
above.
Toxic pollutants - The effluent limitations proposed for
the selected toxic pollutant control parameters are derived from
three sources of information including 1) literature based
treatability estimates (Section 8.1), 2) screening and
verification sampling data, 3) a limited amount of long-term
monitoring data from Plant #251.
The sampling results represent raw process waste pollutants
observed during three days of composite sampling at each of the
plants verified. An assessment of treatment system performance
was not possible in view of the lack of representative effluent
data available in the subcategory. Effluent data obtained
during verification sampling is for treated waste water from the
HF and A1F3 processes combined, since no plant is available
which treats A1F3 wastes alone. Therefore, the screening and
verification data may be used to determine candidate toxic
pollutants for regulation without specifying achievable
concentration limits which represent the A1F3 plant performance
alone. However, review of the combined HF and A1F3 waste
effluent data in Table 15-9 reveals that all toxic pollutants of-
concern are treatable within the levels of treatability defined
in Section 8.1 for lime settling (BPT) . Removal of toxic
pollutants from one waste water or the other would not differ in
light of the similar nature of HF and A1F3 wastes. Therefore,
the literature estimates of treatability discussed in Section
8.1 have been used as 'the basis for determining specific
numerical limitations for toxic pollutants.
539
-------
A. Chromium; The literature treatability value of 0.1
mg/1 from Table 8-11 for lime settling is considered to
represent a maximum 30-day average concentration value for
chromium in view of plant performance data in the HF and
combined HP/A1F3 industries. The unit load limitation was
calculated as follows:
(0.10 mg/1) (11.9 m3/kkg)/ kg/m3 N = 0.0012 kg/kkg
\1000 mg/1 )
Since long-term monitoring data on chromium is not available,
the variability factor ratio (VFR) of 2.0 was selected on the
basis of lead monitoring data from Plant f251 presented in
Tables A-lOa and A-lOc. This is justified by the similarity in
the chemistry of lead, nickel, chromium, and other metals of
concern under BPT treatment conditions. Therefore,
VFR = VF of daily measurements - 3L12
VF of 30-day averages 1.55
VFR =2.0
The daily maximum limitation for chromium was determined as
follows;
(2.0)(0.0012 kg/kkg) = 0.0024 kg/kkg
The proposed effluent limitations on chromium are presented in
Table 15-19 for BPT treatment.
B. Nickel: The raw waste concentration of nickel was
observed as high as 0.29 mg/1 (Section 15.3.3, Table of Maximum
Concentrations Observed) to an average value of 0.22 mg/1 at
Plant 1705 (Table 15-8). The literature treatability value of
0.20 mg/1 from Table 8-11 for lime settling is used for the
purpose of regulation in view of the absence of actual plant
performance data. The limitation is determined as follows:
(0.20 mg/1) (11.9 m3/kkg)/ kg/m3 \ « 0.0024 kg/kkg
V 1000 mg/1 /
Therefore, the 24-hour maximum load limitation is:
(2.0)(0.0024 kg/kkg) = 0.0048 kg/kkg
where 2.0 is the VFR as discussed for chromium.
C. Other metals: The concentration bases for arsenic,
copper, selenium, and zinc are also presented in Table 15-19.
These pollutants are listed to serve as guidance in cases where
these pollutants are found to be of water quality concern. The
540
-------
TABLE 15-19. PROPOSED LIMITATIONS
Aluminum Fluoride
Best Practicable Control Technology Currently Available
Waste Water Plow: 11.9
Pollutant
Effluent Limit
(kg/kkg)
Concentration Basis
Subcategoty (1) (mg/ll
Performance WR
(mg/1) 30-day 24-hr 30-day 24-hr
Avg Max Avg Max
Conventional and
Nonconventional Pollutants:
Total Suspended 97
Solids, TSS
Fluoride 53
Toxic Pollutants:
(2)
2.1
(2)
2.1
(4)
97
53
200
110
1.2 2.4
0.63 1.3
Arsenic
Chromium
Copper
Nickel
Selenium
Zinc
0.50
0.10
0.50
i
0.20
0.20
(3)
(3)
i
(3)
l
(3)
I
(3)
I
(3)
0.50
2.0
2.0
2.0
2.0
2.0
2.!
(5)
(5)
i
(5)
i
(5)
I
(5)
i
(5)
0.50
0.10
0.50
0.20
0.20
0.50
1.0
0.20
1.0
0.40
0.40
0.50
(65 (6)
0.0012 0.0024
(6) (6)
0.0024 0.0048
(6) (6)
(6)
(6)
(1) - VFR: ratio of the 24 hour variability factor to the
30 day variability factor.
(2) - 30 Day maximum average concentration based on the proposed
HP subcategory regulation (Section 12.7,2).
(3) - The lower limit of the literature treatability estimate
(Table 8-11) is used as the basis for the maximum 30-day
average limitation and pubcategory performance, since no plant
is available where BPT treatment can be evaluated for the A1P3
waste water alone.
