440182007
United States
EnvironmentaTPri
Agency c
Water and Waste Management
iuideiines Division
WH-652
Washington DC 20460
Development
Document for
Effluent Limitations
Guidelines and
Standards for the
Inorganic Chemicals
Manufacturing
Point Source Category
-------
-------
DEVELOPMENT DOCUMENT
for
EFFLUENT LIMITATIONS GUIDELINES
NEW SOURCE PERFORMANCE STANDARDS
and
PRETREATMENT STANDARDS
for the
INORGANIC CHEMICALS MANUFACTURING
POINT SOURCE CATEGORY
Anne M. Gorsuch
Administrator
Steven Schatzow
Director
Office of Water Regulations and Standards
m
Jeffery Denit, Acting Director
Effluent Guidelines Division
G. Edward Stigall
Chief, Inorganic Chemicals Branch
Dr. Thomas Fielding
Dwight Hlustick
Elwood E. Martin
June 1982
Effluent Guidelines Division
Office of Water Regulations and Standards
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
-------
TABLE OF CONTENTS
LIST OF FIGURES xx
LIST OF TABLES xxviii
ACKNOWLEDGEMENTS xliv
CONCLUSIONS AND SUMMARY 1
TOXIC POLLUTANTS 1
CONTROL AND TREATMENT TECHNOLOGY 2
COSTS OF ADDITIONAL IN-PLAtfT TREATMENT 2
SUBCATEGORIZATION 2
RESTUDY OF REMANDED REGULATIONS 3
RECOMMENDATIONS 5
INTRODUCTION 17
AUTHORITY 17
The Federal Water Pollution 17
Control Act Amendments
Court Remand of Regulations 19
The Settlement Agreement 21
GENERAL APPROACH AND METHODOLOGY 25
Industry Data Base Development 26
and Subcategorization Review
The Screening and Verification 26
Sampling Programs
Engineering Evaluations 26
Treatment System Cost Estimates 2b
Treatability Study 2/
GENERAL CRITERIA FOR EFFLUENT LIMITATIONS 27
BPT, Effluent Limitations 27
BAT Effluent Limitations 28
BCT Effluent Limitations. 28
New Source Performance 30
Standards
Pretreatment Standards for 30
Existing Sources
Pretreatment Standards for 31
New Sources
-------
TABLE OF CONTENTS - Continued
SOBCATEGORIZATION REVIEW
BASIS FOR SUBCATEGORIZATION
Factors Considered
General Conclusions
SECONDARY SUBCATEGORIZATION
Chlor-Alkali
Titanium Dioxide
Hydrogen Cyanide
REVIEW OF POSSIBLE INTEGRATION OF
SUBCATEGORIES
Hydrofluoric Acid and Aluminum
Fluoride
SUMMARY
SCREENING AND VERIFICATION SAMPLING PROGRAMS
SCOPE AND METHODOLOGY
Selecting Plants and Making
Preliminary Contacts
Screening and Verification
Sampling
Analytical Methodology for
Toxic Pollutants
Quality Assurance Provisions
SUMMARY OF ANALYTICAL RESULTS
PROCESS AND WASTE TREATMENT INFORMATION
DEVELOPMENT AND EVALUATION
INDUSTRY DATA BASE DESCRIPTION
Literature Review
Plant Visits
Telephone and Direct Contact
308-Questionnaire Responses
PROCESS WASTE SOURCES AND CURRENT
TREATMENT PRACTICES
Data Acquisition
Evaluation of Data
Model Plant and BPT Treatment
System Specification
Dissolved Solids in Wastewater
Effluents
38
39
41
41
41
42
45
50
53
59
59
59
59
59
59
60
60
60
62
63
-------
TABLE OF CONTENTS - Continued
Page
7. ASSESSMENT OF TECHNOLOGY FOR ADVANCED TREATMENT 65
AND CONTROL
INTRODUCTION 65
HYDROXIDE PRECIPITATION 66
FERRITE COPRECIPITATION 70
SULFIDE PRECIPITATION 71
THE XANTHATE PROCESS 74
ION EXCHANGE 76
REDUCTION PROCESSES 2B
OXIDATION PROCESSES 79
MEMBRANE PROCESSES 82
ADSORPTION 83
FLUORIDE REMOVAL 87
CHLORINE REMOVAL 88
8. TREATABILITY ESTIMATES AND LONG-TERM DATA 91
ANALYSIS
THE DEVELOPMENT OF TREATABILITY ESTIMATES 91
Preliminary Analysis 91
Final Analysis 101
SELECTION OF TOXIC METAL CONTROL 109
PARAMETERS
Control Parameters for 109
Hydroxide Precipitation
Control Parameters for 109
Sulfide Precipitation
THE USE OF HISTORICAL POLLUTANT 111
DATA
Determination of Limitation 111
Guidelines Based Upon
Historical Performance
Assumptions Concerning Daily 112
Pollutant Level Measurements
Assumptions Concerning 30-Day 117
Average Pollutant Level
Observation
9. TREATMENT TECHNOLOGY APPLICATIONS FOR TOXIC 125
POLLUTANT REMOVAL
SELECTION OF POLLUTANTS TO BE CONTROLLED 125
APPLICATION OF ADVANCE LEVEL TREATMENT 125
AND CONTROL ALTERNATIVES
General Design Objectives 125
Pretreatment Technology 128
111
-------
TABLE OF CONTENTS - Continued
11.
New Source Performance
Standards
ESTIMATED ACHIEVABLE PERFORMANCE
CHARACTERISTICS FOR ADVANCED LEVEL
APPLICATIONS
Advanced Level Removal of BPT
Pollutants
Advanced Level Removal of Toxic
Pollutants
POLLUTION CONTROL PARAMETERS TO BE
REGULATED
Conventional Pollutants
Nonconventional Pollutants
Toxic Pollutants
COST OF TREATMENT AND CONTROL SYSTEMS
INTRODUCTION
Purpose of Cost Data
General Approach
Cost References and Rationale
Definition of Levels of
Treatment and Control Cost
Development
Treatment and Disposal
Rationale Applied to Cost
Development
Expression of Costs
COST ESTIMATES FOR EACH SUBCATEGORY
CHLOR-ALKALI INDUSTRY
INDUSTRY PROFILE
General Description
Subcategorization
MERCURY CELL PROCESS INDUSTRY
PROFILE
General Description
General process Description
WATER USE AND WASTEWATER SOURCE
CHARACTERISTICS
Water Use
Waste Sources
DESCRIPTION OF SPECIFIC PLANTS
Screening Program
Verification
128
129
129
129
130
130
133
133
133
134
134
134
135
135
142
145
145
145
145
146
146
146
149
149
151
153
153
153
IV
-------
TABLE OF CONTENTS - Continued
Page
—^•h.^AdM,
Descriptions of Plants Not 158
S ampled
Summary of the Toxic Pollutant 163
Data
POLLUTION ABATEMENT OPTIONS 165
Toxic Pollutants of Concern 165
Prevailing Control and 166
Treatment Practices
Process Modifications and 166
Technology Transfer Options
Best Management Practices 170
Advanced Treatment Technologies 170
SELECTION OF APPROPRIATE TECHNOLOGY AND 170
EQUIPMENT
Technologies for Different 170
Treatment Levels
Equipment for Different 171
Treatment Levels
TREATMENT COST ESTIMATES 172
General Discussion 172
Chlorine Bearing Wastes 172
Model Plant Treatment Costs 175
BASIS FOR REGULATIONS 175
Basis for BPT Limitations 1/5
Basis for Final BAT Effluent 176
Limitations
Basis for BCT Effluent 192
Limitations
Basis for New Source 192
Performance Standards
Basis for Pretreatment 192
Standards
DIAPHRAGM CELL PROCESS INDUSTRY PROFILE 192
General Description 193
General Process Description 193
WATER USE AND WASTE WATER SOURCES 196
Water Use 196
Waste Sources 195
DESCRIPTIONS OF SPECIFIC PLANTS 199
Screening 199
Verification 200
Descriptions of Plants Not 200
S ampled
Toxic Pollutant Concentrations 202.
POLLUTION ABATEMENT OPTIONS 218
Toxic Pollutants of Concern 218
v
-------
TABLE OF CONTENTS - Continued
12.
Prevailing Control and
Treatment Practices
Process Modifications and
Technology Transfer Options
Best Management Practices
Advanced Treatment Technologies
SELECTION OF APPROPRIATE TECHNOLOGY AND
EQUIPMFNT
Technologies for Different
Treatment Levels
Equipment for Different
Treatment Levels
TREATMENT COST ESTIMATES
General Discussion
Model Plant Treatment Costs
BASn FOR REGULATIONS
Basis for BPT Limitations
Basis for BAT Effluent
Limitations
BCT Limitations
Basis for New Source
Performance Standards
Basis for Pretreatment
Standards
HYDROFLUORIC ACID INDUSTRY
INDUSTRY PROFILE
General Description
Subcategorization
General Process Description and
Raw Materials
WATER USE AND WASTE SOURCE
CHARACTERISTICS
Water Use
Waste Sources
DESCRIPTION OF PLANTS VISITED AND SAMPLED
Screening
Verification
Summary of the Toxic Pollutant
Data
POLLUTION ABATEMENT OPTIONS
Toxic Pollutants of Concern
Process Modifications and
Technology Transfer Options
Best Management Practices
218
223
224
224
225
225
22P
227
727
233
233
233
249
254
255
256
261
261
261
261
264
267
267
267
272
272
272
278
282
282
282
285
VI
-------
TABLE OF CONTENTS - Continued
13,
14,
Prevailing Control and
Treatment Practices
Advanced Treatment Technologies
SELECTION OF APPROPRIATE TECHNOLOGY AND
EQUIPMENT
Technologies for Different
Treatment Levels
Equipment for Different
Treatment Levels
TREATMENT COST ESTIMATES
General Discussion
Model Plant Control Costs for
Existing Sources
BASIS FOR REGULATIONS
Evaluation of BPT Treatment
Practices
Basis for BPT Effluent
Limitations
Basis for BCT Effluent
Limitations
Basis for BAT Effluent
Limitations
Basis for the New Source
Performance Standards
Basis for the Pretreatment
Standards
HYDROGEN PEROXIDE INDUSTRY
SUMMARY OF DETERMINATIONS
ASSESSMENT OF THE WATER POLLUTION
POTENTIAL
Production Processes and
Effluents
Plants
Toxic Pollutants
STATUS OF REGULATIONS
TITANIUM DIOXIDE INDUSTRY
TITANIUM DIOXIDE, CHLORIDE PROCESS
INDUSTRY PROFILE
General Description
Subcategorization
General Process Description and
Raw Materials
286
287
287
288
293
293
295
295
295
303
316
31 fi
323
323
327
327
327
327
327
329
329
331
331
331
331
331
VI i
-------
TABLE OF CONTENTS - Continued
WATER USE AND WASTE SOURCE 335
CHARACTERISTICS
Water Use 335
Waste Sources 335
DESCRIPTION OP PLANTS VISITED AND SAMPLED 338
Screening 338
Verification 338
Toxic Pollutant Concentrations 340
POLLUTION ABATEMENT OPTIONS 345
Toxic Pollutants of Concern 345
Process Modification and 345
Technology Transfer Options
Best Management Practices 345
Prevailing Control and 347
Treatment Practices
Advanced Treatment Technologies 348
SELECTION OF APPROPRIATE TECHNOLOGY AND 348
EQUIPMENT
Technologies for Different 348
Treatment Levels
Equipment for Different 349
Treatment Levels
TREATMENT COST ESTIMATES 349
General Discussion 349
Model Plant Control and 354
Treatment Costs
BASIS FOR REGULATIONS 354
Evaluations of BPT Treatment 354
Practices
Basis for BPT Effluent 36!
Limitations
Basis for BCT Effluent 370
Limitations
Basis for BAT Effluent 371
Limitations
Basis for New Source 374
Performance Standards
Basis for Pretreatment 377
Standards
TITANIUM DIOXIDE - SULFATE PROCESS 377
INDUSTRY PROFILE
General Description 377
General Process Description and 381
Raw Materials
WATER USE AND WASTE SOURCE 384
CHARACTERISTICS
Water Use 384
Waste Sources 384
Vlll
-------
TABLE OF CONTENTS - Continued
DESCRIPTION OF PLANTS
Screening
Verification
Other Plant Descriptions
Toxic Pollutant Concentrations
POT"T)TION ABATEMENT OPTIONS
Toxic Pollutants of Concern
Process Modifications and
Technology Transfer Options
Best Management Practices
Prevailing Control and
Treatment Practices
Advanced Treatment Technologies
SELECTION OF APPROPRIATE TECHNOLOGY AND
EQUIPMENT
Technologies for Different
Treatment Levels
Equipment for Different
Treatment Levels
TREATMENT COST ESTIMATES
General Discussion
Model Plant Control and
Treatment Costs
BASIS FOR REGULATIONS
Evaluation of BPT Practices
Basis for BPT Effluent.
Limitations
Basis for BCT Effluent
Limitations
Basis for BAT Effluent
Limitations
Basis for New Source
Performance Standards
Basis for Pretreatment
Standards
TITANIUM DIOXIDE - CHLORIDE ILMENITE
PROCESS INDUSTRY PROFILE
General Description
General Process Description and
Raw Materials
WATER USE AND WASTE SOURCE
CHARACTERISTICS
Water Use
Waste Sources
DESCRIPTION OF PLANTS VISTED AND SAMPLED
Screening
Verification Program
Toxic Pollutant Concentration
385
385
387
387
390
395
395
396
396
396
396
396
396
400
402
402
403
404
404
404
417
417
417
424
427
427
427
4 3D
430
430
431
431
431
431
-------
TABLE OF CONTENTS - Continued
15.
POLLUTION ABATEMENT OPTIONS
Toxic Pollutants of Concern
Process Modifications and
Technology Transfer Options
Best Management Practices
Prevailing Control and
Treatment Practices
Advanced Treatment Technology
SELECTION OF APPROPRIATE TECHNOLOGY AND
EQUIPMENT
Technologies for Different
Treatment Levels
Equipment for Different
Treatment Levels
TREATMENT COST ESTIMATES
General Discussion
Model Plant Control and
Treatment Costs
BASIS FOR REGULATIONS
Evaluation of BPT Treatment
Practices
Basis for BPT Effluent
Limitation
Basis for BCT Effluent
Limitations
Basis for BAT Effluent
Limitations
Basis for the New Source
Performance Standards
Basis for Pretreatment
Standards
ALUMINUM FLUORIDE INDUSTRY
INDUSTRY PROFILE
General Description
General Process Description and
Raw Materials
WATER USE AND WASTE SOURCE
CHARACTERISTICS
Water Use
Waste Sources
DESCRIPTION OF PLANTS VISITED AND SAMPLED
Screening
Verification
Summary of the Toxic Pollutant
Data
Page
435
435
436
436
436
437
437
440
441
441
442
452
452
452
456
456
456
460
461
461
461
461
461
461
461
465
465
467
467
x
-------
TABLE OF CONTENTS - Continued
16.
Page
POLLUTION ABATEMENT OPTIONS 476
Toxic Pollutants of Concern 476
Process Modifications and 476
Technology Transfer Options
Best Management Practices 477
Prevailing Control and 477
Treatment Practices
Advanced Treatment Technologies 477
SELECTION OF APPROPRIATE TECHNOLOGY AND 478
EQUIPMENT
Technologies for Different 478
Treatment Levels
Equipment for Different 482
Treatment Levels
TREATMENT COST ESTIMATES 484
General Discussion 484
BASIS FOR REGULATIONS 485
Evaluation of BPT Treatment 485
Practices
BPT Effluent Limitations 485
Basis for the BCT Effluent 495
Limitations
Basis for BAT Effluent 495
Limitations
Basis for New Source 496
Performance Standards
Basis for Pretreatment 498
Standards
CHROME PIGMENTS INDUSTRY 499
INDUSTRY PROFILE 499
General Description 499
Subcategorization 499
General Process Description and 499
Raw Materials
WATER USE AND WASTE SOURCE 508
CHARACTERISTICS
Water Use 508
Waste Sources 508
DESCRIPTION OF PLANTS 513
Screening 513
Verification 513
Toxic Pollutant Concentrations 515
XI
-------
TABLE OP CONTENTS - Continued
17. „
POLLUTION ABATEMENT OPTIONS
Toxic Pollutants of Concern
Process Modifications and
Technology Transfer Options
Best Management Practices
Prevailing Control and
Treatment Practices
Advanced Treatment Technologies
SELECTION OF APPROPRIATE TECHNOLOGY AND
EQUIPMENT
Technologies for Different
Treatment Levels
Equipment for Treatment
TREATMENT COST ESTIMATES
General Discussion
Model Plant Costs
BASIS FOR REGULATIONS
Evaluation of BPT Treatment
Practices
Basis for BPT Effluent
Limitations
Basis for BCT Limitations
Basis for BAT Effluent
Limitations
Basis for New Source
Performance Standards
Basis for Pretreatment
Standards
HYDROGEN CYANIDE INDUSTRY
INDUSTRY PROFILE
General Description
Subcategorization
General Process Description and
Raw Materials
WATER USE AND WASTE SOURCE
CHARACTERISTICS
Water Use
Waste Sources
DESCRIPTION OF PLANTS VISITED AND SAMPLED
Screening
Verification
Toxic Pollutant Concentrations
Page
525
525
526
526 :
527
528
528
528
530
532
532
533
53?
B33
540
548
548
549
550
551
551
551
551
551
554
554
554
557
557
559
562
XII
-------
TABLE OF CONTENTS - Continued
18.
POLLUTION ABATEMENT OPTIONS 568
Toxic Pollutants of Concern 568
Process Modifications and 568
Technology Transfer Options
Best Management Practices 568
Prevailing Control and 568
Treatment Practices
Advanced Treatment Technologies 570
SELECTION OF APPROPRIATE TECHNOLOGY AND 571
EQUIPMENT
Technologies for Different 571
Treatment Levels
Equipment for Different 571
Treatment Levels
TREATMENT COST ESTIMATES 573
General Discussion 573
BASIS FOR REGULATIONS 575
Evaluation of BPT Treatment 575
Practices
Basis for BPT Limitations 576
Basis for BCT Limitations 586
Basis for BAT Limitations 586
Basis for New Source 588
Performance Standards
Basis for Pretreatment 590
Standards
SODIUM DICHROMATE INDUSTRY 593
INDUSTRY PROFILE 593
General Description 593
Subcategorization 593
General Process Description and 593
Raw Materials
WATER USE AND WASTE SOURCE 596
CHARACTERISTICS
Water Use 596
Waste Sources 596
DESCRIPTION OF PLANTS VISITED AND SAMPLED 599
Screening 599
Verification 599
Toxic Pollutant Concentrations 602
and Loadings
xxn
-------
TABLE OF CONTENTS - Continued
POLLUTION ABATEMENT OPTIONS
Toxic Pollutants of Concern
Process Modifications and
Technology Transfer Options
Best Management Practices
Prevailing Control and
Treatment Practices
Advanced Treatment Technologies
SELECTION OF APPROPRIATE TECHNOLOGY AND
EQUIPMENT
Technology for Different
Treatment Levels
TREATMENT COST ESTIMATES
General Discussion
Model Plant Control Costs
BASIS FOR REGULATIONS
BPT Effluent Limitations
BCT Effluent Limitations
BAT Effluent Limitations
NSPS Effluent Limitations
Pretreatment Standards
19. CARBON DIOXIDE INDUSTRY
SUMMARY OF DETERMINATIONS
ASSESSMENT OF THE WATER POLLUTION
POTENTIAL
Production Processes and
Effluents
STATUS OF REGULATIONS
20. CARBON MONOXIDE AND BY-PRODUCT HYDROGEN INDUSTRY
SUMMARY OF DETERMINATIONS
ASSESSMENT OF THE WATER POLLUTION
POTENTIAL
Production Processes and
Effluents
STATUS OF REGULATIONS
21 . COPPER SULFATE INDUSTRY
INDUSTRIAL PROFILE
General Description
Subcategorization
General Process Description and
Raw Materials
606
606
606
609
609
609
610
610
614
614
615
615
615
625
625
625
626
627
627
627
627
627
629
629
629
629
629
631
631
631
631
631
xiv
-------
TABLE OF CONTENTS - Continued
WATER USE AND WASTE SOURCE
CHARACTERISTICS
Water Use
Waste Sources
DESCRIPTION OF PLANTS VISITED AND SAMPLED
Screening
Verification
Toxic Pollutant Concentrations
POLLUTION ABATEMENT OPTIONS
Toxic Pollutants of Concern
Process Modifications and
Technology Transfer Options
Best Management Practices
Prevailing Control and
Treatment Practices
Advanced Treatment Technologies
SELECTION OF APPROPRIATE TECHNOLOGY AND
EQUIPMENT
Technologies for Different
Treatment Levels
Equipment for Different
Treatment Levels
TREATMENT COST ESTIMATES
General Discussion
Model Plant Cost Estimates
BASIS FOR REGULATIONS
Evaluation of BPT Treatment
Practices
Basis for BPT Effluent
Limitations
Basis for BCT Effluent
Limitations
Basis for BAT Effluent
Limitations
Basis for New Source
Performance Standards
Basis for Pretreatment
Standards
Basis for Level 1 Treatment
Performance
634
634
634
637
637,
639
639
643
643
645
645
645
646
22. NICKEL SULFATE INDUSTRY
INDUSTRIAL PROFILE
General Description
Subcategorization
General Process Description and
Raw Materials
646
647
,649
649
651
651
651
652
663
663
664
664
664
673
673
673
673
673
xv
-------
TABLE OF CONTENTS - Continued
23.
WATER USE AND WASTE SOURCE 676
CHARACTERISTICS
Water Use 676
Waste Sources 676
DESCRIPTION OF PLANTS VISITED AND SAMPLED 676
Screening 676
Verification 679
Summary of Toxic Pollutant Data 679
POLLUTION ABATEMENT OPTIONS 685
Toxic Pollutants of Concern 685
Process Modifications and 687
Technology Transfer Options
Best Management Practices 687
Prevailing Control and 687
Treatment Practices
Advanced Treatment Technologies 688
SELECTION OF APPROPRIATE TECHNOLOGY AND 688
EQUIPMENT
Technologies for Different 688
Treatment Levels
Equipment for Different 689
Treatment Levels
TREATMENT COST ESTIMATES 694
General Discussion 694
Model Plant Control Costs 694
BASIS FOR REGULATIONS 695
Evaluation of BPT Treatment 695
Practices
Basis for BPT Effluent 695
Limitations
Basis for BCT Effluent 700
Limitations
Basis for BAT Effluent 7po
Limitations
New Source Performance 708
Standards
Basis for pretreatment 708
Standards
SILVER NITRATE INDUSTRY 709
SUMMARY OF DETERMINATIONS 709
ASSESSMENT OF THE WATER POLLUTION 709
POTENTIAL
Production Processes and 709
Effluents
STATUS OF REGULATIONS 709
xvi
-------
TABLE OF CONTENTS - Continued
24. SODIUM BISULFITE INDUSTRY
INDUSTRY PROFILE
General Description
Subcategorization
General Process Description and
Raw Materials
WATER USE AND WASTE SOURCE
CHARACTERISTICS
Water Use
Waste Sources
DESCRIPTION OF PLANTS VISITED AND SAMPLED
Screening
Verification
Toxic Pollutant Analytical
Results
POLLUTION ABATEMENT OPTIONS
Toxic Pollutants of Concern
Prevailing Control and
Treatment Practices
Advanced Treatment Technologies
SELECTION OF APPROPRIATE TECHNOLOGY AND
EQUIPMENT
Technologies for Different
Treatment Levels
Equipment for Different
Treatment Levels
TREATMENT COST ESTIMATES
General Discussion
Cost Estimates
BASIS FOR REGULATIONS
Evaluation of BPT Treatment
Practices
Basis for BPT Effluent
Limitations
Basis for BCT Effluent
Limitations
Basis for BAT Effluent
Limitations
Basis for New Source
Performance Standards
Basis for Pretreatment
Standards
711
711
714
714
714
714
714
724
724
728
728
728
728
729
733
733
734
734
734
734
746
746
747
747
xvii
-------
TABLE OF CONTENTS - Continued
25
26
SODIUM HYDROSULFITE (FORMATE PROCESS) Industry (Excluded)
SUMMARY OF DETERriMATIOMS
INDUSTRY PROFILE
General Description
Subcategorization
General Process Description and
Raw Materials
WATER USE AND WASTE SOURCE
CHARACTERISTICS
Water Use
Waste Sources
DESCRIPTION OF PLANTS VISITED AND SAMPLED
Screening and Verification
Toxic Pollutant Concentrations
POLLUTION ABATEMENT OPTIONS
Toxic Pollutants of Concern
Prevailing Control and
Treatment Practices
Advanced Treatment Technologies
SELECTION OF APPROPRIATE TECHNOLOGY AND
EQUIPMENT
Technologies for Different
Treatment Levels
Equipment for Different
Treatment Levels
TREATMENT COST ESTIMATES
General Discussion
Cost Estimates
BASIS FOR GUIDANCE
Evaluation of Level 1
Treatment Practices
Basis for Level 1 Treatment
Performance
Basis for Level 2 Treatment
Performance
EXCLUDED SUBCATEGORIES
ALUMINUM SULFATE
AMMONIUM CHLORIDE
AMMONIUM HYDROXIDE
BARIUM CARBONATE
BORAX
BORIC ACID
BROMINE
CALCIUM CARBIDE
CALCIUM CARBONATE
752
752
754
754
754
757
762
762
762
763
763
763
764
767
767
767
767
767
773
777
781
781
782
784
7R4
7P6
788
789
791
791
XVlll
-------
TABLE OF CONTENTS - Continued
•J
CHROMIC ACID
CUPROUS OXIDE
FERRIC CHLORIDE
FERROUS SULFATE
FLUORINE
HYDROCHLORIC ACID
HYDROGEN. „
IODINE
LEAD MONOXIDE
LITHIUM CARBONATE
MANGANESE SULFATE
NITRIC ACID
OXYGEN AND NITROGEN
POTASSIUM CHLORIDE
POTASSIUM DI£HRPMATE
POTASSIUM
-------
LIST OF FIGURES
Page
5-1 Sample flow sheet for metal analysis 47
7-1 Theoretical solubilities of toxic metal 68
hydroxides/oxides as a function of pH
7-2 Theoretical solubilities of toxic metal 73
sulfides as a function of pH
7-3 Electrodialysis process 85
8-1 Cumulative distribution of daily concentrations 115
of mercury in treated effluent from Plant 1251
8-2 Cumulative distribution of daily concentrations 116
of cyanide in treated effluent from Plant 1765
8-3 Statistical distribution for daily pollution 119
measurements
8-4 Cumulative distribution of 30-day averages 121
of total cyanide in treated effluent from Plant
#782
8-5 Cumulative distribution of 30-day averages 122
of ammonia in treated effluent from Plant #782
8-6 Statistical distributions for 30-day average 123
pollution measurements
11-1 General process diagram for production of 150
chlorine/caustic by mercury cells
11-2 General process flow diagram at Plant #299 155
showing the sampling points. Chlorine/caustic
(mercury cell) manufacture
11-3 General process flow diagram at plant #747. 159
showing the sampling points* Chlorine/caustic
(mercury cell) manufacture
XX
-------
LIST OF FIGURES - Continued
11-4 General process flow diagram at plant #167
showing the sampling points. Chlorine/caustic
(mercury cell) manufacture
11-5 General process flow diagram at Plant #317
showing the sampling points. Chlorine/caustic
(mercury cell) manufacture
11-6 Level 1 waste water treatment for
chlorine-mercury cell subcategory
11-7 Level 2 waste water treatment for
chlorine-mercury cell subcategory
11-8 General process flow diagram for production
of chlorine/caustic by diaphragm cells
11-9 General process flow diagram at Plant #014
showing the sampling points. Chlorine/caustic
(diaphragm cell) manufacture
11-10 General process flow diagram at Plant #261
showing the sampling points. Chlorine/caustic
(diaphragm cell) manufacture
11-11 General process flowsheet at Plant #738-A
showing the sampling points. Chlorine/caustic
(diaphragm cell) manufacture
11-12 General process flow diagram at Plant
#738-B showing the sampling points.
Chlorine/caustic (diaphragm cell) manufacture
11-13 General process flow diagram at Plant #736
showing the sampling points. Chlorine/caustic
(diaphragm cell) manufacture
11-14 General process flow diagram at Plant #967
showing the sampling points. Chlorine/caustic
(diaphragm cell) manufacture
11-15 Level 1 waste water treatment for
chlorine-diaphragm cell subcategory
11-16 Level 2 waste water treatment for
chlorine-diaphragm cell subcategory
11-17 Level 3 waste water treatment for
chlorine-diaphragm cell subcategory
161
173
174
198
203
206
207
208
209
210
228
229
230
xxi
-------
LIST OF FIGURES - Continued
Page
12-1 General process flow diagram for production 266
of hydrofluoric acid
12-2 production vs. waste flow data for HF plants 270
12-3 General process flow diagram at Plant #705 276
showing the sampling points. Hydrofluoric acid
manufacture
12-4 General process flow diagram at Plant #251 280
showing the sampling points. Hydrofluoric acid
manufacture
12-5 Level 1 waste water treatment for hydrofluoric 289
acid subcategory
12-6 Level 2 waste water treatment for hydrofluoric 290
acid subcategory
12-7 Level 3 waste water treatment for hydrofluoric 291
acid subcategory
12-8 Level 4 waste water treatment for hydrofluoric 292
acid subcategory
12-9 Level 5 waste water treatment for hydrofluoric 294
acid subcategory
12-10 Fluoride loads and concentrations discharged 308
at selected hydrofluoric acid plants
14-1 General process diagram for production of 336
titanium dioxide (chloride process) from high
grade ores
14-2 General flow diagram at Plant #559 showing 341
the sampling points. (Titanium dioxide -chloride
process manufacture)
14-3 General flow diagram at Plant #172 showing 343
the sampling points. Titanium dioxide (chloride
process) manufacture
14-4 Level 1 (BPT/BAT) waste water treatment for 350
titanium dioxide - chloride process
14-5 Level 2 (NSPS) waste water treatment for 351
titanium dioxide - chloride process
XXI1
-------
LIST OF FIGURES - Continued
14-6 Level 3 waste water treatment for titanium
dioxide - chloride process
14-7 General process flow diagram for production
of titanium dioxide by sulfate process
14-8 General flow diagram at plant #559 showing
the sampling points. (Titanium dioxide -
sulfate process)
14-9a Level 1 waste water treatment for titanium
dioxide - sulfate process
14-9b Level 2 waste water treatment for titanium
dioxide - sulfate process
14-10 Level 3 waste water treatment for titanium
dioxide - sulfate process
14-11 General process flow diagram of the titanium
tetrachloride portion of a. titanium dioxide
plant using the chloride-ilmenite process.
14-12 Level 1 waste water treatment for titanium
dioxide - chloride (ilmenite ore) process
14-13 Level 2 waste water treatment for titanium
dioxide - chloride (ilmenite ore) process
15-1 General process flow diagram for production
of aluminum fluoride
15-2 General process flow diagram at Plant #705
showing the sampling points. (Aluminum
fluoride manufacture)
15-3 General process flow diagram at Plant #251
showing the sampling points. (Aluminum
fluoride manufacture)
15-4 Level 1 waste water treatment for aluminum
fluoride subcategory
15-5 Level 2 waste water treatment for aluminum
fluoride subcategory
15-6 Level 3 waste water treatment for aluminum
fluoride subcategory
Page
352
382
388
398
399
401
426
438
439
463
468
470
479
480
481
XXlll
-------
LIST OF FIGURES - Continued
15-7
16-1
16-2
16-3
16-4
16-5
16-6
16-7
i6-a
Level 4 waste water treatment for aluminum
fluoride subcategory
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
General process diagram for production of
chrome green
General process diagram for production of
zinc yellow
General process diagram for production of
chrome pigment complexes
General waste water treatment process flow
Page
483
503
504
506
507
509
510
512
516
diagram at Plant #002 showing the sampling
points. (Chrome pigment manufacture)
16-9 General waste water treatment process flow
diagram at Plant #894 showing the sampling
points. (Chrome pigment manufacture)
16-10 Level 1 waste water treatment for chrome
pigments
16-11 Level 2 waste water treatment for chrome
pigments
17-1 General process flow diagram for production
o£ hydrogen cyanide by the Andrussow Process
17-2 General waste water treatment process flow
diagram at Plant #765 showing the sampling
points. (Hydrogen cyanide manufacture)
17-3 General waste water treatment process flow
diagram at Plant #782 showing sampling points
(Hydrogen cyanide manufacture)
518
529
531
555
560
564
XXIV
-------
LIST OF FIGURES - Continued
Page
17-4 Level 1 waste water treatment for hydrogen 572
cyanide subcategory
17-5 Level 2 waste water treatment for hydrogen 574
cyanide subcategory
18-1 General process diagram for production of 597
sodium dichromate
18-2 General waste water treatment process flow 600
diagram at Plant #493 showing the sampling
points. (Sodium dichromate manufacture)
18-3 General waste water treatment process flow 603
diagram at Plant #376 showing the sampling
points. (Sodium dichromate manufacture)
18-4 Level 1 waste water treatment for sodium 612
dichromate subcategory
18-5 Level 2 waste water treatment for sodium 613
dichromate subcategory
21-1 General process flow diagram of the 635
manufacture of copper sulfate
21-2 General process flow diagram at Plant #034 640
showing the sampling points. (Copper sulfate
manufacture)
21-3 Level 1 waste water treatment for copper 648
sulfate subcategory - batch process
21-4 Level 2 waste water treatment for copper 650
sulfate subcategory - batch process
22-1 General process flow diagram for nickel sulfate 677
manufacture
22-2 General waste water treatment process flow 680
diagram showing sampling points at Plant #369.
(Nickel sulfate subcategory.)
22-3 General process flow diagram at Plant #572 682
showing the sampling points. (Nickel sulfate
manufacture.)
xxv
-------
LIST OF FIGURES - Continued
Page
22-4 General waste water treatment process flow 683
diagram at Plant #120 showing the sampling
points. (Nickel sulfate manufacture)
22-5 Level 1 waste water treatment for nickel 690
sulfate subcategory - high production model -
batch process
22-6 Level 1 waste water treatment for nickel 691
sulfate subcategory - low and medium production
models - batch process
22-7 Level 2 waste water treatment for nickel 692
sulfate subcategory - high production model -
batch process
22-8 'Level 2 waste water treatment for nickel 693
sulfate - low and medium production models -
batch process
24-1 General process flow diagram at plant #282 718
showing the sampling points. Sodium bisulfite
manufacture
24-2 General flow diagram at Plant #586 showing 719
the sampling points. Sodium bisulfite
manufacture
24-3 General process flow diagram at Plant #987 722
showing the sampling points. Sodium bisulfite
manufacture
24-4 Level 1 waste water treatment for sodium 730
bisulfite subcategory - batch process
24-5 Level 2 waste water treatment for sodium 731
bisulfite subcategory - batch process
24-6 Level 3 waste water treatment for sodium 732
bisulfite subcategory - batch process
25-1 General process flow diagram at Plant #672. 753
(Sodium hydrosulfite manufacture)
25-2 General process flow diagram at Plant #672 756
showing the sampling points. (Sodium
hydrosulfite manufacture)
xxvi
-------
LIST OF FIGURES - Continued
25-3 Level 1 waste water treatment for sodium
hydrosulfite subcategory
25-4 Level 2 waste water treatment for sodium
hydrosulfite subcategory
Page
765
766
XXVI1
-------
LIST OF TABLES
2-1
2-2
2-3
2-4
2-5
3-1
3-2
5-1
5-2
5-3
6-1
7-1
7-2
8-1
Summary of Regulations - Best
Practicable Control Technology Currently
Available (BPT)
Summary of Regulations - Best
Available Technology (BAT)
Summary of Regulations -Pretreatment
Standards for Existing Sources (PSES)
Summary of Regulations - New
Source Performance Standards (NSPS)
Summary of Regulations -Pretreatment
Standards for New Sources (PSNS)
Recommended List of Toxic Pollutants
Scope of Industry Coverage within the
Inorganic Chemicals Manufacturing Point
Source Category
Analytical Detection Limits for Metals
Pollutant Frequency Based on Sampling
Program Results Including Raw Waste
Distribution of Pollutants According
to Subcategory
308-Questionnaire Response Data
Solubility Products of Toxic Metals
Comparison of Reverse Osmosis Concepts
Waste Water Treatment Options and
Performance Data Summary - Antimony and
Arsenic Removal
Page
6
8
10
12
15
23
24
51
55
56
61
69
86
92
XXVlll
-------
8-2
8-3
8-4
8-5
8-6
8-7
8-8
8-9
8-10
8-11
8-12
8-13
8-14
9-1
11-1
LIST OF TABLES - Continued
Page
Waste Water Treatment Options and 93
Performance Data Summary - Beryllium and Cadmium
Removal
Waste Water Treatment Options and 94
Performance Data Summary - Copper Removal
Waste Water Treatment Options and 95
Performance Data Summary - Chromium III and
Chromium VI Removal
Waste Water Treatment Options and 96
Performance Data Summary - Lead Removal
Waste Water Treatment Options and 97
Performance Data Summary - Mercury II Removal
Waste Wat,er Treatment Options and 98
Performance Data Summary - Nickel Removal
Waste Water Treatment Options and 98
Performance Data Summary - Silver Removal
Waste Water Treatment Options and 99
Performance Data Summary - Selenium and Thallium
Removal
Waste Water Treatment Options and 100
Performance Data Summary - Zinc Removal
Estimated Achievable Maximum 30-Day Averages 102
for the Applied Technologies
Industrial Waste Water Treatment System 105
Performance — Summary of Effluent
Concentration Data on Toxic Metals
Estimated Achievable Long-Term Average 108
Concentrations for Toxic Metals with BPT or BAT
Treatment Options
Theoretical Solubilities of Toxic Metal 110
Hydroxides/Oxides at Various pH Values
Prioritization of Toxic Pollutants Found in 126
Each Subcategory
Subcategory Profile Data Summary 147
xxxx
-------
LIST OF TABLES - Continued
Page
11-2 Status of Regulations - Effluent Limitation 148
Guidelines
11-3 Summary of Waste Water Flow Data for 154
Chlorine Mercury Cell Plants
11-4 Pollutant Concentrations and Loads at Plant 156
#299
11-5 Pollutant Concentrations and Loads at 157
Verification Plants
11-6 Average Toxic Pollutant Raw Waste 167
Concentrations and Loads at Verification
Plants
11-7 Summary of Raw Waste Loadings at 168
Verification Plants
11-8 Model Plant Treatment Costs 178
11-9 Model Plant Treatment Costs 179
11-10 Model Plant Treatment Costs 180
11-11 Model Plant Unit Treatment Costs 181
11-12 Mercury Discharges from Selected Chlor- 182
Alkali Mercury Cell Plants
11-13 Residual Chlorine Discharges at Selected 184
Chlor-Alkali Plants
11-14 Comparison of Raw Waste Concentrations 185
of Toxic Pollutants with Treatability
11-15 Effluent Limitations, BAT 187
11-16 Effluent Concentrations of Toxic Pollutants 189
from Verification Sampling
11-17 Subcategory Profile Data Summary 194
11-18 Status of Regulations - Effluent Limitation 195
Guidelines
11-19 Waste Water Flows at Diaphragm Cell Chlorine 201
Plants
XXX
-------
LIST OF TABLES - Continued
11-20 Pollutant Concentrations and Loads at
Screening and Verification Plants
11-21 Results of Asbestos Sampling at Diaphragm
Cell Plants
11-22 Maximum Raw Waste Concentrations of Toxic
Metals Observed at Diaphragm Cell Chlorine
Plants (mg/1)
11-23 Toxic Metal Concentrations and Loads at
Screening and Verification Plants
11-24 Summary of Raw Waste Loadings at Screening
and Verification Metal Anode Plants
11-25 Toxic Metal Concentrations and Loads in
Cell Room Waste Waters at Screening and
Verification Plants
11-26 Raw Waste Toxic Metals Concentration and
Loads in Process Streams Other Than Cell
Room Wastes from Screening and
Verification Plants
11-27 Raw Waste Toxic Organics at a Graphite
Anode Plant
11-28 Raw Waste Toxic Organics by Waste Water
Source at a Graphite Anode Plant
11-29 Model Plant Treatment Costs
11-30 Model Plant Treatment Costs
11-31 Model Plant Treatment Costs
11-32 Model Plant Unit Treatment Costs
11-33 Summary of Unit Flows at Diaphragm Cell
Plants
11-34 Comparison of Raw Waste Concentrations
of Toxic Pollutants with Treatability
11-35 Effluent Limitations, BPCTCA
Page
204
213
214
216
217
219
220
221
222
235
236
237
238
240
241
243
XXXI
-------
LIST OF TABLES - Continued
11-36 Lead and TSS Discharges from Selected
Diaphragm Cell Chlorine Plants
11-37 Comparison of Toxic Pollutants in Diaphragm
Cell Plant Effluents with Treatability
11-38 Effluent Limitations, BAT
11-39 Comparison of Haw Waste Characteristics
at a New Metal Anode Plant with Treatability
of Toxic Metals
11-40 Effluent Limitations, NSPS
12-1 Subcategory Profile Data Summary
12-2 Status of Regulations - Effluent Limitation
Guidelines
12-3 Water Usage in the Hydrofluoric Acid
Subcategory
12-4 Waste Water Flow and Reuse Data for the
Hydrofluoric Acid Subcategory
12-5 Waste Flow from Hydrofluoric Acid
Manufacturing plants
12-6 Flow and Pollution Concentration Data of
Spar Drying and Total Process Waste Water
for Plants #251 and #837 Producing
Hydrofluoric Acid
12-7 Solid Waste Generated at the Hydrofluoric
Acid Plants Sampled
12-8 Gypsum Solids production in the Hydrofluoric
Acid Subcategory
12-9 Flow and Pollutant Concentration Data of
the Sampled Waste Streams of Plant #705
Producing Hydrofluoric Acid
12-10 Flow and Pollutant Concentration Data of
the Sampled Waste Streams for Plants #705,
#251, and #167 Producing Hydrofluoric Acid
12-11 Toxic Pollutant Raw Waste Data
246
247
252
257
259
262
263
268
269
273
274
274
275
277
279
283
XXXll
-------
LIST OF TABLES - Continued
Page
12-12 Summary of Raw Waste Loadings Pound in 284
Screening and Verification Sampling
12-13 Model Plant Treatment Costs 297
12-14 Model Plant Treatment Costs 298
12-15 Model Plant Treatment Costs 299
12-16 Model Plant Unit Treatment Costs 300
12-17 Summary of Waste Water Control and Treatment 302
Technology Employed at Hydrofluoric
Acid Plants
12-18 Summary of Long-Term Monitoring Data from 304
Four Hydrofluoric Acid Plants
12-19 Toxic Pollutant Treated Effluent Data 305
12-20 Development of TSS and Fluoride Limitations 311
12-21 Effluent Limitations, BPCTCA 312
12-22 Summary of Concentrations and Filter 319
Removal Efficiencies of Pollutants Selected for
Study in the Hydrofluoric Acid Subcategory
12-23 Effluent Limitations, BAT 321
12-24 Effluent Limitations, NSPS 325
13-1 Subcategory Profile Data Summary 328
14-1 Subcategory Profile Data Summary 332
(Chloride Process)
14-2 Status of Regulations - Effluent Limitation 333
Guidelines
14-3 Water Usage in Titanium Dioxide-Chloride 337
Process/High Grade Ores Subcategory
14-4 Waste Water Flow for Titanium Dioxide- 339
Chloride Process Subcategpry
XXXlll
-------
LIST OF TABLES - Continued
Page
14-5 Flow and Pollutant Concentration Data of 342
the Sampled Waste Streams of Plant #559
Producing Titanium Dioxide by Chloride-Rutile
Process
14-6 Flow and Pollutant Concentration Data of 344
the Sampled Waste Streams for Plant #172
Producing Titanium Dioxide (Chloride Process)
14-7 Raw Waste Pollutant Data Summary of the 346
Sampled Streams
14-8 Model Plant Treatment Costs 355
14-9 Model Plant Treatment Costs 356
14-10 Model Plant Treatment Costs 357
14-11 Model Plant Treatment Costs 358
14-12 Model Plant Treatment Costs 359
14-13 Model Plant Treatment Costs 360
14-14 Model Plant Unit Treatment Costs 362
14-15 Historical Effluent Monitoring Data Summary 363
with Variability Factor — Daily Measurements,
Plant #559
14-16 Historical Effluent Monitoring Data Summary 364
with Variability Factors — Daily Measurements,
Plant #172
14-17 Treatment Performance Data of Sampled Plants 365
#599 and #172
14-18 Effluent Limitations, BPCTCA 371
14-19 Effluent Limitations, BAT 373
14-20 Effluent Limitations, NSPS 378
14-21 Subcategory Profile Data Summary 379
(Sulfate Process)
14-22 Analysis of Ilmenite Ores 380
XXXIV
-------
LIST OF TABLES - Continued
Page
14-23 Water Usage in Titanium Dioxide - Sulfate 383
Process Subcategory
14-24 Raw Waste Characteristics (Industry Data) 386
for Plant #555 (Production of TiO, by Sulfate
Process)
14-25 Flows and Pollutant Concentrations for 389
the Waste Streams Sampled for Plant #559
Producing Titanium Dioxide
14-26 Process Waste Water Flow at Plants #555, 391
#694 and #559 Titanium Dioxide
(Sulfate Process)
14-27 Summary of Raw Waste Loadings Found in 393
Screening and Verification Sampling
14-28 Toxic Pollutants: Average Raw Waste Loads 394
and Concentrations
14-29 Model Plant Treatment Costs 405
14-30 Model Plant Treatment Costs 406
14-31 Model Plant Treatment Costs 407
14-32 Model Plant Unit Treatment Costs 408
14-33a Historical Effluent Monitoring Data Summary— 410
Treatment Without Iron Removal
14-33b Historical Effluent Monitoring Data Summary— 411
Treatment With Iron Removal
14-34 Verification Results from - Sulfate Process 412
Titanium Dioxide Plant #559
14-35 Effluent Limitations, BPCTCA 415
14-36 Effluent Limitations, BAT 418
14-37 Effluent Limitations, NSPS 420
14-38 Subcategory Profile Data Summary 425
(Chloride-llmenite)
14-39 Average Water Usage for TiO- Production 428
by the Chloride - Ilmenite Process
xxxv
-------
LIST OF TABLES - Continued
14-40 Average Raw Waste Loads for TiO- Production
by the Chloride - Ilmenite Process
14-41 Summary of Raw Waste Loadings Found in
Screening and Verification Sampling
14-42 Toxic Pollutant Average Raw Waste Loads
and Concentrations
14-43 Model Plant Treatment Costs
14-44 Model Plant Treatment Costs
14-45 Model Plant Treatment Costs
14-46 Model Plant Treatment Costs
14-47 Model Plant Treatment Costs
14-48 Model Plant Treatment Costs
14-49 Model Plant Treatment Costs
14-50 Model Plant Treatment Costs
14-51 Model Plant Unit Treatment Costs
14-52 Effluent Limitations, BPCTCA
14-53 Effluent Limitations, NSPS
15-1 Subcategory Profile Data Summary
15-2 Status of Regulations - Effluent Limitation
Guidelines
15-3 Water Usage in the Aluminum Fluoride
Subcategory
15-4 Waste Water Flow at Plants #837, #705 and
#251 for Aluminum Fluoride Subcategory
15-5 Solids Generated at Plant #705 and #251
Producing Aluminum Fluoride
15-6 Flow and Pollutant Concentration Data of
the Sampled Waste Streams for Plant #705
Producing Aluminum Fluoride
Page
429
433
434
443
444
445
446
447
448
449
450
451
454
458
462
464
464
466
466
469
XXXVI
-------
LIST OF TABLES - Continued
Page
15-7 Flow and Pollutant Concentration Data of 471
the Sampled Streams for Plant #251 Producing
Aluminum Fluoride
15-8 Toxic Pollutant Average Raw Waste Loads 473
and Concentrations
15-9 Toxic Pollutant Effluent Concentrations 474
During Sampling
15-10 Summary of Raw Waste Loadings Found in 475
Screening and Verification Sampling
15-11 Model Plant Treatment Costs 486
15-12 Model Plant Treatment Costs 487
15-13 Model Plant Treatment Costs 488
15-14 Model Plant Unit Treatment Costs 489
15-15 Effluent Limitations, BPCTCA 497
16-1 Subcategory Profile Data Summary 500
16-2 Status of Regulations - Effluent Limitation 501
Guidelines
16-3 Water Usage in the Chrome Pigments Subcategory 511
16-4 Summary of Waste Water Flow 514
16-5 Flow, Pollutant Concentration and Load 517
Data of the Sampled Waste Streams for
Plant #002
16-6 Flow, Pollutant Concentration and Load 519
Data for the Sampled Streams at Plant #894
16-7 Toxic Pollutant Raw Waste Data 522
16-8 Summary of Raw Waste Loadings Found in 523
Screening and Verification Sampling
16-9 Toxic Pollutant Treated Waste Data 524
16-10 Model Plant Treatment Costs 534
16-11 Model Plant Treatment Costs 535
xxxvn
-------
LIST OF TABLES - Continued
Page
16-12 Model Plant Treatment Costs 536
16-13 Model Plant Treatment Costs 537
16-14 Model Plant Unit Treatment Costs 539
16-15 Summary of Long Term and Verification 541
Effluent Sampling Results at Plant #894
16-16 Effluent Limitations, BPCTCA 545
17-1 Subcategory Profile Data Summary 552
17-2 Status of Regulations - Effluent Limitation 553
Guidelines
17-3 Water Usage in Hydrogen Cyanide - Andrussow 556
Process Subcategory
17-4 Waste Plow Data for HCN Production by the 558
Andrussow Process
17-5 Flow and Pollutant Data of the Raw and 561
Treated Waste Streams of Plant #765 Producing
Hydrogen Cyanide by Andrussow Process
17-6 Flow and Pollutant Concentration Data of 563
the Sampled Waste Streams for Plant #765
Producing Hydrogen Cyanide
17-7 Flow and Pollutant Concentration Data of 565
the Sampled Waste Streams for Plant #782
Producing Hydrogen Cyanide
17-8 Unit Flow and Unit Pollutant Loading for 566
Raw and Treated Waste Effluents at Plant 1782
17-9 Summary of Pollutant Raw Waste Loading 569
Found in Screening and Verification Sampling
17-10 Model Plant Treatment Costs 577
17-11 Model Plant Treatment Costs 578
17-12 Model Plant Treatment Costs 579
17-13 Model Plant Unit Treatment Costs 580
XXXVlll
-------
LIST OF TABLES - Continued
Page
17-14 Statistical Analysis of Historical Effluent 582
Monitoring Data on Free Cyanide and Ammonia
from Plant #765
17-15 Statistical Analysis of the 28-Day Effluent 585
Sampling Results on Total Cyanide from
Plant #765
17-16 Effluent Limitations, BPCTCA 587
17-17 Effluent Limitations, BAT 589
17-18 Effluent Limitations, NSPS 591
18-1 Subcategory Profile Data Summary 594
18-2 Status of Regulations - Effluent Limitation 595
Guidelines
18-3 Water Usage in Sodium Dichromate Subcategory 598
18-4 Flow and Pollutant Concentration Data of 601
the Sampled Waste Streams for Plant #493
Producing Sodium Dichromate
18-5 Flow and Pollutant Loading Data of the 604
Sampled Waste Streams for Plant #376
Producing Sodium Dichromate
18-6 Flow and Pollutant Loading Data of the 605
Sampled Waste Streams for Plant #398
Producing Sodium Dichromate
18-7 Toxic Pollutant Raw Waste Data 607
18-8 Summary of Raw Waste Loadings Found in 608
Screening and Verification Sampling
18-9 Model Plant Treatment Costs 616
18-10 Model Plant Treatment Costs 617
18-11 Model Plant Treatment Costs 618
18-12 Model Plant Unit Treatment Costs 619
18-13 Effluent Sampling Data from Sodium 622
Dichromate Plants
18-14 Effluent Limitations, BPCTCA 624
XXXIX
-------
LIST OF TABLES - Continued
Page
19-1 Subcategory Profile Data Summary 628
20-1 Subcategory Profile Data Summary 630
21-1 Subcategory Profile Data Summary 632
21-2 Status of Regulations - Effluent Limitation 633
Guidelines
21-3 Water Usage in Copper Sulfate Subcategory 636
21-4 Waste Water Flow for the Copper Sulfate 638
Subcategory
21-5 Flow and Pollutant Concentration Data of 641
the Sampled Waste Streams for Plant #034
Producing Copper Sulfate
21-6 Raw Waste Data 644
21-7 Model Plant Treatment Costs 653
21-8 Model Plant Unit Treatment Costs 654
21-9 Summary of Long-Term Monitoring Data from 655
Plant #034
21-10 Treated Effluent Data 657
21-11 Average Pollutant Levels and Removal 658
Efficiency for Plant #034
21-12 Effluent Limitations, BPCTCA 662
21-13 Effluent Limitations, BAT 665
21-14 Guidance for Effluent Control 668
22-1 Subcategory Profile Data Summary 674
22-2 Status of Regulations - Effluent Limitation 675
Guidelines
22-3 Water Use in the Nickel Sulfate Subcategory 678
22-4 Flow and Pollutant Concentration Data of 681
the Sampled Waste Streams for Plants Producing
Nickel Sulfate
22-5 Toxic Pollutant Raw Waste Data 686
xl
-------
LIST OF TABLES - Continued
Page
22-6 Model Plant Treatment Costs 696
22-7 Model Plant Treatment Costs 697
22-8 Model Plant Treatment Costs 698
22-9 Model Plant Unit Treatment Costs 699
22-10 Toxic Pollutant Treated Effluent Data 701
22-11 Effluent Limitations, BAT 705
23-1 Subcategory Profile Data Summary 710
24-1 Subcategory Profile Data Summary 712
24-2 Status of Regulations - Effluent Limitation 713
Guidelines
24-3 Water Usage in the Sodium Bisulfite 715
Subcategory
24-4 Waste Water Flow at Plants #987 and #282 716
for Sodium Bisulfite Subcategory
24-5 Flow and Pollutant Load Data of the Sampled 717
Waste Streams for Plant #282 Producing Sodium
Bisulfite
24-6 Flow and Pollutant Load Data of the Sampled 720
Waste Streams for Plant #586
24-7 Flow and Pollutant Load Data of the Sampled 721
Waste Streams for Plant #987
24-8 Toxic Pollutant Raw Waste Loads 725
24-9 Summary of Raw Waste Loadings Found in 726
Screening and Verification Sampling
24-10 Toxic Pollutant Concentrations Observed 727
in Treated Effluent During Verification
Sampling
24-11 Model Plant Treatment Costs 735
24-12 Model Plant Treatment Costs 736
24-13 Model Plant Treatment Costs 737
24-14 Model Plant Unit Treatment Costs 738
xli
-------
LIST OF TABLES - Continued
Page
24-15 Comparison of Maximum Raw Waste 739
Concentrations with Treatability
24-16 Effluent Limitations, BPCTCA 742
25-1 Subcategory Profile Data Summary 750
25-2 Status of Regulations - Effluent Limitation 751
Guidelines
25-3 Waste Source Data at Plant #672 755
25-4 Flow, Pollutant Concentration, and Load 758
Data of the Sampled Waste Streams for
Plant 1672 Producing Sodium Hydrosulfite
25-5 Sampling Results and Treatment System 760
Performance for Toxic Pollutants — Plant #672
25-6 Summary of Raw Waste Loadings and 761
Concentrations Found at a Sodium Hydrosulfite
Plant (Formate Process)
25-7 Model Plant Treatment Costs 768
25-8 Model Plant Unit Treatment Costs 769
25-9 Subcategory Performance Evaluation Summary 772
at Plant #672 for Conventional and
Nonconventional Pollutants in the Effluents
25-10 Guidance for Effluent Control - Level 1 776
25-11 Guidance for Effluent Control - Level 2 779
26.2-1 Subcategory Profile Data Summary 783
26.3-1 Subcategory Profile Data Summary 785
26.4-1 Subcategory Profile Data Summary 787
26.6-1 Subcategory Profile Data Summary 790
26.9-1 Subcategory Profile Data Summary 794
26.17-1 Subcategory Profile Data Summary 801
xlii
-------
LIST OF TABLES - Continued
26.23-1
26.24-1
26.27-1
26.32-1
26.35-1
26.36-1
26.37-1
26.38-1
26.40-1
26.42-1
26.43-1
26.44-1
Subcategory
Subcategory
Subcategory
Subcategory
Subcategory
Subcategory
Subcategory
Subcategory
Subcategory
Subcategory
Subcategory
Subcategory
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Profile
Data Summary
Data Summary
Data Summary
Data Summary
Data Summary
Data Summary
Data Summary
Data Summary
Data Summary
Data Summary
Data Summary
Data Summary
Page
807
809
814
819
823
825
827
829
831
833
836
837
xliii
-------
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.
Chester Kaminski, Mr. Mahendra L. Shah, Mr. Dale Newkirk, Mr.
Carl B. Johnston, Mr. Dennis Merklin, Mr. William E. Rorie,
Dr. David Ben Hur, Dr. Ashley Levy, Mr. Milford Walker, Jr.,
Ms. Garnett Gray, 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.
Mr. Steve Schatzow, Director, Office of Water Regulations
S tandards, is gratefully acknowledged for his
:ibutions to the proiect.
and S tandards, is grat
contributions to the project.
Ms. Susan Lepow, Mr. Joseph Freedman, 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. Debra Maness and Mr. Richard Cotz of the Office of
Analysis and Evaluation, Ms, Alexandra Tarnay, Monitoring and
Data Support Division, and Mr. Pred Talcott, Office of
Planning and Evaluation are acknowledged for their assistance.
xliv
-------
SECTION 1
CONCLUSIONS AND SUMMARY
Toxic Pollutants
The following 35 inorganic chemical product subcategories were
screened for the purpose of establishing wastewater effluent
limitations guidelines for existing sources, standards of performance
for new sources, and pretreatment standards for new and existing
sources in this study:
1. Chlor-Alkali
2. Hydrofluoric Acid
3. Titanium Dioxide
4. Aluminum Fluoride
5. Chrome Pigments
6. Hydrogen Cyanide
7. Sodium Bichromate
8. Copper Sulfate
9. Nickel Sulfate
10. Sodium Bisulfite
11. Sodium Hydrosulfite
12. Hydrogen Peroxide
13. Hydrochloric Acid
14. Nitric Acid
15. Sodium Carbonate
16. Sodium Metal
17. Sodium Silicate
18. Sulfuric Acid
19. Carbon Dioxide
20. Carbon Monoxide and
by-product Hydrogen
21. Silver Nitrate
22. Ammonium Chloride
23. Ammonium Hydroxide
24. Barium Carbonate
25. Boric Acid
26. Calcium Carbonate
27. Cuprous Oxide
28. Manganese Sulfate
29. Strong Nitric Acid
30. Oxygen and Nitrogen
31. Potassium Iodide
32. Sodium Hydrosulfide
33. Sodium Silicofluoride
34. Sodium Thiosulfate
35. Sulfur Dioxide
The screening studies showed that only the plant process wastewaters
from subcategories 1 through 11 contain treatable amounts of 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 were present at low level concentrations.
The screening results which indicated the presence of toxic pollutants
in significant amounts were largely confirmed by the results of the
verification program. Verification.sampling accounted for 50 to 75
percent of the current inorganic chemical production rate in the
subcategories covered.
The sources of most of the toxic pollutants found in the raw wastes
and treated effluents were 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.
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 toxic 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.
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.
Subcategorization
A review of the product/process basis for Subcategorization of the
inorganic chemical product subcategories designated for study revealed
that certain modifications may be appropriate in the interest of
developing effective regulations. The toxic pollutant problem per se
impacts sub-categorization directly only in the Chlor-Alkali Industry
where the use of graphite anodes contributes to the generation of
chlorinated hydrocarbons. In the Titanium Dioxide Industry, major
process and raw material differences justify the creation of a
separate segment for the sulfate process, the chloride process, and
for the chloride process using ilmenite ore. Consideration was given
to creating a subcategory for the combined production of hydrofluoric
acid and aluminum fluoride in view of their similiar 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 Organic Chemicals Manufacturing Category. The hydrogen cyanide
subcategory includes only manufacture by the Andrussow process.
-------
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 versus Train, 541 F. 2d 1018 (4th,
Cir. 1976) reversed in part, 430 U.S. 112 (1977}. The factors
affecting the control and treatment of pollutant discharges in those
industries have been studied in response to the remanded issues. It
has been concluded that alternative control and treatment technologies
to those originally considered for BAT and NSPS may be appropriate.
-------
-------
SECTION 2
RECOMMENDATIONS
On the basis of the toxic pollutant screening and verification results
and the evaluation of applicable technologies for discharge control
and treatment, it is recommended that effluent limitation quidelines,
new source performance standards and pretreatment standards for new
and existing sources be promulgated for the following 10 inorganic
chemicals manufacturing subcategories:
Chlor-Alkali
Hydrofluoric Acid
Titanium Dioxide
Aluminum Fluoride
Chrome Pigments
Hydrogen Cyanide
Sodium Dichromate
Copper Sulfate
Nickel Sulfate
Sodium Bisulfite
Table 2-1 summarizes the regulations for Best Practicable Control
Technology Currently Available (BPT). Summaries of regulations for
Best Available Technology (BAT), Pretreatment Standards, and New
Source Performance Standards are given in Tables 2-2, 2-3, 2-4, and
2-5. These tables indicate that Chlor-Alkali has been divided into
two segments and Titanium Dioxide in three segments before listing the
numerical effluent limitations.
The Agency proposed BPT, BCT, and BAT limitations and NSPS, PSES, and
PSNS for the Sodium Hydrosulfite (Formate Process) subcategory. The
proposed regulation basically'1"1 added control of selected toxic metal
pollutants to existing treatment practiced in the industry. The
Agency reviewed the basis for the proposed regulation and concluded
that the total current treated discharge load of only 0.42 pounds per
day total toxic metals from all plants in the subcategory is too
insignificant to justify developing a national regulation.
Accordingly, this subcategory has been excluded from national
'regulation development under Paragraph 8(a)(iv) of the Settlement
Agreement.
