DEVELOPMENT DOCUMENT
for
PROPOSED EFFLUENT LIMITATIONS GUIDELINES,
NEW SOURCE PERFORMANCE STANDARDS,
and
PRETREATMENT STANDARDS
for the
INORGANIC CHEMICALS MANUFACTURING
POINT SOURCE CATEGORY
Douglas M. Costle
Administrator
Robert B. Schaffer, Director
Effluent Guidelines Division
G. Edward Stigall, Chief
Inorganic Chemicals and Service Industries Branch
Elwood E. Martin
Dwight Hlustick
Project Officers
June 1980
Effluent Guidelines Division
Office of Water and Waste Management
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
-------
1 RfcPORT NO
! 2
3 S ACCESSION NO
EPA-440/1-80/007b
4 TlTLh AND SU8TITLL
- -
-
-
—
—¦ -
- - -
S REPORT da'E *"
-
TECHNICAL REPORT DATA
{Plrase read Instructionj on the reverse before c
-------
-------
TABLE OF CONTENTS
Pane
LIST OF FIGURES xxiii
LIST OF TABLES xxxiii
ACKNOWLEDGEMENTS li
1.0 CONCLUSIONS AND SUMMARY \
l.L TOXIC POLLUTANTS ">
1.2 CONTROL AND TREATMENT TECHNOLOGY 2
1.3 COSTS OF ADDITIONAL IN-PLANT TREATMENT 2
!.4 SUECATEGORIZATION 2
1.5 RESTUDY OF REMANDED REGULATIONS 3
2.0 RECOMMENDATIONS 5
3.0 INTRODUCTION 2 3
3.1 AUTHORITY 23
3.1.1 The Federal Water Pollution 23
Control Act Amendments
3.1.2 Court Remand Regulations 26
3.1.3 The Settlement Agreement 28
3.2 GENERAL APPROACH AND METHODOLOGY 36
3.2.1 Industry Data Pased Development 37
and Subcategorization Review
3.2.2 The Screening and Verification 37
Sampling Programs
3.2.3 Engineering Evaluations 37
3.2.4 Treatment Svstem Cost Estimates 38
3.3 GENERAL CRITERIA FOR EFFLUENT LIMITATIONS 38
3.3.1 BPT Effluent Limitations 38
3.3.2 BAT Effluent Limitations 39
3.3.3 BCT Effluent Limitations 39
3.3.4 New Source Performance 42
Standards
3.3.5 Pretreatment Standards for 42
Existing Sources
3.3.6 Pretreatment Standards for 42
New Sources
iii
-------
TABLE OF CONTENTS
- Continued
Page
4.0 SUBCATEGORY AT I ON REVIEW 45
4.1 BASIS FOR SUBCATEGORIZATION 4 5
4.1.1 Factors Considered *5
4.1.2 General Conclusions 49
4.2 SECONDARY SUBCATEGORIZATION 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
4.3.1 Hydrofluoric Acid and Aluminum 51
Fluoride
4.4 SUMMARY 52
5.0 SCREENING AND VERIFICATION SAMPLING PROGRAMS 53
5.1 SCOPE AND METHODOLOGY 5 3
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 BASE 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
iv
-------
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 9 3
7.9 MEMBRANE PROCESSES 96
7.10 ADSORPTION 98
7.11 FLUORIDE REMOVAL 1.0]
7.12 CHLORINE REMOVAL 102
8.0 TREATABILITY ESTIMATES AND LONG-TERM 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 ]17
Guidelines Based Upon
Historical Performance
8.2.2 Assumptions Concerning Daily 118
Pollutant Level Measurements
8.2.3 Assumptions Concerning 30-Day ]23
Average Pollutant Level
Observat ion
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 OF CONTENTS - Continued
Page
9.4 POLLUTION CONTROL PARAMETERS TO BE 136
REGULATED
9.4.1 Conventional Pollutants 136
9.4.2 Nonconventional Pollutants 1?6
9.4.3 Toxic Pollutants 137
10.0 COST OF TREATMENT AND CONTROL SYSTEMS 13°
10.1 INTRODUCTION 139
10.1.] Purpose of Cost Data 139
10.1.2 General Approach 140
10.1.3 Cost References and Rationale 140
10.1.4 Definition of Levels of 1.41
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 151
11.1 MERCURY CELL PROCESS INDUSTRY PROFILE 151
11.1.1 General Description 151
11.1.2 General Process Description and 151
Raw Materials
11.2 WATER USE AND WASTE WATER SOURCE 1.55
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 169
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
vi
-------
TABLE OF CONTENTS
- Continued
Page
11.5
SELECTION
OF APPROPRIATE TECHNOLOGY AND
177
EQUIPMENT
11.5.1
Technologies for Different
177
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
181
IT .6.3
Model Plant Treatment Costs
181
11.7
BASIS FOR
REGULATIONS
18?
11.7.1
Basis for BPT Limitations
182
11.7.2
Basis for Proposed BAT Effluent
194
Limitations
11.7.3
Basis for Proposed BCT Effluent
205
L imi tations
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
Ver if ication
214
11.10.3
Descriptions of Plants Not
221
Sampled
11.30.4
Toxic Pollutant Concentrations
224
11.11
POLLUTION
ABATEMENT OPTIONS
231
11.1 1 .1
Toxic Pollutants of Concern
231
11.11.2
Prevailing Control and
236
Treatment Practices
11.11.3 *
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
vii
-------
TABLE OF CONTENTS - Continued
Page
11.13
TREATMENT
COST ESTIMATES
244
11.13.1
General Discussion
244
11.13.2
Model Plant Treatment Costs
247
1 1 .14
BASIS FOR
REGULATIONS
248
11.14.1
Basis for BPT Limitations
248
1] .14.2
Basis for BAT Effluent
263
L imi tations
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
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
Screeni ng
285
12.3.2
Verification
291
12.3.3
Summary of the Toxic Pollutant
2© i
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
30]
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
viii
-------
TABLE OF CONTENTS
- Continued
Page
12.6.3 Model Plant Control Costs for 317
New Sources
3 2.7 BASIS FOR REGULATIONS 324
12.7.1 Evaluation of BPT Treatment 324
Practices
12.7.2 Basis for Proposed BPT Effluent 329
L imi tations
12.7.3 Basis for Proposed BCT Effluent 340
L imi tat ions
12.7.4 Basis for Proposed EAT 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 3 57
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
14.0 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
ix
-------
TABLE OF CONTENTS
- Continued
Page
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
OF APPROPRIATE TECHNOLOGY AND
379
EQUIPMENT
14.5.1
Technoloaies 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
L imitations
14.7.3
Basis for Proposed BCT Effluent
400
L imi tations
14.7.4
Basis for Proposed BAT Effluent
400
Limitations
1.4.7.5
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.B.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
Scr een i ng
415
14.10.2
Ver if icat ion
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
x
-------
TABLE OF CONTENTS - Continued
Page
14.11.5
Advanced Treatment Technoloaies
426
14.12
SELECTION
OF 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
L imitations
14.14 .4
Basis for Proposed BAT Effluent
452
L imi tations
14.14.5
Basis for Proposed New 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 VISTED AND SAMPLED
461
14.17 .1
Scr eening
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
xi
-------
TABLE OP CONTENTS - Continued
Page
14.19.1
Technologies for Different
468
Treatment Levels
14.19 .2
Equipment for Different
471
Treatment Levels
TREATMENT
COST ESTIMATES
472
14.20.1
General Discussion
472
14.20.2
Model Plant Control and
473
Treatment Costs
BASIS FOR
REGULATIONS
479
14.21.1
Evaluation of BPT Treatment
479
Practices
14.21.2
Basis for Proposed BPT Effluent
479
L imitation
14.21.3
Basis for Proposed BCT Effluent
485
L imi tations
14.21.4
Basis for Proposed BAT Effluent
485
L imi tations
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
xii
-------
TABLE 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
15.7 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
L imitations
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 551
16.1 .INDUSTRY PROFILE 551
16.D.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
xiii
-------
TABLE OF CONTENTS - Continued
Page
16.5.1 Technologies for Different 581
Treatment Levels
16.5.2 Equipment for Different 58]
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
L imi tations
16.7.3 Basis for Proposed BCT 602
L imi tat ions
16.7.4 Basis for Proposed BAT Effluent 602
L imi tat i ons
16.7.5 Basis for Proposed New Source 604
Performance Standards
16.7.6 Easis 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 617
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 6 31
EQUIPMENT
xiv
-------
TABLE OF CONTENTS
- Continued
17.5.1 Technologies for Different
Treatment Levels
17.5.2 Equipment for Different
Treatment Levels
17.6 TREATMENT COST ESTIMATES
17.6.1 General Discussion
17.7. BASIS FOR REGULATIONS
17.7.1 Evaluation of BPT Treatment
Practices
17.7.2 Basis for Proposed BPT
Limitations
17.7.3 Basis for Proposed BCT
Limitations
17.7.4 Basis for Proposed BAT
L imitations
17.7.5 Basis for Proposed New Source
Performance Standards
17.7.6 Basis for Proposed Pretreatment
Standards
18.0 SODIUM DICHROMATE INDUSTRY
18.1
18 .2
18.3
1.8 .4
18.5
INDUSTRY PROFILE
18.1.] General Description
18.1.2 General Process Description and
Raw Materials
WATER USE AND WASTE SOURCE
CHARACTERISTICS
]8.2.1 Water Use
18.2.2 Waste Sources
DESCRIPTION OF PLANTS VISITED AND SAMPLED
18.3.1 Screening
18.3.2 Verification
18.3.3 Toxic Pollutant Concentrations
and Loadings
ABATEMENT OPTIONS
Toxic Pollutants of Concern
Process Modifications and
Technology Transfer Options
4.3 Best Management Practices
4.4 Prevailing Control and
Treatment Practices
Advanced Treatment Technoloaies
OF APPROPRIATE TECHNOLOGY AND
POLLUTION
18.4.1
18.4.2
18
18
18.4.5
SELECTION
EQUIPMENT
18.5.] Technology for Different
Treatment Levels
Page
631
633
635
635
636
636
636
649
649
652
652
655
655
655
655
658
658
658
661
661
664
664
668
668
671
671
671
672
672
672
xv
-------
TABLE OP CONTENTS - Continued
Page
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 70]
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 70 5
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 VISITED 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 7.19
21.4.2 Process Modifications and 719
Technology Transfer Options
xvi
-------
TABLE OF CONTENTS - Continued
Pa9e
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 OF APPROPRIATE TECHNOLOGY AND 721
EQUIPMENT
21.5.1 Technologies for Different 721
Treatment Levels
21.5.2 Equipment for Different 72]
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
L imitations
21.7.3 Basis for Proposed BCT Effluent 733
L imi tations
21.7.4 Basis for Proposed BAT Effluent 733
L imi tations
21.7.5 Basis for Proposed New Source 738
Performance Standards
21.7.6 Basis for Proposed Pretreatment 740
Standards
22.0 NICKEL SULFATE 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
xvii
-------
TABLE OP CONTENTS
- Continued
Page
22.4.4 Prevailing Control and 756
Treatment Practices
22.4.5 Advanced Treatment Technologies 757
22.5 SELECTION OF APPROPRIATE TECHNOLOGY AND 7 57
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 76]
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
L imi tat ions
22.7.4 Basis for Proposed BAT Effluent 770
L imi tations
22.7.5 New Source Performance 777
Standards
22.7.6 Basis for Proposed Pretreatment 777
S tandards
23.0 SILVER NITRATE INDUSTRY 779
23.1 SUMMARY OF DETERMINATIONS 779
23.2 ASSESSMENT OF THE WATER POLLUTATION 779
POTENTIAL
23.2.1 Production Processes and 779
E f fluents
23.3 STATUS OF REGULATIONS 781
24.0 SODIUM BISULFITE INDUSTRY 783
24.] 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
xviii
-------
TABLE OP CONTENTS
- Continued
24.3
24.4
24.5
24.6
24.7
Page
DESCRIPTION OF PLANTS VISITED AND SAMPLED
787
24.3.1
Scr een i ng
787
24.3.2
Ver i f icat ion
787
24.3.3
Toxic Pollutant Analytical
795
Results
POLLUTION
ABATEMENT OPTIONS
800
24.4.1
Toxic Pollutants of Concern
800
24 .4 .2
Prevailing Control anc^
800
Treatment Practices
24.4.3
Advanced Treatment Technologies
800
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
TREATMENT
COST ESTIMATES
805
24.6.1
General Discussion
805
24.6 .2
Cost Estimates
806
BASIS FOR
REGULATIONS
813
24 .7 .1
Evaluation of BPT Treatment
81 3
Practices
24.7.2
Basis for Proposed BPT Effluent
813
L imi tations
24.7.3
Basis for Proposed BCT Effluent
822
L imi tat ions
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
S tandards
25.0 SODIUM HYDROSULFITE INDUSTRY 827
25.1 INDUSTRY PROFILE 827
25.1.1 General Description 827
25.1.2 General Process Description and 827
Raw Materials
2 5.2 WATER USE AND WASTE SOURCE 8 30
CHARACTERISTICS
25.2.1 Water Use 830
25.2.2 Waste Sources 830
2 5.3 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
Page
25.4
POLLUTION ABATEMENT OPTIONS
841
25. 4.:1
Toxic Pollutants of Concern
841
25.4.2
Prevailing Control and
841
Treatment Practices
2 5.4.?
Advanced Treatment Technologies
842
25.5
SELECTION OF APPROPRIATE TECHNOLOGGY AND
842
EQUIPMENT
25.5.1
Technologies for Different
842
Treatment Levels
25.5.2
Eouipment 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
Pract ices
25.7.2
Basis for Proposed BPT Effluent
853
L imitations
25.7.3
Basis for Proposed BCT Effluent
855
L imi tations
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
S tandards
EXCLUDED SUBCATEGORIES
863
26.1
ALUMINUM
i SULFATE
863
26.2
AMMONIUM
CHLORIDE
864
26 .3
AMMONIUM
1 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
Page
26.17 HYDROCHLORIC ACID . 882
26.18 HYDROGEN 885
26.1.9 IODINE 886
26.20 LEAD MONOXIDE 887
26.21 LITHIUM CARBONATE 887
26.22 MANGANESE SULFATE 888
26.23 NITRIC ACID 889
26.24 OXYGEN AND NITROGEN 890
26.25 POTASSIUM CHLORIDE 894
26.26 POTASSIUM DICHROMATE 895
26.27 POTASSIUM IODIDE 896
26.28 POTASSIUM METAL 899
26.29 POTASSIUM PERMANGANATE 899
26.30 POTASSIUM SULFATE 900
26.31 SODIUM BICARBONATE 901
26.32 SODIUM CARBONATE 901
26.33 SODIUM CHLORIDE 902
26.34 SODIUM FLUORIDE 904
26.35 SODIUM HYDROSULFIDE 905
26.36 SODIUM METAL 906
26.37 SODIUM SILICATE 909
26.38 SODIUM SILICOFLUORIDE 912
26.39 SODIUM SULFITE 912
26.40 SODIUM THIOSULFATE 914
26.41 STANNIC OXIDE <>">4
26.42 STRONG NITRIC ACID 916
26.43 SULFUR DIOXIDE 918
26.44 SULFURIC ACID INDUSTRY 919
26.45 ZINC OXIDE 922
26.46 ZINC SULFATE 923
REFERENCES 925
BIBLIOGRAPHY 931
APPENDIX A Analysis of Long-Tern1 Effluent Monitoring A-l
Data for the Inorqanic Chemicals Industry
APPENDIX B pH Control of Industrial Waste Waters in B-1
the Inorganic Chemicals Industry
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
*765.
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 f782.
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 f?99 161
showing the sampling points. Chlorine/caustic
(mercury cell) manufacture
11-3 General process flow diagram at plant flAl ">6 5
showing the sampling points. Chlorine/caustic
(mercury cell) manufacture
xxiii
-------
LIST OP FIGURES - Continued
Page
11-4 General process flow diagram at plant f!67 ]65
showing the sampling points. Chlorine/caustic
(mercury cell) manufacture
11-5 General process flow diagram at Plant f31.7 167
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)
11-1.1 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 f014 215
showing the sampling points. Chlorine/caustic
(diaphragm cell) manufacture
11-16 General process flow diagram at plant f261 218
showing the sampling points. Chlorine/caustic
(diaphragm cell) manufacture
xxiv
-------
LIST OF FIGURES - Continued
Page
11-17 General process flowsheet at Plant ?738-A 21?
showing the sampling points. Chlorine/caustic
(diaphragm cell) manufacture
11-18 General process flow diagram at Plant 220
#738-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
U-20 General process flow diagram at Plant f967 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 f251 293
showing the sampling points. Hydrofluoric
acid manufacture
12-5 Level 1 waste water treatment for hydrofluoric 302
acid subcategory
xxv
-------
Page
303
304
305
307
314
31 5
321
322
330
366
370
373
380
381
382
LIST OP FIGURES - Continued
Level 2 waste water treatment for hydrofluoric
acid subcategory
Level 3 waste water treatment for hydrofluoric
acid subcategory
Level 4 waste water treatment for hydrofluoric
acid subcategory
Waste water treatment new source performance
standard for hydrofluoric acid subcategory
Annual treatment cost versus production for
the hydrofluoric acid subcategory
Annual unit treatment cost versus production
for the hydrofluoric acid subcategory
Annual treatment cost versus production for
the hydrofluoric acid subcategory (NSPS)
Annual unit treatment cost versus production
for the hydrofluoric acid subcategory (NSPS)
Fluoride loads and concentrations discharged
at selected hydrofluoric acid plants
General process diagram for production of
titanium dioxide (chloride process) from high
grade ores
General flow diagram at Plant f559 showing
the sampling points. (Titanium dioxide -chloride
process manufacture)
General flow diagram at Plant f].72 showing
the sampling points. Titanium dioxide (chloride
process) manufacture
Level 1
waste water treatment
for
ti tanium
dixoide
- chloride process
Level 2
waste water treatment
for
ti tanium
dioxide
- chloride process
Level 3
waste water treatment
for
titanium
dioxide
- chloride process
xxvi
-------
7
8
9
10
] 1
12
13
14
15
16
17
1
2
Page
389
390
412
418
428
429
437
438
457
469
470
496
501
LIST OF FIGURES - Continued
Annual treatment cost versus production for
the titanium dioxide subcategory, chloride
process
Annual unit treatment cost versus production
for the titanium dioxide subcategory, chloride
process
General process flow diagram for production
of titanium dioxide by sulfate process
General flow diagram at Plant f559 showing
the sampling points. (Titanium dioxide -sulfate
process)
Level 1 waste water treatment for titanium
dixoide - sulfate process
Level 2 waste water treatment for titanium
dioxide - sulfate process
Annual treatment cost versus production for
the titanium dioxide subcategory, sulfate
process
Annual unit treatment cost versus production
for the titanium dioxide subcategory, sulfate
process
General process flow diagram of the titanium
tetrachloride portion of a titanium dioxide
plant using the chloride-ilmenite process.
Level 1 waste water treatment for titanium
dioxide - chloride (ilmenite ore) process
Level 2 waste water treatment for titanium
dioxide - chloride (ilmenite ore) process
General process flow diagram for production
of aluminum fluoride
General process flow diagram at Plant f705
showing the sampling points. (Aluminum fluoride
manufacture)
xxvii
-------
age
503
513
514
515
5! 6
523
524
527
528
532
533
534
555
556
558
559
LIST OF FIGURES - Continued
General process flow diagram at Plant f251
showing the sampling points. Aluminum fluoride
manufacture
Level 1 waste water treatment for aluminum
fluoride subcategory
Level 2 waste water treatment for aluminum
fluoride subcategory
Level 3 waste water treatment for aluminum
fluoride subcategory
Level 4 waste water treatment for aluminum
fluoride subcategory
Annual treatment cost versus production for
the aluminum fluoride subcategory
Annual unit treatment cost versus production
for the aluminum fluoride subcategory
Effect of variation of pollutant load on
treatment cost at level 1 technology
Effect of variation of pollutant load on
treatment cost at level 4 technology
Effect of variation of hydraulic load on
treatment cost at level 2 technology
Effect of variation of hydraulic load on
treatment cost at level 3 technology
Effect of variation of hydraulic load on
treatment cost at level 4 technology
General process diagram for production of
anhydrous chrome oxide
General process diagram for production of
hydrated chromic oxide
General process diagram for production of
chrome yellow
General process diagram for production of
molybdenum orange
xxviii
-------
LIST OP FIGURES - Continued
Page
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-.ll Level 2 waste water treatment for chrome pigments 583
16-12 Annual treatment cost versus production for 590
the chrome pigments subcategory
16-13 Annual unit treatment cost versus production 593
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 f765 showing the sampling
points. (Hydrogen cyanide manufacture)
17-3 General waste water treatment process flow 624
diagram at Plant #782 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
-------
7
1
2
3
4
5
6
7
1
2
3
4
1
2
Page
641
659
662
665
674
675
682
683
709
714
722
723
745
748
LIST OP FIGURES - Continued
Annual unit treatment cost as a function of
production for the hydrogen cyanide subcategory
General process diagram for production of
sodium dichromate
General waste water treatment process flow
diagram at Plant *493 showing the sampling
points. (Sodium dichromate manufacture)
General waste water treatment process flow
diagram at Plant tf376 showing the sampling
points. (Sodium dichromate itianufacture)
Level 1 waste water treatment for sodium dichromate
subcategory
Level 2 waste water treatment for sodium dichromate
subcategory
Relationship of annual treatment cost to production
for the sodium dichromate subcategory
Relationship of annual unit treatment cost
to production for the sodium dichromate subcategory
General block diagram of the manufacture of
copper sulfate
General process flow diagram at plant f034
showing the sampling points. (Copper Sulfate
manu f acture)
Level 1 waste water treatment for copper sulfate
subcategory - batch process
Level 2 waste water treatment for copper sulfate
subcategory - batch process
General process flow diagram for nickel sulfate
manufacture
General waste water treatment process flow
diagram showing sampling points at Plant f369.
(Nickel sulfate subcategory.)
XXX
-------
¦3
•4
¦5
•6
7
8
•1
2
3
4
5
6
7
8
Page
750
751
758
759
765
766
790
791
794
802
803
804
810
811
LIST OF FIGURES - Continued
General process flow diagram at Plant f572
showing the sampling points. (Nickel sulfate
manufacture.)
General waste water treatment process flow
diagram at Plant fl20 showing the sampling
points. (Nickel sulfate manufacture)
Level 1 waste water treatment for nickel sulfate
subcategory - batch process
Level 2 waste water treatment for nickel sulfate
subcategory - batch process
Relationship of annual treatment cost to production
for the nickel sulfate subcategory
Relationship of annual unit treatment cost
to production for the nickel sulfate subcategory
General process flow diagram at Plant f282
showing the sampling points. Sodium bisulfite
manufacture
General flow diagram at Plant f586 showing
the sampling points. Sodium bisulfite manufacture
General process flow diagram at Plant f987
showing the sampling points. Sodium bisulfite
manufacture
Level 1 waste water treatment for sodium bisulfite
subcategory - batch process
Level 2 waste water treatment for sodium bisulfite
subcategory - batch process
Level 3 waste water treatment for sodium bisulfite
subcategory
Variation of annual treatment cost with production
for the sodium bisulfite subcategory
Variation of annual unit treatment cost with
production (sodium bisulfite subcategory)
xxxi
-------
LIST OP FIGURES - Continued
Page
25-1 General process flow diagram at Plant f'672. 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
xxxii
-------
LIST OP TABLES
Page
2-1 Summary of Proposed Regulations - Pest 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
xxxiii
-------
].
2
3
4
5
6
7
8
Q
10
11
1
-1
-2
-3
-4
age
104
105
106
107
108
109
110
111
1.12
J 13
115
132
152
1.53
159
162
LIST OF TABLES - Continued
Waste Water Treatment Options and Performance
Data Summary - Antimony and Arsenic Removal
Waste Water Treatment Options and Performance
Data Summary - Beryllium and Cadmium Removal
Waste Water Treatment Options and Performance
Data Summary - Copper Removal
Waste Water Treatment Options and Performance
Data Summary - Chromium III and Chromium
VI Removal
Waste Water Treatment Options and Performance
Data Summary - Lead Removal
Waste Water Treatment Options and Performance
Data Summary - Mercury II Removal
Waste Water Treatment Options and Performance
Data Summary - Nickel Removal
Waste Water Treatment Options and Performance
Data Summary - Silver Removal
Waste Water Treatment ODtions and Performance
Data Summary - Selenium and Thallium Removal
Waste Water Treatment Options and Performance
Data Summary - Zinc Removal
Estimated Achievable Maximum 30-Day Averages
for the Applied Technologies
Prioritization of Toxic Metals Found in
Each Subcategory
Subcategory Profile Data Summary
Status of Regulations - Effluent Limitation
Guidelines
Summary of Waste Water Flow Data for Chlorine
Mercury Cell Plants
Pollutant Concentration and Loads at Plant
f 299
xxxiv
-------
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Pa3e
163
172
173
183
184
185
188
] 93
195
196
198
199
202
207
208
213
216
LIST OP TABLES - Continued
Pollutant Concentrations and Loads at
Verification Plants (1.)
Toxic Pollutant Raw Waste Concentrations
and Loads at Verification Plants mg/1
kg/kkg
Summary of Raw Waste Loadings at Verification
Plants
Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Unit Treatment Costs
Estimated Chemical Dechlorination Costs
for the Chlor-Alkali Industry
Mercury Discharges from Selected
Chlor-Alkali Mercury Cell Plants*
Residual Chlorine Discharges at Selected
Chlor-Alkali Plants*
Comparison of Raw Waste Concentrations
of Toxic Pollutants with Treatability
Proposed Limitations, EAT
Effluent Concentrations of Toxic Pollutants
from Verification Sampling
Subcategory Profile Data Summary
Status of Regulations - Effluent Limitation
Guidelines
Waste Water Flows at Diaphragm Cell Chlorine
Plants
Pollutant Concentrations and Loads at
Screening and Verification Plants
xxxv
-------
LIST OF TABLES - Continued
Page
11-22
Results of Asbestos Sampling at Diaphragm
Cell Plants
225
11-23
Maximum Raw Waste Concentrations of Toxic
Metals Observed at Diaphragm Cell Chlorine
Plants (mg/1)
226
11-24
Toxic Metal Concentrations and Loads at
Screening and Verification Plants/ mg/1 \
V kg/kkg /
229
11-25
Summary of Raw Waste Loadings at Screening
and Verification Metal Anode Plants
230
11-26
Toxic Metal Concentrations and Loads in
Cell Room Waste Waters at Screening and
Verification Plants/ mg/1 \
\kg/kkg/
232
1] -27
Raw Waste Toxic Metals Concentration and
Loads in Process Streams Other Than Cell
Room Wastes from Screening and Verification
Plants
233
11-28
Raw Waste Toxic Organics at a Graphite
Anode Plant
234
11-29
Raw Waste Toxic Organics by Waste Water
Source at a Graphite Anode Plant
235
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
Plants
255
11-35
Comparison of Toxic Metals Treatability
with Screening and Verification Sampling
Data
257
11-36
Proposed Limitations, BPCTCA
258
xxxvi
-------
LIST OF 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
12-9
12-10
12-11
Page
Lead and TSS Discharges from Selected Diaphragm 261
Cell Chlorine Plants (1)
Toxic Pollutants in Diaphragm Cell Plant 262
Effluents
Proposed Limitations, BAT 265
Proposed Limitations, BCT 269
Proposed Limitations, NSPS 271
Comparison of Raw Waste Characteristics 272
at a New Metal Anode Plant with Treatability
of Toxic Metals
Subcategory Profile Data Summary 276
Status of Regulations - Effluent Limitation 277
Guidelines
Water Usage in the Hydrofluoric Acid 281
Subcategory
Waste Water Flow and Reuse Data for the 282
Hydrofluoric Acid Subcategory
Waste Flow from Hydrofluoric
Manufacturing Plants
Solid Waste Generated at
Acid Plants Sampled
Gypsum Solids Production
Acid Subcategory
Acid 286
Hydrofluoric 287
288
the
in the Hydrofluoric
Flow and Pollutant Concentration Data of 290
the Sampled Waste Streams for Plant #705
Producing Hydrofluoric Acid
Flow and Pollutant Concentration Data of 292
the Sampled Waste Streams for Plants f705,
f251, and 1167 Producing Hydrofluoric Acid
Toxic Pollutant Raw Waste Data 296
Summary of Raw Waste Loadings Found in 297
Screening and Verification Sampling
xxxvii
-------
LIST OF TABLES - Continued
Page
12-12 Model Plant Treatment Costs 310
12-13 Model Plant Treatment Costs 311
12-14 Model Plant Treatment Costs 312
12-15 Model Plant Treatment Costs 316
12-16 Model Plant Treatment Costs 318
12-17 Model Plant Treatment Costs 319
12-18 Model Plant Treatment Costs 320
12-19 Model Plant Treatment Costs 323
12-20 Summary of Waste Water Control and Treatment 325
Technology Employed at Hydrofluoric Acid
Plants
12-21 Summary of Long-Term Monitoring Data from 327
Four Hydrofluoric Acid Plants
12-22 Toxic Pollutant Treated Effluent Data 328
12-23 Development of TSS and Fluoride Limitations 334
12-24 Proposed Limitations, BPCTCA 335
12-25 Proposed Limitations, BAT 343
12-26 Performance of Alternative Technology, 345
Level 3 Treatment
12-27 Performance of Alternative Technology, 346
Level 4 Treatment
12-28 Toxic Pollutant Raw Waste Data Used to 351
Represent New Sources*
12-29 Proposed Limitations, NSPS 353
13-1 Subcategory Profile Data Summary 359
14-1 Subcategory Profile Data Summary 362
14-2 Status of Regulations - Effluent Limitation 363
Guidelines
xxxviii
-------
LIST OP TABLES - Continued
Page
14-3
Water Usage in Titanium Dioxide-Chloride
Process/High Grade Ores Subcategory
367
14-4
Waste Water Flow for Titanium Dioxide-Chloride
Process Subcategory
369
14-5
Flow and Pollutant Concentration Data of
the Sampled Waste Streams of Plant #172
Producing Titanium Dioxide by Chloride-
Rutile Process
372
14-6
Flow and Pollutant Concentration Data of
the Sampled Waste Streams for Plant fl72
Producing Titanium Dioxide (Chloride Process)
374
14-7
Raw Waste Pollutant Data Summary of the
Sampled Streams
376
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
with Variability Factor
393
14-13
Historical Effluent Monitoring Data Summary
with Variability Factors Daily Measurements
394
14-14
Treatment Performance Data of Sampled Plants
1599 and fl72
395
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
1 4-1?
Analysis of Ilmenite Ores
410
14-20
Water Usage in Titanium Dioxide - Sulfate
Process Subcategory
xxxix
413
-------
LIST OP TABLES - Continued
Page
14-21 Raw Waste Characteristics (Industry Data) 416
for Plant f555 (Production of TiO_ by Sulfate
Process)
14-22 Flows and Pollutant Concentrations for 419
the Waste Streams Sampled for Plant f559
Producing Titanium Dioxide
14-23 Process Waste Water Flow at Plants f555, 421
$694 and f559 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 #559
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 Ti0~ Production 459
by the Chloride - Ilmenite Process
14-37 Average Raw Waste Loads for Ti0~ Production 460
by the Chloride - Ilmenite Process
14-38 Summary of Raw Waste Loadings Found in 464
Screening and Verification Sampling
xl
-------
LIST OF TABLES - Continued
Page
-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
Guidelines
- Effluent Limitation
495
15-3
Water Usage in the Aluminum Fluoride
Subcategory
498
15-4
Waste Water Flow at Plants #837, f705 and
f251 for Aluminum Fluoride Subcategory
499
15-5
Solids Generated at Plant ('705 and f251
Producing Aluminum Fluoride
499
15-6
Flow and Pollutant Concentration Data of
the Sampled Waste Streams for Plant ?705
Producing Aluminum Fluoride
502
15-7
Flow and Pollutant Concentration Data of
the Sampled Streams for Plant ?251 Producing
Aluminum Fluoride
504
15-8
Toxic Pollutant Average Raw Waste Loads
and Concentrations
507
15-9
Toxic Pollutant Effluent Concentrations
During Sampling
508
15-10
Summary of Raw Waste Loadings Found in
Screening and Verification Sampling
509
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
15-16 Model Plant Treatment Costs 530
15-1.7 Model Plant Treatment Costs 531
15-18 Model Plant Treatment Costs 535
15-19 Proposed Limitations, BPCTCA 54]
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 Flow 566
16-5 Flow, Pollutant, Concentration and Load 569
Data of the Sampled Waste Streams for Plant
f 002
16-6 Flow, Pollutant, Concentration and Load 572
Data for the Sampled Streams at Plant f894
16-7 Toxic Pollutant Raw Waste Data 574
xlii
-------
LIST OP 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 63 6
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 f765
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
1782
xliii
-------
LIST OP TABLES - Continued
Page
17-9
Summary of Pollutant Raw Waste Loading
Found in Screening and Verification Sampling
629
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
Sampling Results on Total Cyanide and Ammonia
from Plant f765
645
17-15
Statistical Analysis of Historical Effluent
Monitoring Data on Free Cyanide from Plant
f 765
648
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
Guidelines
657
18-3
Water Usage in Sodium Dichromate Subcategory
660
18-4
Flow and Pollutant Concentration Data of
the Sampled Waste Streams for Plant #493
Producing Sodium Dichromate
663
18-5
Flow and Pollutant Loading Data of the
Sampled Waste Streams for Plant f376 Producing
Sodium Dichromate
666
18-6
Flow and Pollutant Loading Data of the
Sampled Waste Streams for Plant f398 Producing
Sodium Dichromate
667
18-7
Toxic Pollutant Raw Waste Data
669
xliv
-------
LIST OF TABLES - Continued
Paqe
18-8 Summary of Raw Waste Loadings Found in 670
Screening and Verification Sampling
18-9 Model Plant Treatment Costs 678
18-1.0 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 Dichromate 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
Gu ideli nes
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 £034
xlv
-------
LIST OF TABLES - Continued
Page
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 f987 and {282 788
for Sodium Bisulfite Subcategory
24-5 Flow and Pollutant Load Data of the Sampled 789
Waste Streams for Plant #282 Producing
Sodium Bisulfite
24-6 Flow and Pollutant Load Data of the Sampled 792
Waste Streams for Plant #'586
24-7 Flow and Pollutant Load Data of the Sampled 793
Waste Streams for Plant #987
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
807
808
809
812
jmmary for 81.4
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 f672 833
Sampl
ing
24-] 1
Model
Plant
Treatment
Costs
24-12
Model
Plant
Treatment
Costs
24-13
Model
Plant
Treatment
Costs
24-14
Model
Plant
Treatment
Costs
24-15
Plant
Performance Evaluation
xlvii
-------
LIST OF TABLES - Continued
Page
25-4 Flow, 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
1672
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-]
Subcategory
Prof ile
Data
Summary
9H
26.38-1
Subcategory
Prof ile
Data
Summary
913
26 .40-1
Subcategory
Prof ile
Data
Summary
915
26.42-1
Subcategory
Profile
Date
Summary
917
26.43-1
Subcategory
Prof ile
Data
Summary
920
26.44-1
Subcategory
Profile
Data
Summary
921
xlix
-------
Intentionally Blank Page
-------
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.
11 Preceding page blank
-------
Intentionally Blank Page
<
-------
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
19 .
Carbon Dioxide
2.
Hydrofluoric Acid
20.
Carbon Monoxide and
3.
Titanium Dioxide
by-product Hydrogen
4.
Aluminum Fluoride
21.
Silver Nitrate
5.
Chrome Pigments
22.
Ammonium Chloride
6.
Hydrogen Cyanide
23.
Ammonium Hydroxide
7 .
Sodium Dichromate
24.
Barium Carbonate
8.
Copper Sulfate
25.
Boric Acid
9 .
Nickel Sulfate
26.
Calcium Carbonate
10.
Sodium Bisulfite
27.
Cuprous Oxide
11.
Sodium Hydrosulfite
28.
Manganese Sulfate
12.
Hydrogen Peroxide
29.
Strong Nitric Acid
13.
Hydrochloric Acid
30.
Oxygen and Nitrogen
14.
Nitric Acid
31.
Potassium Iodide
15.
Sodium Carbonate
32.
Sodium Hydrosulfide
16.
Sodium Metal
33.
Sodium Silicofluorid<
17.
Sodium Silicate
34.
Sodium Thiosulfate
18.
Sulfuric Acid
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), 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
1
-------
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 time, 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 SOBCATEGORIZATION
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
2
-------
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 (3977). The
factors affecting the control and treatment of pollutant
discharges in those industries have been studied in response to
the remanded issues. It has been concluded that alternative
control and treatment technologies to those originally
considered for BAT and NSPS may be appropriate.
3
-------
Intentionally Blank Page
-------
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 pretreatment standards for new and
existing sources be proposed for the following 11 inorganic
chemical manufacturing subcategories:
Table 2-1 summarizes the proposed regulations for Best
Practicable Control Technology Currently Available (BPT).
Summaries of proposed regulations for Best Available Technology
(BAT), 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.
Chlor-Alkali
Sodium Dichromate
Copper Sulfate
Nickel Sulfate
Sodium Bisulfite
Sodium Hydrosulfite
Hydrofluoric Acid
Titanium Dioxide
Aluminum Fluoride
Chrome Pigments
Hydrogen Cyanide
Hydrogen Peroxide
Hydrochloric Acid
Nitric Acid
Sodium Metal
Sodium Silicate
Sulfuric Acid
Sodium Carbonate
5
-------
TABLE 2-1. SUMMARY OF PROPOSED REGULATIONS -
BEST PRACTICABLE CONTROL TECHNOLOGY
CURRENTLY AVAILABLE (BPT)
Effluent Limitations
Subcategory Parameter
Max 24 -hr
30-day Avq Max pH Range
kg/kkg (or lb/1000 lb.) of product
Chlor-alkali,
Mercury Cells
TSS
Mercury
PH
Chlor-alkali, TSS
Diaphragm Cells Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
pH
Hydrofluoric
Acid
Sodium
Dichronate
Titanium
Dioxide
(sulfate
process)
TSS
Flourice (T)
Antimony (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel(T)
Zinc (T)
PH
TSS
Hexavalent
Chromium
Chromium (T)
PH
TSS
Iron (T)
Arsenic (T)
Antimony (T)
Cadmium (T)
Chromium (T)
0.32
0.00014
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
0.0044
0.00050
30
1.2
0.24
0.38
0.070
0.070
0.64
0.00028
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
0.0088
0.0009
110
4.1
0.46
0.71
0.11
0.13
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
(continued)
-------
TABLE 2-1. Continued
Subcategory
Parameter
Effluent Limitations
Max
30-day Avg
24 -hr
Max
kg/kkg (or lb/1000 lb.) of product
pH Range
Titanium
Dioxide
(sulfate
process)
Titanium
Dioxide
(Chloride
Process)
Titanium Diox-
ide (Chloride
Ilmenite Pro-
cess)
Aluminum
Fluoride
Copper Sulfate
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
PH
TSS
Iron (T)
Chromium (T)
pH
TSS
Iron (T)
Antimony (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
(continued)
7
-------
TABLE 2-1. Continued
Subcategory
Parameter
Effluent Limitations
Max 24 -hr
30-day Avg Max
kg/kkgor lb/1000 lb.) of product
pH Range
Hydrogen
Cyanide
TSS 2.0
Armenia-N 4.3
Cyanide (Free) 0.016
Cyanide (T) 0.23
PH
5.4
12
0.043
0.65
6.0 to 10.5
Nickel Sulfate
TSS
Nickel (T)
pH
0.032
0.0020
0.096
0.0060
6.0 to 9.0
Chrane Pigments
TSS
Antimony (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
PH
3.9
0.051
0.020
0.12
0.042
0.15
0.018
0.12
9.4
0.12
0.048
0.29
0.10
0.36
0.043
0.29
6.0 to 9.0
Sodium Bisul-
fite
TSS
COD
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
pH
0.033
1.2
0.00017
0.00075
0.00045
0.00030
0.00075
0.12
3.6
0.00032
0.0014
0.00086
0.00057
0.0014
6.0 to 9.0
Sodium Hydro-
sulfite
TSS
COD
PH
0.12
13
0.44
46
6.0 to 9.0
8
-------
TABLE 2-2. SUMMARY OF PROPOSED REGULATIONS -
BEST AVAILABLE TECHNOLOGY (BAT)
Subcategory Parameter
Effluent Limitations
Max 24 -hr
30-day Avg Max
kg/kkg (or lb/1000 lb.) of product
Chlor-alkali
Mercury Cells
Arsenic (T)
Cadmium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Silver (T)
Zinc (T)
Total Residual
Chlorine
0.00021
0.00011
0.00011
0.00034
0.00010
0.00021
0.00015
0.00042
0.00042
0.00046
0.00024
0.00024
0.00074
0.00022
0.00046
0.00032
0.00092
0.00071
Chlor-alkali
Diaphragm
Cells
Chrcmium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Total Residual
Chlorine
0.00044
0.0035
0.0019
0.00088
0.0035
0.0018
0.00097
0.00077
0.0042
0.0019
0.0077
0.CQ30
Hydrofluoric
Acid
Fluoride (T)
Antimony (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
1.0
0.023
0.0013
0.0097
0.0020
0.0050
0.017
2.2
0.047
0.0027
0.019
0.0040
0.010
0.035
Sodiun
Dichronate
Chromium (T)
Hexavalent
Chromiun
Nickel (T)
Zinc (T)
0.0022
0.00035
0.0012
0.0033
0.0045
0.00070
0.0024
0.0066
(continued)
-------
TABLE 2-2. Continued
Effluent Limitations
Subcategory Parameter
May 24 -hr
30-day Avg Max
KgTkkg-(orT±7TMorTb7r~oF_pc^uct
Titanium
Dioxide Iron (T) 1.2 4.1
Sulfate
Process
Arsenic (T)
0.24
0.46
Antimony (T)
0.38
0.71
Cadmium (T)
0.070
0.11
Chromium (T)
0.070
0.13
Copper (T)
0.24
0.46
Lead (T)
0.14
0.21
Nickel (T)
0.10
0.18
Zinc (T)
0.24
0.52
Titanim
Dioxide
Iron (T)
0.25
0.84
Chloride
Chromium (T)
0.014
0.027
Process
Titanium
Dioxide
Iron (T)
0.30
1.0
Chloride
Antimony (T)
0.096
0.18
Ilmenite
Arsenic (T)
0.060
0.11
Process
Cadmium (T)
0.012
0.019
Chromium (T)
0.012
0.023
Copper (T)
0.060
0.11
Lead (T)
0.036
0.054
Nickel (T)
0.024
0.046
Zinc (T)
0.060
0.013
Aluminum
Fluoride (T)
0.036
0.75
Fluoride
Chromiim (T)
0.00048
0.00096
Nickel (T)
0.0020
0.0040
Chrcme Pigments
Antimony (T)
0.051
0.12
Cadmium (T)
0.020
0.048
Chrcmium (T)
0.12
0.29
Copper (T)
0.042
0.10
Lead (T)
0.15
0.36
Nickel (T)
0.018
0.043
Zinc (T)
0.12
0.29
(oontinued)
10
-------
TABLE 2-2. Continued
Subcategory
Parameter
Effluent Limitations
Max 24-hr
30-day Avg Max
kg/kkg (or lb/1000 lb.) of product
Copper Sulfate
Antimony (T)
0.00038
0.00072
Arsenic (T)
0.00047
0.00089
Cadmium (T)
0.000047
0.000089
Chrcmium (T)
0.000047
0.000089
Copper (T)
0.00038
0.00072
Lead (T)
0.000047
0.000089
Nickel (T)
0.000094
0.00018
Seleniun (T)
0.000094
0.00018
Zinc (T)
0.00038
0.00072
Hydrogen
Anrnonia - N
4.3
12
Cyanide
Cyanide (Free)
0.16
0.043
Cyanide (T)
0.23
0.65
Total Residual
Chlorine
0.011
0.031
Nickel Sulfate
Antiomony (T)
0.00027
0.00081
Chrcmium (T)
0.000010
0.000034
Copper (T)
0.00027
0.00081
Lead (T)
0.000034
0.00010
Nickel (T)
0.00014
0.00042
Zinc (T)
0.00027
0.00080
Sodium
COD
1.2
3.6
Bisulfite
Chrcmium (T)
0.00017
0.00032
Copper (T)
0.00075
0.0014
Lead (T)
0.00045
0.00086
Nickel (T)
0.00030
0.00057
Zinc (T)
0.00075
0.0014
Sodium
Hydrosulfite
COD
Zinc (T)
Nickel (T)
Lead (T)
Chromium (T)
13
0.0024
0.00094
0.0014
0.00047
46
0.0046
0.0018
0.0027
0.00087
11
-------
TABLE 2-3. SUW1ARY OF PROPOSED REGULATIONS -
PRETREATMEOT STANDARDS FOR EXISTING
SOURCES (PSES)
Effluent Limitations
Subcategory Parameter
rax 30-day 24-hr
Ave Max
(mg/1) or (kg/kkg) (mg/1) or (kg/kkg)
Chlor-alkali
Arsenic (T)
0.10
0.00021
0.22
0.00046
Mercury Cells
Cadmium (T)
0.050
0.00011
0.11
0.00024
Copper (T)
0.050
0.00011
0.11
0.00024
Lead (T)
0.16
0.00034
0.35
0.00074
Mercury (T)
0.048
0.00010
0.10
0.00022
Nickel (T)
0.10
0.00021
0.22
0.00046
Silver
0.070
0.00015
0.15
0.00032
Zinc
0.20
0.00042
0.44
0.00092
Chlor-alkali
Diaphragm
Chromium (T)
0.050
0.00044
0.11
0.00097
Cells
Copper (T)
0.40
0.0035
0.88
0.0077
Lead (T)
0.22
0.0019
0.48
0.0042
Nickel (T)
0.10
0.00088
0.22
0.0019
Zinc (T)
0.40
0.0035
0.88
0.0077
Hydrofluoric
Acid
Fluoride (T)
30
1.0
66
2.2
Antimony (T)
0.70
0.023
1.4
0.047
Chromium (T)
0.040
0.0013
0.080
0.0027
Copper (T)
0.29
0.0097
0.58
0.019
Lead (T)
0.060
0.0020
0.12
0.0040
Nickel (T)
0.15
0.0050
0.30
0.010
Zinc (T)
0.52
0.017
1.0
0.035
Sodium Dichro-
rate
Chromium (T)
0.32
0.0022
0.64
0.0045
Hexavalent
Chrcmium
0.050
0.00035
0.10
0.0070
Nickel (T)
0.17
0.0012
0.34
0.0024
Zinc (T)
0.47
0.0033
0.94
0.0066
Titanium
Dioxide
Sulfate
Iron (T)
2.5
1.2
8.5
4.1
Process
Arsenic (T)
0.50
0.24
0.95
0.46
Antimony (T)
0.80
0.38
1.5
0.71
Cackiium (T)
0.15
0.07
0.24
0.11
Chromium (T)
0.14
0.07
0.27
0.13
Copper (T)
0.50
0.24
0.95
0.46
(continued)
12
-------
TABLE 2-3. Continued
Elf fluent Limitations
Subcategory
Parameter
Max 30-day
Avg
(mg/1) or (kg/kkg)
24-hr
Max
(mg/1) or (kg/kkg)
Lead (T)
0.30
0.14
0.45
0.21
Nickel (T)
0.20
0.10
0.37
0.18
Zinc (T)
0.50
0.24
1.1
0.52
Titanium
Dioxide
lion (T)
2.5
0.25
8.4
0.84
Chloride
Chromium (T)
0.14
0.014
0.27
0.027
Process
Titanium
Dioxide
Iron (T)
2.5
0.30
8.5
1.0
Chloride
Ilmenite
Antimony (T)
0.80
0.096
1.5
0.18
Process
Arsenic (T)
0.50
0.060
0.95
0.11
Cadmiun (T )
0.10
0.012
0.16
0.019
Chraniun (T)
0.10
0.012
0.19
0.023
Copper (T)
0.50
0.060
0.95
0.11
Lead (T)
0.30
0.036
0.45
0.054
Nickel (T)
0.20
0.024
0.38
0.046
Zinc (T)
0.50
0.060
1.1
0.013
Aluminum
Fluoride
Fluoride (T)
30
0.36
63
0.75
Chromium (T)
0.040
0.00048
0.080
0.00096
Nickel (T)
0.17
0.0020
0.34
0.0040
Chrcme
Pigments
Antimony (T)
0.48
0.051
1.2
0.12
Cadmium (T)
0.19
0.020
0.46
0.048
Chrcmium (T)
1.1
0.12
2.6
0.29
Copper (T)
0.40
0.042
0.96
0.10
Lead (T)
1.4
0.15
3.4
0.36
Nickel (T)
0.17
0.018
0.41
0.043
Zinc (T)
1.1
0.12
2.6
0.29
Copper
Sulfate
Copper (T)
0.40
0.00038
0.76
0.00072
Nickel (T)
0.10
0.000094
0.19
0.00018
Arsenic (T)
0.50
0.00047
0.95
0.00089
Selenium (T)
0.10
0.000094
0.19
0.00018
Cadmium (T)
0.050
0.000047
0.095
0.000089
Zinc (T)
0.40
0.00038
0.76
0.00072
(continued)
13
-------
TABLE 2-3. Continued
Subcategory
Parameter
Elf fluent Limitations
Max 30-day
Avg
(mg/1) or (ka/kkg)
24-hr
Max
(mg/1) or (kg/kkg)
Chromiun (T)
0.050
0.000047
0.095
0.00089
Lead (T)
0.05
0.000047
0.095
0.000089
Antimony (T)
0.40
0.00038
0.76
0.00072
Hydrogen
Cyanide
Cyanide (Free)
0.27
0.016
0.74
0.043
Cyanide (T)
4.0
0.23
11
0.65
Ammonia-N
75
4.3
210
12
Nickel Sul-
fate
Antimony (T)
0.40
0.00027
1.2
0.00081
Chrcmiun (T)
0.050
0.00010
0.15
0.000034
Cbpper (T)
0.40
0.00027
1.2
0.00081
Lead (T)
0.050
0.000034
0.15
0.00010
Nickel (T)
0.20
0.00014
0.60
0.00042
Zinc (T)
0.40
0.00027
1.2
0.00080
Sodium Bi-
sulfate
COD
680
1.2
2400
3.6
Chromium (T)
0.11
0.00017
0.22
0.00032
Zinc (T)
0.50
0.00075
•1.0
0.0014
Copper (T)
0.50
0.00075
1.0
0.0014
Lead (T)
0.30
0.00045
0.57
0.00086
Nickel (T)
0.20
0.003
0.38
0.00057
Sodium Hydro-
sulfite
ODD
2700
13
9700
46
Zinc (T)
0.50
0.0024
0.95
0.0046
Nickel (T)
0.20
0.00094
0.38
0.0018
Lead (T)
0.30
0.0014
0.57
0.0027
Chraniun (T)
0.10
0.00047
0.19
0.00089
14
-------
TABLE 2-4. SUMMARY OF PROPOSED REGULATIONS -
NEW SOURCE PERFORMANCE STANDARDS
(NSPS)
Effluent Limitations
Subcategory Parameter
Kax 24-hr
30-day Avg Max pH Range
kg/kkg (°r lb/1000 lb.) of product
Chlor-alkali
Mercury Cells
Chlor-alkali
Diaphragm Cells
Hydrofluoric
Acid
Sodium Dichro-
mate
TSS
Arsenic (T)
Cadmium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Silver
Zinc
Total Residual
Chlorine
pH
TSS
Chrcmium (T)
Lead (T)
Total Residual
Chlorine
PH
TSS
Fluoride (T)
Chrcmium (T)
Nickel (T)
Zinc (T)
PH
TSS
Chrcmium (T)
Hexavalent Chrcm-
ium
Nickel (T)
Zinc (T)
pH
0.32
0.00021
0.00011
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
0.00035
0.0012
0.0033
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
0.00070
0.0024
0.0066
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
(continued)
15
-------
TABLE 2-4. Continued
Effluent Limitations
Subcategory Parameter
Max 24 -hr
30-day Avg Max pH Range
kg/kkg (or lb/1000 lb.) of product
Titanium Diox-
ide (Sulfate
Process)
Titanium Diox-
ide (Chloride
Process)
Titan inn Diox-
ide (Chloride
Ilmenite pro-
cess)
Aluminum
Fluoride
TSS
Iron (T)
Arsenic (T)
Antimony (T)
Cadmiun (T)
Chranium (T)
Cbpper (T)
Lead (T)
Nickel (T)
Zinc (T)
PH
TSS
Iron (T)
Chromium (T)
PH
TSS
Iron (T)
Antimony (T)
Arsenic (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
pH
TSS
Fluoride (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
0.0053
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
(continued)
16
-------
TABLE 2-4. Continued
Subcategory Paraneter
Effluent Limitations
Max 24 -hr
30-day Avg Max pH Range
kg/kkg (or lb/1000 lb.) of product
Chrcme Pigments
Copper Sulfate
TSS
Antimony (T)
Cadnium (T)
Chromium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Zinc (T)
pH
TSS
Copper (T)
Nickel (T)
Arsenic (T)
Selenium (T)
Cadnium (T)
Zinc (T)
Chrcmium (T)
Lead (T)
Antimony (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
9.4
0.10
0.0026
0.013
0.013
0.013
0.0026
0.013
0.0050
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
6.0 to 9.0
Hydrogen Cyan-
ide
Nickel Sulfate
(continued)
TSS
Cyanide (Free)
Cyanide (T)
Ammonia-N
Hotal Residual
Chlorine
pH
TSS
Antimony (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
PH
2.0
0.016
0.23
4.3
0.011
0.032
0.00027
0.000010
0.00027
0.000034
0.00014
0.00027
5.4
0.043
0.65
12
0.031
0.096
0.00081
0.000034
0.00081
0.00010
0.00042
0.00080
6.0 to 10.5
6.0 to 9.0
17
-------
TABLE 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
00D
Chrcmium (T)
Zinc (T)
Copper (T)
Lead (T)
Nickel (T)
PH
TSS
COD
Chrcrniur (T)
Lead (T)
Nickel (T)
Zinc
PH
0.033
1.2
0.00017
0.00075
0.00075
0.00045
0.00030
0.18
13
0.00 047
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. StM-lARY OF PROPOSED REGULATIONS -
PRETREATMEOT STANDARDS FOR NEW SOURCES
(PSNS)
Effluent Limitations
Subcategory Parameter
>2ax 24 -hr
30-day Aver Hax
(mg/1) or (Ka/kkg) (mg/1) or (kg/kkg)
Chlor-alkali
Arsenic (T)
0.10
0.00021
0.22
0.00046
Mercury Cells
Cadmium (T)
0.050
0.00011
0.11
0.00024
Copper (T)
0.050
0.00011
0.11
0.00024
Lead (T)
0.16
0.00034
0.35
0.00074
Mercury (T)
0.048
0.00010
0.10
0.00022
Nickel (T)
0.10
0.00021
0.22
0.00046
Silver
0.070
0.00015
0.15
0.00032
Zinc
0.20
0.00042
0.44
0.00092
Chlor-alkali
Chromium (T)
0.050
0.00 044
0.11
0.00097
Diaphragm
Cells
Lead (T)
0.050
0.00044
0.11
0.00097
Hydrofluoric
Acid
Fluoride (T)
30
0.18
63
0.38
Chromium (T)
0.040
0.00024
0.080
0.00048
Nickel (T)
0.15
0.00090
0.30
0.0018
Zinc (T)
0.50
0.0030
1.0
0.0060
Sodium Di-
chromate
Chromium (T)
0.32
0.0022
0.64
0.0045
Hexavalent
Chrcmium
0.050
0.00035
0.10
0.00070
Nickel (T)
0.17
0.0012
0.34
0.0024
Zinc (T)
0.47
0.0033
0.94
0.0066
Titanium Di-
oxide (sul-
Iron (T)
2.5
1.2
8.5
4.1
fate pro-
Arsenic (T)
0.50
0.24
0.95
0.46
cess)
Antimony (T)
0.80
0.38
1.5
0.71
Cadmium (T)
0.15
0.07
0.24
0.11
Chrcmium (T)
0.14
0.07
0.27
0.13
Copper (T)
0.50
0.24
0.95
0.46
Lead (T)
0.30
0.14
0.45
0.21
Nickel (T)
0.20
0.10
0.37
0.18
Zinc (T)
0.50
0.24
1.1
0.52
Titanium
Dioxide
Iron (T)
1.8
0.18
5.9
0.59
(chloride
Chrcmium (T)
0.05
0.005
0.10
0.01
process)
(continued)
-------
TABLE 2-5. Continued
Elf fluent Limitations
Subcategory
Parameter
Max
30-day Avg
(mg/1) or (kg/kkg)
24-hr
Max
(mg/1) or (kg/kkg)
Titanium
Dioxide
Iron (T)
1.6
0.050
5.4
0,017
Chloride
Ilmenite
Antimony (T)
0.80
0.025
1.5
0.048
Process
Arsenic (T)
0.50
0.016
0.95
0.0.30
Cadmium (T)
0.075
0.0023
0.12
0,0037
Chromium (T)
0.040
0.0012
0.076
0.0023
Copper (T)
0.029
0.0090
0.055
0.017
Lead (T)
0.060
0.0019
0.090
0.0029
Nickel (T)
0.17
0.0053
0.32
0.010
Zinc (T)
0.47
0.015
0.99
0.032
Aluminum
Fluoride
Fluoride (T)
30
0.36
63
0.75
Chromium (T)
0.04
0.00050
0.08
0.0010
Nickel (T)
0.17
0.0020
0.34
0.0040
Chrome
Pigments
Antimony (T)
0.40
0.042
0.96
0.10
Cadmium (T)
0.010
0.0011
0.024
0.0026
Chromium (T)
0.050
0.0053
0.12
0.013
Copper (T)
0.050
0.0053
0.12
0.013
Lead (T)
0.050
0.0053
0.12
0.013
Mercury (T)
0.010
0.0011
0.024
0.0026
Nickel (T)
0.050
0.0053
0.12
0.013
Zinc (T)
0.020
0.0021
0.048
0.0050
Copper
Sulfate
Antimony (T)
0.40
0.00038
0.76
0.00072
: Arsenic (T)
0.50
0.00047
0.95
0.00089
Cadmium (T)
0.050
0.000047
0.095
0.000089
Chromium (T)
0.050
0.000047
0.095
0.000089
Copper (T)
0.40
0.00038
0.76
0.00072
Lead (T)
0.050
0.000047
0.095
0.000089
Nickel (T)
0.10
0.000094
0.19
0.00018
Selenium (T)
0.10
0.000094
0.19
0.00018
Zinc (T)
0.40
0.00038
0.76
0.00072
Hydrogen
Cyanide
Antnonia - N
75
4.3
210
12
Cyanide (Free)
0.27
0.016
0.74
0.043
Cyanide (T)
4.0
0.23
11
0.65
(continued)
20
-------
TABLE 2-5. Continued
Elf fluent Limitations
Subcategory Parameter
Max
24-hr
30-day Avg
Max
(mg/1) or
(ka/kkg)
(mg/1)
or (kg/kkg)
Antimony (T)
0.40
0.00027
1.2
0.00081
Chromium (T)
0.05
0.000010
0.15
0.000034
Copper (T)
0.40
0.00027
1.2
0.00081
Lead (T)
0.05
0.000034
0.15
0.00010
Nickel (T)
0.20
0.00014
0.60
0.00042
Zinc (T)
0.40
0.00027
1.2
0.00080
Nickel
Sulfate
Sodium
Bisulfite
COD 680
Chromium (T) 0.11
Zinc (T) 0.5
Copper (T) 0.5
Lead (T) 0.3
Nickel (T) 0.2
1.2 2400
0.00017 0.22
0.00075 1.0
0.00075 1.0
0.00045 0.57
0.00030 0.38
3.6
0.00032
0.0014
0.0014
0.00086
0.00057
Sodium
Hydrosulfite COD 2700
Zinc (T) 0.50
Nickel (T) 0.20
Lead (T) 0. 30
Chromium (T) 0.10
13 9700
0.0024 0.95
0.00094 0.38
0.0014 0.57
0.00047 0.19
46
0.0046
0.0018
0.0027
0.00089
21
-------
TABLE 2-6. SUMMARY OF PROPOSED REGULATIONS -
BEST CONVE«JTICNAL POLLUTANT CONTROL
TECHNOLOGY (BCT)
Subcategory
Parameter
Effluent Limitations
Max 24-hr
30-day Avg Max
kg/kkg (°r lb/1000 lb.) of product
pH Range
Chlor-alkali
Diaphragm Cell TSS
PH
Hydrofluoric
Acid
TSS
P«
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 ERC 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 1, 1987.
The purpose of these proposed regulations is to provide
effluent limitations guidelines for BPT, 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 final 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, 1.975. 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
terms of quantity, for most
navigable waters attributable
chemicals.
parameters which accounted, in
of the pollution loading of
to the manufacture of inorganic
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. 3 976) , 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 (q) - Definition of process waste water
401.11 (r) - Definition of process waste water
pollutant
Chlor-Alkali
415.63 - BATEA
Hydrochloric Acid
415.72 - BPCTCA
415.7 3 - BATEA
415.75 - New sources
Hydrofluoric Acid
415.82
415.83
415.85
BPCTCA
BATEA
New sources
Hydrogen Peroxide
415.93
415.95
BATEA
New sources
Nitric Acid
415.102
415.103
415 .105
BPCTCA
BATEA
New sources
26
-------
Sodium Carbonate
415.152 - BPCTCA
415.153 - BATEA
415.155 - New sources
Sodium Dichromate
415.17 3 - 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
415.222
415.223
415.225
Applicabili ty
BPCTCA
BATEA
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 this 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 four
environmental groups in Natural Resources Defense Council v.
Train, 8 ERC 2120 (June 8, 1976) modified 12 ERC 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 referred to in the literature as "priority
pollutants").
TABLE 3-1. RECOMMENDED LIST OF TOXIC POLLUTANTS
Compound Name
1. *Acenaphthene
2. *Acrolein
3. *Acrylonitrile
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-Dichloroethane
14. 1,1,2-Trichloroethane
15. 1,1,2,2-Tetrachloroethane
16. Chloroethane
*Chloroalkyl ethers (chloromethyl, 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-Trichlorophenol
22. Parachlorometa cresol
23. *Chloroform (trichloromethane)
24. *2-Chlorophenol
~Dichlorobenzenes
25. 1,2-Dichlorobenzene
26. 1,3-Dichlorobenzene
27. 1,4-Dichlorobenzene
~Dichlorobenzidine
28. 3,3'-Dichlorobenzidine
29
-------
*Dichloroethylenes (1,1-dichloroethylene and
1,2-dichloroethylene)
29. 1,1-Dichloroethylene
30. 1,2-Trans-dichloroethylene
31. * 2,4-Dichlorophenol
*Dichloropropane and dichloropropene
32. 1,2-Dichloropropane
33. 1,2-Dichloropropylene (1,3-dichloropropene)
34. *2,4-Dimethylphenol
*Dini trotoluene
35. 2,4-Dinitrotoluene
36. 2,6-Dinitrotoluene
37. *1,2-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. Trichlorofluoronethane
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. 2,4-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
~Polynuclear aromatic hydrocarbons
72. Benzo(a)anthracene (1,2-benzanthracene)
73. Benzo (a) pyrene (3,4-benzopyrene)
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 (1,2,5,6-dibenzanthracene)
83. Indeno (1,2,3-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. 4,4'-DDE (p,p'-DDX)
94. 4,4'DDD (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
*Hexachlorocyclohexane (all isomers)
A-BHC-Alpha
B-BHC-Beta
R-BHC (lindane)-Gamma
G-BHC-Delta
*Polychlorinated biphenyls (PCB's)
106. PCB-1242 (Arochlor 1242)
107. PCB-1254 (Arochlor 1254)
32
102.
103.
104.
105.
-------
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. *S ilver (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.
Chior-Alkali
33.
Sodium Siiicef1uoride
2.
Hydrofluoric Acid
34.
Sodium Thiosulfate
3.
Hydrogen Peroxide
35.
Sulfur Dioxide
4.
Titanium Dioxide
36.
Bromine
5.
Aluminum Fluoride
37.
Calcium Hydroxide
6.
Chrome Pigments
38.
Chromic Acid
7.
Hydrogen Cyanide
39.
Fluorine
8.
Sodium Dichromate
40.
Hyd rcgen
9.
Carbon Dioxide
41.
Iod ine
10.
Carbon Monoxide/Hydrogen
42.
Potassium Chloride
11.
Copper Sulfate
43.
Stannic Oxide
12.
Nickel Sulfate
44.
Zinc Sulfate
13.
Silver Nitrate
45.
Calcium Carbide
14.
Sodium Bisulfite
46.