(4) - WR based on HP subcategory evaluation.
(5) - WR based on limited long term data.
(6) - No effluent limitation proposed.
541
-------
concentration limitations are also based on literature
treatability levels presented in Table 8-11. However in every
case these treatability levels were above raw waste
concentrations observed for each of these metals.
15.7.3 Basis for Proposed BCT Effluent Limitations
The BCT limitation {applicable only to TSS) was set equal
to BPT because the addition of more treatment technology to
remove conventional pollutants failed to pass the BCT cost
comparison test (44 FR 44501 July 30, 1979) as described in
Section 3.3.3.
15.7.4 Basis for Proposed BAT Effluent Limitations
The Application of Advanced Level Treatment:
The Agency has analyzed the cost effectiveness of the base
level systems (BPT) and the various advanced level options for
conventional, nonconventional and toxic pollutant removal based
on utilizing the cost estimates presented in this report. The
economic impacts on the aluminum fluoride industry have been
evaluated in detail and taken into consideration in the
selection of the technology basis for the proposed BAT
regulations.
For BAT, EPA is proposing limitations based on Level 2
treatment. This treatment option adds dual media filtration to
remove additional toxic metals and fluorides. This level of
treatment removes 300 pounds per year of toxic metals and 62,000
pounds per year of fluorides. Pollutants limited in proposed
BAT regulations are fluoride, chromium, and nickel.
EPA considered limitations based on Level 3 and 4 sulfide
precipitation and use of soda ash to increase recycle,
respectively. These options were rejected because they remove
only small incremental amounts of toxic pollutants in this
subcategory.
Technology Basis
For BAT, the Agency proposes more stringent effluent
limitations on fluoride and the toxic pollutants based on
addition of dual media filtration or its equivalent to the BPT
treatment system (Section 15.7.2).
Flow Basis
The same flow established for BPT in Section 15.7.2 is used
in the development of the BAT effluent limitation. The flow
used is 11.9 m3/kkg of product (Table 15-4).
542
-------
Selection of Pollutants to be Regulated
The Agency has selected fluoride and the same two toxic
pollutants identified in the proposed BPT regulations for the
BAT regulations. The rationale for their selection is discussed
in Section 15.7.2.
Basis of Pollutant Limitations
Nonconventional pollutants - The only nonconventional
pollutantselected is fluoride. The limitation proposed for BAT
is based on the evaluation of plant performance data discussed
in Section 12.7.4 for the HF subcategory. A maximum 30-day
average concentration of 30 mg/1 for total fluoride was
identified in the evaluation and is used here in Table 15-20 for
establishing a numerical limitation. Selection of the
concentration is based on the similarity between the waste
stream from the HF and A1F3 subcategories.
The 24-hour maximum concentration is determined as follows
from the WR and maximum 30-day average concentration;
(2.1)(30 mg/1) = 63 mg/1
The effluent limitation for fluoride is determined as
follows:
(30 mg/1) (11.9 m3/kkg)/ kg/m3 \ = 0.36 kg/kkg
V 1000 mg/1 )
The 24-hour maximum limitation is determined in a similar
manner as follows:
(63 mg/1) (11.9 m3/kkg)/ kg/m3 N » 0.75 kg"/kkg
V 1000 mg/1 )
The variability factor of 2.1 used for the development of
BAT limitations is the same used for BPT in Section 15.7.2.
The estimated performance of Level 3 and Level 4
alternative technologies are presented in Tables 15-21 and
15-22, respectively. The tables present the estimated maximum
30-day average and 24-hour maximum concentrations for the
purpose of comparison with the proposed regulations.
Toxic pollutants - Addition of dual media filtration to the
BPT level of treatment provides additional removal of the
suspended metal hydroxides. • Therefore, BAT provides more
stringent control of the toxic pollutants. Since there is no
directly applicable data on filter performance for the A1F3
543
-------
TABLE 15-20. PROPOSED LIMITATIONS
Aluminum Fluoride
Best Available Techrology
Waste T/feter Flow: 11.9 m3/kj
0.18
0.47
(5)
(5)
2.1 30
2.0
2.0
2.0
2.0
2.0
2.0
0.50
0.04
0.29
0.17
0.18
0.47
63
1.0
0.08
0.58
0.34
0.36
0.94
0.36
0.00048
_J4)
0.0020
__(4)
(4)
0.75
0.00096
_J4)
0.0040
_J4>
(4)
(1) - WR: ratio of the 24-hour variability factor to the 30-day variability
factor.
(2) - Also applicable for pretreatment standards for existing sources (PSES)
which are expressed as concentration.
(3) - 30-day average calculated from the HP subcategory Table 12-21 and
12-25.
(4) - No effluent limitation proposed.
(5) - literature treatability estimates.