-------
TABLE 2-1. SUMMARY OF REGULATIONS -
BEST PRACTICABLE CONTROL TECHNOLOGY
CURRENTLY AVAILABLE (BPT)
Subcategory
Parameter
Effluent Limitations
kg/kkg (or lh/1000 Ib) of product
Chlor-alkali,
Mercury Cells
Chlor-alkali,
Diaphragm Cells
Hydrofluoric
Acid
Sodium
TSS
Mercury
pH
TSS
Copper (T)
Lead (T)
Nickel (T)
pH
TSS
Fluoride (T)
Nickel (T)
Zinc (T)
pH
TSS
Dichromate Hexavalent Chromium
Titanium
Dioxide
(sulfate
process)
Titaniun
Dioxide
(chloride
process)
Titanium
Dioxide (chlor-
ide ilmenite
process)
Chromium (T)
Nickel {T)
pH
TSS
Chromium (T)
Nickel (T)
&
TSS
Chromium (T)
pH
TSS
Chromium (T)
Nickel (T)
pH
0.32
0.00014
0.51
0.0070
0.010
0.0056
5.3
2.9
0.011
0.036
0.22
0.00050
0.0044
0.0034
38
0.21
0.14
6.4
0.030
9.6
0.053
0.035
0.64
0.00028
1.1
0.018
0.026
0.014
11.0
6.1
0.036
0.12
0.44
0.00090
0.0088
0.0068
140
0.48
0.29
23
0.057
35
0.12
0.072
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
6.0 to 9.0
6.0 to 9.0
(continued)
-------
Table 2-1. Continued
Effluent Limitations
SuDcategory
Aluminum
Fluoride
parameter
TSS
Fluoride (T)
Chromiun (T)
Nickel (T)
pH
Max
30-day Avg
kg/kkg (or lb/1000 Ib)
1.2
0.63
0.0045
0.0024
24-hr
Max
of product
2.4
1.3
0.015
0.0079
pH Range
6.0 to 9.0
Copper Sulfate
TSS
Copper (T)
Nickel (T)
Selenium (T)
0.023
0.0010
0.0020
0.00050
0.069
0.0030
0.0060
0.0015
6.0 to 9.0
Hydrogen
Cyanide
Nickel Sulfate
Chrome
Pigments
Sodium
Bisulfite
TSS
Cyanide A
Cyanide (T)
TSS
Nickel (T)
pH
TSS
Chromiun (T)
Lead (T)
Zinc (T)
TSS
COD
Chromium (T)
Zinc (T)
3.2
0.021
0.23
0.032
0.0020
3.8
0.13
0.15
0.13
0.080
0.95
0.00063
0.0015
8.6
0.10
0.65
0.096
0.0060
9.1
0.31
0.36
0.31
0.32
3.8
0.0020
0.0051
6.0 to 10.5
6.0 to 9,0
6.0 to 9.0
6.0 to 9.0
-------
TABLE 2-2. SUMMARY OF REGULATIONS -
BEST AVAILABLE TECHNOLOGY (BAT)
Effluent Limitations
Subcategory
Parameter
Max
30-day Avg
24-hr
Max
kg/kkg (or lb/1000 ib) of product
Ghlor-alkali
Mercury Cells
Mercury (T)
Total Residual
Chlorine
0.00010
0.0019
0.00023
0.0032
Chlor-alkali Copper (T) 0.0049
Diaphragm Cells Lead (T) 0.0024
Nickel (T) 0.0037
Total Residual
Chlorine 0.0079
0.012
0.0059
0.0097
0.013
Hydrofluoric
Acid
Fluoride (T)
Nickel (T)
Zinc (T)
Sodium Chromium (T)
Dichromate Hexavalent Chromium
Nickel (T)
Titanium
Dioxide
(sulfate
process)
Titanium
Dioxide
(chloride
process)
Titanium
Dioxide
(chloride-
ilmenite
process)
Aluminum
Fluoride
Chromium (T)
Nickel (T)
Chromium (T)
Chromium (T)
Nickel (T)
Fluoride (T)
Chromium (T)
Nickel (T)
1.6
0.0060
0.022
0.0044
0.00050
0.0034
0.21
0.14
0.030
0.053
0.035
0.63
0.0045
0.0024
3.4
0.020
0.072
0.0088
0.00090
0.0068
0.48
0.29
0.057
0.12
0.072
1.3
0.013
0.0079
(continued)
-------
Table 2-2. Continued
Effluent Limitations
Subcategory
Parameter
Max
30-day Avg
24-hr
Max
kg/kkg (or lb/1000 UD) of product
Chrome
Pigments
Chromium
Lead (T)
Zinc (T)
0.13
0.15
0.13
0.31
0.36
0.31
Copper Sulfate
Copper {T)
Nickel (T)
Selenium (T)
0.0010
0.0020
0.00050
0.0030
0.0060
0.0015
Hydrogen
Cyanide
Nickel Sulfate
Sodium
Bisulfite
Cyanide A
Cyanide (T)
Total Residual
Chlorine
Copper (T)
Nickel (T)
COD
Chromium (T)
Zinc (T)
0.021
0.23
0.051
0.00024
0.00024
0.95
0.00063
0.0015
0.10
0.65
0.086
0.00074
0.00074
3.8
0.0020
0.0051
-------
TABLE 2-3. SUMMARY OF REGULATIONS -
PRETREATMENT STANDARDS FOR EXISTING
SOURCES (PSES)
Subcategory
Oiler-alkali
Parameter
Mercury (T)
Effluent Limitations
Max 30-day 24-hr
Avg Max
(rog/1) or (kg/kkg) (mg/1) or (kg/kkg)
0.048
0.00010
0.11
0.00023
Mercury Cells
Chlor-aUcali
Diaphragm
Cells
Hydrofluoric
Acid
Sodium
Diohromate
Titanium
Dioxide
(sulfate
process)
Titanium
Dioxide
(chloride
process)
Titanium
Dioxide
(chloride-
ilmenite
process)
Chrone
Pigments
Copper (T)
Lead (T)
Nickel (T)
0.80
1.1
0.64
Fluoride (T) 50
Nickel (T)
Zinc (T)
Chromium (T)
Hexavalent Chromium
Nickel (T)
Chromium (T)
Nickel (T)
Chromium (T)
Chromium (T)
Nickel (T)
Chronium (T)
Lead (T)
Zinc
0.20
0.66
0.50
0.060
0.40
0.44
0.29
0.30
0.44
0.29
1.2
1.4
1.2
0.0070
0.010
0.0056
1.6
0.0060
0.022
0.0044
0.00050
0.0034
0.21
0.14
0.030
0.053
0.035
0.13
0.15
0.13
2.1
2.9
1.6
100
0.66
2.2
1.0
0.11
0.80
1.0
0.60
0.57
1.0
0.60
2.9
3.4
2.9
0.018
0.026
0.014
3.4
0.020
0.072
0.0088
0.00090
0.0068
0.48
0.29
0.057
0.12
0.072
0.31
0.36
0.31
(continued)
10
-------
TABLE 2-3. Continued
Effluent Limitations
Subcategory
Parameter
Max 30-day
Avg
(mg/1) or (kg/kkg)
24-hr
Max
(mg/1) or (kg/kkg)
Copper
Sulfate
Hydrogen
Cyanide
Copper (T)
Nickel (T)
Selenium (T)
Cyanide A
Cyanide (T)
1.1
2.1
0.53
0.36
4.0
0.0010
0.0020
0.00050
0.021
0.23
3.2
6,4
1.6
1.7
11
0.0030
0.0060
0.0015
0.10
0.65
Nickel
Sulfate
Sodium
Bisulfate
Copper (T)
Nickel (T)
0.36
0.36
COD 630
Chromium (T) 0.42
Zinc (T) 1.0
0.00024
0.00024
0.95
0.00063
0.0015
1.1
1.1
2500
1.3
3.4
0.00074
0.00074
3.8
0.0020
0.0051
11
-------
TABLE 2-4. SUMMARY OF REGULATIONS -
NEW SOURCE PERFORMANCE STANDARDS (NSPS)
Subcategory
Parameter
Effluent Limitations
Max
30-day Avg
24-hr
Max
kg/kkg (or lb/1000 Ib) of product
Range
Chlor-alkali
Mercury Cells
TSS 0.32
Mercury (T) 0.00010
Total Residual
Chlorine 0.0019
0.64
0.00023
0.0032
6.0 to 9.0
Chlor-alkali
Diaphragm Cells
Hydrofluoric
Acid
Sodium
Dichromate
Titanium
Dioxide
(sulfate
process)
TSS 0.32
Lead (T) 0.0018
Total Residual
Chlorine 0.0079
TSS
Fluoride (T)
Nickel (T)
Zinc (T)
pH
TSS
Chranium (T)
Hexavalent
Chromium
Nickel '(T)
pH
TSS
Iron (T)
Chromium (T)
Nickel (T)
pH
3.0
1.6
0.0060
0.022
0.22
0.0044
0.00050
0.0034
30
1.2
0.14
0.095
1.0
0.0047
0.013
6.0
3.4
0.020
0.072
0.44
0.0088
0.00090
0.0068
110
4.0
0.27
0.18
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
Titanium
Dioxide
(chloride
process)
TSS
Iron (T)
Chromium (T)
FH
4.0
0.16
0.012
14
0.52
0.023
6.0 to
9.0
(continued)
12
-------
TABLE 2-4. Continued
Subcategory
Titanium
Dioxide
(chloride-
ilmenite
process)
Aluminum
Fluoride
Chrome
Pigments
Copper Sulfate
Hydrogen
Cyanide
Parameter
TSS
Iron (T)
Chromium (T)
Nickel (T)
£H
TSS
Fluoride (T)
Chromium (T)
Nickel (T)
pH
TSS
Chromium (T)
Lead (T)
Zinc (T)
pH
TSS
Copper (T)
Nickel (T)
Selenium (T)
£H
TSS
Cyanide A
Cyanide (T)
Efflue
Max
30-day Avg
kg/kkg (or
2.4
0.096
0.0072
0.010
1.2
0.63
0.0045
0.0024
3.8
0.13
0.15
0.13
0.023
0.0010
0.0020
0.00050
3.2
0.021
0.23
nt Limitations
24-hr
Max
lb/1000 lb) of product
8.4
0.32
0.014
0.020
2.4
1.3
0.015
0.0079
9.1
0.31
0.36
0.31
0.069
0.0030
0.0060
0.0015
8.6
0.10
0.65
pH Fange
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
6.0 to 9.0
Ibtal Residual
Chlorine
0.051
0.086
6.0 to 10.5
Nickel Sulfate
TSS
Copper (T)
Nickel (T)
pH
0.032
0.00024
0.00024
0.096
0.00074
0.00074
6.0 to 9.0
(continued)
13
-------
TABLE 2-4. Continued
Subcategory Parameter
Effluent Limitations
Max
30-day Avg
24-hr
Max
kg/kkg (or lb/1000 Ib) of product
pH Pange
Sodium
Bisulfite
TSS
COD
Chromium (T)
Zinc (T)
0.080
0.95
0.00063
0.0015
0.32
3.8
0.0020
0.0051
6.0 to 9.0
14
-------
TABLE 2-5. SUMMARY OF REGULATIONS -
PRETREATMENT STANDARDS FOR NEW SOURCES (PSNS)
Subcategory
Parameter
Effluent Limitations
Max
30-day Avg
(mg/1) or (kg/kkg)
24-hr
Max
(mg/1) or (kg/kkg)
Chlor-alkali
Mercury Cells
Chlor-alkali
Diaphragm Cells
Mercury
Lead (T)
0.048
0.21
0.00010
0.0018
0.11
0.53
0.00023
0.0047
Hydrofluoric
Acid
Sodium
Dichromate
Titanium
Dioxide
(sulfate
process)
Titanium
Dioxide
(chloride
process)
Titanium
Dioxide
(chloride-
ilmenite
process)
Chrome
Pigments
Fluoride (T)
Nickel (T)
Zinc (T)
Chromium (T)
Hexavalent
Chroniun
Nickel (T)
Iron (T)
Chromium (T)
Nickel (T)
Iron (T)
Chromium (T)
Iron (T)
Chromium (T)
Nickel (T)
Chrcmium (T)
Lead (T)
Zinc (T)
50
0.20
0.66
0.50
0.060
0.40
2.5
0.30
0.20
1.6
0.12
1.6
0.12
0.17
1.2
1.4
1.2
1.6
0.0060
0.022
0.0044
0.00050
0.0034
1.2
0.14
0.095
0.16
0.012
0.096
0.0072
0.010
0.13
0.15
0.13
100
0.66
2.2
1.0
0.11
0.80
8.5
0.57
0.38
5.3
0.23
5.3
0.23
0.33
2.9
3.4
2.9
3.4
0.20
0.072
0.0088
0.00090
0.0068
4.0
0.27
0.18
0.52
0.023
0.32
0.014
0.020
0.31
0.36
0.31
(continued)
15
-------
TABLE 2-5. Continued
Effluent Limitations
Subcategory
Paramater
Max
30-day Avg
(mg/1) or (kg/kkg)
24-hr
Max
(mg/1) or (kg/kkg)
Copper
Sulfate
Hydrogen
Cyanide
Copper (T)
Nickel (T)
Selenium (T)
Cyanide A
Cyanide (T)
1.1
2.1
0.53
0.36
4.0
0.0010
0.0020
0.00050
0.021
0.23
3.2
6.4
1.6
1.7
11
0.0030
0.0060
0.0015
0.10
0.65
Nickel
Sulfate
Sodium
Bisulfite
Copper (T)
Nickel (T)
0.36
0.36
COD 630
Chronium (T) 0.42
Zinc (T) 1.0
0.00024
0.00024
0.95
0.00063
0.0015
1.1
1.1
2500
1.3
3.4
0.00074
0.00074
3.8
0.0020
0.0051
16
-------
SECTION 3
INTRODUCTION
AUTHORITY
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 Pollutant Discharge Elimination System (NPDES) permits
issued under Section 402 of the Act, pretreatment standards were made
enforceable directly against dischargers to POTW (indirect
dischargers).
Although Section 402(a)(l) 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
17
-------
develop a
applicable
501 (a) of
additional
the Act.
list of toxic pollutants and promulgate effluent standards
to all dischargers of toxic pollutants. Finally, Section
the Act authorized the Administrator to prescribe any
regulations "necessary to carry out his functions" under
The EPA was unable to promulgate many of these regulations by the
dates contained in the Act. In 1 976, EPA was sued by several
environmental groups, and in a settlement of this lawsuit EPA and the
plaintiffs executed a "Settlement Agreement" which was approved by the
Court. This Agreement required to 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. versus Train,
9 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.
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 incurred by
and the effluent reduction benefits from a 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 by July 1, 1984, whichever is later, but not later than July 1,
1987.
18
-------
The purpose of these 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. Fgr 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, 1975. Taken
together, the two groups of regulations cover 49 inorganic chemical
subcategories, many of which include more than one specific chemical
product. Although some toxic pollutants 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 control of the pollutant parameters which accounted, in terms
of quantity, for most of the pollution loading of navigable waters
attributable to the manufacture of inorganic chemicals.
Court Remand of Regulations
On March 10, 1976, the United States Court of Appeals for the Fourth
Circuit decided in E.I. du Pont de Nemours & Company, et al. versus
Train, 541 F. 2d 1018 (4th Cir. 1976), to set aside and remand for
reconsideration a number of general definitions and specific discharge
regulations promulgated in 1974. These regulations are all within
Title 40, Parts 401 and 415 of the Code of Federal Regulations and are
listed below:
General Provisions
401.11 (i) - Definition of effluent limitations
401.11 (q) - Definition of process wastewater
401.11 (r) - Definition of process wastewater pollutant
Chlor-Alkali
415.63 - BATEA
Hydrochloric Acid
19
-------
415.72 - BPCTCA
415.73 - BATEA
415.75 - New sources
Hydrofluoric' Acid
415.82 - BPCTCA
415.83 - BATEA
415.85 - New sources
Nitric Acid
415.102
415.103
415.105
BPCTCA
BATEA
New sources
Sodium Carbonate
415.152 - BPCTCA
415.153 - BATEA
415.155 - New sources
Sodium Dichromate
415.173 - BATEA
Sodium Metal
415.182
415.183
415.185
Sodium Silicate
BPCTCA
BATEA
New sources
415
415
415
Sulfuric
415
415
415
415
192 - BPCTCA
193 - BATEA
195 - New sources
Acid
210 - Applicability
212 - BPCTCA
213 - BATEA
215 - New sources
Titanium Dioxide
415.220 - Applicability
415.222 - BPCTCA
415.223 - BATEA
415.225 - New sources
For the most part, the main target of the remand was the zero
discharge regulations from which the industry petitioners sought
relief on grounds of technological infeasibility. During 1975, the
Agency funded a special study of the remand issues (3) and was
prepared to propose amended regulations.
20
-------
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 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.
The Settlement Agreement
A consent decree was issued in a suit filed by four environmental
groups in Natural Resources Defense Council versus Train, 9 ERG 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 (PCB's), cyanide, 13 heavy
metals and asbestos. Table 3-1 lists the 129 toxic pollutants
(sometimes referred to in the literature as "priority pollutants").
The Settlement Agreement also identified 21 point source 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 Calssification (SIC) code
numbers. For the Inorganic Chemicals Manufacturing Point Source
Category, the major industries included are:
SIC (2182 - 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. Most of these
subcategories, A9 in all,- had already been covered by BPT and BAT
discharge regulatTotis; promulgated in 1974 and 1975. Those regulations
established point of discharge control levels for the conventional
^parameters such as pH, TSS, BOD, and oil and grease. In many cases,
specific chemical parameters were regulated, particularly Arsenic,
Chromium, Copper, Mercury, Nickel, Lead, Selenium, Zinc, and Cyanide,
which are now included in the list of toxic pollutants. Other
regulated parameters such as aluminum, barium, iron, ammonia, fluoride
and sulfide are not presently listed as toxic chemicals but are to be
21
-------
treated as nonconventional pollutants
limitations and standards of performance.
under future discharge
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:
A. No. 63, Ferrous Sulfate, is already covered by the Titaniun
Dioxide - Sulfate Process subcategory and does not require
separate consideration.
B. 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
I 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 wastewater 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 subcategor^gs^LTable 3-2, No's. 1 through
35) are covered in this report. This group also includes the 11
subcategories whose final regulations were remanded for restudy in
E.I. du Pont de Nemours and Company, et al, versus Train, supra, and
the four additional subcategories whose interim, final or proposed
regulations were revoked and reserved by the Agency.
It was anticipated by the Agency that a substantial number of the 35
industries to be screened would also qualify for exclusion under
Paragraph 8 on the basis of the analytical results obtained from the
process wastewater 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 results is given under the
individual subcategory sections of this report. The additional
recommended exclusions include the following:
22
-------
TABLE 3-1. FECOMMENEGD LIST OF TOXIC POLLUTANTS
to
LO
17.
1 S .
19.
20,
32.
33.
* acenaphtl^ane
• aero) e in
" acryJonitrile
• benzene
*benzidine
* carbon teirachloride
(teir o chi of om ethane)
•chlorinated benzenes (other than
dichiorol'pnzenes)
chlorooer.zene
l^^-trichiorobenzene
hexachlorobenzene
•chlorinated ethanes (including 1,2-
dichloroethane, 1,1,1-trichloro-
ethane a^id he xachioroe thane)
1,2-dichlor oethane
1,1,1-trichloroeihane
hexachloroethane
l,l-dic~.ioroethane
l,t,:-:ncriioroethane
1,1,2, i-teirachlor oethane
ch!oroe:nane
•chloroalkyl ethers (chlorornethyl,
chloroethyl and mixed ethers
bis(chloromethyl) et'ner^i
bisl2-ch)oroethyl) ether
2-chloroethyl vinyl ether (mixed)
•chlorinated naphthalene
2-chloronaphihaleie
•chlorinated phenols (other than
those listed.elsewhere; includes
trichiorophenols and chlorinated
cresols)
2,ft,6-tric>ilorophenol
p-chJoro-m-cresol
•chioroiorui (trichioromethane)
*2-chlorop>ienoi
• dichiorobenzenes
1,2-dichlorobenzene
1,3-dichiorobenzene
1 ,ft-dichlorobenzene
• di chJorooenz i dine
3,3-dichiorobenzidine
•dichloroethylenes (1,1-dichloroeth-
ylene and 1,2-dicnloroethylenel
1,1-dichioroethylene
*2,iliJOTamtiane( 11,12-
benzofluoranthene)
chrysene
acenaphthylene
anthracene
benzo(ghi)perylene(l,12-
benzoperylene)
fluorene
phenanthrene
dibenzo(a,h)anthracene (1,2,: 5,6-
dibenzanthracene)
indeno (l,2,3-cd)pyrene (2,3-o-
phenylenepyrene)
pyrene
•tetrachloroethylene
•toluene
•trichloroethylene
*vinyl chloride (chloroethylene)
pesticides and metabolites
*aldrin
*dieldrin
"chlordane (technical mixture &
metabolites)
•DDT and metabolites
ft,d'-DDT
V-DDE (p,p'-DDX)
ft,t'-DDD (p.p'-TDE)
•endosulfan and metabolites
a-endosulf an- Mpha
b-endosulf an-Beta
endosulfan sulfate
•endfin and metabolites
endrin
endrin aldehyde
•helptachlor and metabolites
heptachlor
heptachlor epoxide
•hexachlorocyclohexane (all isomers)
a-BHC-A!pha
b-BHC-Beta
g-BHC (Iindane)-Gamm3
d-BHC-Delta
•potycNorinated biphenyls (RGB's)
PCB-12W (Arochlor J2»2>
PCB-1251 (Arochlor 1251)
PCB-1221 (Arochlor 1221)
PCB-1232 (Arochlor 1232)
PCB-12ft8 (Arochlor 12ft8)
PCB-1260 (Arochlor 1260)
PCB-1016 (Arochlor 1016)
•toxaphene
•antimony (total!
•arsenic (total)
"asbestos (fibrous)
•beryllium (total)
'cadmium (total)
•chromium (total)
'copper (total)
'cyanide (total)
-lead (total)
•mercury (total)
»nickel (total)
•selenium (total)
'silver (total)
'thallium (total)
'zinc (total)
"2,3,7,8-tetrachlorodibenzo-p-
dioxin (TCDD)
85.
86.
87.
88.
89.
90.
91.
92.
93.
9ft.
95.
96 .
97.
98.
99.
100.
101 .
102.
103.
10ft.
105.
106.
107.
108.
109.
ilO.
111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
12ft.
125.
126.
127.
128.
129.
,t;^6FR 10723.
t2)oeleted CH/OS/81; t6 FR 2266
•Specific compounds and chemical
classes as listed in the consent
decree.
-------
TABLE 3-2. SCOPE OF INDUSTRY COVERAGE WITHIN THE INORGANIC
CHEMICALS MANUFACTURING POINT SOURCE CATEGORY
Subcategories Designated for Initial Study
1. Chlor-Alkali
2. Hydrofluoric Acid
3. Hydrogen Peroxide
4. Titanium Dioxide
5. Aluminum Fluoride
6. dhrone Pigments
7. Hydrogen Cyanide
8. Sodium Dichromate
9. Carbon Dioxide
10. Carbon Monoxide/Hydrogen
11. Copper Sulfate
12. Nickel Sulfate
13. Silver Nitrate
14. Sodium Bisulfite
15. Sodium Hydrosulfite
16. Hydrochloric Acid
17. Nitric Acid
18. Sodium Carbonate
19. Sodium Metal
20. Sodium Silicate
21. Sulfuric Acid
22. Ammonium Chloride
23. Aimonium Hydroxide
24. Barium Carbonate
25. Boric Acid
26. Calcium Carbonate
27. Copper Oxide
28. Manganese Sulfate
29. Strong Nitric Acid
30. Oxygen and Nitrogen
31. Potassium Iodide
32. Sodium Hydrosulfide
33. Sodium Silicofluoride
34. Sodium Thiosulfate
35. Sulfur Dioxide
36. Bromine
37. Calcium Hydroxide
38. Chromic Acid
39. Fluorine
40. Hydrogen
41. Iodine
42. Potassium Chloride
43. Stannic Oxide
44. Zinc Sulfate
45. Calcium Carbide
46. Calcium Oxide
47. Potassium Metal
48. Potassium Sulfate
49. Sodium Bicarbonate
50. Borax
51. Ferric Chloride
52. Lead Monoxide
53. Sodium Fluoride
54. Aluminum Chloride
55. Aluminum Sulfate
56. Potassium Dichromate
57. Calcium Chloride
58. Sodium Chloride
59. Sodium Sulfite
60. Potassium Permanganate
61. Zinc Oxide
62. Lithium Carbonate
63. Ferrous Sulfate
24
-------
Number
Subcateqory
3.
9.
10.
15.
16.
17.
18.
19.
21 .
22.
23.
24.
25.
26.
27.
28.
29.
30.
31 .
32.
34.
35.
f Hydrogen Peroxide -^.,
f Carbon Dioxide "^
VCarbon Monoxide/Hydrogen ^
5o3iUifi"Hyarosul"fite
Hydrochloric Acid
Nitric Acid
Sodium Carbonate
Sodium Metal
Sulfuric Acid
Ammonium Chloride
Ammonium Hydroxide
Barium Carbonate
Boric Acid
Calcium Carbonate
Copper Oxide (one plant)
Manganese Sulfate (one plant)
Strong Nitric Acid
Oxygen and Nitrogen
. Potassium Iodide
Sodium Hydrosulfide
Sodium Thiosulf^te
Sulfur Dioxide
Silver Nitrate, No. T3, and Sodium Silicofluoride, No. 33, are being
deferred for future study under Phase H of the BAT regulation
development program for Inorganic Chemicals^ This deferrment was
caused by problems with plant access during the course of the present
study.
General Approach and Methodology
Initiating
pollutant
and undertaking a comprehensive study of the toxic
problem in the Inorganic Chemicals Industry was preceded by
of data and
basis for the
identity of
and standards
by documented
an intensive evalutation by the Agency of the kinds
supporting information that should be assembled as a
development of regulations. All major decisions on the
pollutants and the establishment of effluent limitations
of performance for each subcategory had to be suportable
evidence collected from operating production facilities. Similarly,
the necessary information on production trates, processes, raw
materials, water use, waste sources, and treatment technologies in
practice has 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 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.
-------
Industry Data Base Development and Subcategorization Review
Information from individual manufacturers and previous study documents
was 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.
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 and other 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.
Engineering Evaluations
Section 6 describes the procedures and sources used in developing the
industry productions and wastewater generation characteristics that
form the basis of the model plant concept. The sources of detailed
process and waste treatment information are also presented. Section 7
contains an evaluation of treatment technology presently applied in
BPT systems and advanced technologies that may be recommended for BAT
and NSPS applications. Section 8 provides estimates of the
treatablilty 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 statistical analysis of long-term monitoring
data. The statistically derived parameters, including variability
factors for the 24-hour maximum and maximum 30-day average limitations
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 on the basis of the screening and
"verification data and the rationale for the application of advanced
level technologies is presented.
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
26
-------
the advanced level treatment alternatives. Costs of RCRA compliance
and implementation have not been included in these cost estimates.
Figures for RCRA costs are included in a supplementary document,
"Contractor Report on RCRA ISS Compliance coats for Selected Inorganic
Chemicals Industries". RCRA costs are considered in the Economic
Impact Analysis of Pollution Control Technolqcuejs for Segments of_ the
Inorganic Chemicals Manufacturing Industry, EPA 440/2-81-023.
Treatability Study
Data was collected through a treatability study to evaluate the
achievable performance of proposed BAT for the treatment and control
of pollutant discharges, and to provide empirical treatment system
performance information applicable to selected inorganic chemical
subcategories. The study, completed in July 1980, specifically
concentrated on those subcategories in the Inorganic Chemicals
Industry for which analytical data on raw wastewaters and treated
effluents either did not exist or was deficient, and for which data
were needed for purposes of comparison with proposed effluent
limitations. Subcategories for which treatability was studied
include:
Nickel Sulfate
Hydrofluoric Acid \"O
Copper Sulfate
Chlor-Alkali (Diaphragm Cells)
Titanium Dioxide (Chloride Process)
Chrome Pigments
Sodium Dichromate
Sodium Bisulfite
Sodium Hydrosulfite
General Criteria for Effluent Limitations
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
27
-------
Institute v. EPA, 526 F. 2d 1027 (3rd Cir. 1975). In balancing costs
in relation to effluent reduction benefits, EPA considers the volume
and nature of existing discharges, the volume and nature of discharges
expected after application of BPT, the general environmental effects
of the pollutants, and the cost and economic impacts of the required
pollution control level. The Act does not require or permit
consideration of water quality problems attributable to particular
point sources or industries, or water quality improvements in
particular water bodies. Therefore, EPA has not considered these
factors. See Weyerhaeuser Company v. Costle, 590 F.2d 1011 (D.C. Cir.
1978}.
BAT Effluent Limitations
The factors considered in assessing best available technology
economically achievable (BAT) include the age of equipment and
facilities involved, the process employed, process changes, non-water
quality environmental impacts (including energy requirements) and such
other factors as the Administrator deems appropriate. (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 performace 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 BAT
regulations, however, EPA has given substanital weight to the
reasonableness of costs. The Agency has considered the volume and
nature of discharges, the volume and nature of discharges expected
after application of BAT, the general environmental effects of the
pollutants, and the costs and economic impacts of the required
pollution control levels. Despite this expanded consideration of
costs, the primary determinant of BAT is effluent reduction
capability. As a result of the Clean Water Act of 1977, 33 USC 1251
et seq. the achievement of BAT has become the principal national means
of controlling water pollution due to toxic pollutants.
BCT Effluent Limitations
The 1977 amendments added Section 301 (b) (2) (E) to the Act,
establishing "best conventional pollutant control technology" (BCT)
for discharges of conventional pollutants from existing industrial
point sources. Conventional pollutants are those defined in Section
304 (b) (4) -BOD, TSS, fecal coliform, and pH. Oil and grease was
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.
Section 304(b)(4)(B) of the Act requires that BCT limitations be
assessed in light of a two part "cost-reasonableness" test. American
Paper Institute v. EPA, 660 F.2d 954 (4th Cir. 1981). The first test
28
-------
compares the cost for private industry to reduce its conventional
pollutants with the costs to publicly owned treatment works for
similar levels of reduction in their discharge of these pollutants.
The second test examines the cost-effectiveness of additional
industrial treatment beyond BPT, EPA must find that limitations are
"reasonable" under both tests before establishing them as BCT. In no
case may BCT be less stringent than BPT. EPA published its
methodology for carrying out the BCT analysis on August 29, 1979 (44
FR 50732). However, that cost test was remanded by the United States
Court of Appeals for the Fourth Circuit. American Paper Institute y_._
EPA, 660 F.2d 954 (4th Cir. 1981). The Court of Appeals ordered EPA
to correct data errors underlying EPA's calculation of the first test,
and to apply the second cost test. (EPA had argued that a second cost
test was not required). The Agency is Currently developing a new
methodology.
The proposed regulations had set BCT equal to BPT in all subcategories
except for the Chlor-Alkali subcategory (Diaphragm Cell Segment) and
the Hydrofluoric Acid subcategory, either because BPT was set equal to
BAT or because additional TSS removal failed the Agency's original
"cost-reasonableness" test. In the Chlor-Alkali (Diaphragm Cell) and
Hydrofluoric Acid subcategories, additional TSS removal passed our
original test. Pending formulation by the Agency of a new BCT
methodology, the Agency is deferring promulgation of BCT limitations
for the Chlor-Alkali (Diaphragm Cell) and Hydrofluoric Acid
subcategories. BPT is the minimum level of BCT control required by
law. In all other subcategories, we h^ve identified no other
economically achievable technologies which result in significant
additional removal of conventional pollutants. No possible
reassessment of BCT pursuant to the Court's remand could result in BCT
limitations different than those promulgated. Accordingly, the Agency
is promulgating BCT equal to BPT for all oth^r subcategories.
The cost calculations for the Chlor-Alkali (Diaphragm
Hydrofluoric Acid subcategories are presented below.
Cell) and
In the diaphragm cell segment of the Chlor-Alkali Subcategory, the
cost for removal of additional conventional pollutants is $0.53 per
pound. The calculation is as follows:
$0.22
(0.51 kg/kkg - 0.32 kg/kkg)
$0.53 per pound
of TSS removed,
where $0.22 is the increased cost for SAT filtration over BPT
treatment cost in dollars per kkg of production, 0.51 kg/kkg is the
BPT total suspended solids maximum 30-day average limitation from
Table 11-35, and 0.32 kg/kkg is the achievable maximum 30-day average
TSS loading level with filtration. One kilogram (kg) is 2.2 pounds.
In the Hydrofluoric Acid Subcategory, the cost for removal of
additional conventional pollutants is $0.32 per pound. The
calculation is as follows:
29
-------
$1 .46
5.3 kg/kkg - 3.2 kg/kkg
$0.32 per pound
of TSS removed,
where $1.46 is the increased cost for BAT treatment over BPT treatment
cost in dollars per kkg of production from Table 12-16, where 5.3
kg/kkg is the BPT total suspended solids limitation from Table 12-21,
and where 3.2 kg/kkg is the effluent TSS loading achievable by
application of BAT. See Section 12.
In the Hydrofluoric Acid and Chlor-Alkali subcategories where BCT is
deferred, the TSS and pH limitations are the same as BPT.
New Source Performance Standards
The basis for new source performance standards (NSPS) under Section
306 of the Act is the best available demonstrated technology. New
plants have the opportunity to design the best and most efficient
inorganic chemicals manufacturing processes and wastewater 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.
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 also requires pretreatment for pollutants, such as toxic
metals, that limit POTW sludge management alternatives, including the
beneficial use of sludges on agricultural lands. Pretreatment is
required for toxic pollutants that would pass through a POTW in
amounts that would violate direct discharger effluent limitations.
EPA has generally determined that there is pass through of pollutants
if the percent of pollutants removed by a well-operated POTW achieving
secondary treatment is less than the percent removed by the BAT model
treatment system. 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 pretreatment regulations can be found at 40 CFR Part 403, 43 FR
27736 (June 26, 1978).
The Agency is promulgating PSES for the Chlor-Alkali(Diaphragm Cell)
and Chrome Pigments Subcategories and the Agency is amending the
existing PSES for the Copper Sulfate and Nickel Sulfate Subcategories.
The Agency is excluding the Chlor-Alkali(Mercury cell), Hydrofluoric
Acid, Sodium Dichromate, Titanium Dioxide, Hydrogen Cyanide, and
Sodium Bisulfite Subcategories from national categorical PSES under
the provisions of Paragraph 8(b) of the Settlement Agreement because
30
-------
the toxic pollutants in discharges to POTWs from sources in those
subcategories are below treatable levels or are so insignificant as
not to justify developing pretreatment standards. The Agency is not
promulgating PSES for the Aluminum Fluoride Subcategory because a
well-operated POTW with secondary treatment installed achieves better
percent removal of toxic pollutants than is provided by the BAT model
treatment system for this subcategory.
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 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.
31
-------
-------
SECTION 4
SUBCATEGORIZATION REVIEW
Basis for Subcateqorization
Factors Considered
The inorganic chemicals industry is very large and diversified and has
been segmented into subcategories for the purpose of establishing
effluent guidelines. Factors taken into consideration for
subcategorization include: raw materials used, product produced,
manufacturing process employed, geographical location, size and age of
equipment and facility involved, non-water-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.
A. 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 processes 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
wastewater treatment needs, except in the case of trace toxic
materials which may occur in some sources but not in others.
B. Dominant Product
Subcategorization by chemical name of the dominant inorganic
chemical produced involves the least ambiguity in applying
standards to a given point source. This is critical because of
the great variety of product mix, manufacturing processes,
wastewater consitutents, and other factors at existing plants.
Subcategorization by product becomes less useful as product mix
increses in complexity because multi-product wastewater 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 characteristics of the wastewaters are similar and
the same treatment technology can be applied for different
process wastewater waters. If two or more dissimilar processes
33
-------
D
produce wastewater 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
necessary in some
secondary
cases.
subcategorization by process has been
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.
The possibility of ground water contamination may preclude the
use of unlined holding and settling ponds in many locations.
In the northern regions, climatic conditions may necessitate the
inclusion of special provisions to prevent freezing of treatment
system components, particularly biological oxidation units,
clarifiers, ponds, and open collection systems. The costs of
utilizing waste heat sources from the process or providing
various types of thermal protection, such as insulation or burial
of pipes and tanks and building structural shelters, may add
considerably to the capital and O&M cost associated with a
treatment technology.
Thus, the influence of geography, climate, geology, etc. is
reflected, in waste treatment modifications anql 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.
34
-------
H
Plant Age
Plant age can have an important bearing on wastewater 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 wastewater 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 thenselves
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 plant age
is not a reasonable basis for subcategorization.
Non-Water-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 product(s) manufactured and/or the raw
material(s) used. Since both of these elements vary widely
within the inorganic chemicals industry, there is no logic in
subcategorization on the basis of non-water-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
35
-------
of the wastewater. For example, residuals of dissolved heavy
metals will respond to lime precipitation and sedimentation at
high pH without respect to the specific origin of the metals.
This "building block" concept could conceivably result in
selecting various combinations of unit processes to meet the
treatment requirements. However, if the treatment cost must be
expressed in terms of dollars per unit production, this method of
subcategorization crosses product lines and interferes with
comparison of treatment costs based on the production of a
specific chemical. Even if the unit operation is commonly
applicable for treating waste flows of different products, the
cost of treatment will fluctuate because of variations in
quality, loading and flow rates and subcategorization on the
basis of treatment cost is not recommended.
I. Energy Cost
Manufacturing processes in the Inorganic Chemicals Industry
typically have large energy requirements. In contrast,
wastewater treatment processes consume a small fraction of the
total energy used. There appears to be no major energy
requirements for the wastewater treatment facility and
subcategorization on the basis of energy cost is not justified.
J. 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 of costs. Because of the lack of
uniformity within the industry, solid waste generation and
disposal proctices are not a satisfactory basis for
subcategorization.
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 Cyanide which require further
subdivision based on the difference in the quanity and quality of the
wastewater from the processes, and two others, Hydrofluoric Acid and
Aluminum Fluoride, have been reviewed for possible integration (see
Section 4.3).
Secondary Subcateqorization
Chlor-Alkali
36
-------
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
wastewater generated. A principal difference is the presence of
mercury as a contaminant in the wastewaters from the mercury cell
process and asbestos in the diaphragm cell plants 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 produced in the caustic evaporation process. Such water is
not produced in mercury cell plants. The quantity of wastewater
generated from the diaphragm cell plants is several times that of the
mercury cell plants for the same chlorine production capacity. Based
on the quantity and characteristics of the wastewater, further
subcategroization is justified.
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.
The sulfate process uses ilmenite ore and sulfuric acid as raw
materials. The chloride process uses rutile ores or ilmenite ores and
chlorine, with a different process and wastewater characteristics for
each ore. 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 wastewaters 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:
1. Sulfate process
2. Chloride process using rutile ore
3. 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 amount of
dissolved metal chlorides and the treatment technology is expensive.
Solid wastes 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
37
-------
titanium dioxide by the chloride process,
process step generates additional wastes.
and this beneficiation
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 redisues. 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 wastewater 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.
Therefore, further subclassification based on the amount and
characteristics of the wastewater appears to be justified, and the
three process subdivisions indicated above are appropriate for this
purpose.
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
sulfa.tes in addition to cyanide and nitriles.
The primary product in the other process is acrlonitrile (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.
Review of Possible Integration of Subcateqories
Hydrofluoric Acid and Aluminum Fluoride
Aluminum fluoride (A1F3) usually is produced by the reaction of
hydrated alumina (A1203»3H20) with hydrogen fluoride (HF). One plant
produces aluminum fluoride from fluorosilicic acid (H2SiF6), a by-
phophoric acid (H3PO,
and is not
product of
regulation. Two of the aluminum fluoride plants are
integrated with hydrogen fluoride (or hydrofluoric acid
covered by the
known to be
production.
38
-------
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 wastewaters 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 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
wastewaters associated with the two products are similar and a common
treatment facility is normally utillized. In addition, the combined
manufacture of these products does not create a unique or unusual
situation, either with regard to the wastewater treatment requirements
or compliance with discharge regulations. Although the waste gypsum
produced at an HF plant supplies enough calcium for adequate fluoride
removal from neutralized scrubber wastewaters 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, wastewater flow, and the toxic
polllutant loadings are not 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 creaton of
an HF/A1F3 combined product subcategory is not being made at this
time.
Summary
The recommended subcatetjorization with process subdivisions include
the following:
Subcateqory
Chlor-Alkali
Titanium Dioxide
Process Subdivisions
Mercury Cell
Diaphragm Cell
Sulfate
Chloride-Rutile
39
-------
Hydrogen Cyanide
Chloride-IImenite
Andrussow Process
Acrylonitrile By-Product
{Included in the Organic
Chemicals Regulation.)
40
-------
SECTION 5
SCREENING AND VERIFICATION SAMPLING PROGRAMS
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 chemical industry in terms of factual information derived
from the chemical analysis and flow measurement of representative
process raw wastewater 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 wastewater
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 significant. 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 are
presented in the paragraphs below.
Selecting Plants and Making Preliminary Contacts
In each subcategory, plants were selected for screening on
of the following general criteria:
the basis
A. Minimal product mix and no organic product lines which could
increase the potential for interprocess cross contamination
of wastewaters.
B. 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.)
41
-------
C. Manufacture of industrial grade products in
than low volume reagent grade products.
volume, rather
D. Median production capacity with the subcategory.
E. Segregated waste streams to facilitate sampling.
F. NPDES discharges rather than discharges to POTWs, since
treatment for a NPDES discharge is usually more extensive.
G. Geographical clustering of selected plants to facilitate
field logistics, but only to the 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. Information provided by industry for which
confidential treatment was requested has been handled in accordance
with the provisions of 40 CFR Part 2.
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.
Screening and Verification Sampling
A. Collection of Samples for Screening
In the screening phase of the sampling program, the specific
objective was the detection and quantification of waterborne
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 approximately 2-hour
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.
42
-------
Samples were also taken from the composites, or as individual
grabs, for the analysis of the conventional and nonconventional
pollutants.
Volatile organics were collected in teflon-sealed screw cap
vials. Eight 40 ml vials were filled at each sampling site by
grab sampling in pairs at approximately 2-hour intervals. The
individual vials were cooled to 4°C and shipped to the laboratory
where they were used to prepare composites in duplicate just
prior to analysis. Three blank vials prepared and sealed in the
laboratory accompanied each set of samples during collection,
shipment, and storage.
B. 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.
C. 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°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°C.
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. 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:
the date the sample was received, date due, number and type of
each sample, and the analysis required.
43
-------
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 were received, and due dates.
All samples were kept in a laboratory refrigerator at 4°C when
not being handled by the analyst. Upon completion of analysis,
the sample was checked back into the Sample Control Department
and kept in an identified location in the Sample Control
refrigerator. A report of completed samples was then sent to the
EPA Sample Control Center.
D. 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 potentially treatable
concentration or 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 potentially treatable 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 treatmnt and
control of that toxic pollutant.
2. A subcategory which had no potentially treatable raw waste
concentration of any toxic pollutant would not be subject to
verification sampling and would likely be excluded from
regulatory coverage in accordance with the provisions for
exclusion under Paragraph 8 of the Settlement Agreement.
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 verifiction phase of the program was initiated, an important
decision was made with regard to metals analysis. First, in view or
44
-------
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
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 Pollllutants by
Environmental Monitoring and
April, 1977.
U. S. Environmenal Protection Agency,
Support Laboratory, Cincinnati, Ohio,
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.
A. Trace Metal Analysis
Figure 5-1 shows a data flow diagram for metals analysis. Atomic
absorption methods described in 40 CFR 8 136 per Section 304(h)
were used. A set procedure was followed in the laboratory to
generate the analytical values and the quality control data. The
data flow diagram shows the actual sequence employed in
verification analysis and the following notes, which are keyed to
the diagram, provide additional information on the procedures.
1. Blanks—two for each set of analyses digested.
Duplicates—one every seventh sample.
2. Quality Control at Operator Level (Atomic Absorption):
Blanks—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.
Standards—Three different concentrations were run at the
beginning and end of every set analyzed for each metal.
45
-------
3.
4.
Standards were also run every tenth sample during the
analysis of a set.
Spikes—These were made according to the EPA "Method of
Standard Additions," by adding such a volume of standard as
to double the apparent concentration of metal present in the
sample. Extrapolation backwards of the resultant
absorbances allowed correction of absorbance for matrix
effects.
Duplicates—For furnace analysis, the sample was run twice
wherever a low but positive absorbance was obtained. As
well as this, one sample in every seven was run in duplicate
routinely. The average of duplicate measurements was the
taken value; the difference between duplicate measurements
was noted and recorded on control charts. If
reproducibility was outside the limits of ±33 percent, the
measurement was repeated.
UTD = "Unable to Determine" due to matrix interferences.
Criteria Employed in Spike Selection: samples were chosen
to be spiked based upon the following criteria:
All samples where there was any suspicion that
interference or matrix effect was present
All samples containing a measurable concentration of
analyte
In addition, at least one sample in every seven.
The level of spike chosen was controlled by the following
factors:
It should approximately double
concentration
the
apparent
If this results in an absorbance greater than that of
the highest standard, the spiked sample, is suitably
diluted with distilled water.
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. The 13 priority pollutants metals, with the exception of
mercury, were determined by AA spectrophotometry in the
furnace (HGA) mode.
46
-------
t STANDARDS 1 J
i
FURNACE
t Cd.Pt>.SI>.Tt 1
i
i ;
UTO DETERMINATE
(3) VALUE
I
FLAME ANALYSIS
I
* f *
UTD DETERMINATE
O) VALUE
1
FLAME ANALYSIS -""'^
FIELD SAMPLING
I
PRESERVATIVE ADDED
•CEO . AND AIR BHJPED
I
RECEIPT . LOG IN SAMPLES
ANO REFRIGERATE
1
QUALITY CONTROL BLANKS AND
DUPLICATES CREATED (1)
V
PREPARATION 8V
ACK) DIQEBTIOH
I
ATOMIC ABSORPTION ANALYSIS (»)
__1
FLAME VAPOR QENERTION
1 * , _ "*
I i ^—ii
"~t
OFF-SCALE DETERMINATE OFF-SCALE DETERMINATE
RESULT VALUE RESULT VALU6
t \ _
DILUTE . DILUTE
SAMPLE SAMPLE
-^ X
•^ . » . x/
CALCULATES t X/ SELECT** Of
,„ ANALYSIS OF ' ........fl TO nc
- RESULTS •• 8AMM-fc" ' ,
^^- i 1 1
/ FtMAt NV
<^ INSPECTION. IS ^> N° 1. JEUISfSk
N^ DATA VALID ^X BEOWMO
1 YES
TYPING I
MPOOTttta
*
HYDRIDE GENERATION
I A*,S« >
*
* *
OFF-BCALE DETERMINATE
RESULT VALUE
1
DILUTE
SAMPLE
Figtire 5-1. Sample flow sheet for metal analysis.
-------
2, If matrix interference was seen, they were spiked and
redetermined.
3. If difficulties due to matrix interference persisted, or if
metal concentrations appeared high enough, the determination
was repeated in the flame mode.
4. Mercury was determined by the standard cold vapor method.
Certain changes in analytical protocol were instituted during
verification analysis in order to avoid the excessive matrix
interference experienced during screening when the heated graphite
atomizer (HGA) was the primary method applied to the analysis of 12 of
the metals. The modified protocol for metals was:
1. Six elements were determined by flame only, namely, Ag,
Cu, Cr, Ni and Zn.
Be,
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.
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 reproducibility.
2. Selenium and arsenic were determined by hydride generation
using sodium borohydride (NaBH4). This greatly minimized
problems associated with matrix interference. The method is
very reproducible and the detection limits were at levels
well below the verification criteria for 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
48
-------
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 in the subparts per'million range.
See Table 5-1 for a summary of these limts, together with those
of the original protocol.
B. Organic Compound Analysis
The organic toxic pollutants were determined by the standard
protocol (40 CFR 8 136 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 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.
C. Cyanide Analysis
The standard methods for the wet chemical analysis of total
cyanide and cyanide amenable to chlorination (Cyanide A) were
utilized (40 CFR S 136). Cyanide analysis is subject to several
sources of interference, including:
1. 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 wastewater and
during the chemical analysis for cyanide.
2. Oxidizing agents-The presence of free chlorine in the
wastewater 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.
3. Sulfides-Sulfide or bisulfide will interfere in the analysis
of cyanide by reacting with the colorimetric reagents.
The presence of sulfur dioxide or bisulfite in the wastewater sample
should have no appreciable effect on cyanide results. Detection
limits on the order of 1-4 ^g/1 can be achieved by the analytical
49
-------
method employed, but the results have to be interpreted with regard to
the possible interfering components of the sample.
D. Hexavalent Chromium (Cr VI) Analysis
The determination of Cr VI in wastewater samples is also subject to a
number of interferences which can take effect either during sampling
and storage or during analysis.
1. 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.
2. Reducing agents-Samples containing sulfur dioxide,
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).
The detection limits for Cr IV using the diphenylcarbazide
colorimetric method are on the order of 1-3 */g/l in the absence of
substances which interfere with color development.
E. 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).
F. Conventional and Nonconventional Pollutants
All techniques used for the analysis of BPT control parameters
(conventional and nonconventional pollutants) were those recommended
by the Agency- The list of approved test procedures was published in
the Federal Register on October 16, 1973 (38 FR 28758) and may also be
found in Title 40 of the Code of Federal REgulations (40 CFR 136).
Quality .Assurance Provision
The Agency and the contractor's analytical laboratories maintain
consistently high standards for accuracy and quality control. As an
in-house requirement, a minimum of ten percent of all samples are
routinely run in duplicate. Quantitation is based on standards which
are prepared in pure water, at concentrations such that all sample
measurements are greater than the absorbance of the lowest standard,
and less than the absorbance of the highest standard. The standards
are also checked by participation in the EPA Reference Sample Program
50
-------
TABLE 5-1. ANALYTICAL DETECTION LIMITS FOR METALS
CD
Element
Antijnony, Sb
Arsenic, As
Beryllium, Be
Cadmium, Cd
Chromium, Cr
Copper, Cu
Lead, Pb
Mercury, Hg
Nickel, Ni
Selenium, Se
Silver, Ag
Thallium, Tl
Zinc, Zn
Original Screening
Protocol t21
Method (>jg/l)
HGA*
HGA
HGA
HGA
HGA
HGA
HGA
Cold Vapor
HGA
HGA
HGA
HGA
HGA
10
3
0.2
1
1
1
10
0.5
1
9
0.5
2
1
First Modification Second Modification
of Protocol ^ of Protocol (4)
Method (jag/1) Method ; (>ag/l)
HGA
HGA
Flame
HGA
Flame
Flame
HGA
Cold Vapor
Flame
HGA
Flame
HGA
Flame
10
3
15
1
25
20
10
0.5
25
9
15
2
25
HGA
Hydride
Flame
HGA
Flame
Flame
HGA
New Cold
Vapor
Flame
Hydride
Flame
HGA
Flame
10
10
15
1
25
20
10
0.5
25
10
15
2
1
Heated Graphite Atomizer
(1) Assuming no matrix interferences requiring dilution of sample.
(2) EPA Contract ND. 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
51
-------
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.
The quality assurance provisions outlined in the EPA Protocol for
GC/MS Analysis of Toxic Pollutnts are rigorously adhered to with one
added precaution, namely, the use of internal standards as a means of
measuring recovery. Although not required by the protocol for
pesticide analysis, this technique is utilized as an in-house quality
control requirement to ensure 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 standardizaiton 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 sytem 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.
-------
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 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 raw waste 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.
53
-------
-------
TABLE 5-2. POLLUTANT FREQUENCY BASED CM SAMPLING PROGRAM RESULTS
IWCLUDIN3 RAW WASTE *
Pollutants Detected
Antimony
Arsenic
Beryllium
Cackniun
Chroniun
Capper
Cyanide
Lead
Mercury
Nickel
Seleniun
Silver
Thallium
Zinc
Benzene
Carbon Tetrachlorida
Chlorobenzene
1, 2-Oichloroethane
1,1, l-Trichloroethane
Hexachloroethane
1, 1, 2-Trichloroethane
1 , 1 , 2 , 2-Tetrachloroethane
Chloroform
1 , 2-Dichlorobenzene
1, 1-Oichloroethylene
I, 2-Dichloropropylene
2 , 6-Dinitrotoluene
Ethylbenzene
Fluorantnene
Bis(2-Chloroiscpropyl) ether
Methylene chloride
Dichlorobrananethane
TrichlorofluoroDethane
Chlorodibronomethane
Naphthalene
Nitrophenol
Pentachlorophenol
Phenol
Bis(2-Ethylhexyl) phthalate
Butyl benzyl phthalate
Di-n-butyl phthalate
Diethyl phthalate
Dimethyl phthalate
Benzo(a) anthracene
Benzo(a) pyrene
3 , 4-Benzof luoroethane
Chrysene
Anthrac^Be
Pluoi-ene
Phenanthrene
Pyrene
Tetrachloroethylene
Toluene
Trichloroethylene
Nitrobenzene
2 , 4-Dinitrophenol
Pollutant Occurrence Based
on Plant Grouping
5 or <5
Plants
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
>5 but £10
Plants
X
X
>10 Plants
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Pollutant Occurrence Based on
Concentration Classification (ug/1)
iSO
28
38
49
45
20
21
25
46
17
46
45
41
9
6
2
1
2
4
1
2
3
15
1
3
1
7
1
1
11
5
2
2
1
1
2
2
20
3
15
5
2
1
1
1
1
1
1
1
1
4
7
3
>50 but
£500
19
12
4
4
13
16
15
2
20
7
7
11
18
1
1
2
1
1
3
1
1
3
>500 but
&2,500
4
3
4
9
9
7
9
1
1
14
1
1
1
1
1
1
2
2
>2,500
1
10
7
2
6
S
8
12
1
1
* Blanks indicate not detected.
55
-------
TABLE 5-3. DISTRIBUTION OP POLLUTANTS ACCORDING
TO SUBCATEGGRy 1
Pollutants Detected
Subcategory Numbers Where Pollutants Found
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Copper
Cyanide
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Zinc
Benzene
Carbon Tetrachloride
Chlorobenzene
1,2-Dichloroethane
1,1,1-Trichloroethane
Hexachloroethane
1,1,2-Trichloroethane
1,1,2,2-Tetrachloroethane
Chloroform
1,2-Dichlorobenzene
1, l-Dichloroethylene
1,2-Dichloropropylene
2,6-DinitrotDluene
Ethylbenzene
Fluoranthene
Bis (2-Chloroisopropyl) ether
Methylene chloride
Dichlorobratutethane
Trichlorofluoromethane
Oilorodibrononethane
Naphthalene
4-Nitrophenol
Pentachlorophenol
Phenol
Bis (2-Ethylhexyl) phthalate
Butylfcenzyl phthalate
Di-n-butyl phthalate
Diethyl phthalate
Dimethyl phthalate
Benzo(a) anthracene
Benzo (a) pyrene
7, 23, 27, 28, 33
Alf but 7, 23, 27, 28, 33
n n
N »
n n
n ti
" " " II " II
1, 3, 4, 10, U, 25, 32
1, 2
1, 35
1, U, 13, 22, 35
1
4,11
1, 10, 35
1, 3, 4, 10, 13, 15, 19, 21, 22, 25, 32,.35
24
1, U, 13
26
1
1
1
1, 3, 4, 9, U, 21, 25, 32
8
22
1, 4, 8, 9, 12, 13, 19, 21, 22, 25,26, 32,
1, 4, 19, 32
1, 4, 25
10 -51
35
1, SZ
17
2, 3, 4, 8, 15
2, 15, 26, 31, 32
1, 4, 7f 8, 10, 11, 12,13,15,18,24,25,26,30,31,35
1, 2, 12
1, 4, 8, 11, 17, 18, 19, 21, 22, 30, 31f 34, 35
8, 10, U, 19, 31
12, 31
a
For name of subcategory, refer to Table 3-2.
2 "All" means subcategory numbers 1 through 35 of Table 3-2.
(Continued)
56
-------
TABLE 5-3. Continued
Pollutants Detected
Subcategory Nunbers Where Pollutants Found
3,4-Benzofluoranthane
Girysene
Anthracene
Fluorene
Phenanthrene
Pyrene
Tetrachloroethyiene
Toluene
Trichloroethylene
8
8
8
8, 12
8
8
1, 4, 10, 22
1, 3, 4, 10, 11, 15, 18, 32
1, 4, 25
57
-------
-------
SECTION 6
PROCESS AND WASTE TREATMENT INFORMATION DEVELOPMENT
AND EVALUATION
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, industry responses to the Section
308-Questionnaires and information supplied by industry after
proposal. The type of material gathered from these sources is
discussed below.
Literature Review
A review of the literature was conducted to identify and collect
information related to manufacturing processes, raw materials, water
use, wastewater sources, wastewater 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 required for evaluating the
subcategorization of the industries.
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,
wastewater 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.
Telephone 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 communcations.
308-Questionnaire Responses
The basis for much of the work in this study is the responses from
industrial organic chemical firms to the 308 data requests.
59
-------
Data from 284 manufacturers' responses were utilized by the project
team for the development of approriate 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 derivative quantities such as
percent utilization, effluent per ton of product, and conversion to
metric units were compiled.
Process Waste Sources and Current Treatment Practices
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.
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
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
are
60
-------
TABLE 6-1. 308-QOESTIONNAIRE RESPONSE DATA
DATA ELEMENTS
INORGANIC CHEMICALS GUIDELINES STUDY
Datum Reference
Description
Comments
Manufacturer
Product
Plant
Process
Effluent Treatment
Name
Location
EPA Region
Name
Subcategory
Number of other
Products
Capacity
Production
Age
Name
Volume of Process
Effluent
Volume of Noncontact
Effluent
Type
Permit
Major Pollutants
Confidential
Inorganic
Chemicals
Fiscal year
1976
1976
1976
61
-------
Description of individual plants visited, sampled and plant
information from other sources.
Inventory of raw proces wastewater sources and
identification of sampling points.
Process wastewater quality and flow data.
Solid waste generation and disposal.
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 flow) 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.).
Model Plant and BPT Treatment Sytem 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
appropriately designed and sized BPT level wastewater 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.
62
-------
Beginning with Section 11, the model plant and BPT level treatment
system descriptions and specifications for eac:h subcategory include
the following information:
Production rates and mode of opearation.
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
design specifications given for its waste treatment system.
and the
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 wastewater 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.
Dissolved Solids in Wastewater Effluents
Many waste treatment plants discharge final effluent into watercourses
which feed fresh water streams used as sources of water supply by
downstream agencies or industries. Groundwater aquifers which
underlie large portions of the country are tapped to supply fresh
water through wells serving public and industrial water needs. Saline
wastes discharged into streams or into unlined lagoons can
significantly alter the salt content (total dissolved solids) of the
fresh water. Although Federal regulations seldom limit the total
dissolved solids or the various ions such as chloride, sulfate,
bicarbonate, and nitrate, these constituents can be of serious concern
to local water users.
To protect the mineral quality of ground and surface waters State and
local water pollution control agencies typically establish limits on
the discharge of substances which contribute sodium, potassium,
hardness, chloride, sulfate, or conductivity, which is a measure of
total solids in solution. This restriction can affect the chemicals
chosen for waste treatment. For example, alkaline precipitation can
be accomplished by using lime, which forms an insoluble calcium
sludge, or by adding caustic soda, forming a soluble sodium salt.
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.
63
-------
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.
64
-------
SECTION 7
ASSESSMENT OF TECHNOLOGY FOR ADVANCED TREATMENT AND CONTROL
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 Settlement Agreement, 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
often 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
wastewater, and determine whether they can be adopted as viable
technological options.
A list of candidate technologies was complied 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 wastewater characteristics.
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
65
-------
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.
Hydroxide Precipitation
Hydroxide precipitation is the most widely used technology for
removing trace metals from wastewaters, 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 ions = insoluble metal hydroxide
If the pH is below the optimum for hydroxide precipitation soluble
complexes form:
(2a)
+ OH- = M(OH)A
Metal ion + hydroxyl ion = soluble metal complex
If the pH is above the optimum for hydroxide precipitation, the metal
hydroxide may redissolve by forming soluble complex hydroxides:
M(OH)Z + OH- = M(OH)3-
Insoluble metal hydroxide + hydroxyl ion = soluble
metal complex
2b)
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:
M++ + OH- + nR = [M(R)nOH]+
Metal ion + hydroxyl ion = soluble metal
+ organic ions chelate
Such complexes may require unusual treatment to hydrolyze them, and
their presence often explains why some treatment practices yield
relatively poor results.
66
-------
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 toxic metals which form insoluble hydroxides,
while Table ^ 7-1 shows the solubility product constants. For
comparison, the values for sulfides are also given in Table 7-1.
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 wastewater 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 ^g/1 standards, and hydroxide precipitation is often
supplemented by the use of coagulating agents to improve solids
removal, or sulfide coprecipitation 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 expeced 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
wastewater is stirred, and a homogeneous sample is taken and analyzed
to determine the chemical dosage requirements. The chemicals are
added, mixed and stirred for about 10 minutes. After the reaction is
complete, the solids 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 larger 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 wastewater 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
67
-------
Figure 7-1. Theoretical solubilities of toxic metal hydroxides/oxides as a
function of pH.
10 11 12 13
NOTE: Solubilities of metal hydroxides/oxides are from data by M. Pourbaix,
Atlas of Electrochemical Equilibria in Aqueous Solutions,
Pergamon Press, Oxford, 1966.
68
-------
TABLE 7-1. SOLUBILITY PRODUCTS OF TOXIC METALS
Metal
Solubility Product Constant (K—)
Metal Hydroxide Metal Sulf ide
Antimony (III)
Arsenic
Beryllium
Cadmium
Chromium (III)
Copper
Lead
Mercury
Nickel
Selenium
Silver
Thallium (I)
Zinc
1.6 X 10
2.5 X 10
6.3 X 10
2.2 X 10
1.2 X 10
3.0 X 10
2.0 X 10
-22 (1)
-14 (1)
-31 (1)
-20 (1)
-15 (1)
-26 (1)
-15 (1)
2.0 X 10
-8 (1)
1.2 X 10
-17 (1)
3.6 X 10
-29 (2)
8.5 X 10
-45 (2)
-28 (2)
3.4 X 10
2.0 X 10~49 (2j
1.4 X10-24 (2>
1.6 X 10
5.0 X 10
1.2 X 10
-49 (2)
-21 (1)
-28 (2)
NOTE: References for above values are shown below.
(1) Dean, J.A., Ed., Lange's Handbook of Chemistry, 12th ed., McGraw-Hill
Book Co., New York, 1979.