Caicium Oxide
15.
Sodium Hydrosuifite
47.
Potassium Metal
16.
Hydrochloric Acid
48.
Potassium Sulfate
17.
Nitric Acid
49.
Sodium Bicarbonate
18.
Sodium Carbonate
50.
Borax
19.
Sodium Metal
51.
Ferric Chloride
20.
Sodium Silicate
52.
Lead Monoxide
21.
Sulfuric Acid
53.
Sodium Fluoride
22.
Ammonium Chloride
54.
Aluminum Chloride
23.
Ammonium Hydroxide
55.
Aluminum Sulfate
24.
Barium Carbonate
56.
Potassium Dichromate
25.
Boric Acid
57.
Caicium Chicride
26.
Calcium Carbonate
58.
Sodium Chloride
27.
Copper Oxide
59.
Sodium Sulfite
28.
Manganese Sulfate
60.
Potassium Permanganate
29.
Strong Nitric Acid
61.
Zinc Oxide
30.
Oxygen and Nitrogen
62.
Lithium Carbonate
31.
Potassium Iodide
63.
Ferrous Sulfate
32.
Sodium Hydrosuifide
34
-------
Nickel, Lead, Selenium, Zinc, and Cyanide, which are now
included in the list of toxic pollutants. Other regulated
parameters such as Al, Ba, 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.
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 1.1 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 aualify 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
followi ng:
No. Subcategor y
1.
Hydrogen Peroxide
2.
Carbon Dioxide
3.
Carbon Monoxide/Hvdrogen
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
No. 33, are
of the BAT
icals. This
s during the
course of the present study.
Silver Nitrate, No. 13, and Sodium Silicofluoride,
being deferred for future study under Phase II
regulation development program for Inorganic Chem
deferrment was caused by problems with plant acces
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
pollutants 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 qeneration
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 POR 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 f3rd
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. Costle,
590 F.2d 1011 fD.C. Cir. 1978).
38
-------
3.3.2 BAT Effluent Limitations
The factors considered in assessinq best available
technology economically achievable (BAT) include the age of
equipment and facilities involved, the process employed, process
changes, non-water duality 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 reauired 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 USC 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 ]977 amendments added Section 30](b)(2)(El 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 FR 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 lb.) = $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 Dichromate Subcategory, the cost for removal
of additional conventional pollutants is $13.40 per pound.
Thus, 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/2.21b.) = $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 lb.) = $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/kkq) (kg/2.2 lb.) = $1.09/
0.51 kg/kkg - (1.00 - 0.30) (0 .51 kg/kkg) lb.
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 Perforroance 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 for 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 pretreatment regulations can be found at 40
CFR Part 403, 43 FR 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
-------
Intentionally Blank Page
-------
SECTION 4
SUBCATEGORY AT ION REVIEW
4.1 BASIS FOR SOBCATEGORIZATION
4.1.1 Pactors 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 include:
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.
Dcninant 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.
A 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 0 & 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 subcategor ization 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.
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-Alkali, Titanium
Dioxide, and Hydrogen Cvanide 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 SUBCATEGORIZATION
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 of
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 (FeO), 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 (FeS04) 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 Ti02 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 SUBCATEGORIES
4.3.1 Hydrofluoric Acid and Aluminum Pluoride
Aluminum fluoride (A1F3) usually is produced by the
reaction of hydrated alumina (A1203.3H20) with hydrogen
fluoride (HF), although one plant produces aluminum fluoride
from fluorosilicic acid (H2SiF6), a by-product of phosphoric
acid (H3P04) . 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 (CaF2), generated as a solid
waste, is a disposal problem for both the subcategories
because of its moderate toxicity. Only one additional solid
waste, gypsum (CaS04.2H20), 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 qypsum 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
Sulfate
Chlor ide-Rutile
Chloride-Ilmenite
Andrussow Process
Acrylonitrile Bv-Product
Titanium Dioxide
Hydrogen Cyanide
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 Plants 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
extens ive.
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, 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 Screening and Verification Samplinq
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 reduced 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, Envi ronmental 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
procedu res:
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 —
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,
(HGA) , 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.
Duplicates — 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.
3.
4.
UTD = "Unable To Determine"
interferences.
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
-------
• *lO 9*mM0
WM»V*TlVt AOOCO
IC4 0.AMO MA 9MV10
tn
vo
[fli
ftfCCPT LOO M IAWfl(tt
AMO fM'fbO*«ATC
.11
OOAIIT Y CO««TROi •4.ANK# AMO
OU^llCAIIS CMlAfCO (I)
_zn
IAIAMOAHOal ^ j ATOMIC A*K
Wlt^AHtO J I
P«C»AltATlO«
ACO OKXtTKX
3
3"
A»*0«*T»0*
L'UftHACr
( C«W.»T| )
-
¦ I. ..
| UTO I I OCTiftWMATf
, J
t I AMI AMAl VS«9
t I
I UTO I 1 OCTCRWMATI I
111 I »Si« I
[
riAU( ANAl
t
Lmaui
( A9 Cv.MI.2* )
rl
. r~
Off •CAtE |
•f SM f
OAUTl
9AUKC
Dt TCNMMATC
VALUC
OATA VAC©
VA^OA OCNCMTIOM
(*«>
«tt»AT
°"-*CAit oitimimATi
VAIUC
0*.UTI
SAUM
•election of
»AM*£8 TO U
MIRIO (4)
j MTOmOC OtMfftATlON
I < A* S« )
^•=-7
tr scale I dctcauina
RCSUlT I I VAtUl
J -
OfcUTE
sAuni
RtPt AT
ALAl V3<&
ftCOutfttO
V '
Y
[1VP«NO 1
RtPOWTiHO [
I'iqurc 5-1. Sample flw 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 + 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 v/ere determined by flame only, namely, Ag,
Be, Cu, Cr, Ni and Zn.
2. Four 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 were:
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
reproducibili ty.
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 these 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 result 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
chromatography-mass spectrometry (GC/MS). The pesticides were
analyzed by electron capture gas chromatography followed by
61
-------
TABLE 5-1. ANALYTICAL DETECTION LIMITS FOR METALS (15
Original Screenirvg First Modification Second Modification
Element Protocol ^ of Protocol ^ of Protocol ^
Method (pg/1) Method (jjg/l) Method (pg/1)
Antimony, Sb
HGA*
10
HGA
10
HGA
10
Arsenic, As
HGA
3
HGA
3
Hydride
10
Beryllium, Be
HGA
0.2
Flame
15
Flame
15
Cadmium, Cd
HGA
1
HGA
1
HGA
1
Chromium, Cr
HGA
1
Flame
25
Flame
25
Copper, Cu
HGA
1
Flame
20
Flame
20
Lead, Pb
HGA
10
HGA
10
HGA
10
Mercury, Hg
Cold Vapor
0.5
Cold Vapor
0.5
New Cold
Vapor
0.5
Nickel, Ni
HGA
1
Flame
25
Flarr.e
25
Selenium, Se
HGA
9
HGA
9
Hydride
10
Silver, Ag
HGA
0.5
Flame
15
Flame
15
Thallium, T1
HGA
2
HGA
2
HGA
2
Zinc, Zn
HGA
1
Flame
25
Flame
1
*
Heated Graphite Atomizer
(1) Assuming no matrix interferences requiring dilution of sanple.
(2) EPA Contract No. 68-01-4492 (September 29, 1977), Exhibit C,
"Protocol for the Measurement of Toxic Substances", Environmental
Monitoring 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 Fe, 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 Chromiun (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 T6,
1973 (38 FR 28758) and may be also found in Title 40 of the Code
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 cf
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 ANALYTICAL RESULTS
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
-------
TABI£ 5-2. PCLLL.TAOT FREQUENCY BASES CN SAKPLDG PROGRAM RESULTS
SaCIPDC RAW VRSTE
Pollutant Occurrence Based
Pollutant Occurrence Basad or
Pollutants Detected
oo Plant Grouping
Concentration Classification (pg/1)
5 or <5
>5 but 510
>10 Plant*
<5C
>5C but
>5CG but
>2,500
Plants
Plants
<500
i2,500
Antircny
X
28
19
4
1
Arsenic
X
38
12
3
Beryllior.
X
49
4
Ca&ixiun
X
45
4
4
Chraruun
X
20
13
9
10
Oopper
X
21
16
9
Cyanide
X
2
Lead
X
25
15
7
6
Mercury
46
2
5
Nickel
X
* 1
* '
20
8
8
Selcraun
X
46
7
Silver
X
45
7
1
Thallrjn
X
41
:i
A
Zinc
X
9
la
14
12
Benzene
6
l
Carbon Tetrachloride
2
C-.lorcbcr.zene
1
1,2-Dichloroeth^ne
2
1,1,1-Trichlcroeth^r.o
4
i
Hexachlcroethane
1,1, 2-TrichloroerJ-Ar.e
2
1,1,2,2-Tetrachloroethane
X 1
3
OUorcfcnn
1
X
15
2
1
1,2-Dichlorobenzene
X
1
1,l-Dichioroethyle.-e
X
3
1,2-Dichioropropylene
X
2,6-Di/utro toluene
X
EthylJaer.zene
X
7
1
Flusrar.thene
X
1
Bis(2-Oiloroisopropyl) ether
X
1
Methylene chloride
X
U
3
1
Dichlorobrcnomethane
X
5
Trichlorofluorcroe thane
X
2
1
Chlorori i hrcronethane
X
2
Naphthalene
X
1
1
Nitrophenol
X
1
Pentachloropherol
X
2
1
1
Phenol
X
2
3
Bis(2-Ethylhexyl) phthalate
X
2C
1
1
Butyl benzyl phthalate
X
3
Ol-n-butyl phthalate
X
15
Diethyl phthalato
X
5
Di-nethyl phthalate
X
2
Ber.zo(a) anthracene
X
1
Benzola) pyrene
X
3,4-Benzc ! Ijoroethane
X
Otrysene
X
i
1
Anthracene
X
I
1
Fluorene
X
i
1
Phenanthrene
X
1
Pyrene
X
1
Tetrachloroethylene
X
4
Toljene
X
/
1
Trichlcroe thylene
X
3
Nitrobenzene
X
2
2,4-3mitrochenol
X
2
66
-------
TABLE 5-3. DISTRIBUTION OF POLLUTANTS ACCORDING
TO SUBCATEGORY 1
Pollutants Detected Subcategory Numbers Where Pollutants Found
Antimony
All^but 7, 23, 27, 28, 33
Arsenic
II II II II It t|
Beryllium
II tt It |l It |t
Cacknium
tt It It || It ,|
Chromium
W tt It || It ,t
Copper
ii n tt || tt „
Cyanide
Alf but 7, 23, 27, 28, 33
Lead
•
Mercury
II M II „ II ,,
Nickel
ii ii it „ n „
Selenium
ii ii ii „ ii „
Silver
ii it ii „ it
Thallium
" " " " i.
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-Trichloroethane
1
Hexachloroethane
4,11
1,1,2-Trichloroethane
1, 10, 35
1,1,2,2-Tetrachloroethane
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-Chloroisopropyl) ether
22
Methylene chloride
1, 4, 8, 9, 12, 13, 19, 21, 22, 25,26, 32
, 35
Dichlorobromcme thane
1, 4, 19, 32
Trichlorofluorcmethane
1, 4, 25
Chlorodibrancme thane
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,
30,3]
Butylbenzyl phthalate
1, 2, 12
Di-n-butyl phthalate
1, 4, 8, 11, 17, 18, 19, 21, 22, 30, 31,
34,
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.
-------
TABLE 5-3. Continued
Pollutants Detected Subcategory Numbers Where Pollutants Found
3,4-Benzofluoranthane
8
Chrysene
8
Anthracene
8
Fluorene
8,
12
Phenanthrene
8
Pyrene
8
Tetrachloroethylene
1,
4, 10, 22
Toluene
1,
3, 4, 10,
Trichloroethylene
1,
4, 25
69
-------
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, 2.0E7, and
9.4E4 MFL, respectively, where E is exponential on base 10.
All three plants belong to the Chlor-Alkali subcategory.
69
-------
Intentionally Blank Page
-------
SECTION 6
PROCESS AND WASTE TREATMENT INFORMATION DEVELOPMENT
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 reauired 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
Descr ipt ion
Comments
Manufacturer
Name
Location
EPA Region
Conf idential
Product
Name
Subcategory
Number of other
Products
Inorganic
Chemicals
Plant
Capacity
Production
Age
Fiscal year
1976
1976
1976
Process
Name
Volume of Process
E f fluent
Volume of Noncontact
Effluent
Effluent Treatment
Type
Permi t
Major Pollutants
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 subcategory. 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 base level (BPT) and
advanced level (BAT/NSPS) treatment system design
specifications.
75
-------
Beginning with Section I], 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 age, 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
sal t.
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
-------
Intentionally Blank Page
-------
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
character isties.
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 technologies 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 PRECIPITATION
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++ + 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++ + OH- = M(OH)+ (2)
Metal ion + 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++ +0H-
+ nR
= M (R)nOH+
(3)
Metal ion + 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)
/ Cr (OH)
Zn(OH)-
Ni (OH)
Cd (OH)
ZnS
NiS
. CdS
PbS
,-io
-12
CuS
9 10 11 12 13 14
8
6 7
4 5
0 1
2
3
Figure 7-1. Solubility of metal hydroxides and sulfides
as a function of pH.
82
-------
TABLE 7-1. SOLUBILITY PRODUCTS OF TRACE METALS
Solubility Product Constant (log K )
Metal Hylroxide Sulfide Ethyl Xanthate
Cadmium/ Cd
13.6
26.1
13.6
Copper, Cu
18.6
35.2
-
r, +2
Ferrous, Fe
15.3
16.9
7.1
Lead, Pb
16.1
26.6
16.9
Mercury, Hg
25.4
52.2
37.8
Nickel, Ni
14.8
25.7
11.9
Zinc, Zn
15.7
25.2
8.3
+6
Chranium (VI ) ,Cr
8.9
—
-
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
separate 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 + 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 eliminates the
scaling problem. Total dissolved solids in the form of sodium
salts are increased in the caustic soda treated 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 widely 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 PBRRITE COPRECIPITATION
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
landf illi ng.
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 SULFIDE 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++ + Na2S = MS + 2Na+ (5)
85
-------
Metal ion + sodium sulfide = insoluble metal sulfide +
sodium ions
Figure 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 +
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
redaction 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/1, 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).
Metals Concentration
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 XANTHATE 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:
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 pq/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):
2 [ROCS (=S)Na^ + M++ = [ROCS(=S)2M] + 2Na+
(8)
Concentration, mg/1
Metals
Influent
Ef fluent
Cadmium
Chromium
Copper
I ron
Lead
N ickel
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 solution1. 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 capable of exchanging with cations in
solution. Strongly acidic cation exchangers contain functional
groups such as sulfonates, (-S03H and -S03Na), 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 (-R3N0H and -R3NC1), and weakly basic
exchangers contain ammonia functional groups (-NH30H 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++) 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 + HgCl+ = RSHgCl + H+ (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 REDUCTION 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., Cr04= 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:
-------
3S02 + Cr207= + 2H+ = 2Cr+++ + 3S04= + H20
(12)
Sulfur dioxide + dichromate ion + hydrogen ion = trivalent
chromium ion + sulfates and water
3S03= + Cr207= + 8H+ = 2Cr+++ + 3S04 = +4H20 (13)
Sulfite ion + dichromate ion + hydrogen ion = trivalent
chromium ion + water
6Fe++ + Cr207= + 14H+ = 2Cr+++ + 6Fe+++ + 7H20 (14)
Ferrous ion + dichromate ion + 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+++ + 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/1, 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), 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:
-------
4Hg++ + BH4- + 80H- = 4Hg + B(0H)4- + 4H20 (16)
Mercury ion + borohydride ion + 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,
polysulfides, 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 + CI- (17)
Cyanide + chlorine = cyanogen chloride + chloride ion
CNC1 + 20H- = CNO- + CI- + 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
(9, 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- + 40H- + 3C12 = 6C1- + 2C02 + N2 + 2H20 (19)
Cyanate + hydroxyl ion + chlorine = chloride ion +
carbon dioxide + nitrogen + water
CNO- + 2H20 = C02 + 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
-------
03 + CN- = 02 + CN0- (21)
Ozone + cyanide = oxygen + cyanate ion
H202 + CN- = CNO- + H20 (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 + hypochlorous acid = monochloramine + water, etc.
NH2C1 + HOC1 = NHC12 + H20 (24)
NHC12 + H0C1 = NCI3 + H20 (25)
If excess chlorine is added, chloramines can be converted
into nitrogen oxide(s):
2NH3 + 4H0C1 = N20 + 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
Plati ng
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) .'
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 (UF) are similar in
basic concepts. Both are pressure-driven separation processes
that employ high-flux semi-permeable membranes operating under
96
-------
©
0^
r~ +
t 1
1 ~
1
1
PRODUCT
WA3ER
I
OCNCIOTPATE WASTE
Figure 7-2. Electrodialysis 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
respect ively.
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 pretreatment 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
Packing
Water Flux
at 600 psi
Water Output
Per Unit
Sodium
Parasitic Pressure
Losses(psi) Useful
Density
2 3
(ft /ft )
(gal/
day/ft2)
Volume (gal/
day/ft2)
Chloride
Rejection
Feed
Channel
Product
Channel
pH
Range
Ease of
Cleaning
Pla te-and-Frame
150
10
1500
Very good
30
30
2-8
Fair
Large tubes
50
10
500
Very good
50
10
2-8
Very good
Spiral
250
10
2500
Very Good
10
50
2-8
Good to
very good
Polyumide hoilew
fine fibers
5000
1 (400 psi)
5000
Fair
10
50
0-12
Fair
Cellulose acetate
2500
3(250 psi)
7500
Cood
10
50
3-7
Fair
hollow fine
fibers
Source: Weber, Physicochemical Processes, 1972,
-------
Activated carbon is made by charring basic substrates, such
as wood, coke, 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 aqents 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 BAC) 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 FLDORIDB REMOVAL
The conventional method of treating fluorides-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:
Al (OH)3 + F- = Al(OH)2F + OH- (28)
Aluminum hydroxide + fluoride ion =
aluminum monofluorohydroxide + hydroxyl ion, etc.
Al(OH)2F + F- = Al(OH)F 2 + OH- (29)
Al(OH)F 2 + 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 REMOVAL
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.
S02 + 0C1- + H20 = H2S04 + CI- (31)
Sulfur dioxide + hypochlorite ion + water = sulfuric acid
+ chloride ion
Na2S03 + 0C1- = Na2S04 + CI- (32)
Sodium sulfite + hypochlorite ion = sodium sulfate +
chloride ion
Alternatively, hydrogen peroxide, although relatively
expensive may also be used for dechlorination according to
Equation 33.
H 202 + OC1- = H20 + 02 + CI- (33)
Hydrogen peroxide + hypochlorite ion = water + oxygen +
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 OP 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
-------
TABLE 8-1. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMAKY -
ANTIMONY AND ARSENIC REMOVAL
Treatment Technology pH Initial Final Removal References
Concen- Concen- (%)
tration tration
(mg/1) (mg/1)
Antimony
Lime/Filter
11.5
0.6
0.4
28
40
Ferric chloride/Filter
6.2
0.5
0.2
65
40
Alum/Filter
6.4
0.6
0.2
62
40
Arsenic
Lime Softening
-
0.2
0.03
85
9, 10
Sulfide/Fi1ter
6-7
-
0.05
-
9, 10
Lime (260 mg/1)/Filter
10.0
5.0
1.0
80
41
Lime (600 mg/1)/Filter
11.5
5.0
1.4
72
41
Ferric sulfate
5-7.5
0.05
0.005
90
42
Ferric sulfate
6.0
5.0
0.5
90
41
Lime/Ferric Chloride/
Filter
10.3
3.0
0.05
98
9, 10
Activated alunina
(2 mg/1)
6.8
0.4-10
<0.4
96-99+
43
Activated carbon
(3 mg/1)
3.1-3.6
0.4-10
<4.0
63-97
43
Ferric Chloride
-
0.3
0.05
98
9, 10
Ferric Chloride
-
0.6-0.9
<0.13
-
9, 10
104
-------
TABLE 8-2. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
BERYLLIUM AND CADMIUM REMOVAL
Treatment Technology pH Initial Final Removal References
Concen- Concen- (%)
tration tration
(mg/1) (mg/1)
Beryl lium
Lime/Filter
11.5
0.1
0.006
99.4
40
Cadmium
Lime (260 mg/1)/Filter
10.0
5.0
0.25
95
41
Lime (600 mg/1)/Filter
11.5
5.0
0.10
98
41
Liine Softening
5-6.5
0.44-1.0
0.008
92-98
8
Line/Sulfide
8.5-11.3
0.3-10
0.006
98+
44
Ferrous Sulfide (Sulfex)
8.5-9.0
4.0
<0.01
99+
7,8,11
Ferrite coprecipitation/ neutral
240
0.008
99+
5
Filter
105
-------
TABLE 8-3. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
COPPER REMOVAL
Treatment Technology pH Initial Final Removal References
Concen- Concen- (%)
tration tration
(mg/1) (mg/1)
Lime/Filter 8.5-9.0
3.2
0.07
98
8
Lime (260 mg/1)/Filter 10.0
5.0
0.4
92
41
Lime (600 mg/1)/Filter 11.5
5.0
0.5
91
41
Ferric sulfate/Filter 6.0
5.0
0.3
95
41
Lime >8.5
10-20
1-2
90
9,10
Lime 9.5
3.0
0.2
93
45
Alum 6.5-7.0
3.0
0.2
93
45
Lime/Sulfide 5.0-6.5
50-130
<0.5
-
44
Ferrous sulfide (Sulfex)8.5-9.0
3.2
0.02
99
8
Ferrous sulf ide (Sulfex)8.5-9.0
4.0
0.01
99+
7,6,
Ferrite Coprecipitation/
0.01
99+
5
Filter
106
-------
TABLE 8-4. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SLMftRy -
creatnjM in and chkxilm vi removal
Treatmant Technology
P«
Initial
Concen-
tration
(mg/l)
Final
Concen-
tration
(rag/1)
Rencval
(%)
References
Chrnmum
Line (260 rng/1)/Filter
10.0
5.0
0.1
98
41
Line (600 roj/1)/Filter
11.5
5.0
0.1
98
41
Redact ion/Lute
7-8
140 (as
Cr VI)
1.0
9,10
Reduction/Lima
7-8
1300 (as
Cr VI)
0.06 Crlll
—
3,9,10
Lime Softening
10.6-11.3
0.15
98+
46
Lime/Filter
7-9
0.05
47
Lice
9.5
15
0.1
45
Lime
9.5
3.2
<0.1
45
Ferrite ooprecipitation/
Filter
25
0.01
5
Ferric sulfate
6.5-9.3
98+
46
Ferric sulfata/Fliter
5.0
0.05
99
41
Giromiun VI
tetivated cacbon
(pulverized, Pitts-
burgh type HC)
3.0
10
1.5
85
48
Sane as above
2.0
10
0.4
96
48
Activated cartxn
(granular)
6.0
3
0.05
98
41
Ferrite ooprecipitaticn
0.5
not
detectable
—
5
Sulfur rlinxirVi reduction
0.01-0.1
9,10
Bisulfite reduction
_
—
0.05-1.0
9,10
107
-------
TABLE 8-5. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUt-WARY -
LEAD REMOVAL
Treatment Technology
pH
Initial
Concen-
tration
(mg/1)
Final
Concen-
tration
(mg/1)
Removal
(%)
References
Lime (260 mc/1)
10.0
5.0
0.25
95.0
41
Lime/filter
8.5-9.0
189
0.1
99.9
5
Lime (260 mg/1)/Filter
10.0
5.0
0.075
98.5
41
Lime (600 mg/1)/Filter
11.5
5.0
0.10
98.0
41
Ferrous sulfate/Filter
6.0
5.0
0.075
98.5
41
Sodium hydroxide (1 hour
settling)
5.5
1.6
10
Sodium hydroxide (24 hour
settling)
7.0
0.04
10
Sodium hydroxide/Filter
10.5
1700
0.60
99+
49
Sodium carbonate/Filter
10.1
1260
0.60
99+
49
Sodium carbonate/Filter
6.4-8.7
10.2-70.0
0.2-3.6
82-99+
10
Sodium carbonate/Filter
9.0-9.5
5.0
0.01-0.03
99+
9,10
Ferrous sulfide (Sulfex)
8.5-9.0
189
0.1
99.9
8
Ferrite coprecipitation/
Filter
480
0.01-0.05
99.9
5
108
-------
TABLE 8-6. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMT-'ARY -
MERCURY II REMOVAL
Treatment Technology
pH
Initial
Concen-
tration
(ng/1)
Final
Concen-
tration
(ng/1)
Removal
(%)
Refer*
-
0.3-50.0
0.01-0.12
-
9,10
10.0
10.0
1.8
96.4
50
5.5
16.0
0.04
99
50
4.0
36.0
0.06
99.8
50
00
1
00
o
0.3-6.0
0.01-0.125
87-99.2
50
-
6.0-7.4
0.001-0.005
99.9
5
-
0.01-0.05
<0.0005
-
9,10
-
0.02-0.03
0.009
-
46
0.06-0.09
0.006
50
Sulfide
Sulfide
Sulfide/Filter
Sulf ide/Fi1ter
Sulfide/Filter I
Ferrite coprecipitation/
Filter
Activated Carbon
Activated Carbon/Alum
Activated Carbon
109
-------
TABLE 8-7. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
NICKEL REMOVAL
Treatment Technology
pH
Initial
Concen-
tration
(mg/1)
Final
Concen-
tration
(mg/1)
Removal
(%)
References
Lime
8.5-9.0
75
1.5
98
8
Lime (260 mg/1) /Filter
10.0
5.0
0.3
94
41
Lime (600 mg/1)/Filter
11.5
5.0
0.15
97
41
Caustic Soda/Filter
11.0
-
0.3
-
49
Ferrous sulfide (Sulfex)
8.5-9.0
75
0.05
99.9
8,11
Ferrite coprecipitation
-
1000
0.20
99.9
5
110
-------
TABLE 8-8. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
SILVER REMOVAL
Treatment Technology
pH
Initial
Concen-
tration
(mg/1)
Final
Concen-
tration
(mg/1)
Removal References
(%)
Sodium hydroxide
9.0
54
15
72
13
Ferric sulfate (30 mg/1)
6-9
0.15
0.03-0.04
72-83
46
Line Softening
9.0-11.5
0.15
0.01-0.03
80-93
46
Chloride precipitation
(alkaline chlorination
in the presence of
cyanide)
105-250
1.0-3.5
97+
9,10
Ferric chloride/Filter
6.2
0.5
0.04
98.2
40
Sulfide precipitation
5-11
-
-
very high
9,10
111
-------
TABLE 8-9. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
SELENIUM AND THALLIUM REMOVAL
Treatment Technology
PH
Initial
Concen-
tration
(mg/1)
Final
Concen-
tration
(mg/1)
Removal References
(%)
Selenium
Ferric chloride/Filter 6.2 0.1
Ferric chloride/Filter 6.2 0.05
Alum/Filter 6.4 0.5
Ferric sulfate 5.5 0.10
Ferric sulfate 7.0 0.10
Lime/Filter 11.5 0.5
Lime/Filter 11.5 0.06
Thallium
Lime/Filter 11.5 0.5
Ferric chloride/Filter 6.2 0.6
Alum/Filter 6.4 0.6
0.03
0.01
0.26
0.02
0.03
0.3
0.04
0.2
0.4
0.4
75
80
48
82
75
35
38
60
30
31
40
40
40
51
51
40
40
40
40
40
112
-------
TABLE 8-10. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
ZINC REMOVAL
Treatment Technology
pH
Initial
Concen-
tration
(mg/1)
Final
Concen-
tration
(mg/1)
Removal
(%)
References
Lime/Filter
8.5-9.0
3.6
0.25
93
8
Lire (260 rrg/l)
10.0
5.0
0.85
83
41
Lime (260 mg/l)/Filter
10.0'
5.0
0.80
84
41
Lime (600 mg/1)
11.5
5.0
0.35
93
41
Lime (600 mg/l)/Filter
11.5
5.0
1.2
77
41
Lime/Filter
-
16
0.02-0.23
-
5
Sodium hydroxide
9.0
33
1.0
97
13
Sulfide
-
42
1.2
97
5
Ferrous sulfide (Sulfex)
8.5-9.0
3.6
0.02
99+
8,11
Ferrite coprecipitation
-
18
0.02
99+
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
dosage 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
-------
TABLE 8-11. ESTIMATED ACHIEVABLE MAXIMUM 30-DAY AVERAGES FOR THE APPLIED TECHNOLOGIES
Final Concentrations (mg/1)
Ferrite
Lime Lime Sulfide Coprecip- Soda Ash Soda Ash Alun
Settling Filter Filter itation Settling Filter
Filter
Antimony, Sb
0.8-1.5
0.4-0.8
Arsenic V
0.5-1.0
0.5-1.0
0.05-0.1
Beryllium, Be
0.1-0.5
0.01-0.1
Cadmium, Cd
0.1-0.5
0.05-0.1
0.01-0.1
<0.05
Copper, Cu
0.5-1.0
0.4-0.7
0.05-0.5
<0.05
Chrcmium III,
Cr+3
0.1-0.5
0.05-0.5
0.01
Lead, Pb
0. 5-1.6
0.05-0.6
0.05-0.4
0.20
Mercury II,
Hg
0.01-0.05
<0.01
Nickel, Ni
0.2-] .'j
0.1-0.5
0.05-0.5
Silver, Ag
0.4-0.8
0.2-0.4
0.05-0.2
Selenium, Se
0.2-1.0
0.1-0.5
Thallium, Tl
0.2-1.0
0.1-0.5
Zinc, Zn
0.5-].S
0.4-1.2
0.02-1.2
0.02-0
(continued)
-------
TABLE 8-11 continued
Arsenic V, As
Qircmium VI,
Cr+6
Mercury II,
Hg
Silver, Ag
Selenium, Se
Thallium, T1
Cyanide (Free) ,
CNA
Final Concentrations (mg/1)
Ferric Activated SO2 Bisulfite Lime/FeC^ Alkaline
Chloride Carbon Reduction Reduction Filter Chlori-
nation
0.05-0.5 0.3 0.02-0.1
0.1 0.01-0.1 0.05-0.5
0.01
0.05-0.1
0.05-0.1
0.7
0.1-0.5
-------
8.2 THE USE OP 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 shiqhtly
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:
-In (P/A) = S' (Z - S'/2)
119
-------
K>
O
s
30.0
20.0
10.0
0.01
0.1
1 2 5 10 20 30 40 50 60 70 80 90 95 98 99
99.9
PEPCEWACE
Figure 8-1.
Cumulative distribution of daily concentrations of mercury
effluent from plant #251.
in treated
-------
\
to
C3
0. 30
0. 20
0.10
0.0B
0 .00
0.04
0.0 3
0.02
0.01
0.01
0.1 . r> 1 2 S 10 20 30 40 SO 60 70 80 90 95 98 99
99 .9
PERCENTAGE
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. S' is the estimated standard deviation of the
logarithms of pollutant level measurements. In the
calculations which follow, S' 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 A' and S' 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'+ZS', 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'+ZS' 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 VF, is obtained by dividing P by A,
hence,
VF = P/A = exp(S'(Z - S'/2 )), and
In (VF) = In (P/A) = S'(Z - S'/2)
122
-------
To estimate the VF for a particular set of monitoring data,
where the method of moments, is used, S' is calculated as the
square root of ln(l_.0 + (CV) ) , where the sample coefficient of
variation, CV = S/X, is the ratio of sample standard deviation
to sample average.
Example Calculation of Variability Pactors Prom Long-Term Data
Given the following descriptive statistics for a particular
parameter, as might be found for lead fmg/l)in Appendix A.
No Min Avg Max CV
128 0.002 0.068 0.100 0.609
Calculate the estimated standard deviation of logarithms
(S')2 = In fl.O + 0.6092) = 0.315
S' = 0.56
9
Then
In(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 (VF) = A (P/A) = (0.068) (3.15) = 0.214
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
-------
NOFMAL DISTRIBUTION
(MODEL DENSITY OF LOGARITHMS OF POLLUTION VALUES)
ln(P) = A' + 2.33(S')
Y = In (X) = Logarithm (mg/1)
99%
A'
0
\
LOGNORMAL DISTRIBUTION
(MDDEL DENSITY OF
POLLUTION VALUES)
99%
X (mg/1)
L_ P (Perfomrance Standard)
-A (Long Term Arithmetic Average)
SAMPLE DISTRIBUTION OF N MEASUREMENTS
(LONG TERM MONITORING DATA)
Mm
iri
T
X (Sample Average)
X (mg/1)
Note: (a) S' is estimated as (S*)2 = [Ln(l + CV2)J
CV=S/X
S2= I (X-X)2/(N-1)
X=> LX/N
Figure 8-3. 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 = 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/S, 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
VF = 1 + Z(CV) = 1.0 + 1.64(1.03) « 2.69
P = A (VF) = (0.151) (2.69) = 0.406
125
-------
3.00
2.00
1.00
0.01
",l ,:j:l
4
IH jli
Hi
iTrli" ifltlHr !li!
i
20 30 40 SO 60 70 80
9H 99
99.9
PERCENTAGE
Figure 8—4. Ciirwlative distribution of 30—day averages of total cyanide in treated
effluent from plant #782.
-------
700
000
500
100
300
200
100
i;!l
M » t J
:!|;
1
I
T
i
r„
i"!
11 * ¦
-*tri
i
i
!! r
ill"
• •-1
i I !
; i i -
: I :
o.;;i
C.l
1 2
10 20 10 40 SO 60 70 HO 00 95 9H 99
99 .9
PEIOOTACIE
Fiyiire 8-5. Cumulative distribution of 30-clay averaqes of aimonia in treated
effluent from plant 8782.
-------
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
(MODEL DENSITY OF 30-DAY AVERAGE POLLUTION MEASURZMLNTS)
'i '1 X (mg/l)
L P (Performance Standard)
_ A (Long Term Average)
SAMPLE DISTRIBUTION' OF M MONTHLY AVERAGES
(LONG TERM MONITORING DATA)
Note: (a) P/A = 1+1.64(CV)
CV = S^- /X
(S-)2=(l (X-X)/ (M-l) )
X=l X/M
Figure 8-6. Statistical distributions for 30-day average pollution measuranents.
X (Average of 30-Day Averages)
129
-------
-------
SECTION 9
TREATMENT TECHNOLOGY APPLICATIONS
FOR TOXIC POLLUTANT REMOVAL
9.1 SELECTION OP 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 treatability 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.
presents the significant toxic pollutant metals
group. In general, those metals occurring in the
group are of prime concern and may require regulation,
those occurring in the second group are of somewhat less
Table 9-1
found in each
first
while
concern and are not expected to require regulation.
9.2 APPLICATION OF ADVANCE LEVEL TREATMENT AND CONTROL
ALTERNATIVES
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. PRIORITIZATION CF TCKIC METALS FOUND
IN EACH SUBCATEGORY
SUBCATEGORY
Group 1 111
Group 2(2)
Chlcrine-rt laphragm cell
Chromr-ir.
Antimony
Copper
Arsenic
Lead
Cadmiun
Nickel
Mercury
Zinc
Selemun
Thallium
Oilorine-wftrcury cell
Arsenic
Antinony
Cadmium
CJiroiuun
Copper
Thallium
lead
Mercury
Nickel
Silver
Zinc
Hydrofluoric Acid
Antimony
Arsenic
Chromium
Cacbri'jn
Copper
Mercury
Lead
Selenium
Nickel
Thallium
Zinc
Titanium Dioxide -
Chraniun
Lead
Chloride Process
Nickel
Zinc
Titaniun Dioxide -
Antimony
Seleniur.
Sulfate Process
Arsenic
Thall lur.
and
Cadniun
Qvloride Ilnaiute Process
ChramiuTi
Copper
Lead
Nickel
Zinc
(1) Group 1 - dominant raw waste pollutants selected as control parameters
for the proposed effluent Imitations.
(2) Group 2 - secondary raw waste pollutants fomd 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 cn
the Group 1 pollutants.
(continued)
132
-------
TABLE 9-1 - continued
scbcatbxjg:
Group 1
Grc<^ 2
Aljrur.un Fluoride
Copper
Arsenic
Nickel
Ca£ru.vrr.
O-rcm-m
Mercury
Zinc
Chrcne Piperita
Antimcry
Cyanide
Cadmium
Mercury
Chrcmi'jn
Cyanide
Lead
Nickel
Zinc
Hydrogen Cyanide
Cyanide
5odLijn Dichrcrate
3uurujn
Copper
Nickel
Seler.iun
Zinc
Silver
Copper Sulfate
Antimcr.y
Arsenic
Cadruum
Chrsmusn
Copper
Lead
Nickel
Seleni-jn
Zinc
N'icxel Sulfate
Antirony
Arsenic
Chronuum
Cadniirr.
Copper
Mercury
Selenivxn
Nickel
Thallittn
Zinc
Sodiim Bisulfite
Chroru..sii
Antirony
Copper
Cadmus
Lead
Mercury
Nickel
Zinc
Sodian Hydrosulfite
Chrarravan
Copper
Forrate Process
Lead
Pentachlorophenol
Nickel
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 designed 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 Level 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-term 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 the 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 such 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 POTW's 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. POTW's 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.
9o4.3 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
Z inc
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) colorometric 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
cond i tions.
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 the
contractor, and from unit process equipment costs assembled from
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 Def inition 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 1, 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 waste 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 metr
the summaries for each product s
Direct Investment Costs for Land
ic ton of product are shown in
ubcateqory.
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%
From 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
contingencies.
Operation and Maintenance Costs
Annual operation and maintenace costs are described and
calculated as follows:
Labor and supervision 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)
-------
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
Bag
Quicklime Bulk
Ground Limestone
Soda Ash (58% Bulk)
Caustic Soda (58% NaOH)
Sodium Sulfide (60-62%)
Sulfuric Acid
Hydrochloric Acid (32%)
Aluminum Sulfate (56% Alumina)
Flocculant (Polymer)
Sulfur Dioxide (Ton Containers)
$ 80/metric ton
$ 85/metric ton
$ 70/metric ton
$ 13.20/metric ton
$ 85/metric ton
$200/metric ton
$435/metric ton
$ 75/metric ton
$ 70/metric ton
$250/metric ton
$2.00/kg
$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.
Moni tor i ng, analys is and reporti ng - 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
-------
Amortization
Annual depreciation and capital costs are computed as
follows:
CA = BWr(l+r)n AH+r)n -1 f2)
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.
Itens Not Included in Cost Estimates
In some subcategories, a portion of the waste water is
returned to process from an intermediate treatment step. Ir.
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 cost 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-ALKALI INDDSTRY
11.1 MERCURY CELL PROCESS INDUSTRY PROFILE
11.1.1 General Description
Chlorine and its co-product caustic soda (alkali) are used
in large quantities in the production of plastics, orqanic 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 carbonate and sodium hydroxide to
precipitate impurities such as calcium, magnesium and iron.
The precipitated hydroxides and carbonates are then settled
151
-------
TABLE 11-1. SUBCATEGORY PROFILE DATA SUMMARY
SUBCATEGORY
CHLORINE MERCURY rFT.T.
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 range:
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volune per unit product:
Minimum
Maximm
3,545,000 kkg/year
2,750,000 kkg/year
27
15
1,280,600 kkg/year
1,090,000 kkg/year
36 peroent
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 meters/kkg
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Cornerce, Current Industrial
Reports, Deosnber 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessnent 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 CHLORINE MERCURY rTT.T.
SUBPART F (40 CFR 415.60, 3/12/74)
STANDARDS
Product
Process
Para-
meters
BPCTCA
Max.(1) Avg.(2)
(kg/kkc) (kg/kkg)
BATEA*
Max. Avg.
(};g/}*Jcg) (kg/kkg)
NSPS
Max. Avg.
(kg/kkg)(kg/kkg)
Mercury
No discharge
Cell
TSS
0.64
0.32
of pwwp3
0.64 0.32
Process
Hg
0.00028
0.00014
No discharge
0.00028
0.00014
of pwwp
0.00014 0.00007
* Section 415.63 vas resnanded and is presently reserved (41 FR 51601,
Noveirber 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 we.ter 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 WATBR OSS AND WASTE WATER SOURCE CHARACTERISTICS
11.2.1 Water Dse
Water is used at mercury cell plants for
cooling, tailgas scrubbing, cell washing,
maintenance, floor washings and in the deccmpositon
mercury almagam in the denuder to produce sodium
Because most brine systems at mercury cell plants
systems, water use in the brine system is minimal,
water usage at plants was found to range from 7.6 to
noncontact
equ ipment
of sodium-
hydroxide,
are closed
The total
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 quideline 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 Mud
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 tunqsten. 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
-------
ru*rr(cation
GVSTf-K
ru«iri*n ,
BRim
- ,
CHl0£\Hr.
HCTtCtm*
ckm.
Hfwint L k-.ly.,M fcif tmwotr-* .
^ ceu. J *"*"*** ^JjormwroKrn) p
**mntnr
T~TJ
nomcowtact
COOLIWG n*m
tmwocr.n
cr»eno*poKF.*)
I r
soi sooiun
HYDWOXTDK ROIirTTON
OtMIMUMLHEO
WAT**
wowtowtact
COOL ING fc,
*****
**!««
wms
to WAS-ft
KHti nmv.vz
MrmCOKTACT
CcoifTAcrr*)
COOLING WATER
coot**
IZ3I
«:oolr*
(Mm
rittrm
off use
[ sot SOOIUH f
[btowo* icwr 1
SOLUTION
cowofwsatf
L^-
(SCRUBBINC VAST**)
PAtRACflt:
•AritwAKii Kourrion
MCKWAStf W5CTCU!
KAC*WASW
ni/r**
1
ROLtOS
cwijtic (Lime)
AliO WATf.*
1—'
-J TO
ro
5P
t
™ »STr
*rnr«*i.r
TO SAL*S
To
ATHOSfHEftf
rorn.Twr,
towfh
1
m/w txwti
o^t
"TO
i>Mtirft»L
RMUVPIC
i at in
-- TJ
r
ffRTTCLRP
TO CBUL
I
4P*TD WAST*
.TAIL GAS-
J 1 I 1
"•*] COHPUBSSOA J LIQOW1K* fT1
* ' L_» I pi
LIQUID
HLOIUHK
rtotwcT
„T0 SAI.TS
OK USE
somu« con calcium)
_ Nvror«M>*irt
SOLUTION
to use, saijrs,
O* WASTE
com>k*satc
WfTAK StJLTUmC
acid
nr.rmcKHATiow
nrsTtH
TO WASTR OH |»f.U?!C
usro at aomk plants ohlt
1 T
wnwrowrACT
coolihc
WAIT*
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
the residual water from the chlorine gas after
of cooling. In most cases, the acid is used u
concentration of 50-70 percent is reached. The
be regenerated for reuse, used for pH control
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
dryer to remove
the first stage
ntil a constant
spent acids can
in a treatment
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.
Suamary 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 PLANTS
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 Prograa
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
-------
TABLE 11-3. StMM\RY OF WASTE VATER FLOW DATA PDR CHLORINE MERCURY CFTJ. HANTS
SUBCATEGORY
CHLORINE MERCURY CELL
Plant
Mumber
Vtoste Water Flow
(m^/kkg Chlorine)
317
0.51
907
0.36
299
1.6
167
5.6
747
0.69
343
1.6
106
0.67
131
1.7
589
5.8
898
0.98
741
0.51
553
1.0
769
6.3
Average of 13 plants
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 (i'747 , ?167, fl06 and f 317) 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 f747 , 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 fl67, 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 fl67,
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
-------
swriM •*
1U MXOtUlK
(Ml
(1)1
fcfcJJT* tU
RjVUt
tksm
H*IYt&CtKM
Ci
Cl2 (IWF.
AM)
LHWACTJCW
—W
11,0
(#0) 10
-®~
UMMMLIUS
1/ru.mr
h^o
h.o to mxxss
MUMWffUM
IHM HUQ&&
Mints Ati.
HaUus in I ) mo|4uJ Ut
(L'luunkij, uDcit Airln) ton.
Figure 11-2. General process flew diagram at plant #299 showing the sampling points.
Chlorino/caustic (mercury cell) nvinufacture.
-------
TABLE 11-^1. POLLUTANT CONCENTRATIONS AND LOADS AT PLANT # 299
SUBCATEGORY
CHLORINE CEROJRY CELL)
Stream
Number
Stream
Description
TSS
(mg/1) (kg/kkg)
Mercury
(mg/1) (kg/kkg)
Screening Phase: ^
1 Cell waste
Mercury Treatment
Elf fluent
3 Tail Gas
Scrubber
Verification Phase:
(2)
3
4
5
Mercury Treatment
Influent
Mercury Treatment
Effluent
Cell W&ste
Brine Mud
Tail Gas Scrubber
12
NA
91
18
120
13,000
180
0.016
0.007
NA
0.13
0.026
0.17
NA
0.022
0.15 0.0002
0.029 0.00004
0.11
5.9
0.20
11.
0.54
0.17
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
-------
TABLE 11-5. POLLUTANT CONCENTRATIONS AND LOADS AT VERIFICATION PLANTS (1)
SUBCATEGORY
CHLORINE
(MERCURY CELL)
Stream
Stream
TSS
1ercury
Number
Description
(ng/1)
(kg/KkgT
(nig/1)
(kg/.-^g)
Plant 747
1
2
Cell Waste
Treated Waste
700
60
1.6 x 10~*
1.4 x 10
18
0.10
4.3
2.3
x 10"?
x 10~^
3
Acid Input
NA
NA
0.023
3.5
x 10 ^
4
Acid Output
NA
NA
0.003
7.2
x 10 :
5
Dechlor System
9
0.0037
0.035
1.5
x 10?
6
CI2 Condensate
2
2.7 x 10 b
0.27
1.8
x 10 ^
7
Tail Gas Scrubber NA
NA
0.039
8.0
x 10
Plant 167
5
6
7
8
All CI2 Wastes
Cell wash
Brine Process
Treated Waste
560
57
4
2
1-9 -4
5.7 x 10 "
7.1 x 10 ,
1.3 x 10
3.8
0.72
0.005
0.32
1.3
6.7
9.0
1.8
x 10J
x 10"®
x 10 ^
x 10
9
Clarifier
x 10"3
Underflow
5,900
4.0
10.4
8.7
Plant 317
1
Cell W&ste
45
NA
14
NA
2
Brine Mud
Filtrate
520
NA
34
NA
3
Tank Car Wash
18
NA
0.033
NA
4
Collection
_o
Tank
21,000
8.6
123
5.0
x 10
5
Treated
Effluent
110
4.4 x 10"2
0.10
4.3
x 10"°
6
Deionizer
Effluent
18
5.2 x 10~3
0.001
2.9
x 10^
7
N-C Cooling
16
2.2
0.001
1.4
8
Final Effluent
18
2.4
0.002
3.6
x 10
Plant 106
1
Cell wash
79
3.9
2
Treated Cell
wash
20
0.015
4
Final Effluent
2.0
<0.0005
NA
NA = Not available.
(1) = Data based on three 24-hour composites.
163
-------
vs
NjCI
I'J
13
MINE
STSTEN
CI.
J
DRyinc
CO'.OS.
TO PONOS
65* MjSO^ WASIJ
nut*
e
l*IM(
COOK*
ELECTROLYTIC
CEILS
NONCONTACT HjO
fa
m
COKO.
OICHIO'O.
ru«ir
decoh»osition
MOCCSS WASH
TO PONOS
WASHOOVK M O
ANO
cm VAT(R
OTHf • MODUCT
UASTE WATER
FILTE*
REIO«T
• NtOX
ALMll
I
_L
scruise*
COMMISSION
ANO
LIQUtfACTICW
» CHLORINE
moouct
ro*os
PM
adjustment
• OliCKA*C(
uct^a
SAILING ~OIMTS
sonos to,
lANO'lll
NE»CU»T TO M0CESS
Figure 11-3. General process flow diagram at plant #747 showing the sampling points.
Gilorine/caustic (mercury cell) manufacture.
-------
o
cn
3_. 1
HOMCCMIACf N O
10 WASH
1 lllfft
—
COOU*
f tA(Mtt*M
SOt lOS H 0
ro pun# *
W Cm Oft I MAIM
(JfOOft
mo iuoh
10 WASH
10 AOJtfSt-
MHI AMO » »«Al <
OlUHAItU
Mn
MCTC1I N.O
Vim IftlMf
10M AMO IU0H
10 StOftAtt
(OOil*
-
OiTCt
"l.
cm too*
j
(lit toon
Alio iflUl
17
(UCfCU
M.O
COMTMSSOft
AMO
. COOUII
C,2
fAllS
mtovf fty
2
iaus
(0* 10 *,*0^
Mini ' *
(Ooii«
1a ( m,<
M
((11
wash
f iltCfcAt lOq \A
^ «ar n^io,
' »CMC0«IA£ T
•COOt IMC WAHt "j0
COOUft
(OOUt
CAAftOK
| litiiUMiioa
'(MMC0NIAC1) 10 WASH
(0«0(NlAf(
Mint MOUSS WAICi
«
IACOOM
18
ion
AC 11VATtO
CAtftOM
SANO
nin« |**~|~ (
M
SAMfllllC POINTS
UMOtlMOW 10
lAnorin
/ MUl
-Li.
Wt Aft. KjOC I
lu WASH
Figure 11-4. General process flow diagram at plant #167 showing the sanpling points.
Chlorine/caustic (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 31-5.
At Plant fl06, 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 £589, the waste water going to the mercurv
treatment system consists of cell room washdown, brine filter
backwash, leaks, spills, cleanup water, and hydrogen cooling
condensate. The waste waters are reacted with hvdrochloric 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 £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
-------
SALT
Ci
BfUtC
FUVUyiCATIGN
UQUWACTICN
CI
CMUST1C
SOLA STMCV.
IW* CAR wtai
ro IftCTB TfttXTWfT
DC-IGNItO)
rocxtnaer
ootxjnc;
in disoimck
Figure 11-5, General process flow diagram at plant #317 showing the sarrpling points.
Chlorine/caustic (mercury eel1) 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 i324, 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 f416, 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 t'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 mq/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
(ug/l)
Screening
Plant
Verification
Plants
(f299 , f 747 , f1 67,
Pollu tant
(f299)
f 206 , f 317)
Antimony
<
250
770
Arsenic
<
10
400
Cadmium
<
1
790
Chromium
8
180
Copper
350
2,300
Lead
1
1,900
Mercury
150
180 ,000
Nickel
<
100
2,400
S ilver
<
1
870
Thallium
140
440
Z inc
230
34 ,000
Section 5.1.2 of this report
the screening and verification
chlorine mercury cell industry, a
were conducted at Plants #299,
Thirty-two different sampling points were
various raw waste streams and the treated
describes the methodology of
sampling program. In the
total of 18 days of sampling
f 7 47 , f ] 67 , 17 and *106.
involved covering
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 f299 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
Daily loading
per day)
Where:
is determined by:
(as kg of pollutant
(C) (Q)
1000
C is the concentration of the pollutant expressed
units of mg/1 (Note: kg/m3 = 1000 mg/1), and
Q is the waste stream flow
m3/day (m3, a cubic meter,
gallons).
in
rate expressed in units
is equal to 264.2 U.S.
of
170
-------
Similarly, the unit leadings were calculated from the
reported chlorine production rate, the waste stream flew rate,
and the measured pollutant concentration:
Unit loading (as kg of pollutant per (C)(Q)
kkg of chlorine) = 1000P
Where C and 0 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 lbs).
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 1.1-6, the toxic pollutant raw waste data are
presentee 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
24-hour 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
Pollu tant
(kg/year)
Ant imony
1,400
Arsenic
1,000
Cadmium
210
Chromium
360
Copper
960
Lead
880
Mercu ry
44,000
Nickel
820
S ilver
850
Thai1ium
770
Z inc
7 ,200
171
-------
TABLE 11-6. TOXIC POLLUTANT RAW WASTE CONCENTRATIONS AND LOADS AT
VERIFICATION PLANTS . v
ZjngA \
kg/kkg
SUBCATEGORY
CHLORINE CERCURY GILL)
Pollutant
299
Plant
747
ft
167
317
106
Antimony
0.48
0.11
*
*
0.49
0.00077
0.000078
0.00070
Arsenic
0.23
0.030
0.33
0.10
*
0.00037
0.000021
0.0011
0.00005
Cadmium
0.010
0.020
*
0.46
0.031
0.000016
0.000014
0.00023
0.000044
Chrcnium
0.063
0.10
0.12
0.080
0.013
0.00010
0.000071
0.00040
0.000040
0.000019
Copper
0.30
0.38
0.075
1.2
0.12
0.00047
0.00027
0.00025
0.00060
0.00017
Lead
0.060
0.16
0.072
1.4
0.33
0.000096
0.00011
0.00024
0.00070
0.00047
Mercury
5.9
18
3.8
123
3.9
0.0081
0.0043
0.013
0.048
0.006
Nickel
*
0.093
0.060
1.4
0.17
0.000066
0.00020
0.00070
0.00024
Silver
*
0.047
*
0.11
0.58
0.000033
0.000055
0.00083
Thallium
0.18
0.022
*
*
0.38
0.00029
0.000016
0.00054
Zinc
0.27
0.69
0.17
20
0.96
0.00043
0.00049
0.00057
0.010
0.0014
* _
Concentration below significant level as defined in 11.3.4.
172
-------
TABLE 11-7. SUM-CVRY OF RAW VASTE LOADINGS AT
VERIFICATION PLANTS
SUBCATEGORY CHLORINE (MEPCURY CELL)
Pollutant
Daily
Loadings
(kg/day)
Unit
Loadings
(kg/kkc)
Number of
Plants
Averaged*
min.
avg.
nax.
min.
avg.
max.
Antimony
0.044
0.17
0.30
0.000078
0.00052
0.00077
3
Arsenic
0.0054
0.11
0.27
0.000021
0.00038
0.0011
4
Cadmium
0.0062
0.013
0.025
0.000014
0.000076
0.00023
4
Chrcmium
0.0043
0.037
0.098
0.000019
0.00013
0.00040
5
Oopper
0.045
0.10
0.18
0.00025
0.00035
0.00060
5
Lead
0.036
0.070
0.12
0.000096
0.00032
0.00070
r
D
Mercury
1.6
3.1
5.1
0.0043
0.016
0.048
5
Nickel
0.037
0.056
0.075
0.000066
0.00030
0.00070
4
Silver
0.0059
0.082
0.22
0.000033
0.00031
0.00083
3
Thailinn
0.0090
0.086
0.14
0.000016
0.00028
0.00054
3
Zinc
0.14
0.41
1.1
0.00043
0.0026
0.010
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 POLLUTION 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
eauipment materials of constuction. No toxic orqanics were
found at significant levels.
11.4.2 Prevailing 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
net operating efficiently.
11.4.3 Process Modifications and Technology Transfer Options
The following process modifications are being practiced at
one cr 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 refriqerated
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-site, 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 cnlorine 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 OF 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
granular
and any
for this
treatment level is shown in Figure 11-7.
11.5.2 Equipment for Different Treatment Levels
Equipment Functions
The filtered Level ] effluent is passed through a
activated carbon bed where residual metal sulfides
metallic mercury will be removed. The flow diagram
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
-------
nd xrwrm
vnycu: to
vrtmss
mnv*r
(timmiem)-
CO
suuxmir not)
irt/)nr. if**
|4I AfXHfTtHtHT
smiin
ritra
rnxnr.
nmjcp to
HKIRY I
wmvnnr
f* lAwritJ. |
nniTtrr imn) u*«
|«» Imtf i pH nv>ti|tnr(n|, How monitoring *n«4 sampler
I-igure 11 C. Love! 1 waste water treatment for chlorine - mercury cell subcategory.
-------
Nut* nn»
STWTH
/
mxor. to
wrns
AAnrr&txr
at runic* Arm
1
UTtuwy
(TWT*H1NM17>
*+rrr »*ttp
H-ii)IW. TV*
jn.ire *>
»noiPY
IWWJTf
(» LAwrtu
»>f7ri)rr nmw i.tur
Include# pM monitor Ihr. fl»» m««l»orlng and ampler
I-'iguro 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 Requiresents
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 Hater Flow
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/yr, 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 11-8, 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 the 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 reaulations,
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 1l-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,C57 tons per year)
54 metric tons per day (60 tons per day)
Waste water flow 91 cubic meters per day.
LEVEL OF TREATMENT*
FIRST S0CCND
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 15,00 0 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 CC6TS
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 equipnent
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
$134,500
141,300
9,000
56, %0
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
INVESTMENT COST
Construction
$257,700
$2,000
Equipment in place,
including piping,
fittings, electrical
work and controls
213,200
115,000
Monitoring equipnent
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 C06T
$794,860
$163,800
OPERATION AND
MAINTENANCE COST
Labor and supervision.
$112,000
$14,000
Energy
6,400
Chemicals
5,000
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
AMORTIZATION OF
INVESTMENT COST
$109,311
$26,650
TOTAL AhWUAL COST
$381,342
$83,444
~First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
185
-------
100 500 ' 1000 DAILY FLOW
_J i I j (mVday)
LBVgLi *t
50 100 150 200
PRODUCTION (METRIC TON'S/YEAR X 1000)
Figure 11-8. Annual treatment cost vs. production for the chlorine
subcategory (mercury cell process)
186
-------
LEVEL
50 100 150 200
PRODUCTION (METRIC TONS/YEAR X 1000)
Figure 11-9. Annual unit treatment oost vs. production for the Chlorine
Subcategory (Mercury Cell Process)
187
-------
TABLE 11-11. MODEL PLANT UNIT TREATMENT COSTS
Subcategory CHLORINE Mercury cell
Annual Treatment Costs ($/kkg)
LEVEL OF TOEATMEf/T
COST ITEM PRODUCTION FLOW FIRST SECOND* THIRD FOURTH
(kkg/yr) (m3/day)
Annual Operation
and Maintenance
19,100
91
8.21
1.35
95 , 500
455
2.18
0.42
191,000
910
1.42
0.30
Annual
Amortization
19,100
91
1.50
0.18
95,500
455
0.68
0.15
191,000
910
0.57
0.14
Total Cost
19,100
91
9.71
1.53
95 , 500
455
2.86
0.56
191,000
910
2.00
0.44
* = These costs are incremental to first level costs.
188
-------
Too" 150 200
PRODUCTION (METRIC TONS/YEAR X 1000)
Figure 11-10. Annual treatment cost vs. production for the Chlorine
Subcategory (Mercury Cell Process).
189
-------
LEVEL fl COST RANGE
Figure 11-11.
100 150 200
PRODUCTION (METRIC TONS/YEAR X 1000)
Annual unit treatment cost vs. production for the
Qilorine Subcategory (Mercury Cell Process).
190
-------
600
500
o1 400
LEVEL #2 COST RANGE
300
200
100
150
PRODUCTION (METRIC TONS/YEAR X 1000)
100
200
X
Figure 11-12. Annual treatirent cost vs. production for the Chlorine
Subcategory (Mercury Cell Process).
191
-------
LEVEL #2 COST RANGE
i
l
I
I
1
0 100 150 200
PRODUCTION (METRIC TONS/YEAR X 1000)
Figure 11-13.
Annual unit treatment oost vs. production for the
Chlorine Subcategory (Mercury Cell Process).
192
-------
TABLE 11-12. ESTIMATED CHEMICAL DECHLORINATION' COSTS FOR THE CHLOR-ALKALI
INDUSTRY
SUBCATEGORY CHLORINE
(MERCURY CELL)
Chlorine Production (kkg/yr)
19,100
31,850
191,000
A. INVESTMENT COST
Construction
$3,000
$5,000
$10,000
Equipment in place,
including piping,
fittings, electrical
work and controls
20,100
35,000
50,000
Monitoring equipment
in place
—
—
—
Engineering design
and inspection
4,600
8,000
12,000
Incidentals, overhead,
fees, contingencies....
4,600
8,000
12,000
Land
TOTAL INVESTMENT COST
$32,200
$56,000
$84,000
B. OPERATING AND
MAINTENANCE COST
Labor and supervision..
14,000
26,000
28,000
Energy
500
659
1,220
Chemicals (SO2)
1,500
2,000
15,000
Maintenance
3,220
5,600
8,400
Taxes and insurance
966
1,680
2,520
Residual waste
disposal
—
—
—
Monitoring, analysis,
and reporting
7,500
7,500
7,500
TOTAL OPERATING AND
MAINTENANCE COST
$27,686
$43,439
$62,640
C. AMORTIZATION OF
INVESTMENT COST
$5,239
$ 9,111
$13,660
TOTAL ANNUAL COST
$32,925
$52,550
$76,300
COST PER KKG OF
PRODUCT (Dollars)
1.72
1.65
0.40
19 3
-------
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.
Plow 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 C!ILOR-ALKALI MERCURY
CELL PLAOTS*
SUBCATEGORY
CHLORINE (MERCURY CELL)
Mercury Waste Load (kg/kkg)
Plant
Average
Daily Maximum
Maximum 30-day Average
*343
0.000025
0.00094
0.00029
*907
0.000020
0.00026
0.000030
r898
0.000060
0.0025
0.00043
=195
0.000040
0.00073
0.00015
*106
0.000065
0.00022
0.000096
= 589
0.000055
0.00086
0.00049
f?299
0.000040
0.00019
0.000056
r747**
0.000055
0.000083
0.000065
?317**
0.000006
0.000048
0.000010
=195**
0.000022
0.00066
0.00010
-324**
0.00086
0.0022
0.0018
See Reference 3
Fran Plant Long Term Monitoring Data presented in Appendix A.
195
-------
TABLE 11-14. RESIDUAL CHLORINE DISCHARGES AT SELECTED
CHLOR-ALKALI PLANTS*
Plant
Average
Chlorine Waste Load (kg/kxg)
Range
# 207
0.33
1.4 maximum .
# 014
0.04
0 to 1.29
# 819
ND
0.016 to 0.14
# 747
0.002
C to 0.006
* 106
0.001
0 to 0.14
# 589
0.003
0.001 to 0.011
# 747* *
0.0025
ND
# 324* *
3.72
0.38 to 12.2
*See Reference 3
**Frorr, Plant Long Term 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/kkg.
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
the 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 fc&STE CONCENTRATIONS OF TOXIC POLU7TANTS
WITH TREATABILITY
SUBCATEGORY
CHLORINE
(MERCURY CCLL)
Pollutant
Treatability^
(mg/1)
Maxiimm
Plant
Average
(mg/1)
Average of
5 Plants
(mg/1)
Number Plants out
of Five
Exceeding
Treatability
Level
Antimony
(2)
0.49
< 0.28
(2)
Arsenic
0.05
0.33
0.14
3
Cadmium
0.01
0.46
0.11
3
Chromium
(2>
0.12
0.075
_J2)
Copper
0.05
1.2
0.41
5
Lead
0.10
1.4
0.40
3
Nickel
0.05
1.4
0.35
2
Silver
0.05
0.58
0.15
2
Thallium
(2)
0.38
0.17
(2)
Zinc
0.20
20
4.4
4
(1) Literature-based treatability estimates from Section 8.1. Table 8-11,
given as the lower limit of treatability expressed as a 30-day average.
(2) No data available on treatability with sulfide/filter.
198
-------
TABLE 11-16. PROPOSED LI'lTTATIOMS
Chlorine - Mercury Cell
Best Available Technology
V'&ste Vhter Flow; 2.1 n3/k)cn
SUBCATEXRY CHLORINE MERCLTO rrri.
Subcategory Concentration Basis Effluent Unit
Performance ... (mg/1) (tag/Wo;)
Pollutant ingA) VFR 30-day 24-hour 5£-day 24-hour
avg. nax. avg. max.
?tonoa",ver.tjjorval Pollutants:
Total Residual
Chlorine(6)
0.2
1.7
0.20
0.34
0.00042
0.0C071
Toxic Pollutants
' 5)
Antimony
0.23(3)
2.2
0.23
0.51
— <4>
!4)
(e)
Arsenic
0.10(3)
2.2
0.10
0.22
0.00021
0.00046
(C)
CaAuun
0.050!3)
2.2
0.050
0.11
0.00011
0.00024
Chromium
0 . 040(3)
2.2
C.040
0.088
_
_ «>
Copper'5)
0.050(2)
2.2
0.050
0.11
0.00011
0.00024
Lead <5)
0.16<3)
2.2
0.16
0.35
0.CC034
0.00075
Mercury (5)
0.020!2)
2.2
0.048
0.10
0.00010
0.00022
' 5)
Nickel
0.10(2)
2.2
0.10
0.22
0.C0O21
0.00046
Silver*5'
0.070(3)
2.2
0.070
0.15
0.C0015
0.00032
ThalliiW51
0.17(3)
2.2
0.17
0.37
— (4)
_ (4)
Zinc (5)
C.15(3)
2.2
0.15
0.33
0.00032
0.00070
(1) VFR, the variability factor ratio, is the ratio of the variability factor
for daily maasursrwnts to the variability factor for 30-day average.