544
-------
TABLE 15-21. PEEM5KMANCE OF ALTERNATIVE TECHNOLOGY
Alxminum Fluoride
Level of Treatment: 3
feste Water Flow: 11/9 m3/kkg
Pollutant
Nbnoonventional
Pollutants:
Fluoride
Toxic
Pollutants:
Arsenic
Chromium
Copper
Nickel
Selenium
Zinc
Treatability
(mg/1)
25
0.050
0.040
0.050
0.10
0.18
0.20
Concentration Basi§
(1) (rag/1)
™* Max 24-hr
30-day Max
\
3.0 25 75
2.0 0.050 0.10
2.0 0.040 0.080
2.0 0.050 0.10
2.0 0.10 0.20
2.0 0.18 0.36
2.0 0.20 0.40
(1) - VFR: ratio of the 24-hour variability factor to the 30-day variability
factor.
545
-------
T&BuE 15-22. OP SLTEHmTIVE TECHNOLOGY
Aluminum Fluoride
Level of Treatment: 4 /,.,,
K&ste Water Plow: 2.4 mVktag (80% Recycle) u'
Ooncmtration Basis
Bollutant
Nonconventional
Pollutants:
Fluoride
Toxic
Pollutants:
arsenic
Chrcmiim
Copper
Nickel
Selenium
Zinc
Treatability
(rag/1)
30
0.50
0.04
0.29
0.17
0.18
0.47
(1 ) (mg/1)
WR^ '
Max
30-day
Avg
2.1 30
2.0 0.50
2.0 0.04
2.0 0.29
2.0 0.17
2.0 0.18
2.0 0.47
24 -hr
Max
63
1.0
0.08
0.58
0.34
0.36
0.94
(1) - WR: ratio of the 24-hour variability factor to the 30-day variability
factor.
(2) - The effluent flow rate is 20 percent of the average influent or basis
of flow (i.e., 0.20 x 11.9 m3/kkg = 2.4 m3/kkg).
546
-------
industry, the literature treatability studies cited in Section
12.7.4 under "Toxic Pollutants" for the HP industry are used in
the following determinations. The estimated percentage removal
efficiency, presented in the aforementioned section, was applied
to the 30-day average concentrations developed for the BPT
limitations to establish the proposed BAT regulation. The basis
for the BAT limitation on each toxic metal is given below.
A. Chromium: Filtration of the BPT effluent is estimated
to reduce the chromium concentration by 60 percent. Therefore,
the maximum 30-day average concentration would be 0.04 mg/1 by
the following relationship:
BPT 30-day average /100% - 60%\ = BAT 30-day average
concentration \ 100% / concentration
The limitation is determined numerically as follows:
/100 - 60\ 0.10 mg/1 = 0.040 mg/1
\ 100 /
Application of the BAT model plant discharge rate results
in the proposed chromium limitation as follows:
(0.040 mg/1) (11.9 m3/kkg)/ kg/m3 \ = 0.00048 kg/kkg
V 1000 mg/1 /
and, the daily maximum limitation using the VPR value of 2.0
becomes
(2.0) (0.00048 kg/kkg) = 0.00096 kg/kkg
The VFR value of 2.0 used for BPT is similarly used for BAT
because the variability of the filtrate quality is anticipated
to be no greater than the observed variability of the unfiltered
effluent. The variability factor was observed from long-term
data at Plant f251 (Tables A-lOa and A-lOc). Treatability
studies are being conducted by the EPA that assess the proposed
BAT level of treatment.
B. Nickel: Filtration of the BPT effluent is estimated to
reduce the nickel concentration by 14 percent. Therefore, the
maximum 30-day average concentration would be 0.17 mg/1 by the
following calculation:
/100 - 14\ (0.20 mg/1) = 0.17 mg/1
V 100 )
Application of the BAT model plant discharge rate gives the
following load limitation for nickel:
547
-------
(0.17 mg/1) (11.9 m3/kkg)/_kg/in3j\ = 0.0020 kg/kkg
V. 1000 mg/1 )
consequently the 24-hour maximum value is
(2.0)(0.0020 kg/kkg) = 0.0040 kg/kkg
as presented in Table 15-20.
C. Other metals: The concentration basis for arsenic,
copper, selenium, and zinc are also given in Table 15-20
assuming 0, 42, 14, and 6 percent removal efficiency,
respectively, by the addition of filtration to the BPT system.
The values presented in Table 15-20 for these toxic pollutants
are intended for use in cases where they are of concern from a
water quality standpoint. However in all cases the treatability
level was above the raw waste concentration levels observed.
15.7.5 Basis for Proposed New Source Performance Standards
Technology Basis
For NSPS, the Agency proposes the same treatment technology
that is proposed for BAT.
Flow Basis
The same flow established for BPT and BAT is used in the
development of the NSPS effluent limitations.
Selection of Pollutants to be Regulated
The Agency has selected TSS, fluoride and the same two
toxic pollutants identified for the BAT regulations. The
rationale for their selection is discussed in Section 15.7.2.
Basis of Pollutant Limitations
Conventional pollutants -
A. pH: For NSPS, the BPT limitation is retained. Control
of the final effluent within the range of pH between 6.0 and 9.0
is required based on data presented in Appendix B of this report
and the JKB Study (52) .