(2) Vfeast, R.C., Ed., Handbook of Chemistry and Physics, 57th ed., CRC Press,
Cleveland, Ohio, 1976.
69
-------
systems and the use of flocculant aids as improved levels of
treatment.
Soda ash (sodium carbonate, Na2C03) is sometimes found to be the
reagent of choice particularly for lead removal. Lead carbonate,
PbC03, and lead hydroxide/carbonate, 2PbC03«Pb(OH)2, (basic carbonate)
are formed which may afford improved settling properties for a
particular waste. This practice is found in Chlor-Alkali (Diaphragm
Cell) waste treatment.
Hydrated lime suspensions are more commonly used than soda ash or
caustic soda as the hydroxide source because they are cheaper.
However, if there is sulfate ion present in the wastewater, gypsum
will be formed:
Ca(OH
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 elimiates the scaling problem. Total
dissolved solids in the form of sodium salts are increased in the
caustic soda treated wastewaters. 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
wastewaters. Industries that are using hydroxide precipitation
include:
Inorganic Chemicals
Plating and Metal Finishing
Mining
Textiles
Steel and Iron
Non-Ferrous Metal Processing and
Electronics
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 Table 8-1 through 8-10).
Ferrite Coprecipitation
An interesting variation on the theme of hydroxide precipitation is a
process developed in Japan for the removal of heavy metals from acidic
wastewater. The process, known as ferrite coprecipitation, has the
potential for producing a marketable residual by converting the metal
70
-------
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 wastewater, then
neutralizing and oxidizing the complex heavy metal-ferrite
coprecipitate. Particle sizes are reported to be relatively large and
sludges formed can be safely disposed of by landfilling.
Although extensive performance data have not been developed, the
information available indicates that very high 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 wastewater 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
practice of neutralization and aeration may involve the same chemistry
as the ferrite coprecipitation process.
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+ + - + NazS = MS + 2Na+ (5)
Metal ion + sodium sulfide = insoluble metal sulfide
+ sodium ions
In order to calculate the theoretical solubilities of the metal
sulfides as a function of pH, the equilibria involved in solid metal
sulfide dissociation are taken into account:
MS =
+ S=
(6)
Metal sulfide = metal ion + sulfide ion and, depending on pH, the
sulfide ion can react with hydrogen ions to form the bisulfide ion and
hydrogen sulfide.
S= + H+ = HS-
Sulfide ion + hydrogen ion = bisulfide ion
HS- + H+ = Hj.S
7)
(8)
71
-------
Bisulfide ion + hydrogen ion = hydrogen sulfide
The concentration of metal ion in solution will equal the
concentration of sulfide ion, bisulfide ion and hydrogen sulfide.
Knowing the metal sulfide solubility product (Table 7-1) and the acid
dissociation constants of hydrogen sulfide, Kl = 9.1 x 10~8, K2 = 1.1
x 10~12 (see Reference 2 in Table 7-1) the solubility of the metal ion
can be calculated as a function of the hydrogen ion concentration and,
therefore, as a function of pH.
For a divalent metal ion the equation is:
(M++) = [Ksp (1 + (H+)/l.l x TO-12) + ((H+)*/l x
1/2
Using the above information, the theoretical solubilities of the toxic
metal sulfides were calculated and are shown in Figure 7-1.
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 wastewater
to supply sufficient sulfide ions to precipitate metal sulfides which
have lower solubilities than ferrous sulfide. Typical reactions are:
FeS + Cu++ = CuS + Fe++ (10)
Ferrous sulfide + copper ion « insoluble copper sulfide
+ iron ion
FeS + Ni(OH)2 = Fe(OH)2 + NiS (11)
Ferrous sulfide + = ferrous hydroxide +
nickel hydroxide insoluble nickel sulfide
A detention time of 10-15 minutes is sufficient to allow the reaction
to go to completion (7). Ferrous sulfide itself is also a relatively
insoluble compound. Thus the sulfide ion concentration is limited by
the solubility of ferrous sulfide, which amounts to about 0.02 mg/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).
72
-------
Figure 7-2. Theoretical solubilities of toxic metal sulfides as a
function of pH.
10
r T
a
i
8
0)
•a
a
I
i
12 13
pH
73
-------
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 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 wastewater.
After the flash mix, the wastewater 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 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
Copper
Zinc
Nickel
0.01 mg/1
0.01 mg/1
0.01 mg/1
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 using ferrous sulfide 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),
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
wastewaters. Xanthates contain functional groups capable of forming
insoluble complexes with metals, and the sludge so formed can be
separated by conventional means.
74
-------
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 wastewaters (13-20). These are:
Soluble starch xanthate with a cationic polymer,
Insoluble starch xanthate, and
Fibrous cellulose xanthate
The general removal mechanism is as follows:
2 [ROCS (=S)]Na + M++ = [ROCS (=S)J2M + 2Na+ (12)
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 wastewaters, 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
complex ing agent (20). This pilot study demonstrated that the
xanthates can either be added to a reactor to mix with the wastewaters
or be applied as a precoat on a pressure filter (20). Results of
these pilot studies showed that metals were reduced to below 50 »g/l
(ppb).
Another study indicated cellulose xanthate is as effective as starch
xanthate in removing trace metals. The following table summarizes the
result of the study with a cellulose xanthate dosage of 90 mg/1 and a
contact time of 30 minutes (18-19).
Concentration,, mg/1
Metals Influent Effluent
Cadmium
Chromium
Copper
Iron
Lead
Nickel
Zinc
1 .35
0.30
1 .6
3.1
3.9
2.4
1 .0
0.027
0.022
0.06-0.14
0.08-0.36
0.008-0.021
0.077
0.03-0.04
75
-------
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 commerical 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.
Ion Exchange
Ion exchange is a chemical reaction between the ions in solution and
the ionic sites on an exchange resin. Many natural solids (e.g.,
soils, proteins, and zeolites) exhibit such exchange characteristics.
However, synthetic resins are the predominant ones used for ion
exchange applications in modern industrial technology. These resins
contain functional groups that can react with the ions in solution.
Depending on these functional groups, the resins can be classified
into:
Strongly acidic cation exchanger,
Weakly acidic cation exchanger,
Strongly basic anionic exchanger, and
Weakly basic anionic exchanger.
Cation exchangers are 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
<-R3NOH and -R3NC1), and weakly basic exchangers contain ammonia
functional groups (-NH3OH and -NH3C1).
When the functional groups are used up in the reaction, the resins can
usually be regenerated. Cationic resins can be regenerated by sodium
chloride, hydrochloric acid, suIfuric acid or sodium hydroxide.
Anionic resins are regenerated by sodium hydroxide, ammonium
hydroxide, sodium carbonate, sodium chloride, or hydrochloric acid.
The exchanger can either be added to the wastewaters 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
76
-------
follows a four-step
regeneration, and rinse.
cycle; exchange (service), backwash,
During the exchange step, the reaction between the ions in solution
and the ionic sites in the resin takes place as the wastewater 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+
(13)
Resin + mercury ion = insoluble resin complex
+ hydrogen ions
RSH + HgCl + = RSHgCl + H+
Resin + mercuric chloride ion = insoluble resin complex
+ hydrogen ions
14)
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
appropriate regenerant.
the resins can be regenerated with the
(15)
RSHgCl + HC1 = RSH + HgCl2
Insoluble resin complex = regenerated resin
+ hydrochloric acid + 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.
77
-------
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.
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 bisufite, sodium metabisulf ite, sulfur dioxide and
ferrous salts. The reduction is accomplished readily at low pH with
these reagents. Typical reduction reactions are:
Cr207= + 2H+ * 2Cr+++ + 3S04=
H0
(16)
Sulfur dioxide + dichromate ion = trivalent chromium ion
+ hydrogen ion + sulfates and water
3S03= + Crz07= + 8H+ =
Sulfite ion + dichromate ion = trivalent chromium ion
+ hydrogen ion + sulfate + water
Crz07=
4H+ = 2Cr+++
6Fe+++
7H0
(17)
(18)
Ferrous ion + dichromate ion = trivalent chromium ion
+ hydrogen ion + ferric ion + water
78
-------
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}
(19)
Trivalent chromium = insoluble chromium hydroxide
+ hydroxide ion
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 reduced 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) concentraiton 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 was
used in a dichromate plant to remove chromium from its wastewater. 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(OH)4~ +
4H0
20)
Mercury ion + borohydride ion = insoluble mercury metal
+ hvdroxvl ion + borate ion + water
hydroxyl ion
A mercury level of 0.01 mg/1 in the final effluent has
(3).
been reported
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).
Oxidation Processes
The oxidation of organic substances is generally carried out by
thermal processes such as wet oxidation and incineration, or by
79
-------
biological processes such as the activated sludge
filters, biodises, and aerated lagoons.
process, trickling
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
vessel. This
exhaust gas)
for a variety
liquor (26).
phase in a closed, high-temperature, high pressure
reduces some of the problems (such as air pollution from
inherent in incineration. Wet oxidation has been used
of wastes including pulping waste and acrylonitrile
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 harmful species
in dilute waste streams (phenols, mercaptans, polysulf ides, etc.).
Common chemicals used as oxidizing agents included chlorine,
hypochlorite, hydrogen peroxide, potassium permanganate, ozone, and
chlorine dioxide. Air and oxygen are also used.
The most widely used chemical oxidation technology applicable to the
inorganic chemicals industry is the oxidation of cyanide. The
oxidation reaction between chlorine and cyanide is believed to proceed
in two steps as follows:
CN- + Cl, = CNC1 + Cl-
Cyanide + chlorine = cyanogen chloride + chloride ion
CNC1 + 20H- = CNO- + Cl- + H20
Cyanogen chloride = cyanate ion + chloride
+ hydroxyl ion ion + water
(21)
(22)
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.
80
-------
The cyanates can be further decomposed into nitrogen and carbon
dioxide by excess chlorination or acid hydrolysis.
2CNO- + 40H-
= 6C1-
2C0
2H0
(23)
Cyanate + hydroxyl ion = chloride ion + carbon dioxide
+ chlorine + nitrogen + water
CNO
2H0 + = C0
H0
(24)
Cyanate + hydronium ion = carbon dioxide + ammonium ion
+ water
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:
03 + CN- = 02 + CNO- (25)
Ozone + cyanide = oxygen + cyanate ion
H202 + CN- = CNO- + H20 (26) .
Hydrogen perioxide + cyanide = cyanate ion + water
The advantage of using these two oxidizing reagents is that no
dissolved solids are added to the wastewater. In addition, excess
chlorine is not discharged.
A patented process uses hydrogen peroxide and formaldehyde to
decompose cyanide at about 120°F. This has the advantage of
precipitating cadmium and zinc simultaneously (9).
Laboratory studies in one plant currently practicing alkaline
chlorination indicated that the presence of ammonia in the wastewater
reduces the efficiency of cyanide removal. It is well known that
ammonia reacts with chlorine or hypochlorous acid to form chloramines:
NH, + HOC1 = NH,C1 + H»0
(27)
Ammonia + hypochlorous acid = monochloramine + water, etc.
NH2C1 + HOC1 = NHC12 + H20 (28)
NHC12 + HOC1 = NC13 + H20 (29)
81
-------
If excess chlorine is added, chloromines can be converted into
nitrogen oxides(s):
2NH, + 4HOC1 = N,0 + 4HC1 + 3H,0
30)
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 Chemical (Hydrogen Cyanide Production)
Mining
Plating
The free cyanide level after treatment is generally below 0.1 mg/1
(9).
Membrane Processes
Membrane processes have emerged in the last decade as a promising new
technology for the treatment of saline water and wastewaters. 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 7-3).
The electrodialysis membranes are made very thin and are assembled in
stacks. The flow path is the active portion of the cells.
82
-------
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 dynamic flow
conditions (29). In contrast to electrodialysis, these involve the
transport of solvent, not solute, across the membrane.
Osmosis is a process in which solvent from a dilute solution is
transported spontaneously across a semi-permeable membrane into a
concentrated solution. By applying enough pressure to overcome this
osmotic pressure, reverse osmosis, i.e., the passage of solvent from a
concentrated solution to a dilute solution through a semi-permeable
membrane, occurs. The operating pressure of reverse osmosis units is
usually between 350 and 600 psi. Ultrafiltration usually operates at
a much lower pressure (5 to 100 psi). The predominant transport
mechanism is selective sieving through pores. The membrane retains
high molecular weight dissolved solids such as synthetic resins,
colloids, and proteins. The upper and lower molecular weight limit is
generally defined as 500,000 and 500, respectively.
Membranes are usually fabricated in flat sheets or tubular forms. The
most common material is cellulose acetate but other polymers such as
polyamides are used. There are four basic module designs:
plate-and-frame, tubular, spiral-wound, and hollow fiber. Table 7-2
is a comparison between the various reverse osmosis modules. Membrane
processes are effective in removing (concentrating)) inorganic and
organic substances from a wastestream. Usually extensive 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.
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.
83
-------
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.
Activated carbon is made by charring basic substrates, such as wood,
coke, coal, shell, husks, etc., at 600°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 wastewaters through a
contact bed. When the bed is exhausted, the carbon is either
regenerated or sent to landfill. It is economical for large plants to
regenerate the carbon. This can be done either by thermal
regeneration in a rotary kiln or multihearth incinerator, or by
chemical regeneration by using oxidizing agents such as hydrogen
peroxide or acids and bases.
The application of carbon adsorption has been mainly in organic waste
treatment. Recently, there are studies indicating the effectiveness
of carbon adsorption in removing mercury, cadmium, cyanide, chromium,
lead, nickel, zinc, arsenic, and copper (30, 31).
An interesting development in carbon technology is its use after the
wastewater is ozonized. This combination (known as Bacteriologically
Activate 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
research is under way (32).
this technology are known, although
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 inorganic chemical industry
subcategories studied in this report was found.
84
-------
i
PRODUCT
WATER
CCtOMCRATE WASTE
Figtare 7-3. Electrodialysis process,
85
-------
TABLE 7-2. COMPARISON OF REVERSE OSMOSIS CONCEPTS
CD
Plate-and-Frarae
Large tubes
Spiral
Polyamide hollow
fine fibers
Cellulose acetate
hollow fine
fibers
Packing
Density
-------
Fluoride Removal
The conventional method of treating fluoride-bearing wastes is to
precipitate the fluoride as calcium fluoride by the addition of lime.
The reaction is:
Ca(OH)
2F- = CaF
20H~
(31)
Hydrated lime + fluoride ion = insoluble calcium fluoride
+ hydroxyl ion
Using this process alone, it is difficult to remove fluoride to below
8 mg/1 due to the solubility of calcium fluoride (9, 33). Adding alum
with the lime generally improves the removal efficiency. Fluoride
ions are removed as follows*
A1(OH)3 + F- » A1(OH)2F + OH- (32)
Aluminum hydroxide = aluminum monofluorohydroxide
+ fluoride ion + hydroxyl ion, etc.
A1{OH)2F + F- = AKOHJF;. + OH- (33)
Al(OH)Fz + F- = A1F3 = OH- (34)
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 is wastewater (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 waste 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.
87
-------
Chlorine Removal
The removal of residual chlorine (in the form of hypochlorite) in
industrial wastewater is normally accomplished by the addition of
sulfur dioxide or a related reducing agent such as sodium bisulfite or
sodium metabisulf ite. Typical reactions are shown in Equations 35 and
36.
OC1~
= H2S04
(35
Sulfur dioxide + hypochlorite ion = sulfuric acid
•f water + chloride ion
Na^SOj + OC1- = Na?S04 + Cl~ (36)
Sodium sulfite + = sodium sulfate +
hypochlorite ion chloride ion
Alternatively, hydrogen peroxide, although relatively expensive,
also be used for dechlorination according to Equation 37.
may
OC1- = H0
0
Cl-
(37)
Hydrogen peroxide + hypochlorite ion = water + oxygen +
chloride ion
In the chlor-alkali industry, certain wastewater 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 otherwise 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 n
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.
-------
-------
-------
SECTION 8
TREATABILITY ESTIMATES AND LONG-TERM DATA ANALYSIS
The Development of_ Treatability Estimates
Preliminary Analysis
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 wastewaters 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 estimates of average
achievable performance. The sequence of analytical steps is:
1. Review and analyze applicable performance data.
2. Estimate best performance under optimum treatment
conditions.
3. Estimate average achievable performance under expected
industrial operating conditions.
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
91
-------
TABLE 8-1. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA. SUMMARY -
ANTIMONY AND ARSENIC REMOVAL
Treatment Technology
Antimony
Lime/Filter
Ferric chloride/Filter
Alum/Filter
Arsenic
Lime Softening
Sulfide/Filter
Lime (260 rag/1) /Filter
Lime (600 mg/1) /Filter
Ferric sulf ate
Ferric sulf ate
Lime/Ferric Chloride/
Filter
Activated alumina
(2 mg/1)
Activated carbon
(3 mgA)
Ferric Chloride
Ferric Chloride
pH
11.5
6.2
6.4
—
6-7
10.0
11.5
5-7.5
6.0
10.3
6.8
3.1-3.6
-
—
Initial
Concen-
tration
(mg/1)
0.6
0.5
0.6
0.2
-
5.0
5.0
0.05
5.0
3.0
0.4-10
0.4-10
0.3
0.6-0.9
Final
Concen-
tration
(mg/D
0.4
0.2
0.2
0.03
0.05
1.0
1.4
0.005
0.5
0.05
<0.4
<4.0
0.05
<0.13
Removal
(%)
28
65
62
85
-
80
72
90
90
98
96-99+
63-97
98
-
References
40
40
40
9,
9,
41
41
42
41
9,
43
43
9,
9,
10
10
10
10
10
92
-------
TABLE 8-2. WASTE WATER TREA3MENT OPTIONS AND PERFORMANCE DATA SUMVIARY -
BERYLLIUM AND CAEMIUM REMOVAL
Treatment Technology pH
Beryllium
Lime/Filter 11.5
Cadmium
Line (260 mg/1) /Filter 10.0
Lime (600 mg/1) /Filter 11.5
Lime Softening 5-6.5
Lime/Sulf ide 8 . 5-11 . 3
Ferrous Sulfide (Sulfex) 8.5-9.0
Ferrite coprecipitation/ neutral
Filter
Initial
Concen-
tration
(mg/1)
0.1
5.0
5.0
0.44-1.0
0.3-10
4.0
240
Final
Concen-
tration
(mg/1)
0.006
0.25
0.10
0.008
0.006
<0.01
0.008
Removal References
(%)
99.4 40
95 41
98 41
92-98 8
98+ 44
99+ 7,8,11
99+ 5
93
-------
TABLE 8->3. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
COPPER REMOVAL
Treatment Technology
Lime/Filter 8.
Lime (260 mg/1) /Filter
Lime £600 mg/1) /Filter
Ferric sulfate/Filter
Lime >
Lime
Alum 6.
Liroe/Sulfide 5.
Ferrous sulfide (Sulf ex) 8.
Ferrous sulfide (Sulf ex) 8.
Ferrite Coprecipitation/
FH
5-9.0
10.0
11.5
6.0
8.5
9.5
5-7.0
0-6.5
5-9.0
5-9.0
-
Initial
Concen-
tration
(mg/1)
3.2
5.0
5.0
5.0
10-20
3.0
3.0
50-130
3.2
4.0
Final
Concen-
tration
(rng/1)
0.07
0.4
0.5
0.3
1-2
0.2
0.2
<0.5
0.02
0.01
0.01
Removal
(%)
98
92
91
95
90
93
93
-
99
99+
99+
References
8
41
41
41
9,10
45
45
44
8
7,8,11
5
Filter
94
-------
TABLE 8-4. WASTE WHER TSEKO^EtJS OPTICKS AND PERFOIMftNCE DRXA SOfAEQT -
CHHCMIIM III AND GHPCMIIM VI RMDVK.
Treatment Technology
QironduD
Lima (260 mg/1) /Filter
Lima (600 rag/1) /Filter
Reduction/Lime
RBduc&icn/IjjQB
Lime Softening
Limn/Filter
Lima
Lin
Ferrita ooprecipitation/
Filter
Ferric sulfate
Ferric sulf ata/Fllter
ChranJum VI
Activated carbon
(pulverized, Pitts-
burgh type HC)
Same as above
Activated carbon
(granular)
Ferrite coprecipitatiai
Sulfur (Uorifla reduction
Bisulfite reduction
pH
10.0
U.5
7-8
7-8
10.6-11.3
7-9
9.5
9.5
6.5-9.3
3.0
2.0
6.0
—
Initial
Concen-
tration
tng/1)
5.0
5.0
140 (as
Cr VI)
1300 (as
Cr VI)
—
—
15
3.2
25
5.0
10
10
3
0.5
Final
Concen-
tration
ftng/1)
0.1
0.1
1.0
0.06 Crm
0.15
0.05
0.1
<0.1
0.01
—
0.05
1.5
0.4
0.05
not
detectable
0.01-0.1
0.05-1.0
Renewal
98
98
—
—
98+
98+
99
35
96
98
—
References
41
41
9,10
3,9,10
46
47
45
45
5
46
41
48
48
41
5
9,10
9,10
95
-------
TABLE 8-5. WASTE WATER TREA3MEWT OPTIONS AND PERFORMANCE DATA SUKMAIOT -
LEAD REMOVAL
Treatment Technology
Lime (260 rag/1)
Lime/filter
Lijne (260 mg/1) /Filter
Line (600 mg/1) /filter
Ferrous sulf ate/Filter
Sodium hydroxide (1 hour
settling)
Sodium hydroxide (24 hour
settling)
Sodium hydroxide/Filter
Sodium carbonate/Filter
Sodium carbonate/filter
Sodium carbonate/filter
Ferrous sulfide (Sulfex)
Ferrite coprecipitation/
pH
10.0
8.5-9.0
10.0
11.5
6.0
5.5
7.0
10.5
10.1
6.4-8.7
9.0-9.5
8.5-9.0
—
Initial
Concen-
tration
(mg/1)
5.0
189
5.0
5.0
5.0
—
—
1700
1260
10.2-70.0
5.0
189
480
Final
Concen-
tration
(mg/1)
0.25
0.1
0.075
0.10
0.075
1.6
0.04
0.60
0.60
0.2-3.6
0.01-0.03
0.1
0.01-0.05
Removal
{%)
95.0
99.9
98.5
98.0
98.5
99+
99+
82-99+
99+
99.9
99.9
References
41
5
41
41
41
10
10
49
49
10
9,10
8
5
Filter
96
-------
TABLE 8-6. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA. SUMMARY -
MERCURY II REMOVAL
Treatment Technology pH
Sulfide
Sulfide 10.0
Sulfide/Filter 5.5
Sulfide/Filter 4 . 0
Sulfide/Filter 5.8-8.0
Ferrite coprecipitation/
Filter
Activated Carbon
Activated Carbon/Alum
Activated Carbon
Initial
Concen-
tration
-Cng/D
0.3-50.0
10.0
16.0
36.0
0.3-6.0
6.0-7.4
0.01-0.05
0.02-0.03
0.06-0.09
Final Removal
Concen- (%)
tration
(mg/D
0.01-0.12
1.8 96.4
0.04 99
0.06 99.8
0.01-0.125 87-99.2
0.001-0.005 99.9
<0.0005
0.009
0.006
Reference!
9,10
50
50
50
50
5
9,10
46
50
97
-------
TftBTE 8-7. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
NICKEL REMDVAL
Treatment Technology pH
Lime 8.5-9.0
Lime (260 mg/1) /Filter 10.0
Lime (600 mg/1) /Filter 11.5
Caustic Soda/Filter 11.0
Ferrous sulfide (Sulfex) 8.5-9.0
Ferrite coprecipitation
Initial
Concen-
tration
(mg/1)
75
5.0
5.0
-
75
1000
TABLE 8-8. WASTE WATER TREATMENT OPTIONS AND
SILVER REMOVAL
Final
Concen-
tration
(mg/1)
1.5
0.3
0.15
0.3
0.05
0.20
PEKEK3BMANCE
Removal References
98 8
94 41
97 41
49
99,9 8,11
99.9 5
DATA SUMMARY -
Treatment Technology pH
Sodium hydroxide 9.0
Ferric sulfate (30 mg/1) 6-9
Lime Softening 9.0-11.5
Chloride precipitation
(alkaline chlorination
in the presence of
cyanide)
Ferric chloride/Filter 6.2
Sulfide precipitation 5-11
Initial
Concen-
tration
(mg/1)
54
0.15
0.15
105-250
0.5
-
Final
Concen-
tration
(mg/1)
15
0.03-0.04
0.01-0.03
1.0-3.5
0.04
-
Removal References
72 13
72-83 46
80-93 46
97+ 9,10
98.2 40
very high 9,10
98
-------
TABLE 8-9. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
SELENIUM AND THALLIUM REMOVAL
Treatment Technology
Selenium
Ferric chloride/Filter
Ferric chloride/Filter
Alum/Filter
Ferric sulfate
Ferric sulfate
LinieAilter
Lime/Filter
Thallium
Lime/Filter
Ferric chloride/Filter
Alum/Filter
PH
6.2
6.2
6.4
5.5
7.0
11.5
11.5
11.5
6.2
6.4
Initial
Concen-
tration
(mg/1)
0.1
0.05
0.5
0.10
0.10
0.5
0.06
0.5
0.6
0.6
Final
Concen-
tration
(mg/1)
0.03
0.01
0.26
0.02
0.03
0.3
0.04
0.2
0.4
0.4
Removal
(%)
75
80
48
82
75
35
38
60
30
31
References
40
40
40
51
51
40
40
40
40
40
99
-------
TABLE 8-10. WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
ZINC REMOVAL
Treatment Technology
Lime/Filter
lime (260 mg/1)
Lime (260 mg/1) /Filter
Lime (600 rng/1)
Lime (.600 mg/1) /Filter
Lime/Filter
Sodium hydroxide
Sulfide
Ferrous sulfide (Sulfex)
Perrite coprecipitation
pH
8.5-9.0
10.0
10.0
11.5
11.5
-
9,0
-
8.5-9.0
-
Initial
Concen-
tration
fmg/D
3.6
5.0
5.0
5.0
5.0
16
33
42
3.6
18
Final
Concen-
tration
(mg/1)
0.25
0.85
0.80
0.35
1.2
0.02-0.23
1.0
1.2
0.02
0.02
Removal
(%)
93
83
84
93
77
-
97
97
99+
99+
References
8
41
41
41
41
5
13
5
8,11
5
100
-------
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.
The estimated ranges of average achievable performance are presented
in Table 8-11. In formulating the proposed regulations, these values
were used as maximum 30-day averages in cases where there were
insufficient data from sampling or long-term monitoring of the actual
industry discharges.
Statistical evaluation of long-term monitoring data is described in
Section 8.3 and the results are presented in Appendix A where various
derivative quanitites such as long-term averages and standard
deviations are tabulated.
Final Analysis
Following publication of the proposed regulations on July 24, 1980 (45
FR 49450) additional data on performance of the BPT and BAT options
for several subcategories were evaluated and eventually incorporated
into the basis for the final regulations. The sources of additional
data include the following:
A. Treatability Study for the Inorganic Chemicals Manufacturing
Point Source, EPA 440/1-80-103, July, 1980.
B. Industry comments on the proposed regulations - The written
comments received by EPA as well as comments given orally at the
public hearing on proposed pretreatment standards (October 15,
1980) are part of the official public record of the rule making.
The comments are summarized and responses are given in "Responses
to Public Comments, Proposed Inorganic Chemicals Manufacturing
Effluent Guidelines and Standards," which is a part of the Record
for this rule. Individual comment documents or letters are cited
in this report where they are used as sources of information.
C. Treatability Manual, Volume III
Technologies
for
Control/Removal of Pollutants, EPA 600/8-80-042c, July, 1980.
Table 8-12 presents tabular summaries of the available industry
treatment performance data for most of the priority toxic metals.
These include estimated long-term averages in cases where there were
sufficient data given to utilize the Maximum Likelihood Estimation
method for calculating statistical parameters as indicated in the
footnotes. Overall arithmetic medians and averages are also given for
metals where five or more individual data sets were available.
101
-------
TABLE 8-11. ESTIMATED ACHIEVABLE MAXIMUM 30-DAY AVERAGES FOR THE APPLIED TECHNOLOGIES
o
Kl
Antimony, Sb
Arsenic V
Beryllium, Be
Cadmium, Cd
Copper, Cu
Chromium III,
Cr+3
Lead, Pb
Mercury II,
Hg
Nickel, Ni
Silver, Ag
Selenium, Se
Thallium, Tl
Zinc, Zn
Lime
Settling
0.8-1.5
0.5-1.0
0.1-0.5
0.1-0.5
0.5-1.0
0.1-0.5
0.3-1,6
Lime
Filter
0.4-0.8
0.5-1.0
0.01-0.1
0.05-0.1
.0.4-0.7
0.05-0.5
0.05-0.6
Final
Sulfide
Filter
0.05-0.1
0.01-0.1
0.05-0.5
0.05-0.4
Concentrations (mg/1)
Ferrite
Coprecip- Soda Ash Soda Ash Alum
itation Settling Filter
Filter
<0.05
<0.05
0.01
0.20 0.4-0.8 0.1-0.6
0.01-0.05 <0.01
0.2-1.5
0.4-0.8
0.2-1.0
0.2-1.0
0.5-1.5
0.1-0.5
0.2-0.4
0.1-0.5
0.1-0.5
0.4-1.2
0.05-0.5
0.05-0.2
0.02-1.2
0.2-0.5
0.02-0.5
(continued)
-------
TABLE 8-11 continued
H
O
Arsenic V, As
Chronium VI,
Marcury II,
Hg
Final Concentrations (mg/1)
Ferric Activated SO2 Bisulfite Lime/FeCl,
Chloride Carbon Reduction Reduction Filter
0.05-0.5 0.3 0.02-0.1
0.1 0.01-0.1 0.05-0.5
0.01
Alkaline
Chlori-
nation
Silver, Ag
Selenitim, Se
Thallium, Tl
Cyanide (Free),
CNA
0.05-0.1
0.05-0.1
0.7
0.1-0.5
-------
An industry long-term average effluent concentration was then
estimated for each pollutant/treatment option combination for which
sufficient data were available. For copper and nickel, the average
values for lime/settling were adjusted upward from 0.32 to 0.40 mg/1
in order to show a larger decrease when filtration is added. In the
case of chromium, the average with filtration was adjusted to 0.16
mg/1. Plants presently practicing filtration are generally those with
higher raw waste concentrations of metals in comparison to plants
which can achieve adequate treatment without filtration. This tends
to reduce the observed differences in performance with and without
filtration and, therefore, understates the potential benefit of adding
filtration to a particular lime/settling system. The estimated
achievable long-term average concentrations, as shown in Table 8-13,
generally fall within the estimated range of the corresponding maximum
30-day averages in Table 8-11 which were derived from literature data.
Thus, there is substantial agreement between the two sets of estimates
and there is good reason to conclude that the lower limits of the
treatability ranges in Table 8-11 are actually more like long-term
averages than maximum 30-day averages for the inorganic chemicals
industry. The final toxic metal regulations are based on the
estimated achievable long-term average concentrations in Table 8-13 in
cases where there are insufficient industry specific performance data
available. The numerical limitation in each case was obtained by
multiplying the long-term average concentration by the model plant
unit flow rate and an appropriate variability factor. The variability
factors are selected to represent as accurately as possible the actual
full-scale treatment system's variability under normal operating
conditions.
It is understood that in each subcategory plant treatment system
conditions, particularly where chemical precipitation is involved, are
usually optimized for the removal of only one metal. Other metals may
be removed incidentially under the same conditions although their
removal efficiencies may not be optimal. An example is the prevalent
use of sulfide precipitation/filtration technology for the removal of
mercury. The precipitation is normally carried out under neutral to
moderately-acid conditions in order to limit the amount of residual
sulfide in the system and, depending on specific raw waste
characteristics, to obtain desirable solid properties for filtration.
Under these conditions, the incidental removals of other metals such
as nickel and zinc are not at their maximum efficiencies, but are
still effective.
The industry performance data summarized in Table 8-12 for many of the
toxic metal/treatment combinations express an observed incidental
removal rather than an optimum removal. This provides an empirical
basis for estimating practical control levels for metals under
off-optimum pH conditions in either alkaline precipitation or sulfide
precipitation systems. The Agency does not regard the implementation
of more than one optimized metal removal step as necessary to meet the
final BPT/BAT regulations.
104
-------
TABIE 8-12. INDUSTRIAL WRSTE WATER TREATMENT SYSTEM PERFOiWftNCE —
OF EFFLUENT OONCEHTRKTION DATA ON TOXIC METALS
(1)
O
I/I
Lime/Clarification f.i time/Pi ltration>.i
(mg/1) Source1 ' (mg/1) Source1*1
Antimony
Arsenic
Beryllium
CadmiuQ
Mercury
Seleniun
Silver
Thallitm
Cturcmium
III
0.18
0.080
0.060
0.080
0.12
0.040
0.050
0.050
0.070
0.071
0.072
0.080
0.15
0.18
0.26
0.35
0.36
0.43
0.81
1.8
0.
0.
PM*5' ND
NSMt5) 0.30 CP(7t
0.038 CS(8>
ND ND
TDS 0.076 MF
NEM(5>
MF(11)
ND ND
ND ND
ND ND
ND ND
IS<5) 0.037 cp*15*
NEM(5) 0.046 TDCtl5>
CfiD* ' 0.072 SDC' '
TQgdO) Q>20 soc*15*
CAD(15> 0.28 MF(11)
TPC(16> 0.33 CP(7)
TMt5) 0.44 CP(21)
MF1
TDC<15)
(151
SDC113'
(171
CAD(A"
is'5'
(15)
CPU '
(151
SDC1 '
IS<5>
15 median 0.20 median
32 average 0.20 average
Sulfide/FiltrationJ.)
(mg/1) Source"1
0.23 CfiM(6>
0.17 CfiM(9)
0.096 CflM(6)
ND
ND
(Q)
0.020 CAM11"
*1 O\
0.022 CAM' '
(Qt
0.036 CAM1"'
(i ii
0.057 CflM1 '
ND
0.070 CRM<9>
ND
ND
NA
, , . . , (2) . (2)
Lijn(5/Clarificatic«,4j larae^'iltration^;
(mg/1) Source (mg/1) Source
Copper 0.030
0.030
0.038
0.060
0.070
0.070
0.080
0.090
0.10
0.14
0.54
0.70
1.1
1.5
IS(5) 0.035 CP(7)
CAD(14) 0.13 CP<21>
OMD 0.17 CS
OMJ(5) 0.23 CS<15)
NFM15' 0.25 is'5'
SEP*5' 0.37 MF*11'
CM)'5' 0.90 MF*11'
C«D(5>
OMD(5>
(111
MF* '
(171
CAD1 '
(11
N£M
(151
CS1 '
CS<15)
O.OS5 median 0.23 median
0,32
I^ad 0.017
0.10
0.15
0.19
0.20
0.15
0.13
average 0,30 average
HF(19) 0.038 MFtU)
MF(U) 0.11 CPt7)
OMD(5) 0.41 CPt211
HEM(5)
NFM(5)
median HA
average
Sulfide Filtration!?!
(mg/1) Source1 4)
0.033 CAM(6)
0.056 CftM(9>
0.24 CAMtU1
0.43 CAM(17)
NA
0.032 cao'13'
0.12 CftM(9)
0.16 CfiM(6>
0.46 CADtl5)
NA
(Continued)
-------
TABLE 8-12 continued
(2)
Lime/Clarification :..
Enrj/1) Source1 '
NiOfl 0.020
0.050
0.10
0.17
0.20
0.20
0.25
0.26
0.31
0.33
0.50
1.4
0.23
0.32
NOTES:
IS(5)
CAD{14)
NEM<5>
ffi(l5J
C3(15)
CAD<17J
*<5)
HEM(5)
HS(15>
median
average
Lame/FiltrationJ4J. Sulfide/Filtrationj?,
(mg/1) Source1 ' (mg/1) Source1 '
0.090 §DC118J 0.022 CAM*9'
0.11 CS(15) 0.074 CAM(6)
0.13 CS(15>
0.19 US '
0.20 CS(8)
0.59 MF(llt
0.80 NS(20)
0.13 median MA
0.30 average
(21
Lime/Clarification:,:
(ragA> Source1 J
Zinc 0.020 QMD '
0.040 OMOt5)
0.040 FI(5)
0.10 CAD(14J
0.11 TM(5)
0.15 TC6
0.20 NfM* '
0.24 IS(5)
0.25 ISt5)
0.35 CMJ(5)
0.39 MF1U)
0.54 CM)
0.55 HF(19>
0.60 NFM*5'
8.2 PMI5)
0.20 median
0.7S average
(1) Influent or raw waste concentrations of metals are at treatable levels; i.e., higher than the corresponding tn
Lime/FiltrationJ?: Sulfide/Piltrationj^j
(mg/1) Source1" (mg/1) Source14'
0.018 CS*8) 0.090 TM(5J
0.058 CP*7J 0.13 CAM(9)
0.11 SDC(1B1 0.15 CAM16'
0.25 Mf'11*
0.57 MF(11)
0.11 median NA
0.20 average
aatability ranges given in Table 8-11.
All effluent concentrations are measured off treatment and are expressed as total (dissolved plus suspended) for each metal.
(2) Lime/Clarification and Lime/Filtration treatment means equalization of raw waste influent stream(s) followed by alkaline precipitation using lime
or caustic soda, solids removal by sedimentation or clarification, and either discharge of the clarified effluent directly or discharge of the
filtrate after passage of the clarified effluent ttirough a dual media filter or its equivalent.
(3) Sulflde/Filtration refers to a direct treatment of the equalized raw waste influent by aulfide addition (usually in the form of sodium sulfide or
bisulfide) under conditions ranging from pH 5 to 11 followed by settling and/or filtration by filter press or activated carbon column.
(4) Source Codes:
CflD Oilor-Alkali, Diaphragm Cells
CAM Qilor-Alkali, Mercury Cells
CS Copper Sulfate
Cp Chrome Pigments
Fl foundry Industry
HP Hydrofluoric Acid
IS Iron and Steel
MF Metal Finishing (including electroplating)
HEM Nonferrous Metals
NS Nickel Sulfate
CM) tee Mining and Dressing
PM Paint Manufacturing
SEP Steam Electric Power Generating
SEC Sodium Dichronate
TEC Titaniun Dioxide - Chloride Process
TDS Titanium Dioxide - Sulfate Process
1M Textile Mills
(5) U.S. Environmental Protection Agency, Treatabllity Manual, Vol. ill. Technologies for Control/Removal of Pollutants, EPA BOO 8-80 042 c, July, 1980.
(6) This document. Table 11-16.
(7) This docutent, Table 16-9.
(8) This dcxament. Table 21-11.
(Continued)
-------
TABLE 8-12 continued
NOTES; continued
(9) Olin Corporation, Chemicals Group, Charleston, TO. Letter to Mr. Elwood E. Martin, U.S. EPA, Effluent Guidelines Division, Washington, D.C.,
October 20, 1980. Maximuu likelihood estimates of the long term averages from Olin mercury treatment effluent data by Jacobs Engineering Group, Inc.
(10) This docunent. Table 14-30.
(11) Hamilton Standard, Division of United Technologies Corp., Windsor Locks, CT. Letter to Mr. Richard Kinch, U.S. EPA, Effluent Guidelines Division,
Washington, D.C., ttovaiber 25, 1980. Tabulations of statistical parameters derived iron historical data on the metal finishing industry.
(12) The Chlorine Institute, Inc., New York, N.Y. Letter to Mr. G. E. Stigall, U.S. EPA, Effluent Guidelines Division, Washington, D.C., May 28, 1979.
Attachment "C", a tabular summary of mercury treatment effluent data.
(13) PPG Industries, Inc., Pittsburgh, PA. Letter to Mr. Elwood E. Martin, U.S. EPA, Effluent Guidelines Division, Washington, D.C., January 2, 1981.
Maximun likelihood estimates of the long term averages from PPG mercury and lead treatment effluent data by Jacobs Engineering Group, Inc.
(14) This document. Table 11-37.
(15) U.S. Environmental protection Agency, TreatabiUtY Studies for the Inorganic Chemicals tfanufacturincf Point Source Category, EPA 440/1-80/103, July, 1980.
Maximum likelihood estimates of long term averages from treatability data by Jacobs Engineering Group, Inc.
(16) This document. Table 14-12.
(17) Diamond Shamrock Corporation, Dallas, TX. Letter to Mr. Elwood E. Martin, U.S. EPA, Effluent Guidelines Division, Washington, D.C., October 22,1980.
Tabular sunmary of highest values from treatment effluent during one month of monitoring.
(18) This document. Table lfr-13.
(19) This docunent. Table 12-22.
(20) This document, Table 22-10.
(21) This document. Appendix A.
ND = No data available
HA = Not applicable
-------
TABLE 8-13. ESTIMATED ACHIEVABLE LONG TERM AVERAGE
CCNCEMTRATICNS FDR TOXIC METALS WITH
EFT OR BAT TREATMENT OPTIONS
Toxic
Metal
Antimony
Arsenic
Beryllium
Cadmium
Chrcmium
Cbpper
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Zinc
(1> ID:
(2) m.
Lime/Clarification
fag/1)
n>(1>
ID
ND
0.10
0.32
0.40
0.15
ND
0.40
ND
ND
ND
0.80
Lime/Filtration
(mg/1)
ND<2>
ID
ND
ID
0.16
0.30
ID
ND
0.30
ND
ND
ND
0.20
Sulfide/Piltration
(mg/1)
ID
0.15
ND
ND
ND
0.20
0.10
0.034
ID
ND
ID
ND
0.12
Insufficient data for a reliable estimate
NTn s3-i^-* 3^ia -1 1 r^Vll t*
108
-------
Selection of Toxic Metal Control Parameters
Control Parameters for Hydroxide Precipitation
Section 7.2 of this report describes hydroxide precipitation as the
most widely-used technology for removing trace metals from waste-
waters. Out of the thirteen toxic metal pollutants, two have
hydroxide/oxide solubilities independent of the -1-14 pH range
(selenium and thallium) and two have minimum hydroxide/oxide
solubilities over a wide pH range (antimony at pH 2-10.4 and mercury
at pH 4-12). Arsenic is removable by precipitation with lime
(probably as calcium arsentate) in the presence of excess calcium ion
under neutral to alkaline conditions. The remaining eight toxic
metals have minimum hydroxide/oxide solubilities only over narrow pH
ranges (see Figure 7-1). Lead may also be effectively treated with
carbonate (soda ash, Na2C03) to form insoluble basic lead carbonate
precipitates.
It is clear from the range of optimum pH's illustrated in Figure 7-1
that no single pH exists which can effectively provide optimum removal
of all eight of these metals. Relatively effective removal can be
obtained by dividing the eight metals into two groups. Group A
consists of beryllium, chromium, copper, lead, and zinc. Group B
consists of cadmium, nickel and silver. Because they rarely occur at
treatable levels and, therefore, rarely require removal, one metal
from each group (beryllium and silver) can be eliminated from the
selection of an optimum pH range for- each group. The information in
Figure 7-1 was used to determine the solubility of the six remaining
metals at unit pH increments from 8.5 to 11.5. These data are
presented in Table 8-14.
Table 8-14 indicates that control of any metal of Group A in the
8.5-9.5 pH range should control the other members of the group.
Control of any metal of Group B in the 10.5-11.5 pH range should
control the other members of the group. Control of metals from
different groups will depend on the details of each case. Possible
approaches to controlling metals from different groups might involve
the use of the intermediate 9.5-10.5 pH range or the control of one
metal in one group when the theoretical solubilities of the metal or
metals in the other group are low throughout the 8.5-11.5 pH range.
Control Parameters for Sulfide Precipitation
Section 7.2 of this report describes sulfide precipitation as
superior to hydroxide treatment for the removal of several toxic
metals. The main application of sulfide precipitation is in mercury
removal and mercury, therefore, is the obvious choice as the control
metal for this technology. Figure 7-2 points out, however, that
mercury is the most insoluble of the toxic metal sulfides and that the
solubilities of the metal sulfides are strongly dependant upon pH.
Control of mercury in the acid pH range may result in less than
optimum control of the least insoluble metal sulfides. Therefore,
control of a second metal that is present in treatable concentrations
109
-------
TABLE 8-14. THEORETICAL SOLtBILITIES OF TOXIC METAL
HyDROXTEES/OXEDES AT VARIOUS pH VALUES
pH
Metal
Group A
Cr"^
Oa^
Pb^
zn~
Group B
Cd"
N1++
8.5 9.5 10.5 11.5
Concentration
0.030(1) 0.20
0.00010 0. 000080 (1)
8.0 0.50a>
0.60 0.070C1)
>10 1.0
1.0 0.010
(mg/1)
1.0 9.0
0.00050 0.0020
4.0 >10
0.50 3.0
0.010 0.0010(1)
o.ooioa) o.oio
(1)
Lowest value
no
-------
and that is among the least insoluble of the toxic metal sulfides
could give greater assurance that the metals without effluent
limitations were also being removed. However, it could also result in
higher mercury discharges. Operation of sulfide precipitation in the
neutral or slightly alkaline range should result in acceptable removal
of all toxic metal sulfides as well as minimizing the problem of
hydrogen sulfide evolution. Soluble polysulfide formation can be
prevented by avoiding the very alkaline pH range and by close control
of excess sulfide.
The Use Of Historical Pollutant Data
Determination
Performance
of Effluent Limitation Guidelines Based Upon Historical
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, and
offered in comments on the proposed regulations. 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
111
-------
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 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 value 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 of 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 slightly 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.
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 given plant was conducted responsibly and in such a
way that resulting measurements can be considered statistically
112
-------
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 assuming a lognormal
distribution for daily measurements, the plot of the cumulative
distribution of logarithms of daily effluent concentration data 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. To
allow for this variability, performance standards must necessarily be
set above the plant's long-term average performance. However,
guideline limitations must be established at a level low enough to
ensure adequate control. Establishing effluent guidelines that
balance these factors means that occasional, infrequent instances of
non-compliance are statistically predictable at well-operated and
maintained treatment facilities. 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:
ln{P/A) = S'(Z -SV2)
where
A.
"In" represents the
numerical quantity.
natural logarithm (base e) of
113
-------
B. S' is the estimated standard deviation of the logarithms of
pollutant level measurements. In the calculations which
follow, S1 is computed by the statistical procedure known as
the "method of moments".
C. 2 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
ensure that a plant is functioning properly.
The value of 1 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. Use of this percentile
statistically predicts one incident of non-compliance for every TOO
samples for a plant in normal operation. Many plants in this industry
are required by their NPDES permits to self-monitor once per week. At
this frequency, there will be 260 samples analyzed over the 5 year
life of the permit. The use of the 99th percentile to establish daily
maximum limitations statistically predicts 2 to 3 incidents of non-
compliance per pollutant in 5 years. This percentile has been used to
establish daily maximum limitations in all other guidelines proposed
or promulgated, and has been used for daily maximum limitations in
Inorganic Chemicals manufacturing.
A. 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'S1 are the
mean and standard deviation of Y=ln(X) respectively, then the
probability is k percent that an individual Y will not exceed A'+ZS',
where Z is the k-th percentile of the standard normal distribution,
e.g. 2=2.33 is the 99th percentile of the standard normal
distribution. It follows that A'+ZS1 is the natural logarithm of the
k-th percentile of X and that the probability is k percent that X will
not exceed a performance standard P=exp(A'+S1(S'/2)). The variability
factor VF, is obtained by dividing P by A, hence,
VF * P/A = exp(S'(S'/2)), and
ln(VF) = ln(P/A) = S'(Z - SV2)
114
-------
Ul
30.0
20.0
H
f-4
B
2.0
1.0
0.01 0.1 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99
99.9
PERCENTAGE
Figure 8-1. Cumulative distribution of daily concentrations of mercury in treated
effluent from plant #251.
-------
o\
0.30
0.20
^ 0.10
S31
c 0.08
g 0.06
B 0.04
8 0.03
fi
R
§ 0.02
n
0.01
—
-::
=
~ = ! :
.. .
i:: —
", '. Zj
*
•-
E =--:- : •
bbHtttU^^KBitUHUJil
= = •• = = :! =•!: = !:; ! !i!
H 1 1 m-HrfTantTi^ltJi^^S
""ri::: i I-:.1: -..
= ;«::;:: : : : : : : • :
-«::::: :::::;::
£ -
-
_
rt J J t rl-t fi 1 1 T~l ' •.--- —
'
1
1
1 1
i
:• : ::: : : : : : : =
: ;;;•= :•:•:: = =
,----
fi
0.01 0.1 ,512 5 10 20 30 40 50 60 70 80 90 95 98 99
99.9
PERCE2JTAGE
Figure 8-2. Cumulative distribution of daily concentrations of cyanide in treated
effluent from plant #765,
-------
To estimate the VF for a particular set of monitoring data, where the
method of moments is used, S1 is calculated as the square root of
ln(1.0 + (CV)2), where the sample coefficient of variation, CV + S/X,
is the ratio of sample standard deviation to sample average.
B. Example Calculation of Variability Factors From Long-Term
Data
Given the following descriptive statistics for a particular
parameter, as might be found for lead (mg/1) in Appendix A.
No
Min
Avg
Max
cv
128
0.002
0.068
0.100
0.609
Calculate the estimated standard deviation of logarithms
(S1)* = In (1.0 + 0.6092) = 0.315
S * 0.56
Then
ln(P/A) = 0.56(2.33 - 0.56/2) = 1.148
The variability Factor VF is,
VF = P/A = exp(l.148) = 3.15
The performance standard P;
p = A(VF) = A(P/A) = (0.068)(3.15) * 0.21
That is, using the descriptive statistics for a pollutant presented
above and the statistical approach just described, the daily maximum
limitation established for that pollutant in a guideline would be 0.21
mg/1.
The statistical distributions relevant for the anaysis 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 99th 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.
Assumptions Concerning 30-Day Average Pollutant Level Observation
While individual pollution level measurements should be assumed
lognormally distributed, that assumption is not appropriate when
117
-------
analyzing 30-day averages. These averages generally are not
distributed as lognormal quantities. However, for averages of daily
(lognormal) measurements, a statistical 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 Figure 8-4 and 8-5. A straight line plot here on normal
probability paper indicates the validity of this model.
Under these conditions, the 30-day average values (X,, 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 X will not exceed the
performance standard P, where
P = A + 2(S")
The variability factor is
VF = P/A = 1.0 + Z(SVA) and will be estimated by
VF = 1.0 + Z(CV)
Where
A. 1 is a factor derived from 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,
B. CV is the estimated coefficient of variation of the 30-day
averages and is computed by Sx/X, the ratio of standard
error of sample means to overall or grand average of monthly
averages.
Calculation of Variability Factors
A sample calculation of 30-day average variability factor is shown
below. The descriptive statistical data is for zinc (mg/1) from
Appendix A.
No
Min
Avq
Max
CV
30 0.010 0.151 0.815 1.03
VF = 1 + Z(CV) = 1.0 + 1.64(1.03} = 2.7
P = A(VF) = (0.151)(2.7) = 0.41
118
-------
NORMAL DISTRIBUTION
(MODEL DENSITY OF LOGARITHMS OF POLLUTION VALUES)
ln(P) = A' + 2.33(S')
Y = ln(X) = Logarithm (mg/1)
LOGNQPMAL DISTRIBUTION
(MODEL DENSITY OF
POLLUTION VALUES)
X(mg/l)
_ P (Performance Standard)
'—A (Long Term Arithmetic Average)
SAMPLE DISTRIBUTION OF N MEASUREMENTS
(LONG TERM MONITORING DATA.)
f Max^* X(mg/l)
X* (Sample Average)
Note: (a) S' is estimated as (S1)2 = Clnd + CV2)J
S2=
X= ZX/N
Figure 8-3. Statistical distribution for daily pollution measurements.
119
-------
That is, the maximum 30-day average effluent limitation derived from
the descriptive statistics above would be 0.41 mg/1 for that
pollutant.
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, or in other words, the statistics predict up to 3 incidents
of non-compliance with the 30-day average per pollutant over the 5
year (60 month) life of a permit at a well-operated and maintained
treatment facility. This is essentially the same number of predicted
incidents of non-compliance as was predicted for daily maximum
limitations derived using the 99th percentile confidence level (see
above).
In developing the statistical derivatives for monthly averages, in
many cases, a full 30 days of daily average determinations were not
available. In the above example, the monthly average is based on four
data points taken during the month. The standard deviation is then
derived from these "monthly" averages assuming a normal distribution
for the population of averages. Permits are usually written on the
basis of monthly averages obtained from fewer than 30 data points per
month. The use of "monthly" averages rather than 30-day averages
results in a higher variability and, hence, a higher performance
standard than would be attained using 30-day averages based on 30 data
points per month.
Figure 8-6 shows the relationship between the normal probability model
and frequency distribution of 30-day averages.
120
-------
to
3.00
H
2.00
U
W
a
1.00
0.01 0.1 0.5 12 5 10 20 30 40 50 60 70 80 90 95 98 99 99,9
PERCENTAGE
Figure 8-4. Cumulative distribution of 30-day averages of total cyanide in treated
effluent from plant #782.
-------
M
10
H
you
5OO
400
_
_
—
~"
__
— .
._
—
~_
.
"T
.
- - •
-
-- - -
_
• -
—
-
-
—
—
-
_.
--
-
_
—
-
-
-
- : :
-
.. .
.. . . .
-
- - •
- - -
_ -
. . —
—
. . . .
--
• --
- —
—
.
—
—
. . —
. :
•-
-"
-
-
-
•
-
-
-
-
• •-
r - - - -
- - - • - •t
_ . . . .
•
• -
- - - - •
- • • • : -
-- -
..... . . j
... . . i
- . . i
'".: ('-i
- • - - ,
....
. ,
J«
. . . _
• - -
...
. .
• - - -
. . .
• •
t
'
:.:.... K :::::::
'
i . - ...
M - ' • " -
! ' ! • • - . - -
. , .... . . .. _
, . - -
• • • -
. —
-
-
-
^
1— • •
.
!.
/
l-~
—
—
—
— . .
firn
. . _
. . _
-
-
—
-
. _ —
—
—
— .
—
~"
—
0.01 0.1
12 5 10 20 30 40 50 60 70 80 90S 95 98 99
99.9
PERCENTACE
Figure 8-5. Cumulative distribution of 30-day averages of anmonia in treated
effluent from plant #782.
-------
NOEMAL DISTRIBUTION
(MODEL DENSITY OF 30-CAY AVERAGE POLLUTION MEASUREMENTS)
xCmg/1)
— P (Performance Standard)
_ A (long Term Average)
SAMPLE DISTRIBUTION OF M MONTHLY AVERAGES
(LONG TERM MONITORING DATA)
^^ M
Min
:
n
Max
X(mg/l)
(Average of 30-Day Averages)
Note: (a) P/A - 1+1.64 (CV)
CV = S/X
X=E X/M
Figure 8-6. Statistical distributions for 30-day average pollution measurements,
123
-------
-------
SECTION 9
TREATMENT TECHNOLOGY APPLICATIONS
FOR TOXIC POLLUTANT REMOVAL
Selection of Pollutants to be Controlled
In order to determine which toxic pollutants, if any, may require
effluent limitations, the pollutants observed in each subcategory were
evaluated with regard to their 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 priority pollutants which appear at concentration
levels that are readily treatable using available technology.
Group 2 - Other treatable and/or potentially treatable priority
pollutants observed in the subcategory. These include toxic
metals which exist at concentrations below the minimum
treatability limit and above the minimum detection level. The
Group 2 pollutants would be controlled by the same treatment
technology used to control the Group 1 pollutants.
Table 9-1 presents the significant toxic pollutant metals found in
each group. In general, those metals occurring in the first group are
of prime concern and require regulation, while those occurring in the
second group are of somewhat less concern and are not expected to
require regulation. Metals in Group 2 are effectively controlled by
the technologies used to control the metals in Group 1, which are the
two or three dominant metals in the raw waste load and are directly
related to the particular product or process involved.
Application of_ Advance Level Treatment and Control Alternatives
General Design Objectives
Beginning with Section 11 of this document, the selection and
application of toxic pollutant treatment and control technology for
model plant systems for each of the regulated subcategories are
described. Several levels of treatment are indicated. 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
125
-------
TABLE 9-1. PRIORITIZATION OF TOXIC POLLUTANTS POUND IN EACH SUBCATEGORY
SUBCATEGORY
Group 1
Group
Chlorine-diaphragm cell
Copper
Lead
Nickel
Antimony
Arsenic
Cadmium
Chromium
Mercury
Selenium
Thallium
Zinc
Chlorine-mercury cell
Mercury
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Nickel
Silver
Thallium
Zinc
Hydrofluoric Acid
Nickel
Zinc
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Selenium
Thallium
Titanium Dioxide •
Chloride Process
Chromium
Lead
Nickel
Zinc
Titanium Dioxide -
Sulfate Process and
Chloride Ilmenite Process
Chromium
Nickel
Antimony
Arsenic
Cadmium
Copper
Lead
Selenium
Thallium
Zinc
(2)
Group 1 - dominant raw waste pollutants select ed as control parameters
for the effluent limitations or guidance.
Group 2 - secondary raw waste pollutants found less frequently and at
lower concentrations. These pollutants have not been selected
as control parameters but are expected to receive adequate
treatment as a result of controlling the Group 1 pollutants.
126
-------
TABLE 9-1 Continued
SUBCATEGORY
Group 1
Group 2
Aluminum Fluoride
Chrome Pigments
Copper
Nickel
Chromium
Lead
Zinc
Arsenic
Cadmium
Chromium
Mercury
Zinc
Ant imony
Cadmium
Copper
Cyanide
Mercury
Nickel
Hydrogen Cyanide
Sodium Dichromate
Copper Sulfate
Cyanide
Chromium
Nickel
Copper
Nickel
Selenium
None
Copper
Selenium
Silver
Zinc
Antimony
Arsenic
Cadmium
Chromium
Lead
Zinc
Nickel Sulfate
Copper
Nickel
Antimony
Arsenic
Cadmium
Chromium
Lead
Mercury
Selenium
Thallium
Zinc
Sodium Bisulfite
Chromium
Zinc
Antimony
Cadmium
Copper
Lead
Mercury
Nickel
127
-------
for the removal of conventional and nonconventional pollutants, this
is regarded as a secondary design objective.
Pretreatment Technology
Since untreated heavy metal ions will either 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
{such as mercury and lead). Normally the Level 1 and 2 model
treatment processes shown in the following subsections will be
appropriate for pretreatment prior to discharge to a POTW.
Pass-through would occur in the absence of pretreatment when BPT and
BAT treatment would reduce toxic metal concentrations by a greater
percent than is achieved by a POTW,
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.
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.
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
(Section 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 indicated treatment
applications in each subcategory.
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.
128
-------
When established wastewater treatment practices, such as clarification
or filtration, form a part of advanced treatment alternatives, the
specified achievable effuent 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.
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
long-term 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 presented in tabular summaries. By
starting with the estimated achievable long-term averages, the
specific variability factor ratio derived for each pollutant is used
to estimate the maximum 30-day average and 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.
Pollution Control Parameters to be Regulated
are identified as conventional
Conventional Pollutants
Wastewater quality parameters which
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
129
-------
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 of
TSS are specified for both BPCTA and NSPS-based regulations, the
former being largely a function of industry performance and the latter
stemming from treatability estimates with the appropriate
technologies.
Nonconventional Pollutants
The wastewater quality parameters classified as nonconventional
pollutants include the nontoxic metals such as aluminum, boron,
barium, and iron along with chemical oxygen demand (COD), total
residual chlorine, fluoride, ammonia, and nitrate, etc. Of these,
only iron, 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
it is excluded from control in pretreatment regulations because
influent to POTW's is often chlorinated. A similar argument is made
for the control of amminia, that is, POTW's can use ammonia as a
source of essential nutrients. However, since many POTW's are only
capable of about 20 percent ammonia removal, both direct discharge and
pretreatment regulations would specify ammonia limitations.
Similarly, the type of COD found in inorganic chemical industry
discharges may not be 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 calcium fluoride) must be applied to relatively
concentrated wastewater sources. This treatment method achieves
removal levels which at best are still unacceptable for direct
municipal or agricultral 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.
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.
130
-------
The following toxic pollutants have been designated as control
parameters in this point source category:
Cadmium
Chromium (Total)
Copper
Cyanide (amenable to chlorination)
Lead
Mercury
Nickel
Selenium
Zinc
The specific control parameters selected for each subcategory are
presented in the tables entitled "Effluent Limitations" in the
sections of this report dealing with the individual industries. Some
general comments about them are given here.
The most common technology applied in industry for the removal of
chromium from wastewaters 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 an 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 an excess reducing agent in the treated effluent. This
tends to give false low readings for Cr (VI) although in some cases
the opposite may occur as a result of sample preservation and storage
under acidic oxidizing conditions.
Thus, in view of the questionable reliability of the presently
accepted Cr (VI) monitoring procedure, total chromium, Cr (T), is
recommended as the control parameter to be used in the inorganic
chemicals industry. The adequacy of Cr (T) as a control parameter is
predicated on its effectiveness as a surrogate for Cr (VI) control.
Since the concentration of Cr (T) represents the summation of all
forms of chromium normally found in solution or suspension including
Cr (VI), the final concentration of Cr (T) in a treated effluent is
dependent on the effectiveness of both the reduction and the alkaline
precipitation steps. In this way, the use of Cr (T) as the control
parameter assures that adequate removal of Cr (VI) is being achieved
as a direct consequence of the treatment technology required.
131
-------
-------
SECTION 10
COSTS OF TREATMENT AND CONTROL SYSTEMS
Introduction
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.
Plant visits revealed very few treatment plants serving a single
product manufacturing line, therefore, 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.
Based on the level of supporting details and scope definition, the
accuracy range of the cost estimates is expected to be minus 15 to
plus 25 percent.
Actual costs incurred by individual plants may be more or less than
the presented model plant costs. The major causes of variability ares
Wastewater treatment combined with the treatment of other product
effluents.
Site dependent conditions, as reflected in piping lengths,
climate, land availability, water and power supply and the
location of the points of final discharge and solids disposal.
Material (reagent) costs, due to variation in availability and
distance from the source.
Flow rate of wastewater to be treated.
The construction costs are expressed in mid-1978 dollars. The
investment costs and the annual costs given in the preamble to the
regulation are expressed in 1981 dollars, and were updated from 1978
dollars using the Department of Commerce Composite Index for
Construction Costs.
133
-------
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, wastewater 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.
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. Costs are based on similar industrial
installations or engineering estimates. Cost estimates have been
developed from either current costs for similar plants or from general
cost estimates.