(2) Lover limit of treatability for sulfide/filter technology according to
literature treatability data (Table 3-11).
(3) Average effluent concentration from verification sampling.
(4) No load limits proposed; concentration limits are provided for guidance
purposes.
(5) Limits are also applicable to PSES and PSNS and JSPS.
16) Limits axe also applicable to N'SPS.
199
-------
Chlor i ne - 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 mq/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 \ f 1/7 \
\30-day average limit/ \24-hour maximum limit/
= 0.34 mg/1
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
1000 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 \= 0.00071. kg/kkg
V1000 mg/1/
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) {1000 n\g/l\ (2.1 m3/kkg) = 0 .048 mg/1
V kg/m3 )
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 addi tional 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
(sulfide precipitation followed by filtration). Sampling data
for the fifth plant, £299, reflect effluent quality prior to
f iltration.
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 with this.
A. Arsenic: Because the sampling data from five plants
(Table 11-17) indicated an achievable average arsenic
201
-------
TABLE 11-17. EFFLUENT CONCENTRATIONS OF TOXIC POLLUTAiVTS
FROM VERIFICATION SAMPLING
SUBCATEGORY
CHLORINE (MERCURY CELL)
Pollutant
Plant Effluent Concentrations
(mg/1)
Treatability
(mg/1)
(1)
*299
#747
#317
#106
#167
Avg.
Antimony
0.15
<0.25
<0.25
<0.45
<0.065
<0.23
(2)
Arsenic
0.063
<0.010
0.020
<0.005
0.38
<0.096
0.05
Cadnium
0.073
0.120
<0.025
0.016
0.010
<0.050
0.01
Chromium
<0.06
<0.05
<0.05
<0.01
<0.050
<0.044
(2)
Copper
0.038
<0.025
<0.030
0.043
<0.025
<0.033
0.05
Lead
<0.050
0.073
0.170
0.38
0.12
<0.16
0.10
Mercury
0.029
0.10
0.19
<0.0005
0.32
<0.13
0.01
Nickel
<0.050
<0.050
<0.067
0.140
<0.050
<0.074
0.10
Silver
<0.015
<0.015
<0.015
0.260
<0.015
<0.067
0.05
Thallium
0.20
<0.045
<0.25
0.26
0.090
<0.17
(2)
Zinc
0.100
<0.025
0.510
0.088
<0.025
<0.15
0.02
(1) Lower limit frcm literature-based treatability estimates frcm Section 8.1.
(2) No data available for treatability with sulfide/filter.
202
-------
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
V1000 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
V1000 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
V1000 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 0.1.6
mg/1. Thus:
203
-------
(0.16 mo/1)(2.1 m3/kkg) / kg/m3 \ = 0.00034 kg/kkg
V1000 mg/1/
and, applying the VFR 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 mq/1)(2.1 m3/kka) / kg/m3 >\ = 0 .0021 kq/kkg
\J000 mg/1/
and the proposed daily maximum limitation is obtained by
applying the VFR value of 2.2, that is:
(2.2) (0.0002] 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
8-1.1) and is used as the concentration basis for .the proposed
30-day average effluent limitation. Thus:
(0.067 mg/1)(2.1 m3/kkg) ( kg/m3 \ = 0.00014 kg/kkg
V.1000 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)/ kg/m3 N = 0.00032 kg/kkg
U000 mg/1 J
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 iess than 0.17 mg/1 thallium. These relatively high
concentrations are the result of0analytical 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/filter, 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 New Source Performance Standards
For NSPS, 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 pH,
TSS, mercury, arsenic, cadmium, copper, lead, nickel, silver,
zinc and total residual chlorine.
11.7.5 Basis for Proposed Pretreatment 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
zi nc.
205
-------
11.8 DIAPHRAGM CELL PROCESS INDUSTRY PROPILE
11.8.1 General Description
Approximately 65 percent0of 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
-------
TABLE 11-18. SUBCATEGORY PRCFILE DATA SLMMARY
CHLORINE (DIAPHRAGM. CELL)
SUBCATEGORY
Total subcategory capacity rate
"Dotal 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:
Miniimin
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
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 meters/kkg
Sources of data sure Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Cornierce, Current Industrial
Reports, Decenber 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 anc Standards for the Inorganic Chmicals Industry,"
March, 1980.
207
-------
TABLE 11-19. STATUS OF REGULATIONS - EFFLUENT LIMITATION GUIDELINES
SUBCATEGORY CHLORINE (DIAPHRAGM CELL)
SUBPART F (40 CFR 415.60, 3/12/74)
STANDARDS
BPCTCA BATEA NSPS
1 7
Max.x Avg. Max. Avg. Max. Avg.
Product Para- kg/kkg kg/kkg kg/kkg kg/kkg kg/kkc kg/kkg
Process meters (rr.c/1) (mg/1) (rrtg/1) (mc/l) (nc/1) (mc/1)
Diaphragm TSS 0.64 0.32 No discharge 0.64 0. 32
Cell of pwwp
Process
Pb 0.005 0.0025 No discharge 0.00008 0.00004
of pwwp
*
Section 415.63 was remanded and is presently reserved (41 FR 51601,
November 23, 1976).
"Slax. = Maximum of any one day.
2
Avg. = Maximum of daily values for thirty consecutive days.
"^pwwp = 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
sane 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 HATER USE AMD HASTE HATER SOURCES
11.9.1 Hater Ose
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
%
-------
ro
M
O
MINI
IRItICATIOM
- BR1MC MUDS ^tT)
S*STO1
OH ItXTCl* MOMCOMTACT COOLING "ATKH
WOKTMT
COOLIMC
MATfR
1 YD RUG CM
niAntHACM
CRLL
jCflLORIN*
11% TO 121 SOOIUM
ffTDROXIDft
solution
¦AT*R
—t— —1
I rvAronAron U-
—I—
5ULTATF SAL? RRCYCL*
furcr
CHLORINATtt) NATO! CONDCMSAT*
RAI.T
REMOVAL
^ MFAK SULFURIC ACID
sujuwic
CAUSTI
coolimc
SOOIUM
hymoxioc
solution
TO WASTf
TAIL
LtOUCP
FILTER
ATIOW
LKAKS,
srtLU
MASHDO*W
ere.
nuwcontact
COOLIMG
•ATf*
LIQUID
CHUW1ME
MATER
TO fALC8
OR USB
•TO ATMOSFHtRt:
or use
TO ATWKHteHK
A
SOOIUM
soiiumcr hRyrocMLiRim^ to usr
solution salrs, or wast*
T^*
TO USB
Oft SALCS
TO MASTC
DTAfflRAGM
WASH IMC
1
opposition
TO WASTE
UStO $(*& PIAMTS ONLY
• »
w.rrmys uron plakt dcsicm
Figure 11-14. General process flow diagram for production of chlorine/caustic by
diaphragm cells.
-------
maintenance, floor washings and filter backwashing. The
exception at diaphragm cell plants is the use of water 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 Acid
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 Hater
When vapors from caustic evaporators are contact-cooled, a
significant amount of waste water can be generated. Flows 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 0.89 m3/kkg.
Suraary of Haste Hater Plows
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
-------
TABLE 11-20. WASTE WATER FLOWS AT DIAPHRAGM CELL CHLORINE PLANTS
Stream Description
Flow (mVkkg)
Plants with Plant with
Metal Anodes Graphite Anode
min. avg. irax.
Cell roan wastes
and cell wash
Chlorine Condensate
Spent Sulfuric Acid
Tail Gas Scrubber
Caustic Filter Wash
Brine Filter Backwash
Caustic Cooling Blcwdcwn
Brine Mud
0.02 0.38 0.67
0.16 0.49 0.90
0.01
0.17 0.29
0.10
0.82
0.04
NA
NA
0.86 0.89
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 SPECIPIC 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 f014, 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 hydroxide 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 #261, 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 f738 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
-------
NJ
U1
*
"jcHJ
rtmirn-RTftw | 7
1__ 1 H/>
b. _i_
TWI, OASTS
*ALT
mm**
I'H
os/MM! ortuw:
T
1
oom-ifC
%#vrf* li 1
»
jr n
~HT"
t - f
h*»l«
"i
n-Mfitst | m toi«ar
m
cwwne wrrix;
ugt*rKTirtt
MM |4
[mm I
PAT0MT10N I ^
T
HOf
crjnLiHr;
(Mil
ofHiMTwmm
]-
h
mnrntr
comer*
mm to
mi.nr.
v
TT> «AffV«
SOI
t*CM
fvuocntir
OCKOCI9I
WTW W
. t*STK
71% MMI
H)
M SMf>llnq point/*.
Figure 11-15.
General process flew diagram at plant #014 showing the sampling points.
Chlorine/caustic (diaphragm cell) nanufacture.
-------
TABLE 11-21. POLLUTANT CONCENTRATIONS AND LOADS AT
SCREENING AND VERIFICATION PLANTS
SU3CA1UGOr:Y CHiO'TC': DL*J>HRAG4 CEL.,
Plant &
Stream Stream TSS Lead
No. Description (mg/1) (kg/kkg) (mg/1) (kg/kkg)
#014
3
CI 2 oondensate
.2
1.4 x
10~3
0.0055
5.0xl0-6
4
Cell wash
1600
2.4 x
10"2
0.26
3.9xl0-6
5
Brine mud
NA
NA
0.72
1.3xl0~5
6
Bar. condenser
7
3.6
0.005
1.5xl0~3
#261
A
1
Brine mud
NA
NA
0.36
3.0x10
2
Cell wash
4800
1.8 x
10"1
2.0
7.6xl0~5
3
Asbestos filtrate
9
NA
0.075
NA
4
Filter cake
NA
NA
42
NA
5
Bar. condenser
6
NA
< 0.010
NA
#738A1
Cell rocn waste
27
1.4 x
10"3
0.077
3.9xl0~6
2
Asbestos wash
57
7.0 x
10"3
0.031
3.8X10-6
3
Hypo scrubber
290
2.7 x
10"2
0.18
1.7xl0~5
4
CI2 cooling water
35
2.2 x
10"1
0.28
1.3xl0~4
5
Caustic cooling
48
4. 3 x
10"2
0.51
4.5xl0~4
tower
#738B6
Cell rocn waste
95
4.5 x
10" 3
0.067
3.2xl0~6
7
Asbestos wash
72
8.3 x
10" 3
0.13
1.5xl0~5
8
Hypo scrubber
160
1.4 x
10"2
0.20
1.7xl0"5
9
CI2 cooling water
20
1.7 x
10"2
0.20
1.7xl0~5
10
Caustic cooling
4.7
3.8 x
10~3
< 0.010
< 8.2xl0~6
tcMer
/•
11
Chlorate sump
32.
7.0 x
10" 3
< 0.010
< 2.3x10"
12
Plant effluent(B)
63
5.7 x
10"1
0.12
l.lxlO"3
13
Final effluent
58
NA
0.078
NA
(Total)
14
Brine mud
270
NA
0.10
NA
(Continued)
216
-------
TABLE 11-21 (continued)
Plant & Stream TSS Lead
Stream No. Description (mg/1) (kg/kkg) (mg/1) (kg/kkg)
1
Cell wash
934
6.0
x 10~2
0.014
9.1xl0"6
2
Cell rocm drain
283.5
4.6
x 10"3
0.17
2.8xl0-6
3
Brine mud
20,000
33
0.019
3.1xl0-5
4
50% Bar. condenser
32
NA
0.010
NA
5
70% Bar. condenser
21
NA
0.010
NA
6
95% Bar. condenser
90.33
NA
0.010
NA
7
Chlorine condensate
2.4
3.9
1
O
H
X
0.010
1.6xl0~6
)
Cell bldg wastes
1000
1.8
x 10*1
680
1.2xl0-1
2
Lead pond effluent
54
3.0
x 10"2
29
1.6xl0~2
3
Caustic backwash
160
8.6
x 10"1
0.32
1.7xl0"3
4
Brine backwash
13,000
5.8
0.52
2.3xl0"4
5
Cell wash
310
5.6
x 10"2
48
8.6xl0~3
6
Condensate and H2SO4
1100
8.7
x 10"1
0.92
7.3xl0"4
7
Scrubber waste
270
1.2
x 10"2
0.67
2.9xl0~5
NA: Not Available
217
-------
1MIUN
CARBONAH
N«0H
SALT
AW .
I HI
mint
CLARI•
flocculators
*
l\
IDIH( MUOS
AOJACENT PLANT
SANO
f 1 ITERS
1
BACK
WASH
SETTLER
SATURATOR
fur$
_6\K_
CELL
ROOM
•BRINE RECYCLED
tO PROCESS
SLUOGE TO LANDFILL
COOL INC WATER
ORYING
T
HYOROCEN 10 BOILER
COOL INC WATER
MjO CONDENSATE
WASH '
MULTIPLE EFFECT
EVAPORATOR
M
t—*
00
SPENT NjSO^
PURIFICATION
BOTTOM tOVIR
PURIFICATION
TOWER
Clj CONOCNSATC
WATER TO AOJACENT
PLANT FOR USE
LIQUET IER
CAUSTIC
MYPOCHIORITC
TOWf R
-TO VENT
HYPOCHLORITE
SOLUTION TO
AOJACEUT PLANT
COOLING
WATER
STEAM
f CIj STORAGE TANK
CHLORINE AND RAILROAO CAR
TO STORAGE WASHOOVN
7it CAUSTIC
EVAPORATOR
PVRCI TOR DISPOSAL
IT CONTRACT
T
NjOH
0
legend
SAMPLING POINTS
li
-+1t-
FILTER
BAROMETRIC
CONOENSER
FILTRATE TO
' PROCESS SEWER
ASICSTOS TO
LANOFlLl
TO
""^EJECTOR
COOL IMC
WATER
TO BOILER
" *11D WATER
„ TO SU'.E H«E
(fO« LIKE PLANT)
¦PROCESS SEwtR
B.C.
TO EJECTOR
|_ COOLING
WATER
^ PROCESS
SEWER
Figure U-16. General process flow diagram at Plant #261 showing the sampling points.
Chlorine/Caustic (Diaphragm Cell) manufacture
-------
fsj
VO
«p-
tjxnjyr.
MO
OWMKSlrtl
OTtfC
nMirtrATKM
WO
E»r>TUW*TlUN
IMMt I
CJ2J.
RIM
M
• 14
wu«
nw>
V» CMJGTtC
-i
Mien
ftKTHIWR*
i
on.
A.MVMIK
d
nmnr
TO**
*>% KWHRMinM
NO
Ncl wxxvwrr
1 T
»*S7T BWOVTKK*
O
I cntxiwr, jwo
wnroc
»
•v
A.W>RK
SCTTLIMS POO
**pnvi*\c
HjP
I/T710
M r61fM( Ion of All . I
Figure 11-17. General process flowsheet at plant 1738-A showing the sarp^ing points.
Chlorine/caustic (diaphragm call) ranufacture.
-------
to
to
o
OOOLDO
root
gwiyiutfim
V
±
r»Tl "»
' \
bkw | ¦ o
LDC —«
(CNUBTfCt
ou
IMR HyO
aj „
HIM
•1«
0*w«* I RjO *2*>4
i i.
»«
w Bwiwffiqi
m>
0OOUW M»
H
HWD1IUB
DOT OK
muFiuujow
Of WLM10H
Fh
Mm
V
1 i
uamrrHit
*ci.
m&n
"l®4
msn
ccajm
-J?
17
111
nJ*T 070-A
mint mrm f
rum i
msm i
H
• 10
: T1OUT
msn? «*Tt*
^ II)
isz? — O * * —©—
rittw,
oi sumo:
e
IftJM)
Ss^l lnq pint*.
Figure 11-18.
General process flew diagram at plant #738-B showing the sanpling
Chlorine/caustic (diaphragm cell) manufacture.
points.
-------
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 f967 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 Not Sampled
At Plant £999 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 #589, 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 #741, 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
-------
K>
rsj
K>
f fUHrKxrvat
f *'
r
f *****
mwanm m
cm oiFtivK am
V«
*
'^p_>| "sT Jrf
l l Lr * 1
I •)
KMI H£ TO
cesr
oiam
W 0U» MEL
0t'-4TCM.
t1
Oil
ma*
aamct tgrnzmm-
ID :
—C
HO* LJ9JCM
1—U
rJ
M
usr L^JiwowmJC
I (jaiMit*** j 1 snMt
I "
E
QBMCMT 1
m mrm-mi mm*
e
. J' 1 | —I
M|k| utjn^ tlj » I MSfnOTUK |g| I Iwl w^6—
I J " J c&fxxmmm I IaiC32m«3Ui J ^
,-u—.. M " c*rril« )£¦ Lil
* 1 ¦ "•* .'«nci V
cwmc OMK0 «TU
set* to sxmoe wo
Figure 11-19. General process flow diagram at Plant 1736 shewing the sanplinq points
Qilorine/Caustic (Diaphragm Cell) manufacture
-------
K>
W
XI
VMS 4
MMC1
L 1
p^] —
ajuum
wnmeir
am* w»*>
~
* j'M. 4 UMi NMD
i|ti [a*"*
5:
J CKO*
(111
Mom
5fi.
-£>~
r
Qfci SCMJCMSP
»w
I—H *"• *>**j-ffi
0
mwt>
pOMO.
r
1
twuimii
U9#tf»
CI,
j uo»»cno< f-*"!- wwa ~j
*0 Miwwr
NJMT
Figure 11-20. General process flow diagram at Plant 1967 showing the sanpling points.
Chlorine/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 Concentrations
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, f261, f738, 1967, and f736.
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
£014 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 are 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 DIAPHRAS*! CELL PLANTS
Total Asbestos Chrisotile Ampin bole
Plant Stream Fibers (MFL) * MFL MFL
#261
Supply
8.0
108
7.5
CD <
O
H
0.4
Cell Wash
2.1
X
2.1
X
0
Filtered Discharge
1.6
X
103
1.6
X
10J
0
Barometric
Condenser
0.4
0.4
0
#736
Supply
0.7
0.7
io7
0
Cell Wash
2.0
X
107
2.0
X
0
Cell Rocm Waste
2.9
X
102
2.8
X
io2
8
Barometric
Condenser
1.8
0
1.8
Barometric
Condenser
5.3
5.3
0
Barometric
Condenser
1.4
X
102
1.4
X
io2
0
#967
Supply
9.7
X
io2
9.7
X
io2
0
Cell Waste
2.4
X
104
2.4
X
io4
8 X
Pond Effluent
2.4
X
io3
2.4
X
io3
0
Caustic Wash
7.8
X
io3
7.8
X
103
0
Brine Filter
Backwash
8.0
X
io2
6.2
X
io2
1.8
Cathode Wash Waste
3.2
X
105
3.2
X
io5
0
Condensate & Spent
Acid
2.7
X
io2
1.8
X
io2
8.9
Neutralizer Waste
2.1
X
io3
2.1
X
io3
0
~Million fibers per liter
225
-------
TABLE 11-23. MAXIMUM RAW WASTE CONCENT RAT IONS OF TOXIC METALS OBSERVED AT
DIAPHRAGM CELL CHLORINE PLANTS(mg/1)
SUBCATEGORY CHLORINE DIAPHRAGM CELL
Toxic
Plants with
Plant with
Metal
Metal Anodes
Graphite Anode
Antimony
<0.25
<0.065
Arsenic
0.17
0.59
Beryllium
<0.014
<0.001
Cadmium
0.037
0.017
Chromium
7.4
<0.048
Copper
17
0.27
Leac
2.0
44
Mercury
<0.003
0.004
Nickel
22
0.070
Selenium
<0.020
<0.030
Silver
0.018
<0.016
Thallium
<0.25
<0.050
Zinc
3.0
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)
(Flow 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.
227
-------
Unit loading (as kg of pollutant per kkg (C) (Q)
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 lbs).
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 3,1.0 0
Copper 4,400
Lead 470,000
Mercury 48
Nickel 3,600
Silver 5
Zinc 5,10 0
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
-------
TABLE 11-24. TOXIC METAL CONCENTRATIONS AND LOADS AT SCREENING AND
VERIFICATION PLANTS
T
(mg/1)
kg/kkg)
SUBCATEGORY CHLORINE DIAPHRAGM rrr.T.
Plant#
Pollutant 014 261 738A 738B 736 967**
Antimony
*
*
*
*
0.010
0.0000033
0.011
0.00015
Arsenic
*
0.17
0.0000064
*
*
0.011
0.000021
0.057
0.000014
0.30
0.0021
Cadmium
0.002
0.0000018
0.037
0.0000014
*
*
0.025
0.0000061
*
Chromium
0.019
0.000017
1.9
0.000071
0.52
0.0046
0.066
0.0012
0.18 0.004
0.000044 0.000032
Copper
0.015
0.000014
17
0.00064
0.045
0.00039
0.12
0.00023
0.43
0.00011
0.16
0.0011
Lead
0.006
0.0000045
2.0
0.000075
0.082
0.00060
0.11
0.000021
0.016
0.0000039
21
0.015
Mercury
0.002
0.0000018
*
*
*
0.003
0.0000007
0.002
0.000014
Nickel
0.90
0.00081
22
0.00081
0.21
0.0018
0.067
0.00013
0.22
0.000054
0.068
0.00049
Silver
*
0.018
0.0000007
*
*
*
*
Zinc
*
•1.5
0.000054
0.29
0.0021
0.093
0.00018
3.0
0.00074
0.19
0.0014
* Below measurable concentrations
** Graphite Anode plant
229
-------
TABLE 11-25. SIM4AKY OF RAW WASTE LOADINGS AT SCREENING AM) VERIFICATION FETAL ANOCE PLANTS
SUBCATEGORY
Q1LORINE DIAHiKAUl CEII,
K)
Ui
O
Pollutant
Ix>ading
(kg/kkg)
Unit Loading
(kg/kkg)
*Nmber of
Plants
Averaged
(out of 5)
tnin.
avg.
IUX.
rain.
avy.
nax.
Antimony
0.00077
0.00077
0.00077
0.0000033
0.0000033
0.0000033
1
Arsenic
0.0019
0.0084
0.020
0.0000064
0.000017
0.000030
3
CaAiuuo
0.00041
0.00076
0.0014
0.0000014
0.0000032
0.0000061
3
Outxniun
0.0042
0.59
2.8
0.000017
0.00096
0.0046
5
Copper
0.0035
0.12
0.19
0.000014
0.00020
0.00064
5
0.00090
0.094
0.37
0.0000039
0.00016
0.00060
5
Mercury
0.00016
0.00030
0.00044
0.0000007
0.0000012
0.0000018
2
Nickel
0.0066
0.31
1.1
0.000010
0.00057
0.0018
5
Silver
0.00021
0.00021
0.00021
0.0000017
0.0000007
0.0000007
1
Zinc
0.016
2.1
8.0
0.060054
0.00078
0.0021
4
Only those plants where the pollutant was observed at measurable aorvaentratiens.
-------
these two waste mixes were evaluated separately. Table .11-26
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 al
graphite anode plants because of
anode material. They appear
associated with the purification
so occur in waste waters from
the attack of chlorine on the
primarily in waste streams
of chlorine.
231
-------
TABI£ 11-26. TOXIC MTAL OONCQ/TKATlCfC AND IOAC6 IN CF3J. HOCH WASTE HATORS AT SOWWING AND
VERIFICATION PIJWIS / mg/l_\
\ kqAXg )
Plant t
Pollutant
014
261
738A
738B
736
967**
Antimony
•
*
0.050
0.0000081
•
0.038
0.0000031
0.41
0.00015
Arsenic
0.010
0.0000001
0.17
0.0000064
*
»
0.17
0.000014
0.45
0.00017
Cacfcnium
*
0.037
0.0000014
•
•
•
0.016
0.0000059
Chraiius
0.94
0.000014
1.9
0.000071
•
0.075
0.000012
0.54
0.000044
0.086
0.000032
Qapper
0.53
0.0000075
17
0.00064
0.24
0.000042
0.38
0.000061
1.1
0.000090
2.4
0.00089
Le>*d
0.26
0.0000039
2.0
0.000075
0.044
0.0000077
0.11
0.000018
0.047
0.0000038
370
0.14
Mercury
*
•
0.003
0.0000005
*
0.002
0.0000002
0.001
0.0000004
Nickel
54
0.00081
22
0.00081
•
0.061
0.0000098
0.67
0.000055
0.36
0.00013
Silver
•
0.018
0.0000007
•
•
•
*
Zinc
•
1.5
0.000054
0.046
0.0000080
0.46
0.000074
0.58
0.000048
0.92
0.00034
Below detection limits
* * Graphite anode plant
-------
TABLE Ll-27. RAM WASTE TOXIC fETALS OCNQOTRATIGN AM) LQAD6 IN PROCESS STREAMS OTHER THAN CELL ROOM WASTES
FRTH SCKBONr, AM) VERIFICATION PLANTS
Pollutant
|014
I738A
Plant
I738B
(nq/1)
Mtico)
#736
1967
Avg
Antinony
•
*
*
*
»
*
Arsenic
*
•
0.011
•
0.29
0.15
0.000019
0.0020
0.0010
Cacknim
0.002
•
•
0.038
*
0.020
0.0000018
0.0000062
0.0000040
OuomiisB
•
0.53
0.065
•
•
0.29
0.0046
0.00011
0.00014
frmor
"11**
0.004
0.041
0.094
0.090
0.030
0.043
0.0000036
0.00035
0.00016
0.000015
0.00020
0.00014
T*aH
•
0.083
0.11
•
0.40
0.00060
0.00019
0.0027
Marcury
0.002
•
ft
0.003
0.002
0.002
0.0000018
0.0000005
0.000014
0.0000054
Nickel
0.003
0.21
0.067
0.052
0.088
0.0000027
0.0018
0.00012
0.00035
0.00072
Silver
»
•
•
•
•
*
Zinc
•
0.29
0.058
4.3
0.15
1.5
0.0021
0.00010
0.00070
0.0010
0.0037
Below detection limits
-------
TABLE 11-28. RAW WASTE TOXIC ORGANICS AT A GRAPHITE ANODE PLANT
SUBCATEGORY
CHLORINE DIAPHRAGM CELL
Pollutant
Concentration*
(mg/1)
Load
(kg/day)
benzene
0.00040
0.0011
carbon tetrachloride
0.023
0.066
1,2-dichloroethane
0.079
0.23
1,1,1-trichloroethane
0.00014
0.00040
hexachloroethane
0.010
0.029
1,1,2-trichloroethane
0.00040
0.0011
1,1,2,2-tetrachloroethane
0.000044
0.00013
chloroform
0.085
0.24
1,1-dichloroethylene
0.000026
0.000074
2,6-dinitrotoluene
0.000026
0.000074
methylene chloride
0.00056
0.0016
brcmoform
0.000063
0.00018
dichlorobrcmcmethane
0.035
0.10
chlorodibromomethane
0.002
0.0057
hexachlorobutadiene
0.004
0.011
bis(2-ethylhexyl)phthalate
0.00075
0.0022
Ji-n-butyl phthalate
0.00078
0.0022
tetrachloroe thlene
0.036
0.10
toluene
0.0030
0.0086
trichloroethylene
0.020
0.0057
Flow-proportioned concentration
234
-------
TABLE 11-29. RAW WASTE TOXIC ORGANICS BY WASTE WATER SOURCE AT A GRAPHITE
ANODE PLANT
SUBCATEGORY
CHLORINE DIAPHRAGM CELL
Stream
Cell building wastes
Caustic filter backwash
Brine filter backwash
Cell wash
Chlorine condensate and
Spent H2SO4
Scrubber waste
Totals
Total Toxic
Organics
(mg/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 Treataent 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
#967 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 fl95, 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 @ 20
degrees. C) , the waste load would be 0.5 kg/day.
Although 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 #967 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 fl95, 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 noncontact 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 Mod if ied 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" and their 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 life, 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.
Liquefaction 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.
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 eauivalent
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 yater 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
-------
/ "X
* \ yease / *
—\ ^
~"
*+syt i*iu
p-
Q ttaj>nr. v*w
fO
nxxjcuutror.
i
TV
|««51QB ^ _ v -
-------
r~~"Y
J—i
•V
J-
worix m rmrs
K)
Isj
«*CTT!»n I 1
awwtumrt 1 I
fmun'
-<$)
rumurm:
cm. tm
W&lfT,
-8*-
L
Figure 11-22. Level 2 waste water treatment for chlorine
0
-LU-f
(£>
TO MMTtU,
fJI |
diaphragm oell subcategory.
-------
Figure'11-23. Level 3 waste water treatment for chlorine - diaphraqm cell subcategory
-------
11.12.2 Equipment for Different Treatment Levels
Equipnent 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 TREATMENT 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/kkgf 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.
Haste 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/kkg 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 model 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 tail 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.
Chlor i nated organ ic 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 fl95 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.
Although the costs are not included the following information is
provided as guidance. The additional costs for steam stripping
in a plant (such as Plant fl95) 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 orqanic
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 discharqe. All diaphragm cell chlorine
plants are known to be using this technology (Level 1) or its
equ ivalent.
Flow Basis
As described in Section 11.13.1, waste water streams at
diaphragm cell plants are separated into two -types, those that
require treatment for asbestos and metals removal and those that
do not 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 m3/kkg.
248
-------
TABLE 11-30. MODEL PLANT TREATMENT COSTS
Subcategory CHLORINE Diaphragm cell
Production 19,100 metric tons per year ( 21,057 tons per year)
54 metric tons per day ( 60 tons per day )
Waste water flow 68 cubic meters 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 C06T
B. OPERATION AND
MAINTENANCE C06T
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 C06T
TOTAL AhWUAL 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
LEVEL OF TREATMENT*
SECOND
$1,800
17,900
3,940
3,940
$27,580
$14,000
300
2,758
827
7,500
$25,385
$4,487
$29,872
THIRD
$2,250
20,400
4,530
4,530
$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 CC6TS
Subcategory CHLORINE Diaphragm cell
Production 95,500 metric tons per year ( 105,288 tons per year)
272 metric tons per day ( 300 tons per day )
Waste water flew 340 cubic meters 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
AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
FIRST
$148,100
219,700
9,000
75,360
75,360
63,000
LEVEL OF TREATMENT*
SECOND
$590,520
$112,000
4,900
7,500
52,752
17,715
29,000
15,000
$238,867
$85,827
$324,694
$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.
LEVEL OF TREATMENT*
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 C06T
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 AMJUAL 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
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
$44,100
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
251
-------
TABLE 11-33. MODEL PLANT TREATMENT COSTS
Subcategory CHLORINE Diaphragm cell
Annual Treatment Costs ($Akg)
LEVEL OF TREATMENT
COST ITEM PRODUCTION FLOW FIRST SECOND THIRD FOURTH
(kkg/yr) (m3/day)
Annual Operation
and Maintenance
19,100
68
8.83
1.33
1.36
95,500
340
2.50
0.29
0.30
191,000
680
1.66
0.16
0.17
Annual
Amortization
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
Total Cost
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
Not
Applicable
252
-------
soo
EVELS 41 i! 43
400
<2 500
200
100
150
SO
100
200
PRODUCTION (METRIC TONS/YEAR X 1000)
Figure 11-24. Annual treatment cost vs. production for the Chlorine
Subcategory (Diaphragm Cell Process)
253
-------
LEVEL »1 I
i I
MM
i
I i
I . M
! ' i
44—I-
! : I
—rr
i .
i i
50 100 150 200
PRODUCTION (METRIC TCNS/YEAR X 1000)
Figure 11-2 5. Annual unit treatment cost vs. production for the
Chlorine Subcategory (Diaphragm Cell Process)
254
-------
TABLE 11-34.
SUMMARY OF UNIT FLOWS AT DIAPHRAGM CELL PLANTS
SUBCATEGORY CHLORINE DIAPHRAGM CELL
Stream Description Unit Flow Data
(mVWcg) Source
Cell room and cell
wash wastes
1.2
Graphite anode plant
Chlorine condensate
0.78
Graphite anode plant
Tail gas scrubber waste
0.11
Graphite anode plant
Caustic filter wash
5.4
Graphite anode plant
Brine filter wash
0.45
Graphite anode plant
Caustic cooling blowdcwn
0.86
Metal anode plants
average
Spent sulfuric acid
0.01
Metal anode plants
average
Total Unit Flow Discharge 8.8
mVkkg
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.
Conventional 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 #207 (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 of:
256
-------
TABLE 11-35. COMPARISON OF TOXIC METALS TREATABILITY WITH SCREENING
AND VERIFICATION SAMPLING DATA
Treatability
(mg/1)
Level Level
1 2
(1)
Level
3
Maximum
Plant Raw
Waste Average
(mg/1)
Nunnber of Plants
Exceeding
Treatability
Arsenic
0.5
0.5
0.05
0.30
3
Antimony
0.8
0.4
NA
0.011
0
Cadmium
0.1
0.05
0.01
0.037
2
Chrcmiim
0.1
0.05
NA
1.9
4
Copper
0.5
0.4
0.05
17
4
Lead
0.3
0.05
0.05
21
4
Mercury
_J3>
__<3>
0.01
0.003
0
Nickel
0.2
0.1
0.05
22
6
Silver
0.4
0.2
0.05
0.018
0
Zinc
0.5
0.4
0.02
3.0
3
(1) Literature-based treatability estimates from Table 8-11.
(2) Of 6 plants, nunber exceeding treatability by sulfide/filter.
(Level 3)
(3) Treatability with this technology not available.
NA Not Applicable
257
-------
TABLE 11-36. PROPOSED LIMITATIONS
Chlorine - Diaphragm Cell
Best Practicable Control Technology Currently Available
Waste Water Flow: 8.8m3/kkg
Concentration Basis
Effluent
Limit
Pollutant
Max
(kg/kkg)
Subcategory
VFR
^ 30-day
24-hr
Max
30-day
24-hr
Performance
avg.
max.
avg.
max.
(mg/1)
Conventional
Pollutants
TSS
57
2.1
57
120
0.51
1.1
Toxic
Arsenic
0.50(3)
2.6
0.50
1.3
(5)
(5)
Cadmium
0.10{3)
2.6
0.10
0.26
(5)
(5)
Chromium
0.10(3)
2.6
0.10
0.26
0.00088
0.0023
Copper
0.50(3)
2.6
0.50
1.3
0.0044
0.011
Lead
1.1(4)
2.6
1.1
2.9
0.010
0.026
Nickel
0.50(3)
2.6
0.50
1.3
0.0044
0.011
Zinc
0.50(3)
2.6
0.50
1.3
0.0044
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/kkg'N / 1000 mg/^ = 57/mg/l
\ 8.8 m3/kkg/ \ kg/m3 /
and a daily maximum concentration of ]20 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 mq/lN
Q (m3/kkg) ^ kg/m3)
Thus the concentration basis for the maximum 30-day average
for lead is:
/'O.OIO kg/kkg\ / 1000 mq/l\ = \.l 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
(VFR) 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 m3/kkg)/ kg/m3 N = 0.00088 kg/kkg
\1000 mg/1/
and the daily maximum effluent limit is
)/ kg/m3
V1000 mg/1
(0.26 mg/1) (8.8 m3/kkg)f kg/m3 ^ = 0.0023 kg/kkg
C. Copper, Nickel, and Zinc: Raw waste concentrations of
these metals were observed as hiqh 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 FROM SELECTED DIAPHRAGM CELL
CHLORINE PLANTS (1)
SUBCATEGORY CHLORINE - DIAPHRAGM CELL
Plant
Lead Discharge
kg/kkg
Average
Maximum
#589*
0.0020
0.030
#738*
0.0010
0.015
#261*
0.0025
0.019
#014*
0.0060
NA
#967(3)
0.0064
0.026
#207
0.021
0.054
TSS Discharge
kg/kkg
Plant
Average
Maximum
#014*
2.8(2)
NA
#207
0.30
0.57
(1) From Reference 3
(2) Plant has "once-through" barometric condenser water
(3) Long Term Data Appendix A
* Plants with metal anodes
NA: Not Available
261
-------
TABLE 11-38. TOXIC POLLUTANTS IN DIAPHRAGM CELL PLANT EFFLUENTS
Effluent Concentration
( rog/1)
Metal Anode Plant Graphite Anode Plant
#261W #967 . .
Lead Treatment Plant
Pollutant influent effluent influent effluent discharge
Arsenic
0.17
0.12
0.28
0.36
0.30
Cadmium
0.037
0.004
< 0.023
< 0.015
< 0.015
Chromium
1.9
<
0.050
0.10
< 0.050
< 0.050
Copper
17
<
0.025
1.6
0.030
0.031
Nickel
22
<
0.050
0.070
< 0.050
< 0.050
Zinc
1.5
<
0.025
0.93
< 0.10
0.15
^ 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
-------
Using the same VFR 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 kq/kkq
U000 rng/U
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 Limitations
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 proposing 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 2.4-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/kkg, thus the maximum 30-day average is given by:
264
-------
TABLE 11-39. PROPOSED LIMITATIONS
Chlorine Diaphragm Cell
Best Available Technology
Waste Water Flow: 8.8 m3/kkg
Concentration Basis Effluent Limit
Pollutant Treatability VFR(1) Max Max(kg//kkg)
(mg/1) 30-day 24-hr 30-day 24-hr
avg. max. avg. max.
Nonoonventional
Pollutant
Total Residual
Chlorine
0.2
1.7
0.20
0.34
0.0018
0.0030
Toxic Pollutants
Arsenic
0.50(3)
2.2
0.50
1.1
(5)
(5)
Cadmium
0.05(3)
2.2
0.05
0.11
(5)
(5)
Chromium (2)
0.05(3)
2.2
0.05
0.11
0.00044
0.00097
Copper(2)
0.40(3)
2.2
0.40
0.88
0.0035
0.0077
Lead<2)
0.22(4)
2.2
0.22
0.48
0.0019
0.0042
Nickel^
0.10(3)
2.2
0.10
0.22
0.00088
0.0019
Zinc^
0.40(3)
2.2
0.40
0.88
0.0035
0.0077
(1) - VFR: 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
V1000 mg/1 )
The 24-hour maximum limit was calculated similarly,
(0.34 mg/1) (8.8 m3/kkg) / kg/m3 \ = 0.0030 kg/kkg
^1000 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
result 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
V1000 mg/1/
The variability factor ratio (VFR) 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/kka
U000 mg/1 )
266
-------
and, for the daily maximum limitation using the VFR value cf
2.2, one obtains:
(2.2) (0.00044 kg/kkg) = 0.00097 kg/kkg
The corresponding concentration basis is:
(2.2) (0.050 ma/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-dav
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 "N = 0.0035 kq/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-dav average
loading limitation of 0.00088 kg/kkg. That is,
(0.10 mg/1) (8.8 m3/kkg) ( kg/m3 \ = 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 maximumc 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.
Plow 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 1.2 mg/1 using the
diaphragm cell model plant discharge flow rate of 8.8 m3/kkg,
namely:
(12 mg/11 (8.8 m3/kkg) / kg/m3 \ = 0.10 kg/kkg
V1000 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. PROPOSED LIMITATIONS
Chlorine Diaphragm Cell
Best Conventional Technology
Waste Water Flew:8. 8 mVWcg
Pollutant
Treatability
(mg/1)
Concentration Basis
VPrCD Max to?/!)
30-day 24-hr
avg. max.
Effluent Limit
Max(kg/kkg)
30-day 24-hr
avg. max.
Total Suspended
Solids
(2)
12
1.9 12
23
0.10
0.20
(1) - VFR: 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 frcm 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 11-4].
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 are 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 san^e as in BAT
regulations.
C. Total Residual Chlorine: Limitations for total
residual chlorine are the same as in BAT regulations.
270
-------
TABLE 11-41. PROPOSED LIMITATIONS
Chlorine Diaphragm Cell
New Source Performance Standards
Waste Water Flo/: 8.8 m 3/kkg
Pollutant
Treatability
(mg/1)
VFR
(1)
Concentration Basis
Max (mg/1)
30-day 24-hr
avg.
max.
Effluent Limit
Max k^g
30-day 24-hr
avg. max.
Conventional and
Non-Conventional
TSS 12
Total Residual
Chlorine 0.2
1.9
1.7
12
0.2
23
0.34
0.10
0.20
0.0018 0.0030
Toxic Pollutants
Arsenic
0.50
2.2
0.50
1.1
(3)
(3)
Cadmium
0.050
2.2
0.050
0.11
(3)
(3)
Chromium^)
0.050
2.2
0.050
0.11
0.00044
0.00097
Copper
0.40
2.2
0.40
0.88
(3)
(3)
Lead^
0.050
2.2
0.050
0.11
0.00044
0.00097
Nickel
0.10
2.2
0.10
0.22
(3)
(3)
Zinc
0.40
2.2
0.40
0.88
(3)
(3)
(1) - VFR: ratio of the 24 hour variability factor to the 30-day
variability factor
(2) - Also applicable to PSNS limitations
(3) - No effluent limitation proposed
271
-------
TABLE 11-42. COMPARISON OF RAW WASTE CHARACTERISTICS AT A NEW METAL ANODE
PLANT WITH TREATABILITY OF TOXIC METALS
SUBCATEGORY CHLORINE DIAPHRAGM CELL
Conoentrat ion(mg/D
Pollutant Treatability^ Plant #738B(2)
Raw Waste
Arsenic
0.50
0.011
Cadmium
0.050
<0.025
Chromium
0.050
0.066
Copper
0.40
0.12
Lead
0.050
0.11
Nickel
0.10
0.067
Zinc
0.40
0.093
(1)- Literature based treatability estimates using BAT technology
of dual media filtration following alkaline precipitation of
metals (Table 8—11)
(2)- Verification sanpling 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 mq/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 VFR
value of 2.2, one obtains:
(2.2) (0.00044 kg/kkg) = 0.00097 kg/kkg
The concentration basis for the daily maximum is,
(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
HF 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 » CaS04 + 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 HF.
The analysis of a typical upgraded fluorspar is given as:
CaF2 Minimum 97.5-98%
Si02 Maximum 1.0%
S " 0.05%
275
-------
TABLE 12-1
SUBCATEGORY PROFILE DATA SUMMARY
SUBCATEGORY
HYDROFLUORIC ACID
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 range:
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimun
Maximum
Waste water flow range:
Minimum
Maximum
Volune per unit product:
Minimun
Maximum
363,000 kkg/year
261,800 kkg/year
9
8
*
177,000 kkg/year
*
68 percent
7,300 kkg/year
62,000 kkg/year
22,100 kkg/year
15,800 kkg/year
83 percent
7 years
58 years
0 cubic meters/day
4,700 cubic meters/day
0 cubic meters/kkg
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.; 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
-------
TABLE 12-2 -
STATUS OF REGULATIONS - EFFLUENT LIMITATION GUIDELINES
SUBCATEGORY
SUBPART
Hydrofluoric A=id
H (40 CFR 415.80, 3/12/74)
STANDARDS
Product
Process
Para-
meters
BPCICA*
Max.1 Avg.2
kg/Wcg kg/kkg
(mg/1) (mg/1)
batea*
Max.1 Avg.2
kg/kkg kg/kkg
(ng/l) (ing/1)
NSPS*
Max. 1 Avg.2
kg/kkg kg/kkg
lrng/1) (mg/1)
Hydro-
fluoric
Acid
Fluoride (30)
TSS
(50)
(15)
(25)
No discharge
of pwwp3
No discharge
of pwwp
No discharge
of pwwp
No discharge
of p#tip
Sections 415.82, 415.83, and 415.85 were renanded and are presently
reserved (41 FR 51601, November 23, 1976).
Htex. ¦ Maximsa of any one day.
2
Avg. - Average of daily values for thirty consecutive days shall not exceed,
^pwwp - Process wastewater pollutants.
277
-------
H20 " 0.1%
CaC03 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:
S i02 + 2CaF2 + 2H2S04 = SiF4 + 2CaS04 + 2H20 (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 HF 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.
HF 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 HF may discharge from either end. The theoretical
amount of gypsum produced is 3.4 kg/kg of HF produced, but
because of the impurities in the fluorspar the actual amount
of gypsum produced is higher and varies from 3.6 to 4.8 kg/kg of
HF.
One manufacturer uses a patented process to supply
internal heat to the reactor. The heat is supplied by
introducing sulfur trioxide (S03) 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 S03.
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 HF 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 USB 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
-------
OL^H
CONDENSABLES
COLLECTION
ppy
SULTURIC
ACID
fXUORSPAR
SLURRY KATE*
CALCIUM
SULFATE (GYPSUM)
SOLIDS
KILN
t
TO WASTE
NGR-comjcr"
,
«3B!
COOLER
zz:
HATER, SULFURIC
—"IACID & FXUO-
SULTONIC ACID
RECYCLE
DRZP POT
COKE BOX
NONCONTACT
COOLING OR
REFRIGERATION
SYSTEM
WATER
CONDENSER
ACID
SCRUBBER.
CRUDE HYDROGEN
PLUQR2DE
HATER'
EJECTOR
W9I^I«n
TO
TMAIMBIT
if
D^R
T
"HATER
_L_
HATER
SCRUBBER
SULFURIC ACID
VfikSTE VKSX
STORAGE
~
-- DILUTION
1
DISTILLATION-C05®111'5*1^^ TO STORAGE (OR
| RECYCLED TO KILN)
AQOEOOS HYDROGEN
FLUORIDE PRODUCT
~
PACKAGING
1
ANHYDROUS
HYDROGEN PLUORIDE
PRODUCT |
PACKAGING
T
TO SALES
TO SALES
LEGEND
COMMON PRACTICE
INTERMITTENT
PROCESS (OR PRO-
CESS AT ONLY
SOME PLANTS)
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/W
-------
TABLE 12-4. WASTE WATER FLOW AND REUSE DATA FOR THE
HYDROFLUORIC ACID SUBCATEGORY
Plant
Kiln Residue
(1)
Handling
Reuse for
Kiln Residue
(2)
Slurry
(Percent)
Influent to
Treatment
Facility
(m3/k:
-------
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.
Fluorosulfonic 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 Waste Water
Scrubber water is another waste water source, and in
plants which 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 HF, silicon tetrafluoride (SiF4), and hexafluosilicic
acid (H2SiF6). Silica present in the ore as an impurity reacts
with HF forming silicon tetrafluoride as shown in Equation 3.
Si02 + 4HF = SiF4 + 2H20 (3)
In the scrubber, the tetrafluoride is converted to
hexafluosilicic acid according to the following equations:
283
-------
15,000"
12,500 ••
10,000 -•
<*¦>
g 7,500 --
Dry Kiln Waste
Slurrying Kiln Waste
2,500
2,000 -•
1,000 --
25
50
75
100
150
200
HF Production, kkg/day
Figure 12- 2. Production versus waste flow data for HF plants.
284
-------
SiF4 + 2HF = H2SiF6 (4A)
3SiF4 + 2H20 = 2H2SiF6 + Si02 (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 f705 was visited and process waste water samples
were collected and analyzed for conventional, nonconventional
and toxic 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
-------
TABLE 12-5. WASTE FLOW FROM HYDROFLUORIC ACID MANUFACTURING PLAOTS
3
Flow in m /kkg of Hydrofluoric Acid
Plants
Source of
Wastewater #25r ' #987VJJ #753 #426 #120 #722 #167 #705u' #837
Gypsun Slurry
Drip Acid
Scrubber
Waste Water
64.0 Dry
disposal
0.0490 0
14.4 8.30
NA Dry Dry (Total
disposal disposal Recycle)
0 0
2.30 NA
0
0.624
0
(Total
Recycle)
122 (Total 6.50
Recycle)
NA 0.0180 0
40.0 11.3 1.12
Other
0.530 0.530
8.40 NA
5.55
NA
5.20 22.5
NA
(1)
NA =
*
Discontinued HF production
Not Available
Other does not include wasteflews frcm storm water runoff.
-------
TABLE 12-6. SOLID WASTE GENERATED AT THE HYDROFLUORIC ACID PLANTS SAMPLED
Plant Gypsun Solids Going Tto Total Solids Produced
Treatment Facility (kg/kg of HF)
(kgAg of HF)
#705(1) 4.73 4.78
#251(1) 3.81 NA
#167 3.94 NA
(1) Discontinued HF production.
NA = Not Available
287
-------
TABI£ 12-7. GYPSUM SOLIDS PRODUCTION IN THE HYDFCFLL'ORIC ACID SUBCATEGORY
Kiln Residue Produced Kiln Residue
Plant (kgAg of HF) Disposal/Treatment Method
#837
3.86
S
#705 (1)
4.73
S
#167
3.94
S
#722
NA
S
#120
NA
D
#426
4.00
D
#987 (1)
4.13
D
#251 (1)
3.81
S
#753
NA
S
#967
NA
s
#928
NA
s
S * Slurried with water and sent to wastewater treatment facility.
D ¦ Dry disposal.
NA » Not Available.
(1) Discontinued HF production. 288
-------
AllCH)
Mr, nam
aaiuDC rns
CCCLOC TOB)
¦LOCXMN ,HC
IIOM)
W*ata itiejM M«vlod.
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 SAMPLED WASTE
STREAMS OF PLANT #705 PRODUCING HYDROFLUORIC ACID^1)
Stream
No.
Sampled
Stream
Description
(2)
Screening Data
Flow Fluoride
(mVkkg of HF) (kg/kkg of HF)
Total
Suspended
Solids
(kg/kkg of HF)
1
Kiln Slurry
26.6
15
4700
2
Scrubber Waste
Water
10.0
9.6
0.070
3
Surface Drains
Cooling Tcwer
Blowdown
20.0
6.9
3.9
4
Treated Effluent
23.3(3)
1.6
1.9
(1) This plant has discontinued the production of HF since the time of
sampling.
(2) One 72-hour conposite sanple of each waste water stream.
(3) The discharged effluent consists of the treated waste waters frcn
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 f167) 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
#251 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
#251.
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 Suanary 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
-------
TABLE 12-9. FLOW AND POLLUTANT CONCENTRATION DATA OF THE SAMPLED WASTE STREAMS
FOR PLANTS #705, #251, AND #167 PRODUCING HYDROFLUORIC ACID
Verification Data
Total
Plant Stream Sanpled Flow Fluoride Suspended
No. Stream (m /kkg of HF) (kg/kkg of HF) Solids
Description (kg/kkg of HF)
#705(2)
1
Kiln Slurry
26.6
3.8
4700
2
Scrubber Waste
Water
10.0
1.5
0.019
4
Surface Drains
Cooling Tower
Blowdown
20.0
3.4
4.0
5
Treated Effluent
23.3(3)
0.54
0.040
#251(2)
5
AHF Plant
Hosedcwn
1.20
1.9
0.26
6
SO2 Scrubber
Waste
14.4
0.31
0.10
2
Gypsum Pond Inlet
84.7
58
3800
3
Gypsum Pond
Outlet
84.7
27
0.80
#167
1
Kiln Slurry
122
4.9
170
2
Ejector & Absorber
Unit Wastes from
Kilns #1,#2, and
#4
25.0
14
0.36
3
Ejector & Absorber
Unit Wastes from
Kilns #5 and #6
14.6
20
0.41
4
Effluent from
First Lagoon
162
11
22
(1) Three 24-hour ocmposite saitples 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
-------
VENT
DUST
COLLECTION
h2so4
WET
SPAR"
SPAR
DRYING
HOSE DOWN
WATER
HF KILN
r*
AIR
KILN
RESIDUE
HANDLING
LOSSES
ai2o3'3h2o
COOLER
DRIP
ACID
WATER
SLURRY
TREATMENT
0
LEGEND
SAMPLING POINTS.
PLANT
REACTOR
AIPRODUCT
VENT
_L
S02 SCRUBBER
WATER
|— — w<
Uaif
SCRUBBER
LIQUEFACTION
=0
ON
AHF
PURIFICATION
AHF
PRODUCT
DILUTION WATER
#5
5
HOSE DOWN WATER
AHF PLANT
h
WATER
GYPSUM
POND
n
NEUTRALIZATION
SYSTEM
ALKALINE STREAMS AND
ACID FROM OTHER PLANTS
1—
J EFFL
EFFLUENT
Figure 12-4. General process flow diagram at Plant #251 showing the sampling points.
Hydrofluoric acid manufacture.
-------
Maximum Raw Waste Concentrations Observed
(pg/l)
Pollutant Screeninq Verification
Plant f705 Plants *705 , f25"i,
f 167
Copper
770
600
Lead
5200
200
Selenium
25
230
Zinc
8100
13000
Ant imony
70
2800
Arsenic
10
160
Cadmium
2.0
60
Chromium
73
1200
Mercury
2.0
43
Nickel
150
2000
Thallium
5.5
63
Section 5.1.2 of this report describes the methodology of
the screening and verification sampling proqram. In the
Hydrofluoric Acid industry, a total of 12 days of sampling
were conducted at Plants f705 , f251, and f'167. 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 <'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
concentrat ion.
That is,
Daily loading (as kg
per day)
Where:
of pollutant (C) (Q1
= 1000
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) = 1000fP)
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 eoual to
2205 lbs.)
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:
Pollu tant
Waste Load (kg/yearl
Copper
Lead
Selenium
Zinc
Antimony
Arsenic
Cadmium
Chromium
Mercu ry
N ickel
Thallium
6600
10000
260
110000
8900
1400
79
4700
1.30
10000
840
12.4 POLLUTION ABATEMENT OPTIONS
12.4.1 Toxic Pollutants of Concern
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
-------
T\3LE 12-10. TOXIC POLLUTANT RAW WASTE DATA
SUBCATEGORY: HYDROFLUORIC ACID
(1)
Average Daily Fbllutant Concentrations and Loadings at Plants Sampled
(mg/1)
(kg/kkg of Anhydrous HF)
#705(S)
1705(V)
#251 (V)
#167 (V)
Overall
Average
Antimony
0.018
0.010
0.12
0.74
0.22
0.0010
0.00057
0.010
0.12
0.033
Arsenic
0.051
*
0.11
0.028
0.062
0.0029
*
0.0091
0.0046
0.0055
Cadmium
0.0014
0.0060
*
0.0030
0.0035
0.000030
0.00034
*
0.00047
0.00030
Chromium
0.062
0.26
0.47
0.074
0.22
0.0035
0.015
0.040
0.012
0.018
Copper
0.41
0.26
0.12
0.32
0.28
0.023
0.015
0.010
0.051
0.025
Lead
2.47
0.044
0.059
0.062
0.66
0. 14
0.0025
0.0050
0.010
0.039
Mercury
0.00090
0.0053
0.018
0.0010
0.0060
0.000050
0.00030
0.0015
0.00016
0.00050
Nickel
0.062
0.48
1.18
0.15
0.47
0.0035
0.027
0.10
0.025
0.039
Selenium
0.0070
*
0.017
0.0074
0.011
0.00040
*
0.0014
0.0012
0.0010
Thallium
*
*
0.039
0.019
0.029
*
*
0.0033
0.0030
0.0032
Zinc
4.0
0.21
0.28
8.2
3.2
0.23
0.012
0.024
1.3
0.41
S - Screening data £rom 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.
(I) 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. SIMMARY OF RAW WASTE LOADINGS FOUND IN
SCREENING AND VERIFICATION SAMPLING
SUBCATEGORY: HYDROFLUORIC ACID
Pollutant
Daily
Loadings
(kg/day)
Minimun Average Maximum
Unit
Loadings No. 0£
(kg/kkg) Plants
Minimum Averse Maximum Averaged*
Toxic
Antimony
0.023
2.0
6.4
0.00057
0.034
0.12
4
Arsenic
0.012
0.50
1.2
0.00030
0.0055
0.0090
3
Cadmium
0.0031
0.014
0.025
0.000077
0.00030
0.00047
3
Chromium
0.15
1.7
5.4
0.0035
0.018
0.040
4
Copper
0.60
1.4
2.80
0.0096
0.025
0.051
4
Lead
0.10
1.8
5.4
0.0025
0.039
0.14
4
Mercury
0.0021
0.057
0.21
0.000050
0.00050
0.0015
4
Nickel
0.14
4.1
14
0.00035
0.039
0.10
4
Selenium
0.016
0.093
0.20
0.00040
0.0010
0.0014
3
Thallium
0.16
0.31
0.45
0.0030
0.0032
0.0033
2
Zinc
0.49
21
72
0.012
0.41
1.3
4
Conventional 5> Nonconventional
TSS
190000
310000 1
520000
3800
4200
4800
3
Fluoride
13
2900
7900
8.8
34
53
4
* Only those plants vrtiere 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 HF 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 Modifications 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
f luorosulf onic 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 drip acid compared to those which
do not, 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 HF.
HS03F + H20 + 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 #837 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 #987, 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 #722 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 #426 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 slurried 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 HF 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 is 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
-------
LIME
S '
RAW
»do~
WASTE WATER
A j /-*.
EQUAIJ7.AT10N
RFC. YCLE FOR Sl.URRY TRAN.SI-OKT
„ jj l
MIXING
LAGOON
LAGOON
pll ADJUSTMENT
<$>
EFFLUENT
'include • flow monitoring, pH monitoring «n«» * ampler
Figure 12-5. Level 1 waste water treatment for hydrofluoric acid subcategory.
-------
u>
o
OJ
0^1
1
^—==dJi
"l
tccvcuc roi
SUJB1Y TAAKSHOBT4
WAi-TC WATCH
KOUAUZATtOM
u
n
U>TMIXT
¥*l
U riiJKMT
I4GOQ)
ImImA«i Uu« wKolUrtog,
pti wmlWttn »«4 itniflti
Figure 12-6. Level 2 waste water treatment for hydrofluoric acid subcategory.
-------
usu/m
I—-ill—'
---®
-cfc--
1
PWJUWOI
F^ljii]
0
X
IikU4(» 0*w mvfilu»rU|, moalte(t#| Mn^ltr
Figure 12-7. level 3 waste water treatment for hydrofluoric acid subcategory.
-------
8
LH
r
0
J5&-
i
^lill
kxu m»
./J
A
?
<=>
K
Ug-
lMl«ikl flow frunHnfUf. pM nw>lUiiU| ltd ntn|4ii
Figure 12-8. I,evel 4 waste water treatment for hydrofluoric 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 1, 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
o
-------
Q"
U>
o
-J
1 bACKVASlI
CUMIIKR
I
tOUAU?-AT).)f
u/»*
KJL;
~1_W
ADJUSTMCMT
1
1 (^)
1IL"l J
¦ tCTCl.t TO
sciuueti
CI-
Cfl IUfcNT
^ t>UX)M
«on y
TO l>f t> ll L
-------
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
Levels 1, 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 TREATMENT 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.
Waste Hater 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 HF to 166 m3/kkg of
HF. For the model plants, a constant unit flow of 95.4 m3/kkg
of HF was assumed.
308
-------
HF Production
•
In the HF 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
HF 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. .10CEL PLANT TREATMENT COSTS
Subcategory HYDROFLUORIC ACID
(1) (1)
Production 19,100 metric tons per year (21,057 tons per year)
54 metric tons per day (60 tons per day)
Waste water flow 5220 cubic meters per day.
LEVEL OF TREATMQTr*
FIRST
SECOND
THIRD
FOURTH
INVESTMENT COST
Construction
S877.500
$24,500
$25,000
$24,500
Equipment in place,
including piping.
fittings, electrical
vrerk and controls
356,000
89,500
92,000
89,500
Reuse facilities
30,000
Monitoring equipment..
9,000
Engineering design
and inspection
254,500
22,800
23,400
22,800
Incidentals, overhead,
fees, contingencies...
254,500
22,800
23,400
22,800
Land
1,020,000
TOTAL rNVESTMQJT COST
S2,801,500
$159,600
S163.800
$159 , 600
OPERATION AND
MAINTENANCE COST
Labor and supervision.
$56,000
$14,000
$14,000
$14,000
Energy
14,000
1,500
1,800
1,500
Chemicals
534,800
3,400
367,700
Maintenance
172,650
15,960
16,380
15,960
Reus« 0 & M
6,500
Taxes and insurance...
84,045
4,788
4,914
4,788
Residual waste disposal
350,000
Monitoring, analysis
and reporting
15,000
7,500
7,500
7,500
TOTAL OPERATION AND
c
MAINTENANCE COST
1,232,995
$43,748
$47,994
$411,448
AMORTIZATION OF
INVESTMENT COST
$289,850
$25,966
$26,650
$25,966
TOTAL AMJUAL COST
$1,522,845"
$69,714
$74,644
$437,414
•First level represents the base cost of treatnent system.
Other levels represent the incremental cost above base cost.
••Including 511,100 for the reuse of treated effluent to slurry
Kiln residues, etc.
(1) Production year is 350 days.
310
-------
TABLE 12-13. MODEL PLW/T TREATHQJT COSTS
Subcategory HYDROFLUORIC ACID
(1)
Production 38,200 metric tons per year
109 metric tons per day
Waste water flow 10450 cubic meters per day.
(1)
(42,115 tons per year)
(120 tons per day)
A. INVESTMENT COST
Construction
Equipnent in place,
including piping,
fittings, electrical
work and controls
Reuse facilities
Monitoring equipment..
Engineering design
and Inspection
Incidentals, overhead,
fees, contingencies...
Land
B.
TOTAL INVESTMENT COST
OPEJUkTICN AND
MAINTENANCE COST
Labor and supervision.
Energy
Chemicals
Maintenance
Reuse 0 & M
Taxes and Insurance...
Residual waste disposal.
Monitoring, analysis
and reporting
C.
TOTAL OPERATION AND
MAINTENANCE COST
AMORTIZATION OF
INVESTMENT COST
TOTAL AMflJAL COST
FIRST
51,354,500
493,500
40,000
9,000
379,400
379,400
1,944,000
$56,000
19,500
1,069,600
257,580
10,000
137,994
700,000
15,000
2,265,674
LEVEL OF TREATMENT*
SBCCND THIRD
535,000
131,000
$35 , 500
137,500
33,200 34,600
33,200 34,600
514,000
3,100
23,240
6,972
7,500
$54,812
$432,098 $37,811
$14,000
3,400
6,700
24,220
7,266
7,500
$63,086
$39,405
52,697,772" $92,623 $102,491
FOURTH
$35,000
131,000
33,200
33,200
S4,599,800 $232,400 $242,200 $232,400
$14,000
3,100
735,350
23,240
6,972
7,500
$790,162
$37,811
$827,973
'First level represents the base cost of treaenent 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 35Z '!ays.
311
-------
TABLE 12-14.
MODEL. PLA.VT TREAT* D/T COSTS
Subcategory HYDROFLUORIC ACID
Production
Waste water flow
(1)
57,300 metric tons per year
163 metric tons per day
15700 cubic x.eters per day.
(1)
(63,173 tons per year)
(ICO tons per day)
A. INVESTMENT COST
Construction
Equlpnent in place,
including piping,
fittings, electrical
work and controls
Reuse facilities
Monitoring equipment..
Engineering design
and inspection
Incidentals, overhead,
fees, contirtgencies...
Land
TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Energy
Chenicals
Maintenance
Reuse 0 & M
Taxes and insurance...
Residual waste disposal.
Monitoring, analysis
and reporting
TOTAL OPERATION AM)
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
LEVEL OF TREATMENT*
FIRST SECOND THIRD
51,755,500 549,000
848,000
50,000
9,000
532,500
532,500
2,880,000
556,000
28.000
1,604,400
362,350
13,000
198,225
1,050,000
15,000
3,326,975
203,500
50,500
50,500
550,000
215,500
53,100
53,100
FOURTH
549,000
203,500
50,S00
50,500
56,607,500 S353.500 5371,700 5353,500
514,000
4,600
35,350
10,605
7,500
S72.055
5606,464 557,514
514,000
4,900
10,070
37,170
11,151
7,500
584,791
560,475
514,000
4,600
1,103,025
35,350
10,60S
7,500
51,175,080
557,514
53,933,439** 5129,569 5145,266 51,232,594
•First level represents the base cost of treatment systeri.
Other levels represent the Incremental cost above base cost.
••Includes 521,200 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, chemical cost has a significant impact on the additional
annual costs.
313
-------
o
o
o
o
o
o
X
>
i ¦ ¦
| j
i
: ¦ ; 1 ¦ • .—
1 ; . .i ;
—i 1—. r—
¦ :
i ,
.
¦
' ! ;
.
1
1 '
i
i
¦
1 1
i ¦
¦Mi:
1
. ¦ ¦ 1 i : i . ¦ . i
—J . p-p-,
1 !:——n
1 ' i ¦ 1 ¦ 1 '
¦ 1 i;
! i
1 i ¦ i i
¦ 1 1 i ,
i
i
1
¦ ¦ r ¦ 1
i
.... |
1 i
¦ i
i -
1 ¦
i
—h!
, 1
1—!—
—
i
1
. 1 . .
H—i i—-
i . i
i
.
1
-
1
^—:—t-—-
1 TT
' iLEVvEL 44 ¦ X
>_—
1
1 .
i i
1 !