B. TSS: In view of the absence of applicable performance
data concerning TSS, a value of 68 mg/1 was assumed from the HF
subcategory for the maximum 30-day average concentration. The
value was developed by assuming a 30 percent reduction in TSS
over the 30-day average concentration estimated for BPT (97
548
-------
mg/1). The assumption is based on pilot scale studies (41)
which have demonstrated an average removal by filtration of
approximately 30 percent from waste water containing suspended
metal hydroxides after lime treatment.
A VFR of 2.1 is used on the basis of long-term data
presented in Table 12-21 of the HF subcategory. The proposed
30-day average limitation on TSS is determined as follows:
(11.9 m3/kkg)/^ kg/m3 \ - 0.81 kg/kkg
V 1000 mg/1 /
The 24-hour maximum consequently becomes kg/kkg) = 1.7
kg/kkg.
The proposed NSPS limitations are presented in Table 15-23,
Nonconventional pollutants - Fluoride is the only
nonconventional pollutant and is set equal to the BAT limitation
of 30 mg/1 for NSPS.
Toxic pollutants - Waste water sources are expected to be
the same as currently identified for new source A1F3 plants.
Therefore, the proposed toxic pollutant limitations for NSPS
have been set equal to the proposed BAT limitations by the
Agency. BAT limitations for the toxic pollutants is discussed
previously in Section 15.7.4.
15.7.6 Basis for Proposed Pretreatment Standards
Existing Sources
Pretreatment Standards for Existing Sources (PSES) are
proposed by the Agency to equal BAT limitations. The pollutants
to be limited are fluoride, chromium, and nickel.
New Sources
Pretreatment Standards for New Sources (PSNS) are proposed
by the Agency to equal proposed BAT limitations. The pollutants
to be regulated include fluoride, chromium, and nickel.
549
-------
15-23. PKOPSOSED
Aluminum Fluoride
New Source Performance Standards
Pfeste W&ter Flow: 11.9 m3/kkg
Pollutant
Treatability
Cmg/1}
Concentration Basis
(1) (mg/1)
WR
Effluent Limit
(kg/kkg)
Max
30-day
Avg
24-hr
Max
Max
30-day
Avg
24-hr
Max
ODnventional and
Total Suspended
SolMs, TSS
Fluoride, F (5)
Itoxic
lolltatants;
Arsenic
Chromium (5)
Copper
Nickel (5)
Selenium
Zinc
68
30
(2)
(2)
0.50(3>
(3)
0.04
0.29
0.17
0.18
0.47
(3)
(3)
(3)
(3)
2.1 68
2.1 30
2.0
2.0
2.0
2.0
2.0
2.0
0.50
0.04
0.29
0.17
0.18
0.47
140
63
1.0
0.08
0.58
0.34
0.36
0.94
0.81
0.36
0.00050
_J4>
'070020
— (4)
(4)
1.7
0.75
0.0010
_J4)
0.0040
_J4>
(4)
(1) - WSs ratio of the 24-hour variability factor to the 30-day variability
factor.
(2) - 30-day average calculated from the HF subcategory Table 12-29.
(3) - Literature treatability estimates from BM? level treatment.
(4) - No effluent limitation proposed.
(5) - Also applicable for treatment standards for new sources (PSNS) which
are expressed in concentration.
550
-------
SECTION 16
CHROME PIGMENTS INDUSTRY
16.1 INDUSTRY PROFILE
16.1.1 General Description
Chrome pigments are a family of inorganic compounds
primarily used as colorants in a number of industries. These
pigments are used in paints, ceramics, floorcovering products,
ink, paper, and cements. However, certain chromium compounds
(i.e., oxides) may be used as raw materials in the manufacture
of certain metals and alloys. Chrome pigments vary
substantially in their chemical makeup. The various types
include chrome yellow, chrome orange, molybdate chrome orange,
anhydrous and hydrous chromium oxide and zinc yellow. The
industry data profile is given in Table 16-1 and the status of
the regulations are shown in Table 16-2.
16.1.2 General Process Description and Raw Materials
The general manufacturing process for each of the above
compounds is given below.
Chromium Oxide
This pigment consists of two compounds; anhydrous and
hydrated chrome oxide (Guigets Green) , The* amount of the
anhydrous salt oxide produced is approximately ten times the
amount of hydrated chromic oxide produced. It is offered in a
narrow range of shades from light yellowish to dark bluish
green.
Anhydrous oxide is almost pure chromium oxide and the
commercial grade consists of a minimum of 98.5 percent Cr203.
It is prepared by calcination of sodium dichromate with sulfur
or carbon according to the reactions given below:
Na2Cr207 + S = Cr203 + Na2SO4 (1)
551
-------
.TSBEE 16-1. SOBCKEBOOIg HSgHE QKOR.