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) and waste-
water flow rate {i.e., cubic meters/day). Available data indicate
that both pollutant loads and flow rates can vary significantly among
plants manufacturing the same product.
Definition of Levels of Treatment and Control Cost Development
For the purpose of establishing the base level treatment costs, each
industry is assumed to be practicing Best Practicable Control
Technology Currently Available (BPT), for the EPA pollutants
(conventional and nonconventional, as well as some of the toxic metal
pollutants) which are specified for each subcategory. The investment
costs and annual costs of such BPT systems are shown in this report as
either the Base Level, Level 1, or BPT costs. This level of treatment
may also provide incidental removal of additional toxic pollutants not
previously specified in the regulations.
The advanced treatment level (BAT) is 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 (BPT). For example, for Level 2 (BAT)
134
-------
treatment, the incremental cost as given in the table is directly
added to Base Level 1 or BPT cost to obtain the total cost of the
treatment system. The wastewater treatment flow diagrams for the
advanced treatment levels, as given in this report, also include the
flow diagram for Level 1 (BPT) treatment.
Treatment and Disposal Rationale Applied to Cost Development
The following assumptions are employed in the cost development;
A. Noncontact cooling water generally is excluded from treatment
(and treatment costs) provided that no pollutants are introduced.
Water treatment, cooling tower and boiler blowdown discharges are
not considered process wastewater unless such flows contain
significant amounts of pollutants.
B
C,
D-
E.
F,
Sanitary sewage flow is excluded.
The plants are assumed to operate 24-hours per day,
year, except where otherwise noted.
350 days a
G.
Manufacturing plants are assumed to be single product plants.
The inorganic chemical industry extensively uses in-plant control
techniques such as in-process abatement measures, housekeeping
practices, and recycling of process wastewaters 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.
Excluded from the estimates are any costs associated with
environmental permits, reports or hearings required by regulatory
agencies.
Expression of Costs
The estimated costs for all treatment systems are expressed in
mid-1978 dollars to construct appropriate facilities for each single
product manufacturing subcategory at various production rates. Total
costs are given for the BPT and NSPS systems while Incremental costs
are given for a BAT system.
Where a single product plant produces more than one waste stream
requiring treatment, the respective investment and annual costs are
the combined costs of all treatment.
Total annual costs per metric ton of product are shown in the
summaries for each product subcategory.
135
-------
A. Direct Investment Costs for Land and Facilities
Types of direct investment costs for waste treatment facilities and
criteria for estimating major components of the model plants are
contained in the following subsections;
1. Site Development Costs - These include clearing the site,
all earthwork and site improvements. The lagoon costs are
based on the excavation and backfill required to construct
multiple rectangular lagoons with common dikes to permit
alternate dewatering for sludge removal by the clamshell
method. They are also based on reasonably level sites being
available, consisting of sandy loam with high clay content,
and no large rocks or rock formations. Lagoons are unlined,
excepting where contents are highly acidic. Where lining is
required, hypalon or clay is used.
Site improvements include local drainage, fencing, and
roads. Road costs are based on graded and graveled service
roads only within the boundaries of the plant and not for
access. Perimeter fencing is supplied for the lagoons and
for the sludge disposal site.
Equipment
equipment
elsewhere
Costs - This is the installed cost of all
except the monitoring system ( considered
Depending upon the method of treatment,
equipment for wastewater treatment consists of a combination
of items such as pumps, aerators, chemical feed systems,
agitators, flocculant feed systems, tanks, clarifiers,
thickeners, filters, etc. Costs for these items were
obtained from vendors' verbal quotations and were based on
contractors1 experience with procurement of similar items.
Enclosures are provided for critical equipment and controls.
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 day's 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.
Added to the cost of the equipment itself, is the
corresponding installation labor, as well as the material
and labor costs for concrete, structural steel, piping,
instrumentation, and electrical work. The labor costs
include all pro-ratable elements of "indirect" costs such as
fringe benefits, payroll insurance and taxes,
equipment, temporary construction facilities
etc. The hours and unit costs for the labor
construction
, field staff,
are based on
136
-------
Los Angeles
craft labor.
or Gulf Coast type productivity using union
In some subcategories, a portion of the wastewater is
returned to the process from an intermediate treatment step.
In such cases, the estimated investment cost of the reuse
pumps are included as part of the equipment cost. However,
the return piping, accessories, operating and maintenance
costs are considered as water supply costs.
3. Monitoring Equipment - In this report, it is assumed that
flow and pH monitoring equipment will be installed at the
treated effluent discharge point. It will consist of an
indicating sensor and recorder, alarms and controls and an
automatic sampler.
4. 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 wastewater. 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).
B. Investment Costs for Supporting Services
1. Contractor's Overhead and Profit - The construction
contractor's fixed or "overhead" expenses and profit are
estimated as fifteen percent of the installed plant cost.
2. Engineering - This includes the design and inspection
services 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 ground water
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'
137
-------
3.
Soils and groundwater investigation
Laboratory and pilot process work
Engineering design and specifications
Inspection and engineering support
during construction
Operation and maintenance manual
1 to 2%
2 to 4%
7 to 12%
2 to 3%
1 to 2%
From these totals of 14 percent to 25 percent, a midvalue 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. These costs include, in addition to the
professional service hours, the costs for expenses such as
telephone, reproductions, computer services, and travel
fees.
Contingency _ This is an allowance of 10 percent applied to
the estimated total investment cost, excluding land, based
on the status of engineering, design and specifications,
quality of prices used, and the anticipated jobsite
conditions. This covers design development, {but not
scope), errors and omissions, impact of late deliveries and
unusally adverse weather conditions, variations in labor
productivity and other unforeseen difficulties during
construction.
C. Operation and Maintenance Costs
Annual operation and maintenance costs are described and
calculated as follows:
1. 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
wastewater treatment systems, adjustments 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.
2. Engergy 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
138
-------
adjustments are made to suit the production schedule.
cost per horsepower year is computed as follows:
The
Cy
1.1 (0.7457 HP X Hr X Ckw)/(E X P)
1)
Where:
Cy
HP
E
P
Cost per year
Total horsepower rating of motor (1 hp
kw)
Efficiency factor (0.9)
Power factor (1.00)
0.74557
Hr = Annual operating hours (350 X 24 = 8400)
Ckw = Cost per kilowatt-hour of electricity ($0.040)
3.
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 of active ingredient
percentages as follows:
Hydrated Lime (Calcium
Hydroxide)
Quicklime
Ground limestone
Soda Ash (58% Bulk)
Caustic Soda (58% NaOH)
Sodium Sulfide (60-62%)
Sulfuric Acid
Hydrochloric Acid (32%)
Bulk
Bag
Bulk
$ 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
139
-------
Aluminum Sulfate (56% Alumina)
Flocculant (Polymer)
Sulfur Dioxide (Ton Containers
Chlorine (Ton Containers)
Sodium Bisulfide (72-74%)
Ferrous Sulfate
Diatomaceous Earth
Activated Carbon
$250/metric ton
$2.00/kg
$335/metric ton
$220/metric ton
$385/metric ton
$ 70/metric ton
$0.30/kg
$2.00/kg
Maintenance - The annual cost of maintenance is estimated as
10 percent of the investment cost, excluding 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 and type
of waste, and the choice of either on-site disposal or
contract hauling which depends on the size .of the disposal
operation and transport distances. Off-site hauling and
disposal costs are taken as $13.00 per cubic meter ($10.00
per cubic yard) for bulk hauling, with appropriate increases
for small quantities in steel containers. For on-site
disposal from lagoons, a clamshell at $600.00 and front end
loader at $300.00 per disposal day are used. For very large
sludge quantities, lower unit costs have been assumed. The
computed sludge quantities are spread on land valued at
$12,000 per acre.
Monitoring, analysis, and reporting - The manpower
requirements covered by the annual labor and supervision
costs include those activities associated with the operation
and maintenance of monitoring instruments, recorders, and
automatic samplers as well as the taking of periodic grab
samples. Additional costs for analytical laboratory
services have been estimated for each subcategory assuming
that sampling takes place three times a week at the point of
discharge and that an analytical cost of $20.00 per
constituent is incurred. Approximately 10 percent of the
total analytical cost has been added for quality control and
water supply samples. Unless otherwise stated, continuous
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
140
-------
support. Monitoring costs for periodic batch treatments are
reduced in proportion to the number of days per year when
discharges occur.
D. Amortization
Annual depreciation and capital costs are computed as follows:
CA = B (r(1 + r)n)/((l + r)n-») (2)
Where:
CA
B
n
Annual Cost
Initial amount invested excluding cost of land
Annual interest rate (assumed 10 %)
Useful life in years
The multiplier for B in equation (2) is often referred to as the
capital recovery factor, and is 0.1627 for the assumed overall
useful life of 10 years. No residual or salvage value is
assumed.
Items Not Included in Cost Estimates
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, access 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. It was also
assumed that all required utilities are provided at the edge of
the plant site and the existing plant's capacities are capable of
supplying the requirements. RCRA costs have not been included.
RCRA costs are considered in the Economic Impact Analysis o£
Pollution Control Technologies for Segments of. the Inorganic
Chemicals Manufacturing Industry, EPA 440/2-81-023, which were
developed in part from information in "Contractor Report on RCRA
ISS Compliance Costs for Selected Inorganic Chemicals
Industries."
141
-------
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, pipe lines, etc., necessary to deliver raw wastewater to
the treatment plant or to deliver the treated effluent to the
point of discharge are not included in the cost estimates.
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.
Cost Estimates For Each Subcateqory
Estimated costs for the wastewater 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:
1. Investment
2. Annual operation and maintenance
3. 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 (BPT or NSPS) is
expressed as the total cost of the treatment system. The other level
(BAT) represents 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 investment cost,
but not in the total annual costs.
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
specifications together with the basic assumptions on cost
in this section, form the basis of the 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
These
estimating
142
-------
plant specific variables, additional cost elements which may add to
the baseline costs are then considered on a case-by-case basis.
143
-------
-------
SECTION 11
CHLOR-ALKALI INDUSTRY
Industry Profile
General Description
Chlorine and its co-product caustic soda (alkali) are used in large
quantities in the production of plastics, organic and inorganic
chemicals, in the pulp and paper industry, in water and wastewater
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, mercury
cell and diaphragm cell.
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 are available from fully operating
facilities.
Subcategorization
The factor chosen for the primary subcategorization of the inorganic
chemicals point source category was dominant product (see Section 4).
Other factors considered for subcategorization include: raw materials
used, manufacturing process employed, geographical location, size and
age of equipment and facility involved, non-water-quality aspects of
waste characteristics, water pollution control technology, treatment
costs, energy requirements and solid waste disposal. The chlor-alkali
subcategory was further subdivided on the basis of differences in cell
design and in the quantity and quality of wastewater generated.
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
wastewater generated. A principal difference is the presence of
mercury as a contaminant in the wastewaters 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 wastewater is produced in
the caustic evaporation process. Such water is not produced in
mercury cell plants. The quantity of wastewater generated from the
diaphragm cell plants may be more than four times that of the mercury
145
-------
cell plants for the same chlorine production capacity. Based on the
quantity and characteristics of the wastewater, further
subcategorization is justified.
Mercury Cell Process Industry Profile
General Description
Approximately 30 percent of the U.S. production of chlorine is by
mercury cell plants. In 1978 of 27 known plants, 308 data were
available for 15. Table 11-1 presents a summary profile of the
subcategory. Table 11-2 presents the status of discharge regulations
for mercury cell chlorine plants prior to promulgation of these
regulations. Control of pH in the 6.0 to 9.0 range was also included
in those regulations.
General Process Description
A. 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.
Precipitated hydroxides and carbonates are then settled 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, the treated brine is filtered 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.
B. 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 concurrently 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,
NaCKaq) + Hg - Cl, + 2 Na(Hg)
146
-------
TABLE 11-1. SUBCA3EGORY PROFILE .DATA .SUMMARY
SUBCfllEGORY
CHLORINE MERCURY
Ibtal subcategory capacity rate
Ototal subcategory production rate
Number of plants in this subcategory
308 Data on file for
With total capacity of
With total production of
Representing capacity
Representing production
Plant production range:
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
3,550,000 kkg/year
2,750,000 kkg/year
27
15
1,280,000 kkg/year
1,090,000 kkg/year
36 percent
40 percent
19,100 kkg/year
198,000 kkg/year
77,900 kkg/year
70,400 kkg/year
75 percent
2 years
26 years
4 cubic meters/day
2,100 cubic meters/day
< 1 cubic meters/kkg
11 cubic meters/kkg
Sources of data are Stanford Research Institute, Directory of Chemical
producers, U.S.A., 1977, U.S. Department of Ccmnerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Eooncmic 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.
147
-------
TABLE 11-2 STATUS OF REGULATIONS - EFFLUENT LIMITATION GUIDELINES
SUBCATEGORY
SUBPART
CHLORINE MERCURY CELL
F (40 CFR 415.60, 3/12/74)
STANDARDS
Product Parameters
Process
BPCTCA
Max.t1) Avg.
(kg/kkg) (kg/kkg)
(2)
BATEA NSPS
Max. Avg. Max* Avg.
(kg/kkq) (kg/kkg) (kg/kkq) (kg/kkg)
Mercury
Cell
Process TSS
Hg
0.64
0.32
No discharge
of pwwpI3)
No discharge
0.00028 0.00014 __ of pwwp
0.64 0.32
0.00014 0.00007
Section 415.63 was remanded and reserved (41 FR 51601, November 23, 1976).
(1) Max. = Maximum of any one day.
(2) Avg. = Average of daily values for thirty consecutive days shall not exceed.
(3) pwwp = Process wastewater pollutant*
148
-------
The amalgam from the elctrolyzer 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.
C. 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 nonbrine system 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.
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.
Water Use And Wastewater Source Characteristics
Water Use
Water is used at mercury cell plants for noncontact cooling, tailgas
scrubbing, cell washing, equipment maintenance, floor washings and in
the decomposition of sodium-mercury amalgam in the denuder to produce
sodium hydroxide. Because most brine systems at mercury cell plants
are closed systems, water use in the brine system is minimal. The
total water usage at plants was found to range from 7.6 to 204 cubic
meters per metric ton (1800 to 49,000 gallons per short ton), with
noncontact cooling water, which is not covered by this effluent
guideline, comprising approximately 70 percent of the total.
149
-------
H
O
SPKHT BKINt
s »
CHLOf"*
1
BftL * ^S™* »™™ I" AMAU*H •» .^SSL, K- DWIKUUW.
CE"- . 1 (OECOBPOSBWJ COOLER
8YSTE* COO^IHE^HATER HVOJKMtl
' ^-^-™».i.S" (AWP J° ATHOSPHKRE
Dt SODIUM ItOMCOWfACT F1LTKB) *oB. USE
>K SOUITJESM f?fttM.IHG ^
t
COtUUER COHDEHSfttE
1 ff fR^BiinnTiH". u*^Vp*i H*70 Jffr51^
faffetUK UA¥fiD 4n^M
4
Tn tikfiTB - BACKMASU H£CVCLE
AND HECKCLE ** '*
HOMCOWACT
(COWTACT*)
COOHWG HATER c-JIPHBIP
auuuniv. BBCYCLEO ^f-
1 t ACID — MEIVHiBy — ^^ —
i f
S
2 i
E 11
R f f
• ' REFRIGERATIOH
V HEAK SULFURIC SYSTEM
sn. «™tllM 1 I BKCVCLB COUL1HU
FIITEH uruHSi"^1 — |P*C«*fiH*e ) •» TOMER
HCKMAS0 90LUTIOH ^
BACKhUSU A 1
piLven CAUSTIC ILIMB) | f
{ LAHOFILl. |
r i s
— 1 1 SOLIDS hrTO HASTE B
* — HKPOCBU)RITB_-.
" SOLUTION
LIQUID _,, _,._- TO USE, SALES,
CBLORIMB ~~^{at IKE °" W*51*
PRODUCT ""
f HONCOUTACT
TO UASTgOR REUSE COOLING
HATER
USED AT SOME PLAMTS ONLY
Figiare 11-1. General process diagram for production of chlorine/caustic by
mercury cells.
-------
Waste Sources
The following waste sources are or can be contaminated with
mercury and would therefore require treatment if discharged.
A. 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 tungsten. Calcium and iron
are removed as hydroxides. _Bjrine mud is the. majpr_pprtion 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 recycled to the
process as makeup water for the brine. In the mercury cell
process, only 16 percent of the 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.
B. 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.
C. 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 wastewater 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.
D. Spent Sulfuric Acid
Concentrated sulfuric acid is used in the dryer to remove the
residual water from the chlorine gas after the first stage of
cooling. In most cases, the acid is used until a constant
151
-------
concentration of 50-70 percent is reached. The spend acids can
be regenerated for reuse, used for pH control in a nonbrine
treatment system, or sold.
chlorine gas from the
Tail Gas Scrubber Liquid
The tail gas containing the uncondensed ^
liquefaction stage, alcmg wTttr~~gome" air
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, producting the corresponding hypochlorite solution.
The hypochlorite can be used in another process on site, sold,
discharged to treatment or decomposed before discharge or
treatment. The amount of tail gas scrubber water varies from
0.04 to 0.58 cubic meter per metric ton of chlorine.
H
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,
Wastewater 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 merjcury recovery after which it can be
returned to the denuder. Data on the volume of this waste stream
are not available.
Summary of Wastewater 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 (mVkkg} 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 mVkkg, although flows as high as 6.3 mVkkg do
exist.
152
-------
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 wastewater control and treatment.
Screening Program
Plant #299 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 wastewater. The combined wastewater is sent to
JLSurge pond. The effluent from the surge pond is mixed with sodium
%8jS$jlM&lfi and sent to a settling pond. The overflow from the pond is4
pfi adjusted, filtered (in a filter press) and passed through activated
carbon towers before discharge. In the sampling program wastewater
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.
Verification
Four more plants (#747, #167, #106 and #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 #747, 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
wastewater. 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 wastewaters are treated with sodium sulfide (Na2S) and
filtered. Solids are retorted for mercury recovery and the filtrate
is mixed with the other process wastewaters and the pH adjusted before
discharge. A flow diagram of the manufacturing process, including the
wastewater treatment facility, is given in Figure 11-3.
At Plant #167, the wastewater 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
c1arifier and the underflow from the clarifler 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 #167,
including the sampling locations.
153
-------
TABLE 11-3. SUMMARY OF WASTE WATER PLOT DATA FOR CHLORINE
MERCURY rar.T. PLANTS
SUBCATEQORY
CHLORINE MERCURY CKT.L
Plant
Nurber
Waste Water Flow
(m Akg Chlorine)
907
299
343
106
131
589
898
741
553
769
Average of 13 plants
0.36
1.6
1.6
0.67
1,7
5.8
0.98
0,51
1.0
6,3
2,1
154
-------
Ul
SULFIPE
AMITItM
HI
(121
e_ fljncE | seauaa
* KW *- FOND
(Ml £ } H |
HUMEHD
e_|_-J "W« ^ «CTWMH> ^
*
12
4,,
1 1 ' WHBt
1
L _».*«. IK TO
LAHOflU TOJBMJSHH
*
*"* WINE _^ BUM _ff"HE . _,. mlSE ""r^f
"" SMMUtlOU "^ BUBS-H^TIOH »" **"* '*"*
ft
tbOH Hft-p^
Hacuon -,*- \
StOMGe.
•s
V*
i+\ "2°
^^^ (wW-WHjO
• I , ,
>y» "SS^3 .
^ AW
UOlErKTKM
»^ rauriaiT)CN — „»» n_
i
unun
KOttGttl
i
SUCK nw —HJO ID RIVEH- — •»
raw.
DISQMIGE
(131
8
In I \ were aanplad in
, ottw* during
^^wisrero
"** HlVB*
IS
___^»J «
srowce
Figure 11-2. General process flow diagram at plant #299 showing the sampling points.
Chlorine/caustic (irercury cell) manufacture.
-------
TABLE 11-4. POLLUTANT OM23NTKATIONS AND LOADS AT PLANT # 299
SUBCATEGQRY
CHLORINE (MERCURY CELL)
Stream Stream
Number Description
TSS
(rog/1) (kg/kkg)
Mercury
(mg/1) (kgAkg)
Screening Phase:* *
1 Cell Waste
Mercury Treatment
Effluent
12 0.016
5.0 0.0070
Tail Gas
Scrubber
Verification Ptese:
.(2)
Mercury Treatment
Influent
Mercury Treatment
NA
91
NA
0.13
0.15 0.0002
0.029 0.00004
0.11 NA
5,9
NA = Not available.
(1) - Data hased on one 72-hour composite sample of each stream.
(2) = Data based on three 24-hour composite samples of each stream.
0.080
3
4
5
Effluent
Cell Vfeste
Brine Mud
Tail Gas Scrubber
18
120
13,000
180
0.026
0.17
NA
0.022
0.20
11.
0.54
0.17
0.0003
0.015
NA
0.00002
156
-------
TABLE 11-5. POLLUTANT CONCENTKVTIONS AND LOADS AT VERIFICATION PLANTS
SUBCATEGORY CHLORINE (MERCURY CELL)
Stream Stream
Number Description
Plant
1
2
3
4
5
6
7
Plant
5
6
7
8
9
Plant
1
2
3
4
5
6
7
8
Plant
1
2
4
747
Cell Waste
Treated Waste
Acid Input
Acid Output
Dechlor System
Cl2 Condensate
Tail Gas Scrubber
167
All C12 Wastes
Cell Wash
Brine Process
Treated Waste
Clarifier
Underflow 5
317
Cell Waste
Brine Mud
Filtrate
Tank Car Wash
Collection
Tank 21
Treated
Effluent
Deionizer
Effluent
N-C Cooling
Final Effluent
106
Cell Wash
Treated Cell
Wash
Final Effluent
TSS
Ung/J.)
700
60
NA
NA
9.0
2.0
NA
560
57
4.0
2.0
r900
45
520
18
,000
110
18
16
18
79
20
2,0
Uog/Jckg)
1.6 x 10~*
1.4 x 10"^
NA
NA
0.0037
2.7 x 10
NA
1,9 ,
5.7 x 10 ->
7.1 x 10,
1.3 x 10"^
4.0
NA
NA
NA
8.6
'-2
4.4 x 10
5.2 x 10"3
2.2
2.4
Mercury
(mg/i)
18
0.10
0.023
0.0030
0.035
0.27
0.039
3.8
0.72
0.0050
0.32
10.4
14
34
0.033
123
0.10
0.0010
0.0010
0.0020
3.9
0.015
<0. 00050
Ucg/Kkg)
4.3 x 10 j?
2.3 x 10 J?
3.5 x 10~I
7.2 x 10~'
1.5 x 10 J?
1.8 x 10 "2
8.0 x 10
1.3 x 10^
6.7 x 10 J?
9.0 x 10 ,
1.8 x 10
-5
8.7 x 10
NA
NA
NA
-2
5.0 x 10
-5
4.3 x 10
7
2.9 x 10"
1.4 x l(f;
3.6 x 10
NA
NA - Not available.
(1) = Data based on the average of three 24-hour composites.
157
-------
At Plant #317, 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 wastewaters, which includes the brine purge, cell
room liquid wastes and plant area wash water. This is then reacted
with sodium hydrosulfide to precipitate the mercury as mercury sulfide
and then filtered. The solids are sent to a mercury recovery unit and
the filtrate is sent to a holding tank. The effluent from the holding
tank is mixed with de-ionizer waste and noncontact cooling water
before discharge. The process flow diagram showing the waste streams
sampled is given in Figure 11-5.
At Plant #106, mercury-bearing wastes are segregated from other
wastewaters 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 to 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 second
lagoons.
Descriptions of Plants Not Sampled
At Plant #589, the wastewater going to the mercury treatment system
consists of cell room washdown, brine filter backwash, leaks, spills,
cleanup water, and hydrogen cooling condensate. The wastewaters are
reacted with hydrochloric acid and sodium bisulfide and then sent to a
settling basin where mercury sulfide precipitates. The overflow is
passed through a series of effluent filters before discharge.
At Plant #343, the cell room wash water, brine purification sludge,
and chlorine cooling condensate are combined and sent to a pond. The
suspended solids settle in the pond and are dredged out once a year.
The dredged sludge is "Chem Fixed" and disposed of in an appropriate
landfill. The overflow from the pond is reacted with Na2S and the
reacted solution is sent to a clarifier. The clarifier underflow,
consisting mainly of mercury sulfide, is returned to the pond. The
clarifier overflow is discharged.
All contact wastewater 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 amount of mercury 'in the effluent
and for better waste control. Molecular sieves have been installed on
158
-------
£
HVOCM.ORI TE
SOLUTION
^ MINE
& BRINE
t3 SJSUH
-I!
DECHLORO.
fUtlHUV
"3ST"
lti«
IRINE
COHDS.
PROCESS UAiTE i
Att » ,
cii
1*9
A,
Cl 'I*
"-'t. * COMPRESSION
DRVIHG
HOUEFACTIOH
A
cow. T
tit HjSO^ UASTE
TO PONDS
PIMUF.
t
r 4
*r
SULFIOE
OTHER PRODUCT
UASTE WATER
SWKJPG POiWS
SOLIDS 10.^ f
LAHOftLL I
Figure 11-3. General process flow diagram at plant #747 showing the sampling points,
Chlorine/caustic (mercury cell) manufacture.
-------
MttCOMTACT
COOLING UATE* U2°
Mint ptocesi IMIM
10 pH AOJUSI-
H£HI ANft FINAL
OIUUAKCf
IMOOK
+ (T\
n
ION
iUHAHGE
ttfitHP
CA1ION
S**« f tiTEl
ClMtFtU
^
1
^
UNO
fetl.
M
SAMPLING POINTS
UNOFIU
Figure 11-4. General process flow diagram at plant #167 showing the sanpling points.
Chlorine/caustic (mercury cell) manufacture.
-------
sou
SALT 1 I
(XV GAS
U3GEMD
streams eaopled.
DISCIULRGB
Figure 11-5. General process flow diagram at plant #317 showing the sampling points
Chlorine/caustic (mercury cell) manufacture.
-------
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 wastewater 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
wastewater is delivered to large holding tanks, from which it may be
discharged or returned to treatment, depending on its mercury content.
Filter cake, resulting from the filtration of the waste prior to the
coal absorption step, is retorted for mercury recovery.
Waste solids at this facility, including mercury treatment sludges and
brine muds, are deposited in an on-site disposal area. Chlorine
discharges are essentially eliminated by three significant waste
management practices: the chlorine condensate is collected and
returned to the brine system, tail gas scrubbing effluents are used in
the manufacture of another product, and spent sulfuric acid from
chlorine drying is dechlorinated in an air stripper and shipped
off-site for the manufacture of another product. Gases from the air
stripper are returned to the chlorine purification header.
At Plant #324, the barometric condenser on the brine dechlorination
was replaced with an indirect cooler, resulting in a reduction of
chlorinated wastewater. 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
wastewaters which include the cell room wash water, caustic filter
backwash, and brine leaks. The combined wastewater is mixed with
hydrogen processing wastewater, 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 wastewater 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 #416, the cell room wastes are used for bleach manufacture.
The wastewater streams from the chlorine/caustic plant are sent to an
acTjacent paper company.
At Plant #784, the wastewater, 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.
162
-------
The wastes are segregated at Plant #674. The clarification pond is
used for waste streams containing suspended solids. The streams going
to the pond include brine purification muds and spent chlorinated
lime. The mercury-contaminated wastewaters are treated separately.
These include the brine saturation waste, brine filter backwash, cell
room sumps, and tank car washes. The combined mercury-laden
wastewater 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 another pond and the pond overflow is passed through a carbon
adsorption column before final discharge. A part of the treated
effluent is reinjected into the brine well.
At Plant #012, the brine treatment area is paved to trap all spills,
leaks, and rain runoff from that area. The contaminated wastewaters
from the plant are re-injected into the brine wells to keep the
hydraulic balance and maintain pressure in the salt deposits.
Summary of the Toxic Pollutant Data
Presented below are the toxic pollutants
during screening and verification.
found in the raw wastes
Because several waste streams usually contribute to the total raw
waste at mercury cell plants, a calculation was often necessary to
determine the pollutant concentrations that would exist in the streams
before they were mixed prior to treatment. An example of this
calculation is the "mixing" of the following hypothetical streams:
Stream A: 100 gallons per minute, 15 mg/1
Stream B: 10 gallons per minute, 60 mg/1
(Flow x concentration) - (Flow x concentration)
Total Flow
= concentration of mixed streams
- (100 qpm) (15 mq/1 + (10 qpm) (60 mq/1) = 19 mg/1
110 gpm
The maximum raw waste concentrations observed during any single
24-hour sampling period were:
163
-------
Maximum Raw Waste Concentrations Observed
Ug/1)
Pollutant
Screening
Plant
(#299)
Verification
Plants
#299, #747, #167,
#206, #317)
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Thallium
Zinc
250
10
1
8
350
1
150
100
1
140
230
770
400
790
180
2,300
1 ,900
180,000
2,400
870
440
34,000
Section- 5 of this report describes the methodology of the screening
and verification sampling program. In the chlorine mercury cell
industry, a total of 18 days of sampling were conducted at Plants
#299, #747, #167, #317 and #106. Thirty-two different sampling points
were involved covering various raw waste streams and the treated
effluents at these plants. The evaluation of toxic metal content of
these process related waste streams was based on 949 analytical data
points. The screening for toxic organic pollutants at Plants #299 and
#167 generated an additional 490 analytical data points. The daily
raw waste loads were calculatd from the waste stream flow rates
measured or estimated at the time of sampling and the measured
pollutant concentration.
The daily loading is determined by:
Daily loading (as kg of pollutant =
per day)
C)(Q)
1000
where:
C is the concentration of the pollutant expressed
of mg/1 (Note: kg/m3 = 1000 mg/1), and
in units
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:
164
-------
Unit loading (as
kkg of chlorine)
kg of pollutant per
(CHQ)
1000P
where C and Q are the same as described above, and P is the chlorine
production rate expressed in units of kkg/day (kkg is 1000 kg, a
metric ton, which is equal to 2205 Ibs.)
The minimum, average, and maximum values are based on data from those
plants where the particular pollutant was found at a concentration
greater than the analytical detection limits and considered a
"significant concentration". The term "significant concentration"
means an observed concentration in any 24 or 72-hour composite raw
waste sample that is above the analytical detection limit, and
potentially treatable.
In Table 11-6, the toxic pollutant raw waste data are presented as the
average daily concentrations and the unit loadings found at the
individual plants. These averages were derived by averaging the
concentrations and loads based on three 24-hour composite samples from
each plant.
In Table 11-7 daily loadings (in kg/day) and unit loadings in (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:
Pollutant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Thallium
Zinc
Pollution Abatement Options
Toxic Pollutants of Concern
Raw Waste load
(kg/year)
1,400
1,000
210
360
960
880
44,000
820
850
770
7,200
of
Mercury is the major toxic pollutant of concern in the production
chlorine by the mercury cell process. Other toxic metals often found
in significant concentrations in raw wastes include arsenic, antimony,
165
-------
cadmium, chromium, copper, lead, nickel, silver, thallium and zinc.
Sources of these metals are assumed to be impurities in the raw salt
or brine and corrosion products from the reaction between chlorine and
process equipment materials of construction. No toxic organics were
found at treatable levels.
Prevailing Control and Treatment Practices
Specific control and treatment practices at 14 plants were described
above. All known mercury cell plants practice treatment of
mercury-bearing wastes, but control practices such as recycling of
brine mud filtrate or pond overflow, chlorine condensates, hydrogen
condensates and caustic filter backwash, and solids handling vary from
plant to plant. Although all known treatment facilities precipitate
mercury and separate the solids formed by clarification and/or
filtration, sampling data has shown that some treatment systems,
including those with more advanced technologies such as adsorption or
ion exchange, are not operating efficiently.
Process Modifications and Technology Transfer Options
The following process modifications are being practiced at one or more
mercury cell plants and can significantly reduce pollutant loads
discharged.
A. 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.
B. Liquefaction of Chlorine
Utilization of high pressure and refrigeration for chlorine
recovery will reduce the chlorine content of tail gases.
C. 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.
D. Mercury Emissions
Hydrogen gas produced in the denuder can be. refrigerated and
passed through treated carbon or molecular sieves to remove the
mercury escaping with 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.
166
-------
TABLE 11-6. AVERAGE TOXIC POLLUTANT RAW WASTE CCMCENTRATICNS AND
LOADS AT VERIFICATICN PLANTS (1)
\hg/kkgy
SUBCATEGORY
Pollutant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Thallium
Zinc
CHLORINE (MERCURY CELT.)
Plant 1
299
0.48
0.00077
0.23
0.00037
0.010
0.000016
0.063
0.00010
0.30
0.00048
0.060
0.000096
5.9
0.0094
*
*
0.18
0.00029
0.27
0.00043
747
0.11
0.000076
0.030
0.000021
0.020
0.000014
0.10
0.000069
0.38
0.00026
0.16
0.00011
18
0.012
0.093
0.000064
0.047
0.000032
0.022
0.000015
0.69
0.00048
167
*
0.33
0.0018
*
0.12
0.00067
0.075
0.00042
0.072
0.00040
3,8
0.021
0.060
0.00034
*
*
0.17
0.00095
317
*
0.10
0.000051
0.46
0.00023
0.080
0.000041
1.2
0.00061
1.4
0.00071
123
0.063
1.4
0.00071
0.11
0.000056
*
20
0.010
106
0.49
0.00033—
*
- ..
0.031
0.000021 —
0.013
0. 0000087- -
0.12
0.000080
0.33
0.00022
3.9
.0.0026 (
0.17
0.00011 i
0.58
0.00039
0.38
0.00025
0.96
0.00064
A
* - Concentration
(1) Data based on
below treatable level.
the average of three 24-hour composites,
° 7
167
-------
11-7. SUMMARY OF RAW V&SEE LOADINGS AT
VERIFICATION PLANTS
X
SUECA1EGORY
Pollutant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Thallium
Zinc
min.
0.044
0.0054
0.0062
0.0043
0.045
0.036
1.6
0.037
0.0059
0.0090
0.14
CHLORINE
Daily
loadings
(kg/day)
avg.
0.17
0.11
0.013
0.037
0.10
0.070
3-1
0.056
0.082
0.086
0.41
(rtERCURY CELL)
max.
0.30
0.27
0.025
0.098
0.18
0.12
5.1
0.075
0.22
0.14
1.1
S
,_... v._
f Unit
loadings
(kg/kkg)
min.
0,000076
0,000021
0.000014
0.0000087
0,000080
0.000096
0,0026
0,000064
0,000032
0.000015
0.00043
avg.
0.00039
0.00056
0.000070
0.00019
0.00037
0.00031
0.022
0.00015
0.00016
0.00019
0.0025
1 Number of
Plants
Averaged*
max.
0.00077
0.0018
0.00023
0.00067
0.00061
0.00071
0.063
0.00071
0.00039
.0.00029
0.010
3
4
4
5
5
5
5
4
3
3
5
* - Only those plants where the pollutant was observed at "significant
concentrations" are included in the averaging. "Significant
concentrations" is defined in 11.4.4.
168
-------
E. 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
The hypochlorite solution is either
to a wastewater treatment plant, or
Treatment of this waste is a
Decomposition is a common method of
thermal, and chemical methods as
corresponding hypochlorite.
sold, used on-site, sent
discharged without treatment
relatively recent practice.
treatment using catalytic,
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 the
chlorine. Chlorine condensate streams and spent chlorine-drying
acid are most commonly treated by steam or vacuum stripping, with
the chlorine frequently returned to process for purification and
recovery as a product. The tail gas is not generally scrubbed
with water because water does not effectively remove chlorine and
the chlorine concentration in the exhaust will reach 0.1 to 4.5
percent by volume after scrubbing with water. One effective
method of chlorine recovery from the tail gas is by the passage
of the gas through an absorbing material such as carbon
tetrachloride and subsequent recovery of the chlorine. The
process is proprietary and little information is available on its
design or performance.
169
-------
Best Management Practices
A. Area Runoff
Provisions can be made to divert and contain storm runoff from
plant areas. Collected runoff can then be sent to the wastewater
treatment system.
B. 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.
C. 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 precautions are not taken.
D. 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.
Advanced Treatment Technologies
Methods available for the removal of elemental mercury or mercuric
salts from plant wastewaters include precipitation with sodium sulfide
to form insoluble mercuric sulfide, adsorption by activated carbon,
adsorption by ion-exchange and other resins, reduction by borohydride,
hydrazine, sulfite, 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.
Selection of Appropriate Technology and Equipment
Technologies for Different Treatment Levels
Following the evaluation of significant toxic pollutants found in raw
wastewaters, current industry treatment practices and applicable
treatment alternatives, two levels of end-of-pipe treatment were
selectee5, as alternatives for application in the mercury cell chlorine
subcategory.
170
-------
A. Level 1 (BPT)
This treatment consists of sulfide precipitation of
mercury-bearing wastewater followed by pressure filtration. This
level of treatment, which will also reduce other heavy metals,
includes recycle of the brine mud overflow or filtrate 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.
B. Level 2
The filtered Level 1 effluent is passed through a granular
activated carbon bed where residual metal suIf ides and any
metallic mercury will be removed. The flow diagram for this
treatment level is shown in Figure 11-7. Cost estimates for this
level can be obtained from the proposed Development Document
(60).
Equipment for Different Treatment Levels
A. Equipment Functions
In Level 1, typical of existing treatment facilities,
mercury-bearing wastes are equalized in a surge tank, and
following chemical mixing, sulfide precipitates are removed in a
conventional plate and frame filter press followed by final pH
adjustment of the filtrate before discharge. In Level 2 a
conventional granular activated carbon filter is added for
further removal of residual metals before pH adjustment.
B. Chemical Handling
Sodium bisulfide is used with filter aid after pH adjustment to
pH 5-7. Care is needed to prevent escape of toxic and obnoxious
H2S fumes at neutral and acid pH level. 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.
C. 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.
D. Monitoring Requirements
Both levels of treatment include provisions for sampling and
monitoring of the wastewater discharge. Monitoring of heavy
171
-------
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.
Treatment Cost Estimates
General Discussion
To prepare
developed.
treatment cost estimates, a model
The model plant characteristics are:
plant concept was
A. Wastewater Flow
Data presented in Table 11-3 indicate an average wastewater flow
of 2.1 mVkkg for 13 plants, while the average of the five plants
surveyed during this study averaged 1.7 mVkkg.
For effluent limitation calculations (see 11.8.2)
estimation the more conservative unit flow from the
base (2.1 mVkkg) has been used.
B. Chlorine Production
and for
larger
cost
data
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 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).
C. Solid Waste Produced
Brine mud constitutes the major source of solid waste generated
at chlorine plants. Although flows and solids content vary
considerably from plant to plant, a.n average flow of 0.42 mVkkg
at 10 percent suspended solids ga.ve an estimated solids load of
42 kg/kkg to be used for cost estimating purposes. The
implementation of RCRA regulations has not been included in these
estimates, but RCRA costs are considered in the Economic Impact
Analysis of Pollution Control Technologies for Segments of_ the
Inorganic Chemicals Manufacturing Industry, EPA 440/2-81-023.
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
1/2
-------
HJDS1BEW
IMXON
LMXKK
.
erooss
SUUVR1C ACID
lODIHS 1HK
FILTER
AID
SCDIW
D
raxin?
BEIUVI LIW
FIUlfH
SCUDS ID
OR 1AHVIU. |
J
Incluiirt pH monllorlnj(, flow monitoring and
Figure 11-6. tevel 1 waste water treatment for chlorine - mercury cell subcategory
-------
CMNB »H_
STIVMI
UQQCN
-*. BEOICtETO
PKKXSS
Include* pH monitoring, How monitoring and
Figure 11-7. Level 2 waste water treatment for chlorine - mercury cell subcategory.
-------
chlorine/caustic plants, this stream is stripped of chlorine by steam
or vacuum and the chlorine is recycled to the purification operation.
The wastewater is then returned to the process and introduced to the
brine purification unit or sent to the treatment unit. The quantity
of wastewater 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 process wastewater discharges, the cost for the
dechlorination of such streams using sulfur dioxide has been included
because this is the treatment method on which control of total
residual chlorine is based.
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 Tables 11-8, 11-9 and
11-10. The costs of BAT treatment are incremental over BPT costs.
Table 11-11 presents a summary of the unit cost distribution between
amortization and operation and maintenance components.
Basis for Regulations
Basis for BPT Limitations
A. Technology Basis
Existing mercury cell chlorine plants are controlling mercury in
their wastewaters in accordance with existing BPT regulations
which require a discharge of less than 0.00014 kg/kkg of product
as a 30-day average. The BPT regulations presently in effect (40
CFR 415.62 (a)) will not be revised. Pollutants regulated
include TSS, pH 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.
175
-------
The existing regulations apply at the treatment system effluent
for TSS and mercury and at the plant effluent for pH. These
regulations are presented in Table 11-2 and are sustained by the
fact that plants having properly operated BPT technology have
demonstrated the achievability of the effluent limitations based
on available long-term monitoring data. Table 11-12 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.
B. Flow Basis
The existing regulations contain 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.
Basis for Final BAT Effluent Limitations
The original BAT limitations for this subcategory required zero
discharge of process wastewater pollutants. These regulations were
remanded and are not in effect. The final regulations allow for the
discharge of process wastewater following treatment.
A. 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 final BAT limitations.
For BAT, the Agency is promulgating 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 by at least 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 11-13 presents residual chlorine
discharges at plants that have reported the use, sale or
treatment of chlorine-bearing wastewaters. This data indicates
that some plants will be able to meet the residual chlorine
limitations without the application of additional technology.
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.
176
-------
B. Flow Basis
The flow basis for BAT limitations is 2.1 mVkkg 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 mVkkg to 5.6 mVkkg with an average of 1.7
mVkkg.
C. Selection of Toxic Pollutants to be Regulated
The selection of pollutants for which specific effluent
limitations are promulgated is based on the evaluation of raw
waste concentrations found during the sampling program and on the
treatability of toxic pollutants using BAT technology.
Table 11-14 presents the achievable concentrations of toxic
pollutants using the BAT technology of sulfide precipitation
followed by filtration. The first column gives the literature
based treatability data presented in Section 8 and summarized in
Table 8-11 and reflects the lowest level achievable by this
technology. The second column gives actual industrial wastewater
treatment system performance as presented in Tables 8-12 and 8-13
of Section 8. Also presented in.Table 11-14 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 literature-based 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, all the metals in Table 11-14, with the
exception of chromium and thallium for which no sulfide/filter
treatment data are available, are candidates for BAT regulation.
Consideration of Section 8 on the control parameters for sulfide
precipitation, however, leads to the selection of mercury as the
toxic pollutant to be regulated. Antimony, arsenic, cadmium,
chromium, copper, lead, nickel, silver, thallium and zinc are
included for guidance but no limits are set because control of
mercury will control these other metals.
D. Basis of Pollutant Limitations
Limitations are presented as unit loadings (kg/kkg) and/or
concentrations (mg/1) 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 unit loading effluent limitations
(kg/kkg} can be met by well-operated treatment facilities.
177
-------
TABLE 11-8. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Chlorine - Mercury cell
19,100 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal ....ft... .
Contractor's 0 & P .. ..
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
($)
BPT
26,500
93,000
20,000
139,500
20,925
160,425
32,085
192,510
19,251
211,761
21,000
232,761
TOTAL OPERATION AND
MAINTENANCE COST
112,000
1,300
500
21,176
6,983
4,500
15,000
161,459
BAT
0
23,000
0
23,000
3,450
26,450
5,290
31,740
3,174
34,914
0
34,914
14,000
500
1,500
3,491
1,047
0
7,500
28,039
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
34,454
195,912
5,681
33,719
a Represents the incremental cost above that for BPT treatment
Overhead and Profit
178
-------
TABLE 11-9. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Chlorine - Mercury cell
95,500 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal
Contractor's 0 & Pb....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
Land
TOTAL INVESTMENT COST .
B. OPERATION AND
MAINTENANCE COST
Labor and supervision .
Energy
Chemicals
Maintenance
Taxes and insurance ...
Residual waste disposal
Monitoring, analysis
and reporting
($
BPT
79,500
163,000
20,000
262,500
39,375
301,875
60,375
362,250
36,225
398,475
63,000
461,475
TOTAL OPERATION AND
MAINTENANCE COST
112,000
3,700
2,500
39,848
13,844
21,500
15,000
208,392
BAT*
0
40,000
0
40,000
6,000
46,000
9,200
55,200
5,520
60,720
0
60,720
21,000
700
7,500
6,072
1,822
0
7,500
44,594
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
64,832
273,224
9,879
54,473
a Represents the incremental cost above that for BPT treatment
k Overhead and Profit
179
-------
TABLE 11-10. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Chlorine - Mercury cell
191,000 metric tons per year
A. INVESTMENT COST
Site development ,.
Equipment
Monitoring equipment
oUDtOtaJ. •• • • • » ••••••••
Contractor' s 0 & P b.
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
($)
BPT
130,000
246,000
20,000
396,000
59,400
455,400
91,080
546,480
54,648
601,128
123,000
724,128
112,000
6,500
5,000
60,113
21,724
42,500
15,000
262,837
BAT
0
60,000
0
60,000
9,000
69,000
13,800
82,800
8,280
91,080
0
91,080
28,000
1,300
15,000
9,108
2,732
0
7,500
63,640
C. AMORTIZATION OF
INVESTMENT COST
TOT^L ANNUAL COST
97,804
360,640
14,819
78,459
f- Represents the incremental cost above that £or BPT treatment
Overhead and Profit
180
-------
TABLE 11-11. MODEL PLANT UNIT TREATMENT COSTS
Subcategory Chlorine - Mercury cell
Annual Treatment Costs ($/kkg)
LEVEL OF TREATMENT
COST ITEM
Annual Operation
and Maintenance
Annual
Amorti zation
Total Annual
Cost
PRODUCTION
(kkg/yr)
19
95
191
19
95
191
19
95
191
,100
,500
,000
,100
,500
,000
,100
,500
,000
BPT
8.
2.
1.
1.
0.
0.
10.
2.
1.
45
18
38
80
68
51
26
86
89
BAT*
1.
0.
0.
0.
0.
0.
1.
0.
0.
47
47
33
30
10
08
77
57
41
*Represents the incremental cost above BPT
181
-------
TABLE 11-12. MERCURY DISORRGES FROM SELECTED CHLQR-^LKALI
CKT.T. PLANTS*
SUBCATEGOKY
Plant
#343
#907
#898
#195
#106
#589
#299
BHffimrt •
im
dgH*£*
r***££?±':jg
#195**
#324**
Average
0.000025
0.000020
0.000060
0.000040
0.000065
0.000055
0.000040
.^?9iB@|s$f
0.000022
0.00086
CHLORINE (MERCURY CELL)
Mercury
Daily Maximum
0.00094
0.00026
0.0025
0.00073
0.00022
0.00086
0.00019
» ' uldoQa^t
:•----*•-—"- -^ ^"-,"
0.00066
0.0022
Waste Load Ocg/kkg)
Maximum 30 -day Average
0.00029
0.000030
0.00043
0.00015
0.000096
0.00049
0.000056
0.0000001
0.00010
0.0018
* See Reference 3
** Fran Plant Long Term Monitoring Data presented in Appendix A.
182
-------
BAT limitations, which apply at the treatment system effluent for
toxic metals and at the plant effluent for residual chlorine, are
presented in Table 11-15.
1. Chlorine - Total residual chlorine limits are based on an
evaluation of long-term monitoring data for total residual
chlorine as presented in Appendix A (Tables A-la and c).
The -long-term average concentration is 0.64 mg/1. The
average variability factor for daily measurements of total
residual chlorine is 2.3 and the average variability factor
for 30-day averages is 1.4.
The 24-hour maximum concentration is:
(0.64 mg/1) (2.3) = 1.5 mg/1
The maximum 30-day average concentration is:
(0.64 mg/1) (1.4) = 0.90 mg/1
The load limitations for total residual chlorine (kg/kkg)
are calculated based on the unit flow rate of 2.1 m3/kkg,
thus:
(1.5 mg/1) (2.1 mVkkg) ( kq/m3 ) = 0.0032 kg/kkg
(1000 mg/1)
for the 24-hour maximum limit. The maximum 30-day average
limit is calculated similarly, i.e.,
(0.90 mg/1) {2.1 mVkkg) ( kg/tn3 ) = 0.0019 kg/kkg
(1000 mg/1)
2. Toxic Pollutants
The effluent limitations and guidelines for the selected
toxic pollutants are derived from three sources of
information: industrial wastewater treatment system
perfornamce data (Table 8-12 and 8-13), verification
sampling data, and literature based treatability estimates
(Table 8-11).
The results of analysis of treated effluent represent plant
performance observed during three days of screening or
verification sampling. The effluent data for toxic
pollutants found above treatable concentrations in raw
wastes are summarized in Table 11-16. 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
filtration.
a. Mercury
183
-------
TABLE 11-13. RESIDUftL CHLORINE DISCHARGES AT SELECTED
CHLOR-ALKALI PLANTS*
Plant
#207
#014
#819
#747
#106
#589
1747(2)
#324 (2)
Chlorine Waste Load
Average
0.33
0.040
ND(1)
0.0020
0.0010
0.0030
0.0025
3.72
(kg/kkg)
Range
1.4 maximum
0 to 1.29
0.016 to 0.14
0 to 0.0060
0 to 0.14
0.0010 to 0.011
ND
0.38 to 12.2
* See Reference 3
(1) - ND = No data
(2) - Frcm Plant Long-Term Monitoring Data
184
-------
TABLE 11-14. COMPARISON OF RAW WASTE CONCENTRATIONS OF
TOXIC POLLUTANTS WITH TREKTABILITY
SUBCATEGORY
Literature-
Treatability
Pollutant (mg/1)
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Thallium
Zinc
ND
0.050
0.010
ND
0.050
0.050
0.010
0.050
0.050
ND
0.020
CHLORINE
Industrial
(MERCURY CELL)
Number Plants
out of Five
Waste Water Maximum
Treatment System Plant
Performance Average
(mg/1) (mg/1)
0.23(2)
0.15(3)
0.1S(4)
ND
0.20(3)
0.10 (3)
0.034(3)
0.022*41
0.070C4)
ND
0.12<3>
0.49
0.33
0.46
0.12
1.2
1.4
123
1.4
0.58
0.38
20
Average Exceeding
of 5 Literature-based
Plants Treatability
(mg/1) Level
<0.28
0.14
0.11
0.075
0.41
0.40
30.9
0.35
0.15
0.17
4.4
ND
3
3
ND
5
3
5
2
2
ND
5
(1) Estimates from Section 8.1, Table 8-11, given as the lower limit of
treatability.
(2) Data from Table 11-16 (average effluent concentration from verification
sampling).
(3) Estimated achievable long-term average concentrations from Table 8-13.
(4) Data from Table 8-12 (recently submitted by Olin Corporation) .
ND = No data available on treatability witii sulfide/filter.
185
-------
The BAT limitations for mercury, although based on the
same technology, are more stringent than BPT
limitations. The estimated achievable long-term
average concentration for mercury from Table 8-13 is
0.034 mg/1 and is obtained from recently submitted
industry data. A daily variability factor of 3.1 and a
30-day average variability factor of 1.4 are estimated
from data in Appendix A and from the recently submitted
industry data.
The 24-hour maximum concentration is:
(0.34 mg/1) (3.1 ) = 0.11 mg/1
The maximum 30-day average concentration is:
(0.034 mg/1) (1.4) = 0.048 mg/1
The load limitations for mercury (kg/kkg) are
calculated based on the unit flow rate of 2.1 m*/kkg,
thus:
(0.11 mg/1) (2.1 mVkkg) ( kg/in a ) = 0.00023 kg/kkg
(1000 mg/1)
for the 24-hour maximum limit. The maximum 30-day
average limit is calculated similarly, i.e.,
(0.048 mg/1) (2.1 mVkkg) ( kq/m^ ) = 0.00010 kg/kkg
(1000 mg/1)
Comment from 3 companies operating 10 plants stated
that the plants can achieve these limitations.
Zinc
The zinc guidance is based on the estimated achievable
long-term average concentration of 0.12 mg/1 which is
obtained from Table 8-13. A daily variability factor
of 7.6 and a 30-day average variability factor of 1.5
are obtained from the recently submitted industry data
on zinc effluent concentrations.
The 24-hour maximum concentration is:
(0.12 mg/1) (7.6) = 0.91 mg/1
The maximum 30-day average concentration is:
(0.12 mg/1) (1.5) = 0.18 mg/1
c. Antimony
186
-------
TABLE 11-15. EFFLUENT LIMITATIONS
Chlorine-Mercury Cell
Best Available Technology
Wastewater Plow: 2.1
SUBCATEGORY
CHLORINE MERCURY CELL
Pollutant
Subcategory
Performance
(mq/1)
Daily
Variability
Factor
30-day Avg.
Variability
Factor
Concentration
Basis
(mq/1)
Max.
30-day 24-hr,
Avg. Max.
Effluent Limit
Max.
30-day
Avg.
24-hr
Max.
Nonconventional
Pollutants:
Total Residual
Chlorine*7)
Toxic Pollutants:
2.3/1.4
0.90
1.5
0.0019
0.0032
Antimony 0.23<2) 7.6/1.5 0.35 1.7
Arsenic 0.15<3> 6.7/1.4 0.21 1.0
Cadmium Q.050<2> 7.6/1.5 0.075 0.38
Chromium 0.044<2> 7.6/1.5 0.066 0.33
Copper 0.20<3> 7.6/1.5 0.30 1-5
Lead 0.10(3) 4.1/1.3 0.13 0.41
Mercury(6J 0.034<3) 3.1/1.4 0.048 0.11
Nickel 0.10<4> 5.7/1.4 0.14 0.57
Silver 0.067(2) 7.6/1.5 0.10 0.51
Thallium 0.17<2> 7.6/1.5 0.26 1.3
Zinc 0.12(3) 7.6/1.5 0.18 0.91
— IS)
__(5)
— (5)
__(5)
-_(5)
— (5)
0.00010
— (5)
— (5)
— (5)
— (5)
— (5)
— (5)
—(5)
-_(5)
—(5)
— (5)
0.00023
— (5)
— (5)
__(5)
~(5)
( 1 ) Long-term average concentration from Appendix A.
(2) Average effluent concentration from verification sampling.
(3) Estimated achievable long-term average concentration from Table 8-13.
(4) Lower limit of literature treatability for sulf ide/filter technology according
to Table 8-11.
(5) No load limits; concentration limits are provided for guidance purposes.
(6) Limits are also applicable to PSNS, and NSPS.
(7) Limits are also applicable to NSPS.
187
-------
I
Because no industry data is available, the antimony
guidance is based on the average effluent concentration
from the verification sampling. The value of 0.23 mg/1
is used as a long-term average. The variability
factors established for zinc are applied to antimony
since these would approximate the situation where a
metal is incidentally controlled.
The 24-hour maximum concentration is;
(0.23 mg/1) (7.6) = 1.7 mg/1
The maximum 30-day average concentration is:
(Q.23 mg/1) (1.5) = 0.35 mg/1
Arsenic
The arsenic guidance is based on the estimated
achievable long-term average concentration of 0.15 mg/1
which is obtained from Table 8-13. A daily variability
factor of 6.7 and a 30-day average variability factor
of 1.4 are obtained from the recently submitted
industrial data on arsenic effluent concentrations.
The 24-hour maximum concentration is:
(0.15 mg/1) (6.7) = 1.0 mg/1
The maximum 30-day average concentration is:
(0.15 mg/1) (1.4) = 0.21 mg/1
Cadmium
Because no industry data are available, the cadmium
guidance is based on the average effluent concentration
from the verification sampling. The value of 0.050
mg/1 is used as a long-term average. The variability
factors established for zinc are applied to cadmium
since these would approximate the situation where a
metal is incidentally controlled.
The 24-hour maximum concentration is:
(0.050 mg/1) (7.6) = 0.38 mg/1
The maximum 30-day average concentration is:
(0.050 mg/1) (1.5) = 0.075 mg/1
Chromium
188
-------
TABLE 11-16. EFFLUENT CONCENTRATIONS OF TOXIC POLLUTANTS
FROM VERIFICATION SAMPLING
SUBCATEGORY
Pollutant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Thallium
Zinc
CHLORINE (MERCURY
CELL)
Plant Effluent Concentrations
(rag/1)
Plant
#299
0.15
0.063
0.073
<0.060
0.038
<0.050
0.029
<0.050
<0.015
0.20
0.10
#747
<0.25
<0.010
0.12
<0.050
<0.025
0.073
0.10
<0.050
<0.015
<0.045
<0.025
#317
<0.25
0.020
<0.025
<0.050
<0.030
0.17
0.19
<0.067
<0.015
<0.25
0.51
#106
<0.45
<0.0050
0.016
<0.010
0.043
0.38
<0. 00050
0.14
0.26
0.26
0.088
#167
<0.065
0.38
0.010
<0.050
<0.025
0.12
0.32
<0.050
<0.015
0.090
<0.025
Treatability *
(mg/1)
Avg.
<0.23
<0.096
<0.050
<0.044
<0.033
<0.16
<0.13
<0.074
<0.067
<0.17
<0.15
(2)
0.050
0.010
(2)
0.050
0.10
0.010
0.10
0.050
(2)
0.020
(1) Lover limit from literature-based treatability estimates from Section 8.1
(2) No data available for treatability with sulfide/filter.
189
-------
Because no industry data are available, the chromium
guidance is based on the average effluent concentration
from the verification sampling. The value of 0.044
mg/1 is used in place of a long-term average. The
variability factors established for zinc are applied to
chromium since these would approximate the situation
where a metal is incidentally controlled.
The 24-hour maximum concentration is:
{0.044 mg/1) (7.6) = 0.33 mg/1
The maximum 30-day average concentration is:
(0.044 mg/1) (1.5) = 0.066 mg/1
g. Copper
The copper guidance is based on the estimated
achievable long-term average concentration of 0.20 mg/1
which is obtained from Table 8-13. A daily variability
factor of 7.6 and a 30-day average variability factor
of 1.5 are obtained from the recently submitted
industry data on copper ef-fluent concentrations.
The 24-hour maximum concentration is:
(0.20 mg/1) (7.6) = 1.5 mg/1
The maximum 30-day average concentration is:
(0.20 mg/1) (1.5) = 0.30 mg/1
h. Lead
The lead guidance is based on the estimated achievable
long-term average concentration of 0.10 mg/1 which is
obtained from Table 8-13. A daily variability factor
of 4.1 and a 30-day average variability factor of 1.3
are obtained from the recently submitted industry data
on lead effluent concentrations.
The 24-hour maximum concentration is:
(0.10 mg/1) (4.1) = 0.41 mg/1
The maximum 30-day average concentration is:
(0.10 mg/1) (1.3) = 0.13 mg/1
i. Nickel
190
-------
The industrial wastewater treatment system performance
data in Table 8-12 show values for nickel below the
lower limit of literature treatability estimated in
Table 8-11. Because of this, the lower limit, 0.10
mg/1, is used in place of a long-term average as a
basis for the nickel guidance. A daily variability
factor of 5.7 and a 30-day average variability factor
of 1.4 are obtained from the recently submitted
industry data on nickel effluent concentrations.
The 24-hour maximum concentration is:
(0.10 mg/1) (5.7) = 0.57 mg/1
The maximum 30-day average concentration is:
(0.10 mg/1) (1.4) = 0.14 mg/1
j. Silver
Because no industry data are available, the silver
guidance is based on the average effluent concentration
from the verification sampling. The value of 0.067
mg/1 is used in place of a long-term average. The
variability factors established for zinc are applied to
silver since these would approximate the situation
where a metal is incidentally controlled.
The 24-hour maximum concentration is:
(0.067 mg/1) (7.6) = 0.51 mg/1
The maximum 30-day average concentration is:
(0.067 mg/1) (1.5) = 0.10 mg/1
k. Thallium: Because no industry data are available, the
thallium guidance is based on the average effluent
concentration from the verification sampling. The
value of 0.17 mg/1 is used in place of a long-term
average. The variability factors established for zinc
are applied to thallium since these would approximate
the situation where a metal is incidentally controlled.
The 24-hour maximum concentration is:
(0.17 mg/1) (7.6) = 1.3 mg/1
The maximum 30-day average concentration is:
(0.17 mg/1) (1.5) = 0.26 mg/1
191
-------
Basis for BCT Effluent Limitations
While EPA has not yet proposed or promulgated a revised BCT
methodology in response to the American Paper Institute v. EPA
decision mentioned in Section 3, EPA is promulgating BCT limitations
for this subcategory. These limits are identical to those for BPT.
EPA is not promulgating any more stringent limitations since we have
identified no technology option which would remove significant
additional amounts of conventional pollutants. The dechlorination
technology added to BPT for BAT does not remove additional
conventional pollutants. As BPT is the minimal level of control
required by law, no possible application of the BCT cost tests could
result in BCT limitations lower than those promulgated in this
regulation. Accordingly, there is no need to wait until EPA revises
the BCT methodology before promulgating BCT limitations.
Basis for New Source Performance Standards
For NSPS, the Agency is promulgating limitations equal to BPT for TSS
and pH and BAT for other pollutants because of the prohibitive cost of
additional technology. Pollutants to be limited are pH, TSS, mercury,
and total residual chlorine.
Basis for Pretreatment Standards
For pretreatment standards for new sources, the Agency is promulgating
limitations based on BAT technology excluding dechlorination.
Dechlorination is unnecessary for discharges to POTWs because influent
to POTWs is often chlorinated. The pollutant to be limited is
mercury. The PSNS limitations are based on BAT because this provides
better mercury removal than is achieved by a well-operated POTW with
secondary treatment installed and, hence, mercury will pass through a
POTW in the absence of pretreatment.
The Agency is not promulgating PSES for this subcategory. Instead,
the subcategory is excluded from categorical PSES under the provisions
of paragraph 8(b) of the Settlement Agreement because the discharge of
total toxic metals to POTWs from the two existing sources combined is
below treatable levels and amounts to only 40 pounds per year. Both
existing sources have installed and are operating treatment facilities
equivalent to BPT/BAT to control mercury.
Diaphragm Cell Process Industry Profile
General Description
Approximately 65 percent of the U.S. production of chlorine is by
diaphragm cell plants. Of 40 known plants in 1978, Section 308 data
are available for 19. Table 11-17 presents a summary profile of the
subcategory. Table 11-18 presents the status of discharge regulations
prior to promulgation of this regulation for diaphragm cell chlorine
plants.