~ i
1 ¦ ¦
1 1
—7^
l.FVF.M IP k iS
r-
. . . j
:
¦ r~
7TT
L. *1
—r— : ' ¦ 1
; . — . ; 1.
i ¦ .
¦ t
i 1
! ! ¦
1
/ :
. v>r . i ii
.1 i !
// i ,i
; I
^ rt—
1 Y
: : i
1 1
i
.
i
i yS® I i
; i ¦ ¦ 1
1 1 j
1
| '
1 •
r ! ^
yy
1
I 1
—r
¦4*—-
i
1
—1—'—i-i-
1
| ¦ |
— "P
-1«
1
/ —
—¦—; :
1 ¦
' 1
1' , 1
¦ ¦ ¦ ' \
1
i
——;—.—,—i—I—
, i : , '
: ! 1 • ; I
I
i !
: i ¦
1
' i ! i ' !
1
i—1—i—,—i—
i- . .
1
'
. ; . i „ t
i
1 i
i
1
: 1
hn—r~t~
i
i
: i ¦
PRODUCTION (METRIC TONS/YEAR X 1000)
Figure 12-10. Annual treatment cost vs. production for the Hydrofluoric
Acid Subcategory
314
-------
11
l
i
I
!
i ¦ ,
1
•
,
i
i
1 '
i |
.
1
1 ' i
¦
¦
i I ¦
* . '
1
"1
;
1 1
|
I
1
1
i
¦
1
!
1
(
1
i 1
1
i
'
| j
i
|
¦ 1 1
1 1
1
i
¦
,
, 1
i
1
,
,
1 !
V ! ;
1
:
I
:\j ,
i
1 l>-s.
i
!L£VXI
r *
4
i
i
I
/TN
1
1 :
:
i
1
1
1
¦ : 1
i ;
. i . i i i
fi
i , ! ¦
1
' 1 : !
"X ¦
¦ i ' i
\ : '
i
'
,
1 ' '
1
J
i
1 ¦
LE
VELS *2 &
»3i
i
J2\
i
1
,
, | , »
! 1 1
: i
1 ! 1
\ a
i : : 1
i ! : ' 1
II. ! !
| : ; i
1 =
1 . : ! 1 ; i 1
i
i i i
10 20 30 40 50 60
PRODUCTION (METRIC TCNS/YEAR X 1000)
Figure 12-11. Annual unit treatment cost vs. production for the
Hydrofluoric Acid Subcategory
315
-------
TABLE 12-15. MODEL PLANT TREATMENT COSTS
Subcategory HYDROFLUORIC ACID
Annual Treatment Costs ($Akg) of HF
LEVEL OF TREATMENT
COST ITEMS PRODUCTION FLOW FIRST SECOND THIRD FOURTH
(kkg/yr) (m3/day)
Annual Operation
and Maintenance
19,100
5,220
64.55
2.29
2.51
21.54
38,200
10,450
59.31
1.43
1.65
20.68
57,300
15,700
58.06
1.26
1.48
20.51
Annual
Amortization
19,100
5,220
15.18
1.36
1.40
1.36
38,200
10,450
11.31
0.99
1.03
0.99
57,300
15,700
10.58
1.00
1.06
1.00
Total Cost
19,100
5,220
79.73
3.65
3.91
22.90
38,200
10,450
70.62
2.42
2.68
21.67
57,300
15,700
68.65
2.26
2.54
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
(1) (1)
Production 19,100 metric tons per year (21,057 tons per year)
54 metric tons per day (60 tons per day)
Waste water flow 680 cubic meters per day.
LEVEL OF TREATMENT*
FIRST
A. INVESTMENT COST
Construction $64,000
Equipment in place,
including piping,
fittings, electrical
work and controls 327,000
Monitoring equipment
in place 9,000
Engineering design
and inspection 80,000
Incidentals, overhead,
fees, contingencies... 80,000
Land 30,000
TOTAL INVESTMENT COST $590,000
B. OPERATION AND
MAINTENANCE COST
Labor and supervision. $56,000
Energy 6,100
Chemicals 44,000
Maintenance 56,000
Taxes and insurance... 17,700
Residual waste
disposal 742,000
Monitoring, analysis
and reporting 15,000
TOTAL OPERATION AND
MAINTENANCE COST $936,800
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
*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
$91,112
$1,027,912
-------
TABLE 12-17. 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 1370 cubic meters per day.
LEVEL OF TREATMENT*
FIRST
A. 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 AhWUAL 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)
Production 57,300 metric tons per year
163 metric tons per day
Waste water flow 2030 cubic meters per day.
(1)
(63,173 tons per year)
(180 tons per day)
LEVEL OF TREATMENT*
FIRST
A. 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 $2,743,195
~First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
(1) Production year is 350 days.
320
-------
!
i
j
i
¦
1
!
:
¦ i
I
1
i
¦ i i
!
i :
,
! 1
. .
1
|
1 ,
1 , ¦
1
1
1
1 i
1
: 1
I !
1 1
1
1 1
i
; I !
:
! ; ! i .
! :
. .
;
i
i , r
¦
1
! i !
1
i
I !
i
i ¦ i
. i
1
i
' i
¦ , 1
i i
1 : ,
i .
1 1
i !
i
1
1 i ! 1
1
i
! 1
i
¦ i 1
1 &
1
i
: ¦
: !
1 1
1 1
i
i
; i
¦
¦
•
1
i ¦
.
i ¦ ;
1
1
i
i
i ;
1
1
1 ! !
/.
1
1
!
I
1
1
1
f ' Ilev
EL 4l
1
f '
, i
¦
i
i
1
! 1 X
i
!
1 '
!
i
i >*
! jfifi
I
1 1
1
! i
¦
1
!
I
i i
i /¦
i
i
i
/ i
¦
1
1
i
i
r 1
I 1
¦
i /
1 1
1
1 1
• 1
I . .
' ' |
.
/M
1
i
' i
' i
• 1
i i
V
1
, i 1
-
,
i
1 l
: i
i
! 1
1 i
i
i
i
: ! i ! i
I 1 1
1
:
i ¦ il
i : 1
i
|
1
¦ i
1 : 1
1 1 I
i >
; ! i
1 !
i
i |
; i
i ! i
i i i
1
' |
¦ i : : 1 1 1 ' 1 ¦
1
1
1
10 20 30 40 SO
PBCOJCTICN (>ETRIC TCNS/VEAR X 10 00)
60
Figure 12-12. Annual treatment cx>st vs. production for
Hydrofluoric Acid Subcategory (NSPS)
the
321
-------
lieteL" iH
PRCDUCTICN CMETRIC TCNS/YEAR X 1000)
Figure 12-13. Annual unit treatment cost vs. production for
the Hydrofluoric Acid Subcategory (NSPS)
322
-------
TABLE 12-19. MODEL PLANT TREATMENT COSTS
Subcategory HYDROFLUORIC ACID
Annual Treatment Costs ($Akg) of HF
LEVEL OF TREATMENT*
COST ITEM PRODUCTION FLOW FIRST SECCND THIRD FOURTH
(kkg/yr) (m3/day)
Annual Operation
and Maintenance
19,100
680
49.05
38,200
1,370
45.90
57,300
2,030
44.97
Annual
Amortization
19,100
680
4.77
38,200
1,370
3.41
57,300
2,030
2.90
Total Cost
19,100
680
53.82
38,200
1,370
49.31
57,300
2,030
47.87
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
-------
TABU 12-20. SOWWRY Of WASTE WATIR CONTROL MO TMAWBIT TECHNOLOGY tXPLOYES AT mTOROPWOKIC ACID PLANTs'*'
Plant
Product-Ralatad
*aata Natar Sourcaa
Control and Traatiiant
Tachnology laployad
Aaount of
Traatad
Maata Ma tar
Rauaad
Coolar Bottoaa
(Condanaablaa)
Racyclad?
Efiluant Voluoa
In a'/natrlc ton
(gal/ahort ton) of
Actual Production
Long Tara
Avaraga Pollutant
waatalsad Slachargad
!kg/eatrlc ton)
;lb/1000 lb)
fluorlda TSS
I 42( Bydrofluorlc acid
fluoalliclc aclda
production
t $64
Bydrofluorlc acid
production
» 16? Bydrofluoclc acid,
fluorocarbon,
Chlorlna/aodlua
hydroxlda, and
hydrochloric acid
production
Dry raaldua hauling 0 Taa
and duaplngi rvautra-
il ration with cauatlc
of rioncontact cooling
watar and floor
dralnaga
Raaldua alurry, nautra- 94% Yaa
lliatlon with aodlua
carbonat«, aattllng,
racycla
Raaldua alurry, llaa 471 Taa
traataant,aattllng,
racycla
465 (111,397)
lncludaa noncon-
tact cooling
watar
5.71 (1,3(0)
103 (24,200)
1.2 ND
0.10 0.27
1( 0.45 (Nat)'
I 120
Bydrofluorlc acid
productlon
4 W Bydrofluorlc acid,
fluorocarbon, and
aulfurlc acid
production
Plannad dry raaldua 0 Taa
handling, llaa
traataant, clarification
Raaldua alurry, aattllng Praaanti 0 Yaa
(Racycla and p4! Planned ¦ 70*
pollahlng facllltlaa
undar conatructlon.)
to 751
ND
ND
125 (30,000) Praaanti 24 1<
Bxpactad
with 1.6 2.1
additional
facllltlaa
8 928 Bydrofluorlc acid Raaldua alurry, aattllng,
and alualnua racycla (Plocculatlon,
fluorlda production llaa traataant, and
clarification facllltlaa
undar oonatruction.)
83%
Yaa 9.44 (2,260) Praaant: 1 1.7
Bxpactad
with 0.65 0.75
additional
facllltlaa
• $37 All hydrofluoric
acid ganaratad aa
uaad captl*aly for
alualma fluorlda
production
Raaldua alurry, llaa
traataant, aattllng
Yaa
134 (32,200)
1.8 3.1
• 753
Bydrofluorlc acid
production
8 251
8 705
Raaldua alurry, llaa
traataant, aattllng,
racycla, pa pollahlng
65%
(2)
(2)
IP, ALPi, cMorlna/ Raaldua alurry, aattllng,
aodlua hydroxlda, nautrallxatlon
alualnua oxlda, and
fluorocarbon
production
Bydrofluorlc acid Raaldua alurry, llaa
and alualma traataant, aattllng,
fluorlda production racycla, jM pollahlng
Yaa
1 Kiln. Yaa
3 KUnai Ho
30% to 35%
11.0 (2,650)
22.2 x 10'
(553 x 10*)
25.t (6.204)
0.64 0.38
46 530
3.2 0.64
8 722
Bydrofluorlc tnd.
In
fluoborlc,
•eld production
• ffydrofloorlc Acid Dry rssldus hauling
Rssldus slurry, 11m
rscycls, pS pollshin?
92% to 100%
Yss
0-10.3 (0-2,440) 0-0.61 0 to 0.54
8.6
(1) Adapted frod CAlspsn (ftsfsrsnc* 3).
(2) Bydrofluorlc Acid production has discontinued st U\sss plants sines tfcs tlas of sssplln?.
(3) Efflusot lo*dln9 lsss ths lnflusnt losdlnq.
*D • Not dttsr*lnsd. ^_
-------
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 of
326
-------
TA3LE 12-21. SUMMARY OF LONG TERM MONITORING DATA FROM FOUR
(1)
HYDROFLUORIC ACID PLANTS
Treated Waste Load (kg/kkg) or (lb/lOOOlb)
Plant
No. Parameter
Daily Data
Long Term
Avejrage St. Dev.
(X) (S) (S')
30-Day Average Data
(2) Long Tterm (2)
VF Average St. Dev. VF
(XV (S)
#664
Fluoride
TSS
0. 10
0.29
0.090
0.77
4.5
0.10
0.27
0.040
1.7
*753
Fluoride
TSS
0.72
0.38
0.27
0.36
2.2
0.64
0.15
1.4
#722
Fluoride
TSS
0.81
0. 54
0.52
0.37
0.59
0.62
3.3
3.5
—
—
—
#705
(3)
Fluoride
TSS
—
—
—
—
0.49
0.84
0.22
0.37
1.7
1.7
(1)
Based on Reference 3 data.
(2)
In the case of daily measurements, the variability factor, VF,
for a lognormal distribution is found by the expression ln(VF) =
S'(Z - 0.5S'), where S' is the estimated standard deviation of
the logarithm derived from the arithmetic mean, X, and the
arithmetic standard deviation, S, according to the relationship,
(S') 2 = in
•(«"]
1.0 +( S V .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 foind by the expression,
VF = 1.0 + Z ^ S^. When the value of Z is 1.64, the
variability factor is for the 95 percentile. Please refer to
Section 3.2 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 VF.
— Not Available.
327
-------
TABLE 12-22. TOXIC POLLUTANT TREATTO EFFLUENT OATA
SU3CATBGCRY: HYDROFLUORIC ACID
(U
Average tolly ft>llutant Concentrations and Loadings at Plant3 Sampled
(*g/l)
f/erall Average
*705 (S) <705 (VI 1251 (VI U67(V) Average % Removal
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
lercury
Nickel
Selenium
Thai I tun
Zinc
<0.010
<0.00021
<0.0020 <0.17
<0.000042 <0.017
<0.0030 <0.010 <0.020
<0.000063 <0.00021 <0.0020
0.00030 <0.0017 <0.0020
0.0000060 <0.000035 <0.00020
0.014
0.00029
0.10
0.0021
0.0060
0.00012
<0.046
<0.00096
<0.020
<0.00042
<0.022
<0.00046
0.22
0.022
0.070
0.0069
<0.031
<0.0031
<0.00040 <0.00050 <0.0010
<0.0000080 <0.000010 <0.00010
0.050
0.0010
0.033
0.00069
<0.010
<0.00021
<0.0050
<0.00010
0.52
0.052
<0.071
<0.0070
0.0070 <0.0012 <0.0070
0.00015 <0.000025 <0.00069
0.071
0.0015
0.053
0.0011
0.16
0.015
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
<0.033
<0.0091
<0.063
<0.0061
<0.017
<0.0015
<0.0020
<0.00044
<0.17
<0.019
<0.030
<0.0025
74
81
9
9
62
77
97
67
64
Effluent
^Influent
<0.0045 8S
<0.00039
0.55
0.13
83
(S) Screening data from one 72-tx>ur composite sample of treated
effluent.
(V) Verification data from three 24-hour composite samples.
(1) Tie effluent data presented here corresponds to the raw waste
data shovn In Tfcble 12-10. The -nethodology of the sampling
program is described In Section 5.1.2, and the scope of
sampling In the Hydrofluoric Acid Industry is described in
Section 12.3.3.
(21 When averaging values indicated as "less than" (
-------
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
stud ies.
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.
Plow 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
-------
2.0-
1.0-
Percent Reuse
100-
Percent Reuse
I£GPP
• Long-term data
O Expected with treatment system upgrading
~ Screening and verification sanpling results
Figure 12-14 < Fluoride 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 12-12, 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:
M26 and 1120 because kiln residues are handled as a dry
sol id,
$167 , *967 , and f251 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,
332
-------
{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
#664 and #753 and those used for TSS are derived from Plant f722
for daily measurements and Plant f705 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) /1000 mq/l\ = 97 mg/1
(54.6 m3/kkg) V~kg7m3 )
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
Reuse Fluoride TSS
Plant (percent) (kg/kkg of HF) (kg/kkg of HF)
#837
0
1.8
3.1
#753
65
0.72
0.38
#928
83
1.0
1.7
#722
92
0.81
0.54
#664
94
0.10
0.29
Average of four plants
0.66
0.73
practicing effluent reuse
(excluding #837)
Variability Factor for 3.4^ 3.5(5)
Daily Measurements
Variability Factor for 1.6^ 1.7^6^
30-Day Averages
Variability Factor Ratio (VFR) 3.4/1.6 = 2.1(2) 3.5/1.7 = 2.1(2)
Effluent Limitations for BPT
(from Plant #837)
a. Daily Max 3.4 X 1.8 kg/kkg = 6.1 J J 3.5 X 3.1 kg/kkg = 11 ...
b. Max 30-Day Avg 1.6 X 1.8 kg/kkg = 2.9( ' 1.7 X 3.1 kg/kkg = 5.3( '
Effluent Limitations for BAT
(from average of four plants)
a. Daily Max 3.4 X 0.64 kgAkg
b. Max 30-Day Avg 1.6 X 0.64 kg/kkg
NA - Not Applicable
(1) Variability factor average of Plants #664,^722 .one *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 kg/kkg 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.
= 2.2
= 1.0
(3)
(4)
NA
NA
334
-------
TABLE 12-24. PROPOSED UKTDVTICMS
Hydrofluoric Acid
Best Practicable Control Technology Currently Available
waste Water Flow: 54.6 m3Akg of HF (43% Reuse) *
Concentration Basis
Effluent Limit
Subcategory
(1) (mo/1)
(Rg/kka) of HF
Pollutant
Performance
VFR vjax
Max
(mg/1)
30-
0.030
11
6.1
0.088
(5)
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 tern average based on loading data and
variability factors selected from Table 12-21.
(3) - The lower lLnit 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 aarrpluig data
is below this level.
(4) - Average effluent concentration from screening and verification
sampling data.
(5) - No effluent limitation proposed.
• Fror. 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 mq/l\ = 53 mq/1
(54.6 m3/kkg) ^ kg/m3 )
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) /lOOO mq/l\ = 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 f251.
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/l)(54.6 m3/kkg) / kg/m3 \ = 0.030 kq/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 daily measurements = 3.12
VF of 30-day averages 1.55
= 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/1.' 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 mq/1) (54.6 m3/kkg) f kq/m3 \ = 0.0093 kg/kkg,
\1000 mg/V
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 fcr incidental removal and a
higher removal efficiency than can be justified by the use of
this technology. A VFR of 2.0 was used for lead on the basis
of long term data from Plant f 2 51. The proposed maximum 30-
day average limitation is,
(0.30 mq/l)(54.6 m3/kkg) / kg/m3 \ = 0.016 ka/kkq,
\1000 mg/1/
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 cf variation
may be observed when more operating data are collected. The
proposed maximum 30-day average limitation is,
(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/kkq) = 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 VFR 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 mq/1) (54.6 m3/kkq) f kq/m3 "\ = 0.027 kq/kkq,
\1000 mg/1/
and the daily maximum is,
(2.0) (0.027 kcj/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
V1000 mg/1/
and, 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 20 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\
\54.6 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 Liaitations
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 ^llow 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 fl67,
£753, £928 , f664 , and f722) which presently practice reuse as
is shown in Table 12-4.
Plow 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 {753 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) /lOOO mg/l\ = 33 mg/1
(33.4 m3/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 kq/kkg) /1000 mg/l\ = 66 mg/1
(33.4 m3/kkg) Vkg/m3 )
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 LIMITATIONS
Hydrofluoric Acid
Best Available Technology
Waste Water Flow: 33.4 m3Akg of HF (65% Reuse)*
B»»a«aa«g«33aJiaa4Jin ¦ a j i ¦3ixAiAiian=»sa3i3Siaxjga7TTsas3=i3
Concentration Basis Effluent Limit
(1) (mg/1) (kg/kkc of HF)
Pollutant Treatability VFR
(mg/1) 30-day 24-hr 30-day 24-hr
Avg Max Avg Max
Nonconventlonal Pollutants;
(2) (3)
Fluoride, F 33
Toxic
Pollutants:
12)
Antimony 0.70
Arsenic 0.50
Chrcmiun^ 0.040
Copper*2* 0.29
Lead (2) 0.060
Nickel (2) 0.15
Seleniun 0.18
Zinc (2) 0.52
(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 12-21.
(4) - No effluent limitation proposed.
* The effluent flew rate is 35 percent of the average influent
shewn in Table 12-4 (i.e., 0.35 X 95.4 m3/kkg = 33.4 m3/W
-------
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 pollu tants - 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 maximum 30-day average effluent limitation. Application
344
-------
TABLE 12-26. PERFORMANCE OF ALTERNATIVE TECHNOLOGY
Hydrofluoric Acid
Level of Treatment: 3
Waste Water Flew: 33.4 m3/kkg of HF (65% Reuse)
Pollutant
Treatability
VFR(1)
Achievable Concentration
(mg/1)
(mg/1)
Max
30-day
Ave
24-hr
Max
Nonoonventional Pollutants:
Fluoride, F
33
2.1
33
66
Toxic
Pollutants:
Antimony
0.70
2.0
0.70
1.4
Arsenic
0.50
2.0
0.50
0.10
Chrcrdum
0.010
2.0
C.C10
0.020
Copper
0.050
2.0
C. 050
o.io
Lead
0.060
2.0
0.060
0.12
Nickel
0.10
2.0
0.10
0.20
Selenium
0.18
2.0
0.18
0.36
Zinc
0.20
2.0
C-.20
0.20
(1) - VFR: ratio of the 24-hour variability factor to the 30-day
variability factor.
345
-------
TABLE 12-27. PERFORMANCE OF ALTERNATIVE TECHNOLOGY
Hydrofluoric Acid
Level of Treatment: 4
Waste Water Flow: 9.5 m3/kkg of HF (90% Reuse)
VFR(1)
Achievable Concentration
Pollutant
Treatability
(mg/1)
(mg/1)
Max
30-day
Avg
24-hr
Max
Nonoonventional Pollutants:
Fluoride, F
33
2.1
33
66
Toxic
Pollutants:
Antimony
0.70
2.0
0.70
1.4
Arsenic
0.50
2.0
0.50
1.0
Chromium
0.040
2.0
0.040
0.080
Copper
0.29
2.0
0.29
0.58
Lead
0.060
2.0
0.060
0.12
Nickel
0.15
2.0
0.15
0.30
Selenium
0.18
2.0
0.18
0.36
Zinc
0.52
2.0
0.52
1.0
(1) - VFR: 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) ( kq/m3 \ = 0.017 kg/kkg
V1000 mg/V
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 #251 (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. On 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) f kg/m3 A = 0.0020 kg/kkg
V1000 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 VFR 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,
(0.70 mg/1)(33.4 m3/kkg) / kg/m3 \ = 0.023 kg/kkg
V1000 mg/1/
and the daily maximum is,
(2.0) (0.023 kg/kkg) = 0.046 kg/kkg.
The VFR 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 mq/1) (33.4 m3/kkg) ( kg/m3 \ = 0.0097 kg/kkg
\1000 mg/iy
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,
(0.040 mg/1) (33.4 m3/kkg) ( kg/m3 \ = 0.0013 kg/kkg
\1000 mg/ly
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
pH, TSS, 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 #664 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 t'426 , #722, and
f837 were excluded because of incomplete data for scrubber
effluent and Plant fl67 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 water- 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 POLLUTAOT RAW WASTE DATA USED TO
REPRESENT MEW SOURCES*
SUBCATEGORY: HYDROFLUORIC ACID
Concentration
(U (2)
Pollutant Nlaximum Average
(mg/1) (mg/1)
Antimony
0.030
0.014
Arsenic
0.014
0.0090
Cadmium
0.021
0.0080
Chromium
0.41
0.11
Copper
0.12
0.049
Lead
0.029
0.011
^rcury
0.0020
0.0010
Nickel
0.81
CO
•
o
Selenium
0.24
0.068
Thallium
0.0040
0.0020
Zinc
0.45
0.15
* Based on four sets of screening and verification sampling
data from Plants #705, #251, and #167 taking only the
scrubber and "other" vraste sources.
(1) ^laximum 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 JRB 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 VFR 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
\ 1000 mg/1/
and, using the VFR value of 2.1,
(2.1)(0.41 kg/kkg) = 0.86 kg/kkg is the proposed daily
max imum.
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) f kg/m3 \ = 0.18 kg/kkg
\l000 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 tn3Akg (60% Reuse)
Concentration Basis Effluent Limit
(1) (mg/l) (kg/kkg of HF)
Pollutant Treatability WR
(mg/l)
Conventional and
Nonconventional Pollutants:
Total Suspended
Solids, TSS
Fluoride, F
Toxic
Pollutants:
Antimony
\rsenic
(2)
Chromium
Copper
Lead
(2)
Nickel
Selenium
(2)
Zinc
(2)
68
30
0.70
0.5
0.040
0.29
0.060
0.15
0.18
0.52
2.1
2.0
2.0
2.0
2.0
2.0
2.0
2.0
30-day 24-hr 30-day 24-hr
Aug Max Avg Max
2.1 68 143 0.41
63 0.18
0.86
0.38
0.70 1.4 —
2.0 0.5 1.0 —
(3)
(3)
(3)
(3)
0.040 0.080 0.00024 0.00048
(3) (3)
0.29 0.58
0.060 0.23
(3)
(3)
0.15 0.30 0.00090 0.0018
(3) (3)
0.18 0.36 — —
0.52 1.0 0.0031 0.0062
(1) - WR: ratio of the 24 hour variability factor to the
30 day variability factor.
(2) - Also applicable for PSNS limitations.
(3) - No effluent limitations proposed.
353
-------
Toxic pollutants -
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 N = 0.00090 kg/kkg
V1000 mg/1/
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 N = 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 N = 0.0031 kg/kkg
V1000 mg/V
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
-------
Intentionally Blank Page
-------
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 OP 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:
13.3 STATUS OF REGULATIONS
Since no toxic pollutants were found in significant
concentrations, the subcategory is excluded under Paragraph 8.
Pollu tant
Maximum Concentration
Observed (yg/1)
Z i nc
Pentachlorophenol
Bis (2-ethylhexyl)phthaiate
Chloroform
Naphthalene
256
4850
20
11
11
358
-------
TABLE 13- 1
SUBCATEGORY PROFILE DATA SUMMARY
SUBCATEGORY
HYDROGEN PEROXIDE
Tbtal subcategory capacity rate
*K>tal 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:
Minirom
Maxiirum
Average production
Median production
Average capacity utilization
Plant age range:
Mininum
Maximum
Waste water flew range:
Mininum
Maxiirum
Volume per unit product:
Mininum
Maximm
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.
NA = Not Available.
-------
Intentionally Blank Page
-------
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 (FeTi03), 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
-------
TABU2 14- 1
SUBCATEGORY PROFILE DATA SIMMAF&T
SUBCATEGORY
TITANIUM DIOXIDE (CHLORIDE PROCESS)
Ibtal subcategory capacity rate
Ibtal 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:
Mininum
Maxiirum
Volune per unit product:
Minimum
Maxiirum
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 Carrierce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessnent 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
-------
TABID 14-2 - STATUS OF RBGUIATICNS - EFFLUENT UMITATICN GUIDELINES
SUBCATEGORY
SUBPART
Titanium Dioxide
V (40 CFR 415.220, 3/12/74)
STANDARDS
BPCTCA*
BAIEA*
NSPS*
Product
Process
Max. Avg.
Para- kg/kkg kg/kkg
meters (mg/1) (mg/1)
Max. Avg.
kg/kkg kg/kkg
(mg/1) (mg/1)
Max. Avg.
kg /kkg kg /kkg
(mg/1) (mg/1)
Chloride
Process
Sulfate
Process
TSS
Iron
TSS
Iron
4.6
0.72
21.0
**
2.3
0.36
10.5
(100.0) (50.0)
1.7 0.84
(8.1) (4.0)
2.6 1.3
0.36 0.18
10.6 5.3
0.84 0.42
2.6
0.36
1.3
0.18
10.6 5.3
0.84 0.42
Sections 415.220, 415.222, 415.223, and 415.225 were retarded and are
presently reserved (41 FR 51601, November 23, 19761.
Max. - Maxinum of any one day.
2
m Maximun averaqe of dailv values for thirt".v d^vs.
**flow 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 + 2Ti02 + 4C12 = 2TiC14 + C02 + 2C0 (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 TiCl4 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 = T i02 + 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 HASTE 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 fron 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
-------
cm
•CTttf 9U
~
CVLOAHUTION
SOUM
unuurroa [
•olid hajrt
LX9Q10
COOLING
An
C3HDBUATXO*
tttH ^ 71C1,
rn-
,w*:r:cATto*
ajtc
ou?su*?xo«
J
«OLMKiO AJTO
raTKATlCH
D:rr:u-*r:o«
BOTTOM wxrrt
javorrca
tJOj *CII»
!
IPAMtlOi
ikttcmics on
1
TtxatMt I
*T»o«rau
• Kmin vmtk
IXXS 9* CAWTJC 1
i L
i!
vrvocnoMTS
9tCO*90«rriOII
»#«
NATS*
) *
t
TlOj FIOCVT
Figure 14-1. General process diagram for production of titanium dioxide
(chloride process) from high grade ores.
366
-------
TABLE 14-3. WATER USAGE IN TITANIUM DIOXIDE-CHLORIDE PROCESS /HIGH GRADE
ORES SUBCATEGORY
water usage at plants
Water Use
(m^/kkg of TiO^)
Vteter Use
Plant #102
Plant #172
Plant #199
Noneontaot cooling
182
10.66
426
Direct process contact
10.5
15.53
73.2
Indirect process contact
NA
0.72
26.5
Maintenance, equipment
cleaning and work area
washdown
6.65
0.52
2.80
Air pollution control
0.25
7.14
11.3
Nonoontaot ancillary uses
11.60
10.4
9.5
Sanitary & potable water
0.23
0.31
5.6
Total
211.23
45.28
554 .9
NA = 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 OP PLANTS VISITED AND SAMPLED
14.3.1 Screening
Plant f559 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
-------
TABLE 14-4. V&STE WiTER FLOW FOR TITAN UN DIOXIDE-CHLORIDE PROCESS
SUBCATEGORY
SUBCATEGORY
TITANIUM DIOXIDE (Chloride Process)
Plant
Unit Vfeste Water Flow Going to Treatment Plant
#
(m3/kkq of Ti£>2)
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 waste treatment. The average
flow of Plants #559 and #199 was used as the model plant flow for
cost estimating and regulation development.
369
-------
11)11
lUP'lf Wild
NUflCIMl
SU*HT WAT (It
Will
SU'FlT WAT (I
OJ
o
OTMM PRODUCT
MKH At 10)
WASTi WAT(I
SAIIlTAkT
ADO TIO FINISH IMG
ARCA WAS It WAT(K
UTTUnC ?0NO
0TH(« PRODUCT
(STROnC AC 10)
WASTf WATI*
MAC TO*
MAC T0«
A "
6
> S
SlUMlCO
fll SOLI OS
01 ST 111AT i On 10Tion
WASTI UATJ*
e
UCtNQ
sAxniNC joints
SITTIINC fONO
li
-e—
fI HAL f'lUlNT
CH10IU0C MOCISS SC«UMt«
WASTC WATIK
OIMI* MOOUCT
WASH WATik
Figure 14-2. Conoral flow diaqram at Plant #559 showinq the sanplinq 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 fl72 was sampled in the verification 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 f559 . 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 f\12. 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 were:
371
-------
TABLE 14-5. new AND POiAWaNT GOrCENTRATICN tVTA T THE SWIiD V*STE
STEAMS OF PIAMT #172 PKXKJCD*; TITANIUM DIOXIDE BY CHUDRIDE-flirrili PROCESS
FOllutant
STREAM 12
STRD\M #5
CAIdlATED ESTIfWTE
STREW1 16
irr
Pit Sol ids and
Distillation Bottans
Scrubber and
Contact Cooling Vtater
D F F
Total Raw Waste
Treated Effluent
(AxfttlO 3)
(DxExlO-3) (A+D) (OF) (»&dO~3/C)
Unit Flow Cbnc., Uhit Load (Uhit Flow) Gone. Uhit Load Uhit Flow Uhit load Cone. (Jhit Flow Gbrc. liiit Inaq)
10.9
80.1
91
91
TSS
6903
75.2
314
25.2
100.4 1103
23
2.1
Iron
1348
14.7
143
11.5
26.2 288
4.4 0.4
w. Cftromiin
hj
112
1.2
0.11
0.01
1.21 13.3
0.03 0.003
U»d
3.53
0.04
0.009 0.001
0.041 0.5
0.002 0.0002
Nickel
3.46
0.04
0.016 0.001
0.041 0.5
0.005 0.0004
Zinc
2.12
0.02
0.13
0.01
0.03 0.3
0.06 0.005
(1) See Figure 14-2 for location of snnpling points
-------
PROCESS
WASTE WATER
LIME
BOX
NaOH
f MIXING
BASIN
NEUTRALIZE
RAIN RUNOFF
HOLDING POND
FOR
RETREATMENT
RAIN RUNOFF
RETENTION
BASIN
RETENTION
pH ADJUSTMENT
LEGEND
SAMPLING POINTS
THE TOTAL RETENTION TIME
OF WATER IN THE TWO PONDS
IS 5 days.
DISCHARGE
Figure 1^-3. Genera! flow diagram at Plant #172 showing the sampling points.
Titanium dioxide (chloride process) manufacture.
373
-------
TABLE 14-6. FLOW AND POLLUTANT CONCENTRATION DATA CF THE SAMPLED VASTE
STREAMS FOR PLANT #172 PRODUCING TITANIUM DIOXIDE (CHLORIDE PROCESS)
SAMPLED STREAM #1
SAMPLED STREAM #3
Pollutant Raw teste Influent
Treated Effluent
TSS
A B C , D E F,
(A+BxlO) (D+ExlO" )
Unit- Flow Avg. Cone. Unit Load Unit Flow Avg. Cone. Unit Load
(mVkkg) (mg/1) (kg/kkg) (inVkkg) (mg/1) (kg/kkg)
34.7
171
34.7
5.93
6.7
0.23
Iron
Chrornium
Lead
2.9
0.72
0.005
0.10
0.03
0.0002
0.33
0.02
0.002
0.01
0.0007
0.00007
Nickel
0.08
0.003
0.01
0.0C03
Zinc
0.3
0.01
0.09
0.003
374
-------
Maximum Raw Waste Concentrations Observed (pg/1)
Pollutant Screening Verification
Plant f559 Plant #172
Chromium 152,000 1800
Lead 5,150 NS*
Nickel 6,320 NS
Zinc 3,300 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
#559 and fl72. 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 #559 and
#172 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 treatable concentrations at both plants
sampled in the screening and verification phase. Lead, nickel,
and zinc were found in the raw waste of Plant #559 at treatable
levels, but were not present in the Plant fl72 raw waste. At
Plant #559, the chloride process waste effluents are mixed with
375
-------
TABLE 14-7. RAW VftSTE POLLUTANT DATA SUMMARY OF TOE SAMPLED STREAMS
SUBCATEGORY:
TITANIUM DIOXIDE (CHLORIDE PROCESS)
Average Daily Pollutant Concentration and Loadings at Plants Sampled
(kg/kkg of Ti02)
(mg/1)
Pollutant
Plant
#559
Plant
#172
Overall
Average
Toxic :
Iron
Chrcmium
Lead
Nickel
Zinc
Conventional'
TSS
26.2
(288)
1.21
(13.3)
0.041
(0.5)
0.041
(0.5)
0.03
(0.3)
100.4
(1103)
0.10
(2.9)
0.03
(0.72)
0.0002
(0.005)
0.003
(0.08)
0.01
(0.3)
5.93
(171)
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 #172, 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 Modification 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 5172, 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
d ischarge.
At Plant fl99, all the process waste waters are combined,
including storm water and sanitary waste water. The combined
waste 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 f102 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 £605, 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 OP APPROPRIATE TECHNOLOGY AND EQUIPMENT
14.5.1 Technologies for Different Treafent 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
neutralization, using no
and mixer.
In Level 2, second
gravity clarification
necessary pH controls.
379
essentially lagooning with lime
special equipment except a lime feeder
stage lime treatment is followed by
and multi-media filtration, with
-------
LAGOON
LAGOON
RAW
WASTR WATER
EFFLUENT
to
oo
0 •'
ln
-------
RACK WASH
tJ
00
LAGOON
WASTt WATKI
¦A 1AGOCN
I.IMF
LAGOON
LAGOON
filter rnrss
"1
I 0
IkTwjJB
pH AWOSTMTKT
SUMP
CLARIFIER
DUAL
MEDIA
FILTKR
• rrru/KNT
TO LANDFILL
*nclo4e« Hew monitoring, pit monitor Iff »wl ••mplrr.
Figure 14-5. I^vel 2 waste water treatment for titaniun dioxide
process.
— chloride
-------
KKRROUS
SULFATE
SODIUM
HISUI.FIDF
UACKWASH
1J M K
KAW
WASTK WATER
DUAI.
MEDIA
F'll.TER
SUMP
MIXING
-\ UGOON
CLARIF1ER
HI.TKR PRESS
SUMP
TO IANEEILJL
Includes flow monitoring, pH monitoring Ami sampler
Figure 14-6. Level 3 waste water treatment for titanium dioxide — chloride
process.
-------
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 qas 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 Ti02 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 Hater 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
#559 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 Ti02 (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 Ti02 (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 Renoval with BPT Treatment
Treatment Level 1 is equivalent to the proposed BPT in the
Ti02 subcategory (chloride process).
Plants £559 and fl72 practice neutralization and settling
of the raw waste. At Plant f559 , the chloride process raw waste
385
-------
TABLE 14-8. MODEL PLANT TREATMENT C06TS
Subcategory TITANIIM DIOXIDE-Chloride Process
Production
Waste water flow
16,900 metric tons per year
48 metric tons per day
1485 cubic meters per day.
( 18,632 tons per year)
(53 tons per day)
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT C06T
Construction¦
Equipnent in place,
including piping,
fittings, electrical
vork and controls
Monitoring equipment
in place
Engineering design
and inspection
Incidentals, overhead,
fees, contingencies...
Land
TOTAL INVESTMENT C06T
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
$368,500
209,000
9,000
117,300
117,300
192,000
$1,013,100
$56,000
3,700
140,000
82,110
30,393
108,000
15,000
$435,203
$133,592
$568,795
$49,000
389,000
87,600
87,600
6,000
$619,200
$84,000
4,300
34,100
61,320
18,576
9,000
7,500
$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 C06TS
Subcategory TITANIIM 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
INVESTMENT C06T
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
TOTAL INVESTMENT COGT
$1,342,800
$707,120
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
AMORTIZATION OF
INVESTMENT COST
$173,568
$114,072
TOTAL ANJUAL 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 TITANIIM DIOXIDG-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 flew 3980 cubic meters per day.
LEVEL OF
TREATMENT*
FIRST
SECCND
INVESTMENT C06T
Construction
$815,500
$76,800
Equipment in place,
including piping,
fittings, electrical
vrork and controls
283,000
590,000
Monitoring equipment
in place
9,000
Engineering design
and inspection
221,500
133,360
Incidentals, overhead,
fees, contingencies...
221,500
133,360
Land
504,000
6,000
TOTAL INVESTMENT COST
$2,054,500
$939,520
OPERATION AND
MAINTENANCE COST
Labor and supervision.
$56,000
$84,000
Energy
4,600
7,650
Chemicals
374,000
95,000
Maintenance
155,050
93,352
Taxes and insurance...
61,635
28,185
Residual waste
disposal
294,000
20,000
Monitoring, analysis
and reporting
15,000
7,500
TOTAL OPERATION AND
MAINTENANCE COST
$960,285
$335,687
AMORTIZATION OF
INVESTMENT COST
$252,266
$151,883
TOTAL ANNUAL COST
$1,212,551
$487,570
~First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
388
-------
2.0
O
o
o
%
o
o
o
¦L I.
x
CO-
Li vrt-s 12 i. \»\
^BVEL #1 I
10 20 30 40 50
PRODUCTION (METRIC TONS/YEAR X 1000 )
Figure 14-7. Annual treatment cost vs. production for the
Titanium Dioxide Subcategory, Chloride Process
389
-------
60
50
>
40
30
'
I :
¦
i
1 1
. i,
,
i
¦ i
,
'
1
1
1 !
' 1
I ¦ !
1 :
i
1
J
1
i
I
1
1
i
1
:
: I
!
i
i
1 :
i
i
i i
i
i
;
J I
.
i
1
i
i
;
,
1
i
<5>
i
1
1
l ¦
1
\
1
1
¦ i
1
i
1
\
t
i
\
1 "
1
1
; i
i
!
: i i i
¦
. i
1
1 :
V
i ! :
! !
i
\
i
1 1
! 1 i 1
i
|
\
:
i
1
i
1
&
1
1 1
l
. :
.
X
1
i
i
N
I
1
1 !
1
^Sj
1
i 1
i
1
i
1 1 : !
I . |
i
^ ¦ 1.
¦ i :
|
: i ; ! I
i "4
3 ! 1 I
; !
i
1 1
. i v i
1 !
• 1
'
, ;
1 ' !
! j
: ! i
i
1
i
. 1
¦ 1
i
• i
1
i
¦ 1
1 :
I i !
i
i
¦
i i
I
!
! i ¦
i
1
1
1
¦
¦
1
j 1
i
1
1
I
1
1
1 ¦
,
1
1 1
1
1 '
i
\
1 ' ¦
d
EVEL
#i!
j
I
: ! : 1 1
i
1 i
¦ ¦ | ! i
1 )
i
1
' 1 I
1
! !
i
1
: i II! 1
i
1 1
I
¦ 1 '
; ' ! ! ; , !
! ¦ 1 1
i ' l
; i
-
!
1 . i . ! 1 : i
1 ¦
i
PRODUCTION (J-ETRIC TONS/YEAR X 1000 )
Figure 14-8. Annual unit treatment cost vs. production for the
Titanium Dioxide Subcategory, Chloride Process
390
-------
TABLE 14-11. MODEL PLANT TREATMENT C06TS
Subcategory TITANIUM DIOXIDE-Chloride Process
Annual Treatment Costs ($Akg)
COST ITEM
PRODUCTION FLOW
(kkg/yr) (m3/day)
LEVEL OF TREATMENT
FIRST SECOND THIRD FOURTH
Annual Operation
and Maintenance
16,900
1,485
25.75
12.95
13.27
25,500
2,240
23.41
9.82
10.09
45,200
3,980
21.25
7.43
7.65
Annual
Amortization
16,900
1,485
7.90
5.90
6.07
25,500
2,240
6.81
4.47
4.60
45,200
3,980
5.58
3.36
3.47
Total Cost
16,900
1,485
33.66
18.85
19.33
25,500
2,240
30.22
14.29
14.68
45,200
3,980
26.83
10.79
11.12
Not
Applicable
391
-------
water is mixed with the sulfate process waste water for
treatment. Also at Plant f559, the spent ore and coke (solid
residues from the chloride process) are slurried with water and
sent to the treatment facility whereas at Plant fl72, the solid
residues are hauled to a chemical landfill. Long-term treated
effluent data have been submitted by both Plants f559 and fl72.
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 $172 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
i nstalled.
7iow 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 {199 do, in
fact, send the solid residues to the treatment system. The
model plant treatment system is based on an inflow rate of 100
.¦n3/kkg of Ti02 which is an average value of the effluent flow of
Plants #559 and fl99. 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
load ings.
392
-------
TABLE 14-12. HISTORICAL EFFLUENT MONITORING
DATA SU-IIARY WITH VARIABILITY FACTOR
Daily Measurements
Subcategory: Titaniun Dioxide
Chloride Process (Rutile Ore)
Plant #559
April 76 through Septanber 78
Pollutant
TSS Cadmium Chrcmiun Iron Lead Nickel Zinc
Daily Data(1)
No. of Points
Average x, ppm
Standard
Deviation, S
Standard
Deviation, S'
Variability
Factor
30-day(1)
Averages
No. of Points
Standard
Deviation
Variability
Factor
889
21
65.93
1.54
11.0
109
0.058
0.044
0.68
3.85
30 26
21.84 0 .042
3.04
2.4
Variability
Factor Ratio
VFR
(2)
128
0.072
0.054
0 .67
3.81
30
0.038
2.04
854 128 128 128
0.620 0.068 0.08 0.151
3.46 0.041 0.07 0.20
1.86 0.56 0.76 1.02
13.5 3.2 4.4 6.4
28
4.0
30
30
30
0.94 0.04 0.05 0.155
2.1 4.4
3.1
3.6
1.6
1.9
3.4
1.5
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 3C-day averages.
393
-------
TABLE 14-13. HISTORICAL EFFLUENT MONITORING DATA SUMMARY WITH VARIABILITY
FACTORS
DAILY MEASUREMENTS
SUBCATEGORY: TITANIUM DIOXIDE-Chloride Process
(Rutile/Upgraded Ilmenite Ore)
Plant #172
TSS
Pollutant
Chromium
Copper
Zinc
Daily Data^
No. of Points 454
Average x, ppn 5.39
Standard deviation, S 9.13
Standard deviation, S' 1.16
Variability factor 7.6
30-Day Averages
No. of Points 15
Standard deviation, S 6.31
Variability factor 2.92
Variability Factor Ratio ^
VFR 2.6
454
0.008
0.016
1.27
8.6
15
0.012
3.46
2.5
454
0.02
0.03
1.08
6.9
15
0.028
3.29
2.1
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 AND #172
SUBCATEGORY: TITANIUM DIQXIDE-Chloride Process
Pollutant
Plant #559
Plant #172
Pollutant
Concentration
(mg/1)
A
Raw
Waste
B
Treated
Effluent
Percent
Removal
00
Pollutant
Concentration
(mg/1)
D
Raw
Waste
E
Treated
Effluent
Percent
Removal
F=
00
TSS
1103
23
97.9
171
6.7
96.1
Iron
288
4.4
98.5
2.9
0.33
88.6
Chromium
13.3
0.03
99.8
0.72
0.02
97.2
Lead
0.5
0.002
99.6
0.005
0.002
60
Nickel
0.5
0.005
99.0
0.08
0.01
87.5
Zinc
0.3
0.06
80.0
0.3
0.09
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 treatability levels in the raw
waste of Plant #559 and they were: chromium, iron, nickel,
lead, and zinc (Section 14.3.3). A second plant, #172, was
sampled in the verification phase and chromium was the only
pollutant found above treatability levels in the raw waste. At
Plant *559, 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
Ti02 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 Ti02 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 #559 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 f559 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 m3/kkg)/ kg/m3 \
U000 mg/1J
= 6.4 kg of TSS
kkg of T i02
Proposed 24-hour maximum effluent limit
= (230 mg/1) (100 m3/kkg) f kg/m3
\1000 mg/V
= 23 kg of TSS
kkg of Ti02
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 #559 (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 f559 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/1) (13.5) = 8.4 mg/1
Proposed 30-day average effluent limit is:
(2.5 mg/1)(100 m3/kkg)/ m3/kkg \
VlOOO mg/1/
= 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
\1000 mg/1/ 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 pollu tants - 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 f 5 5 9 .
The influent to the treatment system at Plant ?559 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 f559 (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 f559 (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/l)(100 m3/kkg) / kg/m3 \
V1000 mq/lj
= 0,014 kg of chromium
kkg of Ti02
The proposed daily maximum effluent limit is:
(0.27 mg/1)(100 m3/kkg)/ kg/m3 N = 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 f559 intermixes
the chloride and sulfate process waste before treatment. The
presence of these pollutants in the raw waste at Plant f559
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), were considered £or BAT and NSPS but were rejected
on the basis of cost (Level 3 Table 14-11). Level 1, used Cor
BPT, is selected for BAT treatment technology.
Technology Basis
Alkaline precipitation followed by settling used for BPT
(Level 1) is proposed for BAT.
Plow Basis
A unit waste water flow rate of 100 m3/kkg of Ti02 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 pollutants -
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.
9
400
-------
TABLE 14-15. PROPOSED LIMITATIONS
TITANIUM DIOXIDE - CHLORIDE PROCESS (FLTH£ OR UPGRADED IIWEKITE ORE)
Best Practicable Oontrol Technology'Currently Available
Vfeste W&ter Flow: 100 m3/kXg of TiO.,
Subcategory
Pollutant Performance VFR '
Concentration Effluent Limit
Basis (mg/1) (kg/kkg of Ti02)
, ^ . Max Max
lmg/1) 30-day 24-hr 30-day 24-hr
Avg Max Avg Max
Conventional and
Non Conventional"
Pollutants:
Total Suspended Solids 21 ^ 3.6 64 230 6.4 23
Iron 0.62(2) 3'4 2'5 8'4 °*25 °-84
Toxic Pollutants:
Chromium 0.070(2) 1.9 0.14 0.27 0.014 0.027
Lead 0.30(3) 1.9 0.30 0.60 _ (4) _ (4)
Nickel 0.20(3) 1.9 0.20 0.40 _ (4) - (4)
Zinc 0 . 50 ( 3) 1.9 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 LIMITATIONS
TITANIUM DIOXIDE - CHLORIDE PROCESS (R^ILE/UPGRADED ILMENITE ORE)
Best Available Technology
Waste Water Flow: 100 m3/kkg of TiO^
Pollutant
Subcategory,, (2
Performance VFR
Concentration
Basis (mg/1)
Effluent Limit
(kg/kkg of TiO^)
30-day
Avg
24-hour
Max
30-day
Avg
24-hour
Max
Nonconvent ional
Pollutants:
Iron(4)
0.62
3.4
2.5
8.4
0.25
0.84
Toxic
Pollutants:
Chronium(5)
0.070
1.9
0.14
0.27
0.014
0.027
Lead
0.30
1.9
0.30
0.60
(3)
(3)
Nickel
0.20
1.9
0.20
0.40
(3)
(3)
Zinc
0.50
1.9
0.50
1.0
(3)
(3)
(1) See Table 14-14 for details.
(2) WR: 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.
Plow 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 f f 5 59 and fl99) 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.
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 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 Ti02
Nonconventional 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 Ti02
The proposed 24-hour maximum effluent limit is:
(0.84 kg/kkg)(0.70) = 0.59 kg of iron
kkg of Ti02
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 Ti02
(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 cf 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 1.4-16) by 0.40 as follows:
The proposed maximum 30-day concentration basis is:
(0.14 mg/1) (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 Ti02
The proposed maximum daily effluent limit is:
(0.030 kg/kkg) (0.40) = 0.010 kg of chromium
kkg of Ti02
B. Other metals: Treatability studies have indicated
that the following increased removals of lead, nickel, and zinc
can be achieved by filtration (40,41).
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
Additional Removal by
Filtration Using Settled Effluent
(*)
Lead
N ickel
Zinc
60
14
6
405
-------
is:
is:
The proposed 30-day average nickel concentration basis is:
(0.20 mg/1) (0.86) = 0.17 mg/1
The proposed 24-hour maximum concentration basis for nickel
(0.40 mg/1) (0. 86) = 0 . 34 mg/1
The proposed 30-day average concentration basis for zinc
(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 Pretreataent 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
-------
TABLE 14-17. PROPOSED LIMITATIONS
TITANIUM DIOXIDE - CHLORIDE PROCESS
New Source Performance Standards
Waste Water Flow: 100 m3/kkg of Ti^
Pollutant
Treatability VFR
Concentration Effluent Limit
(1) Basis, (mg/1) (kg/kkg of Ti02)
24-hour 30-day 24-hour
Avg Max Avg fax
Conventional and
Nonconvent ional
Pollutants :
Total Suspended
Solids
15
3.6
45
160
4.5 16
Iron
(2)
0.40
3.4
1.8
5.9 0.18 0.59
Toxic Pollutants:
Chrcmium
Lead
Nickel
(2)
Zinc
0.030
0.060
0.17
0.47
1.9
1.9
1.9
1.9
0.060 0.12 0.0060 0.012
0.12
0.17
0.47
0.24
0.34
0.94
(3) (3)
(3) (3)
(3) (3)
(1) VFR: Ratio of 24-hour variability factor to the 30-day variability
factor. J
(2) Also applicable for PSNS limitations.
(3) No effluent limitations proposed.
407
-------
14.8 TITANIUM DIOXIDE - SULFATE 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 Ti02 by the sulfate process
utilizes three important steps:
1. Digestion: Fe0.Ti02 +
2. Precipitation: TiO.SO
3. Calcination: Ti02.H20
2H2S04 = FeS04 + Ti0.S04
+ 2H20
+ 2H20 = T i02.H20 + H2S04
= 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. - SUBCATEGORY PROFILE DATA SUMMARY
SUBCATEGORY TITANIUM DIOXIDE SULFATE 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:
Minimun
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flew 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 cure Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Comnerae, Current Industrial
Reports, Deoentoer 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
-------
TABLE 14-19. ANALYSIS OF 111*1*ITE OMiS1
UN ITU) STKTES
h-»
O
Ovatucal
Constituent
Ti02
FcO
re2°j
SiOj
a12°3
PA
ZrO,
MgO
hUO
CaO
V2°5
cr203
Piney
River
Virginia
44.3
35.9
13.8
2.0
1.21
1.01
0.55
0.07
0.52
0.15
0.16
0.27
Rose land
51.4
37.9
1.6
4.6
0.55
0.17
NA
2.35
0.70
0.59
0.07
NA
New York
44.4
36.7
4.4
3.2
0.19
0.07
0.006
0.80
0.35
1.0
0.24
0.001
Florida
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
CANADA
Ivry Bourget Allard
42.5
39.1
20.7
0.88
1.05
NA
NA
2.0
0.04
0.1
0.36
0.15
22.4
36.9
31.2
1.0
6.01
0.93
NA
1.50
NA
0.55
NA
NA
37.3
26.3
30.0
NA
NA
0.004
NA
NA
0.10
NA
0. 39
NA
^Constituents expressed ac 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 Dse
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
resulting 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
-------
IIMMITE
STRONG-AC10
RECYCLE _
HATER -
STEAM -
EMISSIONS
NATER
HUO SLURRY (MEAK ACID)
EMISSIONS
HATER
HATER
STEAM
EMISSIONS
STEAM
Ef FLUENT
MATER
WEAK-ACID
RECYCLB
STRONG AC It*
HATER
STEAM
4ASTC U1SKBAL «-
HEAR ACID-
RECYCIB
HEAR ACID
EMISSIONS
HATER
COOLING SPRAYS AND ELECl
¦ROSTATIC PRECIPITATORS |
HEAR ACID
EFFLUENT -
HATER
EPPLUENT
RECYCLE
HATER
STEAM
EMISSIONS.
NATER
EFFLUENT-
TITANIUM
DIOXIDE PIGMENT
EFFLUENT
PACKAGING
PRECIPITATION
JET MILLS
FLASH
COOLER
EVAPORATOR
FIRST MOORE
FILTER
SECOND MOORE
FILTER
HBT MILL
DIGESTER
MIST ELIMINATORS
JET MILL CONDENSERS
CONDENSERS
JET MILL SCRUBBERS
CONDENSERS
SPRAY CONDENSERS
AMD VENTUHI SCRUBBERS
— EFFLUENT
TO SALES
Figure 14-9. General process flow diagram for production of titaniun dioxide by
-------
TABLE 14-20. WATER USAGE IN TITANIUM DIOXIDE - SULFATE PROCESS SUBCATEGORY
Uses
Water Usage per Unit of Production
( m^/kkg of Ti02)
Noncontact cooling
Direct process contact
Indirect process contact
(purnps, seals, leaks,
spills, etc.)
Maintenance, equipment
cleaning and work area
washdcwn
Air pollution control
Noncontact 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 OF PLANTS
14.10.1 Screening
Plant £555 was visited and its waste streams sampled in the
screening phase by an EPA Region II team. Tne 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 #555.
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 Verif ication
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 C02 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
-------
TABLE 14-21. RAW WASTE CHARACTERISTICS (INDUSTRY DATA) (1)POR PLANT #555
(PRODUCTION OF Ti02 BY SULFATE PROCESS)
Waste Source
Unit
Flow
(mVkkg
of Ti02)
PH*
Pollutant Waste Loads,(kg/kkg of Ti02)
Acidity NH, Fe TSS TDS
(as H2S04) (as N)
Digestion
115
3.0
20.8
NA
0.042
9.3 35.7
Clarification
3.58
2.5
26.7
NA
8.42
175 40.8
Evaporation
113
4.0
18.7
NA
1.14
3.2 20.2
Cooling
20
6.1
2.49
NA
0.099
0.46 3.09
Strong Acid from 8.49
first Moore Filtration
< 0.5
2.360
NA
139
0.959 2.815
Weak Acid from 12.2
first Moore Filtration
2.0
88.3
NA
3.8
0.23 98.8
Weak Acid from 10.4
second Moore Filtration
1.7
148
NA
0.29
0.13 151
Weak Acid from
first stage
Calcination
12.0
2.0
20.8
NA
0.22
2.0 7.50
Weak Acid frcm
second stage
Calcination
40.0
2.2
19.2
NA
0.64
4.92 33.1
Calcination Mist
Eliminators
38.7
3.0
7.50
NA
0.02
0.21 27.9
Wet Milling Washing
and Drying
11.1
8.0
NA
8.6
0.01
2.13 11.0
Jet-Mill Condenser
27.0
6.5
NA
NA
0.01
1.1 2.7
Jet-Mill Scrubbers
18.0
7.4
NA
NA
0.13
1.7 3.58
Boiler and Water
Plants
16.6
9.0
NA
NA
0.66
5.25 8.92
NA: Not Available
* Value in pH units
(1) - Response to 308 Questionnaire, 1976
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 Plant 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 £696 , 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 #605, 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,
* ' Irom the reactor
and placed in a
__y" 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. Th
filtrate from the centrifuge is recycled to the thickener, an
the thickener overflow is discharged.
e
and
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
-------
OTHfft MOOUCT
VAJTt WMU
V( AX ACID _
WA1TI J1*IA*
WtAK ACID FONt
~0"
*ivi*
lU'HT VAT(II
RUNICI'Al
SU'HV WATU
Mil
lu'nr uati*
mo*c acid .
WASTt 1TMM
1T10NC acid
roxo
/)
-e-
n
-€+
i o
¦0-
/i
&
I".
MACroti
riiTi*
HCACTOHJ
fILTl*
10U01 TO
STOM.CC/
lAHonu
KCACTOKS
riNAi.
JCTTI.INC
POHO
-0-
irnutt
H
OTHCK PRODUCT
VAST! VAT(K
TIO. (SUlfAU 'ftOC(Sl)
' SCHuBM*
UA5H VAT|*
SOIIOS
TO
JTOHACt/lANOflll
UCtxO
Figure 14-10. General flew diagram at plant #559 shewing the sanpling points.
(Titaniun dioxide - sulfate process.)
-------
TABLE 14-22. FLOWS AND POLLUTANT CONCENTRATIONS FOR THE WASTE STREAMS
SAMPLED FOR PLANT *559 PRODUCING TITANIUM DIOXIDE
Stream Sampled Unit TSS Iron
No. ^ Stream Flow Load Load
Description (rnVkkg) (kg/kkg of Ti02) (kg/kkg of Ti02)
Weak Acid 68.i1)(2) 1.23 1.23
Pond Overflow
Strong Acid
Pond Overflow 6.1 205.85 106.34
(1) (2)
Scrubber and 361.4 113.5 51.68
Contact Cooling
Water
(1)(2) (3)
Final Treatment 436 10.0 1.92
Effluent
(1) - The flow is contributed by the sulfate process stream.
(2) - Hie pollutant load was calculated by multiplying the flow contributed
by the sulfate process stream times the concentration of pollutant.
Pollutant Load = (total stream flew) 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 ana
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, f694, and £559. Complete flow data is
not available for Plants #696 and #605.
14.10.4 Toxic Pollutant Concentrat ions
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) = (CM (Q)
1000
Where the concentration (C) of the pollutant is expressed
in units of mg/1 (Note: 1 kg/m3 = 1000 ma/l), 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
Flow in (m /kkg of TiX^)
A B C D = A + B + C
Strong acid Weak acid Scrubber and
contact cooling Total Effluent
water
#555
#694
#559
Average
8.49
16
6.10
10
78.2
67
69
72
362
457
361
393
449
540
436
475
421
-------
Unit loading fas kg of pollutant = (C) (Q)
per day kkg of Ti02) 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 lb).
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 Conce
Screen i
Pollutant (Plants f'555
ntration Observed (pg/1)
ng Verification
& *559) (Plant *559)
Cadmium
340
12
Chromium
124 ,000
31,000
Copper
T , 500
1,000
Lead
3,700
5,200
Nickel
6,400
3. ,300
Z inc
3,800
17 ,000
Antimony
20
1,400
Arsen ic
11
340
Thallium
19
41
Selenium
360
Belov
A summary of daily and unit (per
waste loads for all plants sampled can
Individual plant raw waste loads and
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:
unit of production) raw
be found in Table 14-24.
concentrations found in
422
-------
TABLE 14-24. SUMAKY OF RAH WASTE LOADINGS FOUND IN SOSBnNT. MID VERIFICATION SAMPLING
SUBCATEUORY
TITANIUM DIOOOIK - SUIi'ATK PHOCESS
Pollutant
NJ
u>
TOxic
Antimony
Arsenic
OiLkruun
Ouxmim
Cower
Lead
Nickel
Seleniun
Ttial 1 i 1*0
Zinc
Loading Range,
(kg/day)
Minimis
5.0
1.9
.068
140
8.2
3.0
3.7
7
.47
1.8
Maxims
Minimum
Ifriit Loading,
(kg/kkg)
Average
KiximjD
Conventional and Mormon ventuirwil
TSS
Iron, KG
28
4.0
7.2
530
19
65
23
9.5
1.2
85
.032
.012
.00044
1.1
.065
.024
.029
.0020
.0030
.014
0.11
0.19
0.19
2.0
.085
.18
.080
.031
.0055
.34
320
600
0.22
.032
.057
3.4
.12
.42
0.15
.060
.0080
.55
No. Of
Plants
(1) - Uata are taken only frcxn thoae plants where pollutants were found above detection limits, or, in the
case of TSS arxl Iran, where data are available.
-------
SUBTATVXJOKY
Screening Verification
Plant 1555 Plant 1559 Plant 1559
(mq/1)
(kg/kk
-------
Pollutant
Total Annual Raw
Waste Load (kg/year)
Cadmium
Chromium
Copper
Lead
Nickel
Z inc
Antimony
Arsenic
Selenium
Thallium
5,000
510,000
22,000
47,000
21,000
88,000
29,000
49,000
8,000
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
alternat ives.
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 Manageaent Practices
Storm water runoff from the plant site should be collected
and sent to the treatment facility for the removal of suspended
sol ids.
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 OP APPROPRIATE TECHNOLOGY AND EQUIPMENT
14.12.1 Technologies for Different Treatment Levels
Level 1 (BPT)
In the Level ] 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 landfill. The flow diagram of
this treatment is shown in Figure 14-12.
427
-------
*l*rl*4r« ftov iMB»Oa»lt, fN «*4 nw>l»»
Figure 14-11. Level 1 waste water treatment for titaniun dioxide - sulfate
-------
CHO*JND
UMCSTONC
C4OO.
RAPID Mil AKO SCTTUItO
MCTCIXO KrrUJtWT
(X»! TO
ATMOOMtSftE
--^5)
£*€», IUIIRT
TANK
WO IT» PI M
HEACTOAJ
AF.AATIOM I M1WHO
COj HCMOVAI,
III ITAUK
TUCMtNKH
IMOCTACK
TMCKCHU
»KAK AQD
wastk wATr.1
tr9iA.tr. mt
wiios nstx»AL
' UMSTK
¦a*
*a»tr«atci
n«« «nw<*6»»le®« |N niMiliil«| «
-------
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 Plow
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), 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 Oseage
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/l, 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
"other 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 f559.
Pollutant Renoval 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 f559 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 14-26. 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 flew 42750 cubic meters per day.
LEVEL OF
TREATMENT*
FIRST
SECCND
INVESTMENT COST
Construction
$701,200
$117,500
Equipnent 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
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
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 TITANIIW DIOXIDE Sulfate
Production
Waste water flow
47,700 metric tons per year
136 metric tons per day
64600 cubic meters per day.
(52,589 tons per year)
(150 tons per day)
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT COST
Construction
Equipnent in place,
including piping,
fittings, electrical
work and controls
Monitoring equipnent
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 TITANIIM DIOXIDE Sulfate
Production
74,500 metric tons per year (82,136 tons per year)
212 metric tons per day (234 tons per day)
Waste water flew 100700 cubic meters per day.
LEVEL OF TREATMQ/T*
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 C06T
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
$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
$711,403
~First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
436
-------
LEVEL «Z
20
405060
PRXUCTICN (metric tons/year x looo)
70
Figure 14-13. Annual treatment cost vs. production for the titanium dioxide
subcategory, sulfate process.
437
-------
130
120
110
LEVEL *2
100
90
70 50 5TJ i
FKXUCTICN (METRIC TONSADW X 1000)
Figure 14-14. Annual unit treatment cost vs. production for the titanium dioxide
subcategory, sulfate process.
438
-------
TABLE 14-29. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE Sulfate
PRODUCTION FLOW
(kkg/yr) (m3/day)
Annual Treatment Oasts ($Akg)
LEVEL OF TREATMENT
FIRST SECCND THIRD FOURTH
Annual Operation
and Maintenance
31,800 61,600
47,700 92,600
74,500 144,000
94.50 9.83
90.15 8.82
81.38 7.93
Not Applicable
Annual
Amortization
31,800
61,600
21.77
2.51
47,700
92,600
18.85
2.10
74,500
144,000
15.95
1.62
Total Cost
31,800
61,600
116.27
12.34
47,700
92,600
109.00
10.92
74,500
144,000
97.33
9.55
439
-------
TABLE 14-30. HISTORICAL EFFLUUMT MONITORING DA'IA SIMWRY
Daily Data
No. of RDints
Average, x
Standard ...