SUBCAXBGOIOT
CHBOME PIQ4EH1S
Itotal subcategory capacity rate
Total subcategory production rate
Number of plants in this subcategory
308 Data on file for
With total capacity of
With total production of
Representing capacity
Representing production
Plant production ranges (2)
Miniimm
MaxJsuni
Average production
. Median production
Average capacity utilization
Plant age range:
Minimum
Mscdmsa
Wastewater flow range:
Maxiinira
Volurre per unit product:
Minimum
Maxiirum
63,000 kkg/year
64,500 kfcg/year
12
5
39,800 kkg/year
62 percent
100 kkg/year
18,000 kkg/year
6,300 kkg/year
6,400 kkg/year
78 percent
38 years
60 years
800 cubic meters/day
11,363 cubic meters/day
32 cubic meters/kkg
170 cubic meters/kkg
(1) Sources of data are Stanford Research Institute, Directory of Chemical
Brodacers, U.S.A., 1S77/ U.S. Department of Oonnerce, Current Industrial
Reports, Decanfaer 1977; Energy and Environmental Analysis, Inc.? Draft
Report, "Preliminary Economic Assessment; of Effluent Limitations in the
Inorganic Chanical IMustry," June, 1978 and "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards for the Inorganic Chemicals Industry,"
March, 1980.
(2) Based on production at 11 plants, all other figures are based on 308
Questionnaires.
552
-------
TKBJM 16-2. SEXTOS OF HEGCJLAIICNS - EEFEOEOT LJMTEaTICN GUIDELINES
SOBCSTEQQH21 Chrome Pigments
SUBPAOT
AH
(40CPR
415.340, 5/22/75)
STANDARDS
Product
Process
Chrome
Pigment
Para-
meters
TSS
Cr(T)
Cr*6
Pb
Zn
CN
CM (A)
Fe
BPC3CA * BKEEA HSPS
Max. Avg, Max. Avg, Msec. Avg.
kg/W^ kg/Kkg kgAkq kg/kkg kg/kkg ko/kka
(mg/1) (itq/1) (mg/D W/L) ftnfA) Tmg/a)
5.1
(76.1)*
0.10
(1.5)
0.010
(0.2)
0.42
(6.3)
0.72
(10.8)
0.010
(1.5)
0.10
(0.2)
Oo72
(10.8)
1.7 Reserved Reserved
(25.4)
0.034
(0.5)
0.0034
(0.1)
0.14
(2.1)
0.27
(4.0)
0.0034
(0.5)
0.034
(0.1)
0.27
(4.0)
Sections 415.340, 415.341, and 415.342 were revoked by the Agency
£41 FR 51601, November 23, 19761.
Max, = Maximum of any one day.
2
Avg. * Average of daily values for thirty consecutive days.
* flow basis 67,000 1/kkg.
553
-------
Na2Cr207 + 2C = Cr203 + Na2C03 + CO (2)
The use of sulfur as the reducing agent eliminates C02 and CO
emissions but increases the sulfates in the raw waste as well as
producing SO2 and SOS in the off-gases. In the manufacturing
process using sulfur, the raw materials consisting of sodium
dichromate and sulfur are mixed with water and the resultant
solution is fed to a kiln. The material is heated and the
reacted materials from the kiln are slurried with water,
filtered, washed, dried, ground, screened, and packaged. The
effluent gases from the kiln containing sulfur dioxide and
sulfur trioxide are wet scrubbed before venting to " the
atmosphere.
A general process flow diagram of the preparation of
anhydrous chrome oxide is given in Figure 16-1.
Hydrated chromium oxide, Cr203.2H20 or Cr2O(OH)4, also
known as chromium hydrate and Guigets Green, is a brilliant
bluish green. It is made by reacting sodium dichromate with
boric acid as follows:
2Na2Cr207 + 8H3B03 = 2Cr203.2H20 + 2Na2B407
+ 8H20 + 302 (3)
The raw materials are blended in a mixer and then heated in an
oven at about" 550 degrees C. The reacted material is slurried
with water and filtered. The filtered solids are washed with
water, dried, ground, screened, and packaged. The filtrate and
the wash water are treated with sulfuric acid to recover boric
acid according to the reaction given below:
Na2B407 + H2SO4 + 5H2O = 4H3BO3 + Na2SO4 (4)
A waste stream containing some boric acid and sodium
sulfate is discharged from the boric acid unit. Figure 16-2 is
a generalized flow diagram of the process.
Chrome Yellow and Chrome Orange
Chrome yellow is one of the more important synthetic
pigments. The chrome yellows cover the range of hues from light
greenish yellow to reddish medium yellow and consist mainly of
lead chromate. They are made by reacting sodium dichromate,
caustic soda, and lead nitrate. The reactions are given ass
2HNO3 + PbO = Pb(NO3)2 + H20 (5)
Na2Cr207 + 2NaOH + 2Pb(N03)2 = 2PbCr04 + 4NaN03 + H2O (6)
554
-------
van?