192
-------
General Process Description
A. 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 through the brine system.
B. Diaphragm Cell
The treated brine solution is electrolysed in the diaphragm cell
to form chlorine, hydrogen, and sodium hydroxide according to the
reaction:
2NaCl + 2HZ0 = 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 some cell designs using lead 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 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.
C. 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 and 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 nonbrine system pH
193
-------
TABLE 11-17. SUBCAIEGORY PROFILE DATA
SUBCATEQQRY
CHLORINE (DIAPHRAQ1 CELL)
Total subcategpry 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
Maxinum
Volume per unit product:
Minimum
Maximum
8,270,000 kkg/year
6,430,000 kkg/year
40
19
6,400,000 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 are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S, Department of Commerce, Current Industrial
Reports, December 1977; Energy ani 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.
194
-------
TABLE 11-18. STATUS OF REGULATIONS - EFFLUENT LIMITATION GUIDELINES
SUBCATEGORY
SUBPART
CHLORINE (DIAPHRAGM CELL)
F (40 CFR 415.60, 3/12/74)
STANDARDS
BPCTCA BATEA NSPS
Max "" Avg. Max* Avg. Max. Avg.
Parameters (kgAkg) (kg/kkg) (kq/kkg) (kgAkg) (kg/kkg) (kg/kkg)
Product
Process
Diaphragm
Cell
Process TSS
Pb
0.64
0.005
0.32
0.0025
No discharge
of pwwp(3)
No discharge
of pwwp L£\
0.64
0.32
0.00008 0.00004
Section 415.63 was remanded and reserved (41 PR 51601, November 23, 1976).
(1) Max. = Maximum of any one day.
(2) Avg. = Average of daily values for thirty consecutive days.
(3) pwwp = Process wastewater pollutant.
195
-------
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 11 percent NaOH and a sodium chloride
content as high as 17 percent. The caustic is evaporated to 50
percent NaOH by multiple effect evaporators. Sodium chloride
remains as a solid salt which is then separated from the caustic
and returned to the brine system. Further purification of the
caustic is necessary for some applications (such as rayon
production) and extraction or adsorption techniques have been
used to remove small amounts of impurities. The caustic can be
evaporated further if more concentrated products are required.
The vapor evolved from the last of multiple effect evaporators is
condensed in barometric condensers generating contact cooling
water, or in surface condensers using noncontact cooling water.
The hydrogen gas generated in the process can be vented or cooled
by refrigeration to remove water vapor before sale or use as a
fuel.
Figure 11-8 is a general flow diagram for the manufacture of
chlorine by the diaphragm cell process.
Water Use and Wastewater Sources
Water Use
Water use at diaphragm cell plants is similar to that at mercury cell
plants with one exception. Common uses include noncontact cooling,
tail gas scrubbers, cell wash, equipment 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.
Waste Sources
A. 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.
196
-------
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
(mVkkg), with a solids content of from 2 to 20 percent.
B. 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 some 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.
C. Chlorine Cooling Condensate
Condensate 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 mVkkg.
D. 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 nonbrine system
pH control.
E. 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 mVkkg.
F. 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
wastewater flow of 0.45 mVkkg. Backwashing of filters used to
clarify caustic product at the same plant resulted in an average
flow of 5.4 mVkkg. At some diaphragm cell plants, these
wastewaters 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.
197
-------
BRINE
SULFATE BM>
PURGE
NONCONTACT
COOLING -P>
WATER ^.
PURIFICATION
SYSTEM
*
DIAPHRAGM
CELL
DRINL HUUS ^fM jtgyaj. HOHCOHTACT COOLING HATER
t »
— HYDRC
CHLORINE
1
lit TO lit SODIUM
HYDROXIDE
SOLUTION
1 RECYCLE A
HATER
— BAROMETRIC
^ CONDENSER
COOLER OR USE
t
COOLING
HATER
COOLER •*• CHLORINATED HATER COHDENSATE
t * 1
COOLING
TOHER**
M
M) f
00 f
BLOHDOHN
TO HASTE
HATER
*
LEAKS.
SPILLS
HASIIDOHN
ETC.
r~
t
SALT
REMOVAL
1
SODIUM
HYDROXIDE
SOLUTION
1
FILTER
EUliURIC -p»
ACID
*) ILIME) f 4
uint-KEaiMNi i •- SOLUTION " SALES. OR HAS
f
AT ION
SYSTEM *
* * *
PACKAGING
HOHCOHTACT
COOLING
TO USB
OR SALES
LIQUEFIER f
1 DIAPHRAGM
LIQUID HASHING
CHLORINE n*imsf inw
HnTER
1 t
TO BALES U»CT»
USED SOME PIAHTS ONLY
DEFENDS UPON PLANT DESIGN
Figure 11-8. General process flow diagram for production of chlorine/caustic by
diaphragm cells.
-------
G. 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.
H. Barometric Condenser Wastewater
When vapors from caustic evaporators are contact-cooled, a
significant amount of wastewater can be generated. Flows of from
^,,90^ Jha ^300 mVkkg have^ been reported at facilities where
barometric ^'coriHe'rtser"'"^"i'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 wastewater at two plants ranges from a flow of 0.82
mVkkg to 0.89 mVkkg.
I. Summary of Wastewater Flows
Table 11-19 summarizes unit wastewater flow data available by
specific sources. A separate list of flows at one graphite anode
plant is presented to compare wastewater generation between metal
anode and graphite anode plants.
Descriptions of Specific Plants
The following description of plants includes those plants that were
sampled during the screening and verification program. The discussion
primarily covers plant practices in wastewater 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.
Screening
At Plant 1014, 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 the 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 11-9, the general process flow sheet is presented. The
waste streams sampled and their waste loadings are presented in Table
11-20. The only process waste stream discharged is once-through
barometric condenser cooling water.
199
-------
Verification
Four plants were visited and their waste streams sampled during the
verification program. The results of analysis of the wastewaters are
presented in Table 11-20.
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 filrate 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 wastewaters. Dechlorination of the drying
acid by reaction with sodium bisulfite is planned in the near future.
Figure 11-10 shows the process flow diagram and sampling points.
Plant #738 has two production lines, 738A and 738B, that are almost
identical. At the new plant (738B) the NaOH is not concentrated nor
is the waste from the chlorine disposal system scrubbed. In addition,
the inert gases from the liquefaction step are put through the
chlorine disposal system. The process flow sheets are shown in
Figures 11-11 and 11-12.
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 wastewaters,
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 11-13 shows the process flow diagram
and sampling points.
Plant #967 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 11-14 is a general process flow diagram for Plant #
967.
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 1326, wastewater from the diaphragm cell process is combined
with other process wastewaters. The combined wastewater is sent to
200
-------
TABIE 11-19. WASTE WATER FLOWS AT DIAPHRAC31 rftT.T, CHIORINE PLANTS
Stream Description
Flow (m-VWcg)
Plants with Plant with
Metal Anodes Graphite Anodes
min. avg. max.
Cell room wastes
and cell wash
Chlorine Condensate
Spent Sulfuric Acid
Tail Gas Scrubber
Caustic Filter Wash
Brine Filter Backwash
Caustic Cooling Blowdown
Brine Mud
0.020 0.38 0.67
0.16 0.49 0.90
0.010
0.10 0.17 0.29
NA
NA
0.82 0.86 0.89
0.040 0.42 1.5
1.2
0.78
NA
—
0.
——
5.4
0.45
NA
NA
NA: Not Available
201
-------
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 wastewater 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 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).
Toxic Pollutant Concentrations
A. Analytical Data Base
Section 5 of this report describes the methodology of the
screening and verification sampling program. In the chlorine
diaphragm cell industry, a total of 15 days of sampling were
conducted at Plants #014, 261, 738, 967, and 736. Thirty-seven
different sampling points were involved covering various raw
waste streams and the treated effluents at these plants. The
evaluation of the toxic metal content of these process related
waste streams was based on 975 analytical data points. The
sampling for toxic organic pollutants at Plants #014 and 967
generated 2300 analytical data points. Analysis of waste for
asbestos generated an additional 13 data points.
B. Asbestos
Asbestos, used as a diaphragm sepatating the cell anode and
cathode, is the major toxic pollutant consistently found in
process wastewater from diaphragm cell plants. It occurs
primarily in wastes resulting from activities such as cell room
washdown and cell repair and cleaning.
Table 11-21 presents the results of asbestos determinations of
supply water and wastewaters at three diaphragm cell plants.
202
-------
VEOTGAS
N)
O
BRIMK
BRINE MUD
,annoric
t"
i
REIUJM
1
KM
' B/M-Ofl
OOH
1
NmH ^"l BULGftS^
9»t H-Ki. * mMmffl
* 1
9
C12
i 1 1
WfffB OOOI.IHG ^ i REUSE 751 HfO^ COOLIHO WIHt
ItoCtt 14
1
II
AT1OH
HSOH Hartl ttBIOH
1- | — j— |
f cnoi.iwj 501 71t ..^,,1
HUTMEmiC HJ» NKJM MflttnflUC (aim
CXMJEH3BI „ ,j.™«. CCKDOBER
•..•Mm. ~r. TO WftMB u.«*« «n
tffvniR ID d WTEK TU
MNS1C. I i MSIC
T T ^-^
1 •• Sanplln? polnto.
TJT
PWHT
Figure 11-9. General process flow diagram at plant #014 showing the sampling points.
Chlorine/caustic (diaphragm cell) manufacture.
-------
TABLE 11-20. POLLUTANT OKENTRATIONS AND LOADS AT
SCREENING AND VERIFICATION PLANTS
SUBCATEGORY CHLORINE DIAPHRAGM rrrr.T.
Plant &
Stream Stream TSS
No. Description (mg/1) (kg/kkg)
#014
#261
3
4
5
6
1
2
3
4
5
r?38Al
2
3
4
5
#738B6
7
8
9
10
11
12
13
14
C12 condensate
Cell wash
Brine mud
Bar. condenser
Brine mud
Cell wash
Asbestos filtrate
Filter cake
Bar. condenser
Cell room waste
Asbestos wash
Hypo scrubber
Cl2 cooling water
Caustic cooling
tower
Cell room waste
Asbestos wash
Hypo scrubber
Cl2 cooling water
Caustic cooling
tower
Chlorate sump
Plant effluent(B)
Final effluent
(Total)
Brine mud
.20
1600
NA
7.0
NA
4800
9.0
NA
6.0
27
57
290
35
48
95
72
160
20
4.7
32.
63
58
270
1.
2.
1.
1.
7.
2.
2.
4.
4.
a.
1.
1.
3.
7.
5.
4 x
4 x
NA
3.6
NA
8 x
NA
NA
NA
4 x
0 X
7 x
2 x
3 x
5 x
3 x
4 x
7 x
8 x
0 x
7 x
NA
NA
10
10
10
10
10
10
10
10
10
10
10
-3
-2
-1
-3
~3
-2
-1
-2
-3
~3
-2
10-2
10
10
10
-3
1-3
-1
Lead
(mg/1) (kg/kkg)
0.
0.
0.
0.
0.
2.
0.
42
< 0.
0.
0.
0.
0055
26
72
0050
36
0
075
010
077
031
18
0.28
0.
0.
0.
0.
0.
< 0.
< 0.
51
067
13
20
20
010
010
0.12
0.
0.
078
10
5.
3.
1.
1.
3.
7.
3.
3.
1.
1.
4.
3.
1.
1.
1.
< 8.
< 2.
1.
OxHT6
9xlO"6
3xlQ-5
5xlO~3
OxlO"4
6xlO~5
NA
NA
NA
9X10"6
8X10"6
7xlO"5
3X10"4
SxlO-4
2xlQ-6
5xlO~5
7xlO~5
7xlO"5
2xlO"6
3X10"6
lxlO~3
NA
NA
(Continued)
204
-------
TABLE 11-20 (continued)
Plant &
Stream No
#736
1
2
3
4
5
6
7
#967l
2
3
4
5
6
7
Stream
Description
Cell wash
Cell room drain
Brine mud
50% Bar. condenser
70% Bar. condenser
95% Bar. condenser
Chlorine condensate
Cell bldg wastes
Lead pond effluent
Caustic backwash
Brine backwash
Cell wash
Condensate and I^SO^
Scrubber waste
TSS
(mg/1) (kg/kkg)
934
284
20,000
32
21
90.3
2.4
1000
54
160
13,000
310
1 1100
270
6.0 x 10~2
4.6 x 10~3
33
NA
NA
NA
3.9 x 10"4
1.8 x 10"1
3.0 x 10~2
8.6 x 10'1
5.8
5.6 x 10"2
8.7 x 10'1
1.2 x 10 ~2
Lead
(mg/1) (kg/kkg)
0.014
0.17
0.019
0.010
0.010
0.010
0.010
680
29
0.32
0.52
48
0.92
0.67
9.1X10"6
2.8xlO~6
3.1xlO~5
NA
NA
1.6xlO"6
1.2X10"1
1.6xlO~2
1.7xlO~3
2.3xlO~4
8.6xlO~3
7.3xlO"4
2.9xlO~5
NA: Not Available
205
-------
fOOIUH
CARBONATE
(UOH
SALT
HYDROGEN TO DOILCR
STEAM
ADJACENT PLANT
HI1CAS
CURI-
FLOCCULATORS
f
1RINE HUftS
SAND
FILURS
A
BACK
i i WASH
SETTLER
FILTRATE TO
PROCESS SEWER
SLUDGE TO LANDFILL
COOLIHG WATER
HVPOCHLORITE
TOWER
1 _ fc.
'TO VENT
SOLUTION TO
ADJACENT PLANT
... STORAGE TAMK
CHLORINE ANfi RAILROAD CAR
TO STORAGE WASHDOWN
r
COOLING
WATER
STEAH
\ <
IS
i
73t CAUSTIC
EVAPORATOR
1
PUXGE roH DISPOSAL
IV COH1UCT
NaQH
LEGEND
SAMPLING POINTS
BOILER
FEED WATER
TO SUKE LIME
(FO* LIKE PLANT)
PROCESS SEWER
TO CJECTOH
WATER
PROCESS
SEWER
Figure 11-10. General process flow diagram at Plant #261 showing the sampling points.
Chlorine/Caustic (Diaphragm Cell) manufacture
-------
7«
BRINE
'Cl,
MtSIE WTER
OMC1 + MaOIlt
(Ftolnt 112 - uaate MMp (ccnblnatlon of «11 waste*).I
Figure 11-11. General process flowsheet at plant 1738-A showing the sampling points.
Chlorine/caustic (diaphragm cell) manufacture.
-------
T
OQOUH3 H,O
**«*•*— «MM \
. \
'/I
BRINE ~* •"
WRinCWKM BKOK Q^
ro NO son
O'» , 1
f {
BHKC * CB
HZ> *,
1 '
HJ) ***
I — • ^ NaCl RBCOUBm
fj
OQOUMQ H,0 ' " '
Maiio
1
NU7K
I na »
""^ i (C^I^ M«»l -^*BSS«»l
CCWfCSSlQl H^O H.60, ^ ^^
1 1 II 4
^J ^ CCQUHG MO '±2^ WRtfKKOM ^2^ i i^ia'AM-IlN
OflDW OMWS3KM
i, ^i *
P " Wyl» Hftsrc ware
| H^ HVDHOCMWCNS
U. ASBESTOB *!J3?
5M MUJH H2°
»3iiE!ntO 1 A
nnwOTn 1 /^V\ ^ WttnY
KM) 1 ^*^
'" m&at HMEH f
"a30* in
e~»-« JttWAIJZA'ntMl— J "* I" — ^4-
»JH*UUJ 1 ™ MUBWHI ( ^y
IU
H»SHi WMEH "J |wa;|u
Q— — •
Figure 11-12. General process flow diagram at plant #738-B showing the sampling
points. Chlorine/caustic (diaphragm cell) manufacture.
-------
N)
o
TOQQUHJwaea
i V,
DtSKUKU | ^\;J
HMHCMUM f
^•MICAU
f I
I'll II 11 UlMtMKM
MICK* h*H OMUFIBII hM HUB* r*H CH.L
1 1
1 1 '•«$
^^ , . mit* 11MXMGH f
T™ " TO UU* ICU, (Mt
BRD1B H» TO WfliOW. DOH
0»- Mil. ' «•"«
BMCH
p senuHQ
1 TNK
1 1
ancta ASUKSTDO
UOUt 1O SOLID
to Biwot Wvsre
DISPOSAL
CIO
KtWXtlt TO ATMOSHEI*
OftCWTWEUSB
I MjSOj
fc COftACT
* CCOIA
1 1 1
•M Dram M (nan !*•
t_J I—I «"1W
~ """P'.i-Qit
9I»M T
CCHTAJT IMEfl
TOWVal «»»
+
" tVHOt*
comXcr
wnot
[eatjeotn
HIST -*J BKKMaMC 1
EUMIIhim 1 CQtWBOd I
utiNQ 1 HTKIHJB sxa
KSCKl M— ICGOCHPOCfn
*• *"
OMftm \
udtc sitaaet
lasaccM.
••H nMtut r*-i — ocwSiw
cw. ~
NaCH LICUOH *""
wonnt mxuar t 1 i
fc fHOK
MG(
UO^TIH. U
UO1ID
auawm
TO erowae
7T »u UWA
I*
OKH1CMA
»0f J
M»» -|
1
8WT HI
utracium
HM-wna
nunwrtcH 1 ^,
* '
comer WMCS
loeenuK) KM>
a 4 • ' " 1M
KM """ I OMCBfl
^_
WlMft
«u aUeM* Mniited.
»
oWtterwreR ™.-«.
TO EKITLIW ram ^J2J
«J*K3
; lAnukt HUBM
1O BETIUJC KH>
Figure 11-13. General process flow diagram at Plant #736 showing the sampling points,
Chlorine/Caustic (Diaphragm Cell) manufacture
-------
to
M
O
SOCK*
HIKER
uernui
Figure 11-14. General process flew diagram at Plant #967 shewing the sanpling points.
Chlorine/Caustic (Diaphragm Cell) manufacture
-------
Results are expressed as total fibers per liter (in millions)
well as Chrisotile and Amphibole fibers per liter.
as
There is no standardized 304(h) analytical method for asbestos in
water and because of this, EPA is excluding limitations for
asbestos from these regulations and deferring regulation to a
later date.
C. Toxic Metals
Table 11-22 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 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:
(Flow A x Concentration A) + (Flow B x Concentration B?
(Flow A + Flow B)
= concentration of mixed streams
Substituting numerical values gives:
(100 qpm) (15 mq/I) * (10 qpm) 60 mg/1) = 19 mg/1
(110 gpm)
This method was used to calculate raw waste concentrations of
pollutants as presented in Table 11-22. Barometric condenser
wastewater 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,
211
-------
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
chlorine production rate, the waste stream flow rate, and the
measured pollutant concentration.
Unit loading (as kg of pollutant per kkg =
of chlorine)
(C) (Q).
1000P
where C and Q are as described above, and P is the chlorine
production rate expressed in units of kkg/day (kkg is 1000 kg, a
metric ton, which is equal to 2205 Ibs).
The minimum, average and maximum values were calculated based on
data from those plants where the particular pollutant was found
at a detectable concentration.
In Table 11-23, 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 11-24, plant average daily and unit loadings are
presented as minimum, average, and maximum values based on data
presented in Table 11-23 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 graph i te anode plant, and the
estimated total subcategory production, the estimated total
pollutant raw waste loads generated each year by this subcategory
are as follows:
212
-------
TABLE 11-21. RESULTS OF ASBESTOS SAMPLING AT DIAPHRAGM t-FT.T. PLANTS
Plant Stream
1 261 Supply
Cell Wash
Filtered Discharge
Barometric
Condenser
# 736 Supply
Cell Wash
Cell Rxm waste
Barometric
Condenser
Barometric
Condenser
Barometric
Condenser
# 967 Supply
Cell Waste
Pond Effluent
Caustic Wash
Brine Filter
Backwash
Cathode Wash Waste •
Condensate & Spent
Acid
Neutralizer Waste
Total Asbestos
Fibers (MFL)*
8.0
2.1 X 108
1.6 X 103
0.40
0.70
2.0 X 107
2.9 X 102
1.8
5.3
1.4 X 102
9.7 X 102
2.4 X 104
2.4 X 103
7.8 X 103
8.0 X 102
3.2 X 105
2.7 X 102
2.1 X 103
Chrisotile
MFL
7.5
2.1 X 108
1.6 X 103
0.40
0.70
2.0 X 107
2.8 X 102
0
5.3
1.4 X 102
9.7 X 102
2.4 X 104
2.4 X 103
7.8 X 103
6.2 X 102
3.2 X 105
1.8 X 102
2.1 X 103
Atnphibole
MFL
0.40
0
0
0
0
0
8.0
1.8
0
0
0
8.0 X 102
0
0
1.8 X 102
0
8.9 X 10
0
*Million fibers per liter
213
-------
TftBLE 11-22. MAXIMUM RAW WASTE COTCENTRATIONS OF TOXIC METALS OBSERVED AT
DIAPHRAGM CFT.T. CHLORINE PLANTS (mg/1)
SUBCATEGORY
Toxic
Metal
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Zinc
CHLORINE DIAPHRAQ4 Cftf.!,
Plants with
Metal Anodes
<0.25
0.17
<0.014
0.037
7.4
17
2.0
<0.0030
22
<0.020
0.018
<0.25
3.0
Plant with
Graphite Anode
<0.065
0.59
<0.0010
0.017
<0.048
0.27
44
0.0040
0.070
<0.030
<0.016
<0.050
0.25
214
-------
Raw Waste Load
Pollutant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Zinc
kg/year
483
6,300
41
3, 100
4,400
470,000
48
3,600
5
5,100
Because cell room wastes including cell or cathode wash wastes,
leaks, spills and washdown are usually treated separately at
diaphragm cell plants and because other process wastes such as
filter backwashes, condensates and caustic evaporation wastes are
usually discharged after settling, these two waste mixes were
evaluated separately. Table 11-25 presents average raw waste
concentrations and loads of toxic metals found in cell room
wastes at the six diaphragm cell plants sampled. Table 11-26
presents the similar data from the sampling of other process
wastes at these plants.
D. 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 11-27 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 11-28 presents the concentrations of toxic organics by
individual raw waste streams 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.
215
-------
TABLE 11-23. TOXIC METAL CONCENTRATIONS AND LOADS AT SCREENING AND
VERIFICATION PLANTS
1
(mg/1)
SUBCATEGORY
POLLUTANT
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Zinc
014
*
*
*
*
0.0020
0.0000018
0.019
0.000017
0.015
0.000014
0.0060
0.0000045
0.0020
0.0000018
0.90
0.00081
*
*
*
(kg/kkg)
CHLORINE DIAPHRAGM CELL
261
*
0.17
0.0000064
0.037
0.0000014
1.9
0.000071
17
0.00064
2,0
0.000075
*
*
22
0,00081
0.018
0.00000070
1.5
0.000054
Plant #
738A
*
*
*
*
*
0.52
0.0046
0.045
0.00039
0.082
0.00060
*
*
0.21
0.0018
*
*
0.29
0.0021
738B
*
*
0.011
0.000021
*
*
0.066
0.0012
0.12
0.00023
0.11
0.000021
*
*
0.067
0.00013
*
0.093
0.00018
736
0.010
0.0000033
0.057
0.000014
0.025
0.0000061
0.18
0.000044
0.43
0.00011
0.016
0.0000039
0.0030
0.0000070
0.22
0.000054
*
*
3.0
0.00074
967**
0.011
0.00015
0.30
0.0021
*
*
0.0040
0.000032
0.16
0.0011
$/*••*
0.0020
0.000014
0.68
0.00049
*
0.19
0.0014
Below measurable concentrations.
Graphite Anode plant.
Based on one 72-hour composite or the average of three 24-hour composites.
216
-------
TABLE 11-24. S»«RR3f OF RAW WASTE LOADINGS AT SCREENING AND VERIFICATION JETAI* ANOCE PLANTS
M
SUBCATBQORY
CHLORINE DIAPHRAGM CELL
Loading
(kgAkg)
Pollutant
Antinony
Arsenic
Cadmium
Chromium
Copper
Lead
Ifercury
Nickel
Silver
Zinc
min.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
00077
0019
00041
0042
0035
00090
00016
0066
00021
016
avg.
0.
0.
0.
0.
0.
0.
0.
0.
0.
2.
00077
0084
00076
59
12
094
00030
31
00021
1
max.
0
0
0
2
0
0
0
1
0
8
.00077
.020
.0014
.8
.19
.37
.00044
.1
.00021
.0
Unit Loading
(kgAkg)
min.
0
0
0
0
0
0
0
0
0
0
.0000033
.0000064
.0000014
.000017
.000014
.0000039
.0000007
.000010
.0000017
.000054
avg.
0.0000033
0.000017
0.0000032
0.00096
0.00020
0.00016
0.0000012
0.00057
0.0000007
0.00078
*Ni
I
max. (01
0.0000033
0.000030
0.0000061
0.0046
0.00064
0.00060
0.0000018
0.0018
0.0000007
0.0021
nrber of
Plants
reraged
it of 5)
1
3
3
5
5
5
2
5
1
4
* Cnly those plants where the pollutant was observed at measurable concentrations.
-------
Pollution Abatement Options
Toxic Pollutants of Concern
Lead occurs in high concentrations in the cell room wastewaters of
chlorine plants using lead anchored graphite anodes. Other toxic
metals often found in significant concentrations at diaphragm cell
plants include arsenic, cadmium, chromium/ copper, nickel, and zinc.
Antimony, mercury, and silver were also detected but at concentrations
that are not treatable. These metals are not considered further. The
sources of these metals may be raw material impurities or corrosion
products from the reaction between chlorine or acid and the process
equipment materials of construction.
Toxic organic compounds also occur in wastewaters from graphite anode
plants because of the attack of chlorine on the anode material. They
appear primarily in waste streams associated with the purification of
chlorine.
Asbestos occurs in all wastewaters from diaphragm cell plants, and in
large quantities in cell room wastewaters when cells are cleaned and
repaired.
Prevailing Control and Treatment Practices
Specific control and treatment practices at ten plants were described
above. The prevailing practices at diaphragm cell plants are to
control asbestos wastes by settling or filtering cell wash waters and
to neutralize and settle all wastewaters 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 lead anchored 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 #195, 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 wastewater discharged is approximately 5.7
rnVday. The organic loading in this waste is not known, however, if
218
-------
TABLE 11-25. TOXIC METAL CONCENTRATIONS AND LORDS IN CELL ROOM WASTE WATERS AT SCREENING AND
VERIFICATION PUNTS/ mg/1 \ (1)
\ Jcgi/kkg /
Pollutant
014
261
Plant t
738A
738B
736
967**
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Zinc
*
0.010
0.0000001
*
0.94
0.000014
0.53
0.0000075
0.26
0.0000039
*
54
0.00081
*
*
*
0.17
0.0000064
0.037
0.0000014
1.9
0.000071
17
0.00064
2.0
0.000075
*
22
0.00081
0.018
0.0000007
1.5
0.000054
0.050
0.00000 01
*
*
*
0.24
0.000042
0.044
0.0000077
0.0030
0.0000005
*
*
0.046
0.0000080
*
*
*
0.075
0.000012
0.38
0.000061
0.11
0.000018
*
0.061
0.0000098
*
0.46
0.000074
0.038
0.0000031
0.17
0.000014
*
0.54
0.000044
1.1
0.000090
0.047
0.0000038
0.0020
0.0000002
0.67
0.000055
*
0.5B
0.000048
0.41
0.00015
0.45
0.00017
0.016
0.0000059
0.086
0.000032
2.4
0.00089
<55>
0.0010
0.0000004
0.36
0.00013
*
0.92
0.00034
Below detection limits
** Graphite anode plant
(1) Based on one 72-hour oonposite or the average of three 24-hour conposites.
-------
TABLE 11-26.
RAW WASTE TOXIC METALS CCNGEMTRATICN AND LOADS IN PROCESS STREAMS OTHER THAN rETiT, ROOM WASTES
FROM SCREENING AND VERIFICATION PLANTS (1)
M
O
Pollutant
Antinomy
Arsenic
Cadmium
Chromium
Copper
Lead
ffercury
Nickel
Silver
Zinc
f 014
*
*
0.0020
0.0000018
*
0.0040
0.0000036
*
0.0020
0.0000018
0.0030
0.0000027
*
*
I738A
*
*
*
0.53
0.0046
0.041
0.00035
0.083
0.00060
*
0.21
0.0018
*
0.29
0.0021
Plant!
I738B
(mg/11
(kg/kkq)
*
0.011
0.000019
*
0.065
0.00011
0.094
0.00016
0.11
0.00019
*
0.067
0.00012
*
0.058
0.00010
1736
*
*
0.038
0.0000062
*
0.090
0.000015
*
0.0030
0.0000005
*
4.3
0.00070
1967
*
0.29
0.0020
*
*
0.030
0.00020
0.40
0.0027
0.0020
0.000014
0.052
0.00035
*
0.15
0.0010
Avg
*
0.15
0.0010
0.020
0.0000040
0.29
0.00014
0.043
0.00014
0.0020
0.0000054
0.088
0.00072
*
1.5
0.0037
Below detection limits
(1) Based on one 72-hour composite or the average of three 24-hour ccrrposites.
-------
TABLE 11-27.
RAW WASTE TOXIC ORGANICS AT A GRAPHITE ANODE PLANT
StBCATEGQRY CHLORINE DIAPHRAGM CRT.T.
Pollutant
benzene
carbon tetrachloride
1, 2-dichloroethane
1,1, 1-trichloroethane
hexachloroethane
1,1, 2- trichloroethane
1,1,2, 2- tetr achloroethane
chloroform
1, 1-dichloroethylene
2 , 6-dinitrotoluene
roethylene chloride
bromoform
dichlorobrcnanethane
chlorodibrcmccnethane
hexachlorcbutadiene
bis ( 2-ethylhexyl) phthalate
di-n-butyl phthalate
tetrachloroethlene
toluene
trichloroethylene
Average
Concentration*
(mg/1)
0.00040
0.023
0.079
0.00014
0.010
0.00040
0.000044
0.085
0.000026
0.000026
0.00056
0.000063
0.035
0.0020
0.0040
0.00075
0.00078
0.036
0.0030
0.020
Load
(kg/day)
0.0011
0.066
0.23
0.00040
0.029
0.0011
0.00013
0.24
0.000074
0.000074
0.0016
0.00018
0.10
0.0057
0.011
0.0022
0.0022
0.10
0.0086
0.0057
Flcw^proportioned concentration
221
-------
TABLE 11-28. RAW WASTE TOXIC ORGANICS BY WASTE WATER SOURCE AT A GRAPHITE
ANOCE PLANT
SUBCATEGQRY CHLORINE DIAPHRAO1 rKT.T.
Stream
Cell building wastes
Caustic filter backwash
Brine fi.1t.ftr backwash
Cell wash
Chlorine condensate and
Spent H2SO4
Scrubber waste
Totals
Total Toxic
Qrganics
(mgA)
0.126
0.057
0.0030
0.20
2.2
0.81
0.30*
Total Toxic
Qrganics
(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
222
-------
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
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 #195, where the volume of wastewater 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 stream 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.
Process Modifications and Technology Transfer Options
A. Anode Material
The use of metal anodes rather than graphite anodes increases
cell power efficiency and greatly reduces the pollutant loads of
toxic organics and, in many cases, lead in plant wastewaters.
Approximately half of the diaphragm cell production of chlorine
is now by metal anodes.
B. 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 wastewater 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 wastewater 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.
C. Diaphragm Material
Although not in full scale use at any U.S. chlorine plants,
modified diaphragms have been developed which can reduce power
223
-------
consumption and minimize asbestos discharges. ' The polymer
modified asbestos diaphragm consists 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 diaphragm replacement, the discarded material is
produced in stablized pieces instead of loose asbestos fibers.
Final disposal is thus safer and easier.
D. Liquefaction of Chlorine
Utilization of high pressure and refrigeration for chlorine
recovery will reduce the chlorine content in tail gases.
E. 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, are effective.
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.).
A. 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.
B. Leaks and Spills
Provisions can be made in cell room areas to control and collect
the leaks or spills contaminated with lead or asbestos.
C. Contaminated Solids
Asbestos waste and precipitated metals wastes should be stored in
a lined pond or disposed of in a secure landfill.
Advanced Treatment Technologies
The methods available and currently used in the industry for the
removal of lead and other toxic metals from plant wastewaters include
hydroxide or carbonate precipitation followed by settling or
224
-------
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.
Selection of Appropriate Technology and Equipment
Technologies for Different Treatment Levels
A. 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 anodes to metal anodes will have residual
effects for an extended time after the change—possibly as long
as two years. The asbestos contaminated cell room wastes are
treated with a flocculating agent and a filter aid prior to
f i1trat ion. The sol ids are removed to a landf ill and the
filtrate is sent to a holding tank where it is 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 settling and the filtrate is combined with other
process waste streams such as chlorine condensate,. tail gas
scrubber water, caustic filter backwash and barometric condenser
blowdown waters found to be contaminated with toxic metals at
levels usually below the limits of treatability by alkaline
precipitation. The combined flow is sent to a polishing pond for
additional clarification prior to 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 1.1-15.
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
wastewaters associated with diaphragm cell plants using graphite
anodes. Note that the industry is concerned about the discharge
of asbestos and has taken steps to reduce this discharge.
Treatment level 1 has the effect of reducing total asbestos
discharges even though we do not set specific limits. Plants
utilizing metal anodes or nonlead containing graphite anode cells
are expected to have lower levels of toxic metal emissions and
may not require alkaline precipitation to meet the proposed BPT
limitations. All direct discharge plants in the industry
presently have BPT or equivalent treatment technology or meet the
limits without treatment.
225
-------
B. Level 2 (BAT)
The objective of Level 2 treatment technology is to achieve, at a
reasonable cost, a greater degree of 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 filtered process wastewater
discharge is also included in Level 2 (BAT) treatment. This
assumes treatment by sulfur dioxide or bisulfite to remove total
residual chlorine.
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 BAT limits derived from published treatability data.
In addition, two plants are known to practice dechlorination of
their effluent. The flow diagram for Level 2 is shown in Figure
11-16.
C. 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 IT-17. This option
was not selected due to its relatively high cost per- pound of
additional metal removal obtained. Cost estimates for this level
can be obtained from the Development Document (60).
Equipment for Different Treatment Levels
A. Equipment Functions
Conventional sludge dewatering by filter press is used for
asbestos sludge before disposal and dual-media filter backwash is
returned to the influent surge tank. Level 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.
B. 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
226
-------
reacted with ferrous sulfate, good ventilation is essential to
avoid the hazards associated with hydrogen sulfide gas.
C, Solids Handling
For all three levels of treatment, brine mud solids are
accumulated in lined lagoons on-site. Abestos solids and
precipitated metals wastes are to be sent to suitable chemical
landfills.
Treatment Cost Estimates
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.
A. 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 19,000, 95,500, and 191,000
kkg per year were selected to represent the subcategory
production range.
B. Wastewater Flow
Based on industry flow data (Table 11-19), 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 tai1 gas scrubber wastes. For treatment cost
estimates at all levels of treatment the following flow basis was
used.
1. A brine mud flow of 0.45 mVkkg is sent to lagoons for
solids removal. Solids are disposed of on-site and the
clarified effluent is recirculated to the process.
2. 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 mVkkg to be treated for metals removal. Included in
this flow is the cathode or cell wash waters which are
heavily laden with asbestos. These wash waters are sent to
asbestos removal at a flow rate of 0.07 mVkkg, and then to
treatment with the other wastewaters.
3. Other process wastewater sources account for an additional
7,6 mVkkg which is combined with effluent from the
227
-------
to
03
o
CEU. KXH
HUSK wow
(M3AL UUmilHMlD}
IIOUttNQ WNK
FUmXATIMl
KSXt
OStBDW
i i*sres
I
(HMOl)
tMOIMS 0MC
•*•
L
/
a
pcuatmo
KM)
r '
UKCflU. I
_ _ J
--<*)
I
^rv-i
pll iiiuiiilurlnu, Hiiw rnuiillornm
and namjilc
Figure 11-15. Level 1 waste water treatment for chlorine - diaphragm cell subcategory.
-------
K>
to
CELL BOOM
WiSTE Nh'
mxojuriw:
*CEWT
£*"• BOM
wsi*£ ' ' '•"
(ASBESTOS
U --
RtraCXf TO HOCESS
cniiDt PHXESS
tROTB HMEHS
30-, Oft BISOU-TIE
mi.llu.-injj, lluw .
Figure 11-16. level 2 waste water treatment for chlorine - diaphragm cell subcategory,
-------
ro
w
o
BRINE
MO
D SOU ASII
r
nfaoiunna
KSNT
I (ASBtiTOfl
' CGHDIHINKIU)}
fZ±
RKSOBW
FCPCESB
•*fl^
aooiiM miix PKJCESS
MSTC MITERS
SUOX
HOIOING
™"
30j OH BISU*'m
1 T
-LQ-^aHjmr
1
1
1
u» I I
I UMflU. \
Figxire 11-17. Level 3 waste water treatment for chlorine - diaphragm cell subcategory.
-------
treatment of wastes from the cell room and cathode wash
areas. This brings the model plant total flow rate to an
estimated 8.8 mVkkg. The final, combined process waste
flow is either clarified and discharged as in Level \ (BPT)
treatment or clarified, passed through dual-media
filtration, and dechlorinated prior to discharge as in Level
2 (BAT) treatment.
C. 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 which is disposed of
on-site. Asbestos from cell wash operations and precipitated
solids from metal treatment generate a solid waste of 0.83 kg/kkg
that is disposed of off-site.
D. 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 wastewater is then returned to the
process and introduced to the brine purification unit or sent to
the treatment unit. The quantity of wastewater 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 process wastewater discharges, the cost for the
dechlorination of such streams using sulfur dioxide has been
included because this is the treatment method on which control of
total residual chlorine is based.
1. Chlorinated Organic Wastes
231
-------
The chlorine-bearing waste streams at graphite anode plants
are also those streams carrying the highest concentrations
of toxic organics as indicated in Table 11-28. Section 11
discussed the techniques used to recover and remove organics
from waste streams at Plant #195 and the fact that organic
contaminated streams can exist as either high volume-low
concentration or low volume-high concentration depending on
plant specific factors. Costs for removing organics
(including implementation of RCRA regulations) are not
included in the model plant cost estimates because organics
are not limited in the regulation. Organics 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 #195) which already has a vacuum
vaporizer, would be under $10,000 for modification of the
existing equipment. Steam costs could vary from $1,000 to
$5,000 per year. If a vaporizer is not in place, a steam
stripper to process 5 to 30 mVweek 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 #967) cannot be
reliably estimated in the absence of specific treatability
data on the waste streams in question.
Alternatively, incineration of the chlorinated organic
residuals is an effective means of destroying and disposing
of this material provided that adequate measures are taken
to control the release of HC1 to the atmosphere..
A process evaluation should be made to determine the most
efficient means for isolating and collecting the organic
bearing waste streams prior to treatment.
Incidental removal of chlorinated organics will occur with
the application of model plant treatment levels previously
presented. Such removal, however, is expected to be erratic
and therefore cannot be predicted. Because raw waste
concentrations of these organics vary considerably depending
on plant practices and are marginally treatable at times,
applicable control and treatment technologies will need to
be assessed on a case-by-case basis.
For these reasons, the Agency is not providing specific
numerical discharge limitations for organic pollutants, but
is providing guidance for evaluating control options that
could be applied in the industry.
232
-------
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 Tables 11-29, 11-30
and 11-31. The cost of Level 2 (BAT) is incremental over Level 1
(BPT) costs and provides for higher effluent quality with respect to
toxic pollutants.
Table 11-32 presents a summary of the unit cost distribution between
amortization and operation and maintenance components.
Basis For Regulations
Basis for BPT Limitations
BPT regulations are currently in effect for the diaphragm cell
chlorine subcategory, 40 CFR 815.62(b). The Agency is revising the
limitations, however, based on an increased unit flow rate.
A. Technology Basis
For BPT, the Agency is setting limitations based on equalization,
alkaline precipitation and settling of lead and asbestos-bearing
wastes and neutralization and settling of all wastewaters before
discharge. All diaphragm cell chlorine plants are known to be
using this technology (Level 1) or its equivalent.
B. Flow Basis
As described in Section 11, wastewater 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 11-19, 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 11-33.
Thus the total unit flow discharge used in the development of
effluent limitations is 8.8 mVkkg.
C. Selection of Pollutants to be Regulated
The selection of pollutants for which specific effluent
limitations are being established is based on an evaluation of
raw waste data from screening and verification sampling and on
the treatability of toxic pollutants.
Table 11-34 presents the achievable concentrations of toxic metal
pollutants (found at detectable levels in raw waste streams)
233
-------
using the available treatment technology options. Based on
literature treatability data presented in Section 8 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.
Based on the occurrence of treatable levels of specific toxic
metals, arsenic, cadmium, chromium, copper, lead, nickel and zinc
are selected as candidate toxic pollutants for BPT regulations.
Antimony, mercury, and silver were detected but at less than
treatable levels.
Consideration of (1) the raw waste concentrations in Table M-34
and (2) Section 8 on the control parameters for hydroxide
precipitation, leads to the selection of copper, lead and nickel
as toxic pollutants to be regulated. The operation of an
hydroxide precipitation system at a pH of about 9.5 should be
suitable for the removal of the metals that are present at
treatable levels. Arsenic, cadmium, chromium and zinc are
included for guidance but no limits are set because control of
copper, lead and nickel will adequately control these other
metals.
D. 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 wastewaters. Guidance is
presented as concentrations (mg/1) only.
BPT limitations, which apply at the combined process wastewater
discharge for TSS and toxic metals and at the plant effluent for
pH, are presented in Table 11-35.
1. Conventional Pollutants
a. pH
The treated effluent is to be contolled within the
range of 6.0 to 9.0. This limitation is based on the
data presented in Appendix B of the proposed
Development Document (60) and the JRB Study (52).
b. TSS
The BPT limitations for TSS are based on a summary of
monitoring data from Plant #207 (3). The long-term
average discharge load of 0.30 kg/kkg is used to
develop discharge limitations. Because variability
234
-------
TABLE 11-29. MODEL PLANT TREATMENT COSTS
Subcategory Chlorine - Diaphragm cell
Production 19,100 metric tons per year
A. INVESTMENT COST BPT BATa
Site development 30,000 0
Equipment 136,000 43,000
Monitoring equipment 20,000 0
Subtotal •• -. t 186,000 43,000
Contractor's 0 & P° 27,900 6,450
Subtotal 213,900 49,450
Engineering » 42,780 9,890
Subtotal 256,680 59,340
Contingencies 25,668 5,934
Subtotal 282,348 65,274
Land 27,000 0
TOTAL INVESTMENT COST 309,348 65,274
B. OPERATION AND
MAINTENANCE COST
Labor and supervision 112,000 14,000
Energy 2,200 800
Chemicals 1,500 1,500
Maintenance 28,235 6,527
Taxes and insurance 9,280 1,958
Residual waste disposal .... 5,800 0
Monitoring , analysis
and reporting 15,000 7,500
TOTAL OPERATION AND
MAINTENANCE COST 1*74,015 32,286
C. AMORTIZATION OF
INVESTMENT COST 45,938 10,620
TOTAL ANNUAL COST 219,953 42,906
a Represents the incremental cost above that for BPT treatment
Overhead and Profit
235
-------
TABLE 11-30. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Chlorine - Diaphragm cell
95,500 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment
Subtotal ..............
Contractor's o & Pb
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
BPT
85,000
261,000
20,000
366,000
54,900
420,900
84,180
505,080
50,508
555,588
69,000
624,588
($)
112,000
4,900
7,500
55,559
18,738
29,000
15,000
242,696
BAT
0
70,000
0
70,000
10,500
80,500
16,100
96,600
9,660
106,260
0
106,260
28,000
1,300
7,500
10,626
3,188
0
7,500
58,115
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
90,394
333,091
17,289
75,404
a Represents the incremental cost above that for BPT treatment
k Overhead and Profit
236
-------
TABLE 11-31. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Chlorine - Diaphragm cell
191,000 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ...
Subtotal
Contractor's 0 & Pb.....
Subtotal
Engineering
Subtotal ,
Contingencies ,
Subtotal
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
($)
BPT
140,700
353,000
20,000
513,700
77,055
590,755
118,151
708,906
70,891
779,797
129,000
908,797
TOTAL OPERATION AND
MAINTENANCE COST
112,000
8,000
15,000
77,980
27,264
58,000
15,000
313,244
BAT
0
108,500
0
108,500
16,275
124,775
24,955
149,730
14,973
164,703
0
164,703
28,000
1,800
15,000
16,470
4,941
0
7,500
73,711
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
126,873
440,116
26,797
100,509
a Represents the incremental cost above that for BPT treatment
b Overhead and Profit
237
-------
TABLE 11-32.MODEL PLANT UNIT TREATMENT COSTS
Subcategory Chlorine - Diaphragm cell
Annual Treatment Costs (S/kkg)
COST ITEM
PRODUCTION
(kkg/yr)
LEVEL OF TREATMENT
BPT
BAT*
Annual Operation
and Maintenance
Annual
Amortization
Total Annual
Cost
19r100
95,500
191,000
19,100
95,500
191,000
19,100
95,500
191,000
9.11
2.54
1.64
2.41
0.95
0.66
11.52
3.49
2.30
1.69
0.61
0.39
0.56
0.18
0.14
2.25
0.79
0.53
*Represents the incremental cost above BPT
238
-------
factors for TSS were not available for this plant,
factors obtained from the hydrofluoric acid subcategory
are used. In that subcatergory, 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:
(0.30 kg/kkg) (1.7) = 0.51 kg/kkg
and a 24-hour maximum limitation 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 mgVkkg to obtain a
maximum 30-day average concentration of 58 mg/1 derived
as follows:
0.51 kq/kkq x 1000 mq/1 = 58 mg/1
8.8 mVkkg kg/m^
A 24-hour maximum concentration of 120 mg/1 is derived
from the variability factor ratio (VFR: 3.5/1.7 = 2.1)
as follows:
2.1
(58mg/l
120 mg/1 (carrying two significant
figures)
2.
From the above data, the implicit long-term average
concentration is:
120 mq/1 = 34 mg/1
3.5
The average concentration of TSS after lime addition
and decantation in the Treatability Studies document
(61) was 99 mg/1. This average was calculated from a
limited number of experiments on a nonoptimized system.
Despite these shortcomings, the average falls in
between the maximum 30-day average and the 24-hour
maximum concentration limit for TSS.
Toxic Pollutants
a. Lead
The 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.
239
-------
TABLE 11-33
OF UNIT FLOWS AT DIAPHRftCM CKT.T. PLANTS
SUBCATEGORY
CHLORINE DIAPHRAGM CKT.T.
Stream Description
Unit Flow
Data
Source
Cell room and cell
wash wastes
Chlorine condensate
Tail gas scrubber waste
Caustic filter wash
Brine filter wash
Caustic moling blowdown
Spent sulfuric acid
1.2
0.78
0.11
5.4
0.45
0.86
0.010
Graphite anode plant
Graphite anode plant
Graphite anode plant
Graphite anode plant
Graphite anode plant
Metal anode plants
average
Metal anode plants
average
Total Unit Flow Discharge 8.8 m3Akg
240
-------
TABLE 11-34. COMPARISON OF RAW WASTE CONCENTRATIONS
OF TOXIC POLLUTANTS WITH TREATABILTTY
SUBCATEGORY
Pollutant
Arsenic
Antimony
Cadimium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Zinc
CHLORINE (DIAPHRAGM CELL)
Literature-based
Treatability (D
BPT
0.50
0.80
0.10
0.10
0.50
0.30
_(2)
0.20
0.40
0.50
BAT
0.50
0.40
0.050
0.050
0.40
0.050
»(2)
0.10
0.20
0.40
Maximum Plant
Raw Waste
Average
(mg/1)
0.30
d.oii
0.037
1.9
17
21
0.0030
22
0.018
3.0
(1) Literature-based treatability estiinates from Table 8-11.
(2) Treatability with this technology not available.
241
-------
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 maximum
30-day average limitation for lead is then obtained by
multiplying the variability factor for 30-day averages
by the Jong-term average load; i.e., 1.6 x 0,0064
kg/kkg = 0-010 kg/kkg. Similarly the 24-hour maximum
limitation is 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 load (L),
and unit flow (Q).
C (mg/1) *
L (kq/kkq) x 1000 mg/1
Q (mVkkg) kg/m3
Thus the concentration basis for the maximum 30-day
average for lead is:
0.010 kq/Kkg x 1QOQ mq/1 =1.1 mg/1
8.8 mVkkg kg/m*
The concentration basis for the 24-hour maximum
limitation is obtained by applying the variability
factor ratio (VFR) of 2.6 to the maximum 30-day average
concention:
(1.1 mg/1) (2.6) = 2.9 mg/1
From the above data, the implicit long-term average
concentration is:
2.9 mq/1 * 0.70 mg/1
4. 1
The average concentration of lead after lime addition
and decantation in the Treatability Studies document
(61) was 28 mg/1. This value is much higher than the
values calculated above from the long-term monitoring
data in Appendix A. The conclusion in the Treatability
Studies document, that there was extremely poor
settling of the metal hydroxide sludge, seems
justified.
Monitoring data from six diaphragm cell plants
presented in Table 11-36 indicate that plants using
metal anodes are meeting the BPT lead limitations. One
of two graphite anode plants is meeting the
limitations.
242
-------
TABLE 11-35. EFFLUENT LIMITATIONS
Chlorine - Diaphragm Cell
Best Practicable Control Technology Currently Available
Waste Water Flow: 8.8
Pollutant Subcategory Daily
Performance Variability
(mg/1) Factor
30 -day avg.
Variability
Factor
Concentration
(mg/1)
Max. 24-hr
30-day max.
avg.
-Effluent Limit
(kg/kkg)
Max.
30-day
avg.
24-hr
max.
Conventional Pollutants:
TSS
Toxic Pollutants:
34
(1)
3.5/1.7
58
120
0.51
1.1
Arsenic
Cadmium
Chromiun
Copper (5)
Lead<5>
Nickel (5)
Zinc
0.50 (2)
0.10<3>
0.32(3)
0.50(2)
0.70(1)
0.40 (3>
0.80(3)
4.V1-6
4.1/1.6
2.2/1.1
4.1/1.6
4.1/1.6
4.V1.6
4.V1.6
0.80
0.16
0.35
0.80
1.1
0.64
1.3
2.1
0.41
0.70
2.1
2.9
1.6
3.3
__(4) _J4)
__(4) __(4)
_(4) _(4)
0.0070 0.018
0.010 0.026
0.0056 0.014
~(4) "(4)
(1) Calculated from long-term monitoring load data.
(2) Lower limit of literature treatability for lime/clarification technology
from Table 8-11.
(3) Estimated achievable long-term average concentration from Table 8-13,
(4) No load limits; concentration limits are provided for guidance purposes.
(5) Also applicable for PSES limitations.
243
-------
The limitations and guidelines for the additional toxic
pollutants are derived from two sources—industrial
wastewater treatment system performance data (Tables
8-12 and 8-13) and literature-based treatability
estimates (Table ^8-11). A comparison of these values,
along with effluent data from the sampling of two
diaphragm cell plants, is presented in Table 11-37.
The concentration bases are derived from the industrial
wastewater treatment system performance data unless the
observed concentrations are below literature-based
treatability estimates. In such cases, the lower limit
of the applicable treatability level is used.
b. Copper
The estimated achievable long-term average
concentration for copper (0.40 mg/1 from Table 8-13} is
lower than the literature-based treatability estimate
(0.50 mg/1 from Table 8-11). Therefore, the latter
value is used in place of the long-term average.
Because no long-term monitoring data is available for
copper in this industry, the variability factors used
for lead are employed to obtain the concentration
limits.
The 24-hour maximum concentration is:
(0.50 mg/1) (4.1 ) = 2.1 mg/1
The maximum 30-day average concentration is:
(0.50 mg/1) (1.6) = 0.80 mg/1
The load limitations for copper (kg/kkg) are calculated
based on the unit flow rate of 8.8 mVkkg, thus:
(2.1 mg/1) (8.8 mVkkg) ( kq/m3 ) = 0.018 kg/kkg
(1000 mg/1)
for the 24-hour maximum limit. The maximum 30-day
average limit is calculated similarly, i.e.,
(0.80 mg/1) (8.8 mVkkg) ( kg/m? ) = 0.0070 kg/kkg
(1000 mg/1)
Nickel
The nickel limits are based on the estimated achievable
long-term average concentration of 0.40 mg/1 which is
obtained from Table 8-13. Because no long-term
monitoring data is available for nickel in this
industry, the variability factors used for lead are
employed again.
244
-------
The 24-hour maximum concentration is:
(0.40 mg/1) (4.1) = 1.6 mg/1
The maximum 30-day average concentration is:
{0.40 mg/1) (1,6) = 0.64 mg/1
The load limitations for nickel (kg/kkg) are calculated
based on the unit flow rate of 8.8 mVkkg, thus:
(1.6 mg/1) (8.8 mVkkg} ( kq/m^ ) = 0.014 kg/kkg
(1000 mg/1)
for the 24-hour maximum limit. The maximum 30-day
average limit is calculated similarly, i.e.,
(0.64 mg/1) (8.8 mVkkg) ( kg/m* ) = 0.0056 kg/kkg
(1000 mg/1)
The average concentration of nickel after lime addition
and decantation in the Treatability Studies document
(61) was 0.41 mg/1. This is esentially equal to the
estimated achievable long-term average concentration
which was the basis of the nickel limits.
Arsenic
The industrial wastewater treatment system performance
for arsenic (0.080 mg/1 from Table 8-12) is lower than
the literature-based treatability estimate (0.50 mg/1
from Table 8-11). Therefore, the latter value is used
in place of a long-term average to arrive at the
arsenic guidance. Because no long-term monitoring data
is available for arsenic in this industry, the
variability factors used for lead are employed again.
The 24-hour maximum concentration is:
(0.50 mg/1} (4.1) = 2.1 mg/1
The maximum 30-day average concentration is:
(0.50 mg/1) (1.6) = 0.80 mg/1
Cadmium
The cadmium guidance is based on the estimated
achievable long-term average concentration of 0.10 mg/1
which is obtained from Table 8-13. Because no
long-term monitoring data is available for cadmium in
245
-------
TABLE 11-36. LEAD AND TSS DISCHARGES FROM SELECTED DIAPHRAGM CELL
CHLORINE PLANTS*1)
SUBCATEGORY
Plant
#589*
#738*
#261*
#014*
#967(3>
#207
Plant
#207
#014*
CHLORINE - DIAPHRAGM CELL
Lead Discharge
(kgAkg)
Average
0.0020
0.0010
0.0025<2)
0.0060<2),(4)
0.0064
0.021(2)
TSS Discharge
(kg/kkg)
Average
0.30*2)
2,8(2), (4)
Maximum
0,030
0.015
0.019
NA
0.026
0.054
Maximum
0.57
NA
) From Reference 3
* ) Plant has "once-through" barometric condenser water.
Long Term Data Appendix A.
Total plant discharge including other products. Recent information from
plant indicates current discharge levels much diminished from data given in
Reference 3.
Plants with metal Anodes
NA: Not available
246
-------
TABLE 11-37.
COMPARISON OF TOXIC POLLUTANTS IN
DIAPHRAGM CKT.T. PLANT EFFLUENTS
WITH TREATABILI'IY
Pollutant
Literature-based
Treatability(1)
(mg/1)
Industrial Pfeste Metal Anode Graphite Anode
Water Treatment
System
Performance
(mg/1)
Plant #261
Effluent(2)
(mg/1)
Plant #967
Effluent*3'
(mg/1)
BPT
BAT
BPT
BAT
Arsenic
Cadmiizn
Chromium
Copper
Lead
Nickel
Zinc
0.50
0.10
0.10
0.50
0.30
0.20
0.50
0.50
0.050
0.050
0.40
0.050
0.10
0.40
0.080(4)
0.10(5)
0.32<5>
0.40(5)
0.15<5>
0.40 (5)
0.80(5)
0.17<4>
0.076(4)
0.16(5)
0.3*™
0.19(4>
0.30(5)
0.20C5)
0.12
0.0040
< 0.050
< 0.025
NA
< 0.050
< 0.025
0.30
< 0.015
< 0.050
0.031
NA
< 0.050
0.15
(1) Literature-based treatability estimates from Table 8-11.
(2) Cell wash waste filtered with coagulant to remove asbestos.
(3) Flow-proportioned average discharge, consisting of lead treatment
discharge and untreated filter backwashes, condensates, and scrubber
wastes.
(4) Data from Table 8-12 (value presented or average of values presented).
(5) Estimated achievable long-term average concentrations from Table 8-13.
NA = Not available
247
-------
this industry, the variability factors used for lead
are employed again.
The 24-hour maximum concentration is:
(0.10 mg/1) (4.1) = 0.41 mg/1
The maximum 30-day average concentration is:
(0.10 mg/1) (1.6) = 0.16 mg/1
Chromium
The chromium guidance is based on the estimated
achievable long-term average concentration of 0.32 mg/1
which is obtained from Table 8-13. A daily variability
factor of 2.2 and a 30-day variability factor of 1.1
are obtained from the Treatability Studies document
(61).
The 24-hour maximum concentration is:
(0.32 mg/1) (2.2) = 0.70 mg/1
The maximum 30-day average concentration is:
(0.32 mg/1) (1.1 ) = 0.35 mg/1
The average concentration of chromium after lime
addition and decantation in the Treatability Studies
document (61) was 0.071 mg/1. This value is much lower
than the estimated achievable long-term average
concentration used above. The large difference is
attributable to the very low average chromium
concentration of raw waste in the Treatability Studies
document (0.089 mg/1).
Zinc
The zinc guidance is based on the estimated achievable
long-term average concentration of 0.80 mg/1 which is
obtained from Table 8-13. Because no long-term
monitoring data is available for zinc in this industry,
the variability factors used for lead are employed
again.
The 24-hour maximum concentration is:
(0.80) (4.1) = 3.3 mg/1
The maximum 30-day average concentration is:
248
-------
(0.80) (1.6) - 1.3 mg/1
Basis for BAT Effluent Limitations
Previous BAT regulations called for no discharge of process wastewater
pollutants. The regulations were remanded. The newly promulgated BAT
regulations provide for the discharge of pollutants following
appropriate treatment of process wastes.
A. Technology Basis
Utilizing the cost estimates presented in this report the Agency
has analyzed the cost effectiveness of the base level system
(BPT) and an advanced level option (Level 2) 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 the BAT limitations.
For BAT the Agency is promulgating limitations based on BPT
technology with the addition of dual-media filtration (Level 2)
and dechlorination of all process wastewaters. 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
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.
B. Flow Basis
The flow basis for BAT limitations is the model plant total
discharge of 8.8 mVkkg. (See Table 11-33).
C. Selection of Pollutants to be Regulated
For BAT regulations, the Agency 'has selected the same three toxic
metals identified in the BPT regulations (copper, lead, and
nickel) and total residual chlorine. The other four toxic metals
(arsenic, cadmium, chromium, and zinc) are again included for
guidance.
D. Basis of Pollutant Limitations
For BAT regulations, the Agency is promulgating more stringent
controls on the discharge of the first three toxic metals 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. Plants whose raw waste has toxic
249
-------
metal concentrations below literature treatability levels,
perhaps as a consequence of installing metal anodes, would only
have to sample periodically to show they are meeting the
regulations. Final BAT limitations, which apply at the combined
process wastewater discharge for toxic metals and at the plant
effluent for residual chlorine, are presented in Table 11-38.
1. Nonconventional Pollutant
Chlorine: Total residual chlorine limits are based on the
data from Appendix A that were used to calculate BAT limits
for the chlorine mercury cell subcategory. The chlorine
containing waste streams from both technologies are similar
and the methods for chlorine removal are identical. The
long-term average concentration is 0.64 mg/1. The
variability factor for daily measurements is 2.3 and the
variability factor for 30-day averages is 1.4.
The 24-hour maximum concentration is:
(0.64 mg/1) (2.3) = 1.5 mg/1
The maximum 30-day average concentration is:
(0.64 mg/1) (1.4) = 0.90 mg/1
The load limitations for total residual chlorine (kg/kkg)
are calculated based on the unit flow rate of 8.8 mVkkg,
thus:
1.5 mg/1) (8.8 mVkkg)
kq/m3 ) =0.13 kg/kkg
(1000 mg/1)
2.
for the 24-hour maximum limit. The maximum 30-day average
limit is calculated similarly, i.e.,
(0.90 mg/1) (8.8 mVkkg)
Toxic Pollutants
( kq/m3 ) = 0.0079 kg/kkg
(1000 mg/1)
Dual-media filtration of BPT effluent will significantly
reduce suspended metal precipitates. BAT limitations and
guidelines are derived from two sources—industrial
wastewater treatment system performance data (Tables 8-12
and 8-13) and literature-based treatability estimates (Table
8-11). A comparison of these values is presented in Table
11-37. The concentration" bases are derived from the
industrial wastewater treatment system performance data
unless the observed concentrations are below
literature-based treatability estimates. In such cases, the
lower limit of the applicable treatability level is used.
250
-------
There are insufficient data to calculate daily and 30-day
variability factors for any of the toxic pollutants in this
industry. A daily variability factor of 3.5 and a 30-day
average variability factor of 1.4 will be used to determine
BAT toxic pollutant limitations and guidelines. These
variability factors are the result of evaluating the
Treatability Studies document (61) data from other
industries involving the removal of toxic metals via the
lime/filter method.
a. Copper
The copper limits are based on the literature
treatability level of 0.40 mg/1 (Table 8-11} which is
larger than the estimated achievable long-term average
concentration of 0.30 mg/1 (Table 8-13).
The 24-hour maximum concentration is:
(0.40 mg/1) (3.5) = 1.4 mg/1
The maximum 30-day average concentration is:
(0.40 mg/1) (1.4) = 0.56 mg/1
The load limitations for copper (kg/kkg) are calculated
based on the unit flow rate of 8.8 mVkkg, thus:
(1.4 mg/1) (8.8 mVkkg) ( kq/m^ ) =0.012 kg/kkg
(1000 mg/1)
for the 24-hour maximum limit. The maximum 30-day
average limit is calculated similarly, i.e.,
(0.56 mg/1) (8.8 mVkkg) ( kg/in* ) = 0.0049 kg/kkg
(1000 mg/1)
b. Lead
The lead limits are based on the data in Table 8-12,
The average of the lead values in Table 8-12, 0.19
mg/1, is used in place of a long-term average.