Deviation, S
o Standard (2.
Deviation, S'
Variability(3)
Factor
30-Day Average
No. of Points
Standard (1)
Deviation
Variability^
Factor
Variability
Factor Ratio
SUBCATEGORY - TITANIUM DIOXIDE
sttt.f&tk process ctaot #559
Pollutant
TSS Cadmium Chromium Iron Lead Nickel Zinc
899
21.0
65.93
1.54
11.0
30
21.84
3.04
3.62
109
0.060
0.044
0.68
3.85
26
0.042
2.43
1.58
128
0.070
0.054
0.67
3.81
30
0.038
2.04
1.87
854
0.62
3.46
1.86
13.65
28
0.94
4.00
J.38
128
0.068
0.041
0.56
3.16
30
0.04
2.14
1.48
128
0.08
0.071
0.76
4.39
30
0.048
4.39
1.00
128
0.151
0.204
1.02
6.41
30
0.16
3.05
2.10
(Continued)
-------
TABLE 14-30. Continued
(1) S is the arithmetic standard deviation and is given by
S = ,/V (Xi "
n-1
where xi is the data value for point i
x is the mean value
n is the nunber of data points
(2) S' is the estimated standard deviation
S' =
V <-(+)')
where S is tlie arithmetic standard deviation
x is the mean value
(3) The variability factor (VF) of daily measurements for lognormal distribution
is found by the expression
In (VF) = S' ( Z-0.5S')
where S' is the estimated standard deviation
7,= 2.33 for 99th percentile
(Continued)
-------
TABI.E 14-30. Continued
(4) The variability factor (VF) for 30-day average measurements is found by the
expression
VF = 1.0 + Z (-$-)
^x/
Where x is the mean value
S is the arithmetic standard deviation
7. - 1.64. for 95th percentile
(5) VFR: Ratio of the 24-hour variability factor to the 30-day variability factor
-------
TABLE 14-31. VERIFICATION RESULTS FROM - SULFATE PROCESS
TITANIUM DIOXIDE PIANT *559
Pollutant
Raw Vfeste
Treated Effluent
A
Unit Load
(kg/kkg)
B
Concentrat ion
(mg/1)
C D
Unit Load Concentration
(kg/kkgN (mg/1)
E
Ranoval
Efficiency
(%)
Total Suspended
Solids
116
266
10.0
23
91
Iron
364
835
1.92
4.4
99
Cadmium
0.0 0045
0.0010
0.00004 0
0.00010
90
Chromium
1.3
3.1
0.011
0.025
99
Copper
0.070
0.16
0.002
0.0050
97
Lead
0.040
0.96
0.00090
0.0020
99
Nickel
0.060
0.14
0.0020
0.0050
96
Zinc
0.45
1.0
0.030
0.062
94
Arsenic
0.012
0.028
0.0040
0.010
64
Antimony
0.030
0.074
0.0060
0.015
80
Selenium
0.0020
0.0050
0.0020
0.0050
0
Thallium
0.0030
0.0070
0.0010
0.0030
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.
Plow 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 pollu tant loadi ng - 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 f559 (Table 14-30) was used as the proposed subcategory
performance values. The variability factors for daily and 30-
day average estimated from Plant (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 mg/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
\1000 mg/1/ kkg of Ti02
f rom:
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 limit1
= (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/kko) / kg/m3 \ = 4.1 kg of iron
I 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 \) long-term monitoring data
for Plant 1559, 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 ug/1 in Section 14.10.4).
At Plant t'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 ?4-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)/ kg/m3 \ = 0.38 kg of antimony
\1000 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
\1000 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 ug/1 in Section 14.10.4). The data for
Plant #559 indicated a removal efficiency of 90.0 percent (Table
446
-------
14-31). Thus, the long-term average value of 0.060 ma/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
f559 (Table ]4-30) 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 ^559 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
*559, 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 mq/l)(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
\1000 mg/1/ kkg of Ti02
447
-------
The proposed chromium 24-hour maximum effluent limit is
given by:
(0.27 mq/1)(475 m3/kkq) / kq/m3 N = 0.13 kg o
VlOOO mg/1/ kkg of
of chromium
T iO 2
D. Copper: The value of 0.5 mg/1, 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) / 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) f kg/m3 \ = 0.46 kg of copper
\1000 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 5 59 (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
\]000 mg/1 / kkg of Ti02
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 f559 (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 \ = 0.30 kg of nickel
V1000 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
V1000 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 #559 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
U000 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 \ = 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 £559. 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 mq/1
The proposed arsenic 30-day average effluent limit is given
by:
(0.50 mg/)(475 m3/kkg) / kg/m3 \ = 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 \ = 0.46 kg of arsenic
\1000 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. PROPOSED LIMITATIONS
TITANHM DIOXIDE SL1FATE PROCESS
Best Practical Control Technology Currently Available
'"teste Vfater Flow-. 475 m3/kkg of TiOj
Ftoliutant
Subcategory
Performance
(nc/1)
VFR
(1)
Concentration
Basis
(ng/:.)
flax
Effluent
Li-nit
(kg^xkg of Ti02
30-day 24-hr. 30-aay 24-hr.
Avg Max Avg Max
Conventional and
Nonoonvent ional
Pollutant3
Tbtal Suspended
Solids
21 '2'
3.6
64
230
30
110
Iron
0.62 (2)
3.4
2.5
8.5
1.2
4.1
T^xic ftollutants
Ar.ti.7Dny
0.80(3)
1.9<4!
0.8C
1.2
0.38
0.71
Cadrjun
:.06(2:
1.6
0.15
0.24
0.070
0.11
Chromium
0.07(2!
1.9
0.14
0.27
0.070
0.13
Copper
0.50'3'
1.9(4)
0.50
0.95
0.24
0.46
Lead
0.30(3)
1.5(5)
0.30
0.45
0.14
0.21
Nickel
0.20!3)
1.9«>
C.20
0.37
0.10
0.18
Zinc
3.5C(3)
2.1(5)
0.50
1.1
0.24
0.50
Arsenic
0.5C<3)
1.9
0.50
0.95
0.24
0.46
(1) VFR: Ratio of the 24-hour variability factor to the 30-day variability
factor.
(2) Long-term average baaed on loading data and variability factors of
plant (559 selected from Table 14-30.
(3) The lover liffit of the literature treatability estimate (Table 8-U)
is used as the basis for the 30-day average liiritation.
(4) Variability factor ratio of chranuur. developed frorr. the long-term data
of plant *559 ha9 been used (Table 14-30).
(5) Variability factor ratio estimated for this pollutant from long-tenr.
data of plant *559 has been used.
451
-------
14.14.3 Basis for Proposed BCT Effluent Limitations
The BCT limitation (applicable only to TSS) was set eoual
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 New 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 fPSES), 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
-------
TABLE 14-33. PROPOSED LIMITATIONS
TITANIUM DIOXIDE SULFATE PROCESS
Best Available Technology
Vbste Vbter Flow: 475 m^/kkg of Ti0o
Rjllutant
Subcategory
Performance
(1)
VFR
(2)
Concentration Effluent Limit
Basis (mg/1) (kg/kkg of TiO^)
24- hour 3
-------
TABLE 14-34. PROPOSED LIMITATIONS
TITANIUM DIOXIDE SULFATE PROCESS
New Source Performance Standards
V&ste W&ter Flow: 475 m^/kkg of Ti©2
Subcategory... vfr(2) Concentration Effluent Limit
Pollutant Performance Basis, (mg/1) (kg/kkg of TiO_)
Max Max 1
30-day 24-hour 30-day 24-hour
Avg Max Avg Max
Conventional and
Nonconvent ional
Pollutants
Total Suspended
Solids
21
3.6
64
230
30
110
Iron
0.62
3.4
2.5
8.5
1.2
4.1
Toxic Pollutants
Antimony
0.80
1.9
0.80
1.5
0.38
0.71
Cadmium
0.060
1.6
0.15
0.24
0.070
0.11
Chromium
0.070
1.9
0.14
0.27
0.070
0.13
Copper
0.50
1.9
0.50
0.95
0.24
0.45
Lead
0.3C
1.5
0.30
0.45
0.14
0.21
Nickel
0.20
1.9
0.20
0.37
0.10
0.18
Zinc
0.50
2.1
0.50
1.1
0.24
0.52
Arsenic
0.90
1.9
0.50
0.95
0.24
0.46
(1) For basis, see proposed limitation for HPT Table-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
PROFILE
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 adeauately 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 beneficiat ion 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 chlorination 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
beneficiat ion 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.
45S
-------
TABLE 14-35 - SUBCATEGORY PROFILE EATA SUMMARY
SUBCATEGORY TITANIUM DIOXIDE Chloride Process (Ilmenite Ore)
Total subcategory capacity rate
Total subcategory production rate
Nunber 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
f-toximum
Waste water flow range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
NA
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
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 Conmerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc., Draft
Report, "Preliminary Economic Assessnent of Elf fluent Limitations in the
Inorganic Chemical Industry,n June, 1978, and "Efcorcmic Analysis of
Proposed Revised Effluent Guidelines and Standards for the Inorganic
Chemicals Industry ,"March, 1980.
NA - not available
456
-------
SWB
-------
Once the TiC14 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 USB AND WASTE SOURCE CHARACTERISTICS
14.16.1 Water Dse
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 £7] 3 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 VftTER USAGE FOR TiO- PRODUCTION
BY THE CHLORIDE -miENITE PROCESS
Use
Plant #237
Plant #550
vin^/kkg of Ti©2)
Plant #713
Noncontact
Cooling
73-140
330-390
15-16
Process Contact
and Clearup
100-140(15
47- 59
29-33(2)
Noncontact
Ancillary Uses
(Boilers,
Sanitary, etc.)
9- 11
6- 7
5- 6
Source of data, (55).
(1) The average total flow of 120 ir^/kkg is used as the basis for 3FT.
(2) The average flow of 31m^/kkg is used as the basis for NSPS.
459
-------
TABLE 14-37. AVERAGE RAW WSTE LOADS FOR Ti02 PRODUCTION
BY THE CHLORIDE - ILMIMITE PROCESS
TSS
HC1
FeCl3
Other metal
chlorides
Plant #237 Plant #550 Plant #173*
Cone. Process Dil. Process Cone. Process Dil. Process Cone. Process Dil. Process
Stream Stream Stream Stream Stream Stream
(kg/kkg of Ti02) (kg/kkg of T:^) (kg/kkg of TiO£)
100-150 20-35
200-230 8-10
900-1150 1-3
140-155 Negl.
150-200 15-20
250-300 0.5-0.8
1000-1200 2-3
190-210 Negl.
200-240 5-20
120-240 Negl.
1000-1200 Negl.
120-150 Negl.
* These values 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 OP PLANTS VISITED AND SAMPLED
14.17.1 Screening
Plant (550 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 f!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 f*55 & f559) (Plant f559)
Chromium
124,000
31,000
N ickel
6,400
1,300
Z i nc
3,800
17,000
Lead
3,700
5,200
Copper
1,500
1,000
Cadmium
340
12
Selenium
340
< 20
Antimony
20
1,400
Thallium
19
41
Arsenic
11
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)
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)
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 lbs).
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.
1000
1000(P)
Pollutant
Total Annual Raw Waste Water Load
(kg/year)
Chromium
N ickel
Z i nc
Lead
Copper
Cadmium
Ant imony
Thall ium
Arsenic
S elenium
1,050,000
42,000
178 ,000
94 ,000
44,000
9,900
58 ,000
2,900
99,000
16,000
463
-------
TAflLE 14-38. St»MARY OF RAM WASTE IXWUNGS POUND IN SCNQJJING AND VERIFICATION SAMPLING
SUBCATEGORY
TITAN IIP DIOXHK - SUIFATE PROCESS (Applied to Oiloridc Ilmaiite Process)
Pollutant Loading Range
(kg/day)
Kiniraun Hixicun
Unit Loading
(kg/kkg)
Hiniran Average
ftexirrun
No. of
Plants (1)
Priority
a\
J*.
Antimony
5.0
28
0.032
0.11
0.22
Arsenic
1.9
4.0
0.012
0.19
0.032
CaAmin
0.068
7.2
0.00044
0.019
0.057
Chrcmiua
140
530
1.1
2.0
3.4
Copper
8.2
19
0.065
0.085
0.12
Lead
3.0
65
0.024
0.18
0.42
Nickel
3.7
23
0.029
0 .080
0.15
Seleniun
7.6
9.5
0.0020
0.031
O
O
Thalliim
0.47
1.3
0.0030
0.0055
0.0080
Zinc
1.8
85
0.014
0.34
0.55
Conventional
TSS
Iron
320
600
(1) - rata are taken only fran those plants where pollutants were fowl above detection limits, or
in the case of TSS and Iron, where data are available.
-------
TABie 14-39. TOXIC POLUTTAKT AVEKMX RAH HASTE LOM£ AND CONCENTRATIONS
SUBCATEGORY TlTMilUH DICOOTR - Sulfate Process (Applied to diloridc Ilmenite Process)
Screening Verification
Plant 1555 Plant 1559 Plant 1559
(mj/1) (kg/kJig) (aq/lt (kq/kkg) (rag/1) (kq/kkg)
Antimony
0.77
0.22
0.16
0.080
0.074
0.032
Arsenic
0.11
0.032
0.029
0.014
0.028
0.012
CaAaiin
0.29
0.057
0.002
0.0009
0.0010
0.00044
Oircraiun
3.8
1.1
7.0
3.4
3.1
1.4
Oojjjjer
0.20
0.065
0.25
0.12
11.
0.070
Lead
0.075
0.024
0.20
0.10
0.96
0.42
Nickel
0.091
0.029
0.31
0.15
0.14
0.061
Sclcnim
NA
< 0.06
NA
NA
0.005
0.002
Ittalliun
NA
NA
0.02
0.008
0.007
0.003
Zinc
0.088
0.014
1.1
0.55
1.04
0.45
NA = 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.] 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 Prevailinq Control and Treatment 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
sol ids.
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 he
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) decarbonization 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 exchange, and membrane applications, all of
which were regarded as categorically inappropriate from a
practical and economic point of view.
14.19 SELECTION OP APPROPRIATE TECHNOLOGY 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
-------
TO
I—O
\L
cb
Figure 14-16. Level 1 waste water treatment for titaniun dioxide - chloride
(ilmenite ore) process.
-------
CjU)
-J
O
MkSTV .
SUM*
ISTIUJ*T!OI
aim* »
MKSTt:
AM)
(XHWtT
(TDLIWT.
I«ST«
»#awe;
MTVO/D f>7UIW
CM), SUM*
\snruwj y
Rjro
Wltff
|-Q»—1*
--^ IAOOCH
0
f* MUIETMMT
::—.
r
-O-*1
NUA
rora
"igure 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 , £t 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
i nvolved.
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 Plows
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 cf
T i02
HCl
230
kg/kkg of
T iO 2
I ron
375
kg/kkg of
T i02
Chromium
1.4
kg/kkg of
T i 0 2
Z inc
0.5
kg/kkg of
T i 02
values for TSS, HCl,
and
iron are
based
The loading values for TSS, HCl, 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.
Cheaical Osage
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 PLAI7T 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 plaoe,
including piping,
fittings, electrical
work and controls 696,500
Monitoring equipment
in plaoe 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 15,000
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 oost.
474
-------
TABLE 14-41. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE 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
WDrk 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 OPERATION AND
MAINTENANCE 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. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE Chloride-Il. Ore
Production 113,750 metric tons per year (125,409 tons per year)
325 metric tons per day (358 tons per day)
Waste water flew 39,000 cubic meters per day.
LEVEL OF TREATMENT*
A. INVESTMENT COST FIRST
Construction $508,000
Equipment in place,
including piping,
fittings, electrical
vrork and controls 1,179,500
Monitoring equipment
in place 9,000
Engineering design
and inspection 339,300
Incidentals, overhead,
fees, contingencies... 339,300
Land 780^000
TOTAL INVESTMENT 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 15,000
TOTAL OPERATION AND
MAINTENANCE COST $2,030,163
C. AMORTIZATION OF
INVESTMENT COST __$386i428
TOTAL ANNUAL COST $2,416,591
* First level represents the base cost of treatment system.
Other levels represent the incremental cost above base oost.
476
-------
TABLE 14-43. MODEL PLANT TREATMENT 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 OF TREATMENT*
A. INVESTMENT COST FIRST
Construction $638,000
Equipment in place,
including piping,
fittings, electrical
vrork 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 AND
MAINTENANCE 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 15^000
TOTAL OPERATION AND
MAINTENANCE COST $2,421,946
C. AMORTIZATION OF
INVESTMENT COST $456^43
TOTAL ANNUAL COST $2,878,189
*First level represents the base oost of treatment system.
Other levels represent the incremental oost above base cost.
477
-------
TABLE 14-44. MODEL PLANT TREATMENT COSTS
Subcategory TITANIUM DIOXIDE Chloride-Il. Ore
Annual Treatment Costs ($/kkg)
LEVEL OF TREATMEOT
PRODUCTION FLOW
(kkg/yr) (mVaay)
FIRST
SECOND
TOIRD FOURTH
Annual Operation
and Maintenance
35,000
12,000
26.79
70,000
24,000
20.31
113,750
39,000
17.85
157,500
54,000
15.38
Annual
Amortization
35,000
12,000
6.55
70,000
24,000
4.10
113,750
39,000
3.40
157,500
54,000
2.90
Total Cost
35,000
12,000
33.34
70,000
24,000
24.41
113,750
39,000
21.24
157,500
54,000
18.27
Not Applicable
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 POR 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
regu1 at ions.
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.
Plow Basis
The BPT model plant flow rate is based on the reported
average process contact and clean up waste water flow at Plant
f237 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
N ickel
Lead
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 f559 (Table 14-30) indicates an achievable
long-term average of 21 mg/1 for TSS and 0.62 mq/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 \
V1000 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 proposed limitations 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 Ilrenite
Best Practicable Control Technology Currently Available
Waste Water Flew: 120 n3/kkg
Pollutant
Estimated
Treatability
(mg/1)
VFR
(1)
Concentration Basis
(mg/1)
fax
30-day
Avg
24-hr
max.
Effluent Limit
(kg/kkg)
Max
30-day
Avg
24-hr
max.
Conventional and
Nonconventional
Pollutants
Total Suspended
Solids
Iron
2i(2)
0.62(2)
3.6
3.4
63
2.5
230
8.5
7.6
0.30
27
1.0
Toxic Pollutants
Antimony (6)
0.80(3)
l.g(4)
0.80
1.5
0.096
0.18
Arsenic (6)
0.50(3)
1.9^)
0.50
0.95
0.060
0.11
Cadmium (6)
0.10(3)
1.6(4)
0.10
0.16
0.012
0.019
Chromium (6)
0.10(3)
1.9(4>
0.10
0.19
0.012
0.023
Copper (6)
0.50(3)
1.9(4)
0.50
0.95
0.060
0.11
Lead (6)
0.30(3>
1.5(4)
0.30
0.45
0.036
0.054
Nickel^)
0.20(3)
1.9^)
0.20
0.38
0.024
0.046
Zinc (6)
0.50(3)
2.1(4)
0.50
1.1
0.060
0.013
(1) - VFR: ratio of the 24-hr(daily) variability factor to the 30-day
average variability factor
(2) - Long term average from Plant s559 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 frcm Plant #559(Table 14-30)
(5) - Set equal to the VFR 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:
(0.80 mg/1) (1.20 m3/kkg)/ kg/m3 \
^ 1000 mg/1/
= 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. Cadmium: 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
V1000 mg/1/
= 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 \
Vj.00 0 mg/1/
= 0.012 kg/kkg
and application of the VFR 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.
F. Lead: The lower limit of treatability for lead is
estimated at 0.30 mg/1 as a 30-day average (Table P-ll) . Using
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/m.3 \
V]000 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 N
V000 mq/l)
= 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 \
VJL000 mg/1 J
= 0.060 kg/kkg
and the proposed zinc daily maximum is given by:
(2.1) (0.060 kg/kkg) = 0.13 kg/kkg
The proposed BPT limitations are presented in Table M-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 uset* 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 New 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.
Plow Basis
The reported data on process contact and clean-up waste
water flow at Plant f713 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 averaqe
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)/ kq/m3 \
\1000 mg/1/
= 1.2 kg/kkg
The proposed TSS daily maximum limitation is determined
by multiplying this value by the VFR 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
VFR for BPT.
The proposed NSPS limitations are presented in Table .14-46,
Nonconventional pollutants - The only nonconventional
pollu tant ol concern 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 Prooess Using Ilmenite
New Source Performance Standards*
Waste Water Flew: 32 rtr/kkg
, . . Estimated
Pollutant m , , .... (2)
Treatability v '
(mg/1)
Concentration
T7rr>(l) (mg/1)
Basis
Effluent Limit
(kg/kkg)
Max
30-day
Ava
24-hr
max.
Max
30-day
Avq
24-hr
max.
Conventional and
Nonconvent ional
Total suspended
Solids
40
3.6
40
140
1.2
4.3
Iron
1.6
3.4
1.6
5.4
0.050
0.17
Toxic Pollutants
Antimony^
0.80
1.9
0.80
1.5
0.025
0.048
Arsenic ^ ^
0.50
1.9
0.50
0.95
0.016
0.030
Cadmium (3)
0.075
1.6
0.075
0.12
0.0023
0.0037
Chromium (3)
0.040
1.9
0.040
0.076
0.0012
0.0023
Copper ^ ^
0.29
1.9
0.29
0.55
0.0090
0.017
Lead^-^
0.060
1.5
0.060
0.090
0.0019
0.0029
Nickel ( 3)
0.17
1.9
0.17
0.32
0.0053
0.010
Zinc(3)
0.47
2.1
0.47
0.99
0.015
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 N
V1000 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 *\
\T000mg7T/
= 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
1 imitation:
(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/m? \
U000 mg/1/
= 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 (41).
Thus, for the proposed NSPS maximum 30-dav average
limitations, the concentration basis is given by:
(1.00-0.60) (0.10 mg/1) = 0.040 mg/1
and the proposed NSPS limitation is:
(0.040 mg/1) (31 m3/kkg)/ kg/m3 \
V1000 mg/lj
= 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/1,
and the proposed limitation is:
(0.29 mg/1) (31 m3/kko)/ kg/m3 \ = 0.0090 kg/kkg
U000 mg/1/
The proposed NSPS daily maximum is then obtained by
multiplying the maximum 30-day average by the VFR 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 concer.tr at ion basis for the proposed NSPS maximum
30-day effluent limitation which is:
(0.060 mg/1) (31 m?/kkg)/ kg/m3 ^
V1000 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
(41). 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 i.9. That
is:
(1.9) (0.0053 kg/kkg) = 0.010 kg/kkg
and the daily maximum concentration basis is:
(1.9) (0.17 mq/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 N
V1000 mg/ly
= 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 14-45.
New Sources
Pretreatment Standards for New Sources (PSNS) are being
proposed by the Agency on the basis of NSPS treatment
technology for the Ti02-Chloride-Ilmenite industry. The
pollutants to be limited are iron and the toxic metals as
indicated in Table 14-46.
492
-------
SECTION 15
ALUMINUM PLUORIDE INDUSTRY
15.1 INDUSTRY PROFILE
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 HASTE 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
49 3
-------
TABLE IS-'. .
suBcarEOorcr profile data sutRffl
SUBCA35XC5W
AU^ONIM FLUORIDE
•total subcategory capacity rata
¦total subcategory production rate
Nianber 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:
Miniram
Maxinun
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maxinun
Waste water flew range:
Mininun
Maxinun
Vblune per unit product:
Minima
Maxinun
MA
134,"CO kkg/year
5*
6
204,800 Wcg/year
120,000 kkg/year
MA
MA
38 kXg/year
45,600 kkg/year
24 , 300 !ckg/year
35,500 kkg/year
59 peroent
5 years
21 years
539 cubic meters/day
2,200 cubic rneters/day
5 cubic neters/kkg
12 cubic neters/kkg
Sources of data are Stanford Research Insti tute/ Directory of Chonical
Producers, U.S.A., 1977, U.S. Department of Oxnneroe, Current Industrial
Reports, Deoaaber 1977; Biergy and EnvLicimental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessnent of effluent Limitations in the
Inorganic Chonical Industry." ~ur«, 197a arvd "Eaoncmic Analysis of Proposed
Revised Effluent Guioelo-Tes and Standards for the Inorgar-ic CJ-.enucais Industry,"
ttarcti, 1980.
NA » Not Available
• Seven plants were operating at the beginning of this sfjdy, but two closed do
production after 19~8.
494
-------
TABLE 15-2 .
STATUS OF REGULATIONS - EFFLUEOT LIMITATION GUIDELINES
SUBCATEGORY Aluminum Fluoride
SUBPART W (40 CFR 415.230, 5/22/75
STANDARDS
#
* *
Product Para- kg/kkg
Process meters
BPCTCA BA1EA NSPS
1 2
Max. Avg. Max. Avg. Max. Avg.
-------
WATER
VENT
MKSIC HATH?
NONCONTACT
COOLING WATER
IIYDRATED
ALUMINA
ALUMINUM
FLUORIDE
PRODUCT
HYDROGEN
FLUORIDE
REACTOR
PRODUCT
COLLECTION
AND STORAGE
COOLER
SCRUBBER
Figure 15-1. General process flew diagram for production
of aluminun 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 Hater
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 Bquipment Washings
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 acid, 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 SUBCATEGORY
Source Water use per unit of production
(ir.Vkkg of AIF3)
Plant
# 837
Plant
# 705 {Z)
Plant
* 188
Plant
« 251 (2)
Non-contact cooling
14.5
KA(1)
6.95
NA
Indirect process
contact (punps, seals,
leaks, spills)
12.2
1.15
MP.
NA
Maintenance, e.g.
cleaning and work area
washdcwn
1.13
2.39
NA
1.02
Scrubber
3.45
8.92
3.46
18.7
(1) NA - Not Available
(2) Currently not manufacturing aloninum fluoride.
-------
TABLE 15-4. WASTE WATER FLOW AT PLANTS #837, #705 AND #251
FOR ALUMINUM FLUORIDE SUBCATEGORY
Source
Flew rate per unit of production
( m^/kkg of AlF^)
(1)
Scrubber water
Maintenance equipment
cleaning and work area
washdown
Total raw waste flew
Plant #837
3.45
1.13
4.58
Plant #705
8.92(2)
2.39
11.3
(4)
Plant #251
18.7(3)
1.02
19.7
(4)
Average of above
three flews
11.9
(1) All flew information is frar, 308 Questionnaires and plant visits. Unit
flew is calculated by dividing waste water flow in rnp/day by production
in kkg/day.
(2) From Table 15-6 (see footnotes which, describe basis of information).
(3) Frcm 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 Ckg/kkg of AlF^)
#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
-------
OISTIUMION
KILM
ir
DRIP ACID
su»*ce nvuHS
COOLDC TCMK I
nc
•BmMJzxnON
GKrajNS raw
SETIUNS POND
loneus wwer
Figure 15-2. General process flew diagram at Plant #705 shewing the sanpling points
(aluninun fluoride manufacture).
-------
TABLE 15-6. FLOW AND POLLUTANT CONCFOTRATION DATA OF THE SAMPIED
WASTE STREAMS TOR PI ANT #705 PRODUCING ALUMINUM FLUORIDE
Sairplinq
Phase
Sanpled
Stream
No.
Sanpled
Stream
Description
Unit
Flow
(mVkkg)
Total
Suspended
Solids
(mg/1) ^ (kg/kkg) ^
Fluoride
(3) (4)
(mg/1) (kg/kkg)
Aluminum
(3) (4)
(mg/1) (kg/kkg)
Screening
3
AlF^ scrubber
8.92
13,000
120
530
4.7
780
7.0
4U)
Surface drains,
oooling tcwer,
blcwdown, etc.
2.39
200
0.48
350
0.82
40
0.10
3&4
Tbtal raw waste
load
11.3
(5)
11,000
120
490
5.5
620
7.1
5
(2)
Treated waste
24
80
2.0
70
1.6
10
0.17
Verifica-
tion Sanpling
3
4<1)
AJLF3 scrubber
Surface drains,
oooling tewer,
blowdown, etc.
8.92
2.39
1,400
200
13
0.48
1400
170
12
0.40
460
27
4.1
0.060
3&4
Tbtal load
11.3
1,200(5)
13
1100
13
370
4.1
5
Treated waste ^
24
2.0
0.048
20
0.55
1.0
0.012
(1) Consists of waste water from HF and AlF-, process. Flow indicated is estimated portion of total
flow contributed by AIF3 naintenance ana wash down waste water from 308 Questionnaire.. Tt>tal flow
is 17.8 mVkkg of product for both process wastes combined.
(2) Consists of waste water from HF and AIF3 process. Plant currently not manufacturing AIF3.
(3) Average of three daily oonposibe sairples during verification and single value obtained during
screening.
(4) kg/kkg of AlF-^. (5) Weighted average based on unit flews.
-------
VENT
DUST
COLLECTOR
H2S°i,
WET
spar'
SPAR DRYING
HANDLING
LOSSES
HOSE DOWN
WATER
HF KILN
WASTE
AIR
ai2o3' 3h2o
COOLER
DRIP
ACID
-WATER
SLURRY
TRANSFER
e
LEGEND
SAMPLING POINTS.
A1F.J
PLANT
REACTOR
A1F? PRODUCT
VENT
__L
S02 SCRUBBER
CaTf
WATER
3
SCRUBBER
LIQUEFACTION
ON W
tth _
t?
AHF
PURIFICATION
AHF
PRODUCT
DILUTION
WATER
aif3 PLANT
HOSE DOWN
HOSE DOWN WATER
AHF PLANT
H2
GYPSUM
POND
NEUTRALIZATION
SYSTEM
h
WATER
#6
effluent
TO RIVER
ALKALINE STREAMS 1
AND ACID FROM OTHER PLANTS
Figure 15-3. General process flow diagram at Plant ^251 showing the sampling points,
(aluminum fluoride manufacture).
-------
TABLE 15-7. FI£W AND POLLUTAOT OONCEOTRATION DATA OF niE SAMPLED STREAMS
FOR PLANT #251 PRODUCING ALUMINUM FLUORIDE
Stream
No.
Sanpled
Stream
Description
Unit
Flow
(mVkkg
of A1F3)
Total
Suspended
Solids
(mg/1) (kg/kkg)
Fluoride
(mg/1) (kg/kkg)
Aluminun
(mg/1) (kg/kkg)
Verification
Sampling
4
AlF^ scrubber
water
12.6
1200
16
470
5.90
50
0.60
6
S0_ scrubber
water ^ ^
6.10
0.0
0.0
20
0.14
0.20
0.0010
48,6
Total raw waste
load
18.7
1200
16
320
6.0
50
0.60
2
Gypsun pond
influent
25.1
19,000
470
660
17
26
0.65
3
Gypsum pond
effluent* '
25.1
9.0
0.23
320
8.0
22
0.55
(1) One half flew of SO2 scrubber water is assumed to contribute to the AIF3 process since the
total flew is oorrmon to the AIF3 and HF process.
(2) Consists of hydrofluoric acid and aluninun fluoride waste water. Plant currently not
manufacturing AlF^.
-------
Maximum Raw Waste Concentrations Observed
(>jg/l)
Pollutant Screening Verification
Plant f705 Plant #705 and #251
Arsenic
200
480
Selenium
68
97
Chromium
70
1100
Copper
120
250
Lead
25
91
Mercury
1.6
11
N ickel
150
290
Z inc
450
450
Cadmium
0.70
33
Antimony
0
3
Beryllium
0.80
0
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 f0)
per kkg of aluminum fluoride) = -—-—5-^-
1000P
Where C and 0 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 lbs.)
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:
Pollu tant
Waste Load
(kg/year)
Arsenic
180
Selenium
140
Chromium
400
Copper
94
Lead
20
Mercury
3.
0
N ickel
180
Z inc
140
Cadmium
11
Antimony
0.
70
Beryllium
0.
30
506
-------
TABLE 15-8. TOXIC POLL17EAOT AVERAGE RAW WASTE LOADS AND CONCENTRATIONS
SUBCATEGORY ALUMINUM FLUORIDE
Pollutant
Screening
Verification
Average
Concentration
(mg/1)
Plant #705
(mg/1)(1) (kg/kkg)(2)
Plant
(mg/1)
#705
(kg/kkg)
Plant
(mg/1)
#251
(kg/kkg)
Arsenic
0.18
0.0020
0.18
0.0020
0.020
0.00030
0.13
Seleniun
0.050
0.0010
__ (3)
__(3>
0.050
0.0010
0.050
Chromium
0.030
0.00030
0.44
0.0050
_<3)
__(3)
0.24
Copper
0.10
0.0010
0.070
0.0010
0.010
0.00010
0.060
Lead
0.0050
0.00010
0.020
0.00020
0.010
0.00010
0.012
Mercury
0.00040
0.0000040
0.00040
0.0000050
0.0030
0.000050
0.0013
Nickel
0.11
0.0010
0.22
0.0030
0.010
0.00020
0.11
Zinc
0.16
0.0020
0.080
0.0010
0.020
0.00030
0.090
Cadmium
0.00020
0.0000020
0.010
0.00020
__<3>
__(3)
0.0050
Antimony
_<3>
__<3>
0.00040
0.0000050
__<3)
__<3)
0.00040
Beryllium
0.00020
0.0000020
__<3>
__<3>
_<3>
__<3>
0.00020
(1) Concentrations based on average raw waste loads shown and total process production and waste
flows.
(2) kg/kkg of product.
(3) — belcv/ analytical detection limit.
-------
TABLE 15-9. TOXIC POLLUTANT EFFLUENT CONCENTRATIONS DURING SAMPLING
SUBCATEGORY ALUMINUM FLUORIDE
Pollutant
Plant and Sanpling
Phase
#705
#705
#251
Screening
Verification
Verification
Average
(mg/1)
(mg/1)
(mg/1)
(mg/1)
Arsenic
ND11*
ND
0.0050
< 0.0050
Selenium
ND
ND
0.070
< 0.070
Chromium
0.0070
0.040
0.22
0.090
Copper
0.10
0.0010
0.070
0.060
Lead
0.0020
0.020
0.030
0.020
Mercury
ND
ND
ND
ND
Nickel
0.050
ND
0.45
< 0.25
Zinc
0.0020
0.0010
ND
0.0020
Cadmium
0.0020
0.0010
ND
< 0.0020
Antimony
ND
ND
ND
ND
Berylliun
0.0020
ND
ND
< 0.0020
(1) ND — Not Detected.
508
-------
TABLE 15-10. SUMMARY OF RAW WASTE LOADINGS POUND IN SCREENING AND VERIFICATION SAMPLING
SUBCATEGORY ALUMINUM FLUORIDE
Pollutant
Loading Range,
kg/day
Minimum Maximun
Minimum
Unit Loading,
Average^
kg/kkg
Maximun
No. Of
Plants
Averaged
Tbxic
Arsenic
0.050
0.080
0.00030
0.0013
0.0020
3
Selenium
0.030
0.16
0.0010
0.0010
0.0010
2
Chromium
0.020
0.22
0.00030
0.0030
0.0050
2
Copper
0.020
0.050
0.00010
0.00070
0.0010
3
Lead
0.0030
0.020
0.00010
0.00015
0.00020
3
Mercury
0.026
0.0080
0.0000040
0.000020
0.000050
3
Nickel
0.026
0.12
0.00020
0.0013
0.0030
3
Zinc
0.040
0.080
0.00030
0.0010
0.0020
3
Cadmium
0.00010
0.0070
0.0000020
0.000080
0.00020
2
Antimony
NA<2>
0.00020
NA
0.0000050
NA
1
Beryllium
NA
0.00010
NA
0.0000020
NA
1
Conventional and
Nonconventional
TSS
600
5400
13
50
119.0
3
Fluorine
250
980
5.5
00
•
13.0
3
Aluminum
100
320
0.60
3.9
7.0
3
(1) Avcracjc 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 Treataent 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 f837 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 fl88 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
to 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 SBLECTION OP 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 TO 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
CaF2 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 Treataent 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
-------
0
LDC
—+i2>
(g)- -
I
A
jA
LMXXM
H r
HASTE WATRR _ V
J-+
counuzxricM
I*
Kixnc
uaxH
jj
fH UDMSOOtT
(£)
>0
emxAxr
•Includes flat nvltorlnj, pM nnitorlng and w%4er.
Figure 15-4. Level 1 waste water treatment for the aluminum fluoride subcategory.
-------
BM30A9I
r
-©-—-i
MDUJC
RMj
»MTO
BOUkLIZJOlON
*Inclula* flow mltorlng, |M mltDrlng and aarpler.
Figure 15-5. Level 2 waste water treatment for the aluminum fluoride subcategory
-------
IDWUS MI/XTf
icon* aisuLricr
e
cA>~
=siliil]
Ul
QN.tttrSM
*Wlu4rI now pl< rtnf a«4
h
n
t«
JtftABMXr
(i)
1
.....
I
1
4.
Figure 15-6. Level 3 waste water treatment for the aluninun fluoride subcategory.
-------
0^1
] r
^ I
lUfUl I
H3-1
~
OMtrtm
j
-Q-
tm HiMj—m*
v,,
i
i
-i -
$
>r
(j-
I»rfw4** f)e* menltvpH ifMnltnilitf ind »myl»
Figure 15-7. Level 4 waste water treatment for the aluninim fluoride subcategory.
-------
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 3, 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 Hater 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 A1F3
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-A!F3)
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 Chenicals
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 (35,600 kkg-AlF3/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 PIAVT TREATMENT COSTS
Subcategory ALIMINIM FLUORIDE
(1) (I)
Production 15,900 metric tons par year (17,529 tons per year)
45 metric tons per day (50 sons per day)
Wa3te water flow 540 cuoic oeters per day.
(2)
LEVEL OP TREATMENT
FIRST
\. INVESTMENT COST
Construction
Squlpnent in place,
including piping,
fittings, electrical
wor* and controls
Monitoring equipment
in place
Engineering design
and Inspection
Incidentals, overhead,
fees, contingencies...
Land
TOTAL IWSSTMDfT COST
9. OPERATION \ND
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
rNvss-wnrr cost
TOTAL AftJIAL COST
(1) Production year Is 353
(2) First level represents
Other levels represent
192.330
9,203
48,163
48,163
24,000
S56.000
3,400
35,000
33,712
10,833
5,400
15,000
5159,345
S54.949
5214,194
SECOND
THIRD
S39.800 $19,333 $14,300
$14,300
630
10,923
3,276
7,503
$36,296
$17,766
$54,362
$14,300
900
830
12,320
3,696
7,500
$39,216
S20,044
FOURTH
S20,500
69,003 74,003 172,000
15,600 17,600 38,500
15,600 17,600 38,530
$361,120 $109,200 $123,200 $269,530
S14.000
2,500
9,800
2S.950
8,085
7,500
$68,835
$43,847
$59,260 $112,682
days.
the base cost of treatment system,
the incrmental cost above base cost.
520
-------
TABLE 15-12. MODEL ?'_AMT T3EATMDJT COSTS
Subcategory ALIMINIM FLlDRlDE
(I) (1)
Production 35,633 metric tons per year 139,249 tons per year;
131 metric tons per day C112 tor.s per day;
Waste water flow 12C0 cubic meters per day.
(2)
LIT/EL OF TREATMENT
A. INVESTMENT COST
Construction
Equipment in place.
Including piping,
fittings, electrical
\*ork and controls
Monitoring ecfuianent
in place
Engineering design
and Inspection
Incidentals, overhead,
fees, contingencies...
Land
TOTAL INVESTMENT COST
3. 3f*SATl3N WTO
MAINTENANCE COST
Labor and supervision.
Energy
Chemicals
Maintenance
Taxes and insurance...
Residual waste
disposal
Monitoring, analysis
and reporting
TOTAL OPERATION AND
MAIN rENANCS COST
C. AMORTISATION OF
INVESTMENT COST
TOTAL \.*JiAL COST
(1) Production year is 353
(2) First ievel represents
Other ievels represent
?L*ST SECOND
563,503 $15,300
238,303
9,000
52,120
62,120
42,300
$476,840
$56,003
5,500
30,000
43,484
14,305
12,500
15,003
$226,789
$70,748
$297,537
34,300
'.9,333
.9,333
THI3D
$19,333
93,533
21,900
21,903
$14,330
933
13,363
4,158
$14,333
1,333
1,803
15,333
4,599
',533 7,533
$40,413 $44,529
$22,553
$62,968
$24,941
FOURTH
$34,033
259,000
53,603
53,600
$138,633 $153,303 $413,230
$14,003
3,133
13,330
41,020
12,306
7,533
$96,72e
$66,739
$69,473 $163,465
days.
the bas« cost of treatment system,
the incremental cost above base cost.
521
-------
T\BL£ 15-13. ¦^CTX.L PL\ST TRSA-MEJJT COSTS
Subcategory ALIMIN'JI FLICRIOE
tn in
Production 45,803 metric tons per year (50,494 torj per year)
133 metric tons per day ;144 tens per day)
Waste water flow 155C cubic meters per day.
!2>
LEVEL OF TREATMENT
FIRST
SECOND
THIRD
FOURTH
INVESTMENT COST
Construction
$75,500
$23,503
S24.533
$43,330
Equipment in place,
incljding piping,
fittings, electrical
wor* and controls
281,330
110,000
116,530
317,333
Monitoring equipment
in place
9,230
Engineering design
and inspection
73,333
26,130
29,230
72,000
Incidentals, overhead,
fees, contingencies...
73,333
26,100
28,203
72,300
Land
63,000
TOTM. INVESTMENT COST
5573,100
$192,703
SI 97,400
S534,003
OPERATION V©
MAINTENAJCE COST
Labor and supervision.
$56,333
$14,000
$14,333
$14,300
Energy
7,433
'.,500
1,903
4,300
Chemicals
133,333
2,403
26,400
Maintenance
51,313
19,273
19,743
50,430
Taxes and Insurance...
17,193
5,481
5,922
15,123
Residual waste
disposal
16,000
Monitoring, analysts
and reporting
15,000
7,503
7,530
7,503
TOTAL OPERATION AfTO
MAINTENANCE COST
$262,933
$16,751
$51,462
$117,723
AMORTIZATION OF
INVESTMENT COST
S83,481
$29,725
$32,116
$92,330
TOTAL V*4!AL COST
$346,384
$76,475
S93,579
$199,720
(1) Production year is 350 days.
(2) First level represents the base cost of treatment system.
Other ievels represent the Incremental cost above base cost.
522
-------
I
I
X
-------
20
U
15
>
1
1
1 i
.
i 1 '
:
¦
¦ ¦ i
1
i
1
! 1 ¦ ¦ ; i :
1 .
i
!
1
1 !
i 1 ;
1
;
¦
1
' 1
1 '
.
;
!
1 i
, •
¦ i
\
1 : : : 1
i i
. \
1
i j
!
1 !
! A
i
I ... 1
• ;
: i
1 ! i 1 1
1
I®1 ^
. , i
1 ! '
. : ; 1 i . 1 i
i
: i 1 ! 1 : 1
i V
X
i
1 ¦ : ¦ ! :
i
1
1
1 . i
i
,
¦ 1
'
1 i i
•
NX i
!
,
'NN.
! ¦
; : i \
I !nN
iSl
'
1
! 1 ! 1 !
k ' \\
i
i i
1
! ;
i :
V! \
>
< i
'T 1
rVFr. *4 =
1
,
I X
i X
,V*J !
1 j
1
' , !
i
1 ¦ v
; j
i
1
! ' x
.1 :
^ . Ll
rvzu * 3 :
1
: i
. i
-------
TABLE 15-14. MODEL PtAWT TREATMENT COSTS*31
Subcategory ALUMINiw FLUORI0€
Production
Waste water (low
(U
35,600 metric tons per year
181 metric tons per day
1200 cubic meters per day.
(I)
(39,249 tons per year)
(112 tars per day)
(2)
A. INVESTMENT COST
Construction
Equlpnent In place,
Including piping,
fittings, electrical
work and controls
Monitoring equlpnent
In place..............
Engineering design
and inspection
Incidentals, overhead,
fees, contingencies...
Land
TOTAL INVESTMENT COST
B. OPERATION KJC
MALNTENVCE COST
Labor and supervision.
Energy
Chemicals
maintenance
Taxes and Insurance...
Residual waste
disposal
Monitoring, analysis
and reporting
TOTKL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION Of
INVESTMENT COST
TOTAL ANNLfcL COST
(1) Production year Is 3S0
(2) First level represents
Other levels represent
(3) Sensitivity Analysis -
LEVEL OF TREATMENT
FIRST SECOND THIRD
$82,300 $15,333 $19,030
84,333 93.S00
19,933 21,903
19,303 21,900
$530,833 $138,603 $153,300
241,000
9,300
66,400
66,430
66,330
$S6,333
5,500
130,003
46,480
15,924
19,000
15,000
$287,904
$75,622
$363,526
$14,303
933
13,863
4,158
7,503
$40,418
$22,550
$62,968
$14,333
1,303
1,833
15,333
4,599
7,533
FOURTH
$34,533
270,333
60,933
63,933
$426,333
$14,330
3,130
31,500
42,630
12,789
7,500
$44,529 $111,519
$24,941 $69,359
$69,473 $183,878
days.
the base cost of treaoaent system,
the Incremental cost above base cost,
increased pollutant load.
525
-------
:\3LE 15-15. "MOOSL PLANT TREATMENT COSTS(3)
Subcategory ALIWINUM FLUORIDE
(I) (1)
Prodvetlon 35,603 metric tons per year (39,249 tons per year)
101 metric tons per day (112 tors per day)
Waste water flow 1200 cubic meters per day.
(2)
LEVEL 0* TREATMENT
FIRST SECOND THIRD
A. INVESTMENT COST
Construction $56,933 $15,333 519,330
Equipment In place,
Incljdlng piping,
fittings, electrical
wor* and controls 221,003 34,333 90,533
¦Monitoring ecjilpment
In place 9,003
Engineering design
and inspection 57,383 19,933 21,933
tncidentals, overhead,
fees, contingencies... 57,380 1.9,300 21,933
Land 33,300
TOTAL INVESTMENT COST $431,663 $138,603 $153,333
3. OPERATION AND
•MAINTENAJCS COST
Labor and supervision. $56,333 $14,333 $14,003
Energy 5,503 933 1,333
Chemicals 60,333 1,833
•Maintenance 43,166 13,863 15,333
Taxes and insurance... 12,949 4,158 4,599
Residual waste
disposal 9,333
'Monitoring, analysis
and reporting 15,303 7,503 7,500
TOTAL OPERATION AND
•MAINTENANCE COST $199,615 $43,418 $44,529
C. AMORTISATION OP
INVESTMENT COST $65,350 $22,553 $24,941
TOTAL VNNIAL COST $263,96S $62,963 $69,473
(1) Production year is 353 days.
(2) Fist ievel represents the base cost of treatment system.
Other levels represent the Incremental cost above base cost.
(3) Sensitivity Analysis - decreased pollutant load.
FOURTH
$34,333
259,333
58,633
58,633
$413,203
$14,333
3,133
14,613
41,323
12,336
7,533
$92,536
$66,739
$159,275
526
-------
j
1
j
1
1
1
| |
1
I
, <
1
¦
1 ,
,
1
:
1
1
i •
I
1
:
l
1
i
1
j
1
j
1
i
1
1
1
'
\
I
,
\
:
i
' >
i i
1
\
' ' \
1
^TKrBFAZFT
T)T,T
1TAHT
T
y\
n
1
.
s
v ¦ / 1
v
s
Nl tf i
1
N
\ l
i
I
X
1
i
1
1
¦'
wN! ¦
1
j
! ¦
i
-1
>1 ; V
1
I
1 , 1 ^
1 i
1 ; i " ¦
1
1 i . ! ' •
!
1
np
TRF3
^'1
FD pnr.Tirf
"ant rn5,n
i
1
i
i
I , 1
i
• 1
i
1
i
1 , !
1 .
i ,
! !
' ! .
¦ ;
!
,
:
: i i
¦
i ¦ !
.
• 1
i
30
ALF3 production (metric TONS/YEAR X 1000)
Figure 15-10. Effect of variation of pollutant load on treatment
cost at level 1 technology
527
-------
TOTPF&3r.n pot .tjitant lq^d
A.W? PTflRrWT van
10 20 30 40 50 60
A1F-, PTOUCTICN (METRIC TCNS/YEAR X 1000)
Figure 15-11. Elf feet 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
-------
TA3LS 15-16. MCCEL PLA.VT TREATMENT COSTS!3)
Subcategory ALLMINLM FLUORIDE
!U C.)
Production 35,633 i>etric tons per /ear ( 39,249 tens per vear:
131 .netric tons pec day (112 tons per iav;
Waste water flow 23CO cubic .Deters per day.
(2)
LEVEL OF TREATMENT
FIRST
SECOND
THIRD
FOURTH
INVESTMENT COST
Construction
$66,133
$21,333
$25,333
$43,533
Equipment in place.
including piping.
fittings, electrical
¦work and controls
256,333
117,633
124,333
321,333
Monitoring equlpnent
in olace
9,333
Engineering design
and inspection
66,223
27,723
29,330
72,933
Incidentals, overhead,
fees, contingencies...
66,223
27,723
29,933
72,933
Land
42,333
TOTAL IWESTMfNT CXT
5535,543
$194,343
5239,633
$513,333
OPERATION AND
MAINTENANCE CXT
Labor and supervision.
$56,333
$14,333
514,333
514,333
Energy
7,433
1,533
1,933
4,733
Chemicals
83,333
, 833
13,933
Maintenance
46,354
19,434
23,863
51,333
Taxes and Insurance...
15.166
5,821
6,258
15,339
Residual waste
disposal
12,533
Monitoring, analysis
and reporting
15,333
7,533
7,533
7,533
TOTAL OPERATION AND
MAINTENANCE COST
AMORTIZATION OF
INViS'TiOJT COST
TOTAL, VVUIPKL COST
5232,423
$75,417
5337,837
S48,225
S31,573
579,795
552,319
$33,939
(1) Production year is 353
(2) FlC3t ievel represents
Other ievels represent
(3) Sensitivity Analysis -
days.
trie base cost of treatment system,
the increments! cost above base cost,
increased hydraulic load.
$111,339
S33.325
$85,257 $194,364
530
-------
TVJLE 15-17. MODE!. PIANT TREATIES COSTS
(3)
Subcategory ALIMINIM FLUORIDE
(I) (I)
Production 35,600 i»etrlc torts per y*ar ( 39,249 tons per year)
101 ratrlc tons per day (112 tons per day)
Waste water flow 1020 cubic meters per day.
(2)
LEVEL. OF TREATMENT
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
8. 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
DJVESTiEJT COST
TOTAL ANN UK L COST
(1) Production year Is 353
(2) First level represents
Other levels represent
(3) Sensitivity Analysis -
'TP ST
237,003
3,300
61,920
61,920
42,333
$56,300
5,533
83,333
43,344
14,263
12,500
15,300
S226,537
S73,523
$297,127
SECOND
563,633 514,503
$14,330
633
11,872
3,561
7,530
537,533
THIRD
S13,500
73,300 76,300
16,963 18,900
16,963 18,900
5475,443 $118,723 S132.300
514,033
903
1,803
13,230
3,969
7,503
541,399
519,315 $21,525
$56,848 562,924
FOURTH
$30,300
236,333
47,200
47,200
S330.430
$14,000
2,530
13,800
33,343
9,912
7,530
585,752
553,756
SL39 , 509
days.
the base cost of treatment system,
the Incremental cost above base cost,
decreased hydraulic load.
531
-------
20
»»¦ i ft
20 30 40 SO
AlF, PRODUCT!CN (METRIC TONS/YEAR X 1000)
Figure 15-12. Effect of variation of hydraulic load on treatment
cost at level 2 technology
532
-------
EASED-,
HYpRMJI.
DB23
Figure
10 20 30 40 50 60
A1F3 PRCDUCTICN (METRIC TONS/YEAR X 1000)
15-13. Effect of variation of hydraulic load on treatment
cost at level 3 technology
533
-------
20
U
v>
15
10
¦asfd KvnRAi
I
SEH-HYD
10 20 30 40 SO 60
aif3 PRODUCTION (METRIC TONS/YEAR X 1000)
Figure 15-14. Effect of variation of hydraulic load on treatment
cost at level 4 technology
534
-------
TABLE 15-18. *10DEL PLAOT TREATMENT COSTS
Subcategory ALIMINLM FLUORIDE
Annual Treatment Costs/Metric Ton of Product
'OST ITEMS PRODUCTION FLOW
(kkg/yr) (m3/day)
LEVEL OF TREATMENT
FIRST SECOND THIRD FOURTH
Annual Operation
and Maintenance
Annual
Amortization
Totai Cost
15,900
690
10.02
2.28
2.47
4.33
35,500
1,550
6.37
1.14
1.25
2.72
45,800
1,990
5.74
1.32
1.12
2.57
a
35,600
1,550
8.09
1.14
1.25
3.13
b
35,630
1,550
5.58
1.14
1.25
2.60
c
35,600
2,203
6.53
1.35
1.47
3.13
d
35,630
1,064
6.37
1.05
1.16
2.41
15,900
690
3.45
1.12
1.26
2.76
35,600
1,550
1.99
0.63
0.70
1.87
45,800
1,990
1.82
0.65
0.70
1.79
a
35,600
1,550
2.12
0.63
0.70
1.95
b
35,600
1,550
1.84
0.63
0.70
1.87
c
35,600
2,203
2.12
0.89
0.95
2.33
d
35,600
1,064
1.98
0.54
0.60
1.51
15,900
690
13.47
3.40
3.73
7.09
35,600
1,550
8.36
1.77
1.95
4.59
45,800
1,990
7.56
1.67
1.82
4.36
a
35,600
1,550
13.21
1.77
1.95
5.08
b
35,600
1,550
7.41
1.77
1.95
4.47
c
35,600
2,203
8.65
2.24
2.42
5.46
d
35,600
1,064
8.35
1.60
1.77
3.92
a Increased poliutant 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,
Plow 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 f837, f251,
and f705. Plant fl88 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 ]1.9 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 #705 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 Linitations
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 ^7 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) = (0) (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
\ 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 nO/kkg)/' kg/m3 \ = 0.63 kg/kkg
V 1665 mg/1 )
The 24-hour maximum unit loading is determined by
multiplying 2.1 times the 30-day average unit loading determined
above.
Tox ic 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 HF/A1F3 industries. The unit load limitation was
calculated as follows:
(0.10 mg/1) (11.9 m3/kkg)/ kg/m3 \ = 0.0012 kg/kkg
VlOOO 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 = 3.12
VF of 30-aay 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 #705 (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 Flow: 11.9 m3Akg
Concentration Basis Effluent Limit
Subcategory (1) (mg/ll (kg/kkg)
Pollutant
Performance
(mg/i )
VFR
30-day
Avg
24-hr
Max
30-day
Avg
24-hr
Max
Conventional and
Nonconventional
Pollutants:
(2)
Total Suspended
97
2.1
97
200
1.2
2.4
Solids, TSS
(2)
(4)
Fluoride
53
2.1
53
110
0.63
1.3
Toxic Pollutants:
(3)
(5)
(6)
(6)
Arsenic
0.50
2.0
0.50
1.0
—
—
(3)
(5)
Chromium
0.10
2.0
0.10
0.20
0.0012
0.0024
(3)
(5)
(6)
(6)
Copper
0.50
2.0
0.50
1.0
—
—
(3)
(5)
Nickel
0.20
2.0
0.20
0.40
0.0024
0.00 48
(3)
(5)
(6)
(6)
Selenium
0.20
2.0
0.20
0.40
—
—
(3)
(5)
(6)
(6)
Zinc
0.50
2.0
0.50
0.50
—
—
(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
HF subcategory regulation (Section 12.7.2).
(3) - The lower limit of the literature treatability estimate
(Table 9-11) is used as the basis for the maximum 30-day
average limitation and subcategory performance, since no plant
is available where BPT treatment can be evaluated for the A1F3
waste water alone.
(4) - VFR based on HF subcategory evaluation.
(5) - VFR 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
regulat ions.
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).
Plow 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 Linitations
Nonconventional pollutants - The only nonconventional
pollutant selected is fluor ide. 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 VFR 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_N = 0.36 kg/kkg
V 1000 mg/1J
The 24-hour maximum limitation is determined in a similar
manner as follows:
(63 mq/1) (11.9 m3/kkg) / kg/m3 = 0.75 kg/kkg
\ 1000 mg/1 J
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 Technology
W&ste Water Flow: 11.9 m^/kkg
Pollutant
Subcategory
Performance
(nc/1)
Concentration Basis
(1) (mg/1)
VFR
Max
30-day
Ava
24-hr
Max
Effluent Limit
(kg/kkg)
Max
30-day
ftW
24-hr
Max
Nonconventional
Pollutants:
Fluoride
30(3)
2.1
30
63
0.36
Toxic
Pollutants:
Arsenic
0.50(5)
2.0
0.50
1.0
_J4>
(2)
Chromium
0.04 (d)
2.0
0.04
0.08
0.00048
Copper
0.29(5)
2.0
0.29
0.58
__<4>
Nickel^
0.17(5)
2.0
0.17
0.34
0.0020
Selenium
0.18(5)
2.0
0.18
0.36
-J4*
Zinc
0.47(5)
2.0
0.47
0.94
_J4>
0.75
_J4>
0.00096
_J4>
0.0040
_J4>
(4)
(1) - VFR: 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 frcn the HF subcategory Table 12-21 and
12-25.
(4) - No effluent limitation proposed.
(5) - Literature treatability estimates.
544
-------
TABLE 15-21. PERFORMANCE OF ALTERNATIVE TECHNOLOGY
Aluminum Fluoride
Level of Treatment: 3
W&ste Vfater Flow: 11/9 mVkkg
Pollutant
Treatability
(rog/1)
(1)
Concentration Basis
(ng/1)
VFR
Max
30-day
—Awg—
24- hr
Max
Nonoonvent iona1
Pollutants:
Fluoride
25
3.0
25
75
Toxic
Pollutants:
Arsenic
0.050
2.0
0.050
0.10
Chrariium
0.040
2.0
0.040
0.080
Copper
0.050
2.0
0.050
0.10
Nickel
0.10
2.0
0.10
0.20
Selenium
0.18
2.0
0.18
0.36
Zinc
0.20
2.0
0.20
0.40
(1) - VFR: ratio of the 24-hour variability factor to the 30-day variability
factor.
545
-------
TABLE 15-22. PERFORMANCE OF ALTERNATIVE TECHNOLOGY
Aluminum Fluoride
Level of Treatment: 4
W&ste W&ter Flow: 2.4 mVkkg (80% Recycle)
Pollutant
Treatability
(n>g/l)
VFR(1)
Concentration Basis
(mg/1)
Max
30-day
Avg
24-hr
Max
Nonconvent ional
Pollutants:
Fluoride
30
2.1
30
63
Toxic
Pollutants:
Arsenic
0.50
2.0
0.50
1.0
Chrcmiim
0.04
2.0
0.04
0.08
Copper
0.29
2.0
0.29
0.58
Nickel
0.17
2.0
0.17
0.34
Selenium
0.18
2.0
0.18
0.36
Zinc
0.47
2.0
0.47
0.94
(1) - VFR: 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 mVkkg = 2.4 mVkkg) •
546
-------
industry, the literature treatability studies cited in Section
12.7.4 under "Toxic Pollutants" for the HF 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:
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 VFR 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 iffo J
Application of the BAT model plant discharge rate gives the
following load limitation for nickel:
0.10 mg/1
0.040 mg/1
547
-------
(0.17 mg/1) (11.9 m3/kkg)^ kg/m3 = 0.0020 kg/kkg
^ 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 Hew Source Performance Standards
Technology Basis
For NSPS, the Agency proposes the same treatment technology
that is proposed for BAT.
Plow 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 JRB 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) f kg/m3 *\ = 0.81 kq/kkq
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 Pretreatnent 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
-------
TABLE 15-23. PROPRDSED LIMITATIONS
Aluminum Fluoride
New Source Performance Standards
waste water Flow: 11.9 mVkkg
Pollutant
Treatability
(mg/1)
Concentration Basis
(1) (mg/1)
VFR
Max
30-day 24-hr
Avg Max
Effluent Limit
(kg/kkg)
Max
30-day
Avg
24 -hr
Conventional and
Nonconvent ional
Pollutants:
Total Suspended
Solids, TSS
68(2)
2.1
68
140
0.81
Fluoride, F (5)
30(2)
2.1
30
63
0.36
Toxic
Pollutants:
Arsenic
0.50(3)
2.0
0.50
1.0
_J4>
Chrotniun (5)
0.04(3)
2.0
0.04
0.08
0.00050
Copper
0.29(3)
2.0
0.29
0.58
(4)
Nickel (5)
0.17(3)
2.0
0.17
0.34
0.0020
Selenium
0.18(3)
2.0
0.18
0.36
_J4>
Zinc
0.47(3)
2.0
0.47
0.94
_J4)
1.7
0.75
_(4)
0.0010
_J4)
0.0040
_J4>
(4)
(1) - VFR: ratio of the 24-hour variability factor to the 30-day variability
factor.
(2) - 30-day average calculated frcm the HF subcategory Table 12-29.
(3) - Literature treatability estimates from BAT 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, molvbdate 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.
Chrcaium 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 + Na2S04 (1)
551
-------
TOLE 16-1. SUBCAIBOOre' PSCTHE DATA StMWBf
(1)
SUBCATEGORY
CHPDME PIOEOTS
Tbtal subcategory capacity rate
Total subcategory production rate
Nisnber of plants in this subcategory
308 Data cn file for
with total capacity of
With total production of
Representing capacity
Representing production
Plant production range:'2)
Minimis
Maxima
Average t»v»fc>ct'inn
Median production
Average capacity utilization
Plant age range:
Minima
Maxima
Wastewater flow range:
Minimn
Maxima
Volune per unit product:
Minima
Maxima
63,000 kkg/year
64,500 kkg/year
12
5
39,800 kkg/year
62 percent
100 kkg/year
18,COO kkg/year
6,300 kkg/year
6,400 kkg/year
78 percent
38 years
60 years
800 cubic meters/day
11,363 cubic nwters/day
32 cubic raeters/kkg
170 cubic meters/kkg
(1) Sources of data are Stanford Research Institute, Directory of Chenical
Producers, U.S.A., 1977, O.S. Department of Oonnerce, Current Industrial
Reports, Ocoanber 1977; Energy and Environnental Analysis, Inc.; Draft
Report, "Preliminary Eoonanic Assesonent of Effluent Limitations in the
Inorganic Chemical Industry," June, 1978 and "Economic Analysis of Proposer!
Pevised Effluent Guidelines and Standards for the Inorganic Oierucals Industry,"
March, 1980.
!2) 3ased on production at 11 plants, all other figvaws are based on 308
Ques tionnaires.
552
-------
TA3LE 16-2. STATUS OF REGULATIONS - EFFLUENT UMTTATICN GUIDELINES
SDBCATEQOTOf
SUBPART
Chrome Pigments
AH (40CFR 415.340, 5/22/75)
STANDARDS
BPCTCA*
baiea
NSPS
Product
Process
Max.
Para- kg/kkg
meters (mg/1)
Avg.
kg/kkg
(mg/1)
Max. Avg.
kg/kkq kg/kka
(mg/1) (mg/f)
Max. Avg.
ka/kkg kq/kkg
(mg/1) Wl)
Chrome
Pigment
TSS
Cr (T)
Cr*6
Pb
Zn
CN
CN(A)
Fe
5.1
1.7
(76.1)*
(25.4)
0.10
0.034
(1.5)
(0.5)
0.010
0.0034
(0.2)
(0.1)
0.42
0.14
(6.3)
(2.1)
0.72
0.27
(10.8)
(4.0)
0.010
0.0034
(1.5)
(0.5)
0.10
0.034
(0.2)
(0.1)
0.72
0.27
(10.8)
(4.0)
Reserved
Reserved
*
Sections 415.340, 415.341, and 415.342 were revoked by the Agency
£41 FR 51601, November 23, 1976)..
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 reducinq agent eliminates C02 and CO
emissions but increases the sulfates in the raw waste as well as
producing S02 and S03 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 Cr20(0H)4, also
known as chromium hydrate and Guiqets Green, is a brilliant
bluish green. It is made by reacting sodium dichromate with
boric acid as follows:
2Na2Cr207 + 8H3B03 = 2Cr203.2H2O + 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 + H2S04 + 5H20 = 4H3B03 + Na2S04 f4)
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.