1
U1
01
Sf
SCRUBBER **
UQOTJ 3CH.
i
WRIER
SODIUM DICHBOMATE T
J2
IJHBER """"" """""I
I
«p T®
MTVFTT ^ KT^" *^ ^^!Tr ^ FTTiTRR ^ HRVRR ^ ^^ TMT_r ^^
SULFUR ^
CHROME CKIDE
Figure 16-1. General process diagram for production of anhydrous chrome oxide.
-------
SODIUM
DICHROMftTE
BORIC ftCID
(Ji
tn
en
HTOBftllC CHK»E
C5XIDE TO GRINDINS,
MID
VRSTE TRUER
Figure 16-2. General process diagram for production of hydrated diromic oxide.
-------
Lead chromate is formed as a precipitate during the
reaction. It is filtered and treated with chemicals for
development of desired specific pigment properties, dried,
milled, and packaged. The filtrate from the filtration
operation is sent to the waste water treatment facility, A
flow diagram of the chrome yellow manufacturing process is shown
in Figure 16-3.
Molybdenum Orange
Molybdenum orange is made by the coprecipitation of lead
chromate (PbCr04) and lead molybdate (PbMo04). The resulting
pigments are more brilliant than chrome oranges.
The process consists of dissolving molybdic oxide in
aqueous sodium hydroxide and adding sodium chromate. The
solution is mixed and reacted with a solution-of lead nitrate.
The precipitate from the reaction is filtered, washed, dried,
milled and packaged. The filtrate is sent to the treatment
facility.
The reaction is given as follows:
Mo03 + 2NaOH = Na2Mo04 + H20 (7)
PbO + 2HN03 = Pb(NO3)2 + H20 (8)
Na2MoQ4 + Pb (NO3)2 = PbMo04 + 2NaNO3 (9)
Na2Cr04 + Pb (N03)2 = PbCr04 + 2NaN03 (10)
PbMo04 + PbCr04 = PbCr04.PbMo04 fll)
A simplified flow diagram for the manufacture of molybdenum
orange is given in Figure 16-4.
Chrome Green
Chrome greens are a coprecipitate of chrome yellow and iron
blues. They include a wide variety of hues from very light to
very dark green. Iron blues are manufactured by reaction of
aqueous solution of iron sulfate and ammonium sulfate with
sodium hexacyanoferrate. The precipitate formed is separated
and oxidized with sodium chlorate 'or sodium chromate to form
iron blues (Fe NH4 .Fe CN 6) . Chrome green is produced by
mechanically mixing chrome yellow and iron blue pigments in
water. The coprecipitate formation of chrome green is given bys
PbCr04 + Fe(NH4) .Fe(CN)6 = PbCr04Fe(Nfi4) .Fe(CN) 6 (12)
557
-------
(Jl
Ul
00
IBU) GXHE
HKBBH
tcraic ACID
SCDIIH HYDHOXIDE
SODIUM
CHRCHE YELiCW
1DDRONS, MHMNG
AH) PACKAGING
HRSffi WIER
Figure 16-3. (feneral process diagram for prod.tcrtd.on of dirone yellow.
-------
VENT
I
SODIUM
CHROMATE WATER
Mjbxuuu; uxjJJis
WATER
CAUSTIC SODA fc
(J\ '
VO
I£AD OXIDE
NITRIC flCID
DISSOLVER
VENT
UifctlULiVJSK
f f
^^ FILTRATION
WASHING
*
WASTE WATER
DRYING
MULDG
AND
ESCKAGIN3
OF
tOLYBEENtM ORRNGE
(PbCr04.PbMo04)
PRODUCT
Figure 16-4. General process diagram for production of molybdenum orange.
-------
Figure 16-5 gives a process flow diagram for the
manufacture of chrome green.
Zinc Yellow
Zinc yellow, also called zinc chromate, is a greenish
yellow pigment. It is a complex compound of zinc, potassium,
and chromium which has the approximate composition
4ZnO.K20.4Cr03.3H20. It is made by the reaction of zinc oxide,
hydrochloric acid, sodium dichromate, and potassium chloride.
Zinc yellow is formed as a precipitate and is filtered, washed,
dried, milled, and packaged for sale. The reactions are given
as:
2KC1 + 2HC1 + 2Na2Cr207.H20 = K2Cr4013 + 4NaCl
+ 3H20 , (13)
4ZnO + K2Cr4013 + 3H20 = 4ZnO.K20.4Cr03.3H20 (14)
A general flow diagram of the manufacturing process is
given in Figure 16-6.
16.2 WATER USE AND WASTE SOURCE CHARACTERISTICS
16.2.1 Water Use
In the chrome pigments industry, water is used primarily
for noncontact cooling, washing the precipitated product, and as
boiler feed for steam generation. In some cases, water is
introduced into the reactor along with the raw materials.
In addition, substantial quantities of water may be used in
cleaning equipment. This occurs during product changes at
plants manufacturing a number of pigments. This partially
accounts for the increased unit water use at larger plants,
since these plants have the most complex product mix.