The 24-hour maximum concentration is:
(0.19 mg/1) (3.5) = 0.67 mg/1
The maximum 30-day average concentration is:
(0.19 mg/1) (1.4) = 0.27 mg/1
251
-------
TABLE 11-38. EFFLUENT LIMITATIONS
Chlorine Diaphragm Cell
Best Available Technology
3
Waste Water Flow: 8.8 m /kkg
Pollutant
Treatability
CngA
Daily
Variability
Factor
30-day avg.
Variability
Factor
Concentration
Basis
(mq/1)
Max. 24-hr.
30-day max.
avg.
Ef f luent Limit
(kg/kkg)
Max. 24-hr.
30-day max.
avg.
Nonconventional Pollutant
Total Residual
Chlorine
Toxic Pollutants
0.64(1)
2.3/1.4
Arsenic
Cadimium
Chromium
Copper
Lead
Nickel
Zinc
0.50(2)
0.076(3)
0.16(4)
0.40(2)
0.19(3)
0.30(4)
0.40(2)
4.1/1.6
3.5/1.4
5/1.4
5/1-4
5A-4
5/1.4
3.5A-4
0.90
0.80
0.11
0.22
0.56
0.27
0.42
0.56
1.5 0.0079
2.1
0.27
0.56
1.4
0.67
1.1
1.4
-(5)
-(5)
-(5)
0.0049
0.0024
0.0037
-(5)
0.013
-(5)
-(5)
-(5)
0.012
0.0059
O.(fo97
-(5)
(1) Long-term average concentration from Appendix A.
(2) Lower limit of literature treatability for lime/filter technology from
Table 8-11.
(3) Industrial waste water treatment system performance data from Table 8-12,
(4) Estimated achievable long-term average concentration from Table 8-13.
(5) No load limits; concentration limits are provided for guidance purposes.
252
-------
The load limitations for lead (kg/kkg) are calculated
based on the unit flow rate of 8.8 mVkkg), thus:
(0.67 mg/1) (8.8 mVkkg) ( kg/in* ) = 0.0059 kg/kkg
(1000 mg/1)
for the 24-hour maximum limit. The maximum 30-day
average limit is calculated similarly, i.e.,
{0.27 mg/1) (8.8 mVkkg) ( kq/m3 ) = 0.0024 kg/kkg
(1000 mg/1)
Nickel
The nickel limits are based on the estimated achievable
long-term average concentration of 0.30 mg/1 which is
obtained from Table 8-13.
The 24-hour maximum concentration is:
(0.30 mg/1) (3.5) = 1.1 mg/1
The maximum 30-day average concentration is:
(0.30) (1.4) = 0.42 mg/1
The load limitations for nickel (kg/kkg) are calculated
based on the unit flow rate of 8.8 mVkkg, thus:
(1.1 mg/1) (8.8 mVkkg) ( kq/m3 ) - 0.0097 kg/kkg
(1000 mg/1)
for the 24-hour maximum limit. The maximum 30-day
average limit is calculated similarly, i.e.,
(0.42 mg/1) (8.8 mVkkg) ( kq/m3 ) = 0.0037 kg/kkg
(1000 mg/1)
Arsenic
Because the BPT and BAT literature treatability levels
for arsenic are identical, the BAT guidance for arsenic
is set equal to the BPT guidance.
Cadmium
The cadmium guidance is based on the data in Table
8-12. The cadmium value in Table 8-12, 0.076 mg/1, is
used in place of a long-term average.
The 24-hour maximum concentration is:
(0.076 mg/1) (3.5) = 0.27 mg/1
253
-------
The maximum 30-day average concentration is:
(0.076 mg/1) (1.4) = 0.11 mg/1
f. Chromium
The chromium guidance is based on the estimated
achievable long-term average concentration of 0.16 mg/1
which is obtained from Table 8-13.
The 24-hour maximum concentration is:
(0.16 mg/1) (3.5) = 0.56 mg/1
The maximum 30-day average concentration is:
(0.16 mg/1) (1.4) = 0.22 mg/1
g. Zinc
The zinc guidance is based on the literature
treatability level of 0.40 mg/1 (Table 8-11) which is
larger than the estimated achievable long-term average
concentration of 0.20 mg/1 (Table 8-13).
The 24-hour maximum concentration is:
(0.40 mg/1) (3.5) = 1.4 mg/1
The maximum 30-day average concentration is:
(0.40 mg/1) (1.4) = 0.56 mg/1
BCT Limitations
EPA has determined that the BAT technology for this subcategory is
capable of removing significant amounts of conventional pollutants.
However, EPA has not yet proposed or promulgated a revised BCT
methodology in response to the American Paper Institute v. EPA
decision mentioned earlier. Thus, it is not now possible to apply the
BCT cost test to this technology option. Accordingly, EPA is
deferring a decision on the appropriate BCT limitations until EPA
proposes the revised BCT methodology. However, the Agency has
calculated the expected TSS effluent quality after application of
dual-media filtration and the cost of the additional TSS removal. As
described in Section 3, this cost was calculated to be $0.53 per Ib
TSS removed.
254
-------
Basis for New Source Performance Standards
A. 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 which exclude the use of lead in
cell construction. The conversion to metal anodes has largely
eliminated the source of lead in wastewaters, 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 wastewaters.
B. Flow Basis
The flow basis of 8.8 mVkkg used for BPT and BAT limitations is
conservatively being used for new sources.
C. 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. Table 11-39 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. Application of the information in Section 8
on the control parameters for hydroxide precipitation leads to
the selection of lead as the toxic pollutant to be regulated.
Control of this pollutant within the limitations being set will
assure that other toxic metals are being controlled. Other
metals are presented on a concentration basis for guidance
purposes only.
D. Basis of Pollutant Limitations
For NSPS regulations the Agency is promulgating more stringent
controls on the discharge of lead on the basis of lower raw waste
loads generated at plants using metal anodes. New plants that
have raw waste lead concentrations below estimated industry
long-term averages would only have to sample periodically to show
they are meeting the regulations. The NSPS regulations, which
apply at the combined process wastewater discharge for TSS and
lead and at the plant effluent for residual chlorine and pH, are
shown in Table 11-40.
1
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
255
-------
data presented in Appendix B of the proposed
Development Document (60) and the JRB Study (52).
b. TSS
Limitations for TSS are the same as BPT regulations.
c. Total Residual Chlorine
Limitations for total residual chlorine are the same as
in BAT regulations.
2. Toxic Pollutants
a. Lead
The concentration bases for lead is based on the
estimated achievable long-term average of 0.15 mg/1
(Table 11-39). Variability factors used to obtain BAT
limits for lead are used for the NSPS limits. The
concentration bases for lead are determined as follows:
The 24-hour maximum concentration is:
(0.15 mg/1) (3.5) = 0.53 mg/1
and the corresponding effluent limit is:
(0.53 mg/1) (8.8 mVkkg) ( kq/m3 ) = 0.0047 kg/kkg
(1000 mg/1)
The maximum 30-day average is:
(0.15 mg/1) (1.4) = 0.21 mg/1
and the corresponding effluent limit is
(0.21 mg/1) (8.8 mVkkg) ( kq/m a ) = 0.0018 kg/kkg
(1000 mg/1)
b. Other Toxic Pollutants
Guidance for the other toxic metals are the same as in
BAT regulations.
Basis for Pretreatment Standards
A. Existing Sources
For Pretreatment Standards for Existing Sources (PSES), the
Agency is promulgating the same limitations as for BPT (excluding
256
-------
TABLE 11-39. COMPARISON OF RAW WASTE CHARACTERISTICS AT A NEW METAL ANODE
PLANT WITH TKEATABILITY OF TOXIC METALS
StBCATEGORY
CHLORINE DIAPHRAOt rKT.T.
Pollutant
Arsenic
Cadmium
Chromium
Copper
Lead
Nickel
Zinc
Treatability
0.50^)
0.050(1)
oae(3)
(3)
0.301^
0.15<3>
0.30(3)
0.20 (3)
Concentration(mg/l^
Plant #738B(2)
Raw Waste
0.011
<0.025
0.066
0.12
0.11
0.067
0.093
(1)- Literature-based treatability estimates using BAT technology
of dual media filtration following alkaline precipitation of
metals (Table 8-11)
(2)- Verification sampling at new metal anode facility
(3)- Estimated achievable long-term average from Table 8-13.
257
-------
TSS and pH) based on the identical treatment technology (see
Table 11-35). The pollutants to be limited are copper, lead and
nickel. Pretreatment is necessary because BPT provides better
removal of toxic metals than is achieved by a well-operated POTW
with secondary treatment installed. PSES are being set equal to
BPT, because the POTW achieves equal or greater removal of the
toxic metals remaining after application of BPT (PSES) technology
than is achieved by the filtration required by BAT. Thus, a
filter is not required for PSES because there is no pass-through.
B. New Sources
For Pretreatment Standards for New Sources (PSNS), the Agency is
promulgating the same limitations as for NSPS based on the
identical treatment technology without dechlorination being used
for indirect dischargers (see Table 11-40). Dechlorination is
unnecessary because influent to publicly-owned treatment works is
often chlorinated. The pollutant to be limited is lead. The
other pollutants (copper and nickel) are not being limited based
on the assumption that all new plants will have effluent levels
equal to the levels of a new metal anode plant. Pretreatment is
necessary because NSPS provides better removal of toxic metals
(when present at treatable levels) than is achieved by a POTW and
hence these pollutants would pass through a POTW in the absence
of pretreatment.
258
-------
TABLE 11-40. EFFLUENT LIMITATIONS
Chlorine Diaphragm Cell
New Source Performance Standards
Waste Water Flow: 8.8 m /kkg
Pollutant Treatability Daily
(mg/1) Variability
Factor
30-day avg.
Variability
Factor
Concentration
Basis
(mg/1)
Max. 24-hr.
30-day max.
avg.
Effluent Limit
(kg/kkg)
Max.
30-day
avg.
24-hr.
max.
Conventional
and
Nonconventional
TSS
Total Residual
Chlorine
34
0.64
3.5/1.7
2.3/1.4
58
120
0.90 1.5
0.51 1.1
0.0079 0.013
Toxic Pollutants
Arsenic
Cadmium
Chromium
Copper
Lead (1)
Nickel
Zinc
0.
0,
0,
0.
0,
0.
0.
50
076
16
40
15
30
20
4.
3.
3.
3.
3.
3.
3.
VI.
5/1.
5/1.
5/1.
5/1.
5/1.
5/1.
6
4
4
4
4
4
4
0
0
0
0
0
0
0
.80
.11
.22
.56
.21
.42
.28
2
0
0
1
0
1
0
.1
.27
.56
.4
.53
.1
.70
-(2)
-(2)
-(2)
-(2)
0.0018
-(2)
-(2)
-(2)
-(2)
-(2)
-(2)
0.0047
-(2)
-{2)
(1) Also applicable to PSNS limitations.
(2) NO load limits; concentration limits are provided for guidance purposes.
259
-------
-------
SECTION 12
HYDROFLUORIC ACID INDUSTRY
Industry Profile
General Description
Hydrofluoric acid (hydrogen fluoride or 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. HF is used
in the production of aluminum, in the refining and enriching of
uranium fuel, in 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 prior to
promulgation of these new regulations is given in Table 12-2.
Subcategorization
Hydrogen fluoride is usually used in the production of aluminum
fluoride (A1F3) by reacting with hydrated alumina (A1203»3H20). Two
aluminum fluoride plants are integrated with hydrofluoric acid
production.
For both products (HF and A1F3), process wastewaters 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 fluoro anions. Calcium fluoride (CaF2), generated as
a solid waste, is a disposal problem for both subcategories because of
its moderate toxicity.
However, 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 wastewaters generated from both industries do not have a common
production basis. In addition, the combined manufacture of these
products does not create a unique or unusual situation, either with
regard to the wastewater treatment requirements or compliance with
discharge regulations.
Due to process differences and the variations in the pollutant
loadings for the two industries, different variability factor ratios
were used in obtaining the effluent limitations of the toxic
pollutants. In the case of the hydrofluoric acid subcategory, nickel
and zinc are chosen as the control toxic pollutants for the effluent
limitations. On the other hand, chromium and nickel are the toxic
metals controlled in the aluminum fluoride subcategory. These
261
-------
TABLE 12-1 - SUBCATEGORY PROFILE DATA SUMMARY
SUBCATEGORY HYDROFLUORIC ACID
Total subcategory capacity rate 363,000 kkg/year
Total subcategory production rate 261,800 kkg/year
Number of plants in this subcategory 9
308 Data on file for 8
With total capacity of *
With total production of 177,000 kkg/year
Representing capacity *
Representing production 68 percent
Plant production range:
Minimum 7,300 kkg/year
Maximum 62,000 kkg/year
Average production 22,100 kkg/year
Median production 15,800 kkg/year
Average capacity utilization 83 percent
Plant age range:
Minimum 7 years
Maximum 58 years
Waste water flow range:
Minimum 0 cubic meters/day
Maximum 4,700 cubic meters/day
Volume per unit product:
Minimum 0 cubic metersAkg
Maximum 86 cubic metersAkg
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, 1380.
* Data inconplete because certain plants did not respond to this question.
262
-------
TABLE 12-2. STATUS OF REGULATIONS - EFFLUENT LIMITATION GUIDELINES
SUBCATEGORY
SUBPART
HYDROFLUORIC ACID
H (40CFR 415.80, 3/12/74)
STANDARDS
Product
Process
BPCTCA*
1 7
Max. ' Avg. *•
Parameters ( mg/1 ) ( mg/1 )
BATEA*
1 ">
Max. Avg. *•
(mg/1) (mg/1)
NSPS*
Max. ' Avg.2
(mg/1) (mg/1)
Hydrofluoric
Acid Fluoride (30)
TSS (50)
No discharge
(15) of pwwp3
No discharge
(25) of pwwp
No discharge
of pwwp
No discharge
of pwwp
Section 415.82, 415.83, and 415.85 were remanded and reserved (41 FR 51601,
November 23, 1976).
Max. = Maximum of any one day.
Avg. = Average of daily values for thirty consecutive days.
pwwp — Process wastewater pollutants.
263
-------
differences are due to the different relative concentrations for
chromium and zinc in the raw waste loads from the two products.
Furthermore, the opportunities for drip acid recycle (or the
hydrolysis of complex fluoride prior to treatment) and scrubber water
recycle are a function of plant design and age, rather than product
mix.
An additional solid waste, gypsum (CaS04»2H20), is generated from the
hydrofluoric acid manufacture and supplies enough calcium for adequate
fluoride removal from neutralized scrubber wastewaters generated by
both HF and A1F3 production. However, the applied treatment
technology is essentially the same as that applied by manufacturers of
either product alone.
In view of these considerations, hydrofluoric acid and aluminum
fluoride remain separated as two distinct subcategories.
General Process Description and Raw Materials
HF is the most important manufactured compound of the fluoride 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:
CaF
S
H20
CaC03
Minimum 97.5-98%
Maximum 1.0%
Maximum 0.05%
Maximum 0.1%
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:
Si0
2CaF2 +2H2S04 = SiF4 + 2CaS04 + 2H20
264
-------
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 to produce sulfuric acid is used for the
heat required for HF generation. Thus a part of the sulfuric acid is
supplied as SO3.
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 refigerated
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 wastewater treatment
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 condehsate, 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.
265
-------
OLEDH
SLOSRY NK.TER
ACID
FLOORSPAJt
CM^T™ ,..,,
SCLFATE (GYP SUM)
_
F WB-ccwacr~*
nflrt nas ^^L_
*•!"!••• IMP ^^M^
COOLIMG OR ^
SYSTEM £
KftXER
1 1—
MHDEN
COLLB
9Q
t i
COO
1
DRX1
COX
1
CONDI
:EODE :
FLD<
STOR]
SABI^S
CTIOH
F
•
r mi
LES
f
t POT
r
E BOX
'
OISEK
IYDROGE
IRXDB
USE
"ER, SQL
hCZO t
SOLFOMI
•
B
1
PI
f
C
s
ntzc
ao-
ACZD
RBCY
AC
scum
•UUTJRX
CLE
CD
SBER.
t
: ACID
WATER'
EJECTOR
HtSSEtBOER
TO
TREATMEHT
DISTllIATIOM-C0!IDBHS*rE^ TO STORAGE
-------
Water Use and Waste Source Characteristics
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 wastewater 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.
Waste Sources
A. Gypsum Solids
Gypsum solids are generated as a by-product. The amount produced
is in the range of 3.6 to 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 wastewater,
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
wastewater. When kiln residue is disposed of as a solid waste,
scrubber wastewater is the major source of waste. Table 12-4
gives the data for the direct and indirect process contact
wastewater 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
wastewater flow to inplant treatment facilities for plants whose
wastewater includes the gypsum slurry and for those practicing
disposal of kiln residue as a solid waste.
B.. 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
267
-------
TABLE 12-3. WATER USAGE IN THE HYDROFLUORIC ACID SU8CATEGORY
Water Usage at Plants
(ra3/kkg of HF)
(1) (1) (1)
Source #987 #251 #753 #426 #120 #722 #167 #705
Non-contact 154 NA 63.5 110 NA 13.6 116 30.0
Cooling
Gypsum Slurry NA 64.0 NA * NA 22.5 41.6 30.0
Transport
Maintenance, NA 2.40 2.11 NA 0.1 12.2 5.00 16.9
Equipment and
Area Washdown
Air Pollution 7.90 14.4 4.23 NA 0.586 14.5 40.0 11.3
Control
NA = Not Available
* = Not Applicable
(1) Discontinued HF production.
268
-------
TABLE 12-4. WASTE WATER FLOW AND REUSE DATA FOR THE
HYDROFLUORIC ACID SUBCATEGORY
Plant
#120
#426
(3)
1987
#837
#967
(31
#251
(3)
#705
#167
#753
#928
#664
#722
Kiln Residue
(D
Handling
D
D
D
S
S
S
S
S
S
S
S
S
Reuse for
Kiln Residue
(21
Slurry
(Percent)
_„,
(4)
(41
0
0
0
30.0-35.0
47.0
65.0
83.0
94.0
92.0-100
Influent to
Treatment
Facility
(m3/kkg)
HF
9.10
0
13.6
120
125
84.7
58.2
166
31.4
55.5
96.6
120
Treated
Effluent
Di scharged
(m3Akg)
HF
9.10
Not available
13.6
120
125
84.7
39.3
38. 2
11.1
9.40
5.80
7.20
Averages: (S only)
(1) D = Dry disposal
42.8 percent 95.4 m
S = Slurried to treatment
54.6 m
(21 Percent of waste water flow reused for residue slurry after
treatment.
(3) Dicontinued HF production.
(4) Not Applicable.
269
-------
15,000
12,500 t
1,000
Dry Kiln Waste
Slurrying Kiln Waste
75 100 150
HF Production, kkg/day
200
Figure 12- 2. Production versos waste flow data for HF plants.
270
-------
D
reaction between hydrofluoric acid and suIfuric 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 Wastewater
Scrubber water is another wastewater source, and in plants which
practice dry disposal of gypsum, scrubber water constitutes the
predominant and major source of wastewater. 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.
SiO
4HF = SiF
2H»0
(3)
In the scrubber, the silicon tetrafluoride is converted to
hexafiuosilicic .acid according to the following equations:
SiF4 + 2HF = H2SiF6
3SiF4 + 2H20 - 2H2SiF6
Distillation Wastes
Si0
(4A)
(4B)
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 wastewater 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.
271
-------
F. Spar Drying Wastewater
Wet fluorspar is first processed through a dryer to remove
essentially all of the water. The steam and dust generated is
passed through a scrubber from which the wastewater is sent to
the gypsum neutralization tank. As shown in Table 12-6, the
contribution of the spar drying wastewater is insignificant
compared to the total amount of process wastewater and hence can
be neglected. The data is taken from 308-Questionnaire
responses.
G. 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 12-7 gives the amount of suspended solids
generated from the process and the quantity of total suspended
solids generated at the wastewater 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 12-8 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.
Description of Plants Visited and Sampled
Screening
Plant #705 was visited and process wastewater 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
wastewater treatment facility and the gypsum produced from the reactor
is slurried with water and also sent to the treatment facility. The
wastewaters from the HF production facility are combined with aluminum
fluoride plant wastewaters. The combined raw wastewater 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 12-9 gives the flow data and the total suspended solids (TSS)
and fluoride emissions.
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 12-10 gives the TSS and fluoride load
summary of the sampled streams.
Two more HF plants (Plant #251 and #167) were sampled in the
verification phase. The drip acid at both facilities is also sent to
the waste treatment plant and the hydrofluoric acid wastewaters are
272
-------
TAME 12-5. WASTE HOW FTCM HYDROFLUORIC ACID MAHUFACttJRING PLANTS
to
-4
Source of
Waste Water
Gypsum Slurry
Drip Acid
Scrubber
Haste Water
Other
Plow in m /Wog of Hydrofluoric Acid
Plants
1251 (1) #987 (1) |753 |426 1120 1722
64.0 Dry NA Dry Dry (Total
disposal disposal disposal Recycle)
0.0490 0 000 0
14.4 8.30 2.30 NA 0.624 (Total
Recycle)
0.530 0.530 8.40 NA 5.55 NA
1167 1705 (1) 1837
122 (Total 6.50
Recycle)
NA 0.0180 0
40.0 11.3 1.12
5.20 22.5 NA
(1) Discontinued HP production
NA = Not Available
* Other does not include wasteflows from storm water runoff.
-------
TABLE 12-6. FLOW AND POLLUTION CONCENTRATION DATA OF SPAR IKYING
AND TOTAL PROCESS WASTE WKEER FOR PLANTS #251 AND
#837 PRODUCING HYDROFLUORIC ACID
Stream
Plant Description
# 251 Spar drying
waste water
., Flow
(m /kkg of HF)
1.20 (1.8%) (1)
Total
Susperided Solids
(kg/kkg of HF)
70.8 (1.8%) (1)
# 837
Total process waste
water
Spar drying waste
water
Total process waste
water
66.9
1.11 (1%) (1)
114
4000
16.7 (0.5%)
3140
(1)
Nuiber in parentheses is tiie percentage contribution of a pollutant
parameter of the spar drying waste water to that of the total
process waste water.
TABLE 12-7. SOLID WASTE (DERATED AT THE HYDROFLUORIC ACID PLANTS SAMPLED
Gypsum Solids Going To Total Solids Produced
Plant Treatment Facility (kgAg of HF)
(kg/kg of HF)
#705(1) 4,73 4.78
#251(1) 3.81 NA
#167 3.94 NA
(1) Discontinued HF production.
NA = Not Available
274
-------
TABLE 12-8. G¥PSIK SOLIDS PRODUCTION IN THE HYDROFLUORIC ACID SUBCATEQQRY
Plant
*837
#705 (1)
#167
#722
#120
#426
#987 Cl)
#251t:L)
#753
#967
#928
Kiln Residue Produced
(kg/kg of HF)
3.86
4.73
3.94
NA
NA
4.00
4.13
3.81
NA
NA
NA
Kiln Residue
Disposal/Treatment Method
S
S
S
S
D
D
D
S
S
S
S
S = Slurrxed with water and sent to waste water treatment facility.
D = Dry disposal.
NA = Not Available.
(1) = Discontinued HF production.
275
-------
ro
-j
FlilQBSiftg
'W
t f
aOUBBER
KIUI
s.
FMSH
IATCR
MIPKID
'
UMB
e.
14
flUWJCE DBMNS
e
IJUJO
BtFNMB KW|>}ad.
^
12
Uf PBOOJCT
Figure 12-3. General process flew diagram at plant #705 showing the sampling points.
Hydrofluoric acid manufacture.
-------
TABLE 12-9. FI0W AND POLLUTANT CONCENTRATION DATA OF THE SAMPLED WASTE
STREAMS OF PLANT #705- PRODUCING HYDROFLUORIC
Stream
No.
1
2
3
4
Sarrpled
Stream
Description*
Kiln Slurry
Scrubber Waste
Water
Surface Drains
Cooling Tower
Slowdown
Treated Effluent
Screening
Flow
(m3Akg of HF)
26.6
10.0
20.0
23.3<3>
Data(2)
Fluoride
(kg/kkg of HF)
15
9.6
6.9
1.6
Tbtal
Suspended
Solids
(kg/kkg of HF)
4700
0.070
3.9
1.9
(1) This plant has discontinued the production of HF since the time of
sairpling.
(2) One 72-hour conposite sanple of each waste water stream.
(3) Ihe discharged effluent consists of the treated waste waters from
hydrofluoric acid and aluminum fluoride plants.
277
-------
combined with aluminum fluoride plant waste for treatment. In
addition to drip acid, Plant #251 wastewater 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 wastewater originating in the HF plant can be intercepted.
Summary of the Toxic Pollutant Data
Eleven toxic pollutants were found in the raw waste samples from HF
Plant #705. They were also verified at three other typical HF plants
practicing BPT~- .treatment. No organic toxic pollutants were found at
detectable levels.\ The results were:
( Maxlmum\Raw Waste Concentrations Observed
\ 1 Ug/1)
Screening
Plant #705
Verification
Plants #705,
#251, #167
Copper
Lead
Selenium
Zinc
Antimony-
Arsenic
Cadmium
Chromium
Mercury
Nickel
Thallium
770
5200
25
8100
70
10
2.0
73
2.0
150
5.5
600
200
230
11000
2800
160
60
1200
43
2000
63
Section 5 of this report describes the methodology of the screening
and verification sampling program. In the Hydrofluoric Acid industry,
a total of 12 days of sampling were conducted at Plants I 705, #251,
278
-------
TABLE 12-10. FLOW AND POLLUTANT CONCENTRATION DATA OF THE SAMPLED WASTE STREAMS
FOR PLANTS #705, §251, AND #167 PRODUCING HYDROFLUORIC ACID
Plant
#705 (2)
#251 C2)
#167
Stream
No.
1
2
4
5
5
6
2
3
1
2
3
4
Verification Data
Sanpled 3 Flow
Stream (m /kkg of HF)
Description
Kiln Slurry
Scrubber Waste
Water
Surface Drains
Cooling Tower
Slowdown
Treated Effluent
AHF Plant
Hosedown
SO- Scrubber
Waste
Gypsum Pond Inlet
Gypsum Pond
Outlet
Kiln Slurry
Ejector & Absorber
Unit Wastes from
Kilns #1,#2, and
#4
Ejector & Absorber
Unit Wastes from
Kilns #5 and #6
Effluent from
First Lagoon
26.6
10.0
20. Q
23.3^
1.20
14.4
84.7
84.7
122
25.0
14.6
162
(1)
Fluoride
(kg/kkg of HF)
3.8
1.5
3.4
0.54
1.9
0.31
58
27
4.9
14
20
11
Total
Suspended
Solids
(fcg/kkg of HF)
4700
0.019
4.0
0.040
0.26
0.10
3800
0.80
170
0.36
0.41
22
(1) Three 24-hour composite samples of each waste water stream.
(2) These plants have now discontinued their HF production.
(3) Consists of the combined flow from hydrofluoric acid and aluminum
fluoride plants.
279
-------
00
o
VENT
DUST
COLLECTION
WET
SPAR"
•ARI L»-
SPAR
DRYING
HOSE DOWN
WATER
r
HF KILN
AIR
HANDLING
LOSSES
-t
DRIP
AC1D
WATER
SLURRY
TREATMENT
e
LEGEND
SAMPLING POINTS.
AlFj PRODUCT
LIQUEFACTION
ON *J
AHF
PURIFICATION
1
AHF
DILUTION WATER
DOWN WATER
AHF PLANT
#2
GYPSUM
POND
^
NEUTRALIZATION
SYSTEM
ALKALINE STREAMS AND
ACID FROM OTHER PLANTS
Figure 12-4. General process flow diagram at Plant #251 showing the sampling points
Hydrofluoric acid manufacture.
WATER
EFFLUENT
-------
and #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 ft705
generated an additional 635 analytical data points. The daily raw
waste loads were calculated from the waste stream flow rates measured
or estimated at the time of sampling and the measured pollutant
concentration.
That is,
Daily loading {as kg of pollutant 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 unit of m3/day.
(m3, a cubic meter, is equal to 264.2 U.S. gallons)
Similarly, the unit loadings were calculated from the reported
hydrofluoric acid production rate, the waste stream flow rate, and the
measured pollutant concentration.
Unit loading ( as kg of pollutant - (C)(Q)
per kkg of hydrofluoric acid) 1000 (P)
where C and Q are the same as described above, and P is the
hydrofluoric acid production rate expressed in units of kkg/day. (Jckg
is 1000 kg, a metric ton, which is equal to 2205 Ibs.)
The minimum, average, and maximum values are based on data from those
plants where the particular pollutant was found at concentrations
greater than the analytical detection limits and significant in that
it could conceivably be treated by an available treatment technology
regardless of economic considerations.
In Table 12-11, 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 12-12 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:
281
-------
Pollutant Waste Load (kg/year)
Copper 6600
Lead 10000
Selenium 260
Zinc 110000
Antimony 8900
Arsenic 1400
Cadmium 79
Chromium 4700
Mercury 130
Nickel 10000
Thallium 840
Pollution Abatement Options
Toxic Pollutants of Concern
Toxic pollutants in raw wastewaters and slurries typical of the HF
industry include the heavy metals often found as impurities in
fluorspar. These metals are zinc, .lead, nickel, antimony, chromium,
arsenic, copper, and selenium. Raw wastewaters 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 wastewater treatment facilities. Although the fluorosulfonate
anion is found in HF wastes containing drip acid, organic compounds
are not anticipated in wastewaters from this industry. No toxic
organic pollutants were found at significant levels.
Process Modifications and Technology Transfer Options
A. Gypsum produced in the kiln can be disposed of as a solid waste
instead of being slurried with water and sent to the wastewater
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 transportation
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.
B. 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
wastewater 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
282
-------
TABLE 12-11. TOXIC POLLUTANT RAW WASTS DATA.
SUBCATEGQRY: HYDROFLUORIC ACID
(1)
Average Daily Pollutant Concentrations and Loadings at Plants Sampled
(rog/l)
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Tnallium
Zinc
(i
1705 (S)
0.018
0.0010
0.051
0. 0029
0.0014
0. 000080
0.062
0.0035
0.41
0.023
2.47
0.14
0.00090
0.000050
0.062
0.0035
0.0070
0.00040
*
4.0
0.23
kgAkg of Anhydrous HP)
$705 (V) *23L(V>
0.010
0.00057
* J-
*
0,0060
0.00034
0.26
0.015
0.26
0.015
0.044
0.0025
0.0053
0.00030
0.48
0.027
*
*
*
0.21
0.012
0.12
0.010
0,11
0.0091
*
0.47
0.040
0-12
0.010
0.059
0.0050
0.018
0.0015
1.18
0.10
0.017
0.0014
0.039
0.0033
0.28
0.024
ti67(\n
0.74
0.12
0.028
0.0046
0.0030
0.00047
0.074
0.012
0.32
0.051
0.062
0.010
0.0010
0.00016
0.15
0.025
0,0074
0.0012
0.019
0.0030
8.2
1.3
Overall
Average
0.22
0.033
0.062
0.0055
0.0035
0.00030
0.22
0.018
0.23
0.025
0.66
0.039
0.0060
0.00050
0.47
0.039
0.011
0.0010
0.029
0.0032
3,2
0,41
S - Screening data from one 72-hour composite sample of
individual or combined raw waste streams.
V - Verification data from three 24-hour composite samples, averaged,
from each raw waste sampling .point.
* - Concentration below significant level.
(1) The methodology of the sampling program is described in Section
5.1.2, and Section 12.3.3 presents the scope of sampling in the
Hydrofluoric Acid industry.
283
-------
TABLE 12-12. SUMMARY OF RStf WASTE LOADINGS FOUND IN
SCREENING AND VERIFICATION SAMPLING
SUBCATEGORY:
Pollutant
Toxic
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Thallium
Zinc
Conventional
HYDROFLUORIC
ACID
Daily
loadings
(kg/day)
Minimum Average Maximun
0.023
0.012
0.0031
0.15
0.60
0-10
0.0021
0.14
0.016
0.16
0.49
2.0
0.50
0.'014
1.7
1.4
1.8
0.057
4.1
0.093
0.31
21
6
1
0
5
2
.4
.2
.025
.4
.80
5.4
0
14
0.
0.
72
.21
20
45
Minimum
0.
0.
0.
0.
00057
00030
000077
0035
0.0096
0.
0.
0.
0.
0.
0.
0025
QOQ05Q
00035
00040
0030
012
Unit
Loadings No. Of
(kgAkg) Plants
Average Maximum Averaged*
0,
0,
0.
0.
0.
0.
0.
0.
0.
0.
0.
034
0055
00030
018
025
039
00050
039
0010
0032
41
0.
0.
12
0090
0.00047
0.
0.
0.
0.
0.
0.
0.
1.
040
051
14
0015
10
0014
0033
3
4
3
3
4
4
4
4
4
3
2
4
& Nonconvetxtional
TSS 190000
Fluoride
13
310000
2900
520000
7900
3800
8.
8
4200
34
4800
58
3
4
* Only those plants **iere the pollutant was observed at treatable
levels were included.
284
-------
best alternative for operation of the recycle system at a 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 wastewater, 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.
C. Two out of a total of 11 plants manufacturing hydrofluoric acid
send the drip acid to the wastewater treatment facility. The
rest of the plants recycle it to the reactor. When discharged to
the waste treatment system, the fluorosulfonic acid does not
hydrolyze and leaves with the treated effluent as a complex
fluoride in soluble form. The total fluoride concentration of
the effluent will be higher for the plants discharging 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 = HZS04 + 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.
Best Management Practices
A. Runoff can be collected from raw material and product storage,
process, and impoundment areas. It should be treated with other
process waste at the wastewater 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.
B. 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.
285
-------
Prevailing Control and Treatment Practices
Plant #705 combines the hydrofluoric acid wastes, including the gypsum
slurry, with aluminum fluoride waste. The combined wastewater, 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
above.
Plant #837 combines the gypsum slurry and plant area hosedown
wastewater 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 wastewater
from scrubbers of both hydrofluoric acid and aluminum fluoride plants
is sent to an adjoining facility for use.
Plant #251 also combines the hydrofluoric acid and aluminum fluoride
wastewater. The suspended solids in the combined wastewater are
removed in the gypsum ponds. The overflow from the gypsum ponds is
neutralized and the pH adjusted with the wastewater 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 wastewaters.
Two plants, #120 and #987, dispose of the kiln residue as a solid
waste after lime addition. The wastewater 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 wastewater (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 wastewaters 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.
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.
286
-------
Selection of Appropriate Technology and Equipment
Technologies for Different Treatment Levels
A. Level 1 (BPT and BAT)
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
(0-35 percent) 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.
For BAT, 65 percent effluent reuse is incorporated into the Level
1 treatment system.
B. Level 2
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.
C. 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.
D. 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 the 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
dicharge or process recycling. (Figure 12-8.)
E. Level 5
Level 5 treatment is dry handling and off-site chemical landfill
for the kiln waste and two-stage alkaline precipitation with
287
-------
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.)
Incremental cost and performance estimates for the alternative
levels of treatment and control were evaluated in detail as part
of the rule-making process which lead to the proposed regulations
{45 FR 49450, July 24, 1980). This material is presented in the
proposed Development Document (60).
Equipment for Different Treatment Levels
A. 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 Level 5,
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 wastewater treatment,
and the cost is not included in the cost estimates.
B. 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 HZS 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,
with no unusual safety hazards or special handling problems. In
Level 5, only lime, soda ash and hydrochloric acid are used,
introducing no special problems of safety or handling.
C. 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
288
-------
OD
UME
RAW
RECYCLE FOR SLURRY TRANSPORT
LAGOON
<*>
WASTE WATER
-»\ JQ /'—"»
EQUALIZATION
""""""PjCy'*™™™^'^
U.
MIXING
LAGOON
^^*****PP
pH ADJUSTMENT
EFFLUENT
Include! flow monitoring, pH monitoring and tamplar
12-5. Level 1 waste water treatment for hydrofluoric acid subcategory.
-------
vO
o
f— nukx
LJ-*n
HAW
WA6TK WATCH
kl/
KOUAIIZATION
r~-©
i
i
I
db
1
1
|i
M1UHO .
RCCVCLB FOB
0U4VBV TRANSVOBl
.A t'AlUOH
~1
nr*itj
-H_h
BUMP LQJ
fii
1WTMKMT
X
Uttw HUM!luring.
Figure 12-6. Level 2 waste water treatment for hydrofluoric acid subcategory
-------
to
\O
H
h^Qn
z
J^
fU Kousaan
0
MI10MB
•Lh-a^1
ln«luile> How moolktrlnf, pH nWnllutln) and uuu
Figure 12-7. Ijevel 3 waste water treatment for hydrofluoric acid subcategory.
-------
r
i
1 IttMAM
M
,
w3w
IKTU
«•)
_ /
^MMLuOnM 1
1 — ^
f
UXIW
t
Li
*
._ A tiflOM L^
n n
•
'
:
—^t UUUGM /.»
»j [M
«
1
.rim
X
T "
K
1
HJ
'
g)
Wtft
VC
^V
lsiitom ftft* munMadtt), ptl
Figiore 12-8. Level 4 vraste water treatment for hydrofluoric acid subcategory
-------
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 model are not subjected to treatment, except for
nominal application of lime before hauling in dry form to an
approved chemical landfill.
Treatment Cost Estimates
General Discussion
To prepare
developed.
treatment cost estimates, a model plant concept was
The 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.
In the BAT treatment, 65 percent of the BPT treated effluent is reused
to transport kiln wash to the treatment facility as a slurry.
Wastewater 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 mVkkg of HF to 166 mVkkg of HF. For the model
plants, a constant unit flow of 95.4 mVkkg of HF is assumed.
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 wastewater 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.
Wastewater Pollutant Load
The amount of kiln residue varies from 3.8 to 4.1 kg/kg of HF
produced. The wastewater going to treatment model plants is assumed
to contain 3.8 kg of solid kiln residue per kg of HF. Fluoride
emissions in wastewater have been shown to vary as indicated below:
293
-------
H» ADJUSTMENT
UME
UXK
WLN
WASTE
|*-LID
**r •
i i X 1-AGOON
J- LAGOON
TO LAt.TlbtLI.
Include* On* mttMtlarlni, pit nMnilt«rln| fend
Figure 12-9. Level 5 waste water treatment for hydrofluoric acid subcategory
-------
Source of Data
Fluoride, (kq/kkq)
Reference 3
Reference 3
Screening and Verification
Phase Sampling
(Tables 12-9 and 12-10)
20
37
3.8 to 58
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 the BAT system which may add to or
modify the existing BPT system to meet more stringent pollutant
removal requirements.
Treatment costs at Levels 2, 3, 4, and 5 are given in the proposed
Development Document (60).
The estimated costs for three models having different production
levels are given in Tables 12-13, 12-14, and 12-15. For these models,
both the hydraulic and the pollution loads per unit of production are
held constant over the entire range of production.
Table 12-16 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 level of treatment, the cost estimate is based on 65
percent of the wastewater flow being recirculated.
Model Plant Control Costs for Existing Sources
For the model plant control costs for existing sources at the BPT and
BAT levels 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 signficant impact on the total
annual costs.
Basis for Regulations
Evaluation of BPT Treatment Practices
Control and treatment practices for eleven plants producing HF are
presented in Table 12-17. Also indicated are other product-related
wastewater sources and pollutant loads discharged.
295
-------
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:
A. 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 wastewater generated at most plants, greatly
reducing the raw waste loads to be treated. The only sources of
wastewater remaining are from air pollution control and washdown.
B. Effluent Reuse
Reuse of treated wastewater 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, the fact that five plants do practice reuse
demonstrates that the practice is both technologically and
economically feasible.
C. 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.
D. 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.
In addition to the above factors, the design and operation of the
treatment facilities affect the 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.
E. Pollutant Removal with BPT Treatment
Treatment level 1 is BPT in the Hydrofluoric Acid industry.
Table 12-18 presents a summary of long term effluent monitoring
296
-------
TABLE 12-13. MODEL PLANT TREATMENT COSTS
Subcategory Hydrofluoric acid
Production 19,100 metric tons per year
A. INVESTMENT COST
Site development ......
Equipment
Monitoring equipment ,.
Subtotal
Contractor's 0 & P b....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
BPT
410,000
429,500
20,000
859,500
128,925
988,425
197,685
1,185,110
lift, 511
1,304,721
1,020,000
2,324,721
TOTAL OPERATION AND
MAINTENANCE COST
55,000
14,000
534,300
130,472
59,742
350,000
15,000
1,170,014
BAT
0
45,000
0
45,000
6,750
51,750
10,350
52,100
5,210
58,310
0
58,310
14,000
1,500
0
5,831
2,049
0
7,500
31,880
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
212,273
1,382,292
11,114
42,994
Represents the incremental cost above that for BPT treatment
Overhead and Profit
297
-------
TABLE 12-14. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Hydrofluoric acid
38,200 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal
Contractor's 0 5t P*3....
Subtotal
Snginhering
Subtotal
Contingencies
Subtotal
Land
TOTAL INVESTMENT COST .
B. OPERATION AND
MAINTENANCE COST
Labor and supervision .
Energy
Chemicals
Maintenance
Taxes and insurance ...
Re s i d ua 1 wa st e d i spo sa1
Monitoring , analysis
and reporting
BPT
735,000
551,500
20,000
1,306,500
195,975
1,502,475
300,495
1,802,970
180,297
1,991,2*7
1,940,000
3,923,2*7
5*,000
19,500
1,0*9,000
193,327
TOTAL OPERATION AND
MAINTENANCE COST
700,000
15,000
2,175,525
($)
BATa
0
70,000
0
70,000
10,500
80,500
1*,100
9*,*00
9,**0
10*,2*0
0
10*,2*0
14,000
3,100
0
10,*?*
0
7,500
38,414
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
322,*78
2,498,202
17,2S9
55,702
a Represents the incremental cost above that for BPT treatment
k Overhead and Profit
298
-------
TABLE 12-15. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Hydrofluoric acid
57,300 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment
Subtotal .. *... .....*..
Contractor's 0 & Pb
Subtotal
Engineering .
Subtotal
Contingencies
Subtotal
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
($)
BPT
1,050,000
928,000
20,000
1,998,000
299,700
2,297,700
459,540
2,757,240
275,724
3,032,964
2,890,000
5,922,964
56,000
28,000
1,604,000
303,296
177,689
1,050,000
15,000
3,233,985
BAT0
0
100,000
0
100,000
15,000
115,000
23,000
138,000
13,800
151,800
0
151,800
14,000
4,600
0
15,180
4,554
0
7,500
45,834
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
493,463
3,727,449
24,698
70,532
a Represents the incremental cost above that for BPT treatment
y.
Overhead and Profit
299
-------
TABLE 12-16.MODEL PLANT UNIT TREATMENT COSTS
Subcategory Hydrofluoric acid
Annual Treatment Costs ($/kkg)
COST ITEM
PRODUCTION
(kkg/yr)
LEVEL OF TREATMENT
BPT BAT*
Annual Operation
and Maintenance
Annual
Amortization
Total Annual
Cost
19,100
38,200
57,300
19,100
38,200
57,300
19,100
38,200
57,300
61.26
56.95
56.44
11.11
8.45
8.61
72.37
65.40
65.05
1.67
1.01
0.80
0.58
0.45
0.43
2.25
1.46
1.23
*Represents the incremental cost above BPT
300
-------
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
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
12-11 with the corresponding treated effluent data presented in
Table 12-19. 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 BPT 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 BPT wastewater control and treatment technology allows for
the discharge of process wastewater 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
sampling, seven of the 11 plants operating were meeting the
proposed fluoride limitations and eight were meeting the proposed
TSS limitations according to the data available (60). Although
there was practically no long term monitoring data available to
support the additional proposed limitations on toxic metals, the
screening and verification data indicated that all three plants
sampled were meeting the proposed limitations on antimony,
copper, and lead, while two of the plants were meeting the
proposed zinc limitation and one plant was meeting the proposed
chromium and nickel limitations. With the limited amount of
toxic metal data, it was not possible to estimate compliance or
noncompliance on a statistical basis. The Agency has conducted
additional treatability studies (61) and has utilized the results
of this work in formulating the final regulations as described in
the following sections.
301
-------
TABLE 12-17. SOHMARY OF WASTE HATER CONTROL AMD TREATMENT TECHNOLOGY EMPLOYED AT HYDROFLUORIC ACID PLANTS
(1)
Plant
Product-Related
Waste Hater Sources
Control and
Technology
Treatment
Employed
Amount of
Treated
Haste Hater
Reused
Cooler Bottoms
(Condensablea)
Recycled?
Effluent Volume
in IB* /»etrie ton
(gal/ftbort ton) of
Actual Production
Long Tern
Average Pollutant
Hasteload Discharged
(kg/aetric ton)
{lb/1000 Ib)
6«4
167
120
' 967
928
! 837
* 753
Fluoride
I 251
(21
I 705
(2)
* 722
Hydrofluoric acid
production
Hydrofluoric acid,
floorocarboo,
Chiorine/sodi urn
hydroxide, and
hydrochloric acid
production
Hydrofluoric acid
production
Hydrofluoric acid,
fluorocarbon, and
•olfuric acid
production
Residue slurry, neutra- 94%
Illation with •odium
carbonate, settling,
recycle
Residue slurry, lime 47%
treatment,settling,
recycle
Planned dry residue 0
handling, li«*
treatment, clarification
Residue slurry, settling Present: 0
(Recycle and pa Planned! 70%
polishing facilities bo 75%
under construction.)
5.78 (1,360)
103 (24,200)
Tes
Hydrofluoric acid Residue slurry, settling,
and aluminum recycle {Plocculation,
fluoride production lime treatment, and
clarification facilities
under construction.)
83%
All hydrofluoric
acid generated as
used captively for
aluminum fluoride
production
Hydrofluoric acid
production
Residue slurry, lime
tr«et»*nt, settling
Residue slurry, lime
treatment, settling,
recycle, pH polishing
Yes
65%
HF, AIT], chlorine/ Residue slurry, settling,
sodium hydroxide, neutralization
aluminum oxide, and
fluorocarbon
production
Hydrofluoric acid Residue slurry, line
and aluminum treatment, settling,
fluoride production recycle, pB polishing
1 Kiln: Tea
3 Kilns.- No
30% to 35%
134 (32,200)
11.0 (2,650)
22.2 x 10s
(553 x 10*)
25.9 (6,204)
TSS
426 Hydrofluoric acid
fluosilicic acids
production
'Dry residue hauling 0
and dumping; neutra-
lization with caustic
of noncontact cooling
water and floor
drainage
Yes 465 (111,397)
includes noncon-
tact .cooling
water
1.2 SD
0.10 0.27
18 0.45 (Net)
(3)
125 (30.000) Present: 24 16
Expected
with 1.8 2.1
additional
facilities
9.44 (2,260) Present: 1 1.7
Expected
with 0.65 0.75
additional
facilities
1.8 3.1
0.64 0.38
46 530
3.2 0.64
Hydrofluoric and,
in recent peat,
fluoboric,
acid production
Residue slurry, lime
treatment, settling,
recycle, pB polishing
92% to 100%
» 967l ' Hydrofluoric Acid Dry residue hauling
Tes
0-10.3 (0-2,460) 0-0.81 0 to 0.54
8.8
tl) Adapted fros Calspen (Reference 3).
(2) Hydrofluoric Acid production has been discontinued at these plants since the time of sampling.
(3) Effluent loading less the influent loading.
HD - Not determined.
302
-------
Basis for BPT Effluent Limitations
A. Technology Basis
For BPT, the Agency is promulgating 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 necessary to meet the fluoride
limitations.
B. Flow Basis
The reuse of treated wastewater 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 wastewater ranging from 30 to 100 percent of the
plant flow as shown in Tables 12-4 and 12-17.
The practice of reusing wastewater in this manner has two
opposing effects on the plant effluent:
1. A decrease in the net discharge unit flow rate (m3/kkg), and
2. 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 12-10. The apparent reason
for the increase in fluoride concentration with reuse is a
calcium deficiency which may result from the buildup of sulfate
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 wastewater
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 12-13, 12-14, and 12-15).
The model plant BPT treatment system is based on an inflow rate
of 95.4 mVkkg derived from the average of nine plants which
303
-------
TABLE 12-18. StWARY OF LONG TERM "WNITORING DMA FRCM FOUR
(1)
tnEROPLUORIC ACID PLANTS
Treated Waste toad
or (Ib/lOOOlb)
Plant
No. Etararaeter
Daily Data
Long Term
Average St.Dev.
(X) (S) (SM
30-Day Average Data
(2) Long Term (2)
W Average St.Dev. W
(X) (S)
1664
*753
1722
*705
Fluoride
TSS
Fluoride
TSS
Fluoride
TSS
(3)
Fluoride
TSS
0.
0.
0.
0.
0.
0.
—
—
10
29
72
38
81
54
0.090
—
0.27
—
0.52
0.37
—
—
0.77
—
0.36
—
0.59
0.62
—
—
4.5
—
2.2
—
3.3
3.5
—
—
0.
0.
0.
—
—
0.
0.
10
27
64
49
84
0.
—
0.
—
_
— =•
0.
0.
040
15
22
37
1.7
—
1.4
— —
_
—
1.7
1.7
(I)
Based on Reference 3 data.
(21
In the case of daily measurements, the variability factor, VF,
Cor a lognomal 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 l.O +(_Sj\21.When the value of Z is 2.33, the
In fl.
L
variability factor for the 99 percentile is obtained.
For 30-day average measurements, a noraal distribution is
obtained and the variability factor is found by the expression,
VF » 1.0 + Z ..• 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
of long term data.
analysis
(31
M.though Plant V705 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 bailable.
304
-------
TABLE 12-19. TOXIC POLLUTANT TREATED EFFLUENT DATA
SU3CATEGOTX: HYDROFLUORIC PCID
(1)
Average Daily Pollutant Concentrations and Loadings at Plants Sampled
(mg/D
(kg/kkg of Anhydrous HP)
(2)
Overall Average
*705{S) 1705 (V) f251(V) H67(V) Average % Removal
Antimony
Arsenic
Cadmium
Chromium
Copper
lead
Mercury
Nickel
Selenium
Thallium
Zinc
<0.010
<0. 00021
<0.0030
<0, 000063
0.00030
0.0000060
0.014
0.00029
0.10
0.0021
0.0060
0.00012
<0. 00040
<0. 0000080
0.050
0.0010
0.033
0.00069
0.0070
0.00015
0.071
0.0015
<0.0020
<0. 000042
<0.010
<0. 00021
<0. 0017
<0. 000035
<0.046
<0. 00096
<0.020
<0. 00042
<0.022
<0. 00046
<0. 00050
<0. 000010
<0.010
<0. 00021
<0.0050
<0. 00010
<0.0012
<0. 000025
0.053
o.oon
<0.17
<0. 017
<0.020
<0.0020
<0.0020
<0. 00020
0.22
0.022
0.070
0.0069
<0.031
<0.0031
<0.0010
<0.00010
0.52
0.052
<0.071
<0.0070
<0.0070
<0. 00069
0.16
0.015
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
<0.0026
<0. 0045
<0. 00039
0.55
0.13
74
81
9
9
62
77
97
67
64
Effluent
>Influent
85
«3
(S) Screening data from one 72-hour composite sample of treated
effluent.
(V) Verification data from three 24-hour composite samples.
(1) The affluent data presented here corresponds to tfte raw waste
data showi in Table 12-41. The methodology of the sampling
program is described in Section 5.1.2, and the scope of
sampling in the Hydrofluoric Scid industry is described in
Section 12.3.3.
(2) When averaging values indicated as "less than" (O, the
absolute value was used and the resulting average was indicated
as a "less than" value.
305
-------
handle the kiln residues in a slurry system as shown in Table
12-4. The treated effluent flow rate is 54.6 mVkkg 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.
C. Selection of Pollutants to be Regulated
The selection of pollutants for which specific numerical effluent
limitations are promulgated is 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. 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 technology options. The other metals,
chromium, thallium, and mercury exhibited maximum concentrations
that were considerably lower.
Total subcateqory 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 12-11 and the daily and
unit loadings are summarized in Table 12-12. 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.
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 incidential 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. Based on the argument
provided in Section 8, the control of zinc and nickel, having the
highest loadings and concentrations in the wastewater, would
306
-------
effectively control the other toxic metals in the treatment
system. Thus, zinc and nickel were selected as the control
parameters of toxic pollutants for BPT regulations.
D. Basis of Pollutant Limitations
1. 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 the proposed Development Document
(60) and the JRB Study (52).
b. TSS and Fluoride: The data presented in Tables 12-12 and
12-18 were used for the development of TSS and fluoride
limitations. However, because of the wide range of product
mixes, significant differences in residue handling,
wastewater treatment, reuse practices, and dilution with
other product waste streams, it was necessary to select only
those plants where the effect of BPT technology could be
clearly observed. The plants excluded are:
#426 and 1120 because kiln residues are
solid.
handled as a dry
#167, 1967, and #251 because the combined treatment of HF
wastes along with the wastewaters from other major products
generated high fluoride loadings in the large volume
discharges with fluoride at its minimum treatability
concentration,
#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
12-20 which summarizes the development of the regulations
for total suspended solids and fluoride. Since the BPT
level of treatment does not require the reuse of treated
wastewater for slurrying kiln residues, the performance of
Plant #837 was used as the long term average unit loading
basis for the TSS and fluoride limitations. The variability
factors used for fluoride are based on the long term data
from Plants 1664 and #753 and those used for TSS are derived
from Plant #722 for daily measurements and Plant #705 for
30-day average measurements as indicated in Table 12-20.
The maximum 30-day average TSS limitation was obtained by
multiplying the variability factor for 30-day averages from
Table 12-20 by the long term average waste load; i.e., 1.7 x
3.1 kg/kkg = 5,3 kg/kkg. Similarly, the 24-hour maximum TSS
limitation was obtained by multiplying the variability
307
-------
100
Percent Reuse
IEGENP
• tang-term data
O Expected with treatment system upgrading
D Screening and verification sanpling results
Figure 12-10, Fluoride loads and concentrations discharged at
selected hydrofluoric acid plants.
308
-------
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 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 = t 6.1 kg/kkg for the 24-hour maximum
limitation. These' computations are shown on Table 12-20 and
the BPT limitations are presented in Table 12-21.
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 (mVkkg). (Note: kg/m3 = 1000 mg/1.)
Thus, the concentration basis for the maximum 30-day average
TSS limitation is:
(5.3 kq/kkq) x
(54.6 mVkkg)
(1QQQ mg/1) = 97 mg/1
(kg/m3)
and the concentrations basis for the 24-hour maximum
limitation is obtained by a similar calculation or simply by
applying the variability factor ratio, VFR, from Table 12-21
to the maximum 30-day average concentration; that is:
(VFR) (max. 30-day average concentration or loading)
= 24-hour maximum concentration or loading
In the same manner, the concentration basis for the maximum
30-day average fluoride limitation is:
(2.9 kq/kkq) x (1000 mg/1) = 53 mg/1
(54.6 mVkkg) (kg/m3)
and the 24-hour 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
12-21 which was obtained by calculating the concentration
directly from the 24-hour maximum limitation; i.e.,
(6.1 kq/kkq) x
(54.6 mVkkg)
In either case,
taken. )
(1000 mg/1) =112 mg/1
(kg/m3)
only two significant figures should be
309
-------
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 wastewater.
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.
2. Toxic pollutants
The effluent limitations set 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), and 3) a limited amount of long-term monitoring
data from Plant #251.
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 12-19. The
average values shown for each pollutant are interpreted as
being approximately equal to a long-term 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 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 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 12-19. This high degree of incidental
310
-------
TABLE 12-20. DEVELOPMENT OF TSS AND FLUORIDE LIMITATIONS
Plant
1837
1753
#928
#722
#664
Reuse
(percent)
0
65
83
92
94
Long
Waste
Fluoride
(kgAkg of HF)
1.8
0.72
1.0
0.81
0.10
Term Average
Load Discharged
TSS
(kgAkg of HF)
3.1
0.38
1.7
0.54
0.29
Variability Factor for
Daily Measurements
Variability Factor for
30-Day Averages
3.5(5)
1.6
Variability Factor Ratio (VFR) 3.4/1.6 = 2.1
(1)
(2)
3.5/1.7 - 2.1
(2)
Effluent Limitations for BPT
(from Plant #837)
a. Daily Max 3.4 X 1.8 kgAkg ~ 6.1
b. Max 30-Day Avg 1.6 X 1.8 kg/kkg = 2.9v
Effluent Limitations for BAT
(fron Plant #928)
a. Daily Max 3.4 X 1.0 kgAkg = 3.4
b. Max 30H3ay Avg 1.6 X 1.0 kgAkg = 1.6
(3)
(3)
(4)
3.5 X 3.1 kgAkg = 11
(3)
1.7 X 3.1 kg/kkg = 5.3
MA
NA
(4)
NA - Not Applicable
(1) Variability factor average of Plants #664,#722 and *753 from Table 12-18,
(2) Ratio of the daily (24-hr) variability factor to the 30-day
average variability factor. This value appears on the Proposed
Limitations tables.
(3) The long term average loading in kgAkg multiplied by the
variability factor for daily measurements as shown.
(4) The long term average loading in kgAkg multiplied by the
variability factor for 30-day measurements as_shown.
(5) Variability factor from Plant #722, Table 12-18.
(6) Variability factor from Plant #705, Table 12-18.
311
-------
Table 12-21. Effluent Limitations
Hydrofluoric Acid
Best Practicable Control Technology Currently Available
Waste Water Flow: 54.6 m3/kkg of HF (43% Reuse)*
Subcategory
Pollutant Performance
Conventional and
Nonconventional
Pollutants:
Total Suspended
Solids
Fluoride
Toxic Pollutants:
Antimony
Arsenic
Chromium
Copper
Lead
Nickel
Selenium
Zinc
(mg/1)
(?}
57 Uj
33(2>
0.80 (3>
0.50<3)
0.32(3>
0.32(3)
0.13<3>
0.17<4>
0.20(3)
0.55<4>
Concentration Basis Effluent Limit
m (mg/1) (kg/kkg of HF)
VFET '
3.5/1.7
3.4A-6
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
Max
30 -day
Avg
97
53
0.96
0.60
0.38
0.38
0.16
0.20
0.24
0.66
24-hr
Max
200
110
3.1
2.0
1.3
1.3
0.51
0,66
0.78
2.2
Max
30-day 24-hr
Avg Max
5.3 11
2.9 6.1
_(5> .(5)
_<5) _<5)
..(5) _(5)
__(5) __(5)
-(5) __(5)
0.011 0.036
-(5) _<5)
0.036 0.12
(1) - VFR: ratio of the 24-hour variability factor to the 30-day variability
factor.
(2) - Long term average based on loading data and variability factors selected
from Table 12-18.
(3) - The lower limit of the literature treatability estimate (Table 8-11) and
industrial waste water treatment system performance (Table 8-12) are used
as the basis for the long term average vAien the observed average of the
sampling data is below tJiis level.
(4) - Average effluent concentration from screening and verification sampling
data.
(5) - No effluent limitation required.
* - From Table 12-4.
312
-------
removal supports the contention that by applying effluent
1 imitations just to the dominant metal pollutant(s),
effective control of the other metals will also be assured.
In Table 12-21, the concentration bases for the BPT
limitations are derived from the averaged effluent sampling
data unless the observed pollutant concentration is actually
below the literature treatability level. In some 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 BPT limitations on the control metals,
zinc and nickel is given below. For the other toxic metals,
the concentration bases are derived below and are intended
to serve as guidance in cases where these pollutants are
found to be of serious concern.
a.
Zinc
The raw waste concentrations of zinc ranged as high as
11.3 mg/1 {Section.12, Table of Maximum Concentrations
Observed) and averaged about 3.2 mg/1 (Table 12-11} 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 12-19. This level of
performance 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 0.036 kg/kkg using the model plant flow
of 54.6 mVkkg (Table 12-4). This limitation was
achieved by all but one of the plants sampled. Using
the model plant flow of 54.6 mVkkg from Table 12-15,
the limitation was calculated as follows:
(0.66 mg/1) (54.6 mVkkg) ( kq/m3 ) = 0.036 kg/kkg
(1000 mg/1)
where 0.66 is the maximum 30-day average concentration
calculated by multiplying the average concentration
(0.55) by the 30-day variability factor, 1.2:
(0.55 mg/1}(1.2) =0.66 mg/1
313
-------
Since the long term monitoring data on zinc are not
available from the HF industry, the variability factor
ratio (VFR) of 3.3 was derived from the nickel sulfate
industry data. The same VFR was used for the other
toxic metals in the HF effluent. This is supported by
the fact that nickel is a dominant toxic metal in the
HF process wastewater, and is assumed to have equal or
better performance characteristics compared with the
wastewater from the nickel sulfate industry. Thus,
VFR = VF of daily measurements = 3.9
VF of 30-day averages 1,2
= 3.3
and the 24-hour maximum limitation for zinc is:
(3.3) (O.D36 kg/klcg) =0.12 kg/kkg.
The effluent limitations on zinc and the other metals
of concern are given in Table 12-21.
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. Using a monthly
variability factor of 1.2, a value of 0.20 mg/1 is used
as the concentration basis for the maximum 30-«3ay
average effluent limitation of 0.011 kg/kkg. A VFR of
3.3 was used following the same rationale described for
zinc. Thus, the maximum 30-day average limitation is:
(0.20 mg/1)(54.6 m3/kkg) ( kg/m* ) = 0.011 kg/kkg
(1000 mg/1)
and the 24-hour maximum limitation is:
(3.3) (0.011 kg/kkg) = 0.036 kg/kkg.
c. Lead
The observed average raw waste concentration of lead
(0.66 mg/1) was not far above the 0.13 mg/1 estimated
long term average treatability according to industry's
performance data in Table 8-12. Using a monthly
variability factor of 1.2, the concentration basis for
the maximum 30-day average is:
(1.2) (0.13 mg/1) =0.16 mg/1
Based on a 24-hour variability factor of 3.9, the
concentration basis for the 24-hour maximum is:
314
-------
(3.9) (0.13 mg/1) = 0.51 mg/1
No effluent: limitation is set for lead.
Antimony
According to literature treatability data (Table 8-11),
the lower limit of treatability for antimony is 0.80
mg/1 as a long term average. Based on VFR of 3.9/1.2,
the concentration basis for the maximum 30-day average
is:
(1.2) (0.80 mg/1) =0.96 mg/1
and the concentration basis for the 24-hour maximum is:
(3.9) (0.80 mg/1) =3.1 mg/1
No effluent limitation is set for antimong.