Chrone 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 as:
2HN03 + PbO = Pb(N03)2 + H20 (5)
Na2Cr207 + 2NaOH + 2Pb(N03)2 = 2PbCr04 + 4NaN03 + H20 (6)
554
-------
van-
SCDIUH DiatlORTE
screening am>
PSCKRGING CF
Ntmmais
CHHOHE QXIDE
PHDOUCT
SULFUR
SO.
HIXER
KILN
tfeSH TOTER
Figure 16-1. General process diagram for production of anhydrous chrome oxide.
-------
sooiw
IfYDRATED CHRCME
oxn*: to ouroiNG
SCREB4ING AM)
WCXAfiING
BORIC ACID
SULFURIC ACID
BORIC ACID
RHXVKRY
UNIT
VftSTE WTER
Figure 16-2. General process diagram for production of hydra ted chromic 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.
Molybdenuo 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 (N03) 2 + H20 (8)
Na2Mo04 + Pb (N03)2 = PbMo04 + 2NaN03 (9)
Na2Cr04 + Pb (N03)2 = PbCr04 + 2NaN03 (10)
PbMo04 + PbCr04 = PbCr04.PbMo04 (11)
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 hexacyanoferr ate. 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 by:
PbCr04 + Fe(NH4).Fe(CN)6 = PbCr04FefNH4).Fe(CN)6 (12)
557
-------
UN> GBdTB
KITRIC ACLD
U1
U1
00
t^TER
SOD UK HYUCOUU
SCDltM DICHHttWTE
CHOC YHXCM
(KCrt>4)
TO OWING, HIUJNG
AM) PACMCrNG
VTSTC tATCK
Figure 16-3. General process diagram for production of chrcre yellow.
-------
vror
tn
cn
vo
snonM
anowTF. VRTEP
CAUSTIC SUOft
DRY INT,
HIIJ.ING
AM)
PACKAGING
»'
MOLYBCFNLH ORANQ:
(PbCrO4.PtMo04)
PRODUCT
VENT
VASTE WVTO*
HITTUC ACm
MIXER
Figure 16-4. General process diagram for production of molybdenum orange.
-------
Figure 16-5 qives 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.K2O.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)
4 ZnO + 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
-------
VOTER
IRON BUJR
tn
-------
if©
Ln
ro
»
iOCTION 1MK
FILTRATION
id ^
oRirnc
CASHING
¦CI ^
T
.kiij.dc, iiovcdc:
CF WE Zltc YOJOi
IKfi-«ZnO-4CiOj • 3H JO)
wairr
t*9re whh»
Figure 16-6. General process diagram for production of zinc yellcw.
-------
TABLE 16-3. WATER USAGE IN TOE CHROME PI<3*ENTS SU3CAGEG0KY(1)
USE
#464
UNIT FLOW (mVkkg)
Plant Designation
#436
#214
Nonoontact cooling
9.50
6.45
NA
Direct process contact
18.6
147
32.6
Indirect process contact
7.18
na(2)
NA
Maintenance
12.0
1.78
0.152
Scrubbers
3.30
9.56(3)
NA
Boiler Feed
2.52
11.1
0.152
Total
53.1
176
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 summary 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 ]6-4 represent actual plant
d ischarges.
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 1214. Data based on 308 questionnaire submission.
Chrome pigment and iron blue production and flows were included.
Plant M36. 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 1894. Data based on three days of sampling. Chrome
pigment, iron blue, and organic pigment n.5%) 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 OP PLANTS
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
-------
VATEK
RAW ^TERIALS
U1
MIU.ING MD
SCREENING
VfcSTE VftTCK
(BY-PRODUCT SALTS,
UNREACTtD WkTERIALS,
FIC.)
NON-OOMBCT
STOW
pioerr
PARTICULATE
VRSTE
Figure 16-7. General process diagram for production of dirome pigment aonplexes.
-------
TABLE 16-4. SUNMARY OF WASTE WATER FLOW
SUBCATEGORY: CHROME PIGMENTS
Plant Designation Waste Water Flcx/^
(m3/kkg)
#464
41.1
#214
32.8
#436
149
#002
78.4
#894
170(
(3)
Weighted Average Flow 105
(1) includes waste water fran all pigment product mixes.
(2) Includes organic pigments.
(3) Weighted on the basis of production since unit waste flew is
directly related to plant production:
Weighted average = Z. I (unit flow)(production)]
Z (production)
i.e. = Qi(P ) + Q2(P2) + Q3(p3)+----K>n(pn)
Pl+P2+P3+..+Pn
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 f2, half the sample was filtered
through a glass fiber filter on a Buechner funnel to simulate
the filtration process which was being bypassed at the time of
sampling. Analyses were carried out on the filtered and
unfiltered samples in order to make possible a comparison of the
total and dissolved concentrations.
A review of the sampling data indicates that the waste
treatment facility was not functioning properly during the
period of sampling at Plant f002. The inadequacies observed in
the treated effluent quality have been related to deficiencies
in the treatment system design including 1) inadequate
equalization and S02 contact facilities, 2) inadequate
clarification which in turn caused blinding of the filter and
the subsequent need for filter bypass.
Plant <1894 was also visited during the verification phase.
The treatment system has been previously discussed. The major
problem at this plant is the high unit water use rate. However,
this is the only plant found with an adequately designed and
operated treatment system in this subcategory.
During the verification phase, only certain pollutant
parameters were analyzed. These were pH, cyanide, suspended
567
-------
CAUSTIC
RAW WASTE S02 ACID
#1
CHROME
fREATMENT
TANK
pH 3.0
1
CAUSTIC
ADDITION
THROUGH
pH
8.5
±
FILTER FEED
TANK
i
FILTER AID
r
OUTFALL
TO SEWER
LAB FILTERED
BACKWASH
6
FILTERS
-J
n-F
^(FILTERS NOT WORKING SO
| WERE BEING BYPASSED.
. THIS WOULD BE THE FLOW
1 PATTERN IF FILTERS WERF
OPERATING.)
LEGEND
^ SAMPLING POINTS.
Figure 16-8. General waste water treatment process flew diagram at plant #002
showing the sairpling points. (Chrome pigment manufacture.)
568
-------
TABLE 16-5. FLOW, POLLL7TANT CONCENTRATION AND LOAD
DATA OF THE SAMPLED WASTE STREAMS FOR PLANT # 002
SUBCATEGORY: CHROME PIGMENTS
Conventional
and Nonconventiona1
Pollutants
(mg/1)
(kg/kkg of chrome pigments)
Stream
#
Stream
Description
Flow
(mvkkg)
TSS
Fe
Cr (VI)
1
Raw Waste
78.4
700
55
1.6
0.13
300
24
2-U
Unfiltered
Treated
Waste
78.4
970
76
2.3
0.18
120
9.4
2-F
Filtered
Treated
Waste
78.4
NA(1)
0.06
0.0047
na'11
(1) NA - Not available
569
-------
solids, and toxic metals. No organics were analyzed during
verification.
Figure 16-9 shows the treatment system flew diagram with
the sampling points indicated. Table 16-6 gives waste flows and
pollutant loadings.
16.3.3 Toxic Pollutant 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
ug/1) was naphthalene at 14 yg/1. It should be noted however
that some nitrobenzene (56 ug/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 f002. 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 as kg 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
0 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 chrome pigments production rate, the waste stream flow
rate, and the measured pollutant concentration.
570
-------
SLAKED
LIME
WASTE
h2so k
SO.
WASTE
WATER
#1
(EPA SAMPLE
POINT ALSO)
Li
BLEND
TANK
pH 2.5-3.0
tn
-J
EQUALIZATION
TANK
NEUTRALIZATION
TANK
pH 6.2-6.5
NEUTRALIZATION
TANK
pH 8.0-8.3
HOLDING
HOLDING
TANK
TANK
POLYMER
feACWASH
HOLDING
TANK
BACKWASH
(2)
SAND
FILTERS
e-
#5
CLARIFIER
EFFLUENT
HOLDING
TANK
(3) CLARIFIERS
#2 (EPA SAMPLE
POINT ALSO)
FINAL
DISCHARGE
TO RIVER
#3
e
SLUDGE
LEACHATE
GRAB SAMPLE
SLUDGE
LANDFILL
.SLUDGE
(3)
FILTER
PRESSES
SLUDGE
HOLDING
TANK
FILTRATE
e
LEGEND
SAMPLE POINTS.
Figure 16-9. General waste water treatment process flow diagram at plant #894
shewing the sanpling points. (Chrome pigment manufacture.)
-------
TABLE 16-6. FLOW, POLLUTANT, CONCENTRATION AND LOAD DATA FOR TOE SAMPLED
WASTE STREAMS AT PLANT # 894
SUBCATEGORY: CHROME PIOENTS
(3)
Conventional and Nonconventiona1 Pollutants
(mg/1)
(kg/kkg of chrarve pigments)
Stream
#
Stream
Description
Flow
(m3/kkg)
TSS
Fe
Cr (VI)
1
Raw waste
170
770
130
48
8.2
nd(2)
2
Final
Discharge
170
3.9
0.66
0.30
0.051
0.023
0.0039
3
Leachate
NA(1)
nd(2)
0.04
NA
nd(2)
5
Sand Filter
Influent
170
11
1.9
1.0
0.17
nd(2)
(1) Not Applicable
(2) Not Detected
(3) Verification sampling which involves three 24-hour composite samples.
572
-------
Unit loading (as kg of pollutant
per kkg of chrome pigments)
(C) (Q)
1000(P)
Where C and Q are the same as described above, and P is the
pigment production rate expressed in units of kkg/day.
(kkg is 1000 kg, a metric ton, which is equal to 2205 lbs.)
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 16-7, 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 16-8 along with the corresponding minimum and maximum
values. The toxic pollutant concentrations in the treated
effluent are presented in Table 16-9 for the two plants visited
during verification sampling.
Based on the total annual production from Table 16-1 of
this subcategory and the average waste load generated per unit
product from Table 16-8, the estimated total toxic pollutant raw
waste loads generated each year for this subcategory are as
follows:
Pollutant
Waste Load (kg/year)
Antimony
Cadmium
Chromium
Copper
Lead
Nickel
Z inc
Mercury
Cyanide
Phenol (1)
Phenolics (1)
230
34,000
900
1/500,000
1,030,000
48,000
250,000
1,200
310,000
37,000
7,100
(1) From organic pigment process
573
-------
TABLE 16-"?. TCXIC POLLLTIANT SAW MASTS DAXA
SUBCATEGORY: OHDC PIOEJTS
Average Daily Pollutant Concentrations and loadings at Plants Sanpled (1)
(mct/1)
(kgAkg of Oirorn Piiynants)
Pollutant
*894(S)
Plant Designation
(2)
*894 (V)
(3)
#002 (V)
Overall
Average
Antiiicny
Cadruun
Gircmi'jn
Ospper
Lead
Nickel
Zinc
Mercury
Cyanide, CN
Cyanide, CD (A)
7.7
1.5
0.79
0.15
55
10
7.5
1.4
36
6.8
0.16
0.030
4.1
0.78
3.6
0.68
0.76
0.13
0.88
0.15
32
14
4.1
0.70
4.8
0.82
0.017
0.0028
4.2
0.71
0.042
0.0072
4.9
0.84
0.88
0.15
1.4
0.U
0.20
0.016
310
24
1.4
0.U
54
4.2
0.32
0.025
163
13
0.00043
0.000034
0.71
0.056
3.3
0.58
0.62
o.u
150
16
4.3
0.74
32
3.9
0.17
0.019
57
4.8
0.014
0.0036
3.1
0.53
0.38
0.15
(1) The otathodology of the smpling program is described in Section 5.1.2, ar.d
Section 16.3.3 presents the scope of sarpling in the chime pignents industry.
(2) S - Screening data from one 72-hour canposite sample of individual or
carbined raw waste streams.
(3) V - verification data fran three 24-hour oanposite samples, averaged ,
fron each raw waste sampling point.
Concentration below detection or no data available.
574
-------
TAM£ 16-8. SLFMARY OF RAW fcRSTE Uya>INT»T> FUN) IN SCREENING AM) VERIFICATION SAMPLDti
SUBCATDUORY C»«CME PlOfWS
Oi
-J
i_n
FYillutujit Loading ferny e,
(kg/day)
Hininaan Mixiraun
1»>xie
Anturemy 6.0 98
Cactaiun 0.87 10
Chrcmnm^' 700 1300
Copper 6.1 %
Load 55 4 59
Nickel 0.19 2.0
Zinc 48 714
Mercury 0.0019 0.48
Cyunide, CN 3.1 56
Cyanide, CN(A) 9.8
Phenol 0.93
Phenol ics 8.8
(Jhit Ukxiiny,
(kg/kkg)
Minimum Average
0.11
0.016
10
0.11
0.82
0.0028
0.71
0.000034
0.056
0.58
0.11
16
0.74
3.9
0.019
4.8
0.0036
0.53
0.15
0.014
0.13
Mixinun
1.5
0.15
24
1.4
6.8
0.030
13 13
0.0072
0.84
t*>. of
Plants
Averaged
(2)
Conventional and Nonaonventiona 1
It) Lai
Suspended
9^1 ids, TSS
Fe
Ikscavalent
CtuXTOLLin
Cr+6
3100
7.1
8800
550
(1)
1300
5r>
0.13
93
4.2
130
8.2
(1) liexavalent chraniun is only one valent form of chnmiun.
(2) Only tlioac plants where tic pollutant was observed at significant level3 were included.
-------
TABLE 16-9. TOXIC POLLUTANT TREATED VPiSTE D&TA(1)
SUBCATEGORY: CHROME PIGMENTS
(2)
Pollutant Plant Designation Overall
(mg/1) #894 #002 Average Concentration
Antimony
0.30
0.43
0.37
Cadnium
0.0084
0.12
0.064
Chromium
0.33
130
65
Copper
0.035
0.077
0.056
Lead
0.11
1.5
0.81
Nickel
0.021
0.083
0.052
Zinc
0.058
117
59
Mercury
nd(3)
ND
ND
Cyanide, CN
0.065
*
0.065
Cyanide, CN(A)
0.0067
*
0.0067
(1) Verification sampling concentration data, average of three 24-hour
ocmposite samples.
(2) Average of two plants shown during verification sampling.
(3) Not detected.
* No data
576
-------
16.4 POLLUTION ABATEMENT OPTIONS
16.4.1 Toxic Pollutants of Concern
The toxic pollutants found in significant amounts are
mostly the heavy metals found in the products as well as the
chromium ore and other raw materials. These metals are cadmium,
chromium, copper, lead, zinc, antimony and nickel. In addition,
some cyanide was found in raw wastes and treated effluents.
This cyanide is a result of the manufacture of iron blues and,
at one plant site, HCN. However, these guidelines do not apply
to iron blues; they will be included in Phase II of the
Inorganic Chemicals regulation development. There is
significant removal of the cyanides in the chrome pigments
treatment, however, probably due to the precipitation of
ferrocyanides. The HCN manufacturing process is also regulated
by another guideline (see Section 17). Some organic toxic
pollutants were found during the screening phase. This was
believed to be an anomaly caused by the sampling procedure,
since they were also found in the raw intake water, treated
effluent, or in the raw waste. In addition, any organics
present are probably caused by organic pigments manufacture
which is not regulated by this guideline, but will be regulated
under the Organic Chemicals Category.
All the waste waters generated in the chrome pigments
subcategory contain dissolved chromium and pigment
particulates.
Additional pollutants that may be anticipated are given
below for each major pigment group.
Chrome Yellow and Chrome Orange
The raw waste waters contain sodium acetate, sodium
chloride, sodium nitrate, sodium sulfate, and lead salts.
Chrome Oxide
The aqueous process effluent contains sodium sulfate. If
boric acid is used in the preparation of hydrated chromic oxide
then the waste water will contain sodium borate and boric acid.
Chrome Yellow and Chrome Orange
Additional pollutants present in the raw waste water from
chrome yellow and chrome orange manufacture include sodium
acetate, sodium chloride, sodium nitrate, sodium sulfate, and
lead salts.
577
-------
Molybdenum Orange
Process waste effluents from the manufacture of molybdenum
orange contain sodium chloride, sodium nitrate, sodium sulfate,
chromium hydroxide, lead salts, and silica.
Chrome Green
The raw waste water contains sodium nitrate. If iron blue
is manufactured on site as part of the process for chrome green
manufacture, the waste water also contains sodium chloride,
ammonium sulfate, ferrous sulfate, sulfuric acid and iron blue
pigment particulates.
Zinc Yellow
The raw wastes contain hydrochloric acid, sodium chloride,
potassium chloride, and soluble zinc salts.
16.4.2 Process Modifications and Technology Transfer Options
The major process problem in the industry is the high rate
of water use in some cases. This can be alleviated in a number
of ways.
1. Close attention to product quality in conjunction with
reduction of product rinses.
2. Reduction in equipment cleaning rinses by the following
methodologi es:
a. Recycle of rinse waters.
b. Minimizing of product changes by the use of better
planning and increased number of units.
Equipment cleaning is known to contribute approximately 20
percent of the waste load volume at one plant (f002).
3. Use of parallel treatment for individual product lines.
This will allow the reuse of rinse waters and the recovery of
products presently lost in waste sludges.
4. The use of ion exchange and/or reverse osmosis on
isolated waste waters. This will allow total recovery of
product as well as total reuse of waste water. This system is in
use on one line at Plant ?409.
578
-------
The above options were reviewed, but except for option 1
were not considered for inclusion in the treatment models due to
the engineering required and their capital intensive nature.
16.4.3 Best Management Practices
1. All storm water and surface area runoff from the plant
site should be collected and sent to a treatment facility if the
water is contaminated from process wastes. This contamination
can be minimized by storage of chemicals indoors, proper air
pollution control, and elimination of all spills.
2. If the solids from the treatment plant are disposed of
on-site, provision should be made to control leachates and
permeates. It is possible to monitor the metal concentrations
and when concentrations approach predetermined limits, the
leachate can be pumped back to the treatment system for further
treatment.
16.4.4 Prevailing Control and Treatment Practices
A description of the individual treatment facilities for
those plants visited is given in 16.3.1 and 16.3.2. In
addition, the following information was obtained for the
remaining plants.
Plant f214 manufactures pigments and other chemicals. The
plant does not have a waste water treatment facility. After pH
adjustment, waste is discharged to a POTW. Part of the process
waste is recycled.
Plant f593 manufactures organic and inorganic chemicals.
Existing combined waste water treatment plant consists of
lagoon, aeration, clarifiers, and filters. The sludge disposal
is on-site landfill.
Plant f464 manufactures both organic and inorganic
pigments. After pH adjustment, waste water is discharged to
POTW.
Plant flOl manufactures inorganic ceramic pigments, color
and porcelain. The existing combined waste water facility
consists of a series of settling basins. Sludge disposal is to
off-site landfill. After pH adjustment, the final discharge is
to a POTW.
Plant f502 manufactures both organic and inorganic
pigments, of which chrome pigments are a small part. Treatment
consists of pH adjustment prior to discharge.
579
-------
Plant #436 manufactures several chemicals in addition to
chrome pigments. The treatment system consists of
neutralization with caustic and clarification in settling
lagoons prior to discharge. Sludge is contract-hauled
approximately once every three years.
Plant M09 manufactures specialty chemicals and inorganic
pigments. The existing waste water treatment facility consists
of S02 reduction, clarification, filters and pH adjustment.
Sludge disposal is to an off-site location.
Plant £997 manufactures chromic oxide and sulfuric acid.
Production data is not available. The existing waste water
treatment facility consists of pH adjustment, SO? reduction and
lagoons.
Plant t'962 manufactures inorganic pigments (chrome
yellow). Existing waste treatment plant consists of
flocculation, clarification and filters. After pH adjustment,
the effluent is discharged to a POTW. Sludqe is recycled to
process.
Plant f200 manufactures and imports small quantities of
chrome pigments. Treatment is unknown.
In summary, a review of the existing treatment system
descriptions indicates that the prevailing treatment practices
appear insufficient except for the system at Plant f894. The
major problems besides total lack of treatment is lack of
sufficient residence time, lack of critical treatment units, and
failure to collect all waste streams. As previously stated,
only Plant £894 has a properly designed and operated treatment
system. This system is basically the same as the Level 1
treatment system shown in Figure 16-10.
16.4.5 Advanced Treatment Technologies
The treatment technologies in use in the industry consist
of segregation, equalization, S02 reduction, alkaline
neutralization, clarification, and filtration. In addition, the
following technolgies were reviewed for model plant development:
sulfide precipitation, ion exchange, reverse osmosis, and the
xanthate process.
580
-------
16.5 SELECTION OP APPROPRIATE TECHNOLOGY AND EQUIPMENT
16.5.1 Technologies for Different Treataent Levels
A careful review of the end-of-pipe treatment methods
available to industry was made. As a result, the following two
methodologies were chosen as treatment levels. The following
considerations were made in establishing the models:
]. Effective reduction of pollutants.
2. Established treatment practices in the industry.
3. The cost of technology.
4. The adaptability of the model to different situations.
Level 1 (BPT/BAT)
Consists of equalization, S02 reduction, alkaline
precipitation, clarification, and filtration.
Level 2
For better removal of the trace metals, sulfide
precipitation is incorporated ahead of the BPT dual media
f ilter.
The flow diagrams for these two levels are shown in Figures
16-10 and 16-11._
16.5.2 Equipment for Different Treataent Levels
Equipment Functions
In both levels, the incoming wastes are acidified in a
holding tank and then treated with sulfur dioxide solution in a
reactor to convert hexavalent chromium to trivalent chromium.
Caustic soda is then added as a precipitant and a polymeric
coagulant is added to help settle the heavv metal hydroxides in
a clarifier. The settled effluent is then filtered in a dual
media filter and discharged aftet pH adjustment to the range 6
to 9. In Level 2, ferrous sulfide is added ahead of the dual
media filter for more effective precipitation of all the
residual heavy metals, including antimony. As in Level 1, the
filter effluent is adjusted to a pH between 6 to 9 before
discharge.
581
-------
BACKWASH
cn
oo
to
SUIJ'MRIC
ACID
I 0
RAW
WASTE WATER
SUI.FUR
I I dioxide
Q-jIl
HO!.DING TANK
CAUSTIC SODA
POLYMER
ADJUSTMENT
EFFLUENT
SUMP
FILTER
CI^ARIFIER
REACTION MIX
TANK TANK
+-
-------
FFRROUS SULKATF SODIUM HI.SUI.FIDF
RAW
WASTE WAT
BACKWASH
C.AUbTK
SODA
SIII. F U R
nioxiDf
II1 Al'.Itl.STMKMT
POLYMER
I HOI-PING TANK
SUMP FI1.TFR
CI.AKIFIER
EKKUIKN'l
REACTION MiX
TANK TANK
Ul
00
to
IhfrLJ—»
SUJDGE
TO I.ANDFU.I.
In< luHr* flow monitoring, |>H monitoring and sampler.
Figure 16-11. Level 2 waste water treatment for chrcme pigments.
-------
Chemicals and Handling
Sulfuric acid and caustic soda solutions are common
industrial chemicals which are readily handled with conventional
liquid feeding equipment. Sulfur dioxide is received as a
compressed gas which is dissolved in water by a modified gas
chlorinator and fed to the reactor to maintain consistent
reducing conditions. Polymer is fed by a standard package of
holding tank, mixer, and feeder. With normal precautions there
are no unusual hazards in handling chemicals for treatment of
chrome pigment wastes.
Separation and Disposal of Solids
Solids from the clarifier, including reci
backwash solids, are dewatered in a filter press
chemical landfill. Sludge filtrate is returned
holding tank.
Monitoring Requirements
Internal process monitoring consists of maintaining proper
pH levels in the holding tank and final effluent, using
conventional field equipment. A reducing environment is
maintained in the reactor, using an oxidation-reduction
potential instrument and/or analysis for excess S02. Periodic
effluent analyses for chromium and heavy metals should be made
on composite samples by atomic absorption methods, for official
reporting purposes. Sulfide monitoring is generally unnecessary
because dissolved sulfides should not exist in the presence of
excess ferrous iron and oxygen.
rculated filter
and hauled to a
to the influent
16.6 TREATMENT COST ESTIMATES
16.6.1 General Discussion
To prepare cost estimates, a model plant concept was
developed and plant criteria developed for both Level 1 and
Level 7.
Waste Water Plow
The data for five plants with usable flow data is
summarized in Table 16-4. This information was used on a
production weighted basis to determine the average flow in the
industry. This average was computed to be 105 m3/kkg (25,200
gal/ton). This value was used for sizing the model plants.
584
-------
Chromium Pigment Production
Production in the chrome pigment subcategory ranges from a
low of 100 kkg/year to a high of approximately 18,000 kkg/year.
The mean production is approximately 7200 kkg/year. For the
purposes of estimating treatment costs, four production levels
were selected as model plants. These are 1500 kkg/year, 4000
kkg/year, 6000 kkg/year, and 18,000 kkg/year. These cover the
entire range of production rates. Most plants produce manv
chrome pigment products on a continuous basis so the operational
mode selected was continuous and assumed to run 350 days per
year. Chrome pigments are usually produced in integrated
facilities with the necessary flexibility to shift from one
product or combination of products to another. The model plant
was selected to reflect this type of complexity.
Waste Hater Pollution Load
For the model plants, the loads are based on verification
plant data. This data indicated an average loading of 16 kg/kkg
chromates as chromium (Table 16-8) . Total toxic metals loadings
ranged from 12 kg/kkg to 47 kg/kkg. Total suspended solid
loadings ranges from 55 kg/kkg to 130 kg/kkg (Table 16-8). The
overall solid waste generation is expected to be 85 kg/kkg to
150 kg/kkg (dry solids). For the purpose of determining solid
waste generation, a value of 105 kg/kkg (dry solids) was
selected.
The costs shown at each level of treatment correspond to
the model plant BPT system (Level 1) and an alternative system
incorporating sulfide precipitation into the BPT model in order
to meet more stringent toxic pollutant requirements.
The estimated costs for the four models is given in Tables
16-10, 16-11, 16-12, and 16-13. For these models, both
hydraulic and pollution loads per unit of production were held
constant over the entire range of production. Annual treatment
costs as a function of production is shown graphically in Figure
16-12, while unit treatment costs as a function of production is
given in Figure 16-13.
In order to determine the accuracy of the treatment model,
an attempt was made to compare the model costs against actual
industry costs. Cost data were received on two plants, one with
treatment installed and one in the design stage. No attempt was
made to compare costs item by item since these specific costs
may differ for the following reasons:
1. Variations in land costs.
585
-------
TABLE 1£-10. MODEL PLANT TREATMENTT COSTS
Subcategory CHRCME PIOlENTS
Production
Waste water flow
1,500 metric tons per year
4 metric tons per day
454 cubic meters per day.
(1)
(l,^*5 tons per year9^
(4 tons per day)
A. INVESTMENT COST
Construction
Equipment in p^ce,
including piping,
fittings, electrical
work and controls
Monitoring equipnent
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 v*este
disposal
Monitoring, analysis
and reporting
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
LEVEL OF TREATMEhTI(2)
FTRST SECOND
$3£,800
280,^50
9,000
£5,290
£5,290
£,000
S4£3,030
$112,000
7,350
51,000
45,"701
n,R90
5,000
15,000
$251,941
$74,358
$32£,301
$],000
10,000
2,200
2,200
$15,400
SI 4,000
300
2,200
1 ,540
4£2
¦\500
S2£,002
$2,505
$28,50^
(1) 350 days per year
(2) First level represents the base cost of treatment system.
Other levels represent the incremental cost, above base cost.
586
-------
TABLE 15-11. MODEL PLANT TREATMENT COSTS
Subcategory CHRCME PICTIENTS
Production
Waste water flow
4,000 metric tons per ye?r
11 metric tons per day
1219 cubic meters per day.
(1)
(4,410 tons per year^"'
(12 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
Ma intenance
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
(2)
FIRST
$5"*, 900
510,000
9,000
114,580
114,580
12,000
$814,050
$1.12,000
]5,000
141,300
80, 20*1
24,421
15,000
15,000
$402,927
$130,495
$533,422
SECOND
$2,000
15,000
3,400
"*,400
$23,R00
$14,000
"*00
5, °00
2, 380
n\A
"7,500
$30,794
$3, 872
$^4,«55
(1) 350 days per year
(2) First level represents the base cost of treatment system.
Other levels represent, the incremental cost above base cost.
587
-------
TABLE 16- 12. MODEL PLANT TREATMENT COSTS
Subcategory CHRCME PIGMENTS
Production 6,000 metric tons per year^ (6,615 tons per year>^
17 metric tons per day (18 tons per day)
Waste water flow 1820 cubic meters per day.
(2)
LEVEL OF TREATMENT
FIRST SECOND
A. INVESTMENT COST
Construction
$71,400
$5,000
Equipment in place,
including piping,
fittings, electrical
work and controls
667,000
20,000
Monitoring equipnent
in place
9,000
Engineering design
and inspection
149,480
5,000
Incidentals, overhead,
fees, contingencies...
149,480
5,000
Land
12,000
TOTAL INVESTMENT COST
$1,058,360
$35,000
OPERATION AND
MAINTENANCE COST
Labor and supervision.
$112,000
$14,000
Energy
20,200
300
Chemicals
211,500
R, 800
Maintenance
104,636
1,500
Taxes and insurance...
31,750
1,050
Residual waste
disposal
20,000
Monitoring, analysis
and reporting
15,000
7,500
TOTAL OPERATION AND
MAINTENANCE COST
$515,086
$35,150
AMORTIZATION OF
INVESTMENT COST
$170,242
$5,694
TOTAL ANNUAL COST
$685,328
$40,844
(1) 350 days per year.
(2) First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
588
-------
TABLE 15-13. MODEL PLANT TREATMENT COSTS
Subcategory CHRCME PIOIENTTS
Production IP,000 metric tons per year^ (19,845 tons per year9^
51 metric tons per day (56 tons per day)
Waste water flow 5450 cubic meters per day.
LEVEL OF TREATMENT (2)
FIRST SECOND
A. INVESTMENT COST
Construction
Equipment in place,
including piping,
fittings, electrical
work and controls
Monitoring equipment
in place
Engineering design
and inspection
Incidentals, overhead,
fees, contingencies...
Land
TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Energy
Chemicals
Maintenance
Taxes and insurance...
Residual waste
disposal
Monitoring, analysis
and reporting
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
5205,500
1,495,500
9,000
342,000
^42,000
18,000
52,412,000
5112,000
28,000
635,000
239,400
72,360
60,000
15,000
51,161,760
5389, 50
51,551,263
54,000
60,000
12,800
12,800
589,600
514,000
600
26,400
8,960
2,688
7,500
560,148
514,577
574,725
(1) 350 days per year
(2) First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
589
-------
I
LEVEL, » I OR *2
c ? T2 rr
PRXOCna." (MTTRIC TCNS/23VR X lOOO)
Figure 16-12. Annual treatment cost vs. production for the chrome pigments
subcategory.
590
-------
240
2ZC
ZOO
180
U
160
140
120
LEVEL iZ
80
PRXUCTICN («TWC TCNSAEXR X 1000^
Figure 16-13. Annual unit treatment cost vs. production for the chrome pigments
subcategory.
591
-------
2. Variations in hydraulic loading.
3. Varying costs of solid waste disposal.
The following overall results were obtained:
Annual Costs
($/kkg)
Model Plant 86.18
Plant f002 85.38
Plant f 894 91 .03
The above data indicate a very good correlation between the
model plant and site specific engineering estimates.
Table 16-14 presents a summary of the unit cost
distribution beween amortization, operation, and maintenance
cost components at various production and levels of treatment.
For the model plant, the primary sources of waste water are
from product washing, slurrying of reaction products, scrubbing
of reactor vent gases, and washing of equipment due to product
charges.
16.6.2 Model Plant Costs
The major costs for the Level 1 model plant
labor, and chemical costs. Engineering design
maintenance are also fairly large. The majority
cost is tied up in operation and maintenance,
approach 50% of the total capital cost.
The second level of treatment has a much lower incremental
cost than the first. However, the cost breakdown is quite
similar to Level 1.
The cost of transporting and disposal of 30% solids sludge
is included in the cost estimates.
16.7 BASIS FOR REGULATIONS
16.7.1 Evaluation of BPT Treatment Practices
A number of factors are anticipated to contr
variation in the effluent quality at chrome
treatment facilities. Consideration of these
included in establishing limitations in that the
are equipment,
and equipment
of the annual
This cost can
ibute to a wide
pigment plant
variations is
performance of
592
-------
TABLE 16-14. MODEL PLANT TREATMENT COSTS
Subcategory: CHROME PIGMENTS
Annual Treatment Costs ($/kkg)
LEVEL OF TREATMENT
COST ITEM
PRODUCTION
FLOW
FIRST
SEOONI
(kkg/yr)
(m3/day)
Annual Operation
and Maintenance
1,500
454
167.96
17.33
4,000
1,219
100.73
7.70
6,000
1,820
85.85
5.86
18,000
5,460
64.54
3.34
Annual
Amortization
1,500
454
49.57
1.67
4,000
1,219
32.62
0.97
6,000
1,820
28.37
0.95
18,000
5,460
21.64
0.81
Total Cost
1,500
454
217.53
19.00
4,000
1,219
133.36
8.67
6,000
1,820
114.22
6.81
18,000
5,460
86.18
4.15
THIRD FOURTH
Not Applicable
593
-------
the plant on which limitations are based is a large complex
plant that encounters all of these factors. These include the
followi ng:
Product Changes
Changes in products require that equipment he thoroughly
cleaned prior to reuse. Therefore, frequent product changes
will result in higher waste flows.
Product Application
The final disposition of the product will affect the
quality required. The higher the Quality, the more water
required for rinsing.
Air Pollution Control
Equipment will be required in many cases for control of the
environment as well as off-site air compliance. Scrubbers will
add some waste flow to the treatment system. This flow,
however, is generally small.
Other Related Products
Many plants manufacture other types of pigments including
iron blues and organic pigments. These products generate
significant quantities of waste water which tend to dilute
chrome pigment wastes. However, these waste waters were
included in the computation of the unit waste flow. Therefore,
the use of parallel treatment for existing facilities producing
other pigments is not required at this time as long as chromium
pigment production is the majority of the overall production.
The following guidelines should be used in applying these
requlations:
1. When determining the effluent loadings, the total
production of a facility will be used as long as the chrome
pigment production is in the majority.
2. When the chromium production is the minority of the
overall production, the total production should be used for
computing the effluent limits under the following conditions:
the remaining production (other than chrome pigments) generates
a waste water containing significant amounts of toxic metals
which will be removed by a chrome pigment treatment system.
3. For those facilities (existing sources) where chromium
production is in the minority and the wastes from other sources
do not contain metals above accepted levels of treatability,
594
-------
segregation and parallel treatment of chromium pigment wastes
are recommended. However, the permitting authority or POTW must
consider the following balancing factors:
a. The economic impact on the facility balanced against
b. The environmental benefits of parallel treatment.
In addition to the above factors, the design and operation
of the treatment facilities affect effluent quality. Important
factors are equalization, S02 contact time and pH depression,
S02 dose, proper neutralization, and adequate solids removal.
Table 16-15 is a summary of verification sampling and long-
term effluent monitoring data at Plant f894 for the major
pollutants of concern. Plant £002 sampling results are excluded
from the subcategory performance evaluation, since the treatment
system was not functioning properly as previously discussed.
The long-term monitoring data in Table 16-15 is for the maximum
30-day average long-term monitoring results. Sufficient data
was not available to estimate long-term daily maximum values.
Plant #894 is the only known plant with Level 1 treatment
system installed and operating. Table A-lla sets forth means,
variability factors, and the 95 percent monthly average.
Maximum daily performance (99%) was not computed since the
discrete sampling data was not available at the time of the
evaluation. The performance evaluation in Table 16-5 is
utilized for the development of proposed regulations for TSS and
applicable toxic metals.
As previously stated, only one plant of the existing twelve
is known to have a Level 1 treatment system installed. This
plant represents approximately 30-35 percent of total
production. Most other plants have some type of treatment
installed, but none of these appear to be adequate. This
technology is expected to remove 3,200,000 pounds per year of
toxic metals.
The Agency is conducting additional treatability studies
for the subcategory, the data from which will be available
before promulgation of a final regulation.
16.7.2 Basis for Proposed BPT Effluent Limitations
Technology Basis
For BPT, the Agency is proposing limitations based on
equalization, reduction of hexavalent chromium followed by
alkaline precipitation, and dual media filtration. Reduction of
595
-------
TABLE 16-15. SUMMARY OF LCNG TERM AND VERIFICATION EFFLUENT SAMPLING
RESULTS AT PLANT *894
SUBCATEGORY: CHROME PIGMENTS
Pollutant
(3)
Verification Sampling
(mg/l) (kg/kkg)
Achievable Performance
Max 30-^.ay Avrj
(mg/1)
(kg/kkg)
Total Suspended
Solids, TSS
3.9
0.66
23
3.9
Iron
0.30
0.051
na(2)
NA
Antimony
0.30
0.051
NA
NA
Arsenic
ND(1)
ND
0.16
0.027
Cadmium
0.0084
0.0014
0.12
0.020
Chromium
0.33
0.056
0.73
0.12
Copper
0.035
0.0060
0.25
0.42
Lead
0.11
0.019
0.87
0.15
Mercury
ND
ND
0.0016
0.00027
Nickel
0.021
0.0036
NA
NA
Zinc
0.058
0.0099
0.074
0.013
Cyanide (CN-A)
0.065
0.011
0.068
0.012
Cyanide (Total)
0.0067
0.0011
0.31
0.053
Chromium (VI)
0.023
0.0039
0.30
0.051
(1) ND, Not Detected.
(2) NA, Not Available.
(3) Frcm Table 16-9.
(4) Frcm Table A-lla, "Historical Effluent Monitoring Data Sunmary."
596
-------
flow by the methods given in 16.4.2 was considered but not used
since their application is site specific. However, they are
quite viable options in most cases and could result in
substantial treatment cost savings.
Plow Basis
The basis of flow for the proposed BPT limitations is
estimated from data provided in the 308 questionnaires and plant
visits during sampling. Table 16-4 presents the plant flow data
used for the purpose of regulation. A weighted average flow was
determined based on plant production. In other words, plants
producing a greater quantity of chrome pigment product have a
waste flow which has a greater influence on the average flow
calculation. This approach for the determination of the average
flow is substantiated by the unit waste flow which is related to
the plant production rate.
Since plants in the chrome pigments subcategory do not
segregate waste waters from the various pigment processes for
treatment, the basis of flow for the purpose of regulation
includes all process related waste water combined. The flow
basis is 105 m3/kkg from Table 16-4. This flow does not include
any recycle or reuse of waste waters other than some incidental
recycle being done at five plants included in the data base.
Selection Basis for 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 at Plants #002 and f894 provided
additional pollutant raw waste concentration data needed to
assess the magnitude of the pollution potential.
Results of the screening and verification sampling are
tabulated in Section ]6.3.3 for the raw process waste streams.
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 significant concentration levels. Pollutants from
this list were considered as candidates for regulation if their
concentrations appeared to equal or exceed in at least one
instance the lowest level estimated as treatable using any
available technology appropriate for their removal, ignoring
economic considerations.
597
-------
The relative significance of the candidate pollutants was
estimated based on the total annual raw waste load for each
pollutant which appears in a Table in Section 16.3.3. The total
annual load is based or. the average concentration observed
during screening and verification which is tabulated in Table
16-8 in addition to the estimated annual production of 64,500
kkg of product for the industry.
Specific numerical effluent loading limitations were
proposed only for those candidate pollutants which appeared at
average concentration levels (Table 16-7) considered to be
treatable for at least one plant visited during sampling.
On the basis of concentration and total annual raw waste
loads determined during sampling, chromium, zinc, lead, copper,
antimony, cadmium, nickel and mercury have been identified in
the raw waste stream and are also candidates for regulation.
Organic pollutants and cyanide are not included, since they are
considered products of iron blue, organic pigments, or HCN
production as discussed under 16.3.1. In addition, these
parameters will be covered by future regulations in other
subcategor ies.
In view of the treatment technology currently practiced and
the related nature of the candidate pollutants, control of the
more significant toxic pollutants should ensure adequate control
of those metals which may occasionally appear at treatable
levels.
Consideration of direct hexavalent chromium limitations
has been dropped due to problems with the analytical procedure.
Studies have shown significant inaccuracies in the measurement
of hexavalent chromium in chrome pigment wastes. It does not
appear that this problem will be overcome in the near future.
However, hexavalent chromium will be adequately controlled by
the total chromium limit. This is because almost all the
chromium must be converted to the trivalent state in order to be
removed from solution by alkaline precipitation. Limitations on
hexavalent to some degree may be considered redundant.
Hexavalent chromium should be excluded from consideration
in the proposed regulations. The complexity and subsequent
accuracy of the analysis may cause misleading conclusions if
used as an effluent monitoring parameter. S02 reduction under
acidic conditions should convert hexavalent chromium to its
trivalent form which can be conveniently verified by analysis of
total chromium in the treated effluent. Chromium can not be
removed by alkaline precipitation unless it is in the trivalent
598
-------
form. Therefore, if the S02 reduction step fails to reduce the
hexavalent chromium it will become apparent in the effluent
total chromium concentration.
Basis of Pollutant Linitations
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. Total Suspended Solids (TSS): Review of the long-term
monitoring and verification sampling data in Table 16-3 5
indicates a maximum 30-day average TSS discharge of 3.9 kg/kkg
for the purpose of the proposed limitation determination. The
30-day average concentration basis is then determined as
follows:
/ 3.9 kg/kkg\ /1000 mq/l\
\105 m3/kkg/ \ kg/m3 J
37 mg/1
The 24-hour maximum loading limitation is determined by the
following relationship:
/ 30-day average ^ (VFR)
^loading or concentration/
24-hour maximum
loading or
concentration
The variability factor ratio (VFR) is estimated from the
Titanium Dioxide Sulfate Process Subcategory based on 30-day
average and daily variability factors for zinc. The long-term
monitoring data on zinc showed daily average concentrations
ranging from 0.010 to 1.14 mg/1 during a period of more than two
years (Tables A-9a-l and A-9c-l in Appendix A). This range of
values for zinc nearly spans the observed range of toxic metal
concentrations found in the effluent from Chrome Pigments Plant
f894 (Table 16-15) . The VFR of 2.4 for zinc in the Ti02 Sulfate
Process reflects the overall metal removal performance of
alkaline precipitation followed by settling and discharge
without filtration. Therefore, this VFR is applied to the
Chrome Pigments industry as a conservative estimate of the
performance of a similar treatment technology which does include
a final filtration step. Therefore, the 24-hour maximum
limitation becomes,
(3.9 kg/kkg) (2.4) = 9.4 kg/kkg
C. Other pollutants: The concentration basis for iron is
also presented in Table 16-15. This concentration is intended
to serve as guidance in cases where iron is found to be of
serious concern.
599
-------
Toxic Pollutants
The effluent limitations proposed for the selected toxic
pollutant control parameters are derived from three sources of
information including 1) screening and verification sampling
data, 2) literature based treatability estimates (Section 8.1) i
and 3) a limited amount of long-term monitoring data at Plant
£894.
The sampling results represent plant performance observed
during three days of sampling. The sampling data was used
primarily to select the pollutants of concern, and in the case
of antimony and nickel the sampling results were used to
estimate the 30-day average concentration in view of the lack of
long-term monitoring data for these two pollutants.
The sampling data for Plant f894 appears to demonstrate
that in some cases the effluent quality for metal pollutants are
considerably better for BPT treatment than indicated by
literature treatability data in Section 8.1. 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.
The VFR used to determine the proposed 24-hour maximum
limitations is based on long-term data for zinc in the Titanium
Dioxide Subcategory.
A. Chromium: The raw waste concentration for chromium was
observed as high as 370 mg/1 and averaged 150 mg/1 during
sampling (Table 16-7). The long-term monitoring results
indicate a maximum 30-day average discharge of 0.12 kg/kkg which
is the basis of the proposed limitations. The concentration
basis then becomes,
/ 0.1 2 kg/kkg\ ( 1000 mg/A
\105 m3/kkg / V kg/m3 /
1.1 mg/1
The 24-hour maximum is determined as follows,
(0.12 kg/kkg) (2.4) = 0.29 kg/kkg
where the VFR is set equal to 1.9 based on data from the
Ti02 subcategory.
B. Zinc: Proposed zinc limitations were set equal to
chromium. Tables 16-7 and 16-9 indicate that the removals of
zinc and chromium are similar at Plant f002 where zinc is found
at very high raw waste concentrations.
600
-------
C. Lead: The raw waste concentration for lead was
observed as high as 69 mg/1 and averaged 32 mg/1 during
sampling. The long-term monitoring results indicate a maximum
30-day average discharge of 0.15 kg/kkg which is used as the
30-day average limitation. The concentration basis then
becomes,
/0.15 kg/kkg"\/l000 mq/l\ = 1.4 ma/1
\10 5 m3/kkg J\ kg/m3 /
The 24-hour maximum proposed limitation then becomes,
(0.15 kg/kkg) (2.4) = 0.36 kg/kkg.
D. Copper: The raw waste concentration for copper was
observed as high as 6.2 mg/1 and averaged 4.3 mg/1 during
sampling. The long-term monitoring results indicate a maximum
30-day average discharge of 0.042 kg/kkg which is used for the
proposed limitations. Therefore, the proposed concentration
basis becomes,
/0.042 kg/kkg\ (1000 mg/l\ = 0.40 mg/1
\ 105 m3/kkg/ \ kg/m3 J
The 24-hour maximum proposed limitation then becomes,
(0.042 kg/kkg) (2.4) = 0.10 kg/kkg.
E. Antimony: The raw waste concentration for antimony was
observed as high as 7.7 mg/1 and averaged 3.3 mg/1 during
sampling. The verification sampling results indicate an average
discharge of 0.051 kg/kkg which is used as the 30-day average
limitation. The concentration basis then becomes,
/ 0.051 kg/kkg\ /1Q00 m/ql\ = 0.48 mg/1
V 105 m3/kkg/ \ kg/m3 /
The 24-hour maximum is then,
(0.051 kg/kkg) (2.4) = 0.12 kg/kkg.
F. Cadmium: The raw waste concentration for cadmium was
observed as high as 1.3 mg/1 and averaged 0.62 mg/1 during
sampling. The long-term monitoring results indicate a maximum
30-day average discharge of 0.020 kg/kkg which is used as the
30-day limitation. The concentration basis then becomes,
/ 0.020 kg/kkg\/1000 mg/l\ = 0.19 mq/1
V 105 m3/kkg ) \ kg/m3 )
601
-------
The 24-hour maximum is then,
(0.020 kg/kkg) (2.4) = 0.048 kg/kkg.
G. Nickel: The raw waste concentration for nickel was
observed as high as 0.74 mg/1 and averaged 0.17 mg/1 during
sampling. The verification sampling results indicate an
achievable concentration of 0.021 mg/1 which compares to a
literature treatability value of 0.17 mg/1. This was estimated
by application of a 14 percent removal to 0.2 mg/1 from Table 8-
11 as demonstrated in Section 15.7.4 for nickel. Therefore, the
proposed 30-day average limitation is based on 0.17 mg/1 as
follows:
(0.17 mg/1) O.0 5 m3/kkg) ( kg/m3 \ = 0.018 kg/kkq
V1000 mg/1/
The 24-hour maximum then becomes,
(0.018 kg/kkg) (2.4) = 0 .043 kg/kkg.
The proposed limitations are summarized in Table 16-16 for
BPT.
H. Mercury: The raw waste concentration for mercury was
observed as high as 0.078 mg/1 and averaged 0.014 mg/1 during
sampling. The 30-day average long-term monitoring data
indicates a maximum 30-day average discharge of 0.00027 kg/kkg.
At the unit flow rate of 105 m3/kkg, this reflects a discharge
concentration of 0.0026 mg/1. Althouah significant coincidental
removal of mercury is observed with a larqe scale BPT system,
the treatment technology is not specifically oriented for the
treatment of mercury. Therefore, the concentration basis for
mercury is indicated in Table 16-16 for use in cases where it is
found to be of serious concern.
16.7.3 Basis for Proposed BCT Limitations
The BCT limitation (applicable only to TSS and pH) was set
equal to BPT because BAT is equal to BPT.
16.7.4 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 various advanced level options for
conventional, nonconventional, and toxic pollutant removal. The
economic impacts on the Chrome Pigments Industry have been
602
-------
TABLE 16-16. PROPOSED LIMITATIONS
Chrcme Pigments
Best Practicable Control Technology Currently Available
Waste Water Flew: 105 m3/kkg
Pollutant
Subcategory
Performance
(mg/1)
(1) Concentration Basis
^ (rrg/l)
Effluent Limit
(kg/kkg)
Max
30-day
Avg
24-hr
Max
Max
30-day
Avg
24-hr
Max
Conventional
and Nonoonventional
Pollutants:
Total Suspended
Solids, TSS 23
2.4,5)
37
89
3.9
9.4
Iron
0.30(4)
2.4
0.49
1.2
—
—
(2)
Toxic Pollutants:
Antiomony
0.30(4)
2.4
0.48
1.2
0.051
0.12
Cadmium
0.12(3)
2.4
0.19
0.46
0.020
0.048
Chromium
0.73(3)
2.4
1.1
2.6
0.12
0.29
Copper
0.25(3)
2.4
0.40
0.96
0.042
0.10
Lead
0.87(3)
2.4
1.4
3.4
0.15
0.36
Mercury
0.0016
2.4
0.0026
0.0062
—
—
Nickel
0.021(4)
2.4
0.17(6)
0.41
0.018
0.043
Zinc
0.074(3)
2.4
1.1
2.6
0.12
0.29
(1) VFR: ratio of the 24-hour variability factor to the 30-day variability
factor.
(2) Also applicable to BAT and PSES which are set equal to BPT by the Agency.
(3) Long term 30-day average monitoring data frcm Table 16-15.
(4) Verification sampling results based on three, 24-hour composite effluent
samples.
(5) VFR selected from long term data evaluation in the Titanium Dioxide
subcategory.
(6) Lower limit of treatability estinate (Table 8-11).
603
-------
evaluated in detail and taken into consideration in the
selection of the technology basis for the proposed BAT
regulat ions.
The Agency is proposing BAT limitations based on treatment
consisting of Level 1 technology which is equivalent to BPT.
The implementation of BPT/BAT will remove 3,200,000 pounds of
toxic metals annually.
Technology Basis
For BAT, the Agency is proposing the identical technology
basis discussed for BPT in Section 16.7.2. BAT includes no
additional treatment because there are insufficient data to
confirm performance and the added cost is not offset by better
effluent quality.
Plow Basis
The unit flow of 105 m3/kkg is also proposed for BAT.
Selection of Pollutants to be Regulated
The basis of pollutant selection is discussed for BPT under
Section 16.7.2. For BAT, the toxic metals shown in Table 16-1.6
are proposed for regulation. These include chromium, zinc,
lead, copper, antimony, cadmium, and nickel.
Basis of Pollutant Limitations
The basis of the limitations are discussed in detail under
BPT Section 16.7.2. Table 16-16 summarizes the proposed
limitations for BAT which are designated by footnote 2.
16.7.5 Basis for Proposed New Source Performance Standards
Application of Advanced Level Treatment
Chrome pigment industry wastes primarily contain toxic
metal pollutants which are particularly amenable to removal by
alkaline precipitation and sulfide precipitation. Almost all
plants combine waste water from the chrome pigment process with
waste water from unrelated processes. The Agency proposes that
for new sources, the waste water from the chrome pigments
process be segregated from waste water from other processes
unless the other waste water contains toxic metal pollutants.
Segregation and separate treatment of the waste waters can
conceivably reduce treatment costs, and simplify the treatment
of metals without complications from unrelated waste water
constituents not amenable to metals treatment.
604
-------
Technology Basis
For New Source Performance Standards (NSPS), the Agency is
proposing limitations based on more stringent removal of metals
by sulfide precipitation before filtration in addition to BPT
(Level 2). The Agency also proposes that all unrelated waste
water sources which are not amenable to metals treatment, be
segregated before treatment as previously discussed.
Flow Basis
The basis for the unit flow used for the purpose of
proposing limitations is 105 m3/kkg and does not differ from
BPT.
Selection of Pollutants to be Regulated
The same conventional, nonconventional, and toxic
pollutants selected for BPT Section 16.7.2 are also considered
here for the proposed NSPS limitations. These include TSS, pH,
iron, and the same eight toxic metal pollutants.
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 the data presented in Appendix B of
this report and the JRB Study (52).
B. TSS: For NSPS, the proposed BPT limitation is
retained. Addition of sulfide precipitation is not anticipated
to significantly improve or degrade the suspended solids since
this treatment is not specifically intended to improve TSS
removal efficiency. Therefore, the 30-day averaqe limitation of
3.9 kg/kkg is retained based on the 30-day average long-term
monitoring data (Section 16.7.2).
Nonconventional pollu tants - The only nonconventional
pollutant considered is iron. Iron should be controlled
adequately by the proposed treatment technology and is included
in Table 16-17 on a concentration basis only. The proposed
concentration basis is presented as auidance in cases where iron
may be of serious concern.
Toxic pollu tants - The addition of sulfide treatment to the
proposed base level treatment is anticipated to provide more
stringent removal of toxic metals. The proposed NSPS
limitations are based on literature treatability estimates
605
-------
TABLE 16-17. PROPOSED LIMITATIONS
Chrcme Pigments
New Source Performance Standards
Waste Water Flew: 105 m3/kkg
Pollutant Treatability VFR
Concentration Basis Effluent Limit
(D rog/1 (kg/kkg)
, /n Max Max
vmg/J-' 30-day 24-hr 30-^ay 24-hr
Avg Max Avg Max
Conventional and Nonconventional
Pollutants:
Total Suspended
Solids, TSS
23 (3)
2.4
37
89
3.9
9.4
Iron
0.30(3)
2.4
0.49
1.2
__(5>
__(5>
Toxic Pollutants:
* i.- (2)
Antimony
0.40(3)
2.4
0.40
0.96
0.042
0.10
Cadmium^
0.01(4)
2.4
0.010
0.024
0.0011
0.0026
Chromium^
0.05(3)
2.4
0.05
0.12
0.0053
0.013
Copper(2)
0.05(4)
2.4
0.05
0.12
0.0053
0.013
Lead(2)
0.05(4)
2.4
0.05
0.12
0.0053
0.013
Mercury ^ ^
o.oi(4)
2.4
0.01
0.024
0.0011
0.0026
(2)
Nickelv '
0.05(4)
2.4
0.05
0.12
0.0053
0.013
Zinc(2)
0.02(4)
2.4
0.02
0.048
0.0021
0.0050
(1) VFR: ratio of the 24-hour variability factor to the 30-day variability
factor.
(2) Also applicable to PSNS limitations.
(3) Proposed BPT limitations are retained.
(4) Lower limit of literature treatability as per the discussion in Section 8.1
and presented in Table 8-11.
(5) No effluent limitation proposed.
606
-------
(Table 8-11) since no plant in the industry currently utilizes
sulfide precipitation of metals on which to base specific
numerical limitations.
The variability factor ratio (VFR) for the pollutants of
concern are retained from the BPT limitations. The VFR is based
on the Titanium Dioxide Subcategory for similar pollutants.
A. Chromium: The proposed limitation for chromium is
based on the literature treatability estimate of 0.05 mq/1,
since sulfide treatment is not expected to improve significantly
the removal efficiency other than coincidental removal. The 30-
day average limitation is therefore,
(0.05 mq/1) (105 m3/kkq) ( kg/m3 \ = 0.0053 kg/kkq
\1000 mg/1)
and the 24-hour maximum limit becomes,
(0.0053 kg/kkg) (2.4) = 0.013 kg/kkg
B. Zinc: The 30-day average zinc concentration is
expected to achieve 0.02 mg/1 in view of the proposed technology
basis and treatability values as was reported in the literature
(Table 8-11) . The proposed load limitation is then,
(0.02 mg/1)(105m3/kkg) ( kg/m3 \ = 0.0021 kg/kkg
\1000 mg/ly
The 24-hour maximum concentration becomes,
(0.0021 kg/kkg) (1.9) = 0.0040 kg/kkg
C. Lead: The 30-day average lead concentration is
expected to achieve 0.05 mg/1 in view of the proposed technology
basis and treatability values reported in the literature (Table
8-11). The proposed load limitation is,
f0.05 mq/1)(305 m3/kkq) ( kg/m3 \ = 0.0053 kq/kkg
VJ.000 mg/1/
The 24-hour maximum concentration becomes,
(0.0053 kg/kkg) (2.4) = 0.013 kg/kkg.
D. Copper: The 30-day average copper concentration is
anticipated to achieve 0.05 mg/1 based on literature
treatability estimates. Therefore, the proposed limitation
becomes,
607
-------
(0.05 mg/l)(105 m3/kkg) / kq/m3 = 0.0053 kg/kkg
VlOOO mg/1)
The 24-hour maximum concentration is then,
(0.0053 kg/kkg) (2.4^ = 0.013 kg/kkg.
E. Antimony: The proposed BPT limitation for antimony is
retained since sulfide treatment is not expected to improve
significantly the removal of efficiency other than coincidental
removal. Therefore, the proposed 30-day average limitation is
0.042 kg/kkg in Table 16-17.
(0.40 mg/1) (105 m3/kkg) / kg/m3 "N = 0.042 kq/kkg
VI000 mg/1/
The 24-hour maximum is then,
(0.042 kg/kkg) (2.4) = 0.10 kg/kkg.
F. Cadmium: The 30-day average cadmium concentration is
anticipated to achieve 0.01 mg/1 based on literature
treatability in Table 8-11. Therefore, the proposed limitation
becomes,
(0.0.1 mg/1) (105 m3/kkg) f kg/m3 \ = 0.0011 kg/kkg
^1.000 mg/1/
The 24-hour maximum concentration is,
(0.0011 kg/kkg) (2.4) = 0 .0026 kg/kkg.
G. Nickel: The 30-day average nickel concentration is
expected to achieve 0.05 mg/1 based on literature treatability
estimates in Table 8-11. Therefore, the proposed limitation
becomes,
(0.05 mg/1)(105 m3/kkg) / kg/m3 \ = 0.0053 kg/kkg
\ 1000 mg/1/
The 24-hour maximum becomes,
(0.0053 kg/kkg\ (2.4) = 0.013 kg/kkg.
H. Mercury: Sulfide precipitation of mercury can achieve
approximately a 0.01 mg/1 concentration based on literature
treatability. Therefore, the proposed 30-day average load
limitation is,
608
-------
(0.010 mg/1)(105 m3/kkg) ( kg/m3 "N = 0.0011 kq/kkq
V1000 mg/1 )
The 24-hour maximum limitation is,
(0.0011 kg/kkg) (2.4) = 0.0026 kg/kkg.
16.7.6 Basis for Proposed Pretreatment Standards
Existing Sources
There are currently nine indirect discharge chrome pigment
plants in the subcategory. For Pretreatment Standards for
Existing Sources (PSES), the Agency is proposing limitations
based on BAT described in Section 16.7.4. The pollutants to be
limited are chromium, zinc, lead, copper, antimony, cadmium, and
nickel as presented in Table 16-16.
New Sources
For Pretreatment Standards for New Sources (PSNS), the
Agency is proposing limitations based on NSPS. The pollutants
are indicated in Table 1.6-17.
609
-------
Intentionally Blank Page
-------
SECTION 17
HYDROGEN CYANIDE INDUSTRY
17.1 INDUSTRY PROFILE
17.1.1 General Description
Over 50 percent of the Hydrogen Cyanide manufactured is
produced by the Andrussow process, while about 40 percent is a
by-product from acrylonitrile manufacture. A major portion of
the production is used in the manufacture of methyl
methacrylate, plexiglass molding and extrusion powders, and
surface coating resins. It is also used as a fumigant for
orchards and tree crops. The industrial data profile for this
industry is given in Table 17-1, while the status of regulations
is given in Table 17-2.
17.1.2 General Process Description and Raw Materials
The hydrogen cyanide subcategory in this study is confined
to the Andrussow process, in which air, ammonia and methane are
reacted to produce hydrogen cyanide.
The raw materials are reacted at elevated temperatures
(900-1000 degrees C) over a platinum catalyst. The reaction is
given as:
2CH4 + 2NH3 + 302 = 2HCN + 6H20 (1)
The source of methane is natural gas containing 50 to 100
percent methane by volume. In addition to hydrogen cyanide, the
reacted gases contain ammonia, nitrogen, carbon monoxide, carbon
dioxide, hydrogen and small amounts of oxygen, as well as traces
of organic nitriles formed from nonmethane hydrocarbon
components of natural gas. The reactor gases are cooled and
then scrubbed in one of two processes which are used to remove
the unreacted ammonia. In one patented process the gases are
scrubbed with phosphate liquor, the resulting solution is
decomposed and the phosphate solution is recirculated. The
recovered ammonia is recycled to the reactor. In the second
611 «
Preceding page blank
-------
TABLE 17-1.
SUBCATEGORY PROFILE DATA SLNMARY
SUBCATEGORY
HYDROGEN CYANIDE*
Total subcategory capacity rate
Total subcategory production rate
No. of plants in this subcategory
Plant age range:
Minimum
Maximum
308 Data** on file for
With total capacity of
With total production of
Representing capacity
Representing production
Average production
Average capacity utilization
Waste water flew per unit product
Minimum
Maximum
289,000 kkg/year
165,500 kkg/year
7
5 years
30 years
2
178,500 kkg/year
115,500 kkg/year
62 percent
70 percent
57,750 kkg/year
65 percent
10 mVkkg of HCN
57 mVkkg of HCN
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Corrmerce, 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.
* Lncludes data from plants using Andrussow Process and from plants
recovering HCJC as a byproduct from the manufacture of acrylor.itrile.
**Includes data from plants using Andrusscw Process.
-------
TABLE 17-2. STATUS OF REGULATIONS - EFFLUEOT LIMITATION GUIDELINES
SUBCATEGORY
SUBPART
HYDROGEN CYANIDE
AP (40 CFR 415.420, 5/22/75)
STANDARDS
Product
Process
3PCTCA*
M 1 2
Max. Avg.
P=r=m- kg/kkg kg/kkg
eterT (rv3/1} (m9/1>
BATEA
Max. Avg.
kg/kkg kg/kkg
(mg/1) (mg/1)
NSPS
Max. Avg.
kg/kkg kg/kkg
(mg/1) (mg/1)
Andrusscw
Process
TSS
CN
CN (A)
BOD.
nh3-n
2.4
1.2
(48.0)**
(24.0)
0.005
0.025
(1.0)
(0.5)
0.005
0.005
(0.1)
(0.05)
3.6
1.8
(72.0)
(36.0)
0.36
0.18
(7.2)
(3.6)
~
Sections 415.420, 415.421, and 415.422 were revoked by the Agency
( 41 FR 10681, February 23, 1977).
. = Maximum of any one day.
Avg. = Average of daily values for thirty consecutive days shall not exoeed.
**
flew basis 50,000 1/kkg.
613
-------
process sulfuric acid is used to absorb ammonia from the reactor
gases. At one plant the resulting ammonium sulfate solution is
used for the manufacture of another product.
The hydrogen cyanide is removed from the ammonia scrubber
effluent gases by absorbtion in cold water, and the waste gases
are vented to the atmosphere. The absorbed solution containing
hydrogen cyanide, water, and other contaminants is distilled to
produce HCN gas of over 99 percent purity.
The water produced during the initial reaction (Equation 1)
of the formation of hydrogen cyanide is purged with the
distillation bottom stream and is either recycled to the
absorber or discharged to the treatment facility. In order to
be recycled, the distillation bottom water has to be cooled by
refrigeration prior to reuse in the HCN absorber unit. At plant
locations where cold water is readily available in large
quantities, it can be used on a once-through basis with a
significant savings in energy costs. Figure 17-1 presents a
general block diagram for the manufacture of hydrogen cyanide by
the Andrussow process.
17.2 WATER USE AND WASTE SOURCE CHARACTERISTICS
17.2.1 Water Use
Water is used in noncontact cooling in the absorber, pump
seal quenches, flare stack flushes, for washdown and cleanup of
tank cars, for absorption of the product from reactor gases and
for washing equipment and cleaning up leaks and spills. Table
17-3 gives the detailed water consumption at one plant and also
the total consumption at two plants. There is a pronounced
difference in water usage at these two plants due to the use of
refrigeration at Plant £782 which makes possible the recycling
of absorber water from the distillation unit back to the
absorber. This practice is energy intensive but is required in
locations where an abundant supply of cool water is not
available. Plant f765 has such a supply and uses absorber water
on a once-through basis. In this case, a much larger flow must
be treated prior to discharge.
17.2.2 Waste Sources
The following are sources of waste water produced from the
manufacture of hydrogen cyanide by the Andrussow process:
Distillation Bottoms
The waste water contains ammonia, hydrogen cyanide and
small amounts of organic nitriles. The water consists of the
water produced by the reaction plus scrubber water used for the
614
-------
cold vnrr
Htm CASES
ACID 4
1 xl
7
MtCNIA
ABSOf&ffl
1CM ABSOftfTICN
90CTC*
DI9IU1ATIGN
o> WR ^
tn
usa> POM HB
Mwureciuw op
onot ncoucrs
OR MnClED.
a blqi) is son
to KB wsn
iMKnetr rum.