In anhydrous and hydrated chrome oxide manufacture, water
is used for slurrying of the reaction product and in scrubbing
the reactor vent gases. Table 16-3 is a summary of water usage
at different pigment plants in the chrome pigments subcategory.
16.2.2 Waste Sources
Some plants produce different pigment products
sequentially in the same equipment. At a few plants, the
different pigment products are manufactured concurrently and the
waste waters combined and treated at a single facility. A
generalized flow diagram applicable to all chrome pigment plants
560
-------
VRTER
IRON BUS
RESLtMW
Ul
01
LEAD NITRATE ^^. • |
SODIUM CHBDMRTE ^
SODUM SOHB.TE ^
BfflOTW
V 1
PIHER
AM)
1-
SHADE
TRNK
FILTER
I
DRYER
•» ORIKDING
BIfflDINQ
MO
PACKING
CHECMB GREEN
PKODUCT
WASTE WATER
WASTE WATER
Figure 16-5. General process diagram for production of chrome green.
-------
'ZnO
warn
Ul
o\
to
^
mi fc
RCI te
i -
HBSCfKW 18NK
FH.TRATICN
ASHING
DR-mc
pHUUHG, PACKW3DK
OFfflE ZQC MliOW
(K,0'4ZnD-4ClO,'3H.O)
2 J 4
HASTE VftlER
Figure 16-6.- General process diagram for production of zinc yellow.
-------
16-3. IN THE CHROME SUBCAGEGOKy
(1)
USE
#464
UNIT FLCW (m3/kkg)
Plant Designation
#436 #214
Noncontact cooling
Direct process contact
Indirect process contact
Maintenance
Scrubbers
Boiler Feed
Total
. 9.50
18.6
7.18
12.0
3.30
2.52
53.1
6.45
147
m(2)
1.78
9.56^
11.1
176
NA
32.6
NA
0.152
NA
0.152
32.9
(1) Includes all chrome pigment product mixes. Values indicated only
for those plants that reported complete information.
(2) NA - Not applicable.
(3) Iron blue pigment process.
563
-------
is given in Figure 16-7. The waste water sources are similar
for all pigment products except that at chrome oxide plants, an
additional scrubber waste is generated. Table 16-4 gives the
waste water flow data Nummary for several plants. The quantity
of waste water and the pollutants vary for the different pigment
products since the pollutants are dependent on the raw materials
used. The figures in Table 16-4 represent actual • plant
discharges.
The data sources for the plants used in the determination
of unit flow values presented in Table 16-4 are outlined below:
Plant f464. Data based on 308 questionnaire submission.
Only chrome pigment production and flows were included.
Plant §214, Data based on 308 questionnaire submission.
Chrome pigment and iron blue production and flows were included,
Plant f436. Data based on 308 questionnaire submission.
Chrome pigment production and flows were included.
Plant f002. Data based on three days of sampling. Chrome
pigment and organic pigment (20%) productions and flows were
included.
Plant |894. Data based on three days of sampling. Chrome
pigment, iron blue, and organic pigment (15%) productions and
flow were included.
As previously discussed, various plants make several chrome
pigments sequentially or concurrently. Thus the unit hydraulic
load going to the treatment facility will be an average of all
the waste loads from the different processes. The raw waste
from a complex plant may contain nearly all of the following
substances: sodium acetate, sodium chloride, sodium nitrate,
sodium sulfate, potassium chloride; lead, iron, and zinc salts;
soluble chromium and pigment particulates.
16.3 DESCRIPTION OF PIA8TS
16.3,1 Screening
Plant f894 was visited during the screening phase of the
program. The samples for this plant were analyzed for.all toxic
and conventional pollutants.
This plant produces over 100 products including organic
pigments such as copper phthalocyanine, and all the wastes are
combined and,treated together. Treatment consists of chromium
564
-------
o\
Ul
WATER WASH WATER ^^ ^j, ^^
1 * T . -
RAW MATERIALS ^
• ^
REACTOR
FILTER
DRYER
MILLING MB
SCREENING
*» PIGMENT
PRODUCTS
TO '
I 41 1 PACKAGING
WASTE WATER NON-CONTACT PAKnOTATE
(BY-PRODUCT SALTS, STEAM WRCTO
UNREflCTED MATERIALS,
ETC.)
Figure 16-7. General process diagram for production of chrome pigment complexes.
-------
TSBIE 16-4. SUMYRKY OF W&SIE WMER FI£X?
SUBCMEGOKT: PIGMENTS
Plant Designation
#464
#214
#436
#002
#894
Waste Water Flow^
(m3/kkg)
41.1
32.8
149
78.4(2)
170 <2>
Weighted Average Flow , 105^ '
(1) mcludes waste water from all pigment product mixes.
(2) Includes organic pigments.
(3) Weighted on the basis of production since unit waste flow is
directly related to plant production:
Weighted average = 21 [(unit flow)(production)]
2 (production)
i.e. =
Where Q = Unit flow and P = production (which is considered
confidential information) .