Copper: With 0.32 mg/1 as the average treatability for
copper (Table 8-12), the concentration basis for the
maximum 30-day average is:
(1.2) (0.32 mg/1) =0.38 mg/1
and the concentration basis for the 24-hour maximum is:
(3.9) (0.32. mg/1) = 1.25 mg/1
No effluent, limitation is set for copper.
Chromium: Similar to copper, an average treatability of
0.32 mg/1 is used for chromium. The concentration for
the maximum 30-day average is:
(1.2) (0.32 mg/1) = 0.38 mg/1
and the concentration basis for the 24-hour maximum is:
(3.9) (0.32 mg/1) = 1.3 mg/1
No effluent limitation is set for chromium.
Other Metals
The concentration bases for arsenic and selenium are
also presented in Table 12-21.
315
-------
Basis for BCT Effluent Limitation
EPA has determined that the BAT technology for this subcategory is
capable of removing significant amounts of conventional pollutants.
However, EPA has not yet proposed or promulgated a revised BCT
methodology in response to the American Paper Institute v. EPA
decision mentioned earlier. Thus, it is not now possible to apply the
BCT cost test to this technology option. Accordingly, EPA is
deferring a decision on the appropriate BCT limitations until EPA
proposes the revised BCT methodology. However, the Agency has
calculated the TSS loading based on the estimated performance of BAT
treatment, and the cost of the additional TSS removal. As described
in Section 3, this cost was calculated to be $0.32 per pound of TSS
removed.
By adjusting the loading to account for the decrease in effluent flow
rate from BPT (54.6 mVkkg) to BAT (33.4 m3/kkg), the TSS maximum
30-day average effluent loading becomes:
(5.3 kg/kkg) x (33.4 mVkkq) = 3.2 kg/kkg
(54.6 mVkkg)
The corresponding 24-hour maximum effluent loading is then obtained by
applying the VFR value of 2.1 (Table 12-21). That is:
(2.1)(3.2 kg/kkg) =6.7 kg/kkg
Basis for BAT Effluent Limitations
A. 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 BAT regulations.
For BAT, the Agency is promulgating limitations based on
treatment consisting of Level 1 technology. The Agency
considered the use of treatment Level 2 (addition of a dual-media
filter) but did not adopt it because the installation of the
filter is not cost effective in the removal of TSS, fluorides and
toxic metals. In addition, the use of treatment Level 3
(addition of sulfide precipitation) was considered but rejected
due to lack of performance data. EPA also considered Level 4, a
variation of Level 2, that would substitute soda ash in the lime
precipitation step and allow 90 percent recycle of effluent.
This option was rejected due to being prohibitively expensive.
Pollutants limited by the final BAT regulation are fluoride,
nickel, and zinc.
316
-------
B. Technology Basis
For BAT, the Agency is setting the effluent limitations on
fluoride and the toxic metals based on the BPT treatment system,
coupled with the requirement of at least 65 percent effluent
reuse for kiln residue slurrying.
In Section 7.0 "Hydrofluoric Acid Subcategory" of EPA1s
Treatabllity Studies for the Inorganic Chemicals Manufacturing
Point Source Subcateqory (EPA 440/1-80/103) (61), the conclusion
states:
"A major conclusion that can be drawn from this study is that the
addition of dual-media filtration after alkaline precipitation
and settling is not particularly effective for the reduction of
final TSS and total fluoride concentrations. Further, dual-media
filtration does not appear to be justified on the basis of
additional toxic metal removal judging by the results presented
in Table 7-3."
A summary of the filter removal efficiencies for the pollutants
selected for the treatability study is given in Table 12-22.
The minimum reuse rate of 65 percent was selected because it is
typical of the five plants (Plants #167, #753, #928, # 664, and #
722} which presently practice reuse as is shown in Table 12-4.
C. Flow Basis
With the model plant inflow rate of 95.4 mVkkg and the reuse of
65 percent of the treated effluent, the quantity discharged is
33.4 mVkkg; i.e., (1.00 -0.65X95.4 mVkkg) = 33.4 mVkkg.
D. Selection of Pollutants to be Regulated
For the BAT regulations, the Agency has selected fluoride and the
same two toxic metals identified in the BPT regulations. The
rationale for their selection is discussed above.
E. Basis of Pollutant Limitations
1. Nonconventional pollutants
The only nonconventional pollutant found in the wastewater
in this subcategory is fluoride. The limitation for BAT is
based on the performance of the four plants shown in Table
12-20 that presently reuse at least 65 percent of their
treated effluent. The long term average effluent loading
taken from Table 12-22 is 1.0 kg/kkq for Plant #928, since
the percent reuse for this plant is intermediate in range
for the three plants considered for BAT (Plants #753, # 928,
and #722). Plant #664 is excluded from consideration for
BAT because the high degree of recycle practiced requires
317
-------
the extensive use of soda ash. 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 mVkkg
results in a calculated maximum 30-day average concentration
of 50 mg/1 total fluoride. Thus, the maximum 30-day average
limitation is:
(1.6)(1.0 kg/kkg) =1.6 kg/kkg
and its concentration basis is:
(1.6 kq/kkq)
(33.4 mVkkg)
x 1000 mq/1 = 50 mg/1
(kg/m3)
This represents a six percent reduction in fluoride
concentration in going from BPT (43 percent reuse) to BAT
(65 percent reuse). The use of a fixed loading limitation
allows the permissible concentration to increase as a
function of percent reuse. The 24-hour maximum limitation
on fluoride is obtained by utilizing the long term average
and variability factor for daily measurements,
(3.4)(1.0 kg/kkg) = 3.4 kg/kkg
and the concentration basis is:
(3.4 kg/kkq) x (1000 mg/2) = IGQ-mg/l
(33.4 mVkkg) (kg/m3)
The variability factors used for the BAT limitations on
fluoride are the same as for BPT shown in Table 12-20. The
BAT limitations for the Hydrofluoric Acid Subcategory are
presented in Table 12-23.
2". Toxic pollutants
For BAT regulations, the EPA is proposing more stringent
controls on the discharge of the two toxic metals of concern
on the basis of a reduced volume of discharge. 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 undoubtly involved
during the treatment process and probably account for a
substantial portion of the removal of certain toxic metals.
The bases for the BAT limitation on the two control metals,
zinc and nickel, are given below. The concentration bases
of other metals are derived below and are intended to serve
as guidance in cases where these pollutants are found to be
of serious concern.
a.
Zinc
318
-------
TABLE 12-22. SU4MAHY OF CCNCENTRATIONS AND FILTER REMOVAL EFFICIENCIES
OF POLLUTANTS SELECTED FOR STUDY IN THE HYDROFLUORIC ACID
SUBCATEG3KY
Total Total Total Free Total
Chromium Zinc Nickel Fluoride Fluoride TSS
Average raw 0.30 1.09 1.23 219.8 — 3315
vraste concentration
(mg/1)
Average supernatant 0.073 0.10 0.46 45.9 116 333
concentration
(ng/1)
Average filter 0.067 0.05 0.41 42.2 91 134
effluent concentra-
tion (mg/1)
Average filter 6 36 17 6 20 53
removal efficiency
319
-------
The same value of 0.55 mg/1 used for BPT is used as the
treatability limit for zinc. As in the BPT
regulations, a variability factor ratio (VFR) of
3.9/1.2 was used for zinc and the other metals in
setting the BAT limitations and concentration bases.
Thus, the concentration basis for the maximum 30-day
average effluent limitation is:
(1.2) (0.55 mg/1) =0.66 mg/1
and the maximum 30-day average effluent limitation is:
(0.66 mg/1) (33.4 mVkg) (kg/m* ) = 0.022 kg/kkg
(1000 mg/1)
The concentration basis for the 24-hour maximum
limitation is:
(3.9) (0.55 mg/1) = 2.2 mg/1
and the 24-hour maximum limitation is:
(2.15 mg/1) (33.4 mVkkg) ( kq/m3 ) = 0.022 kg/kkg
(1000 mg/1)
This represents an overall reduction of 40 percent from
the BPT loading limitation. The BAT limitations on
zinc and nickel are included in Table 12-23.
b. Nickel
The same value of 0.17 mg/1 used for BPT is used as the
treatability limit for nickel. The concentration basis
for the maximum 30-day average effluent limitation is:
(1.2) (0.17 mg/1) =0.20 mg/1
and the maximum 30-day average effluent limitation is:
(0.20 mg/1) (33.4 mVkkg) ( kq/m^ ) = 0.0060 kg/kkg
(1000 mg/1)
The concentration basis for the
limitation is:
24-hour maximum
(3.9) (0.17 mg/1) = 0.66 mg/1
and the 24-hour maximum effluent limitation is:
(0.66 mg/1) (33.4 mVkkg) ( kg/m3 ) =0.20 kg/kkg
(1000 mg/1)
320
-------
TABLE 12-23. EFFLUENT LIMITATIONS
HYDROFLUORIC ACID
Best Available Technology
Wastewater Flow: 33.4 m3/kkg of HF {65% Reuse)
Concentration Basis
(mg/1)
Pollutant
Treatability
(mg/1)
VFR
(1)
Max.
30-day
Avg»
24-hr.
Max.
Effluent Limit
(kg/kkg of HF)
Max
30-day 24-hr,
Ayg. Max.
Noriconvetitional
Pollutants:
Fluoride/ F
Toxic Pollutants
31
(2)
3.4/1.6
50
100
1.6
3.4
Antimony
Arsenic
Chromium
Copper
Lead
Nickel
Selenium
zinc
0.80
0.50
0.32
0.32
0.13
0.17
0.20
0.55
3.9/1,2
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
0,96
0.60
0.38
0.38
0.16
0.20
0,24
0.66
3.1
2.0
1.3
1.3
0.51
0.66
0.78
2.2
— (3)
— (3)
— (3)
— (3)
0.0060
— (3)
0.022
— (3)
— (3)
— (3)
— (3)
0.020
— (3)
0.072
(1) VFR: ratio of the 24-hour variability factor to the 30-day variability
factor.
(2) 30-day average calculated for the model plant based on data in Table 12-20.
(3) No effluent limitations.
* The effluent flow rate is 35 percent of the average influent shown in
Table 12-4 (i.e., 0.35 X 95.4 m3A^g = 33.4
321
-------
This represents an overall 44 percent decrease from the
corresponding BPT level.
Lead
Using a treatability limit of 0.13 mg/1 for lead, the
maximum 30-day average concentration basis is:
(1.2) (0.13 mg/1) =0.16 mg/1
The 24-hour maximum concentration basis is:
(3.9) (0.13 mg/1) =0.51 mg/1
No effluent limitation is set for lead.
Antimony
Using a treatability limit of 0.80 mg/1 for antimony,
the maximum 30-day average concentration basis is:
(1.2) (0.8 mg/1) = 0.96 mg/1
The 24-hour maximum concentration basis is:
(3.9) (0.8 mg/1) = 3.1 mg/1
No effluent limitation is set for antimony.
Copper
Using a treatability limit of 0.32 mg/1 for copper, the
maximum 30-day average concentration basis is:
(7.2) (0.32 mg/1) = 0.38 mg/1
The 24-hour maximum concentration basis is:
(3.9) (0.32 mg/1) = 1.3 mg/1
No effluent limitation is set for cooper
Chromium.
Using a treatability limit of 0.32 mg/1 for chromium,
the maximum 30-day average concentration basis is:
(1.2) (0.32 mg/1) =0.38 mg/1
The 24-hour maximum concentration basis is:
322
-------
(3.9) (0.32 mg/1) =1.3 mg/1
No effluent limitation is set for chromium.
g. Other Metals
The concentration bases for arsenic and selenium are
also given in Table 12-23.
Basis for the New Source Performance Standards
BAT was selected as the basis for NSPS limitations. The pollutants to
be controlled for NSPS are pH, total suspended solids, fluorides,
nickel and zinc. The NSPS limitations are given in Table 12-24.
In addition to the toxic pollutants controlled by the BAT, a total
suspended solids limitation was developed for NSPS. Based on the data
in Table 12-20, a long term average of 1.7 kg/kkg (Plant t928) was
used for the TSS. The variability factor ratio (VFR) used is 2.1,
3.5 = 2.1
1 .7
Thus, the 24-hour maximum effluent limitation for TSS is:
(3.5) (1.7 kg/kkg} = 6.0 kg/kkg HF
The maximum 30-day average limitation for TSS is:
(1.7) (1.7 kg/kkg) = 3.0 kg/kkg HF
Basis for the Pretreatment Standards
A. Existing Sources
The Agency is not promulgating PSES because there are no indirect
dischargers in the subcategory. Instead, we are excluding the
subcategory from categorical PSES under the provisions of
paragraph 8(b) of the Settlement Agreement.
B. New Sources
For Pretreatment Standards for New Sources (PSNS), the Agency is
setting limitations equal to NSPS because NSPS provides better
removal of nickel and zinc than is achieved by a well-operated
POTW with secondary treatment installed and therefore these
pollutants would pass through a POTW in the absence of
pretreatment. The pollutants to be regulated are fluoride,
nickel, and zinc as indicated in Table 12-24.
323
-------
-------
Table 12-24. Effluent Limitations
Hydrofluoric Acid
New Source Performance Standards *
Waste Water Flow: 33.4 n^/kkg of HF (65% Reuse)*
Pollutant
Treatability
(mg/1)
Concentration Basis
,U)
VFR'
Max
30-day 24-hr
Avg Max
Effluent Limit
(kg/kkg of HF)
Max
30-day 24-hr
Avg Max
Conventional and
Nbnconventianal
Pollutants;
Total Suspended
Solids, TSS
Fluoride, F
Toxic Pollutants:
53
31
<3>
3.5/1.7
3.4A-6
90
50
180
100
3.0
1.6
6.0
3.4
Antimony
Arsenic
Chromium
Copper
Lead
Nickel (2)
Selenium
2incC2)
0.80
0.50
0.32
0.32
0.13
0.17
.0.20
0.55
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
3.9/1.2
0.96
0.60
0.38
0.38
0.16
0.20
0.24
0.66
3.1
2.0
1.3
1.3
0.51
0.66
0.78
2.2
_V*' \"*l
(A) Ml
* ' i— ^ '
_(4) _(4)
i 4) f 4)
I A\ f4)
0.0060 0.020
_J4) _<4)
0.022 0.072
(1)
(2)
(3)
(4)
*
- VFR: ratio of the 24-hour variability factor to the 30-day variability
factor.
- Also applicable for PSES limitations.
- 30-day average calculated for the model plant based on the data in
Table 12-20.
- No effluent limitation.
is
shown in
t -
Table 12-4 (i.e., 0.35 X 95.4 m3/kkg = 33.4 m^/kkg) .
This table is also applicable to the PSNS effluent limitations except
for the TSS limitation.
325
-------
-------
SECTION 13
HYDROGEN PEROXIDE INDUSTRY
Summary of 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 exclusion 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 Settlement Agreement.
Assessment of the Water Pollution Potential
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 persulfate 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 the manufacture of hydrogen peroxide
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 wastewater sources
include contact cooling (barometric-condenser) water, purification
washing of the organic working solutions, regeneration waste from the
deionizers, and leaks and spills.
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 Section 308 letters
is given in Table 13-1.
327
-------
13-1
SUBCATEGQRY PROFILE DATA SUMMARY
SGBCATEQORy
HYDROGEN PEROXTCE
Total subcategory capacity rate
Ttotal subcategory production rate
Number of plants in this subcategory
308 Data on file for
With total capacity of
With total production of
Representing capacity
Representing production
Plant production range:
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volume per unit product:
Mi Tvjrmm
Maximum
85,700 kkg/year
4
4
102,200 kkg/year
57,000 kkg/year
66 percent
5,560 kkg/year
28,730 kkg/year
NA
NA
NA
15 years
27 years
NA
NA
NA
NA
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Commerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chemical Industry, " June, 1978 and "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards f
-------
Three plants product hydrogen peroxide by the organic process.
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 its
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:
Pollutant
Zinc
Pentachlorophenol
Bis(2-ethylhexyl)phthalate
Chloroform
Naphthalene
Status of Regulations
Maximum Concentration
Observed Uq/1)
256
4850
20
11
11
Since no toxic pollutants were found in significant concentrations,
the subcategory is excluded under Paragraph 8 of the Settlement
Agreement•
329
-------
-------
SECTION 14
TITANIUM DIOXIDE INDUSTRY
INDUSTRY PROFILE
General Description
Titanium dioxide (TiO2) is manufactured by a chloride process, a
sulfate process, and a chloride-ilmenite process. This subcategory is
subdivided into three segments, one for each process because of the
difference in raw materials used, wastewater flows, and raw waste
characteristics. TiOz 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
laquers. 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-S, while the status of regulations prior to promulgation of
this now regulation is given in Table 14-2.
Subcategorization
The Titanium JHoxide Subcategory is divided into three different
processes requiring separate consideration. Two processing techniques
including sulfate arad tihe chloride process using two major ores
including rutlie and ilmenite dictate a division into the sulfate
process, chloride process/rutile ore, and chloride process/ilmenite
ore
-------
TABLE 14-1
SDBGKEEX3QRY' PROFILE DKEA
SOBCKEEGQEQ?
TTIM1IUM DIOXIDE (OEOKTDE PROCESS)
Total subcategory capacity rate
Ibtal subcategory production rate
ftnfcer of plants in this subcategory
308 Data on file for
With total capacity of
With total production of
capacity
Representing production
Plant production ranges
Average production
Median production
Average capacity rfr'*
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volume per unit product:
Minimum
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 Besearch Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S, Department of Gcnroerce, 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 Chemcals
Industry," March, 1980.
.332
-------
TABLE 14-2. STATUS OF REGULATIONS - EFFLUENT LIMITATION GUIDELINES
SUBCATEGORY
SUBPART
TITANIUM DIOXIDE
V (40 CFR 415.220, 3/13/74)
STANDARDS
BPCTCA*
Product
Process
Chloride
Process
Sulfate
Process
Max.1
kg/kkg
Parameters (mg/1)
TSS 4,6
Iron 0.72
TSS 21.0
(100.0)**
Iron 1 . 7
(8.1)
Avg. 2
kg/kkg
(mg/1)
2,3
0.36
10.5
(50.0)
0.84
(4.0)
BATEA*
Max.
kg/kkg
(mg/1)
2,6
0.36
10,6
0.84
Avg.
kg/kkg
(mg/1)
1.3
0. 18
5.3
0.42
NSPS*
Max.
kg/kkg
(mg/1)
2.6
0.36
10.6
0.84
Avg.
kg/kkg
(mg/1)
1.3
0.18
5.3
0.42
Sections 415.220, 415.222, 415.223 and 415.225 were remanded and reserved
(41 FR 51601, November 23, 1976).
Max. = Maximum of any one day.
2 Avg. = Average of daily values for thirty consecutive days,
** - Flow basis 210,000 1/kkg.
333
-------
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 liquified titanium tetrachloride. Iron
chloride and small amounts of vanadium, zirconium, and other trace
metal chlorides are removed by centrif ugation 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 sulf ide, 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 TiCl4 stream. Organic complexing agents aid in
separation of the TiCl* 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, lf^JLetrachlor ide contains impurities such as
, ,— ; ff-%fn., .-^--^x^^i^^^^^^^-''fr"<*!^^ , , • L -i ,
aluminum chlorrae, 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:
TiCl4 + 02 = TiOz + 2C1? (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 discharges, or by use of fluidized beds. After the oxidation
titanium dioxide, ,f p,r.m,s_, a_ so, 114 an,d_ is,7separa±.e
-------
Water Use and Waste Source Charateristics
Water Use
Water is used in noncontact cooling, for scrubbing the tail gases from
the purification and oxidation reactors 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 TiO2 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 TiOz
produced.
Waste Sources
A. Wastes from Cooling Chlorinator Gas
The waste consists of solid particles of unreacted ore, coke,
iron, and small amounts of vanadium, zirconium, chromium, and
other heavy metal chlorides. They are either dissolved in water
and sent to the wastewater treatment facility or disposed of in
landfills as a solid waste.
B. 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. In a
second stage, they are scrubbed with caustic soda to remove
chlorine as hypochlorite.
C. 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.
D. Oxidation Tail Gas Scrubber Wastes
The gases from the oxidation unit are cooled by refrigeration to
liquefy and recover chlorine. The uncondensed off-gasses 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 wastewater treatment facility, or
discharged without treatment. The scrubber waste stream also
contains titanium dioxide particulates.
335
-------
aoj ?XGKMr
Figure 14-1. General process diagram for production of titanium
(chloride process) from high grade ores.
336
-------
TABIE 14-3. WATER USAGE IN TITANIUM DIOXCDE-CHIORlbE PROCESS/HIQI GRADE
ORES SUBCATEGORY
Water Use
Water Usage at Plants
(m3/kkg of
Water Use
Plant #102
Noncontact cooling
Direct process contact
Indirect process contact
Maintenance, equipment
cleaning and work area
washdcftm
Air pollution control
Noncontact ancillary uses
Sanitary & portable water
Total
182
10.5
NA
6.65
0.25
11.6
0.23
211
Plant #172
10.7
15.5
0.72
0.52
7.14
10.4
0.31
45.3
Plant #199
426
73.2
26.5
2.80
11.3
9.5
5.6
555
NA = Not available
337
-------
E. Finishing Operations Waste
The liquid wastes from the finishing operation contain titanium
dioxide as suspended solid and dissolved sodium chloride formed
by the neutralization of residual HC1 with caustic soda.
The range of wastewater 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.
Description Of Plants Visited And Sampled
Screening
Plant #559 was visited and the waste effluents were sampled in the
screening phase of the program. Plant £ 559 makes titanium dioxide
using both the sulfate and the chloride processes. The wastewaters
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 TiCl^)
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
wastewaters from both the sulfate and chloride processes are also
combined at this point. The combined wastewater 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.
Problems were encountered during the sampling of the pit solids and
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.
Verification Plant -JtLZi^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 wastewater
338
-------
TABLE 14-4. WASTE WATER FLOW FOR TITANIUM DIOXIDE-CHLORIDE PROCESS
SUBCATEGORy
SUBCATEGORy
TITANIUM DIOXIDE (Chloride Process)
Plant
Unit Waste Water Flow Going to Treatment Plant
(m3/kkg of Ti02>
#102
#172
#559
#199
29.3
34.7
91.0
110.0
(1)
(1)
(2)
(2)
(1) Off-site 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 flew for
cost estimating and regulation development.
339
-------
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 runoff and sent to the first of two retention basins
arranged in series. The overflow from the second retention basin 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.
Toxic Pollutant Concentrations
Chromium was identified in the raw wastewater for Plant #172 at a
concentration considered treatable based on information summarized in
Section 8. Four toxic metal pollutants were identified above
treatable levels at Plant #559, although it is likely that the levels
observed included some contributions from the sulfate process, since
the two raw waste effluents are intermixed before treatment. No toxic
organic pollutants were identified above treatable levels in the raw
wastes at either plant. The maximum concentration of toxic metal
pollutants observed during screening and verification are presented as
follows:
Maximum Raw Waste Concentrations Observed (ug/1)
Pollutant
Screening
Plant #559
Verification
Plant #172
Chromium
Lead
Nickel
Zinc
152,000
5,150
6,320
3,300
1800
14
85
660
The screening and verification sampling program and the methodology
used have been described in Section 5 of this report. A total of six
days of sampling was conducted at Plants 1559 and #172. 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
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:
34C
-------
JUWf WATEft
sum* WATEN
WILL
UAH*
^T
SANITARY
ANft TIO, fINISHINC
AREA IlkSfE WATER
OfHEN fHODUCT
(STKOHC ACIOl
WASTE UA!£«
iUILIHC
OIHCK roooucr
(UCAK ACIO) »*
WAIFI WAtEft
,
ILUKMIED — ^
fH SOLIDS
rCHD
9"
t
»i,
1C AC ION
1
ft
fl
^
t
^_ PISTILLATION I01TOH
^^ UA$TE IMTCI
UCtMD
•4 SAHPiINC rOINTS
(
tTH
WA
EN 1
HE
mis
* CHLONIOC PNOCCSS SCHUIICR
WASU WATER
floouci
WATER
Fioure 14-2. General flow diagram at Plant #559 showing the sanpling points.
(Titanium dioxide — chloride process manufacture.)
-------
TABt£ 14-5. FUCW AND POLLUTANT CONCENTRATION DATA OF THE SAILED WASTE
STREAMS OF PIANT #559 PRODUCING TITANIUM DIOXIDE BY
CHLORIDE-RUTILE PROCESS
(1)
Pollutant
TSS
Iron
Chromium
Lead
Nickel
Zinc
STREAM #2
Pit Solids arri
Distillation Bottoms
Unit
Flow Cone.,
(m3/kkg) (mg/1)
10,9
6903
1348
112
3.53
3.46
2.12
Unit
Load
(kg/kkg)
75.2
14.7
1.2
0.04
0.04
0.02
STREAM #5
Scrubber and
Contact Cooling Water
Unit Unit
Flow Cone., load
(m3/kkg) (mg/1) (kg/kkg)
80.1
314 25.2
143 11.5
0.11 0.01
0.009 0.001
0.016 0.001
0.13 0.01
CALCULATED ESTIMATE
Total Raw Waste
Unit
Flew Cone.,
(m3/kkg) (mg/1)
91
1103
288
13.3
0.5
0.5
0.3
Unit
Load
100.4
26.2
1.21
0.041
0.041
0.03
STREAM #6
Treated Effluent
Unit Unit
Flow Cone., Load
(m3/kkg) (mg/1) (kg/kkg)
91
23 2.1
4.4 0.4
0.03 0.003
0.002 0.0002
0.005 0.0004
0.06 0.005
w
**
10
(1) See Figure 14-2 for location of sampling points.
-------
PROCESS
WASTE WATER
HOLDING POND
FOR
RETREATMENT
RETENTION
BASIN
RETENTION
BASIN
'#3
DISCHARGE
NaOH
MIXING
BASIN
NEUTRALIZE
RAIN RUNOFF
RAIN RUNOFF
.pH ADJUSTMENT
LEGEND
SAMPLING POINTS
* THE TOTAL RETENTION TIME
OF WATER IN THE TWO-PONDS
IS 5 DAYS.
Figure
General flow diagram at Plant #172 showing the sampling points.
Titanium dioxide (chloride process) manufacture.
343
-------
TABLE 14-6. FLOW AND POLLUTANT CONCENTRATION DATA OF THE SAMPLED WASTE
STREAMS FOR PLANT #172 PRODUCING TITANIUM DIOXIDE
(CHLORIDE PROCESS)
Pollutant
TSS
Iron
Chromium
Lead
Nickel
Zinc
STREAM #1
(Raw Waste
Avg. Cone.
(mg/1)
171
2.9
0.72
0.005
0.08
0.3
Influent) (1)
Unit Load(2)
(kg/kkg)
5.93
0.10
0.03
0,0002
0.003
0.01
STREAM
#3
(Treated Effluent) (1)
Avg. Cone.
(mg/1)
6.7
0.33
0.02
0.002
0.01
0.09
Unit Load
(kg/kkg)
0.23
0.01
0.0007
0.00007
0.0003
0.003
(1) Unit flow is 34.7 m /kkg.
(2) Unit load equals the product of the pollutant concentration in iflg/1 and
the unit flow in m3/]ckg divided by a conversion factor of 1000.
344
-------
Waste Load (kg/year)
241,000
8,200
8,500
7,800
Pollutant
Chromium
Lead
Nickel
Zinc
Pollution Abatement Options
Toxic Pollutants of Concern
The dominant toxic pollutant observed in the raw waste effluents for
the Titanium Dioxide (Chloride Process) Subcategory is chromium.
Chromium was found at 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 #172 raw waste. At Plant #559, the chloride
process waste effluents are mixed with the sulfate process waste
effluents before treatment. It is likely that the three major toxic
pollutants found were contributed by the sulfate process wastes since
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 is probable that most of the toxic metal
pollutants are present in this solid waste and hence do not appear in
the wastewaters.
Process Modification and Technology Transfer Options
A. 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.
B. 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 in view of analagous data from the chlorine
subcategory presented in Section 11.
C. 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 six existing plants are presently
treating storm water runoff.
345
-------
TABLE 14-7. RAW TOSTE POLUJTANT DATA SUM-JARY OF THE SAMPLED STREAMS
SUBCATEGORY:
Average
Pollutant
Iron
ChronLum
Lead
Nickel
Zinc
TSS
TTTANUM DICKIDE (CHLORIDE
PROCESS)
Daily Pollutant Concentration and loadings at Plants
kg/kkg of TiO.
(mg/1)
Plant
#559
26.2
(288)
1.21
(13.3)
0.041
(0.5)
0.041
(0.5)
0.03
(0,3)
100.4
(1103)
Plant v J }
#172 v '
0.10
(2.9)
0.03
(0.72)
0.0002
(0.005)
0.003
(0.08)
0.01
(0.3)
5.93
(171)
Sampled
Overall
Average
13.15
0.62
0.021
0.022
0.02
53.17
346
-------
D. Prevailing Control and Treatment Practices
At Plant #172, the solid wastes consisting of spent ore and coke
are hauled to an off-site landfill. Process wastewaters
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 wastewaters
from both chloride and sulfate processes 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 wastewater is then mixed
with the neutralized strong acid waste (from the sulfate process)
and scrubber waters (from both the chloride and sulfate
processes), and neutralized with lime in a reactor and sent to a
final settling pond. The overflow from the pond is the final
discharge.
At Plant #199, all the process wastewaters are combined,
including storm water and sanitary wastewater. The combined
wastewater 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 wastewater streams at Plant #102 are received in two
tanks, neutralized with lime, and then sent to a settling basin.
The settled solids are retained in the settling lagoons. The
plant has future plans for treating boiler blowdown, cooling
tower blowdown, leaks, and spills with the process wastewater.
At Plant #605, the unreacted ore and coke are disposed of as a
solid waste in the pit. The wastewater 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.
347
-------
E. 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.
Selection of Appropriate Technology and Equipment
Technologies for Different Treatment Levels
A. Level 1 (BPT/BAT)
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.
B. Level 2 (NSPS)
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 selected as the basis for
the proposed BAT limitations because it provides an economical
method for the removal of additional toxic metals. For the
reasons discussed below under Basis for New Source Performance
Standards, the Agency did not select this option for BAT in the
promulgated regulation, but did select it for NSPS.
C. 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.
348
-------
Equipment for Different Treatment Levels
A. Equipment Functions
BPT treatment is essentially lagooning with lime neutralization,
using no special equipment except a lime feeder and mixer.
In Level 2, second stage lime treatment is followed by gravity
clarification and multi-media filtration, with necessary pH
controls.
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.
B. Chemicals and Handling
Lime and hydrochloric acid are fed with conventional equipment at
all levels, and ferrous sulfide is prepared on-site by mixing
ferrous sulfate with sodium bisulfide. When normal dust control
and good ventilation are used, there should be no adverse effects
from handling these chemicals, although care should be taken that
hydrogen sulfide gas is not generated.
C. 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 operatirig 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.
Treatment Cost Estimates
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 BPT and NSPS were calculated
for each of the model plant production ranges using the unit flow and
unit pollutant loads. The unit flow of 100 mVkkg used for the model
plant in regulation development has been selected to be representative
of the subcategory and it is assumed that the imreacted ore and coke
are slurried and sent to the treatment system instead of being
disposed of in a landfill as a solid waste.
>49
-------
LIME
RAW
WASTE WATER
LAGOON
LAGOON
I
MIXING
LAGOON
LAGOON
EFFLUENT
w
ui
o
Includes flow monitoring, pH monitoring and sampler.
Figure 14-4. Level 1 (BPT/BAT) waste water treatment for titanium dioxide —
chloride process.
-------
BACKWASH
U)
UI
P.AW
WASTE WATB
Q pH ADJUSTMENT
* EFFLUENT
SUMP
TO LANDFILL
Include* flow monitoring, pit monitoring and •ampler,
Figure 14-5. Level 2 (NSPS) waste water treatment for titanium dioxide — chloride
process.
-------
FERROUS
SULFATE
SODIUM
BISULFIDE
HAW
WASTE WATER
SUMP
TO LANDFILL
Includes flow monitoring, pH monitoring and sampler.
Figure 14-6. Level 3 waste water treatment for titanium dioxide — chloride
process.
-------
A. Wastewatec Flow
The unit waste effluent flow varies from 29.3 to 110.0 mVkkg of
TiOz 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 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 mVkkg) and #199 (unit flow of 110 mVkkg) send 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 mVkkg of Ti02 has
been used for the model plants, which is an average of the unit
flows of Plants #559 and #199.
B. Pollutant Load
The primary pollutants occurring in the wastewater 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 too low 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 plant. 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 (kq/kkq of Tip,)
Chromium 0.63
Lead 0.021
Zinc 0.020
Iron 13.15
Nickel 0.022
C. Production Rates
Six plants produce titanium dioxide from rutile ore or upgraded
ilmenite ore using the chloride process at a total production
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
353
-------
mean of 28,400 kkg/year and a median of 25,600 kkg/year. For
wastewater treatment cost estimates, three production levels were
selected as model plants. These are 16,900 kkg/year, 25,500
kkg/year, and 45,200 kg/year. This range of production includes
all United States plants.
Model Plant Control and Treatment Costs
The estimated costs for the three models having different production
levels are given in Tables 14-8, 14-9, and 14-10. The costs shown at
each level of treatment correspond to the model plant BPT (Level 1}
system and higher level (NSPS) system which may add to or modify the
existing BPT system to meet more stringent toxic pollutant removal
requirements. The higher level (NSPS) also furnishes a better
effluent quality with respect to the conventional and nonconventinal
parameters. Tables 14-11, 14-12, and 14-13 present the estimated NSPS
treatment costs for the three models having different production
levels. 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 NSPS treatment
level, amortization, chemicals, and labor constitute a major portion
of the additional annual costs. Table 14-14 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.
Basis For Regulations
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 six 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 wastewater at any of the plants.
A. Pollutant Removal with BPT Treatment
Treatment Level 1 is equivalent to BPT in the Ti02 subcategory
(chloride process).
Plants #559 and #172 practice neutralization and settling of the
raw waste. At Plant #559, the chlride process raw wastewater is
mixed with the sulfate process wastewater for treatment. Also,
at Plant #559 the spent ore and coke (solid residues from the
chloride process) are slurried with water and sent to the
treatment facility, whereas at Plant #172 the solid residues are
hauled to a chemical landfill. Long-term treated effluent data
have been submitted by both plants 4559 and #172. The derivation
354
-------
TABLE 14-8. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Titanium dioxide - Chloride process
16,900 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment .,
Subtotal .........
Contractor's O & Pb....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
($)
BPT
175,600
205,000
20,000
400,600
60,090
460,690
92,138
552,828
55,283
608,111
192,000
800,111
TOTAL OPERATION AND
MAINTENANCE COST
56,000
5,900
140,000
60,811
24,003
108,000
15,000
409,714
BAT
0
0
0
0
0
0
0
0
0
0
0
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
98,940
508,654
Represents the incremental cost above that for BPT treatment
Overhead and Profit
355
-------
TABLE 14-9. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Titanium dioxide - Chloride process
25,500 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal £ • • •
Contractor's 0 & P 9...
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
Land
TOTAL INVESTMENT COST .
B. OPERATION AND
MAINTENANCE COST
Labor and supervision .
Energy
Chemicals
Maintenance
Taxes and insurance ...
Residual waste disposal
Monitoring, analysis,
and reporting
BPT
244,100
215,000
20,000
479,100
71,8*55
550,965
110,193
661,158
66,116
727,274
276,000
1,003,274
TOTAL OPERATION AND
MAINTENANCE COST
56,000
6,800
211,000
72,727
30,098
164,000
15,000
555,626
($)
BAT
0
0
0
0
0
0
0
0
0
0
0
a
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
118,327
673,953
a Represents the incremental cost above that for BPT treatment
Overhead and Profit
356
-------
TABLE 14-10. MODEL PLANT TREATMENT COSTS
Subcategory Titanium dioxide - Chloride process
Production 45,200 metric tons per year
A. INVESTMENT COST BPT BAT
Site development 395,000 0
Equipment 275,000 0
Monitoring equipment 20,000 0
Subtotal 690,000 0
Contractor's 0 & P b 103,500 0
Subtotal 793,500 0
Engineering 158,700 0
Subtotal 952,200 0
Contingencies 95,220 0
Subtotal 1,047,420 0
Land 504,000 0
TOTAL INVESTMENT COST 1,551,420 0
B. OPERATION AND
MAINTENANCE COST
Labor and supervision 56,000 0
Energy 86,000 0
Chemicals 374,000 0
Maintenance 104,742 0
Taxes and insurance 46,543 0
Residual waste disposal .... 294,999 0
Monitoring, analysis,
and reporting 15,000 0
TOTAL OPERATION AND
MAINTENANCE COST 977,284 0
C. AMORTIZATION OF
INVESTMENT COST 170,415 0
TOTAL ANNUAL COST 1,147,699 0
a Represents the incremental cost above that for BPT treatment
Overhead and Profit
357
-------
TABLE 14-11
MODEL PLANT TREATMENT COSTS
Subcategory
Production
Titanium dioxide - Chloride process
16,900 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal
Contractor's 0 & Pa....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
La nd
TOTAL INVESTMENT COST .
B. OPERATION AND
MAINTENANCE COST
Labor and supervision .
Energy
Chemicals
Maintenance
Taxes and insurance ...
Residual waste disposal
Monitoring, analysis,
and reporting
(S)
NSPS
175,600
724,000
20rOOO
919,600
137,940
1,057,540
211,508
1,269,048
126,905
1,395,953
192,000
1,587,953
TOTAL OPERATION AND
MAINTENANCE COST
140,000
11,500
174,000
139,595
47,639
117,000
15,000
644,734
AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
227,122
871,855
Overhead and Profit
358
-------
TABLE 14-12. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Titanium dioxide - Chloride process
25,500 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal
Contractor's 0 & P3 ....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
($)
NSPS
244,100
853,000
20,000
1,117,100
167,565
1,284,665
256,933
1,541,598
154,160
1,695,758
276,000
1,971,758
TOTAL OPERATION AND
MAINTENANCE COST
140,000
13,500
262,000
169,576
59,153
175,000
15,000
834,229
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
275,900
1,110,128
Overhead and Profit
359
-------
TABLE 14-13. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Titanium dioxide - Chloride
45,200 metric tons per year
process
A. INVESTMENT COST
Site development ......
Equipment
Monitoring eguipment ..
Subtotal
Contractor's 0 & Pa....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
(S)
NSPS
395,000
1,215,000
20,000
1,630,000
244,500
1,874,500
374,900
2,249,400
224,940
2,474,340
276,000
2,750,340
TOTAL OPERATION AND
MAINTENANCE COST
140,000
19,000
469,000
247,434
82,510
314,000
15,000
1,286,944
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
402,575
1,689,519
Overhead and Profit
360
-------
of the variability factors for daily maximum and 30-day
for both plants are given, in Tables 14-15 and 14-16.
averages
The concentration of the raw waste and treated effluent along
with the percent removal of the pollutants by the treatment
system for Plants #559 and #172 sampled in the screening and
verification program are given in Table 14-17.
In the chloride process, most iron is in the ferric state and is
readily removed by alkaline precipitation. Toxic metal removal
is also improved, since ferrite coprecipitation will occur which
tends to provide a better effluent quality than alkaline
precipitation alone, as observed at Plant #559.
Basis for BPT Effluent Limitations
A. Technology Basis
For BPT, the Agency is basing limitations on equalization,
neutralization, and settling or clarification. All plants in
this segment of the industry have BPT technology installed.
B. Flow 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 model plant treatment system,
the solid residues from the manufacturing process are assumed to
be slurried with water and sent to the treatment system. Plant
#559 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 mVkkg of Ti02 which is an average value of
the effluent flow of plants #559 and #199. The treated effluent
flow is assumed to be the same as the influent flow. The water
added or removed in the treatment system through chemical
addition, precipitation, and evaporation have been neglected, as
it varies from plant to plant and is dependent on the selection
of treatment chemicals as well as climatic conditions and is
insignificant in comparison to the total flow.
C. Selection of Pollutants to be Regulated
The selection of pollutants for regulation is based on an
evaluation of the waste data from the screening and verification
sampling program. The two major factors considered were the
individual plants raw waste concentrations and the total
subcategory pollutant loadings.
361
-------
TABLE 14-14.MODEL PLANT UNIT TREATMENT COSTS
Subcategory Titanium dioxide - Chloride process
Annual Treatment Costs ($/kkg)
LEVEL OF TREATMENT
COST ITEM PRODUCTION
(kkg/yr)
Annual Operation
and Maintenance
Annual
Amortization
Total Annual
Cost
16,900
25,500
45,200
16,900
25,500
45,200
16,900
25,500
45,200
BPT
24.24
21.79
21.62
5.85
4.64
3.77
30.10
26.43
25.39
BAT*
NA
NA
NA
NA
NA
NA
NA
NA
NA
NSPS
38.15
32.71
28.47
13.44
10.82
8.91
51.59
43.53
37.38
*Represents the incremental cost above BPT
362
-------
1ABLE 14-15. HISTORICAL EFFLUENT 1CNITORING
DATA SUMMARY WETH VARIABILITY FACTOR
TREATMENT WTTH IPDN REMOVAL
Daily Measurements
Subcategory: Titanium Dioxide
Chloride Process (Rutile Ore)
Plant 1559
April 76 through SepteKiber 78
TSS
Daily Data(1)
No. of Points 889
Average x, ppn 21
Standard gg g
Deviation, S "
Standard , ^A
Deviation, S'
Variability 11 n
Factor •"""
30-day (1)
Averages
No. of Points 30
Standard 21 8
Deviation
Variability . 04
Factor *'™
f2)
Variability
Factor Ratio
Pollutant
Cadmium Chrcmium Iron Lead
109 128 854 128
0.058 0.072 0.620 0.068
0.044 0.054 3.46 0.041
0.68 0.67 1.86 0.56
3.85 3.81 13.7 3.2
26 30 28 30
0.042 0.038 0.94 0.04
2.4 2.04 4.0 2.1
Nickel Zinc
128 128
0.080 0.15
0.07 0.20
0.76 1.02
4.4 6.4
30 30
0.05 0.16
4.4 3.1
VFR
3.6
1.6
1.9
3.4
1.5 1.0 2.1
(!) 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 rconthly averages.
363
-------
TABLE 14- 16. HISTORICAL EFFLUENT MONTTORJKG DATA SUMMARY WITH VARIABILITY
FACTORS
DAILY MEASUREMENTS
SUBCATBQQRY: TTEANIUM DIOXIDE-Chloride Process — Plant 1172
(Rutile/Upgraded Ilmenite Ore)
Pollutant
Daily Data '
No. of Points
Average x, ppn
Standard deviation, S
Standard deviation, S'
Variability factor
30-Day Averages
No. of Points
Standard deviation, S
Variability factor
O
Variability Factor Ratio *
VFR
TSS
454
5.39
9.13
1.16
7.6
15
6.31
2.92
!)
2.6
Chromium
454
0.0080
0.016
1.27
8.6
15
0.012
3.46
2.5
Copper
454
0.020
0.030
1.08
6.9
15
0.028
3.29
2.1
Zinc
454
0.020
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 monthly averages.
364
-------
TABLE 14-17. TREATMENT PERFORMANCE DATA OF SAMPLED PLANTS #559 AND #172
SOBCATEGORY: TITANIUM DICKIDE-Chlaride Process
Pollutant
TSS
Iron
Chromium
Lead
Nickel
Zinc
Plant #559
Pollutant
Concentration
(mg/U
Paw Treated
Waste Effluent
1103 23
288 4.4
13.3 0.030
0.50 0,002
0.50 0.005
0.30 0.060
Percent
Rsnoval
97.9
98.5
99.8
99.6
99.0
80.0
Plant #172
Pollutant
Concentration
Cmg/D
Raw Treated
Waste Effluent
171 6.7
2.9 0.33
0.72 0.020
0.005 0.002
0.08 0.010
0.30 0.090
Percent
Removal
96.1
88.6
97.2
60.0
87.5
70.0
365
-------
Raw Waste Concentration
Plant #559 was visited in the screening phase for sampling
of the raw and treated wastewater. 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 (see
above). 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 wastewater from the chloride process is mixed with the
sulfate process wastewater 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 since the sulfate process for the
manufacture of TiO, uses
an
ore of lesser purity. In
D.
addition, control of chromium should also provide adequate
removal of nickel, lead, and zinc which may coincidentally
occur at treatable levels. For these reasons, nickel, lead,
and zinc will not require effluent limitations. The
discussion of wastewater treatment chemistry in Sections 7
and 8 presents the basis for selecting one toxic metal
control parameter which effectively ensures the control of
the other metals present. The conventional and toxic
pollutants of concern include chromium and TSS. Iron, a
nonconventional pollutant, is also present as a major
impurity in the rutile or upgraded ilmenite ore and was
found at treatable levels in Plants #559 and #172.
2. 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 as shown in Table
14-17.
Basis of Pollutant Limitations
1. Conventional and Nonconventional Parameters
366
-------
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 the proposed
Development Document (60) and the JRB Study (52).
TSS Long-term effluent data is availabe for TSS from
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
intermixing of sulfate waste), the long-term average
concentration of 21 mg/1 (Table 14-15) calculated from
the monitored data submitted by Plant #559 is selected
as the treatment performance basis for the subcategory.
The variability factors of the daily measurements and
monthly averages {11 and 3.04) derived from the
long-term data of plant #559 and given in Table 14-15
are used to calculate the concentration basis. The
unit effluent limitations are calculated using the
model plant unit flow of 100 mVkkg. The calculations
are given below:
The maximum 30-day average concentration is:
(21 mg/1) (3.04) = 64 mg/1
The 24-hour maximum concentration is:
{21 mg/1) (11) = 230 mg/1
The maximum 30-day average effluent limit is:
(64 mg/1) (100 mVkkg) ( kg/m3 ) =6.4 kg/kkg
(1000 mg/1)
The 24-hour maximum effluent limit is:
(230 mg/1) (TOO mVkkg) ( kq/m^ ) = 23 kg/kkg
(1000 mg/1)
Iron
The Agency has decided not to promulgate limitations on
iron for existing sources because of the increased cost
of treatment when iron is controlled and because two
plants operate both the chloride process and the
sulfate process and send wastewater from both processes
to the same treatment facility. In order for these two
plants to treat the chloride process wastewater to
remove iron, they would either also have to treat the
367
-------
2.
sulfate process wastewater to remove iron or undertake
a massive reconstruction of the treatment facility.
Such a reconstruction or removal of iron in the sulfate
process would wipe out the recycle benefits and
treatment cost reduction associated with the final
BPT/BAT limitations for the sulfate process, and would
probably result in the closure of the two sulfate
process lines, with attendant increase in unemployment.
Control of toxic metal pollutants will be adequate at
existing plants even without an iron removal step. We
are providing guidance for use by permit writers in
cases where control of iron is warranted by
water-quality considerations.
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
14-15).
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 variability factor of
the daily measurement of 4.0 and the variability factor
of the monthly averages of 13.7 estimated from the
long-term monitored effluent data of Plant #559 for
iron (Table 14-15), and the model plant unit flow of
100 m3/kkg, the concentration basis is determined. The
limitations can be developed in the event that iron
becomes a concern. The concentrations are developed as
follows:
The maximum 30-day average concentration basis is:
(0.62 mg/1) (4.0) = 2.5 mg/1
The 24-hour maximum concentration basis is:
(0.62 mg/1) (13.7) = 8.5 mg/1
The 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 follwed by settling of acid mine
drainage waste-, and may not be appropriate for this
subcategory.
Toxic Pollutants
Chromium is the only regulated toxic pollutant because of
its presence at treatable levels in the raw waste of the
plants surveyed.
368
-------
Chromium:
The chromium limitation is based primarily on long-term
monitoring data available for Plant 1559. A waste treatment
system influent concentration for chromium of 13.3 mg/1 was
observed at Plant #559 during screening. The long-term data
summarized in Table ]4-15 indicates a current treatment
performance of 0.070 mg/1 and a monthly average and daily
measurement variability factor of 2.0 and 3.8, respectively.
Review of long-term average treatment data for chromium in
Tables 8-12 and 8-13 indicates that 0.070 mg/1 shows
excellent removal when compared to the same treatment
technology for chromium removal in other industries. A
long-term average of 0.15 mg/1 for chromium is used as the
basis for the final limitation. This value is in agreement
with the observed performance data from other industries
summarized in Tables 8-12 and 8-13. The monthly average and
daily measurement variability factors of 2.0 and 3.8,
respectively, are representative of BPT treatment
performance and are used for the purpose of establishing the
limitations.
The 30-day average concentration basis is:
(0.15 mg/1) (2.0) = 0.30 mg/1
The daily maximum concentration basis is:
(0.15 mg/1) (3.8) = 0.57 mg/1
The 30-day average effluent limit is:
(0.30 mg/1) (100 mVkkg) ( kg/m* ) = 0.030 kg/kkg
{1000 mg/1)
The daily maximum effluent limit is:
(0.57 mg/1) (100 mVkkg) ( kq/m^ ) * 0.057 kg/kkg
1000 mg/1)
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. The concentration limitations for
the three pollutants are given and are intended to serve as
guidance in cases where the pollutants are found to be of
concern.
The long-term average and corresponding variability factors
are based on long-term data from Plant #559 and data
presented in Table 8-12 for lead, nickel, and zinc. In all
369
-------
cases, the long-term average observed for Plant 1559 was
adjusted after evaluation of treatment performance for other
industries summarized in Table 8-12. The long-term averages
for lead, nickel, and zinc are 0.15 mg/1, 0.10 mg/1, and
0.20 mg/1, respectively, based on review of the table.
Selection of variability factors was made directly from
Table 14-15 for the appropriate pollutants. The variability
factors corresponding to chromium (the primary pollutant of
concern) were selected for nickel.
The concentration basis is determined as follows:
The 24-hour maximum concentration for lead is:
(0.15 mg/1) (3.2) - 0.48 mg/1
The maximum 30-day average concentration for lead is:
(0.15 mg/1) (2.1) = 0.32 mg/1
The 24-hour maximum concentration for nickel is:
(0.10 mg/1) (3.8) = 0.38 mg/1
The maximum 30-day average concentration for nickel is:
(0.10 mg/1) (2.0) = 0.20 mg/1
The 24-hour maximum concentration for zinc is:
(0.20 mg/1) (6.4) = 1.28 mg/1
The maximum 30-day average concentration for zinc is:
(0.20 mg/1) (3.1) = 0.62 mg/1
The limitations for BPT presented in Table 14-18.
Basis for BCT Effluent Limitations
While EPA has not yet proposed or promulgated a revised BCT
methodology in response to the American Paper Institute v. EPA
decision mentioned earlier, EPA is promulgating BCT limitations for
this subcategory. These limits are identical to those for BPT because
the only technology option that removes significant amounts of
conventional pollutants is not economically achievable. See the
discussion of iron under BPT limitations above. Removal of
significant additional amounts of conventional pollutants can be
achieved in this subcategory only if iron is also removed.
As BPT is the minimal level of control required by law, no possible
application of the BCT cost tests could result in BCT limitations
lower than those promulgated today. Accordingly, there is no need to I
370
-------
TABLE 14-18. EFFLUENT LIMITATIONS
TITANIUM DIOXIDE - CHLORIDE PROCESS (RUTLLE OR UPGRADED ILMENITE ORE)
Best Practicable Control Technology Currently Available
Waste Water Flow: 100 m3Akg of TiO0
Subcategory
Pollutant Performance 1
(mg/1)
Concentration
p. Basis Cmg/1)
Max
30-day
Avg
24-hr
Max
Effluent Limit
(kg/kkg of TiO-)
Max
30-day
Avg
24-hr
Max
Conventional and
Nonconventional 'Pollutants:
Total Suspended
Solids
Iron
Toxic Pollutants:
21
(2)
0.62
(2)
11.0/3.0
13.7/4.0
64
2.5
230
8,5
6.4
23
(4)
(4)
Chromium
Lead
Nickel
Zinc
m
0.15u;
(3}
0.15^'
(3)
0.10U'
(3)
0.20U'
3.8/2.0
3.2/2.1
3.8/2.0
6.4/3.1
0.30
0.32
0.20
0.62
0.57 0.030
f a\
0.48 —w
(A.}
0.38 — w
/^\
1.28 — w
0.057
(4)
—
M\
—
(4)
(1) VFR: Ratio of the variability factor of the daily measurements to the
variability factor of the monthly averages.
(2) Long term average based on data and variability factors of plant #559
selected from Table 14-15.
(3) Limitation based on long term data at plant #559 and review of Table 8-12,
(4) No effluent limitation.
371
-------
wait until EPA revises the BCT methodology before promulgating BCT
limitations.
Basis for 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 for BAT and NSPS but were rejected on the basis of cost.
Level 1, used for BPT, is selected for BAT treatment technology.
A. Technology Basis
Alkaline precipitation followed by settling used for BPT (Level
1) is the same as for BAT.
B.
Flow Basis
A unit wastewater flow rate of 100 m3/kkg of TiO2 used for the BPT
model plants has been selected for BAT.
C. Selection of Pollutants to be Regulated
Chromium is the pollutant identified for regulation.
1. Nonconventional Pollutants
The concentration basis was developed for iron which can be
used for guidance in the event that iron becomes an
environmental concern.
2. Toxic Pollutants
a. Chromium
The limitations for BAT are the same as selected for
BPT.
b. Other Metals
Lead, nickel, and zinc are not limited. However,
achievable concentration levels are contained
in this document for use as guidance if these pollutants are
found to be of concern. The values are the same as those selected for
BPT.
Table 14-19 gives the limitations for BAT.
372
-------
TABLE 14-19. EFFLUENT LIMITATIONS
TITANIUM DIOXIDE-CHLORIDE PROCESS (RUTILE/UPGRADED ILMENITE ORE)
Best Available Technology
Wastewater Flow: 100 m /kkg of TiO2
Pollutant
Subcategory
Performance*1
(mq/1)
Concentration Effluent Limit
Basis (mg/1) ^kg/kkg of TiOg)
Max. Max.
30-day 24-hr. 30-day 24-hr,
Avg. Max. Avg. Max.
Nonconvent ional
Pollutants:
Iron
Toxic Pollutants:
0.62
13.7/4.0
2.5
8.5
Chromium
Lead
Nickel
Zinc
0.15
0.15
0.10
0.20
3.8/2.0
3.2/2.1
3.8/2.0
6.4/3.1
0.30
0.32
0.20
0.62
0.57 0.030
0,48 — (3)
0.38 -<3)
1.28 -(3)
0.057
— (3)
— (3)
— (3)
(1) See Table 14-17 for details.
(2) VFR: Ratio of the variability factor of the daily measurements to the
variability factor of the monthly averages.
(3) No effluent limitation.
373
-------
Basis for New Source Performance Standards
A. 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.
B. Technology Basis
For new plants, the recommended wastewater 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 neutralization and settling, but only
published treatability data and the results of our treatability
study are available on the preformance of dual-media filters.
C. Flow Basis
The raw effluent flow rate is the same as that used for BAT,
namely 100 m3/kkg of TiOz. It is assumed that the unreacted ore
and coke are slurried with water and sent to the treatment
system. The selected flow .value is an average of the unit
effluent flow rate of two plants (#559 and #199) practicing this
method of solids disposal.
D. Selection of Pollutants to be Regulated
The pollutants regulated for BPT are also regulated for NSPS in
addition to iron. 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 a
concern.
NSPS are based on the addition of an iron removal step to BPT/BAT
treatment. New plants can achieve significant reductions in
wastewater flow, thus reducing overall treatment costs, even with
the inclusion of the iron-removal step. See also the discussion
of NSPS for the Sulfate Process in this section.
T. 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
equired on the basis of data presented in Appendix B of
the proposed Development Document (60) and the JRB
Study (52).
374
-------
TSS
Since there is no in-plant performance data available on the
addition of filtration to the treated wastewater effluent,
treatability studies were performed on a chloride process
raw wastewater to estimate the effectiveness of dual-media
filtration for the removal of the selected pollutants (61).
Results of these studies support the use of filters to
achieve a 38 percent additional .removal for TSS. This
reduction is applied to the selected BPT maximum 30-day
average of 64 mg/1. The maximum 30-day average
concentration basis is then given by: (64 mg/1} (1.00-0.38)
= 40 mg/1. Likewise, the 24-hour maximum concentrations and
unit effluent limitations are obtained from the BPT
limitations (Table 14-18) as shown below.
The maximum 30-day average concentration is:
(64 mg/1)(0.62) = 40 mg/1
The 24-hour maximum concentration is:
(230 mg/l)(0.62) = 140 mg/1
The maximum 30-day average effluent limit is:
(6.4 kg/kkg)(0.62) = 4.0 kg/kkg
The 24-hour maximum effluent limit is:
(23 kg/kkg)(0.62) = 14 kg/kkg
Nonconventional Pollutants
a. Iron
The Treatability Study (61) which tests the
effectiveness of dual-media filtration on the removal
of iron from the lime treated and clarified wastewater
from BPT treatment support a removal efficiency of 38
percent. The concentration basis and effluent
limitation for NSPS are obtained by multiplying the
selected BAT (or BPT) limitations (Table 14-18) by 0.62
as follows:
The 30-day average concentration basis is:
(2.5 mg/1)(0.62) = 1.6 mg/1
The 24-hour maximum concentration basis is:
(8.5 mg/1)(0.62) - 5.3 mg/1
375
-------
The 30-day average effluent limt is:
(0.25 kg/kkg)(0.62) =0.16 kg/kkg
The 24-hour maximum effluent limit is:
(0.84 kg/kkg) (0.62) = 0.52 kg/kkg
3. Toxic Pollutants
a. Chromium
Results of the Treatability Study (61) indicate that an
additional 60 percent removal of chromium can be achieved
using a dual-media filtration polishing step. Therefore,
the limitation is developed by application of a 0.60
reduction factor to the BPT/BAT limitation in Table 14-16 as
follows:
(BPT/BAT limitation) (1.00-0.60) = 0.40 (BPT/BAT limitation)
The maximum 30-day cencentration basis is:
(0.30 mg/l)(0.40) =0.12 mg/1
The daily maximum concentration basis is:
(0.57 mg/1)(0.40) =0.23 mg/1
The maximum 30-day average effluent limit is:
(0.030 kg/kkg) (0.40) = 0.012 kg/kkg
The maximum daily effluent limit is:
(0.57 kg/kkg) (0.40) = 0.23 kg/kkg
b. Other Metals
Treatability studies have indicated that the following
increased removals of lead, nickel, and zinc can be achieved
by filtration (40,41,61).
Additional Removal by
Filtration Using Settled Effluent
Lead
Nickel
Zinc
80!
The additional levels of removal are applied to the
corresponding BAT (or BPT) concentration for the above
metals to get the NSPS concentrations.
376
-------
The maximum 30-day average lead concentration basis is:
(0.32 mg/1)(0.20) = 0.064 mg/1
The 24-hour lead concentration basis is:
(0.48 mg/1)(0.20) = 0.096 mg/1
The maximum 30-day average nickel concentration basis
is:
(0.20 mg/1)(0.86) =0,17 mg/1
The maximum 24-hour concentration basis for nickel is:
(0.38 mg/l)(0.86) = 0.33 mg/1
The maximum 30-day average concentration basis for zinc
is:
(0.62 mg/1)(0.94) =0.58 mg/1
The 24-hour maximum concentration basis for zinc is:
(1.28 mg/1)(0.94) =1.20 mg/1
The conventional, nonconventional, and toxic pollutant
limitations for NSPS are given in Table 14-20.
Basis for Pretreatment Standards
Existing Sources
Since there are no indirect dischargers in this subcategory, we are
not promulgating PSES but are instead excluding this subcategory from
categorical PSES under the provisions of paragraph 8(b) of the
Settlement Agreement.
New Sources
For Pretreatment Standards for New Sources (PSNS), the Agency is
basing limitations on NSPS. The pollutants to be regulated are iron
and chromium ('See Table 14-20). These pollutants are regulated
because NSPS provides better removal or iron and chromium than is
achieved by a well-operated POTW with secondary treatment installed
and therefore iron and chromium would pass through the POTW in the
absence of pretreatment.
Titanium Dioxide - Sulfate Process Industry Profile
377
-------
TABLE 14-20. EFFLUENT LIMITATIONS
TITANIUM DIOXIDE - CHLORIDE PROCESS
New Source Performance Standards
Waste Water Flow: 100 m3/kkg of TiO2
Pollutant Treatability
Conventional and
VFRC1)
Concentration Effluent Limit
Basis (mg/1) (kg/kkg of TiO-)
Max
30-day
Avg
Max
24-hr 30-day
Max Avg
24-hr
Max
Nonconventional Pollutants:
Total Suspended
Solids
iron'2'
Toxic Pollutants:
(2)
Chromium
Lead
Nickel
Zinc
13
0.38
0.060(4)
{&}
0.030 w
0,086(4)
(&\
0.13t4)
n.o/3.0
13.7/4.0
3.8/2.0
3.2/2.1
3.8/2.0
6.4/3.1
40
1.6
0.12
0.064
0.17
0.58
140 4.0
5.3 0,16
0.23 0,012
CVi
0.096 — ^
0.33 -<3>
I 1\
11 \ -5/
.2 —
14
0.52
0.023
( 3)
—
-J3>
/ri\
(1) VFR: Ratio of variability factor of the daily measurements to the
variability factor of the monthly averages.
(2) Also applicable for PSNS limitations.
(3) No effluent limitations.
(4) Long term average from Table 14-18 multiplied by the appropriate percent
removal factor indicated in Section 14.7.5.
378
-------
TABLE 14- 21. SUBCATEGORY PROFILE DATA SUMMARY
SUBCATEGORY
TTTANIUM 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:
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volume per unit product
Minimum
Maximum
401,000 kkg/year
259,000 kkg/year
4
5
320,000 kkg/year
246,000 kkg/year
80 percent
95 percent
31,000 kkg/year
74,500 kkg/year
49,000 kkg/year
43,000 kkg/year
76 percent
23 years
54 years
35,000 cubic meters/day
125,000 cubic meters/day
300 cubic meters/kkg
780 cubic meters/kkg
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of 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. Jr
379
-------
TABLE 14-22. ANALYSIS OP ILMENITE ORES1
Chemical
Constituent
Ti02
PeO
Fe203
Si02
A1203
P2°5
w
00
Zr02
MgO
MnO
CaO
V2°5
Cr203
Virginia
Piney
River Roseland
44.3
35,9
13.8
2.0
1.21
1.01
0.55
0.07
0.52
0.15
0.16
0.27
51.4
37.9
1.6
4.6
0.55
0.17
NA
2.35
0.70
0.59
0.07
NA
UNITED STATES
New York Florida
44.4
36.7
4.4
3.2
0.19
0.07
0.006
0.80
0.35
1.0
0.24
0.001
64.1
4.7
25.6
0.3
1.5
0.21
NA
0,35
1.35
0.13
0.13
0.1
California
48.2
39.1
10.4
1.4
0.2
NA
0.05
0.6
0.1
0.1
0.05
0.03
Ivry
42.5
39.1
20.7
0.88
1.05
NA
NA
2.0
0.04
0.1
0.36
0.15
CANADA
Bourget
22.4
36.9
31.2
1.0
6.01
0.93
NA
1.50
NA
0.55
NA
NA
Allard
37.3
26.3
30.0
NA
NA
0.004
NA
NA
0.10
NA
0.39
NA
Constituents expressed as weight percent.