TIB DlSTTUAnOM BCTIOM
IS EITIKJt MCtOiS (A
MO IS DISdlMCED) OR
SWT ID UK TREADOtr
FUdUTY.
HCM PKXUCT
Figure 17-1. General process flew diagram for production of hydrogen cyanide
by the Andrussow process.
-------
TABLE 17-3. WATER USAGE IN HYDROGEN CYANIDE - ANDRUSSCW PROCESS
SUBCATEGORY
Plant Water Usage, (mV^kg of HCN)
Total Consumption Noncontact Cooling
?f782(1) 29.5 18.9
?765 58.3 8.00
(1)
Detail water usage (mVkkg) at Plant #782 is:
Xoncontact cooling = 18.9
Direct process contact = 7.45
Indirect process contact = 0.71
(pumps, seals, leaks,
spills, etc.)
Maintenance, e.g. cleaning = 0.31
and work area washdcwn
Xoncontact ancillary uses = 0.67
(boilers, utilities, etc.)
Exported steam = 1.44
616
-------
absorption of HCN. The absorption water bottoms are either
recycled to the HCN absorber or discharged to the treatment
facility. Even if the distillation bottom stream is recycled to
the absorber, a portion of it is discharged to stop the buildup
of impurities.
Scrubber Streans
If the ammonia scrubber liquid is recycled, a portion of it
has to be purged to control the accumulation of impurities. The
bleed contains the acid used from scrubbing and minor amounts of
organic nitriles. The scrubber solution can also used for the
manufacture of other products in which case nothing is
discharged to the treatment plant.
Other Waste Water
This includes leaks and spills, equipment and tank car
washings, noncontact cooling water blowdown and rainfall runoff.
The tank cars are washed out with dilute acid or alkali to
remove any contaminants present, which, if allowed to remain in
the tank car, can polymerize the hydrogen cyanide causing safety
hazards due to possible explosion during shipment. The
noncontact cooling water may be contaminated with the product as
a result of leaks. The recirculated cooling water is monitored
for cyanide and the cooling tower blowdown is discharged to the
waste water treatment facility. During shutdown, the equipment
is drained to avoid freeze-up and the resulting waste water is
discharged to the treatment facility.
The quantity of waste water produced and treated at two
plants producing hydrogen cyanide by the Andrussow process is
given in Table 17-4. The large variation in flow exists because
the water used to absorb the hydrogen cyanide from the reactor
gases in Plant f765 is not recycled. As discussed earlier, that
plant is situated where sufficient cold water is available for
once-through use. Since the cold water is readily available at
a low cost, the water used for absorption is discharged. A
similar plant practicing recycling, in the absence of available
cold water, can achieve a total waste effluent of 7.1 m3/kkg of
HCN.
17.3 DESCRIPTION OP PLANTS VISITED AND SAMPLED
17.3.1 Screening
Plant f765 was visited and the waste water sampled during
the screening phase of the program. The combined wastes consist
of distillation bottoms, ammonia recovery purge liquor, tank car
washings, leaks, spills and equipment clean out, purge fronr^the
617
-------
TABLE 17-4. WASTE FLOW DATA FOR HCN PRODUCTION BY THE ANDRUSSCJW
PROCESS
Plant Total waste going to the treatment facility (mVkkg)
#765 57
#782 9.9*
The breakdown and flow of the different waste streams comprising the total
is given below:
Source Unit Flow (mVkkg)
Recovery and purification 6.3
Pump seal quenches 0.58
Flare stack flushes 0.09
Sanple hoods 0.02
NH3 stripper caustic 0.24
Steam condensate from NH^ stripper 0.90
Freeze protection 0.06
Washdowns and cleanup 0.25
Boiler blcwdown and condensate 1.48
618
-------
noncontact cooling water system and stormwater runoff. These
combined wastes are commingled with the other cyanide product
waste waters and sent to the alkaline chlorination treatment
facility. The first unit of the treatment facility is a trench
where the p>H of the waste water is raised to the range of 8.5 to
11 with dilute caustic soda. The caustic is added under
controlled mixing conditions with continuous automatic pH
recording and caustic feed adjustment. The pH-adjusted waste
water is sent to two 8-hour retention ponds. Chlorination is
accomplished by adding sodium hypochlorite at the pond entrance.
The chlorinate waste water from the 8-hour ponds are alternately
discharged to another small pond having one hour of detention
and equipped with baffles and agitators. Caustic and chlorine
are added as required in the one-hour pond to achieve the low
levels of cyanide desired. The effluent from the pond is
discharged to a POTW. The pond contains a flow
controller/analyzer, which will block the discharge from the
pond when a high cyanide level is detected in the treated
effluent. Figure 17-2 is a flow diagram of the treatment
process indicating the sampling location used during the
screening program.
Composite sampling conducted consisted of one 48-hour
composite sample for nonvolatile organics, metals and mercury
and one 24-hour composite sample of B0D5, TSS, TDS, NH3, Fe, Cr,
Zn, Cu and settleable solids. Grab samples for volatile
organics, cyanide, phenols, temperature and pH were collected on
two consecutive days at each sampling location. Table 17-5
gives the flow data and concentration and unit loads of ammonia-
nitrogen, total cyanide and thallium, for the sampled streams.
It is believed that thallium is not contributed by the hydrogen
cyanide manufacturing process.
17.3.2 Verification
Plant |765 was sampled again in the verification phase.
One additional stream of hydrogen cyanide waste water was
sampled in the verification phase at a point upstream of mixing
with other cyanide product waste water. This stream is
identified in Figure 17-2. The variation in the flow of the
streams in the two sampling phases was small. Table 17-6 gives
the flow and pollutant data of the sampled streams.
The second hydrogen cyanide plant sampled in the
verification phase was Plant f782. The waste water from the
hydrogen cyanide plant mainly consists of blowdown from the
distillation column which is combined with a portion of the
other product waste water and sent to an ammonia stripper.
Effluent from the ammonia stripper is mixed with the rest of the
process waste water from other products and sent to a single
stage biological system. The primary treatment facility
consists of oil skimmers, grit removal and pH adjustment. The
619
-------
HYDROCtN CYANIDE
WASTE WATER
OIHFR CYANIDE PHUXJCT
VftSTE WATER
(ii AUJUSlMm*
RIIJLrrt CAUSTIC
SODIUM
HYKXHLORITC
CAUSTIC'
FOND
Vteste strtams sdn«iled
FINAL TREATED
FFFLUQir
Waste streaii was
san))led in the
verification program
since it is free frun
other cyanide wastes.
Figure 17-2. General waste water treatment process flow diagram at plant #765
shewing the sampling points. (Hydrogen cyanide manufacture.)
-------
TABLE 17-5. FLOW AND POLLUTANT DATA OF TKF RAV' AND TRFATED WASTE
STREP'S OF PLANT <1765 PRODUCING HYDROGEN CYANIDF BY
AMJPUSSO-; PROCESS
Stream Unit NH^-N rvanf'V Thallium
Description Flow . . (1) . ^ ,. (1) /iv^\ /L, (X)
(m^Akg) (?) (^j (?) (£gj (f)
12 r x / x f(3)
Influent 51(2> 7.8 4.4 (3) 107 6.1 (3> .028 0.0016
to
Treatment
Treatment 57(2) 35 2.0 (3) 0.36 0.02 (3) .010 0.00057
(Alkaline
Chlorination)
Effluent
' Unit Load = Unit Flew (57 m^ j x pollutant x /1000 mg/l>
i/ii kkg' concentration I , ,3 .
in kg/kkg y ^ nyg/1 V kg/m-* V
(2) The stream is a cortmingled waste water. The flew given is the
amount contributed by the HCN process.
(3) The pollutant load was calculated by apportioning the mass emitted
between the two waste streams on the basis of measured flews. This
is clearly a very approximate process and the results must be used with
caution.
621
-------
TAHUi 17-6. FLOW AND CQLIi/TAf/T (DNTTWrKATION DATA Of' TIE SAMPLED WASTE STHEAMS KX< 1'IJW #765 I'WJtWINC
UYTXOfHJ CYAN UK
0.82
0.39 (2>
NA
1.6
1.6
0.00015
(1)
(2)
(3)
Tt>e bticam is a ouniuir»jltxl waste water. The flow given is the amount contributed by the IICN process.
Hie [ol lutoiit load was calculated by apjxirtioning the mass emit tod between the two waste stieams on
the basis ot measured flows. Ttlis is cloarly a very ajjpioxinute process and tJ>e results must be used
with cautiCr..
1tie addition or Iocs of water from rainfall, addition of chemicals and evaporation has not been
est invited.
NA = Not Available
-------
effluent from primary treatment goes through an API separator
and into an aerated lagoon. Effluent from the lagoon is
flocculated and sent to a clarifier. The overflow from the
clarifier is sent to a final settling basin before final
discharge. The surface drainage consisting of runoff, wash
down, etc., from the hydrogen cyanide and other process areas is
collected separately. The water is sent first to a surface pond
where it undergoes a two-stage pH adjustment and then is piped
to a trickling filter. It then merges with the treated process
waste waters in the clarifier. A general flow diagram of the
treatment process including streams sampled is shown in Figure
17-3.
Table 17-7 gives flow and concentration data of the sampled
streams. In Table 17-8, the unit waste flow and unit pollutant
loads are given for the raw and treated effluent. Because of
intermixing of various product waste water streams, the unit
pollutant loads (especially for treated effluent) were
calculated based on hydraulic loadings and the method used is
only an approximation. The principal process waste water from
the hydrogen cyanide plant is the waste from the recovery and
purification operation and has a loading of 6.3 m3/kkg of HCN.
The total waste water going to the treatment facility from the
hydrogen cyanide plant has a loading of approximately 9.9 m3/kkg
of HCN, consisting of both process contact and noncontact
effluents.
In calculating the pollutant loads, (Table 17-8) the loss
or gain of water to the treatment system such as evaporation,
loss through filtered solids, precipitation and the water
introduced by treatment chemicals has not been included because
it was considered insignificant in comparison to other factors.
17.3.3 Toxic Pollutant Concentrations
Total cyanide and thallium were the toxic pollutants
detected in the raw waste from Plant f765 which was sampled in
the screening phase. It is believed that thallium in the waste
water is not contributed from the hydrogen cyanide process.
The HCN waste water at Plant f765 is mixed with other
product waste waters and the combined flow was sampled upstream
of the treatment system. It is probable that thallium is
contributed from these other product waste waters.
The raw waste stream was not analyzed for free cyanide.
The same plant was sampled again with another plant in the
verification phase. In addition to total cyanide, free cyanide
was found in significant concentrations in the raw process waste
sources from the two HCN plants. Free cyanide in the waste
623
-------
#1
DISTILLATION
BOTTOM PURGE
n
OTHER PRODUCT
WASTE WATERS
AMMONIA
STRIPPER
5
#3
i
A
PRIMARY
TREATMENT
OTHER PRODUCT
WASTE WATER
SURFACE DRAINS
BIOLOGICAL
TREATMENT
CLARIFIER
SETTLING
POND
O'5
DISCHARGE
CHEMICAL AND
BIOLOGICAL
TREATMENT
LEGEND
^ SAMPLING POINTS
Figure 17-3. General waste water treatment process flew diagram at plant #782
shewing sampling points. (Hydrogen cyanide manufacture.)
624
-------
TABLE 17-7. FLOW AND POLLLTAKT CONCENTRATION DATA OF THE SAMPLED WASTE
STREAMS FOR PLANT #782 PRODUCING HYDROGEN CYANIDE
Stream Waste Flew CN (T)
No. Stream rcr'day
Description
CN (F)
(mg/1)
nh3-n
TSS
1 Distillation(6.3)^
bottom purge
2 Armenia strippir^ 5400
influent
(3)
3 Airmonia stripper 5400
effluent
4 Influent to^
primary treatment
facility
5 Final treated
effluent
(3)
6400
NA
71
167
51
31
2.2
62
145
41
886
410
41
7.0
1.7
1380
5.6
24
76
162
110
74
(1)- The total waste is composed of the blowdcwn frcm three distillation
columns. Three 24-hour composite samples were collected for each unit.
The pollutant concentration value (given in mg/1) is an average of the
three composited sarrples for the three waste stream sources.
3
(2)- The value given is the total unit flow in m /kkg of HCN for the three
purge streams.
(3)- The stream is a combined waste water. It includes the waste effluents
from hydrogen cyanide and other products.
625
-------
TABLE 17-8. UNIT FLDW AND UNIT POLLUTANT LOADING FOR RAW AND
TREATED WASTE EFFLUENTS AT PLANT #782
Stream
Unit
Flow
(m3/kkg)
Unit Pollutant Loadina (kg/kkg) ^
Total Free Armenia- Total
Cyanide Cyanide N Suspended
CN
(T)
CN
(F)
nh3-n
Solids
TSS
Process raw
waste water
(distillation
bottan purge)
Process
waste water
treated
effluent
Total HCN
waste water
treated
effluent
(2)
6.3
6.3
(2)
9.9(3)
0.45
0.014
0.022
0.39
0.011
0.017
5. 6
0.035
0.055
0.15
0.47
0.74
(1) Unit pollutant load = unit flew pollutant concentration , kq/n3 \
(m^/kkg) (in ma/1 frar. Table 17-7) ^1000 mg/y
(2) The pollutant load was calculated by apportioning the mass emitted from
the total treated effluent (which includes other product waste water) on
the basis of measured flew contributed by the HCN process. This is clearly
an approximate process and the results must be used with caution.
(3) Ihe waste water flew consists of direct process contact and noncontact
effluent from the HCN plcint going to the treatment system.
626
-------
water consists of hydrogen cyanide, sodium or potassium cyanide
and cyanogen chloride which may be present as a result of
chlorination (especially in the treated effluent). Total
cyanide includes the free cyanide and cyanides found in metal
complexes (such as sodium ferrocyanide or sodium ferricyanide).
No toxic organic pollutants were found in significant
concentrations in the HCN plant raw waste sampled. The
concentrations of the toxic pollutants found in the raw waste
water in the screening and verification were:
Maximum Raw Waste Concentration Observed
(ug/l)
Screening Verification
Pollutant Plant f765 Plants f765, f782
Thallium 25 Not Determined
Cyanide (Total) 166,000 186,000
Cyanide (Free) Not Determined 172,000
The general sampling methodology used in the screening and
verification program is described in Section 5.1.2. A total of
nine days of sampling was conducted at Plants *'765 (sampled
twice) and £782. Thirteen waste water sampling points were
involved which included the raw waste water, combined waste
water and combined treated effluent streams. The evaluation of
the toxic metal and toxic organic pollutant content of these
process streams was based on total analytical data points from
both the screening and verification phases.
The daily toxic pollutant waste load in the raw waste was
calculated from the effluent waste flow rate and the measured
pollutant concentration of the toxic pollutant.
This is given by:
Daily loading (as kg of pollutant per day) = (C)(Q)
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). Similarly, the unit loadings were calculated
from the reported hydrogen cyanide production rate, the
waste stream flow rate, and the measured pollutant
concentration.
627
-------
Unit loading (as kg of pollutant _ (C)(Q)
per kkg of hydrogen cyanide) ~ 1000P
Where C and Q are the same as described above, and P is the
hydrogen cyanide production rate expressed in units of
kkg/day (kkg is 1000 kg, a metric ton, which is equal to
2205 lbs).
In the case of two or more process waste streams going to
the treatment system the daily raw waste load of the toxic
pollutant was calculated by determining the combined pollutant
load of the individual streams.
The unit raw waste loading for a pollutant (toxic,
conventional or nonconventional) was calculated by dividing the
daily pollutant load with the average daily production of
hydrogen cyanide at the plant.
Unit pollutant Pollutant Load (in kg/day)
Load in the raw waste = Average Daily HCN Production
(kg/kkg of HCN) (kkg/day)
Table 17-9 gives the toxic, conventional and
nonconventional pollutant loadings of the raw waste for Plants
#765 and #782 which were sampled in the screening and
verifcation phases. The overall average pollutant loads for the
sampled plants are given in the last column of the table.
The approximate toxic pollutant generated per year by the
entire subcategory is estimated by multiplying the overall
average unit pollutant loading (Table 17-9) with the hydrogen
cyanide subcategory production from Table 17-1 (165,500 kkg/yr),
Pollutant Waste Load (kg/year)
Cyanide (Free) 100,000
Cyanide (Total) 450,000
17.4 POLLUTION ABATEMENT OPTIONS
17.4.1 Toxic Pollutants of Concern
The toxic pollutants of concern in the HCN raw waste are
free (or oxidizable) cyanide and total cyanide. No organic
toxic pollutants of significance were found in the raw waste of
the sampled plants.
17.4.2 Process Modifications and Technology Transfer Options
Process
subcategory.
modifications
have not
628
been identified for the
-------
TABLE 17-9. SUMMARY OF PQLLUTAOT RAW WASTE LOADING FOUND IN SCREENING AND
VERIFICATION SAMPLING
SUBCATEGORY
HYDROGEN CYANIDE
Average Daily Pollutant Loading and Concentrations at Plants Sampled
kg/kkg of HCN
(mg/1)
Pollutant
#765(s)
# 765(v)
# 782(v)
Overall
Average
TOXIC
Free Cyanide
NA
Total Cyanide 6.1
(110)
Conventional
and Nonconventional
0.82
(14)
1.6
(29)
0.39
(62)
0.45
(71)
0.61
2.7
TSS
nh3-n
NA
4.4
(78)
2.0
(35)
27
(480)
0.15
(24)
5.6
(890)
1.1
12
(S)
(V)
Sampled in screening phase
Sampled in verification phase
629
-------
17.4.3 Best Management Practices
No best management practices have been identified for the
subcategory.
17.4.4 Prevailing Control and Treatment Practices
Out of a total of seven plants currently producing hydrogen
cyanide by the Andrussow Process, 308 data is available for only
two. The production at these two plants constitutes more than
70% of the total subcategory production. Since the two plants
produce a significant amount of the total subcategory
production, their waste water treatment technologies are taken
as the subcategory treatment practices. The two plants were
visited to review the treatment systems and to collect waste
effluent samples.
Plant #765 has a high volume effluent because the water
used to absorb the reactor gases is not recycled since low cost
cold water is readily available at the site. The waste water
consisting of scrubber purge, absorption water, and plant run-
off is mixed with other cyanide product waste waters and sent to
an alkaline chlorination system. The pH of the waste water is
raised to about 10 with dilute caustic in a small pond which has
a retention time of two hours and then it is discharged to two 8-
hour ponds where sodium hypochlorite is added to oxidize the
cyanide to cyanate. The chlorinated waste water is transferred
to a small pond equipped with agitators and baffles before final
discharge to a POTW. Caustic or chlorine is added to the final
pond to achieve the desired low levels of cyanide. The
treatment system is shown in Figure 17-2.
Plant ?782 uses a single-stage biological treatment system
for the treatment of effluent from the hydrogen cyanide plant.
The process waste water from the HCN plant consists mainly of
distillation column blowdown and is combined with other cyanide
product waste water and sent to an ammonia stripper. The
effluent from the stripper combines with other product waste
waters and is treated by means of an oil separator, a grit
chamber, a compactor, a second API separator, an' aerated lagoon,
a flocculator and a final clarifier. The overflow from the
clarifier is sent to the final settling basin before discharge.
The run-off from the HCN plant and other product manufacturing
areas is combined and sent to a pond for a two-stage pH
adjustment. The effluent from the pond is treated by a
trickling filter and clarifier, and the clarifier effluent is
mixed with the treated process waste water. A general block
diagram of the treatment system is shown in Figure 17-3.
17.4.5 Advanced Treatment Technologies
The three pollutants of concern in hydrogen cyanide plant
effluents are cyanide, ammonia and chlorine. The treatment
630
-------
technologies for cyanide removal include alkaline chlorination,
biological treatment, ozonation, wet air oxidation,
electrolytic decompostion, wet thermal decompostion,
acidification, activated carbon, permanganate oxidation, lime
reaction with sulfur, radiation, evaporative recovery,
catalytic oxidation ana ion exchange. Except for alkaline
chlorination and biological treatment, the remaining treatment
technologies are not effective or advantageous for one or more
of the followina reasons:
A. The technology has low cyanide removal efficiency.
B. The technology cannot treat waste water with high
cyanide concentrations.
C. The technology has air pollution problems.
D. The technology has high operating costs.
The free cyanide in the raw waste is readily oxidizable and
exerts a chlorine demand. Sufficient chlorine is added to react
with ammonia and to oxidize cyanide. The presence of large
amounts of ammonia will increase the cost of chlorination. if
costs are too extensive, residual ammonia in the raw waste
effluent can be reduced by steam or air stripping before
alkaline chlorination to reduce the amount of chlorine required.
17.5 SELECTION OF APPROPRIATE TECHNOLOGY AND EQUIPMENT
17.5.1 Technologies for Different Treatment Levels
Level 1 (BPT)
Two-stage alkaline chlorination followed by pH adjustment
was chosen for the removal of cyanide from the raw waste
effluents. The technology is being practiced in the industry.
The flow diagram of the treatment system is shown in Figure 17-
4.
Level 2 (BAT)
The treatment is the same as BPT (Level 1) except that
residual chlorine is reduced to a lower level by treatment with
sulfur dioxide. Chlorine in adequate amounts is added to remove
ammonia and to oxidize cyanide. Where practiced, steam or air
stripping of ammonia has not been considered as a part of the
treatment system since the value of the recovered ammonia is the
justification for doing it. It has been assumed to be process
631
-------
RAW
WASTE-
HATER
U>
N>
CAUSTIC 300ft
&
(*)
CHLOKDC
^-1
1
ri)
H7LOINH AM) 1ST
SI7\GE AIRLINE
Ofl/KSATION
JTXXMD SIMX
ALRM.INE aijORINATION
Q pi! Aliltl?
<$)
AIMIIRIMFNT
EFFLUJOT
Incltidt-s flow monitoring, pH monitoring And sampler.
Figure 17-4. Level 1 waste water treatment for hydrogen cyanide subcategory.
-------
related. The general flow diagram of the treatment process is
given in Figure 17-5.
17.5.2 Equipment for Different Treatment Levels
Equipment Functions
In level 1, the raw waste water enters a holding tank
equipped with an external pump and recirculation system.
Caustic soda and chlorine are added and the tank contents are
mixed by the recirculation pump. Following this first stage
alkaline chlorination, the waste water is chlorinated further in
a second tank which is equipped with automatic pH control The
final effluent is neutralized to pH 6-9 before discharge. In
Level 2, using the same equipment as in Level 1, the chlorine
feed to the second stage alkaline chlorination system is
increased. To remove excess chlorine before release, sulfur
dioxide is fed by a modified gas chlorinator, with oxidation-
reduction potential control. As in Level 1, the effluent is
then adjusted to pH 6-9 before discharge.
Chemicals and Handling
Caustic soda solution, chlorine, sulfur dioxide, and
sulfuric acid are used in the waste treatment process. Caustic
soda and sulfuric acid are common industrial chemicals which
pose no special hazards when handled by conventional corrosion-
resistant feeding equipment. Chlorine and sulfur dioxide are
received in one-ton containers as compressed gases, and are fed
as water solutions by vacuum-controlled equipment designed for
the specific chemical. No unusual chemical feeding or handling
problems are anticipated, provided precautions are taken to
prevent gas leaks and to guard against corrosive attack.
Separation and Removal of Solids
Since few solids are produced in the treatment process,
there is no significant sludge disposal problem.
Monitoring Requirements
Internal process monitoring is done largely with automatic
sensing and control equipment for regulating pH and
chlorine/sulfur dioxide residuals. Field tests for cyanide
and/or chlorine in the effluent should be made regularly by the
operator, and 24-hour composite effluent samples should be
collected and analyzed for cyanide as required in local or NPDES
permi ts.
633
-------
CHLORINE
CAUSTIC
SODA
RAW
waste WATER *
*Q>—1
HOLDING AND 1ST STAGE
ALkAIJNE CHLORINATION
6
I
| SULFUR
| DIOXIDE
.SECOND STAGE
ALKALINE CHLORINATION
Q>H ADJUSTMENT
OKP
1-MD-
-EFFLUENT
Includes flow monitoring. pM monitoring .ind sampler.
ORP Oxidation Rechjct K>n Potential Control
Figure 17-5. Level 2 waste water treatment for hydrogen cyanide subcategory.
-------
17.6 TREATMENT COST ESTIMATES
17.6.1 General Discussion
A model plant concept was developed as a basis for
estimating treatment costs. For conceptual design a
representative unit waste flow (cubic meters per kkg of HCN) was
selected, together with three different HCN production rates.
The latter were chosen to cover most of the subcategory
production range. The selected daily HCN production for the
model plant was multiplied by the selected unit flow to obtain
the volume of waste water passing to the treatment system. The
selected unit raw waste pollutant load was 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.
Waste Water Flow
The unit process waste water flow for the two plants
visited in this study are 6.3 m3/kkg of HCN (Plant f782) and 57
m3/kkg of HCN (Plant f765). The difference results from the
different absorption water discharge practices at the two
plants. (See Section 17.2.2). The model plant has been
developed using the larger unit flow rate of 57 cubic meter/kkg
of HCN, since this is a more conservative approach. The Agency
considered developing effluent limits for two different levels
of flow but rejected it because of the cost, complexity, and
difficulty in implementing the approach.
For waste water treatment cost estimates, three production
levels were selected for the model plant. These are 31,800,
50,900 and 63,600 kkg/yr.
Waste Water Pollutant Load
The three pollutants of concern in the subcategory are
cyanide (oxidizable and total), ammonia and chlorine. Chlorine
is not present in the raw waste but is added during alkaline
chlorination treatment. The average value of 0.61 kg of free
cyanide/kkg of HCN and 12 kg of NH3/kkg of HCN (Table 17-9)
developed from the screening and verification results were used
for the model plant raw waste loads.
Chemicals Used
At the BPT level of treatment, alkaline chlorination
requires 33 kg of chlorine and 5.0 kg of caustic per kkg of HCN.
For BAT treatment, 9.0 kg of S02 per kkg of HCN is used for
635
-------
dechlorination in addition to the chemicals used for BPT
treatment.
Solids Generated
Few, if any, solids are produced in treating HCN production
wastes.
The costs shown in Table 17-10 at each level of treatment
correspond to BPT (Level 1) with incremental costs to meet the
more stringent BAT requirements.
The estimated costs for the three model plants at different
production levels are given in Table 17-10, 17-11, and 17-12.
As mentioned earlier, both the hydraulic and pollutant loads per
unit of production are held constant over the entire range of
production.
Annual treatment cost as a function of production and
treatment cost per ion of HCN produced are shown graphicallv in
Figures 17-6 and 17-7, respectively.
Table 7-13 presents a summary of the unit cost distribution
between amorization, operation and maintenance cost components
at various production rates and levels of treatment.
17.7 BASIS FOR REGULATIONS
17.7.1 Evaluation of BPT Treatment Practices
A total of seven plants produce hydrogen cyanide by the
Andrussow Process. At one facility the raw wastes from the
hydrogen cyanide plant is combined with the waste from an
organic cyanide product and sent to a biological treatment
system to reduce organic and cyanide pollutants. Five of the
other seven HCN producers (using the Andrussow Process) use
alkaline chlorination for treatment of raw waste effluents.
There is no available information concerning the treatment
practices at the other two plants.
17.7.2 Basis for Proposed BPT Limitations
Technology Basis
The predominant treatment practice for raw waste effluent
in the HCN subcategory is alkaline chlorination. The Agency is
therefore proposing BPT effluent limitations based on alkaline
-------
TABLE 17-10. MODEL PLANT TREATMENT COSTS
Subcategory HYDROGEN CYANIDE
Production 31,800 metric tons per year ( 35,059 tons per year)
90 metric tons per day (100 tons per day)
Waste water flew 5,100 cubic meters per day.
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT COST
Construction $65,500 $15,000
Equipment in place,
including piping,
fittings, electrical
vrork arid controls 810,500 120,000
Monitoring equipment
in place 9,000
Engineering design
and inspection 177,000 27,000
Incidentals, overhead,
fees, contingencies 177,000 27,000
Land 3,000
TOTAL INVESTMECT OOST $1,242,000 $18?,000
B. OPERATION AND
MAINTENANCE COST
Labor and supervision ... 84,000 14,000
Energy 9,000 3,100
Chemicals 296,000 97,000
Maintenance 123,900 18,900
Taxes and insurance 37,300 5,700
Residual waste
disposal
Monitoring, analysis
and reporting 15,000 7,500
TOTAL OPERATION AND
MAINTENANCE OOST $565,200 $146,200
C. AMORTIZATION OF
INVESTMENT OOST $202,100 $ 30,800
TOTAL ANNUAL OOST $767,300 $177,000
* First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
637
-------
TABLE 17-11. f'CDEL PLANT TREATMENT COSTS
Subcategory HYDROGEN CYANIDE
Production 50,900 metric tons per year (56,117 tons per year)
145 metric tons per day (160 tons per day)
Waste water flew 8,200 cubic meters per day.
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT COST
Construction $105,000 $ 20,000
Equipment in place,
including piping,
fittings, electrical
work and controls 1,246,500 120,000
Monitoring equipment
in place 9,000
Engineering design
and inspection 272,100 28,000
Incidentals, overhead,
fees, contingencies 272,100 28,000
Land 3,000
TOTAL INVESTMENT COST $1,907,700 $196,000
B. OPERATION AND
MAINTENANCE COST
Labor and supervision ... $ 84,000 $ 14,000
Energy 9/800 3,100
Chemicals 476,000 154,000
Maintenance 190,500 19,600
Taxes and insurance 57,200 6 500
Residual waste
disposal
Monitoring, analysis
and reporting 15,000 7,500
TOTAL OPERATION AND
MAINTENANCE COST $832,500 $204,700
C. AMORTIZATION OF
INVESTMENT OOST $310,400 $ 31,900
TOTAL ANNUAL COST $1,142,900 $236,600
* First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
638
-------
TABLE 17-12. MODEL PLANT TREATMENT OOSTS
Subcategory HYDROGEN' CYANIDE
Production 63,600 metric tons per year (70,119 tons per year)
181 metric tons per day (200 tons per cay)
Waste water flew 10,300 cubic meters per day.
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT COST
Construction $117,000 $40,000
Equipment in place,
including piping,
fittings, electrical
work and controls 1,505,000 160,000
Monitoring equipment
in place 9,000
Engineering design
and inspection 326,200 40,000
Incidentals, overhead,
fees, contingencies 326,200 40,000
Land 3,000
TOTAL INVESTMENT COST $2,286,400 $280,000
B. OPERATION AND
MAINTENANCE COST
labor and supervision $ 84,000 $14,000
Energy 11,500 4,600
Chemicals 592,000 191,000
Maintenance 228,400 28,000
Taxes and insurance 68,600 9,200
Residual waste
disposal
Monitoring, analysis
and reporting 15,000 7,500
TOTAL OPERATION AND
MAINTENANCE COST $999,500 $254,300
C. AMORTIZATION OF
INVESTMENT COST $372,000 $ 45,700
TOTAL ANNUAL COST $1,371,500 $300,000
* First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
639
-------
I
i
1
.
1
;
1
, i
i
I
J j
i
j
; i
;
! 1
1
! i !
i
i : ¦
1
.
I !
j
:
! i ¦
, 1 !
i i i
¦ ® ! LEi
'EL #2| ! 1
! ' i
/ ¦
; 1
;
i
1
' : i
; 1
/'
'
'
i i
/
1
i
1 i ¦ :
i
l/ ¦
¦ 1
' ' 1
1
; :
! , |
! 1
A i
1
! i
: i
! i
! !
1 . :
; :
. :
¦ 1 ¦
A
1 1
;
! 1
i
V
i
1
1
:
i l
>
i i
j 1
1
!
1 ' <
i 1 .
i
1
1
, i
1
. :
1
/
! ¦ i i . i ;
1 I
i
/
: ! .fS) E.EVBL :#l '
i
: ^ i
1
1
i /
¦ ! ¦.
| i
1 ' '
, XI,!
. : ¦ ! i
1
h
/ 1
1f \ j
, i i ; 1
,
(9
i I 1 !
i
i1
i
/
I 1
! I i
,
¦ i.
\st
i 1
¦ ! | :
| i :
y
i i !
¦ I
y
1 1 . .
1 1
' -XI
f
i
' 1
¦ i i :
1
!
i i
i
:
1 :
1
j I
!
' ' i !
i
1 !
I
!
;
; ! I
i : ;
i :
1
i
!
i :
1 '
1 •
1 "
i
! ¦ i
1
1
1 !
¦ : i
i '
1
1
i
1 i
!
i
30
40
60
70
80
HCN PRCDUCTICN CMETRIC TONS/YEAR X 1000)
Figure 17-6. Annual treatment cost as a furction of production for the
Hydrogen Cyanide Subcategory
640
-------
16
!
¦ i
1
1 f
i
i
h
: ¦
¦
¦ 1
i
i ,
i
1
'
¦
1
¦ I
r—!-
——X
V-
1
/ i
i
1 1
r—1
1
—-i—¦—1
'
L ¦
j
fx
¦ ¦
1
l
i
-K:
—J
i
1
I ¦ ^
1 1 .
:
!
|
1 i 1 1
|
¦
LE
I'EL tZ
f
I
¦
i .
-r-H—
——1
1
1 *
1 :
j
—r-
i
—
—1—,—i
¦ i—
TH
i
uJ
I
i
i
i .
¦ i
1 1
T
i
i
H—
1
i .
—1
i
I '
i
'ill
1 I j
, ¦ ! j
—i—
-u
l_
i
l !
i
. >
i
—+H
X ..
i
1—
. 1 .
¦ i
—f~
1—
^
V—
i
! '
—4
-J
I
~*i—:—1
N~
i .
i
r1
¦
i
—i—1—
1
1 ¦
' : ! ; ¦
¦ ' ¦ ¦ t ¦
;
T—
i
1—h-;
——i—
. 1
1_
i .
i
i
:
i ^
1 .
¦
i 'ill
¦
i
—J
1
lx^el *i
1
—j—
i
i
i
—i—
i
1
i
. 1
i
i
i
¦
-1
¦
1
—M
. i
1 i i : : 1 .
. i
—I—
,i , . . , i , . , , .
! ! ' 1 1 1 ' ! ' :
1 1 1 |
1
i i
30 40 SO 60 70 80
HCN PRODUCTION (METRIC TONS/YEAR X 1000)
Figure 17-7. Annual unit treatment oost as a function of
production for the Hydrogen Cyanide Subcategory
641
-------
TABLE 17-13. MODEL PLACT TREATMENT COSTS
Subcategory HYDROGEN CYANIDE
Annual Treatment Costs/Metric ton of Product
COST ITEK
PRODUCTION
M tons
FLOW
(rr.Vday)
LEVEL OF
FIRST
TREATMENT
SECOND
Annual Operation
and Maintenance
31,800
5,100
17.78
4.60
50,900
8, 2C0
16.35
4.02
63,600
10,300
15.72
3.4C
Annual
0.9^7
Amortization
31,800
5,100
6.35
50,900
8,200
6.10
0.63
63,600
10,300
5.85
0.72
Tbtal Cost
31,800
5,100
24.13
5.57
50,900
8,200
22.45
4.65
63,600
10,300
21.57
4.12
642
-------
chlorination to destroy cyanide amenable to treatment by
chlorination, followed by clarification.
Flow Basis
The proposed effluent limitations are based on the high
flow (57 m3/kkg of HCN) model; that is no recycle of absorber
water. A low flow basis (7 m3/kkg of HCN based on the flow of
Plant f782) was rejected as being too energy intensive due to
the need for refrigerative cooling of the recycled absorber
water. The water going to the model plant treatment system is
assumed to consist of process and contact cooling waste
effluents, leaks and spills, and storm water run-off. The
boiler blowdown and noncontact cooling water (once through or
blowdown discharge in case of closed loop) are not included in
the flow basis.
Selection of Pollutants to be Regulated
The selection of pollutants on which specific limitations
are proposed are based on the evaluation of raw waste
composition as determined during the screening and verification
programs.
Raw waste pollutant concentrations - Plant f765 was sampled
during the screening phase and the presence of toxic pollutants
in significant concentrations established the need for
verification sampling. Two plants were sampled in the
verification phase. Free cyanide, total cyanide, and ammonia
were found in the raw waste at concentrations high enough to be
treatable (Table 17-9) using available treatment technology
options. These were therefore selected for regulation.
Chlorine concentrations in the effluent are not affected by BPT
treatment technology and therefore no BPT limit is proposed for
this parameter. Thallium is best controlled by management
practices developed by the permit authority on a case-by-case
basis.
Total subcategory raw waste pollu tant loading - The average
unit loading of the pollutants found in significant amounts were
calculated from the raw waste loads of the plant sampled during
screening and verification. The unit pollutant load values
(Table 17-9) were multiplied by the estimated production rate of
165,500 kkg/year to estimate the total annual production loading
rates for the subcategory (Section 17.3.3). The prevalent
treatment technologies (alkaline chlorination and biological
treatment) are implemented for removal of the regulated
pollutants.
64 3
-------
Basis of Pollutant Limitations
Conventional and nonconventional parameters -
A. pH: The treated effluent is to be controlled within
the pH 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 concentration of suspended solids found
during sampling of the raw waste water was low. No additional
solids are produced in the treatment technology and no provision
presently exists in the existing or model treatment systems for
the removal of solids. The maximum concentration of 35 mg/1 of
TSS found in the raw waste during screening and verification
sampling (Table 17-9) was taken as the concentration basis for
the proposed maximum 30-day average effluent limitation. In the
absence of long-term monitoring data for TSS, the variability
factor ratio of 2.7 estimated for free cyanide is used to
calculate the 24-hour concentration basis and effluent limit.
The proposed total suspended solids (TSS) maximum 30-day
average effluent limit is given by:
(35 mg/1)(57 m3/kkg) / kg/m3 N = 2.0 kg/kkq
^1000 mg/1J
The proposed TSS 24-hour maximum concentration is given by:
(35 mg/1) (2.7) = 95 mg/1
The proposed TSS 24-hour maximum effluent limit is given
by:
(2.0 kg/kkg)(2.7) = 5.4 kg/kkg
C. Ammonia: Plant C765 conducted a 28-day sampling study
of the treated effluent for pollutants which were not monitored
on a long-term basis. The Agency is proposing regulation of
ammonia in the discharge, based on the 28-day sampling results
of Plant f765. Plant #765 uses a proprietary process for the
removal of ammonia, however, the same performance can be
achieved by steam stripping. The 28-day test data of ammonia in
the discharge effluent was reported by Plant f765 on a unit
product basis; i.e. kg/kkg. The average ammonia effluent
loading of 3.6 kg/kkg (Table 17-14) from the 28-day sampling
test is multiplied by the 30-day average variability factor
(also determined from the 28-day test data) of 1.2 (Table 17-14)
to calculate the 30-day average unit effluent limit. The
variability factor of 2.7 (Table 17-14) estimated from the
sampling study is used to calculate the proposed 24-hour maximum
effluent limit. The corresponding proposed concentration
644
-------
TABLE 17-14. STATISTICAL ANALYSIS OF THE 28-DAY EFFLUENT
SAMPLING RESULTS ON TOTAL CYANIDE AND
AMMONIA FROM PLANT =765
POLLUTANT
Total Cyanide Anmonia-N
Daily Data
No. of points 25 26
Average Unit Load 0.192 3.634
kg/kkg of HCN
Std. Deviation S^ 0.128 3.312
Std. Deviation 0.61 0.58
Variability Factor^) 3.44 3.26
30-Day Average Data
The Standard error
of the mean (A) ^) = 0.023 0.422
Coefficient of
variation for the mean(CV)' ' 0.119 0.116
Variability factor^) 1.19 1.19
Variability Factor Ratio
V.F.R.(7> 2.9 2.7
(1) S = Arithmetic Standard Deviation
(X. - X )
n-1
X
X.
1
n
is the mean value
is the data point value
is the no. of points
(2) S = is the estimated_standard deviation of the logarithm derived frcm
the arithmetic mean, X, and the arithmetic standard deviation, S, accor-
ding to the relationship
(S') = In
!.° +(-=-
645
Continued.
-------
TABLE 17-14 Continued
(3) In case of daily measurements, the variability factor, VF, for a
lognormal distribution is found by the expression
In (VF) = S* (Z-0.5S1)
when the value of Z is 2.33, the variability factor for the 99
percentile is obtained.
(5) Coefficient of variation for the mean CV = Standard error of the mean
(4)
\TW
Arithmetic standard deviation
/30~
Mean Value
= A
X
(6) Variability factor for 30-day average
= 1 + Z (CV)
Where the value of Z is 1.64, the variability
factor is for the 95th percentile
CV —> coefficient of variation for the mean
(7) VFR: ratio of the 24-hour variability
factor to the 30-day average variability factor
646
-------
limitations are calculated by using the model plant flow of 57
m3/kkg.
The proposed maximum 30-day average effluent limit for
ammonia-N is given by:
(3.6 kg/kkg)(1.2) = 4.3 kg/kkg
The proposed 24-hour maximum effluent limitation is given
by:
(4.3 kg/kkg) (2.7) = 12 kg/kkg
The corresponding 30-day average concentration basis is
calculated as follows:
(4.3 kg/kkg) (57 m3/kkg) / kg/m3 \ = 75 mg/1
V1000 mg/1J
and the 24-hour maximum concentration basis is:
(12 kg/kkg)(57 m3/kkg) f kg/m3 ^ = 210 mq/1
V1000 mg/V
Toxic Pollutants - The toxic pollutants proposed for
regulation are free cyanide and total cyanide.
A. Free Cyanide: Plant f765 practices alkaline
chlorination and has submitted two years of monitoring data on
the treated effluent for free cyanide. The samples were
properly stabilized before analysis. The variability factors
for the daily data and 30-day averages were calculated from the
long-term data as shown in Table 1.7-15. The long-term average
concentration of 0.15 mg/1 (Table 17-15) was used as the basis
for the proposed limitations. The estimated variability factors
and model plant flow rate were used in calculating the proposed
concentration bases and effluent limitations.
The proposed 30-day average concentration basis for free
cyanide is given by:
(0.15 mg/1)(1.8) = 0.27 mg/1
The proposed 24-hour maximum concentration is given by:
(0.15 mg/1)(4.9) = 0.74 mg/1
The proposed maximum 30-day average effluent limitation is
calculated by:
647
-------
TABLE 17-15. STATISTICAL ANALYSIS OF HISTORICAL EFFLUENT
MONITORING DATA ON FREE CYANIDE FRCT: PLANT #765
PERIOD: SEP. 1976 - AUG. 1978
MONITORING
FREQUENCY
N
No.
X
Mean
(ng/1)
:(D
Std Dev
(mg/1)
CV(2)
Goeff. of
Variation
VF
Variability
Factor
Daily 585
30 Day Average 24
0.15
0.15
0.15
0.076
1.0
0.5
4.9
1.8
(3)
(4)
For free cyanide, the long-term monitoring data were screened out-
lines. In the first place, values recorded as zero were interpreted to mean
"inability to measure pollutant" and were rejected prior to the statistical
analysis. For the remaining data, the reported measurements of oxidizable
CN were screened by the use of the t-statistic
t= max ( (X max-x)/s, (X-Xmin)/s)
for extreme values as outliers. Screening was performed on a month-by-month
basis, and any da tun with a calculated t value exceeding the 99% confidence
limics from the t distribution was concluded to be an outlier. Giva-i rejec-
tion of a value, reoomputation of statistical measures for that month was
performed.
(1) Arithmetic standard deviation, S
where S2 = £(X-X)2/(N-l)
For 30 day averages, this is the standard error of the mean
(2) CV = S/X
(3) For daily measurements, VF is calculated by
In (VF) = S' (2.33-5 1/2)
Where (S')2 = In (1 + (CV) ). S1 is the moments estimator of the scale
parameter of the lognormal distribution
and 2.33 is the Z value corresponding to
99th percentile
(4) For 30-day average data, VF = (l + 1.64 (CV))
Where 1.64 is the Z value for the 95th percentile.
648
-------
(0.27 mg/1)(57 m3/kkg) ( kg/m3 N = 0.015 kg/kkg
V1000 mg/1/
The proposed 24-hour maximum effluent limitation is given by:
(0.74 mg/1)(57 m3/kkg) ( kg/m3 = 0.042 kg/kkg
V.1000 mg/1/
B. Total Cyanide: The variability factors for total
cyanide for daily data and 30-day averages were estimated from
the 28-day study data conducted by Plant £765 and are given in
Table 17-14. The proposed limitations for total cyanide are
derived from the average unit effluent load (0.19 kg/kkg given
in Table 17-14), variability factors estimated from 28-day test
and model plant flow of 57 m3/kkg.
The proposed maximum 30-day average effluent for total
cyanide limitation is calculated by:
(0.19 kg/kkg) (1.2) = 0.23 kg/kkg
The proposed total cyanide 24-hour maximum effluent
limitation is given by:
(0.19 kg/kkg)(3.4) = 0.65 kg/kkg
The total cyanide maximum average concentration basis is:
(0.23 kg/kkg) (57 m3/kkg) / kg/m3 "\ = 4.0 mg/1
V1000 mg/V
The proposed total cyanide 24-hour maximum concentration
basis is:
(0.65 kg/kkg)(57 m3/kkg) ( kg/m3 N = 11 mg/1
V1000 mg/V
The proposed effluent limitations for Hydrogen Cyanide
produced by the Andrussow Process are summarized in Table 17-16
for toxic, conventional, and nonconventional pollutants.
17.7.3 Basis for Proposed BCT Limitations
The BCT limitation (applicable only to TSS) was set equal
to BPT because the dechlorination technology added for BAT does
not impact conventional pollutants.
17.7.4 Basis for Proposed BAT Limitations
The Agency considered different advanced level
technologies and their cost effectiveness relative to the base
649
-------
TABLE 17-16. PROPOSED LIMITATIONS
HYDROGEN CYANIDE (ANDRUSSOW PROCESS)
Best Practicable Control Technology Currently Available
Waste Water Flew: 57 mVkkg of HCN
Pollutant
Subcategory
Performance
(mg/1)
VFR
Concentration Basis
(1) lES/±l
Max
30-day
Avg
24-hr
Max
Effluent Limit
l!S2AlS2i
Max
30-day 24-hr
Avg
Conventional and Nonconventional
Pollutants;
Total Suspended 35(2) 2.7 35 95 2.0 5.4
Solids
Armenia -N(5) 42(4) 2.7 75 210 4.3 12
Toxic Pollutants:
Free Cyanide(5) 0.15(3) 2.7 0.27 0.74 0.015 0.042
Total Cyanide(5) 3.4(4) 2.8 4.0 11 0.23 0.65
(1) VFR: Ratio of the 24-hour variability factor to the 30-day variability
factor.
(2) Maximum effluent concentration from screening and verification sampling
data.
(3) Average based on two years of long term monitoring data submitted by
Plant #765 (Table 17-14)
(4) Average based on the 28-day comprehensive sampling data submitted by
Plant #765 (Table 17-15)
(5) Also applicable for PSES and PSNS limitations
650
-------
level systems (BPT) for the removal of toxic, conventional, and
nonconventional pollutants. For BAT, the Aqency is proposing
Level 2 technology which includes dechlorination before final
discharge.
The Agency also considered break point chlorination for
essentially complete destruction of cyanide. However, the
operational costs were too high. The reduction of effluent load
to the treatment system by recycling the absorber water was also
considered and was found to be too energy intensive and too
costly. Therefore the only cost effective treatment technology
beyond BPT was found to be dechlorination.
Technology Basis
For BAT, the Agency is proposing limitations based on BPT
with the addition of dechlorination (Figure 17-5, Level 2) .
Control of chlorine in the discharge in uniformly inadequate in
this industry. Its control in BAT is believed to be appropriate
because of its well-documented toxicity to aquatic life. The
basis for the chlorine limit is transfer of technology from the
electric utility industry (58). This transfer is appropriate
because the chlorine in both streams is amenable to the same
treatment for removal and removal is not inhibited by the
presence of other chemicals in either of the waste streams.
Flow Basis
The BPT effluent discharge rate of 57 m3/kkg of HCN has
been used as the basis for the BAT model plant.
Selection of Pollutants to be Regulated
For the BAT regulation, the Agency has selected chlorine in
addition to the pollutants identified in BPT.
Basis of Pollutant Limitations
Nonconventional pollutants - The two nonconventional
pollutants proposed for regulation are ammonia-N and total
residual chlorine. The BAT limitations for ammonia are the same
as those proposed for BPT. For total residual chlorine the BAT
regulation is based on the chlorine discharge limits for the
Steam Electric Generating Point Source Category. The maximum
30-day average in that industry is 0.20 mg/1, for the BPT (58).
The same value is proposed for this BAT regulation. The
variability factors used for free cyanide (Table 17-15) and the
model plant flow of 57 m3/kkg are used to calculate the
concentration and unit effluent limitations.
651
-------
The proposed 24-hour maximum concentration basis is given
by multiplying the VFR (4.9/1.8 = 2.7) from Table 17-15 by the
maximum 30-day average concentration as follows:
(2.7)(0.20 mg/1) = 0.54 mg/1
The proposed maximum 30-day average effluent limitation for
total residual chlorine is:
(0.20 mg/1)(57 m3/kkg) / kg/m3 \ = 0.011 kg/kkg
\1000 mg/1/
The proposed 24-hour maximum effluent limitation is given
by:
(0.54 mg/1)(57 m3/kkg) / kg/m3 \ = 0.031 kg/kkg
V1000 mg/1/
Toxic Pollutants - The Agency has selected the same
limitations for free cyanide and total cyanide as those proposed
for BPT because Level 2 technology does not affect either of
these pollutant parameters.
The nonconventional and toxic pollutant limitations for BAT
are summarized in Table 17-17.
17.7.5 Basis for Proposed New Source Performance Standards
Level 2 treatment technology (also proposed for BAT) was
selected as the basis for NSPS limitations. The pollutants to
be controlled for NSPS are pH, total suspended solids, total
residual chlorine, ammonia-N, free cyanide, and total cyanide.
The proposed NSPS limitations are given in Table 17-18.
17.7.6 Basis for Proposed Pretreatment Standards
Existing Sources
The Agency is proposing Pretreatment Standards for Existing
Sources (PSES) based on BAT technology excluding dechlorination
which consists of alkaline chlorination. Dechlorination is not
required because it is common practice for a POTW to treat
influents with chlorine. One plant (f765) discharges to a POTW.
The pollutants to be limited are ammonia, free cyanide, and
total cyanide as indicated in Table 17-16.
New Sources
For Pretreatment Standards for New Sources (PSES), the
Agency is proposing limitations based on NSPS. The pollutants
to be regulated are ammonia, free cyanide, and total cyanide as
summarized in Table 17-16.
652
-------
TABLE 17-17. PROPOSED LIMITATIONS
HYDROGEN CYANIDE (ANDRUSSOW PROCESS)
Best Available Technology
Waste Water Flow: 57 m-^/kkg of HCN
Concentration Basis,(mg/1) Effluent Limit
_ , . (ko/kkc of HCN)
Subcategory VFR^) ^ -
Pollutant ^ __ 30--day 24-hr 30-day 24-hr
Performance
y 24-hr JO-day
Avg Max Avg Max
Nonconventional Pollutants:
Armenia-N 42(2) 2.7 75 210 4.3 12
Total Residual ...
Chlorine 0.20(^ 2.7 0.20 0.54 0.011 0.031
Toxic Pollutants:
Free 0.15(4) 2.7 0.27 0.74 0.015 0.042
Cyanide
Tota.l 3.4{2) 2. 8 4.0 11 0.23 0.65
Cyanide
(1) VFR: Ratio of the 24-hour variability factor to the 30-day variability
factor.
(2) Average based on 28-day comprehensive sampling data of treated effluent
submitted by Plant #765 (Table 17-15).
(3) Regulation is based on the chlorine discharge limits in the utility
industry.
(4) Average based on two years of long-term monitoring data submitted by
ftant #765 (Table 17-14).
653
-------
TABLE 17-18. CONTROL PARAMETER LLMITATIONS
HYDROGEN CYANIDE (ANDRUSSOW PROCESS)
NEW SOURCE PERFORMANCE STANDARDS
WASTE WATER FI£W: 57 m3/k) m (kgAkg of HCN)
Max Max
Pollutant (mg/1) 30-day 24-hr 30-day 24-hr
Ava Max Avg Max
Conventional and Nonoonventional
Pollutants;
Total Suspended
Solids.TSS, 35 2.7 35 95 2.0 5.4
Total Residual
Chlorine 0.2 2.7 0.2 0.54 0.011 0.031
Anmonia - n 42 2.7 76 210 4.3 12
Toxic Pollutants :
Free 0.15 2.7 0.27 0.74 0.01S 0.042
Cyanide
Total 3.4 2.8 4.0 11 0.23 0.65
Cyanide
(1) VFR: Ratio of the 24-hour variability factor to the 30-day variability
factor.
654
-------
SECTION 18
SODIUM DICHROMATE INDUSTRY
18.1 INDUSTRY PROFILE
18.1.1 General Description
Most of the sodium dichromate produced is used in the
chromic acid and pigment industries. It is used for leather
tanning, and metal treatment as well as a corrosion inhibitor.
The industry profile data for this subcategory are given in
Table 18-1, and the status of regulations is given in Table 18-
2.
18.1.2 General Process Description and Raw Materials
The starting materials for the preparation of sodium
dichromate are chromite ore, limestone and soda ash. When the
above materials are reacted, sodium chromate is formed which is
reacted with sulfuric acid to produce sodium dichromate. The
reactions are given as:
4FeCr204 + 8Na2C03 + 702 = 8Na2Cr04 + 2Fe203 + 8C02 (1)
2Na2Cr04 + H2S04 = Na2Cr207 =
h2o
+ Na2S04
(2)
Chromite ore is a chromium iron oxide containing ferrous
chromite (FeCr204 or Fe0Cr203). Small amounts of aluminum,
silica and magnesia are present. For the preparation of sodium
chromate and finally, sodium dichromate, high grade chromite
ores are used containing approximately 50 percent Cr203. These
ores are imported from South Africa.
At the plant site, the ore is ground to a fine powder,
mixed with soda ash and calcined in rotary kilns at 1100 to 1150
degrees C. The reacted product is leached with hot water in a
leachate tank. The thickener underflow is filtered and the
filtrate recycled to the leachate tank or thickener. The solid
filter cake is dried in rotary kilns. The aluminum present in
-------
TABLE 18- :
SUBCATEGORY PROFILE DATA SUMAHSf
SUBCRTEOOref
soorjy. DIQUCMA1E
Ttotal subcategory capacity rate
Total subcategory production rate
Ntrrber 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:
Mininun
Maxinun
Average production
Median production
Average capacity utilization
Plant age range:
Mininun
Maxima
Waste water flow range:
Mininun
Maximum
Vblrne per unit product:
Minimus
Maxiaun
140,000 Wcg/year
136,500 kkg/year
3
3
NA
112,000 *kg/year
MA
82 peroent
20,700 kkg/year
66,800 kkg/year
37,300 kkg/year
24,800 kkg/year
77 peroent
7 years
28 years
4S5 cubic meters/day
720 cubic roters/day
4 cubic meters/Wo;
8 cubic neters/lckg
Stxirces of data ara Stanford Research Institute, Directory of Chmical
Producers, U.S.A., 1977, U.S. Department of Ountneiue, Current Industrial
Reports, Deosnber 1977; Energy arid ERvirannental 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 Qianicals Industry,"
March, 1980.
NA * Not Available
656
-------
TABLE 18-2 -
STATUS OF REGULATIONS - EFFLUENT LIMITATION GUIDELINES
SUBCATEGORY
SUBPART
Sodium Dichranate
Q (40 CFR 415.170, 3/12/74)
STANDARDS
BPCTCA
batea*
NSPS
Max. Avg.
Max. Avg.
Max. Avg.
Product
Para- kg/kkg kg/kkg
kg/kkg kg/kkg
kq/kkg kg/kkg
(mg/1) (mg/1)
Process
meters (mg/1) (mg/1)
(mg/1) (mg/1)
No
discharge
of
pwwp-^
NO
discharge
of
pwwp
No
discharge
of
pwwp
Na2Cr2°7
TSS
Cr*
Cr (T)
0.44
(52)
0.009(4)
(0.11)
0.0088
(1.0)
0.22
(26)
0.0005
(0.060)
0.0044
(0.50)
0.30
0.15
0.009(4) 0.0005
0.0088 0.0044
* Section 415.173 was remanded and is presently reserved (41 FR 51601,
November 23, 1976).
^Max. - Maximum of any one day.
2
Avg. *• Maximum average of daily values for thirty consecutive days.
"^pwwp - Process wastewater pollutants.
^The published value in 40 CFR 415.172 and 415.175 is incorrect and should be
0.0009 kg/kkg.
657
-------
the thickener overflow is hydrolyzed and removed from the
chromate solution as precipitated aluminum hydrate in slurry
form. The solution is centrifuged and the centrate is
evaporated, to give a concentrated solution of sodium chromate,
which is reacted with sulfuric acid to give sodium dichromate
and sodium sulfate. Sodium sulfate crystallizes as anhydrous
sodium sulfate from the boiling solution, and the crystals are
removed by filtration. The filtrate is concentrated in multiple
effect evaporators. The residual sodium sulfates separate out
as solids from each of the evaporators while the hot
concentrated solution of sodium dichromate from the last effect
of the evaporator is fed to a water-cooled crystallizer. Sodium
dichromate crystallizes out and is centrifuged. The centrate,
or mother liquor, is returned to the evaporator. The sodium
dichromate crystals separated in the centrifuge are dried in a
rotary drum dryer and then packaged for sale or stored for use.
Figure 18-1 presents a generalized flow diagram for the
production of sodium dichromate.
18.2 WATER USE AND WASTE SOURCE CHARACTERISTICS
18.2.1 Water Use
Water is used for noncontact cooling, in leaching, for
scrubbing vent gases and for process steam for heating. Water
use information provided in 308 Questionnaires is given in Table
18-3. It is possible that the figures given in the 308
Questionnaires may be the amount going to each unit operation
and not the amount added as makeup water. The quantities seem
unusually high for an industry practicing extensive recycling of
water, as this one does.
18.2.2 Waste Sources
Spent Ore
The unreacted ore is removed from the process as a sludge.
The solids contain chromium and other impurities originally
present in the ore. The waste is disposed as a solid waste in a
suitable landfill or is slurried with water and sent to the
treatment facility.
Noncontact Cooling Water and Cooling Tower Blovdown
The noncontact cooling water is either used on a once-
through basis and discharged or is recycled and the blowdown
discharged to the treatment facility, in addition to dissolved
sulfate and'chloride, it may contain chromates.
658
-------
1
CHJHBS
AM)
omm
BOM MM
i_n
vo
Fanoj! ¦
wont
FFCM
PROCESS
v»
L
KILNS
Ma» VAFCK
senna
t
LOCH
Fiunst MS
r
—»—»n
_L
1
I
~
ID t*9K
rasrouK
mrioz
TES
VCWB RBIDUE
PILTB)
GOGLUC TCMER
MO BOIUR
eUMXM)
I
I
I
T
*
DRH*
SODIUM
suuwre
BYPWDUCT
TO SMiS OR U6K
riLTOt
OR
cumuiuz
N0N3QNDCT
i COOLING I
WTCR f
EVAPORATION
SIWM
¦SULFURIC
ACID
JCIDIFIBI
to t*9re
CRM SOOIU4
i
i
t
oucium
RfKMU.
— soonM
arocmic
OVOfVIS 9CU/nCN
~UCIUM SLUDGE
CDOLDC IDBt MO *4
B0HJ3I f
H£MXMN TO t*9I«
aorsnauzB)
camuiucE
NUNJUKXMCT
OOOIJNG
WflTB
DRYER
1
SCDRJM
Diooo^re
mooter
LIQUID
water
i
PMRACING
I
T
ID SAL£S
TO SALES
Figure 18-1.
General process diagram for production of sodiun didironute.
-------
TABLE 18-3.
WATER USAGE IN SODIUM DICHRCMATE SUBCATEGORY
Source
Water usage at plants, Jn^/kkg of Na2^207)
Nonoontact coolinq
Noncontact ancillary
uses
Direct process contact
Plant #398
277
0.5
(1)
Indirect process contact
(pumps, seals, leaks and
spills)
Maintenance, e.g.
cleaning and work area
washdcwn
Air pollution control
5.7<2)
0.9(2)
0.5
2.5
(2)
(2)
Plant #376(3)
11.39
NA
]
7.83
(4)
4.16
NA
Plant #493
5.7
3.12
2.85
0.2
0.2
1.0
Total contact waste
water influent to
treatment
9.6
(2)
11.59
4.25
NA = Not Available
(1) Up to 50 percent solids
(2) Total recovery and recycle is practiced at this plant.
(3) Plant is no longer in operation.
(4) Due to a high evaporation rate, there is no discharge fran the primary
pond during 9 to 10 months of. the year. There was no primary pond
effluent at the time of sampling and only 4.16 mVkkg of the indirect
contact sources were being treated and discharged.
660
-------
Boiler Blowdown
The steam used for heating is recovered as condensate,
while the boiler blowdown is discharged to the treatment
facility. It may become contaminated with chromium escaping
from the process area and hence should be sent to the waste
water treatment facility for treatment.
The majority of aqueous streams resulting from the
manufacture of sodium dichromate are recycled. Streams recycled
include condensates from product evaporation and drying; product
recovery filtrates; air pollution control scrubber effluents
from product drying, leaching and roasting kilns; filter wash
waters; and equipment and process area washdowns. At two plants
the waste water, consisting of boiler and noncontact cooling
tower, is used to slurry the spent ore residue to the waste
water treatment facility. At one plant, the only waste water
resulting from process operations is the noncontact cooling
water, which is used on a once-through basis.
18.3 DESCRIPTION OP PLANTS VISITED AND SAMPLED
18.3.1 Screening
Three sodium dichromate plants were visited and the waste
water streams sampled. Plant f493 was sampled in the screening
phase and Plants f376 and f398 were sampled in the verification
phase.
At Plant f493, the waste water going to the treatment
facility includes the boiler and cooling tower blowdown and a
small volume of effluent from a scrubber on a by-product sodium
sulfate operation. The total waste includes the spent ore
residue, which is also sent to the treatment facility. At the
treatment facility, the alkaline waste waters are reacted with
imported acidic industrial waste (pickle liquor containing
ferrous iron) at an elevated temperature in a reactor. The
chromium is reduced and precipitated during the reaction. The
reacted waste is sent to clarifiers via holding tanks. In the
clarifiers, large quantities of water are used to wash the
precipitated solids in a countercurrent fashion. The final
clarifier overflow, which is the treated effluent, is filtered
and discharged and the clarifier underflow is disposed of in a
quarry. Figure 18-2 is a block diagram of the treatment process
and indicates which streams were sampled. Table 18-4 gives the
flow data and pollutant emissions of the streams sampled.
661
-------
RAW WASTE WATER
WATER
0
#1
REACTORS
HOLDING TANKS
T
#3
SLUDGE TO
LAND DISPOSAL
IMPORTED ACID
INDUSTRIAL WASTE
CLARIFIERS
#2
TREATED EFFLUENT
LEGEND
^ SAMPLING POINTS.
Figure 18-2.
General waste water treatment process flow diagram at Plant #*~ 93 showinc
the sampling points. (Sodium dichromate manufacture).
662
-------
TABLE 18-4. FLOW AND POLLUTANT CONCENTRATION DATA OF THE SAMPLED V&STE
STREAMS FOR PLANT #493 PRODUCING SODIUM DICHRCMATE
Chromium
Stream No. W&ste Stream Unit Flow TSS Load Cr+^ Load Load
Discription (m^/kkg (kg/kkg (kg/kkg (kg/kkg
of Na2Cr2()7) of Na20:207) of Na2Cr207) of
Raw W&ste 4.25 183 3.5 3.30
Water
Treated 28.91* 0.018 0.0001 0.072
Effluent
* This value includes the flow frcm the sodium dichronate plant, imported
acid used for neutralization, and the water used for washing the solids.
663
-------
18.3.2. Verification
At Plant (376, sodium sulfide is used for simultaneous
chromate reduction and precipitation. The waste waters at this
plant are segregated into two streams. One stream consists of
the cooling tower and boiler blowdown and is used for slurrying
the spent ore residue to the treatment facility. The second
waste stream consists of stormwater runoff from both the solids
disposal areas and the production areas. The first waste water
stream is mixed with sodium sulfide during transportation and
sent to a diked containment and settling pond system. The
sulfide reduces the hexavalent chromium to trivalent chromium,
which in turn is precipitated as chromium hydroxide. The solids
are settled in the pond, and the overflow from the ponds is
mixed with the second waste stream and reacted with sufficient
alkaline sodium sulfide to reduce the chromate and precipitate
chromium hydroxide. The reacted solution is sent to a settling
pond where the suspended solids are settled and the overflow
discharged. A simplified flow diagram of the waste water
treatment process is given in Figure 18-3. Table 18-5 gives the
flow data and pollutant emissions for the streams sampled.