566
-------
VI reduction, equalization and neutralization, followed by
clarification and filtration. Sulfur dioxide is added to reduce
the hexavalent chromium to the trivalent state at a low pH prior
to hydroxide precipitation. The backwash from the sand filters
is recycled to the equalization tank, while the sludge from the
clarifiers is passed through filter presses and then hauled to a
landfill. The landfill has a bottom consisting of two clay
layers sandwiching a gravel layer to allow for collection of
leachate drainage. Any water from the sludge is trapped in the
gravel layer, and is pumped out and returned to the plant for
retreatment.
16.3.2 Verification
Two plants were visited during the verification phase of
the program. The first plant, f002, has a rather large product
mix. However, one of the larger continuous units can have a
major impact on the raw waste characteristics. This unit either
produces lead chromate or zinc chromate. During the sampling
period, zinc chromate was being produced. All process waste
waters are treated continuously. First, the wastes are treated
in an S02 reactor to convert .hexavalent chromium to the
trivalent state. The pH is then adjusted to 8.5 and then the
waste is passed through precoated filters, followed by discharge
to the sewer. Figure 16-8 shows the treatment flow diagram and
sampling points. Table 16-5 shows the waste flows and pollutant
loadings. At sample point
-------
RAW WASTE S02 ACID
#1
CHROME TREATMENT
TANK
pH 3.0
CAUSTIC
CAUSTIC ADDITION
THROUGH
pH 8.5
LAB FILTERED
1
OUTFALL
TO SEWER
FILTER FEED
TANK
I
FILTER AID
BACKWASH
J
(FILTERS NOT WORKING SO
WERE BEING BYPASSED.
THIS WOULD BE THE FLOW
PATTERN IF FILTERS UERF
OPERATING.)
LEGEND
^| SAMPLING POINTS.
Figure 16-8.
General waste water treatment process flow diagram at plant #002
"showing the sampling points. (Chrome pigment manufacture.)
568
-------
16-5. FICW, PODJOTSNT
OF THE SAMPLED WASTE STREAMS FOR PLBNT # 002
SOBCATEGORY: CHROME PIQfflTS
Conventional and Nonconventional Pollutants
(mg/1)
(kg/kkg of chrome pigments)
Stream
*
1
2-0
2-F
Stream
Description
Raw Waste
Unfiltered
Treated
Waste
Filtered
Treated
Waste
Flow
(ia3/kkg) TSS
78.4 700
55
78.4 970
76
78.4 NA/ *
Fe
1.6
0.13
2.3
0.18
0.06
0.0047
Cr(¥I)
300
24
120
9.4
m
(1) NA - Not available
569
-------
solids, and toxic metals. No organics were analyzed during
verification.
Figure 16-9 shows the treatment system flow diagram with
the sampling points indicated. Table 16-6 gives waste flows and
pollutant loadings.
16.3.3 Toxic Polliitant Concentrations
The toxic pollutants found above treatable concentrations
in the raw wastes during sampling are given in the table below.
Screening data was obtained at Plant f894. Verification was
completed at Plants f894 and |002. The only organic pollutant
found in the raw waste above the protocol detectable limit (10
lig/1) was naphthalene at 14 pg/l. It should be noted however
that some nitrobenzene (56 pg/1) and phthalates at levels up to
220 yg/1 were found in the treated effluent and one raw water
intake. Since they were not present in the raw wastes, it is
presumed they are present as a results of sample contamination?
i.e., plasticizer in Tygon Tubing. No organic pollutant
sampling was made during verification.
Section 5.1.2 of this report describes the methodology of
the screening and verification sampling program. In the chrome
pigments industry, 9 days of sampling were conducted at Plants
#894 and 1002. This involved 5 different sampling points for
raw and treated waste streams. The evaluation of toxic metals
content of these process related waste streams was based on 195
analytical data points. The screening at Plant £894 for organic
pollutants generated another 228 data points. The daily raw
waste loads were calculated from the waste stream flow rates
measured or estimated at the time of sampling and the measured
concentration.
That is,
Daily loading askg of pollutant = (C)(Q)
day 1000
Where:
C is the concentration of the pollutant expressed in
units of mg/1 (Note: kg/m3 = 1000 mg/1), and
Q is the waste-stream flow rate expressed in units of
mS/day. (m3, a cubic meter, is equal to 264.2 U.S.
gallons.)
Similarly, the unit loadings were calculated from the
reported chrome pigments production rate, the waste stream flow
rate, and the measured pollutant concentration.
570
-------
WASTE
WASTE (_
WATER
(EPA SAMPLE
POINT ALSO)
en
«j
H
SLAKED
LIME
EQUALIZATION
TANK
1
NEUTRALIZATION
TANK
pH 6.2-6.5
NEUTRALIZATION
TANK
pH 8.0-8.3
BACKWASH
HOLDING
TANK
BACKWASH
(2)
SAND
FILTERS
-9-
|