Plant #376 has recently discontinued its production of
sodium dichromate. At the time of sampling, the data obtained
from this plant was considered a valid part of the data base for
assessing the pollution potential of the industry and evaluating
viable treatment options. The chromate reduction technology
being used was evidently subject to periodic problems associated
with the hazard of H2S gas production. This has been confirmed
in treatability studies currently being conducted by the Agency.
With proper operation of the treatment system this problem can
be avoided.
At Plant #398, the only effluent produced is the noncontact
cooling water. The noncontact cooling water is used on a once-
through basis and is discharged without treatment through two
outfalls. The solid waste residuals from the leaching process
are trucked to a state-licensed hazardous waste landfill area.
The amount of solid waste residue disposed of is approximately
290 kg/kkg of product. Table 18-6 gives the unit flow data and
pollutant emissions for the process effluent.
18.3.3 Toxic Pollutant Concentrations and Loadings
Toxic pollutants detected in the raw wastes during sampling
were as follows:
664
-------
COOLING TOWER
BLOWOOUN
WASTE
SODIUM SULFIDE
HUD
SLURRY
i
-------
TABLE 18-5. FLOW AND POLLUTANT LOADING DATA OF THE SAMPLED V&STE
STREAMS FOR PLANT #376 PRODUCING SODILM DICHRQMATE
Average Observed Loadings
Stream
No.
Waste Stream Unit Flow TSS Load Cr+^ Load Chromium
Load
(kg/kkg (kg/kkg (kg/kkg
of Na2Cr207) of Na2Cr207) of Na^^Oy) of Na2Cr207)
(m^/kkg
Mid Slurry
waste
7.85
3988
0.407
1.041
Primary Pond* N&
Effluent
0.591
NA
0.808
4
5
Surface Runoff
Reactor
Effluent
Pond Effluent
.4.16
0.621 0.057
7.942
NA
0.046 < 0.00004
0.55
0.77
0.0034
* EXae to a high evaporation rate, there is normally no discharge frcm the
primary pond for 9 or 10 months of the year.
NA = Not available
666
-------
TABLE 18-6. FLOW AM) POLLUTANT LOADING DATA OF THE SAMPLED VASTE
STREAMS FOR PLANT #398 PRODUCING SODIUM DICHRCMATE
Average Observed Loadings
Stream Waste Stream Unit Flow TSS Load Cr+^ load Chromium
No. Description Load
(mVkkg (kg/kkg (kg/kkg of (kg/kkg of
of Na2Cr20-7) of Na2Cr207) ^20:207)
1 Nonoontact 71 0.426 NNI* NNI*
cooling water
2 Nonoontact 206 0.55 NNI* NNI*
cooling water
* NNI= No net increase of the pollutant load, oanpared to the intake
source.
667
-------
Maximum Concentrations Observed (ug/1)
Verification
Pollutants Screening (2 Plants)
Chromium (Total) 250,000 310,000
Chromium (Hexavalent) 150,000
Nickel 13,000 1,300
Zinc 580 1,200
Copper 35 240
Lead 9 24
Silver <0.5 230*
Arsenic <10 <5
Selenium < 5 140**
* Found at one plant only
** Noncontact cooling water at one plant only
Individual plant average raw waste loads per unit product
found in sampling can be found in Table 18-7. A summary of daily
and unit product raw waste loads for all plants sampled can be
found in Table 18-8.
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
for this subcategory are as follows:
Total Subcategory Raw Waste Load Generation
Pollutant Waste Load (kg/year)
Chromium (Total)
290,000
Cr (Hexavalent)
210 ,000
Nickel
3,700
Zinc
330
Copper
55
S ilver
20
Lead
< 8.2
Selenium
4
Arsenic
< 5
18.4 POLLUTION ABATEMENT OPTIONS
18.4.1 Toxic pollutants of Concern
The most significant toxic pollutants found are the primary
pollutant, chromium, and the common heavy metals often present
668
-------
TABLE 18-7. TOXIC P0LU7EANT RAW WASTE DATA
SUBCATEGORY
SODIUM DICHRCMATE
POLLUTANT
AVERAGE RAW WASTE INFLUENT
PLANT #493
PLANT
#376
(mg/1)
(kg/kkg)
(mg/1)
(kg/kkg)
Chromium, Cr
250.0
0.94
420.0
3.30
(topper, Cu
0.035
0.00013
0.085
0.00067
Lead, Pb
0.009
0.00003
0.011
0.00009
Nickel, Ni
1.25
0.0047
0.64
0.0050
Zinc, Zn
0.580
0.0022
0.318
0.0025
Silver, Ag
< 0.005
< 0.00002
0.036
0.00028
Selenivm, Se
< 0.005
< 0.00002
< 0.005
< 0.00004
Arsenic, As
< 0.010
< 0.00004
< 0.005
< 0.00004
669
-------
TABLE 18-8.
SUMMARY OF RAW WASTE LOADINGS FOUND IN
SCREENING AND VERIFICATION SAMPLING
SUBCATEGORY 90DILM DICHROMATE
Pollutant Unit Loading, (kg/kkg)
No. of
Minimum Average Maximun Plants
Toxic
Chranium, total
0.94
2.12
3.30
2
Chromium,
Hexavalent
0.47
1.6
2.6
3
Oopper
0.00013
0.0004
0.00067
2
Nickel
0.0047
0.027
0.050
2
Silver
0.00002
0.00015
0.00028
2
Zinc
0.0022
0.0024
0.0025
2
Seleniun
*
< 0.00003
*
2
Arsenic
*
< 0.00004
*
2
Conventional
TSS
140
2100
4000
2
* Ooncentraticns were at or below the detection limits
670
-------
as impurities in the chromium ore, notably zinc and nickel. In
controlling these metals by the processes chosen for the
treatment models, incidental removal of other trace toxic metals
may also occur.
The existing BPT regulations control pH, TSS, and chromium
(Table 18-2). Effluent limitations on nickel and zinc are being
added under the proposed BAT-based regulations. Although
copper, silver, selenium, lead, and arsenic were detected in
trace quantities (Section 18.3.3 and Tables 18-7 and 18-8),
these five toxic pollutants did not occur at treatable
concentrations and, therefore, no regulations on them are being
proposed.
18.4.2 Process Modifications and Technology Transfer Options
Appropriate process modifications can be made where
opportunities exist for recycle of chrome-bearing waste waters
for recovery and reuse in the process or for use in other
product manufacturing operations. Plant tf398 currently
practices extensive recovery of chromium values for use in other
processes and has no discharge of direct process contact waste
waters.
18.4.3 Best Management Practices
Extensive recycle and reuse of process contact waste water
limit effluent generation at sodium dichromate plants. At two
facilities, cooling water blowdown streams are used to slurry
spent ore residues and the resultant waste stream is treated for
the removal of chromium prior to discharge. At the remaining
plant, ore residues are removed as a solid waste and only once
through noncontact cooling water is discharged.
18.4.4 Prevailing Control and Treatment Practices
At the time of verification sampling, Plant £376 was using
alkaline sodium sulfide (or bisulfide) for the reduction of
hexavalent chromium, followed by precipitation of metal sulfides
and hydroxides. Problems experienced by the plant included
intermittent, low level H2S gas generation and incomplete
reduction of the chromates. These problems were mitigated by
the physical layout of the treatment system and lagoons and the
long retention time afforded by the evaporation ponds during
most of the year. This plant, however, is no longer in
operation.
At present, Plant £493 is the only plant in the industry
which has a process contact waste water discharge. The
treatment technology employed is the reduction of chromate
wastes with an acidic ferrous iron solution (waste pickle
liquor), followed by lime addition for metal hydroxide
671
-------
precipitation, settling, and filtration. Overall, this
technology is roughly equivalent to the sulfide
reduction/alkaline precipitation technique previously used by
Plant f376 and has the advantage of not risking operator
exposure to hydrogen sulfide gas.
18.4.5 Advanced Treatment Technologies
In addition to the chromate reduction and metal removal
techniques practiced in the sodium dichromate industry,
consideration was given to other advanced treatment technologies
considered to be equal to or better than the proposed BAT.
These technologies include:
The use of sulfur dioxide for chromite reduction.
Ferrite coprecipitation i.e., the addition of ferrous iron
(e.g., waste pickle liquor) and aeration at about pH 5-6
for both chromate reduction and metals precipitation.
Ion exchange systems.
Xanthate precipitation.
These options are not considered viable at this time
because there is not sufficient information on performance and
cost effectiveness.
18.5 SELECTION OP APPROPRIATE TECHNOLOGY AND EQUIPMENT
18.5.1 Technology for Different Treatment Levels
Alkaline precipitation or reaction with sulfide will
separate nickel and zinc from solution. Hexavalent chromium
must be reduced to its trivalent form before it can be
precipitated as the hydroxide. Although ion exchange or
xanthates can remove metals from clarified solutions they are
inappropriate for treating raw waste slurries from this
industry.
Level 1 (BPT)
The system utilizes sodium bisulfide added to the raw
wastes to reduce hexavalent chromium to its trivalent form and
partially to precipitate some of the metals as metallic
sulfides, along with inert ore solids in a first-stage lagoon.
The lagoon effluent is then subjected to alkaline precipitation
of trivalent chromium, followed by solids separation in a
clarifier and by pH adjustment of the overflow before discharge.
672
-------
Other reducing agents may be utilized instead of sodium
bisulfide for the reduction of hexavalent chromium such as
ferrous iron or sulfur dioxide. Using either of these reagents,
chromate reduction under acid conditions would be followed by pH
adjustment with lime or caustic to obtain alkaline precipitation
of the metal hydroxides. This would obviate the need for
bisulfide addition for metal precipitation and avoid the
potential risk of operator exposure to hydrogen sulfide gas.
This level of treatment was selected as a basis for BPT because
it was typical of industry practice at the time. The flow
diagram for the sulfide-based option for Level 1 is shown in
Figure 18-4.
Level 2 (BAT)
Dual media filtration is added to achieve a higher level of
suspended solids removal, including metallic hydroxides and
sulfides which may have passed through the clarifier. The
effluent is adjusted to a pH range of 6 to 9 as in Level 1.
These technologies are uniquely appropriate for wastes of the
sodium dichromate industry because the sodium bisulfide
pretreatment performs the dual function of converting hexavalent
chromium to a potentially settleable form, as well as reacting
with other heavy metals to form insoluble metallic sulfides.
Level 2 was selected as a viable BAT treatment basis because it
was being practiced by one plant in the industry and it provides
a cost effective method of removing additional quantities of
toxic metals from the waste water with negligible impact on
solid waste handling and disposal requirements. The flow
diagram for the sulfide-based option for Level 2 is shown in
Figure 18-5.
Equipnent Functions
The raw waste flows into an equalizing lagoon where the
influent flows are measured by a magnetic flow meter which
controls application of sodium bisulfide solution into the
influent pipeline. Hexavalent chromium is converted to the less
toxic trivalent form and together with trace metal sulfides and
inert solids passes to the first-stage lagoon. A second
application of sodium bisulfide is made in the lagoon outflow,
and lime is added to precipitate trivalent chromium and residual
trace metals prior to clarification. in Level 1 the clarifier
effluent is adjusted to pH 6 to 9 and released. In the Level 2
system a dual media filter is added to remove additional
suspended material from the overflow. Clarifier underflow and
filter backwash are returned to the equalizing lagoon influent,
to be settled in the lagoon.
673
-------
SODIUM
IUSULFIDE
RAW |
WASTE
MAGNETIC
MKTt'K
SODIUM
BISULFIDE
l.IMK
LAGOON
^1
LAGOON
dXi
0 f---$*>
I
MIX
TANK
QpH ADJUSTMENT
i>
> »~-
EFFLUENT
CLAR1FIFR
-
-------
FWKWlSH
5UUIIK
bisjjjj-ji*: |
rOl
I
RAW I
WSTK ^ ^
i
MWWCTIC
MFHTS
IAOOGN
LAOOOM
SODIIM
Bisn-Fwe
i,»c
0
l ? ! J
Juri 1WH
" KlX SMP FILTE
TONK
I
SW FILTER
OARIFIF3?
Q p»l NMVJJJ1MNT
:s-t
(£)
i
»>niwr
Inrludos flow monitoring, pH monitoring •""> s;imj»lrr.
Figure 18-5. Level 2 waste water treatment for sodium dichramatc subcategory.
-------
Chemicals and Handling
Sodium bisulfide, lime, and hydrochloric acid can be used
in the treatment process. When used, the first application of
sodium bisulfide is made into the influent pipeline in
proportion to flow, minimizing the release of hydrogen sulfide
at times when the influent pH may be low. The second
application of sodium bisulfide is also into a closed pipeline
to ensure adequate mixing with the settled lagoon effluent.
Lime slurry is fed through conventional equipment ahead of the
clarifier. Hydrochloric acid is used (instead of sulfuric acid)
to minimize the formation of gypsum scale which could result
from heavy use of lime followed by sulfuric acid. The only
unusual hazard involved in the handling of chemicals for the
proposed treatment, is some hydrogen sulfide qeneraticn. This
may be unavoidable even under carefully controlled conditions.
Because of the high toxicity of this gas, all appropriate
measures to protect workers must be taken, and consideration of
alternative reduction methods given.
Separation and Disposal of Solids
As a basis for estimating model plant costs, influent
suspended solids, metallic hydroxide and sulfide precipitates,
and filter backwash are returned to or left in the influent
lagoon(s). As each lagoon becomes filled with solids it is
replaced by another, on a ten-year cycle. Liquid is decanted
from each filled lagoon and the solid material must be
periodically removed to a chemical landfill.
Monitoring Requirements
Internal process monitoring should include both routine
testing to maintain reducing conditions and a pH above 7 in the
influent lagoons, and simple field determination of pH to assure
that the optimum level is reached for precipitation of chromic
hydroxide. Routine testing of the effluent should also be
performed at the site to show that hexavalent chromium is being
consistently reduced to trivalent chromium and that total
chromium in the final effluent does not exceed the allowable
limit. Periodic composite effluent samples should be analyzed
for total chromium by the atomic absorption method, for official
reporting purposes.
18.6 TREATMENT COST ESTIMATES
18.6.1 General Discussion
Model plant specifications were selected for the purpose of
cost estimation. The rationale for the selection of model plant
characteristics is as follows:
676
-------
Production
At the time of sampling, five industrial plants produced
sodium dichromate at a total production rate of approximately
140,000 kkg/year. Two of these plants have discontinued
production. Production and waste water flow data, from which
model plant characteristics are derived, are on file for three
plants which produce a total of 112,000 kkg/year, or
approximately 80 percent of the United States production. For
waste water treatment cost estimates, three production levels
were selected. These are 20,000 kkg/year, 50,000 kkg/year and
70,000 kkg/year.
Waste Water Plow
Unit waste flows for three plants either treating or
recycling their waste waters are approximately 9.6, 11.59, and
4.25 m3/kkg of product. For the model plant, 8.5 m3/kkg of
sodium dichromate was used as the waste water flow.
Pollutant Loading
For the model plant, it is assumed that the spent ore
residues are slurried and transported to the treatment facility,
since this is the prevalent practice at two plants. The spent
ore waste-generated residue at Plant f969 is 290 kg/kkg of
Na2Cr207. The hexavalent chromium loading in the waste water
varies from 0.5 to 14 kg/kkg of Na2Cr207. Pollutant loadings
used for the model plants are suspended solids (spent ore
residue) at 290 kg/kkg Na2Cr207 produced, and hexavalent
chromium at 5 kg/kkg.
Chemicals Required
To reduce Cr+6 to Cr+3, a sodium bisulfide dosage of 168
mg/1 is needed, but to allow for reaction with other metals, a
model dosage of 200 mg/1 was used. This is equivalent to 1.7
kg/kkg of product in a unit flow of 8.5 m3/kkg. To raise the pH
to 9.5, 100 mg/1 of lime is needed, equivalent to 0.7 kg/kkg of
product. For final neutralization, HCl is used in the amount of
10 percent of the lime dosage.
Solids Generated
Total dry solids produced from treatment are 260 kg/kkg of
sodium dichromate.
18.6.2 Model Plant Control Costs
The cost estimates of three models having different
production levels are presented in Tables 18-9, 18-10, and
18-11. Annual treatment costs as a function of production are
677
-------
TABLE 18-9. MODEL PLANT TREATMENT COSTS
Subcategory SODIUM DICHRCMATE
Production
Waste water flow
20,000 metric tons per year
57 metric tons per day
400 cubic meters per day.
(22,050 tons per year)
(63 tons 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
$615,250
168,500
9,000
158,550
158,550
156,000
$1,265,850
$56,000
2,500
17,000
110,985
37,975
15,000
$239,460
$180,572
$420,032
$4,700
33,200
7,580
7,580
$53,060
$14,000
600
5,306
1,591
7,500
$28,997
$8,632
$37,629
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
678
-------
TABLE 18- 10. MODEL PLANT TREATMENT COSTS
Subcategory SODIUM DICHRCMATE
Production 50,000 metric tons per year (55,125 tons per year)
142 metric tons per day (157 tons per day)
Waste water flow 1000 cubic meters per day.
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT COST
Construction $1,375,800 $8,600
Equipment in place,
including piping,
fittings, electrical
work and controls 302 , 50 0 80 , 500
Monitoring equipment
in place 7,000
Engineering design
and inspection 337,060 17,820
Incidentals, overhead,
fees, contingencies... 337,060 17,820
Land 252,000
TOTAL INVESTMENT COST $2,611,420 $124,740
B. OPERATION AND
MAINTENANCE COST
Labor and supervision. $56,000 $14,000
Energy 2,800 1,000
Chemicals 42,000
Maintenance 235,942 12,474
Taxes and insurance... 78,342 3,742
Residual waste
disposal
Monitoring, analysis
and reporting 15,00 0 7 , 500
TOTAL OPERATION AND
MAINTENANCE COST $430,084 $38,716
C. AMORTIZATION OF
INVESTMENT COST $383,877 $20,295
TOTAL AhNUAL COST $813,961 $59,011
~First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
679
-------
TABLE 18-11. MODEL PLANT TREATMENT COSTS
Subcategory SODIUM DICHRCMATE
Production 70,000 metric tons per year (77,175 tons per year)
200 metric tons per day (220 tons per day)
Waste water flow 1400 cubic meters per day.
LEVEL OF TREATMENT*
FIRST SECOND
A. INVESTMENT COST
Construction $1,742,950 $12,200
Equipment in place,
including piping,
fittings, electrical
work and controls 390,500 91,500
Monitoring equipment
in place 9,000
Engineering design
and inspection 428,490 20,740
Incidentals, overhead,
fees, contingencies... 428,490 20,740
Land 324,000
TOTAL INVESTMENT COST $3,323,430 $145,180
B. OPERATION AND
MAINTENANCE COST
Labor and supervision. $56,000 $14,000
Energy 2,800 1,000
Chemicals 58,000
Maintenance 299,943 14,518
Taxes and insurance... 99,702 4,355
Residual waste
disposal
Monitoring, analysis
and reporting 15,000 7,500
TOTAL OPERATION AND
MAINTENANCE C06T $531,445 $41,373
C. AMORTIZATION OF
INVESTMENT COST $488,007 $23,620
TOTAL AJWUAL C06T $1,019,452 $64,993
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.
680
-------
shown graphically in Figure 18-6. Treatment cost per metric ton
of product is shown in Figure 18-7.
Table 18-12 gives a summary of the unit cost distribution
between amortization, and the operation and maintenance cost
components at various production rates and levels of. treatment.
At the first level of treatment, investment costs are high
because sludge lagoons costs are provided for a ten-year period.
Therefore, amortization is the major portion of the total annual
costs, in place of annual cost for the residual waste (sludge)
disposal, a large investment in land is shown. At the second
level of treatment, labor and amortization have significant
impact on the additional annual costs.
18.7 BASIS FOR REGULATIONS
18.7.1 BPT Effluent Limitations
Technology Basis
BPT regulations for the Sodium Dichromate Subcategory are
presently in effect, 40 CFR 415.172 (Table 18-2). The
technology basis for the existing BPT is sulfide reduction of
hexavalent chromium, followed by alkaline precipitation of
metals and clarification. As an alternative to the use of
sodium bisulfide, the reduction of hexavalent chromium may be
accomplished by reaction with ferrous iron or sulfur dioxide
under acidic conditions. All three plants in this subcategory
have installed BPT technology and are meeting the limits.
Necessary to the achievement of good effluent quality after
precipitation of heavy metals, is the control of suspended
solids. In the Sodium Dichromate Subcategory, it can be assumed
that chromium is a significant constituent in the suspended
solids discharged. For this reason, only one advanced treatment
alternative, addition of a filtration unit for solids control,
has been recommended.
Response to Renand issues
The zero discharge requirement originally promulgated as
BAT for sodium dichromate production was remanded on the basis
of inadequate technical and economic justification for the
evaporative technology required to eliminate discharge. A
control and treatment alternative, which allows waste water
discharge, has been identified and the performance levels
achievable have been demonstrated at one facility.
681
-------
11
10
o
o
o
*
O 9
O *
X
>
1
i
' 1 i
1
i
,
' 1
.
1
: 1 i
1
i
J
¦
i 1
i .
1
i
I ! !
'
i
' i
• \ \ '
i
1 ¦ ! i :
1 : '
¦ i
i
¦
1
1
;
i .
' I 1
1
I
1
j 1
' l
i ¦ , ! ¦ 1
I
i
i
i i
.
i
: : i
.
:
1 j
r.Fvri. §?
^:
¦
I 1 .
¦ 1
/V : 1
1
I :
i ; 1 1
! ¦
r X LIEYE^ #1
!
I 1
i
! 1 :
i . , j
! I X
/ ¦ ! 1 , 1
|
' 1
! i
! ! '
I
! /¦
i ;
T
i
i
I 1
' 1
1 1
!
' 1
1 1
' i
/ /
1
i
1
l
i
1 . 1
I !
, |
, I
i /
i
1
!
i
!
, 1
i i
1 i
/' Ci
5
;
1
' i '
1 i
/ . /
! i
1
' 1 !
/
y
¦
)
¦ ,
{1
/ i
i
I
1 ! ;
¦ i
j . 1 /
I
! 1
1 »
: 1
i
! A
/
' 1
I 1
1 1
i |
I
| | i
i i
/ /
I
' !
1
! ! !
/ /
1
1 ¦
: |
1
i .
/: /\
1 ¦
"
1 ,
1 i
/ 1/
1 i
1
/! i
¦ 1
! /
/ *
' 1
i ¦ j •
¦ i /
/I !
¦ . 1 i i
/ /
r i 1 ;
i
i ¦ i
/ / i ! !
; i
! 1
1
1
i i
/
/ = :
1 ¦
i !
1
it- j
¦ i
r
/ I
1
, i
i :
1 1
I ¦ ^
i,
r\ 1
!
7
! i ¦
i I
1
,. *
5
i
i
¦ , j
i
i
¦ i ¦
1 1
i
i
.
1 '
1
!
! j
¦ !
i ' ¦ !
i
¦ 1
1 1
| ,
i
1 1
i
1 , .
i
! i '
i |
20 30 40 50 60 70
PRCDUCTICN (METRIC TCNS/YEAR X 1000 )
Figure JL8-6. Relationship of annual treatment oost to production
for the Sodium Dichrorate Subcategory
682
-------
LEVEL '2
20 30 40 50 60 70
pfqducticn (Metric tcns/year x 1000 )
Figure 18-7. Relationship of annual unit treatment cost to production
for the Sodium Dichraiate Subcategory
683
-------
TABLE 18-12. MODEL PLANT TREATMENT COSTS
Subcategory SODIUM DICHRCMATE
Annual Treatment Costs ($Akg)
00 ST ITEM
PRODUCTION FLOW
(kkg/yr) (m3/day)
LEVEL OF TREATMENT
FIRST SECOND THIRD FOURTH
Annual Operation
and Maintenance
20,000
400
11.97
1.45
50,000
1,000
8.60
0.77
70,000
1,400
7.59
0.59
Annual
Amortization
20,000
400
9.03
0.43
50,000
1,000
7.68
0.41
70,000
1,400
6.97
0.34
Total Cost
20,000
400
21.00
1.88
50,000
1,000
16.28
1.18
70,000
1,400
14.56
0.93
Not Applicable
684
-------
Flow Basis
The model plant waste water flow rate is based on the raw
waste influent data obtained from three plants as shown in Table
18-3. The flow rate selected, 8.5 m3/kkg, is the average of the
flows for these three plants. All three plants are included in
the flow averaging because the waste sources were typical for
the industry at the time of sampling and represent the range of
inflow rates expected to be handled by a BPT treatment system.
Selection of Pollutants to be Regulated
For BPT regulations the Agency is retaining the pollutants
that are presently limited under 40 CFR 415.172. These are pH,
total suspended solids (TSS), hexavalent chromium (CrVI), and
total chromium (Cr). The significance of these pollutants is
substantiated by the screening and verification data presented
in Section 18.1.1.
The available treatment technology for the removal of
chromium from waste water necessitates the reduction of
hexavalent chromium (chromate or dichromate) to the trivalent
state which can then be precipitated as chromic hydroxide,
Cr(OH)3. Thus, from the regulatory point of view, an effluent
limitation on the discharge of total chromium effectively limits
hexavalent chromium as well. But, placing limitations on both
forms of chromium in the Sodium Dichromate Subcategory is
consistent with the primary objective of controlling
specifically the highly toxic hexavalent form by means of a two-
step treatment process. in light of the potential analytical
difficulties associated with the measurement of hexavalent
chromium discussed in Section 5.1.3, monitoring both the
hexavalent chromium and the total chromium content of the
treated effluent provides an additional assurance that high
chromate levels would not go undetected. As treatment system
performance data are accumulated, support may develop for a
decreased monitoring requirement or the elimination of effluent
limitations on hexavalent chromium.
Basis for Pollutant Limitations
Conventional Parameters -
A. pH: After final pH adjustment, the BPT treated
effluent is to be held within the pH range of 6 to 9. The pH
limitation is based on Appendix B of this report and a study
report, "An Assessment of pH Control of Process Waters in
Selected Plants" by JRB Associates, inc. (52).
B. TSS: The present study substantiates the basis for the
existing BPT limitation on total suspended solids. The treated
685
-------
effluent sampling data from two plants presented in Table 18-13
suggest that the TSS concentrations found in sampling represent
achievable performance of a well operated BPT system. This is
in agreement with the 26 mg/1 TSS which is the concentration
basis for the existing maximum 30-day average effluent
limitation (Table 18-2) . For comparison, Table A-lla summarizes
the long-term data available from another subcategory where a
similar BPT is applied. Plant *376 discharged an average TSS of
11 mg/1 without filtration. Monitoring data from Plant $493
shown at the bottom of Table 18-13 indicates that 25 mg/1 is an
achievable maximum 30-day average for TSS with filtration.
Thus, individual plant performance can be seen as a function of
a very large number of operating variables and waste
characteristics. in general, the available performance data
support the achievability of the existing regulations.
The variability factor ratio (VFR) of 2.0 is derived from
the long-term data on chromium as presented in Tables A-9a-l,
and following. This VFR value is used for TSS and chromium
because a significant proportion of the TSS is composed of
suspended metal hydroxides resulting from BPT treatment. For
TSS, the maximum 30-day average limitation is related to the
concentration basis and the model plant flow as follows:
(26 mg/1)(8.5 m3/kkg) f kg/m3 \ = 0.22 kg/kkg
1^1000 mg/1)
and the daily maximum limitation is obtained by multiplying by
the VFR;
(2.0)(0.22 kg/kkg) = 0.44 kg/kkg
Toxic Pollutants -
A. Chromium: For BPT, the Agency is retaining the
existing limitations on total and hexavalent chromium as given
in 40 CFR 415.172 (Table 18-2). The verification sampling data
from Plant *376 (Table 18-13) provide support for the 30-day
average concentration bases used for total and hexavalent
chromium. The observed performance level of 0.81 mg/1 of total
chromium falls between the maximum 30-day average and the daily
maximum concentration limits of 0.50 and 1.0 irg/l, respectively,
for the model plant.
For hexavalent chromium, the observed performance level of
less than 0.01 mg/1 was below the accepted lower limit of
treatability (0.05 mg/1) from Table 8-11. The treatability
level was the basis of the 30-day average concentration basis of
0.060 mg/1 used for the existing BPT regulations (Tables 18-2
and 18 — 14) .
686
-------
TABLE 18-13. EFFLUENT SAMPLING DATA FRCM
SODIU-1 DICHRQMATE PLANTS
Pollutant
Screening & Verification Data
Plant #376 Plant #493
(rog/1)
(kg/kkg)
(rog/1)
(kg/kkg)
Total Suspended
Solids, TSS
11
0.046
2.0
0.0085
Hexavalent
Chromium, Cr (VI)
< 0.01
< 0.00004
0.004
0.00002
Total
Chromium, Cr (T)
0.81
0.0034
2.5
0.011
Copper, Cu
0.012
0.00005
0.016
0.C0C07
Nickel, Ni
0.20
0.00083
0.090
0.00038
Selenium, Se
< 0.005
< 0.00002
0.10
0.00043
Silver, Ag
0.015
0.00006
< 0.007
< 0.00003
Zinc, Zn
0.008
0.0003
0.11
0.00047
Flow (m /kkg)
4.16
4.25
Long Term Monitoring Data-Maximum 30-Day Averages
Plant #493{1)(2)
(mg/1) (kg/kkg)
TSS
Cr (VI)
Cr (T)
25
0.023
0.072
0.11
0.00010
0.00031
(1) Filtered effluent data reported in response to 308 questionnaire
(12-22-76)
(2) The nunber of samples is unknown.
687
-------
The VFR of 2.0 used for total chromium is confirmed by
long-term data (Tables A-9a-l, and following) on alkaline
precipitation of chromium in another subcategory where a 1.8
value was determined for a similar BPT technology.
The existing 24-hour maximum effluent limitation that was
published for hexavalent chromium is in error as it appears in
40 CFR 415.172. The correct value is 0.0009 kg/kkg reflecting
an overall VFR value of approximately 1.8 for the residual
hexavalent chromium remaining after the two-step treatment
process.
For total chromium, the maximum 30-day average limitation
is,
(0.50 mg/l)(8.5 m3/kkg) f kg/m3 = 0.0044 kg/kkg
^1000 mg/lj
and, by applying the VFR value of 2.0, the daily maximum is,
(2.0) (0.0044 kg/kkg) = 0.0088 kg/kkg
For hexavalent, the maximum 30-day average limitation is,
(0.060 mg/1) (8.5 m3/kkg) /kg/m3__\ = 0.0005 kg/kkg
llOOO mg/l)
and the daily maximum is obtained using the VFR value of 1.8,
that is,
(1.8) (0.0005 kg/kkg) = 0.0009 kg/kkg
B. Other Metals: The concentration bases for nickel and
zinc are also given in Table 18-14. These and other similar
metals will be effectively removed by the BPT alkaline
precipitation step. Copper, silver, selenium, arsenic, and lead
did not occur at concentrations high enough to be treatable and
are therefore not regulated. An adequate removal of these
metals is expected with a BPT treatment system specifically
designed to provide optimum conditions for the precipitation of
chromic hydroxide.
18.7.2 BCT Effluent Limitations
For the control of conventional pollutants, the Agency is
setting BCT equal to BPT because the addition of more treatment
technology to increase the removal of TSS failed to pass the BCT
cost comparison test described in Section 3.3.3 of this report.
688
-------
TABLE 18-14. PROPOSED LIMITATIONS
Sodixm Dichronate
Best Practicable Control Technology Currently Available
waste Water Flow: 8.5 irr/kkg
Concentration Basis Effluent Limit
Subcategory (mg/1) (kg/kkg)
Pollutant Performance VFR^1'
(mg/1) 30-day 24-hr 30-day 24-hr
Avg Max Avg Max
Conventional Pollutants:
Total Suspended 11(3) 2.0 26 52 0.22 0.44(2)
Solids
Toxic Pollutants:
Total Chromium 0.81(3) 2.0(5) 0.50 1.0 0.0044 0.0088(2)
Hexavalent 0.050(4) 1.8(6) 0.060 o.ll 0.0005 0.0009(2)
Chromium
Nickel 0.20(3) 2.0 0.20(4)1.0
Zinc 0.50(4) 2.0 0.5 1.0
(1) VFR: ratio of the 24 hour variability factor to the 30 day
variability factor.
(2) Existing regulations, 40 CFR 415.72 (Table 18-2)
(3) Verification sampling averages from Plant #376 (Table 18-13) .
(4) Dower limit of treatability (Table 8-11).
(5) The VFR used in original regulation is confirmed by long term data on
alkaline precipitation of chrcmium in another subcategory (Tables A-9a-l,
etc.)
(6) VFR used in original regulation.
689
-------
18.7.3 BAT Effluent Limitations
Technology Basis
For BAT, the Agency is proposing limitations based on
technology that includes BPT treatment plus the addition of dual
media filtration to remove additional toxic metals from the
effluent. One plant has installed BAT treatment and is
presently meeting the proposed limitations.
Flow Basis
The 308 data was collected from three plants of which two
are still operating. For BAT, the model plant flow rate
selected is 7.0 m3/kkg which is the average of the two plants
still operating, i.e., Plants f398 and f493. Plant f376 which
is the third plant has shut down its sodium dichromate
production facilities.
Selection of Pollutants to be Regulated
For the BPT regulations previously developed, the pollutant
parameters of concern were identified as pH, TSS, hexavalent
chromium, and total chromium. The selection of toxic pollutants
for control at the BAT step is based on the results of the
screening and verification sampling program reported in this
document. in Section 18.3.3 a tabular summary of the maximum
observed raw waste concentrations is presented to show the
relative importance of the metals that were found. No
detectable concentrations of toxic organic substances were
found. Of the metals found, chromium, nickel and zinc were by
far the dominant pollutants in terms of maximum concentrations,
while copper, silver, lead, and selenium were found at lower
levels. Because of the high percentage of total chromium in the
hexavalent state it was concluded that the limitation of both
hexavalent and total chromium was advisable to assure reduction
of chrome (+6) to chromium (+3) which is part of the technology
basis for the regulation. The total subcategory raw waste
loadings are also shown in Section 18.3.3. These are based on
the average observed concentrations and loadings presented in
Table 18-8.
The estimated total loadings for the subcategory confirm
importance of chromium, nickel and zinc and these three metals
have been selected as the control parameters for the BAT
regulat ions.
Basis of Pollutant Limitations
Nonconventional Pollutants - No nonconventional pollutants
have been identified for control in the Sodium Dichromate
Subcategory.
690
-------
Toxic Pollutants -
A. Chromium: The addition of dual media filtration to BPT
is expected to achieve the removal of an additional 60 percent
of the total chromium in the treated effluent. This estimate is
based on literature treatability data (41) on chromium presented
in Table 8-11 and discussed further in Section 12.3.3. Thus,
using the sampling data obtained from Plant #376 (Table 8-13),
the concentration basis for the 30-day average effluent
limitation becomes 0.32 mg/1 total chromium with BAT treatment.
That is,
(1.00 -0.60)(0.81 mg/1) = 0.32 mg/1
The hexavalent chromium contribution to the total chromium
concentration is negligible when the chromate reduction step is
properly designed and operated. However, for BAT the
designation of hexavalent chromium as a control parameter is
retained and the proposed maximum 30-day average limitation is
based on the accepted lower limit of treatability derived from
literature data (Table 8-11), that is, 0.05 mg/1. The observed
performance at Plants (376 and *493 shown in Table 18-13
supports the achievability of this concentration on a 30-day
average basis.
The VFR value of 2.0 used for BPT is supported by long-term
data (Tables A-9a-l, and following) and is also used for the
proposed BAT regulations. This value applies to both the total
and hexavalent forms of chromium. Thus, for total chromium, the
proposed maximum 30-day average limitation is,
(0.32 mg/1)(7.0 m3/kkg) f kg/m3 \= 0.0022 kg/kkg
^1000 mg/1/
and, applying the VFR value of 2.0, the proposed daily maximum
is,
(2.0)(0.32 mg/1)(7.0 m3/kkg)f kg/m3 = 0.0045 kg/kkg
^1000 mg/y
For hexavalent chromium, the proposed maximum 30-day
average limitation is:
(0.050 mg/1)(7.0 m3/kkg) f kg/m3 \ = 0.00035 kg/kkg,
\i000 mg/y
and the proposed daily maximum is,
(2.0)(0.00035 kg/kkg) = 0.00070 kg/kkg
The proposed limitations are shown in Table 18-14.
691
-------
B. Nickel: Starting with the BPT concentration basis
shown in Table 18-14, and dual media filtration will remove an
additional 14 percent (41) of the nickel from the treated
effluent, the proposed 30-day average limitation for BAT was
determined to be (1.00 -0.14)(0.20 mg/1) = 0.17 mg/1. The VFR
value of 2.0 that was used for chromium was also applied to
nickel because of the similarity in the treatment chemistry of
these metals. Thus, the proposed maximum 30-day average nickel
limitation is,
(0.17 mg/1)(7.0 m3/kkg) f kg/m3 \ = 0.0012 kg/kkg,
\1000 mg/1/
and, the corresponding daily maximum limitation is:
(2.0)(0.0012 kg/kkg) = 0.0024 kg/kkg
C. Zinc: For BAT, the proposed zinc limitation is
based on the lower limit of treatability by alkaline
precipitation and a filter efficiency of 6 percent (41). Thus,
the concentration basis for the maximum 30-day average effluent
limitations was set at 0.47 mg/1 as follows:
(1.00 -0.06)(0.50 mg/1) = 0.47 mg/1
The VFR value of 2.0 was also used on zinc for the same
reasons given in the discussion of the chromium and nickel
limitations. Thus, the proposed maximum 30-day average
limitation is,
(0.47 mg/1)(7.0 m3/kkg) / kg/m3 \ = 0.0033 kg/kkg
\TM^mg7y
and the proposed daily maximum limitation is,
(2.0)(0.0033 kg/kkg) = 0.0066 kg/kkg.
D. Other Metals: Other toxic metals detected in the raw
waste waters include copper, lead, silver, selenium, and
arsenic. None of these occured at maximum concentrations
considered treatable by the applied technology for BAT or NSPS,
and therefore specific numerical limitations are not proposed.
Should any of these toxic metals be found at treatable raw waste
concentrations, effluent limitations would be established on a
case-by-case basis by applying the appropriate lower limits of
treatability from Table 8-11 as the concentration bases for
maximum 30-day average limitation.
692
-------
18.7.4 NSPS Effluent Limitations
Technology Basis
For new sources the EPA is proposing to replace the
existing NSPS regulations (40 CFR 415.174) with a new NSPS
regulation based on BAT.
Flow Basis
For the new NSPS, the model plant flow rate is the same
rate that was applied to the model plant BAT systems, i.e., 7.0
m3/kkg. The basis for this flow rate is described in Section
18.7.3.
Selection of Pollutants to be Regulated
For NSPS the Agency is proposing to regulate the same
conventional parameters presently controlled under the existing
BPT regulation. These are pH and TSS. No nonconventional
pollutants have been selected for regulation.
For the control of toxic metals, the Agency has selected
total and hexavalent chromium, nickel, and zinc on the basis of
screening and verification data. The bases for toxic pollutant
selection for NSPS are the same as those discussed for BAT in
Section 18.7.3.
Basis for Pollutant Limitations
Conventional Pollutant Parameters -
A. pH: For NSPS, the Agency is proposing a pH limitation
identical to the existing BPT regulation. The treated effluent
is to be held within the range of pH 6 to 9. This limitation is
supported by the results of studies presented in Appendix B of
this report and the JRB Associates, inc. report previously cited
(52) .
B. TSS: For NSPS, the EPA is proposing a total suspended
solids limitation achievable with BAT treatment. The
concentration basis for the maximum 30-day average is equal to
the 25 mg/1 derived from long-term data at Plant f493 (Table 18-
13) where the equivalent of BAT treatment is practiced.
Thus, the proposed maximum 30-day average limitation is:
Zkg/mSX
\1000 mg/y
(25 mg/1)(7.0 mg/kkg)
= 0.18 kg/kkg
693
-------
and the corresponding daily maximum is obtained by applying the
VFR value of 2.0. That is,
(2.0)(25 mg/1)(7.0 m3/kkg) { kg/m3 >\ = 0.35 kg/kkg
umJT"ig7T;
Nonconventional Pollutants - no nonconventional
pollutants have Been identified for control under NSPS
regulations.
Toxic Pollutants - For NSPS, the proposed BAT limitations
on total chromium, hexavalent chromium, nickel, and zinc apply.
The bases for BAT limitations are discussed in Section 18.7.3.
The proposed NSPS limitations are presented in Table 18-16.
18.7.5 Pretreatment Standards
Existing Sources
The Agency is proposing pretreatment standards for existing
sources (PSES) based on BAT treatment. The pollutants limited
by the proposed PSES are total chromium, hexavalent chromium,
nickel and zinc. Table 18-15 presents the PSES limitations.
New Sources
There is an existing pretreatment standard for new sources
(PSNS) in effect (40 CFR 415.176) which is based on BPT
treatment. The Agency is proposing to amend this regulation by
substituting new PSNS limitations based on BAT. The pollutants
limited by the proposed new PSNS are total chromium, hexavalent
chromium, nickel and zinc. Table 18-15 presents the proposed
new PSNS limitations. At present, there are no indirect
dischargers in this subcategory.
694
-------
TABLE 18-15. PROPOSED LIMITATIONS
Sodium Dichranate . .
Best Available Technology
Waste Water Flow: 7.0 m3/kkg
Pollutant
Treatability VFR
(mg/1)
(2)
Concentration Basis
(mg/1)
30-day
Avg
24-hr
Max
Effluent Limit
(kg/kkg)
30-day
Avg
24-hr
Max
Toxic
Pollutants:
Total Chromium
Hexavalent
Chromium
Nickel
Zinc
0.32
(3)
0.050
(4)
0.17
0.47
(4)
(4)
2.0 0.32 0.64 0.0022 0.0045
2.0 0.050 0.10 0.00035 0.00070
2.0 0.17 0.34 0.0012 0.0024
2.0 0.47 0.94 0.0033 0.0066
(1) Including pretreatment standards for existing sources (PSES)and
pretreatment standards for new sources(PSNS) expressed as concentration
limitations but with mass equivalents as an alternate.
(2) VFR: ratio of the 24 hour variability factor to the 30 day variability
factor.
(3) BPT performance basis with additional 60 percent removal by filtration.
(4) Estimated lower limit of treatability with filtration.
695
-------
Pollutant
TABLE 18-16. CONTROL PARAMETER LIMITATIONS
Sodium Dichronate
New Source Performance Standards
Vfaste Water Flow: 7.0 m3/kkg
Treatability
(mg/1)
Concentration Basis
(mg/1)
VE^1)
30-day 24-hr
Avg Max
Effluent Limit
(kg/kkg)
30-day 24-hr
Avg Max
Conventional Pollutants:
Total Suspended 25
Solids
Toxic Pollutants:
Total Chromium- 0.32
Hexavalenf
Chroniun
(2)
(2)
0.050
,(3)
(2)
Nickel 0.17'
Zinc 0.47
(3)
2.0
25
2.0 0.17
2.0 0.47
50
2.0 0.32 0.64
2.0 0.050 0.10
0.34
0.94
0.18
0.0012
0.0033
0.35
0.0022 0.0045
0.00035 0.00070
0.0024
0.0066
(1) VFR: Ratio of the 24 hour variability factor to the
30 day variability factor.
(2) Maximum 30-day average performance at Plant #493 (Table 18-13)
This plant employs treatment equal to BAT.
(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. The sampling data are presented in
Table 18-10, Plant #493.
696
-------
SECTION 19
CARBON DIOXIDE INDUSTRY
19.1 SUMMARY OF DETERMINATIONS
It has been determined that no further effort will be given
to developing BPT, BAT, NSPS, and Pretreatment regulations for
the carbon dioxide subcategory. The basis for this
determination is that no toxic pollutants were found at
significant levels in the process related waste water during the
screening of one plant. The subcategory is excluded under
Paragraph 8 of the Consent Decree.
19.2 ASSESSMENT OP THE WATER POLLUTION POTENTIAL
19.2.1 Production Processes and Effluents
Carbon dioxide is produced in gaseous, liquid, or solid
form. Most of the carbon dioxide is produced as a by-product of
ammonia production. A major portion of the carbon dioxide is
used captively for producing urea and secondary recovery of oil
and natural gas. It is also used for refrigeration, in the food
industry for the carbonation of beverages, in fire extinguishing
equipment, and oil well stimulation.
The process waste water is derived from gas scrubbing and
condensation. The only toxic pollutant found at a significant
concentration is the raw waste during screening at one plant was
zinc (910 ug/1). When the data was reviewed with plant
personnel, it was discovered that the zinc level was due to zinc
corrosion inhibitors and was not process related. Control of
zinc from this type of source is best achieved by management on
a case-by-case basis by the permitting authority. The
subcategory profile data is given in Table 19-1.
Maximum concentration of toxic pollutants found in screening
at one plant were:
697
-------
TABLE 19-1.
SUBCATEGORY PRCFILE DATA SUMMARY
SUBCATEGORY
CARBON DIOXIDE
It»tcil 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 range:
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flew range:
Mininum
Maxinum
Volume per unit product:
Minimum
Maximum
12,194,000 kkg/year
1,819,000 kkg/year
105
12
713,947 kkg/year
558,667 kkg/year
59 percent
31 percent
1,600 kkg/year
155,000 kkg/year
NA
NA
NA
6 years
50 years
NA
NA
NA
NA
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Comerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chenical Industry," June, 1978 and "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards for the Inorganic Chemicals Industry,"
March, 1980.
698
NA = Not Available
-------
Pollutant yq/1
Zinc 910
Copper 75
Chromium 31
19.3 STATUS OF REGULATIONS
Subpart AF has been reserved for this subcategory.
699
-------
Intentionally Blank Page
-------
SECTION 20
CARBON MONOXIDE AND BY-PRODUCT HYDROGEN INDUSTRY
20.1 SUMMARY OF DETERMINATIONS
It has been determined that no further effort be given to
developing BAT, NSPS, and Pretreatment regulations for the
Carbon Monoxide and By-Product Hydrogen Subcategory. The basis
for this determination is that no toxic pollutants were found at
significant levels in the process related waste water during the
screening of one plant. The subcategory is excluded under
Paragraph 8 of the Consent Decree.
20.2 ASSESSMENT OP THE WATER POLLUTION POTENTIAL
20.2.1 Production Processes and Effluents
Carbon monoxide is produced as a result of production of
hydrogen by refining natural gas. It is also recovered from
several gas sources including partial combustion of oil or
natural gas, coke oven gas, blast furnace gas, water gas, and
methane reformer gas.
The major use of carbon monoxide is for the manufacture of
methanol. It is also used in the production of ammonia, acetic
acid, zinc white pigments, and for reducing oxides for special
steels and nickel refining.
The industry profile data is given in Table 20-1.
Toxic pollutants detected in the raw waste during
screening at one plant were:
The only pollutants of significance in terms of waste loads
are chrome and zinc. However, those result from the additives
Pollu tant
Concentration (yg/H
Chromium
Z inc
S ilver
Mercury
2590
820
1.4
1.2
701
-------
TABLE 20-1
SUBCATEGORY PRCFILE DATA SUMMARY
SUBCATEGORY
CARBON MONOXIDE AND BY-PRODUCT HYDROGEN
Totcil 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 range:
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Miniiraxn
Maximum
Waste water flow range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
277,200 kkg/year
5
5
112,400 kkg/year
40 percent
47 kkg/year
63,000 kkg/year
NA
NA
NA
8 years
19 years
NA
NA
NA
NA
Sources of data cure Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Carrierce, Current Industrial
Reports, Deconber 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 Chanicals
Industry," March, 1980.
NA = Not Available
702
-------
used in cooling water to inhibit corrosion, and are not process
related. Control of zinc and chromium from this type of source
is best achieved by best management practices on a case-by-case
basis by the permitting authority.
20.3 STATUS OP REGULATIONS
Subpart AG has been reserved for this subcategory.
703
-------
Intentionally Blank Page
-------
SECTION 21
COPPER SULFATE INDUSTRY
21.1 INDUSTRIAL PROFILE
21.1.1 General Description
Most of the copper sulfate produced is sold in the merchant
market, consequently captive use is very small. Copper sulfate
is produced either as a liquid solution or dried crystals. It
is used in agriculture as a pesticide, and as an additive to
copper-deficient soils. It is also used in electroplating and
petroleum refining, and as a preservative for wood. Of the 16
plants in this industry, four plants produce copper sulfate in
significant quantities and account for 70% of the total U.S.
production. Two of these facilities account for over 50%.
The industrial profile data for this subcategory are given
in Table 21-1. The status of regulations is summarized in Table
21-2.
21.1.2 General Process Description and Raw Materials
Copper sulfate is produced by reacting copper with sulfuric
acid, air and water. The general reaction is:
Cu + 1/2 02 + H2S04 = CuS04 + H2 (1)
Various forms of copper feed material are used, from pure
copper to copper slag. The purity of raw materials
significantly effects the quality and quantity of raw waste
generated. One plant does not start with copper metal but uses
a waste stream from a copper refinery which consists of copper,
sulfuric acid, and a small amount of nickel. The solution needs
to be strengthened by the addition of more copper but the same
general equation applies.
Copper metal and/or copper refinery waste stream, steam,
water, sulfuric acid and air are treated in oxidizer tanks at
100° C to produce a solution of copper sulfate. This solution
is partially concentrated by evaporation.
705
-------
TABLE 21-1. SUBCATEGORY PROFILE DATA SUMMARY
SUBCATEGORY
COPPER SULFATE
Ttotal 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 range:
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maxiirun
Waste water flew range:
Minimum
Maxinum
Volume per unit product:
Minimum
Maximum
Indeterminate
27,300 kkg/year
16
10
33,850 kkg/year
21,420 kkg/year
78 percent
45 kkg/year
9,100 kkg/year
2,100 kkg/year
790 kkg/year
63 percent
3 years
52 years
0 cubic meters/day
45 cubic meters/day
0 cubic meter/kkg
23 cubic meter/kkg
Sources of data are Stanford Research Institute, Directory of Chanical
Producers, U.S.A., 1977, U.S. Department of Connerce, Current Industrial
Reports, Decanber 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 Chanicals Industry
March, 1980.
706
-------
TABLE 21-2. STATUS OF REGULATIONS - EFFUUD7T LIMITATION GPTnCT.TTCS
SUBCATEGORY Copper Sulfate
SUBPART AJ (40 CFR 415.360, 5/22/75)
STANDARDS
BPCTCA
BATEA
NSPS
Product
Process
Para-
meters
Max.(1)
(kg/kJcg )
(mg/1)
Avg,(2)
(kg/Wcg)
(mg/1)
Max. Avg.
(kg/kkg) (kg/kkg)
(mg/1) (mg/lf
Max. Avg.
Pure Raw
Materials
Process
Cu
0.0006
0.0002
Recovery
Process
TSS
CU
Ni
Se
0.069
0.003
0.006
0.0015
0.023
0.001
0.002
0.0005
(1)- Max. = Maximum of any one day.
(2)- Avg. = Maximum average of daily values for thirty consecutive days.
707
-------
If pure copper is used as a raw material, the resulting
copper sulfate solution is pure enough to be either sold, or fed
to crystallizers producing copper sulfate crystals. If impure
copper feed, or copper refinery waste is used, the concentrated
copper sulfate solution is filtered to remove other metal
impurities. This purified solution can be sold as is or fed to
the crystallizer. Copper sulfate crystals are recovered by
centrifugation, dried at ^110° C, screened and then packed dry
for sale. The mother liquor is recycled to the evaporator or
crystallizer with some being purged to prevent impurities
buildup. The purges are usually sold for metal recovery.
Figure 21-1 shows a general process flow diagram for the
manufacture of copper sulfate.
21.2 WATER USE AND WASTE SOURCE CHARACTERISTICS
21.2.1 Water Dse
Water is used in copper sulfate production as the reaction
medium, and it may be evaporated to the atmosphere during
crystallization or it becomes part of the dry product as its
water of crystallization (hydration). Noncontact cooling water,
including steam condensate, consitutes the major water use.
Water is also used for pump seals and washdowns. Table 21-3
gives a summary of plant water usages found in this study for
facilities where information was available from 308
Questionnaire responses and previous documents.
21.2.2 Waste Sources
Noncontact Cooling Water
Noncontact cooling water is used to cool the crystallizers
and constitutes one of the main wastes. This waste stream
should not be contaminated by process leaks, and therefore can
be discharged without treatment.
Washdowns, Leaks, and Spills
Washdown, pump seal leaks, and spills are sources of
contact waste water. These flows, however, are relatively small
and intermittant, and do not represent a major waste source.
Waste waters emanating from this source are either combined with
the mother liquor, or treated and discharged.
708
-------
WATER
ELECTROLYTE
FRCM COPPER
REFINER OR "
SHOT COPPER
SULFURIC
ACID
WASH FILTER
WATER CAKE
CRYSTAL-
LI ZER
CENTRIFUGE
EVAPORATOR
DRYER
FILTER
LIQUOR
I3IiEFl>
Fiqure 21-1. (General block diaqram of the manufacture of copper sulfate.
-------
TABLE 21-3. WATER USAGE IN COPPER SULFATE SUBCATEGORY
Source #034
Process * 1.21^
Contact
Noncontact 19.6 0
Cooling
#571
24.8 3.30 0.075
37.3' 105 0
Water Usage at Plants (rn^/kkg)
#284 #313(1) #069
(2)
Maintenance 1.25 0.35 0.28 3.77 0.017
Cleaning and
Washaown, Pumps
Seals and Leaks
Steam 38.6 0 0 0 0
Air Pollution 0 0.52 0 0 0
Control
(1) Includes uses for other processes
(2) Maxiumum - includes groundwater infiltration
* Utilizes feed solution frcm another industry
710
-------
Mother Liquor Purges
A small portion of the mother liquor is purged periodically
from the process to prevent buildup of metal impurities. The
amount of purge is variable and depends on the purity of
feedstock. These purges are processed to separate metallic
salts, particularly those of copper and nickel, from the
impurities. These recovered metallic salts are used for other
processes while the impurities are disposed of at an approved
landfill.
Steam Condensate
A few plants use evaporators to concentrate the production
solution. Steam condensate is an additional noncontact waste
water formed in the process. This can also be discharged
without treatment.
Sludge
Solid waste is generated in product purification by the
filtration step. This is necessary only for plants utilizing
impure copper, or copper refinery waste, as raw material. These
filter sludges contain metallic impurities or copper sulfides
and need disposal at an approved landfill.
Plants that produce copper sulfate in liquid form have no
contact waste streams from the process. Plants utilizing pure
copper feedstock are able to recycle most contact waste waters
and generally have no discharge of contact wastes. Table 21-4
summarizes the quantities of waste water that go to the
treatment facility, their sources, and the handling practices
for plants which do not discharge waste waters. The data was
taken from 308 Questionnaire responses, previous development
documents, and industry visits.
21.3 DESCRIPTION OF PLANTS VISITED AND SAMPLED
21.3.1 Screening
Plant #034 was visited and process waste water and effluent
samples were collected and analyzed for conventional and toxic
pollutants. The process used at this plant is similiar to that
described earlier, for one which utilizes a waste stream from a
copper refining facility as its feedstock. The feedstock is
strengthened by the addition of copper shot. The filter cake
and wash water are sent to a settler where the cake and wash
water are finally separated. The decant of the settler is
recycled back to the reactor, while the settled sludge is sent
711
-------
TABLE 21-4. WASTE WATER FLOW FOR THE COPPER SULFATE SUBCATEGORY
Avg. Waste Waste Water
Water Flow to Handling
Treatment Practice
Plant (rr.Vkkg of CuSO^j)
SO 34
#284
#313
#069
#571
#885
#458
#100
#969
*050
0.94
0.52
23.4*
4.01
0
0
0
0
0
0
Segregated treatment of
CuSO^ waste
(lime treatment)
Waste streams and treatment
are contained with other
mining, milling and man-
ufacturing prooess wastes.
Waste streams and treatment
are combined with other
metal prooess wastes.
Waste streams are contained
with waste from other re-
agent grade processes and
discharged to sewer.
No discharge of waste frcm the
process (recycle)
No discharge of waste frcm the
process (recycle)
No discharge of waste frcm the
prooess (recycle)
No discharge of waste from the
prooess (recycle)
No discharge of waste frcm the
process (recycle)
No discharge of waste frcm the
prooess (recycle)
* Flow is for the contained waste from all process per kkg of CuSO..
Actual amount of flow contributed by CuSO^ process is unavailable.
712
-------
to another process for melting. Mother liquor purges from the
centrifuge are also sent to other processes. Leaks, spills and
washdown water flow down to a sump in the basement of the
facility where it collects with contaminated ground water, and
is then pumped to holding tanks. About one quarter of this
waste water volume is comprised of contaminated ground water
from the immediate area. From the holding tanks, the waste goes
to the treatment facility where it is treated with lime,
filtered and discharged to a collection tank.
The uncontaminated steam condensate from the evaporator,
and noncontact cooling water from the crystallizer, are combined
with the effluent from the lime treatment in a collection tank.
The combined stream passes through a cloth filter for final
polishing and is discharged to a sewer. The filter residue from
the filter press is hauled to an approved landfill site. Figure
21-2 shows the general process and treatment flow diagram with
the location of the sampling points. Table 21-5 presents flow
data, total suspended solids (TSS) , and copper and nickel
emissions for the various waste streams sampled during
screening.
21.3.2 Ver ification
Plant #034 was sampled again during the verification phase.
Prior to this, the system was changed so that only the effluent
from lime treatment goes to the collection tank and through the
cloth filter. This effluent then combines with the steam
condensate and noncontact cooling water waste streams after the
cloth filter and discharges to the sewer.
Figure 21-1 also shows this change, and the subsequent new
sample points for verification phase sampling. Table 21-5 also
gives flow and discharge data for various waste streams sampled
during verification.
Plant #034 was the only plant sampled for the copper
sulfate subcategory. During the program, an attempt was made to
locate other candidates for sampling. A search was conducted
using the 308 questionnaires, published materials and the
telephone. Out of the 17 other facilities, 11 have no discharge
of process waste waters (practice recycle); four plants were
large multi-product complexes with combined waste treatment
systems where segregation of copper sulfate process wastes was
impossible; and two plants produced only reagent grade product,
and are therefore low volume producers.
713
-------
mrraxjrn rmj«
»*»«**, CU9HJ*. (MlNWna
(icrsmiux
ui
icnoi uum
10 emeu rmiss
,
-------
TABLE 21-5. FDCW AND POLLUTANT CONCENTRATION DATA OF THE SAMPLED WASTE
STREAMS FOR PLANT #034 PRODUCING COPPER SULFATE
Stream Sampled Unit Flew TSS Cu Ni
No. Stream (m3/kkg of Cu90^ (all in kg/kkg of CuS04)
Description
Screening (1)
1
2
1
2
CuS04 waste * 1.25 0.087
Effluent from 1.25 0.078
lime treatment
Steam Condensate 0.209
Verification (2)
Cu9D4 waste * 1.25
Elf fluent from
lime treatment
Nonoontact
Cooling Water
and Steam
Condensate
1.25
14.2
0.00021
1.8
0.030
0.11
4.2
0.010
0.00016
5.0
0.0042
0.024
0.25
0.00053
0.000025
0.20
0.00038
0.0020
(1) From grab samples ccrposited during the period of batch manufacturing
and treatment process.
(2) Average of three daily grab samples composited during the period of
batch manufacturing and treatment process.
* Infiltration of ground water into the collection sump was suspected at
the time of sampling.
715
-------
21.3.3 Toxic Pollutant Concentrations
The following toxic pollutants were found at detectable
concentrations in the raw waste samples at copper sulfate Plant
$034 during screening and verification sampling.
Maximum Raw Waste Concentration Observed
(ug/1)
Pollutant Screening Verification
Antimony
330
1,300
Arsen ic
3,500
127,000
Cadmium
870
2,500
Chromium
140
940
Copper
1,850,000
3,940,000
Lead
180
2,200
Nickel
112,000
136 ,000
Z i nc
11,000
17 ,000
1,1,1-tr ichloroethane
240
NA
NA = Not analyzed
A large portion of the raw waste water at this plant
consists of ground water which seeps and collects in the
basement, along with leaks and washdown water from the process.
The ground water is contaminated from the surrounding area which
is heavily industrialized. The trichloroethane is presumed to
be external contamination because this chemical is not used in
the process.
No other organic toxic pollutants were found at significant
concentrations during screening sampling. Consequently, no
organic toxic pollutants were analyzed for in the verification
phase.
Section 5.1.2 of this report describes the methodology of
the screening and verification sampling program. In the copper
sulfate industry, a total of 6 days of sampling were conducted
at Plant f034. Five different sampling points were
involved covering the various raw wastes, and the intermediate
and treated effluent streams. The evaluation of toxic metal
content of these process related waste streams was based on 221
analytical data points. The screening for toxic organic
pollutants at Plant f034 generated an additional 456 analytical
data points. The unit loadings were calculated from the waste
stream flow rates measured or estimated at the time of sampling,
the measured pollutant concentration, and the reported copper
sulfate production rate.
716
-------
That is,
Unit loading (as kg of pollutant per
kkg of copper sulfate) = (C) (Q)
1000 (P)
Where:
C is the concentration of the pollutant expressed in
units of mg/1 (Note: kg/m3 = mg/1),
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), and
P is the copper sulfate production rate expressed in
units of kkg/day (kkg is 1000 kg, a metric ton, which
is equal to 2205 lbs).
The average values are based on data from Plant #034 where
the particular pollutant was found at concentrations greater
than the analytical detection limits and in significant
concentrations since it could be treated by an available
treatment technology.
In Table 21-6, the toxic pollutant raw waste data are
presented as the average daily concentrations and the unit
loading found during each sampling at Plant #034. The overall
averages are also shown. It is this overall average which is
used as the average raw waste load from the copper sulfate
process in various calculations.
Based on the total annual production rate of this
subcategory and the average waste load generated per unit
product, the estimated total pollutant raw waste loads generated
each year for this subcategory are as follows:
Pollutant
Waste Load (kg/year)
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
124,000
30
6200
700
26
1400
74
15
N ickel
Z inc
717
-------
TABLE 21-6. RAW WASTE DATA
Subcategory: Copper Sulfate
Average Daily
Pollutant
iraqc
Antimony, Sb
Arsenic, As
Cattni'jn, Cd
Copper, Cu
Lead, %
Nickel, Ni
Zinc, Zn
Chrcruirr., Cr
Seleniun, Se
CCNVENTICNAL
TSS
Pollutant Concentrations anc},Loadings fovx.d during Sampling of
Plant t034v"'
Screening
0.31
0.00069
3.5
0.0078
0.37
0.0019
1900
4.2
0.18
0.00039
110
0.25
11.0
0.024
0.14
0.000030
< 0.011
< 0.000024
39.0
0.087
(wq/l)
(kg/kkg of CuSC4.5H2G)
(2)
Verification
!3)
Overall Average
(4)
0.54
0.0012
44.0
0.097
1.6
0.0035
2200
5.0
0.78
0.0018
91.0
0.20
12.0
0.027
0.36
0.000080
< 0.0050
< 0.000011
0.44
0.00095
24.0
0.052
1.2
0.0027
2 000
4.5
0.48
o.oon
1C2
0.23
12.0
0.026
0.25
0.00055
< 0.008
< 0.000018
790
1.80
410
0.92
iy>e methodology of the sarpling program is described in Section 5.1.2,
and Section 21.1.2 presents the scope of sarplina ir. the Copper
Sulfate industry.
Screening data frcn ore 72-hour grab aarposite sample of individual or
oorbined raw waste stream.
(2)
(3)
(4)
Verification data frtsr three 24-hour grab ocnposite sarples, averaged.
fchan averaging values indicated as "Less than" (<), the absolute value
was used and the resulting average was indicated as a "less than" value.
718
-------
21.4
POLLUTION ABATEMENT OPTIONS
21.4.1 Toxic Pollutants of Concern
The principal pollutant of concern is copper. The other
toxic pollutants found in plant waste waters are closely related
to the purity of the copper and acid sources. The heavy metals;
cadmium, nickel, zinc and, to a lesser extent, antimony,
chromium and lead, which were found during field sampling, may
originate as trace impurities in copper scrap and other copper
sources. Plants utilizing pure copper shot would not experience
a buildup of these impurities in the mother liquor, and
consequently would not generate a waste stream
|