NA: Not Available
-------
General Description
The industrial profile for the Sulfate Process Segment of the Titanium
Dioxide Subcategory is presented in Table 14-21 and the status of
regulations prior to promulgation of this regulation is shown in Table
14-2
General Process Description and Raw Materials
A. Sulfate Process - General Description
Among the various titanium ores, ilmenite is availabe in abundance.
Ilmenite is a low-grade titanium ore with a Ti02 content varying from
45 to 60 percent. JJLmenite 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 14-22 gives the
analysis of various ilmenite ores. The preparation of TiO2 by the
sulfate process utilizes three important steps:
1. Digestion:
FeO.TiOz + 2HZS04 = FeS04 + TiO.SO4 + 2H20
2. Precipitation:
TiO.S04 + 2H20 = TiOz.H20 + H2S04
3. Calcination:
TiOz.H20 = TiOz + HZ0
The ore is dried, ground, and then reacted with sulfuric acid. The
reaction takes place at 160° 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° C when hydrated ferrous sulfate (FeS04.7HzO)
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. The solution
is filtered, and the filtrate, known as strong acid, is separated and
either discharged or recycled. The Ti02.HzO 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 is known as weak acid. The product is then rinsed to remove
381
-------
OJ
00
to
STRONG-ACID *
HATER — »
DIGESTER
HATER , n.
CLARIFIER
*
J EVAPORATOR
4
STEAM _*
PRECIPITATION
t
— EMISSIONS »•
HATBR ^* AND VEHTURI SCRUBBERS »•»-—•-
w EMISSIONS — ,_^ ' " "- .^
HATER CONDENSERS — EFFLUENT -^
COOLER P MI"0iUBB """*; CONDENSERS U WPLUBHT ~»
WEAK -AC ID 1 1
RBCYCLB ^t * 1_ J —
HATER— to
FIRST MOORE
rn.TBn
HATER—*
SECOND MOORE
FILTER
STEMl ^— ' I '— ' •
HATBR fr>
STEAM ft»
DIl
CALCIHER
i
HBT HILL
rBHlBajuna » COOLING SPRAYS AND ELECfc fc HH.K ncm.
HATER fc FROSTATIC PRECIPITATORS |
EMISSIONS
*
HATER »J MIST ELIMINATORS j~ EFFLUENT -•*•
*
1
JET MILLS
WATER ^^^-toi 1
1
TITANIUM
JXIDB PIGMENT
t
PACKAGING
MATER » JET MILL SCRUBBERS 1— . ». EFFLUENT
TO SALES
Figure 14-7. General process flow diagram for production of titanium dioxide by
sulfate process.
-------
TABLE 14-23. WATER USAGE IN TITANIUM DIOXIDE - SUI*FATE PROCESS SUBCATEGOKY
Uses
Water Usage per Unit of Production
of Ti02)
Noncontact
cooling
Direct process contact
indirect pr
ocess contact
Plant #555
47.8
390
6
Plant #694
408
588
1.6
Plant #696
149
297
4
(pumps, seals, leaks,
spills, etc.)
Maintenance, equipment
cleaning and work area
washdown
Air pollution control
Noncontact ancillary
uses (boilers, utilities,
etc.)
1.8
258
36
78
33
81
NA
NA: Not Available
383
-------
iron and unreacted acid. Residual acids in the precipitate are
removed along with the water of hydration by calcination. The
resulting TiOz 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:
pulping, 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 14-7.
Water Use and Waste Source Characteristics
Water Use
Water is used in the preparation of titanium dioxide by the sulfate
process for noncontact cooling, air emission control, and for process
reactions. In the process, water is used to leach the soluble sulfate
salts from the reaction mass and to convert the titanyl sulfate to
titanium dioxide hydrate. Water is also used to wash the titanium
dioxide hydrate precipitate free from residual acid and iron. Water
is used for air emission control during the drying of ore, on digester
units, and for the cleaning of the kiln gases before they are vented
to the atmosphere. In the digester unit, water seals are used to
maintain a vacuum on the digester units. Large amounts of water are
also used in the finishing operations. Table 14-23 is a summary of
water usage in the titanium dioxide subcategory using the sulfate
process.
Waste Sources
A. Digester Sludge
After the digestion of the ore in suIfuric 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.
B. 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 contains small
amounts of adsorbed sulfuric acid.
C. Strong Acid Waste
When water is added to the titanyl sulfate solution after
the removal of copperas, sulfuric acid and the hydrate of
384
-------
titanium dioxide 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.
D. 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.
E. Scrubber Wastes
Scrubber wastewater results from the scrubbing of vapors
emitted during the drying of the ore, during digestion, and
during kiln drying. The amount of wastewater generated
depends on the amount of water used and type of emission
controls practiced. The scrubber water contains titanium
dioxide particulates, acid mist, sulfur trioxide and sulfur
dioxide. Of all the waste produced from titanium
dioxide-sulfate process manufacture subcategory, the
scrubber wastewater constitutes the major portion.
F. 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 wastewater from wet finishing operation,
therefore, contains titania, sodium sulfate, and other
agents added to improve or achieve desired properties in the
final product.
Description of Plants
Screening
Plant #555 was visited and its waste streams sampled in the screening
phase by an EPA Region II team. The pigment manufacturing operation
385
-------
TABLE 14-24. HAW WASTE CHARACTERISTICS (INDUSTRY DATA) ^FOR PLANT 1555
(PRODUCTION OF TiO2 BY SULFATE PROCESS)
Waste Source
Unit
Flow
(m3/kkg
of TiC>2)
pH
Pollutant Waste Loads, (kg/kkg of Ti02)
Acidity NH3
(asH2S04) (asN)
Fe TSS TDS
Digestion 115 3.0
Clarification 3.58 2.5
Evaporation 113 4.0
Cooling 20 6.1
Strong Acid fron 8.49 < 0.5
first Moore Filtration
Weak Acid from 12.2 2.0
first Moore Filtration
Weak Acid from 10.4 1.7
second Moore Filtration
Weak Acid from 12.0 2.0
first stage
Calcination
Weak Acid from 40.0 2.2
second stage
Calcination
20.8
26.7
18.7
2.49
2.360
88.3
148
20.8
19.2
NA 0.042 9.3 35.7
NA 8.42 175 40.8
NA 1.14 3.2 20.2
NA 0.099 0.46 3.09
NA 139 0.959 2.815
NA 3.8
NA
0.64
NA: Not Available
* Value in ffl units
(1) - Response to 308 Questionnaire, 1976
0.23 98.8
NA 0.29 0.13 151
NA 0.22 2.0 7.50
4.92 33.1
Calcination Mist
Eliminators
Wet Milling Washing
and Drying
Jet-Mill Condenser
Jet-Mill Scrubbers
Boiler and Water
Plants
38.
11.
27.
18.
16.
7
1
0
0
6
3.
8.
6.
7.
9.
0
0
5
4
0
7.50
NA
NA
NA
NA
NA
8.6
NA
NA
NA
0.
0.
0.
0.
0.
02
01
01
13
66
0.
2.
1.
1.
5.
21
13
1
7
25
27.9
11.0
2.7
3.58
8.92
386
-------
utilizes a titania slag for the production of TiOz by the sulfate
process. After digestion of the slag in sulfuric acid the residual
gangue material is filtered out and the clear liquor is concentrted 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 14-24 presents raw waste flows and pollutant
characteristics for Plant #555.
Wastewater samples were collected at five points and analyzed for the
conventional, nonconventional, and toxic pollutants. These sampling
points were designated as 1) the 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.
Verification
Plant #559 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 COZ 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 to raise the pH and precipitate more
calcium sulfate. Air is also introduced to convert the ferrous iron
to ferric iron. 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 wastewaters. These include the scrubber,
finishing, and cooling wastewaters. The combined water is neutralized
with slaked lime before it is sent to a final settling pond, the
effluent from which is discharged. Figure 14-8 gives the flow diagram
of the treatment process and shows the sampling locations for both
screening and verification. Table 14-25 gives the flow data for the
waste streams and conventional and nonconventional pollutant
emissions.
387
-------
WEAK ACIt - fc
UAJTI ITftlJM
OTMCR PRODUCT
WASH UATU
VCAR ACID POND
n
HI VI* ^~1
SUPPLY VATIR " '" ™'^J "
H
S
\
OJ
03
UASTC STMAH
UPPLV UATfh
UPPLY UATER
PQHO
n
8
10
n
UACTORS
FILTIR
SOLIDS
TO
i
REACTORS
riu»
I
— ~ ~~* *"" FINAL ^-^
h
SOLI OS TO
STORA6E/ f
,
POHO ~m*
, , "
*
j OTHER PRODUCT
^ WASTI WATCR
TIO, (SULfATC PROCESS)
1 SCRUIICR
WASTE WATCR
STflHACI/LANOF.Ul
LECENO
mm SAMPLING POINTS
Figure 14-8. General flow diagram at plant #559 showing the sanpling points.
(Titanium dioxide - sulfate process.)
-------
TABLE 14-25. FLOWS AND POLLUTANT CONCENTRATIONS FOR THE WASTE STREAMS
SAMPLED FOR PLANT #559 PRODUCING TITANIUM DIOXIDE
Sampled
Stream Stream
(4)
No
Description
Unit
Flow
(m3/kkg)
TSS Iron
Load Load
(kg/kkg of TiO2) (kg/kkg of TiO2)
Weak Acid
Pond Overflow
1.23
1.23
Strong Acid
Pond Overflow
6.1
205.9
106,3
Scrubber and
Contact Cooling
Water
361.4<1>(2> 113.5
51.68
Final Treatment
Effluent
436(1)(2)(3) 10.0
1.92
(1) The flow is contributed by the sulfate process stream.
(2) The pollutant load was calculated by multiplying the flow contributed by the
sulfate process stream times the concentration of pollutant. Pollutant Load
= (total stream flotf) x (fraction contributed by sulfate process waste) x
stream pollutant concentration.
(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 14-8 for sampling point location.
389
-------
Other Plant Descriptions
At Plant 1694, 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 were sent to thickeners to remove the
suspended solids and the overflow was discharged. Depending on the
titanium content, the underflow from the thickeners was 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, which is the coarse gypsum
slurry, is separated from the reactor effluent at a concentration of
85 to 90 percent, and placed in a self-draining dewatering system.
The "dry" solids are finally trucked to a landfill. The second
fraction separated in the hydrocyclone is a fine gypsum slurry which
is recycled to the neutralization reactor. The residual gel slurry
forms the third fraction, and this is sent to a thickener after C02
degassing. A flocculating agent is added to the flow to the thickener
to promote solids separation and thickening. The underflow from the
thickener is centrifuged and the solids landfilled. The filtrate from
the centrifuge is recycled to the thickener, and the thickener
overflow is discharged.
The volume and characteristics of wastewater streams from different
sulfate process titanium dioxide plants do not differ greatly. Some
variations, however, are noted as a result of differences in ore
qualities, in location and in process details. The majority of the
dissolved pollutants in wastewater from this segment of the Ti02
industry consist of acidity and iron. Segregation of the wastewater
is important for control and treatment practices and aids in
developing economically feasible treatment systems. Generally, weak
and strong acid streams are segregated from each other as well as from
the less contaminated wastewaters 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 14-26.
The average total effluent flow rate is 475 mVkkg (Table 14-16) for
Plants #555, #694, and #559. Complete flow data is not available for
Plants #696 and #605.
390
-------
TABLE 14-26. PROCESS WASTE WATER FLOW AT PLANTS # 555, #694 and #559
TITANIUM DIOXIDE (SULFATE PROCESS)
Plant
#555
#694
#559
Average
Strong
8.5
16
6.1
10
Flow in (irr/kkg
acid Weak acid
78.2
67
69
72
of Ti02)
Scrubber and
contact cooling
water
362
457
361
393
Total Effluent
449
540
436
475
391
-------
Toxic Pollutant Concentrations
Section 5 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 wastewaters, 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 0.0070 mg/1.
This is well below the treatability level for phenol, therefore,
phenol is not considered a significant or process related pollutant.
Daily raw waste loads were calculated from the flow rates measured or
estimated at the time of sampling and the measured pollutant
concentrations. That is,
daily loading (as kg of pollutant per day) - (C)(Q)
1000
where the concentration (C) of the pollutant is expressed in
units of mg/1 (Note: 1 kg/m3 = 1000 mg/1), and the flow rate (Q)
expressed in units of mVday ( a cubic meter, m3, 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 = (C)(Q)
per day kkg of TiOa) 1000(P>
where C and Q are expressed in the same units described above,
and the production (P) is expressed in units of kkg/day {kkg is
1000 kg, a metric ton, which is equal to 2205 Ib).
The maximum concentration of toxic pollutants found in the raw waste
at concentrations above the treatability level in the screening and
verification program were:
392
-------
TABLE 14-27. SIM-1ARY OF RAW HASTE LOADINGS fOCND IN SCREENING AND VERIFICATION SAMPLING
W
U>
U)
SUBCftTEGORY
Pollutant
Toxic
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Nickel
Selenium
Thallium
Zinc
Conventional
TSS
Iron, Pe
TITANHH DIOXIDE - SULFATE PROCESS
Loading Range,
(kg/day)
Miniimm Maximum
5.0 28
1.9 4.0
.068 7.2
140 530
8.2 19
3.0 65
3.7 23
7.0 9.5
0.47 1.2
1.8 85
and Nanconventional
Minimum
0.032
0.012
0.00044
1.1
0.065
0.024
0.029
0.0020
0.0030
0.014
Unit Loading,
(kg/kkg)
Average
0.11
0.19
0.019
2.0
0.085
0,18
0.080
0.031
0.0055
0.34
320
600
Maximum
0.22
0.032
0.057
3.4
0.12
0.42
0.15
0.060
0.0080
0.55
No. of
Plants W
3
3
3
3
3
3
3
2
2
3
1
1
(1) - Data are taken only from those plants where pollutants were found above detection limits, or, in the
case of TSS and Iron, where data are available.
-------
TABLE 14-28. TOXIC POLLUTANTS; AVERAGE RAW, WASTE LOADS AND OONGEKTRATIONS
w
•£>
SUBCATEGORY
TTTANIIM DIOXIDE - SULEKTE PROCESS
Screening
Antimony
Arsenic
Cadmium
Chrcmiun
Copper
Lead
Nickel
Selenium
Thallium
Zinc
(m3/l)
0.77
0.11
0.29
3.8
0.20
0.075
0.091
NA
NA
0.088
Plant I5S5
(kg/kkq)
0.22
0.032
0.057
1.1
0.065
0.024
0.029
< 0.06
NA
0.014
Plant
(mg/1)
0.16
0.029
0.0020
7.0
0.25
0.20
0.31
NA
0.020
1.1
1559
(kg/kkg)
0.080
0.014
0.0009
3.4
0.10
0.15
NA
0.0080
0.55
verification
Plant
(irci/1)
0.074
0.028
0.0010
3.1
0.96
0.14
0.0050
0.0070
1.04
1559
(kg/kkg)
0.032
0.012
0.00041
1.4
0.070
0.42
0.061
0.0020
0.0030
0.45
Average
(nw/1)
0.34
0.06
0.10
4.63
0.20
0.41
0.18
0.005
0.014
0.74
NA = Not Available
-------
Maximum Concentration Observed (ng/1)
Screening Verification
Pollutant (Plants #555 & #559) (Plant #559)
Cadmium 340 12
Chromium 124,000 31,000
Copper 1,500 1,000
Lead 3,700 5,200
Nickel 6,400 1,300
Zinc 3,800 1,700
A summary of daily and unit (per unit of production) raw waste loads
for all plants sampled can be found in Table 14-27. Individual plant
raw waste loads and concentrations found in sampling are given in
Table 14-28.
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:
Pollutant Total Annual Raw
Waste Load (Kg/year)
Cadmium 5,000
Chromium 510,000
Copper 22,000
Lead 47,000
Nickel 21,000
Zinc * 88,000
Antimony 29,000
Arsenic 49,000
Selenium 8,000
Thallium 1,400
Pollution Abatement Options
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 wastewaters. These values are shown above. Using
cadmium as an example of a borderline case, its maximum observed
concentration of 0.34 mg/1 (340 ^g/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 ultimately selected as a basis for regulations
may not be as effective as the most advanced technology considered at
this stage of the evaluation of alternatives.
395
-------
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 maximum
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 14-22. 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 the 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 wastewaters.
Process Modifications and Technology Transfer Options
Specific process modification recommendations are not made. However,
several areas for further investigation suggest themselves. They are:
A. 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.
B. 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.
C. If markets could be developed for the sale of ferrous sulfate
(copperas), solid waste disposal problems wouId be reduced.
Currently, a portion is sold and the rest disposed of as a solid
waste.
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.
Prevailing Control and Treatment Practices
The treatment practices of the plants producing TiOz by the sulfate
process are given, above.
396
-------
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 costs incurred are high
because of the large quantity of water (more than 400 m3/kkg of Ti02)
that must treated. The sulfate process is one of two subcategories
(the other being the Soda Ash Solvay Process) in the Inorganic
Chemicals Industry studied in this report that generates the largest
quantities of waste effluent.
Selection of Appropriate Technology and Equipment
Technologies for Different Treatment Levels
A. Level 1 (BPT/BAT)
In the Level 1 treatment, the strong and weak acid streams are
independently equalized in lined lagoons and neutralized in a reator
with ground limestone to a pH range of 5 to 5.5, precipitating calcium
sulfate and forming carbon dioxide, which remains in solution, along
with large quantities of ferrous iron. Solids are generated in a
thickener at the unit rate of 3 kkg per kkg of products. The
thickener overflow joins the large but relatively unpolluted flow of
"other wastes", and the combined wastewaters are treated to pH 6 to 9
and passed through a polishing lagoon prior to discharge. Level 1
treatment allows recovery of gypsum solids without contamination by
iron and toxic metals pollutants. The flow schematic for Level 1 is
presented in Figure 14-9a.
B. Level 2 (NSPS)
In the Level 2 treatment, the blended strong and weak acid streams are
neutralized with calcium carbonate. Most of the toxic metal
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 toxic metal pollutants. The combined stream is then
given lime treatment to pH 9 and settled in polishing lagoons before
discharge. This three-step system is patterned after an existing
system which separates 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 more
toxic 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
additional heavy metals present, very similar to the ferrite
coprecipitation method described in the Treatment Technology
397
-------
lO
CO
MCTE HUTCH
WOTS
&
SCUDS OZSPOSALOf
SHE OR
BJUJBfT
* InclulM flat BocUterlrq, fti ncrUtarlnq
•nd
Figure 14-9a. Level 1 waste water treatment for titanium dioxide - sulfate process.
-------
EFFLUENT
WASTE mi'
CM SHE
Includes flow iiioniloi-ing, pH monitoring and sampler
Figure 14-9b. Level 2 waste water treatment for titanium dioxide
sulfate process.
-------
Assessment section. The flow diagram of the treatment system is shown
in Figure 14-9b.
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).
C. Level 3
Level 3 for the sulfate process employs the described NSPS treatment
for stong acid, weak acid, and 55 percent of the "other wastes". The
remaining other wastes receive soda ash treatment and settling, to
permit recycling a nonscaling effluent for scrubbers and miscellaneous
uses. Toxic metal pollutants in the separated recycle stream are
settled as hydroxides or carbonates, plus any calcium arsenate formed,
and periodically removed to secure landfill. The flow diagram of this
treatment is shown in Figure 14-10.
Euipment for Different Treatment Levels
A. Equipment Functions
Treatment of wastewater 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 and one stage thickening to recover gypsum.
Calcium saturated thickener overflow and .miscellaneous other
wastes are subjected to pH adjustment and one-day settling in a
polishing pond prior to discharge.
In Level 2 or NSPS, wastewater treatment requires the use of
mechanical aerators, lime addition, and second stage thickening
equipment in addition to equipment specified for BPT treatment to
remove additional toxic metals and iron.
In Level 3, to reduce the raw waste load of toxic metals, only 55
percent of the NSPS "other waste" flow joins the treated acid
waste stream, for Level 2 treatment as described above. However,
the remaining 45 percent of "other waste" is given separate
treatment with soda ash and 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 NSPS model.
B. Chemicals and Handling
First stage neutralization employs ground limestone, while lime
is used in cases where second stage and final alkaline
4CO
-------
__. BECTfCUC
(MISC. USES)
UUESTONE
C.OO.
OTHER
WABTKS
m
RAPID UIX AND 8ETTUNO
MECYCLBO BrrutENT
O
crruiFHT
Figure 14-10. Level 3 waste water treatment for titaniun dioxide - sulfate process.
-------
precipitation are employed. Oxygen is supplied from atmospheric
air, and polymer is added to assist in the second stage settling
of iron hydroxide when required. Aside from the bulk handling of
large amounts of these common chemicals, there are no special
hazards involved in their use.
C. Separation and Removal of Solids
Large quantities of thickener underflow are pumped to spreading
areas for consolidation of the solids, which are later pushed to
18 foot high piles on land provided for 10 years of operation.
Solids from occasional draining of the polishing lagoon and the
Level 3 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 are sources of pure and impure
gypsum for future byproduct recovery.
Treatment Cost Estimates
General Discussion
To prepare
developed.
treatment cost estimates, a model plant concept was
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 TiQ2
(Sulfate Process) subcategory. 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.
A. Wastewater Flow
Waste effluent data is available for three plants and is given in
Table 14-26. For the model the average value of the three plants
data has been used. The unit flow data for strong acid ranges
from 6.1 to 15 mVkkg of Ti02. (Table 14-26). For the model
plant the average value of 10 mVkkg has been used. Unit flows
for the weak acid stream range from 67 to 78 m3/kkg. For the
model plant, a unit flow of 72 m3/kkg of Ti02 is used. The third
segregated stream includes contact cooling water, scrubber water,
and finishing operation wastewater. 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 total flow which consists of strong acid, weak acid,
scrubber effluents, and product finishing wastewaters of 475
mVkkg of Ti02 was used.
402
-------
B. Production
Four 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 kg/yr,
and 74,500 kkg/yr.
C. Wastewater Pollutant Load
As stated before, the principal pollutants occur in the strong and
weak acid streams and include high acidity (sulfuric acid),
suspended solids, iron and other heavy metal sulfates. The other
wastewaters contain titanium dioxide and small amounts of other
heavy metals on suspended solids. Iron concentrations vary
depending on the grade of ilmenite ore used.
Model Plant Control and Treatment Costs
The average raw waste pollutant loadings given in Table 14-27 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 spended
ferric hydroxide and one-third (200 kg/kkg of TiO2) was soluble
ferrous iron. The unit sulfate and suspended solid loadings for the
different wastewater streams for the model plant were:
Stream
Weak Acid
Strong Acid
Other Wastewater
E. Chemical Usage
Sulfate Loading
(kg/kkg of Ti02)
2,300
1,800
Negligible
TSS Loading
(kg/kkg of 1i02)
300
200
113
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 TiOz. For iron removal (NSPS), 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.
F. Solids Produced
Existing plants have attempted to produce two grades of saleable
gypsum from the strong and weak acid streams. At present, one plant
has successfully identified a market that will partly offset the cost
of solids handling and disposal. The plant currently employs BPT
treatment which generates gypsum solids of sufficient purity to be
sold. NSPS treatment generates iron oxide laden gypsum which is, at
403
-------
present, unsuitable for sale. The total solids produced in the model
plant are assumed to be 5,200 kg/kkg of Ti02 for BPT and 5,500 kg/kkg
of Ti02 for NSPS.
Treatment Costs
The estimated costs for three models having different production
levels are given in Tables 14-29, 14-30, and 14-31. The costs
represent the total cost required for treatment system implementation
and assume that the plant has no treatment system in place.
Table 14-32 presents a summary of the unit cost distribution between
amortization and operation and maintenance cost components at
different production levels and at the BPT and NSPS levels of
treatment.
For existing sources at the first level of treatment, the costs
reflect disposal of sludge on-site, hence land requirements are fairly
large. However, this cost can be offset by a revenue on the sale of
gypsum, currently practiced at one plant. These off-setting revenues
are not included in the costs presented in Tables 14-29, 14-30, 14-31,
and 14-32, These off-setting revenues are included in the Economic
Impact Analysis of_ Pollution Control Technologies for Segments of the
Inorganic Chemicals Manufacturing Industry, EPA 440/2-81-023.
Amortization, chemicals, labor, and residual waste disposal costs have
significant impact on the annual costs.
The total annual costs presented in Table 14-32 would increase by 20
to 40 percent if the sludge dewatering equipment is to be installed
for gypsum recovery.
Basis For Regulations
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 three plants practice partial neutralization and
settling. The BPT limitations are based on available long-term data
from Plant #559 and treatment performance information in Section 8.
A. Pollutant Removal with BPT Treatment
Treatment Level 1 is equivalent to the BPT in the Titanium
Dioxide (Sulfate Process) industry. Means, standard deviations,
and variability factors were calculated from data submitted by
Plant #559 for final effluent quality, and the results are given
in Tables 14-33a and 14-33b. The performance characteristics are
utilized for the development of the BPT regulations.
The ability of the treatment system to remove conventional and
toxic pollutants was estimated by comparing the treated effluent
4C4
-------
TABLE 14-29. MODEL PLANT TREATMENT COSTS
Titanium dioxide - Sulfate process
31,800 metric tons per year
Subcategory
Production
INVESTMENT COST
BPT
Site development 232,500
Equipment 958,600
Monitoring equipment .... 7, 800
Subtotal 1,198,900
Contractor's 0 & P 179,835
Subtotal 1,378,735
Engineering 275,800
Subtotal 1,654,535
Contingencies 275,800
Subtotal 1,930,335
Land 636,000
TOTAL INVESTMENT COST 2,566,335
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.... 252,000
Energy 34,000
Chemicals 1,760,000
Maintenance 193,032
Taxes and insurance 76,989
Residual waste disposal.. 141,000
Monitoring, analysis,
and reporting 15,000
TOTAL OPERATION AND
MAINTENANCE COST 2,472,021
C. AMORTIZATION OF
INVESTMENT COST 314,063
TOTAL ANNUAL COST 2,786,084
BAT'
0
0
0
0
0
0
0
0
0
0
0
NSPS
316,500
2,551,000
20,000
2,887,500
433,125
3,320,625
664,125
3,984,750
398,475
4,383,225
1,372,000
5,755,225
504,000
96,000
1,589,000
438,323
172,657
210,000
15,000
0 3,024,979
0 713,151
0 3,738,130
Represents the incremental cost above that for BPT treatment
Overhead and Profit
405
-------
TABLE 14-30. MODEL PLANT TREATMENT COSTS
Subcategory Titanium dioxide - Sulfate process
Production 47,700 metric tons per year
A. INVESTMENT COST BPT BATa NSPS
Site development 316,200 0 439,000
Equipment 1,213,200 0 3,288,500
Monitoring equipment . 7,800 0 20,000
Subtotal 1,537,200 0 3,747,500
Contractor's 0 & P 230,580 0 562,125
Subtotal 1,767,780 0 4,309,625
Engineering 353,560 0 861,925
Subtotal 2,121,340 0 5,171,550
Contingencies 353,560 0 517,155
Subtotal 2,474,900 0 5,688,705
Land 960,000 0 1,920,000
TOTAL INVESTMENT COST 3,434,900 0 7,608,705
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.... 336,000 0 672,000
Energy 51,000 0 138,000
Chemicals 2,640,000 0 2,384,000
Maintenance 247,492 0 568,871
Taxes and insurance 103,047 0 228,261
Residual waste disposal.. 209,000 0 315,000
Monitoring, analysis,
and reporting 15,000 0 15,000
TOTAL OPERATION AND
MAINTENANCE COST 3,601,539 0 4,321,132
C. AMORTIZATION OF
INVESTMENT COST 402,669 0 925,552
TOTAL ANNUAL COST 4,004,208 0 5,246,684
Represents the incremental cost above that for BPT treatment
Overhead and Profit
406
-------
TABLE 14-31. MODEL PLANT TREATMENT COSTS
Subcategory Titanium dioxide - Sulfate process
Production 74,500 metric tons per year
A. INVESTMENT COST B?T BATa NSPS
Site development 418,300 0 631,000
Equipment 1,622,700 0 4,352,000
Monitoring equipment 7,800 0 20,000
Subtotal 2,048,800 0 5,003,000
Contractor's O & P ...... 307,320 0 750,450
Subtotal 2,356,120 0 5,753,450
Engineering 471,200 0 1,150,690
Subtotal 2,827,320 0 6,904,140
Contingencies 471,200 0 690,414
Subtotal 3,298,520 0 7,594,554
Land 1,470,000 0 2,940,000
TOTAL INVESTMENT COST 4,768,520 0 10,534,554
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.... 336,000 0 672,000
Energy 71,000 0 199,000
Chemicals 4,100,000 0 3,719,000
Maintenance 329,854 0 759,455
Taxes and insurance 143,056 0 316,037
Residual waste disposal.. 283,000 0 420,000
Monitoring, analysis,
and reporting 15,000 0 15,000
TOTAL OPERATION AND
MAINTENANCE COST 5,277,910 0 6,100,492
C. AMORTIZATION OF
INVESTMENT COST 536,672 0 1,235,634
TOTAL ANNUAL COST 5,814,582 0 7,336,126
a Represents the incremental cost above that for BPT treatment
i_
Overhead and Profit
407
-------
TABLE 14-32 MODEL PLANT UNIT TREATMENT COSTS
Subcategory Titanium dioxide - Sulfate process
Annual Treatment Costs (S/kkg)
COST ITEM
PRODUCTION
(kkg/yr)
LEVEL OF TREATMENT
BPT BAT* NSPS
Annual Operation
and Maintenance
Annual
Amortization
Total Annual
Cost
31,800
47,700
74,500
31,800
47,700
74,500
31,800
47,700
74,500
77,74
75.50
70.84
9,88
8.44
7.20
37.61
83.95
78.05
NA
MA
NA
NA
NA
NA
NA
NA
NA
95.13
90.59
81.89
22.43
19.40
16.59
117.55
109.99
98.47
*Represents the incremental cost above BPT
408
-------
qualities with the raw waste qualities of the sampled waste
streams.
Basis for BPT Effluent Limitations
A. Technology Basis
For BPT, the Agency is basing limitations on equalization,
limestone neutralization and clarification, followed by pH
adjustment before final polish and discharge of the effluent.
This technology is chosen because it has been installed and
operated successfully by a plant in the industry.
/
B. Flow Basis
Waste flow data is available for three plants and the average
value of 475 mVkkg of Ti02 (Table 14-26) is taken as the inflow
for the model plant treatment system. The treatment plant
effluent flow is taken to be the same as the influent flow and
the loss or addition of water through chemicals, evaporation,
precipitation, and through solids removal have been neglected.
C. Selection of Pollutants to be Regulated
Selection of pollutants for limitation was based on a number of
criteria including raw waste pollutant concentrations during
sampling, total subcategory raw waste pollutant loading, and
experience related to the control of pollutant removal by
alkaline treatment technology presented in Section 8. The Agency
decided not to regulate the nonconventional pollutant iron under
BPT and BAT because of increased cost of the treatment when iron
is controlled and because the gypsum solids produced by the
treatment with iron removal can be reused only if dissolved iron
is not controlled. One plant has developed a market for reuse of
the gypsum. The Agency estimates that requiring iron removal at
existing plants would increase treatment costs by up to 40
percent and generate large quantities of waste solids for
disposal. Based on our economic impact analysis (63), the Agency
believes that two of the four existing Ti02 sulfate process
plants would close, with the attendant increase in unemployment,
if the plants were required to control iron. Control of toxic
metal pollutants wil.l be adequate at existing plants even without
an iron removal step.
Review of Tables 14-27, 14-28, and 14-34 indicates that six
pollutants were consistently identified at treatable
concentrations. Total suspended solids, iron, chromium, lead,
nickel, and zinc were all identified at treatable levels on the
average during sampling.
The toxic pollutants can be divided into two groups exhibiting
similar behavior depending on the treatment pH. Review of
alkaline treatment in Section 8 shows that chromium, lead, and
4G9
-------
TABLE 14-33a. HISTORICAL EFFLUENT M3NTIORING DATA SUMMARY
SUBCATEGQRY - TITANIUM DIOXIDE
SULFATE PROCESS PLANT 1559 - TREATMENT WITHOUT
IRON ^1'
(2)
Parameter v '
Daily Data
No. of Points
Average, x
Standard Deviation, S
Standard Deviation,
S1 (of logs)
Variability Factor
Monthly Averages
No. of Points
Standard Deviation, S_
Variability Factor
Variability Factor
Ratio
Pollutant (mg/1)
TSS Iron Ghromiun Nickel
30(3) 84
50 330
59 73
0.95 Q.20
5.8 1.5
20 20
44 42
1.6 1.2
3.6 1.3
81
0.23
.0.20
0.75
4.4
20
0.11
1.9
2.3
82
0.14
0.12
0.74
4.3
20
0.099
2.1
2.0
Zinc
54
0.50
0.60
0.95
5.8
13
0.21
1.7
3,4
(1)
(2)
(3)
Lead and cadmiun concentrations vrere less than 0.10 rag/1 for the 20
months.
Refer to Table 14-33b for explanation of statistical parameters.
Based on one month of data for April 1981.
410
-------
TABLE 14-33b.HISTORICAL EFFLUENT r-UNlTORINC DATTl
SJUZATLODRY. - TITANIUM DIOXIDE
SULTATE PROCESS P1/MT 1559 - TTCA"T!UJT KITH HOI
Pollutant
TSS Cadnium Chrcnuum Iron Lead Nickel
DaiJy Data
No. of Points 899 109 128 854 128 128
Average, x 21.0 0.060 0.070 0.62 0.06S 0.08
Deviation S(1J 65-93 0.044 0.054 3.46 0.041 0.071
Euindard 1.54 0.68 0.67 1.86 0.56 0.76
Deviation, S' '
Factor^ ^ 11'° 3"85 3'B1 13-65 3'16 4'39
Month lv Averages
Ito. of Points 30 26 30 28 30 3D
Standard 21.84 0.042 0.038 0.94 0.04 0.048
Variability 2U 4 39
Factor
Variabil ity
Factor Ratio
VFRE5) 3.62 1.58 1.87 3.38 !..« 1.00
Zinc
128
0.151
0.204
1.02
6.41
30
0.16
3.05
2.10
(1) S is the ajrit>metic stairiaj-d deviation and is given, by
S =
i=n , -, 2
(x: - x)
n-1
xi is the data value for point i
x is the mean value
n is the number of data, points
(2) S* is the estimated standard deviation of tl* logarithms
S'
S is tlie arithmetic standard deviation
jT is the mean value
(3) Tie variability factor [VF) of daily measurements for kgnonnal distribution
is found by the expression
In (VF) = S' ( Z - 0.5 S1)
where S' is defined in Note 2 above.
2= 2.33 for 99th percentile
(4) The variability factor (VF) for 30-day average measurements is found by the
expression
VF = 1.0 + Z
x is the mean value of the monthly averages
m
S is the arithmetic stan^oi'd deviation of the monthly averages
in
2 =1.64 for 95th r«rctntile
(5J VFR: Ratio of the 24-hour variability factor to the 30-day variability factor
411
-------
TABLE 14-34. VERIFICATION RESULTS FRCM - SULFATE PROCESS
TITANIUM DIOXIDE PLANT #559
Waste Stream
Pollutant Raw Waste Treated Ef f luent
Unit Load Concentration Unit Load Concentration Perceval
Efficiency
-------
zinc can be removed together at pH 8.0-9.5, while nickel requires
a higher pH. Therefore, chromium and nickel were selected based
on their relative significance in the raw waste and due to their
representative nature regarding overall toxic metal removal.
Lead, zinc, cadmium, copper, arsenic, and antimony are of lower
significance on the basis of sampling results. No toxic organic
pollutants were identified at significant concentrations.
D. Basis of Pollutant Limitations
1. 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 the proposed Development Document
(60) and the JRB Study (52).
b. TSS and Iron
The long-term average values of 50 mg/1 for TSS and 330 mg/1
for iron derived from the monitoring data of Plant 1559
(Table 14-33a) was used as the subcategory performance
values. The variability factors for daily measurements and
monthly averages estimated from Plant #559 long-term data
(Table 14-33a) were used in calculating the concentration
basis and effluent limitations as described below.
Total Suspended Solids
The TSS maximum 30-day average concentration is:
(50 mg/1)(7.6) = 80 mg/1
The 24-hour maximum TSS is:
(50 mg/1)(5.8) = 290 mg/1
The TSS maximum 30-day average effluent limit was
obtained by using the model plant unit flow of 475
m3/kkg as follows:
(80 mg/1) (475 mVkkg) ( kq/m^ ) = 38 kg/kkg
(1000 mg/1)
The 24-hour maximum TSS effluent limit is:
(290 mg/1) (475 mVkkg) ( kg/m3 ) = 140 kg/kkg
(1000 mg/1)
413
-------
Iron
The maximum 30-day average iron concentration is
determined similarly from Table 14-33a as follows:
(330 mg/l)(l.2} = 400 mg/1
The 24-hour maximum iron concentration is:
(330 mg/l)(l.5) = 500 mg/1
The concentrations are presented in Table 14-35 as guidance.
Toxic Pollutants
The effluent limitations for chromium and nickel were established
on the basis of long-term monitoring data for Plant #559 which
are presented in Table 14-33a.
a. Chromium
The chromium limitation is based primarily on long-term
monitoring data available for Plant 1559. A treatment
system raw waste influent concentration for chromium of 3.1
mg/1 was observed at Plant #559 during sampling. The
long-term data summarized in Table 14-33a indicates a
current treatment performance of 0.23 mg/1 and a monthly
average and daily measurement variability factor of 1.9 and
4.4, respectively.
The chromium maximum 30-day average concentration is given
by:
(0.23 mg/1)(1.9) =0.44 mg/1
The chromium 24-hour maximum concentration is given by:
(0.23 mg/1}{4.4) =1.0 mg/1
The chromium maximum 30-day average effluent limit is
developed using the model plant flow of 475 m3/kkg and
concentration limit above as follows:
(0.44 mg/1) (475 mVkkg ( kg/in3 ) = 0.21 kg/kkg
(1000 mg/1)
The chromium 24-hour maximum effluent limit is given
by:
(1.0 mg/1) (457 mVkkg) ( kg/m3 ) =0.48 kg/kkg
(1000 mg/1)
-------
TABLE 14-35. EFFLUENT LIMITATIONS
TITANIUM DIOXIDE SULFATE PROCESS
Best Practicable Control Technology Currently Available
Waste Water Flew: 475 m3/kkg of TiO2
Concentration Basis Effluent Limit
(rag/1) (kg/kkg of Ti02)
Subcategory
Performance
Pollutant (mg/1)
Mav
/T \ £*JdA*
VFRU' 30-day
Avg.
Max.
24-hr 30-day 24-hr
Max. Avg. Max
Conventional and Nonconventional Pollutants
Total Suspended
Solids
Iron
Toxic Pollutants
Chromium
Lead
Nickel
Zinc
Cadmium
Copper
Antimony
Arsenic
50(2>
330 (2)
0.23 t2>
0.15<3)
0.14<2>
0.50(2)
0.10(3)
0.40 (3'
0.80(5)
0.50(5)
5.8 80
1.6
1.5 400
1.2
4.4 0.44
1.9
3.2 0.32
2.1
4.3 0.29
5.8 0.85
1.7
3.9 0.24
2.4
3.8 0.80
2.0
3.8 1.6
2.0
3.8 1.0
2.0
290 38 140
500 -<4' -(4)
1.0 0.21 0.48
0.48 -<4> _«>
0.60 0.14 0.29
(4) (4)
2.9 — w —^'
f4} (4)
0.39 — w — w
1.5 — —
3.0 — (4) — (4)
(4) (4)
1.9 — V4' — v*'
(1) VFR: Ratio of the variability factor of the daily measurements to the
variability factor of the monthly averages.
(2) Long-term average based on loading data and variability factors of
Plant #559 selected frcm Table 14-33a.
(3) Limitation based on long-term data at Plant #559 and review of Table 8-12,
(4) No effluent limitation.
(5) The lower limit of the literature treatability estimate (Table 8-11) is
used as the basis for the long-term average limitation.
-------
b. Nickel
The long-term average concentration for nickel at Plant #559
is 0.14 mg/1 from Table 14-33a for the treated effluent.
The average nickel concentration observed during sampling
was 0.18 mg/1 for Plant #555 and #559 (Table 14-28). The
long-term average and variability factors observed for Plant
#559 were used to develop the limitations as follows:
The maximum 30-day concentration for nickel is:
(0.14 mg/1) (2.1) = 0.29 mg/1
The 24-hour maximum concentration is:
(0.14 mg/1) (4.3) = 0.60 mg/1
The nickel maximum 30-day average effluent limit is:
(0.29 mg/1) (475 m3/kkg) ( kq/m3 ) = 0.14 kg/kkg
(1000 mg/1)
The nickel 24-hour maximum effluent limit is:
(0.60 mg/1 (475 mVkkg) ( kg/in^ ) =0.29 kg/kkg
(1000 mg/1)
c. Other Metals
Lead, zinc, cadmium, copper, arsenic, and antimony were also
present in the raw wastewater. A concentration level is
developed for these toxic pollutants which can be used to
develop a limitation in the . event that they become a
concern. Development of the concentration level is
primarily made on a similar basis as established previously
for the primary pollutants. Information used to establish
the concentration level for each pollutant is summarized as
follows:
416
-------
Summary
Pollutant
Zinc
Cadmium
Copper
Arsenic
Antimony
Lead
(1)
(2)
Long-term
Average
0.50 mg/l(5)
0.10 mg/1(1,2)
0.40 mg/l(3)
0.50 mg/l(4)
0.80 mg/1(4)
0.15 mg/l(l,2)
Variabilityd )
Factor of the
Monthly Average
1 /7
2.4
2.0
2.0
2.0
2.1
Variabilityd )
Factor of the
Daily
Measurements
5.8
3.9
3.8
3.8
3.8
3.2
Table 14-33b
Table 8-12
(3) Table 8-13
(4) Table 8-11
(5) Table 14-33a
The concentration levels are developed for all the pollutants as
follows:
The maximum 30-day = (Long-term average
concentration concentration)
(Monthly average
variability factor)
The 24-hour maximum = (Long-term average X (Daily measurement
concentration) variability factor)
Summaries of the conventional and toxic pollutant limitations for BPT
are presented in Table 14-35.
Basis for BCT Effluent Limitations
Wh i 1 e EPA has not yet proposed or pr omu 1 gated a r evi sed BCT
methodology in response to the American Paper Institute v. EPA
decision mentioned earlier, EPA is promulgating BCT limitations for
this subcategory. These limits are identical to those for BPT because
the only technology option that removes significant amounts of
conventional pollutants is not economically achievable. See the
discussion of iron under BPT above. Removal of significant additional
amounts of conventional pollutants can be achieved in this subcategory
only if iron is also removed.
As BPT is the minimal level of control required by law, no possible
application of the BCT cost tests could result in BCT limitations
lower than those promulgated today. Accordingly, there is no need to
wait until EPA revises the BCT methodology before promulgating BCT
limitations.
417
-------
TABLE 14-36. EFFLUENT LIMITATIONS
TITANIUM DIOXIDE SULFATE PROCESS
Best Available Technology
Wastewater Flow: 475 m3A)cg of TiO2
Concentration Effluent Limit
Subcategory( "" '
Pollutant Performance
Nonconventional
Pollutants :
Iron 330
Toxic Pollutants:
Chromium 0.23
Lead 0.15
Nickel 0.14
Zinc 0.50
Cadmium 0.10
Copper 0.40
Antimony 0.8Q(4>
Arsenic 0.50(4)
VFR<2>
1.5/1.2
4.4/1.9
3.2/2.1
4.3/2.1
5.8/1.7
3.9/2.4
3.8/2.0
3.8/2.0
3.8/2.0
Basis
Max.
30 -day
Avg.
400
0.44
0.32
0.29
0.85
0.24
0.80
1.6
1.0
(mg/1) (kg/kkg of TiO;>)
Max. ~
24-hr . 30-day 24-hr .
Max. Avg. Max.
500 —(3) —(3)
1.0 0.21 0.48
0.48 — <3) — (3)
0.60 0.14 0.29
2Q \ J ) __\-3j
• 3
0.39 —0) -0)
1-5 -0) — <3>
o » U ™*^ — ^
L9 —(3) __(3)
(1) Limitations for BPT Table 14-35.
(2) VFR: Ratio of the variability factor of the daily measurements to the
variability factor of the monthly averages.
(3) No effluent limitation.
(4) The lower limit of the literature treatability estimate (Table 8-11) is used
as the basis for the long-term average limitation.
418
-------
Basis for BAT Effluent Limitations
For BAT, the Agency is basing limitations on treatment consisting of
Level 1 technology, and is the same as BPT. Treatments requiring
aeration for iron removal (Level 2) and 55 percent recycle through use
of soda ash precipitation (Level 3) were considered but rejected
because of costs. The limitations for BAT are given in Table 14-36.
Basis for New Source Performance Standards
For NSPS, the Agency is basing limitations on equalization, limestone
neutralization, clarification, aeration, alkaline precipitation and
settling, followed by pH adjustment before final polish in a lagoon
and discharge of the effluent (Level 2). This technology is chosen
because it has been installed and operated successfully by Plant #559.
A. Flow Basis
Waste flow data is available for three plants and the average
value of 475 mVkkg of Ti02 (Table 14-26) is taken as the
influent to the model treatment system.
B. Selection of Pollutants to be Regulated
The same pollutants considered for BPT are also considered for
the NSPS regulations. These pollutants include total suspended
solids (TSS), iron, chromium, lead, nickel, and zinc which were
identified at treatable levels on the average during sampling.
Iron was selected in addition to TSS, chromium, and nickel for
regulation on the basis of additional removal of large quantities
of iron and additional TSS and toxic metal removal beyond the
BPT/BAT level with the application of aeration and alkaline
precipitation to BPT treatment.
An iron removal step added to BPT/BAT technology is used for
NSPS/PSNS because the additional treatment provides better
removal of toxic metals, and the more stringent standards are
unlikely to pose a significant barrier to entry. New sulfate
process plants are unlikely because the two alternate processes
are-more economical. However, if a company did want to construct
a new sulfate process plant, a process change involving the
recycle and reuse of the strong acid wastewater would likely be
adopted. This process change would reduce production costs, and
would also reduce the amount of strong acid wastewater treated
and discharged by 70 to 90 percent. Reducing the flow of the
strong acid wastewater reduces treatment costs substantially and
also substantially reduces the amount of gypsum solids produced
by treatment. With the smaller amount of gypsum solids produced,
disposal of the solids as waste is competitive with sale of the
solids for reuse, when cost of sales is considered. At least one
company using the sulfate process is actively developing the"
recycle/reuse technology. One other company using the chloride
419
-------
TABLE 14-37. EFFLUENT LIMITATIONS
TITANIUM DIOXIDE SULFATE PROCESS
New Source Performance Standards
Wastewater Flow: 475 m3/kkg of TiO2
Pollutant
Subcategory
Performance
(mg/1)
Concentration
Basis (mg/1)
Max.
30-day 24-hr.
Avg. Max.
Effluent Limit
(kgAkg of TjQ7)
Max.
30-day 24-hr.
Avg. Max.
Conventional and
Nonconventional
Pollutants
Total Suspended
Solids
Iron
(6)
Toxic Pollutants
0.62
<2>
11.0/3.0 64
13.7/4.0 2.5
230
8.5
30
1.2
110
4.1
Chromium ' ° '
Lead
Nickel (6)
Zinc
Cadmium
Copper
Antimony
Arsenic
0.15(3)
0.15(3)
0.10<3>
0.20<3)
0.10*3)
0.40<3>
0.80<5>
0.50<5>
3.8/2.0
3.2/2. 1
3.8/2.0
6.4/3.1
3.9/2.4
3.8/2.0
3.8/2.0
3.8/2.0
0.30
0.32
0.20
0.62
0.24
0.80
1.6
1.0
0.57
0.48
0.38
1.3
0.39
1.5
3.0
1.9
0.14
— (4)
0.095
— (4)
— (4)
__(4)
— (4)
— (4)
0.27
— (4)
0.18
— (4)
— (4)
— (4)
— (4)
— (4)
(1) VFR: Ratio of the variability factor of the daily measurements to the
variability factor of the monthly averages.
(2) Long-term average based on loading data and variability factors of Plant #559
selected from Table 14-33b.
(3) Limitation based on long-term data at plant #559 and review of Table 8-12.
(4) No effluent limitation.
(5) The lower limit of the literature treatability estimate {Table 8-11) is used
as the basis for the long-term average limitation.
(6) Applicable to PSNS regulations.
420
-------
ilmenite process has recently built a new plant that discharges
50 percent less total wastewater than comparable plants.
C. Basis of Pollutant Limitations
1. 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 the proposed
Development Document (60) 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 1559 (Table 14-33b) were used as the subcategory
.performance values. The variability factors for daily
measurements and monthly averages estimated from Plant
#559 long-term data (Table 14-33b) were used in
calculating the concentration basis and effluent
limitations as determined below
Total Suspended Solids
The TSS maximum 30-day average concentration is:
(21 mg/1) (3.04) - 64 mg/1
The 24-hour maximum TSS is:
(21 mg/1) (11.0) = 230 mg/1
The TSS maximum 30-day average effluent limit was
obtained by using the model plant unit flow of 475
m3/kkg as follows:
(64 mg/1) (475 mVkkg ( kq/m3 ) = 30 kg/kkg
(1000 mg/1)
The 24-hour maximum TSS efffluent limit is:
(230 mg/1) (475 mVkkg) ( kq/m3 ) = 110 kg/kkg
(1000 mg/1)
Iron
The maximum 30-day average iron concentration is
determined similarly from Table 14-33b as follows:
421
-------
(0.62 mg/1) (4.0) = 2.5 mg/1
The 24-hour maximum iron concentration is:
(0.62 mg/1) (13.7) = 8.5 mg/1
The maximum 30-day average iron effluent limit is:
(2.5 mg/1) (475 mVkkg) ( kg/m^ ) =1.2 kg/kkg
(1000 mg/1)
The 24 hour maximum iron effluent limit is:
(8.5 mg/1) (475 mVkkg) ( kq/m3 ) =4.1 kg/kkg
(1000 mg/1)
The conventional and nonconventional pollutants
limitations are presented in Table 14-37 for treatment
with iron removal.
2. Toxic Pollutants
The effluent limitations for chromium with iron removal is
based primarily on long-term monitoring data for Plant #559
and review of treatment performance information in Tables
8--12 and 8-13 for other industries utilizing the same
treatment technology for removal of the pollutant.
a. Chromium
The chromium limitation with iron removal is based
primarily on long-term monitoring data available for
Plant #559. A treatment system raw waste influent
concentration for chromium of 3.2 mg/1 was observed at
Plant #559 during sampling. The long-term data
summarized in Table 14-33b indicates a current
treatment performance of 0.070 mg/1 and a monthly
average and daily measurement variability factor of 2.0
and 3.8, respectively.
Review of long-term average treatment data for chromium
in Tables 8-12 and 8-13 indicates that 0.070 mg/1 shows
excellent removal when compared to the same treatment
technology for chromium removal in other industries. A
long-term average of 0.15 mg/1 for chromium is used as
the basis for the final limitation based on observed
performance data from other industries summarized on
the tables in Section 8. The final limitations are
developed below for chromium.
The chromium maximum 30-day average concentration is
given by:
422
-------
Basis for Pretreatment Standards
A. Existing Sources
The only wastewater discharged to the POTW from the one existing
indirect discharger is a small portion of the weak acid stream.
Based on data provided in response to Section 308 request, it is
unlikely that this wastewater contains any toxic metals at
treatable levels. Therefore, -we are excluding this subcategory
from categorical PSES under the provisions of Paragraph 8(b) of
the Settlement Agreement.
B.. New Sources
Tor pretreatment standards for New Sources (PSNS), the Agency is
basing limitations on NSPS. The pollutants to be regulated are
iron, chromium,, and nickel as indicated in Table 14-37.
Pretreatment is required because NSPS provides better removal of
iron and toxic metals than does a well-operated POTW with
secondary treatment installed and therefore chromium, nickel, and
iron would pass through the POTW in the absence of pretreatment
Titanium Dioxi.de - Chloride - Ilmenite Process Industry Profile
General Description
Total subcategory production capacity is given in Table ,14-38 Profile
Data Summary. Tiie 30S .data available for the TiOz Subcategory does
not adequately cover the one-step chloride-ilmenite process; however,
supplementary information iteas been -submitted by industry (55). The
status of regulations prior to promulgation of this regulation is
presented -in Table T4-2. Additional information on the
chloride-ilmenite process industry is given in this section.
General Process Description and Raw Materials
For the manufacture of titanium dioxide by the combined ore
beneficiation-chloride process, a generalized process flow diagram
including the waste streams is shown in Figure 14—11.
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 document as the chloride-ilmenite process. Processes which
involve a separate ore beneficiation step (either at the plant or at
the ore source) resulting in an upgraded or a synthetic rutile product
to be used as feed material for a chloride process would not be
classified as a chloride-ilmenite process. A separate ore
beneficiation process would fall within the Ore Mining and Dressing
Category for regulatory purposes, and the manufacture of Ti02 from an
upgraded ilmenite or synthetic rutile would be in the same
classification as a chloride process using natural rutile ore.
427
-------
TABLE 14-39. AVERAGE VftTER USAGE FOR TiO? PROTECTION
BY THE CHDORIDE -HMENTTE PROCESS
Use Plant #237 Plant #550
, 3 - .
Cm /kkg of TiO2>
Noncontact 73-140 330-390
Cooling
Process Contact 100-140 (1) 47- 59
and Cleanup
Noncontact 9- 11 6-7
Ancillary Uses
(Boilers,
Sanitary, etc.)
Plant #713
15-16
60 (2)
5- 6
Source of data, (55).
(1) Tha average total flow of 120 m /kfcg is used as the basis for 8PI.
3
(2) The average flow of 60m /kkg is used as the basis for NSPS.
428
-------
TABLE 14-41. SUM4ARY OF RAW HASTE LOADINGS FOUND IN SCREENING AND VERIFICATION SAMPLING
W
SUBCATEOORY TITAN I LW DIOXIDE - SULFATE PROCESS (Applied to Chloride Ilirenite Process)
Pollutant Loading Range
(kg/day)
Mininun Maxinun
Priority
Antimony 5.0 28
Arsenic 1.9 4.0
Cadmiun 0.068 7.2
Chromium 140 530
Copper 8.2 19
Lead 3.0 65
Nickel 3.7 23
Selenium 7.0 9.5
Thalliun 0,47 1.2
Zinc 1.8 85
Conventional
TSS
Iron
Unit Loading
Minimum Average Maximum
0.032 0.11 0.22
0.012 0.19 0.032
0.00044 0.019 0.057
1.1 2.0 3.4
0.065 0.085 0.12
0.024 0.18 0.42
0.029 0.080 0.15
0.0020 0.031 0.06
0.0030 0.0055 0.0080
0.014 0.34 0.55
320
600
No. of
Plants (1)
3
3
3
3
3
3
3
2
2
3
1
1
(1) - Data are taken only from those plants where pollutants were found above detection limits, or
in the case of TSS and Iron, where data are available.
-------
TABLE 14-42. TOXIC POLLUTANT AVERAGE RAW HASTE LOADS AND OCNCENTRftTICHS
U>
•U
SUaCATEGORY
TITANIUM DIOXIDE - Sulfate Process
(Applied to Chloride Ilmenite
Screening
Plant 1555
-------
Daily raw waste loads were calculated from the flow rates measured or
estimated at the time of sampling and the measured pollutant
concentrations. That is,
Daily loading (as kg of pollutant per day) = (C) (Q)
1000
where the concentration (C) of the pollutant is expressed in units of
mg/1, (Note: 1 kg/m3 = 1000 mg/1 ) , and the flow rate (Q) is
expressed in units of mVday (m3, a cubic meter, is equal to 264 U.S.
gallons) .
Similarly, the unit loading were calculated from the reported
production rate (P), the waste stream flow rate (Q), and the measured
pollutant concentration (C).
Unit loading (as kg of pollutant = (C) (Q)
per 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 Ibs).
A summary of daily and unit per unit of production raw waste loads for
the plants sampled is presented in Table 14-41 and the individual
plant averages are given in Table 14-42.
The estimated total annual raw wastewater load of toxic pollutants
generated by the Chloride-Ilmenite Process is given below.
Pollutant
Chromium
Nickel
Zinc
Lead
Copper
Cadmium
Antimony
Thallium
Arsenic
Selenium
Total Annual Raw Wastewater Load
(kg/year)
1 ,050,000
42,000
178,000
94,000
44,000
9,900
58,000
2,900
99,000
16,000
Section 5 of this report described the scope and methodology of the
sampling program and this section 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.
435
-------
Pollution Abatement Option
Toxic Pollutants of Concern
Rationale for selection of the toxic pollutants of concern are
presented in this section 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 14-22. In the sulfate process the
unwanted iron remains largely in the ferrrous state and may be
crystallized out of the acid waste streams and sold as copperas
(ferrous sulfate}. In the chloride-ilmenite process, the same ore
impurity is oxidized to some extent to the ferric state during the
chlorination step. This appears in the acid waste streams as ferric
chloride (FeClj) in the amounts indicated in Table 14-41.
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.
Process Modifications and Technology Transfer Options
The comments made in regard to the Titanium Dioxide-Chloride Process
for rutile and upgraded ores in this section are generally applicable
to the Chloride-ilmenite Process.
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.
Prevailing 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, 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 is.
436
-------
therefore, included in the raw waste influent to the model plant
wastewater treatment system.
In practice, one plant disposes of the entire metal chloride, HC1, and
TSS waste by ocean dumping. The remainder of the plants dispose of
the concentrated waste by deep well injection after use of surface
lagoons for removal of settleable solids.
The dilute process waste streams are segregated to the extent possible
from noncontact sources and treated in conventional in-plant systems
utilizing equalization and spill diversion facilities followed by lime
neutralization/coagulation, solid separation in a settling pond, and
final discharge of the treated effluent. Chemical coagulating agents
such as ferric chloride and alum may be used either before or after pH
control as an aid in the removal of metal hydroxides and other
suspended solids.
Advanced Treatment Technology
Advanced treatment technology options for the 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:
A. Aeration for a) decarbonization if limestone is used for
neutralization, and b) ferrite coprecipitation, assuming that
sufficient ferrous iron is already present or is added to the
system as needed (the latter may^also be accomplished by adding
scrap iron to the aqid wastes).
B. An alkaline precipitation step under optimum conditions for metal
hydroxide precipitation, i.e., pH 9-10.
C. Dual-media filtration for additional removal of suspended solids
including toxic metal hydroxides.
D. Sulfide precipitation for additional toxic metal removal followed
by filtration.
E. Other metal removal technologies including xanthate
precipitation, ion exchange, and membrane applications, all of
which were regarded as inappropriate from a practical and
economic point of view,
Selection of Appropriate Technology and Equipment
Technologies for Different Treatment Levels
A. Level 1 (BPT)
Figure 14-12 shows the model treatment system for treatment with
and without iron removal. Calcium carbonate (limestone) is used
to neutralize the concentrated acid waste stream. Aeration is
4-27
-------
00
Figure 14-12.
Level 1 waste water treatment for titanium dioxide - chloride
(ilmenite ore) process.
-------
Figure 14-13.
Level 2 waste water treatment for titanium dioxide - chloride
(ilmenite ore) process.
-------
then used to primarily remove C02 prior to pH adjustment to
reduce demand for treatment chemicals, and mixing with dilute and
miscellaneous plant wastes. The combined stream is given lime
treatment to pH 9-10 for iron and toxic pollutant removal or 6-9
during treatment without iron and toxic metajl removal. All
wastewater passes through a clarifier and final' polishing lagoon
prior to discharge.
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 can remove greater than 95
percent of the major pollutants of concern including toxic metals
according to preliminary treatability estimates. No iron removal
is required for BPT.
B. Level 2 (NSPS)
Level 2 treatment adds iron removal as described in Level 1 and
dual-media filtration to the Level 1 technology for additional
removal of iron, suspended solids, and toxic metal hydroxides
following the alkaline precipitation and settling steps. The
flow diagram for Level 2 is shown in Figure 14-13. This level of
treatment was selected as the basis for NSPS because it provides
a relatively economical method for removing additional toxic
metals.
Equipment for Different Treatment Levels
A. Equipment Functions
Unlike treatment of the wastewaters from the Ti02 Sulfate
Process, limestone neutralization of the Chloride-Ilmenite
Process wastewaters does not generate large quantities of solids
(e.g., gypsum) which require mechanized separation and transfer
to sizable on-site or off-site disposal areas. The solids that
are generated from TSS and metal precipitate separation can be
collected in moderate-sized lagoons and periodically transferred
to appropriate chemical landfill disposal sites in accordance
with the Resource Conservation and Recovery Act (RCRA) (as
amended, 42 USC 6901, et seq.). The Level 1 treatment model
includes rail car deliveries of ground limestone and lime, 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.
B. Chemicals and Chemical Handling
First stage neutralization utilizes ground limestone while lime
is used for second stage neutralization and final alkaline
440
-------
precipitation. Oxygen is supplied as air to oxidize any ferrous
iron remaining in addition to COZ removal, and treatment
chemicals may be added as required for removal of precipitated
metals and other suspended solids. Aside from the large scale
bulk chemical handling requirements for limestone and lime, there
are no particular hazards involved.
C. Disposal of Solids
Periodic removal of solids from settling impoundments will
require compliace with RCRA regulations as applicable to on-site
or off-site chemical disposal site operation. However, this
subcategory has not been listed under RCRA-ISS for hazardous
pollutants.
Treatment Cost Estimates
General Discussion
To prepare treatment cost estimates, a model plant concept was
developed. Cost estimates have been prepared for Level 1 (BPT)
treatment, and for Level 2 (NSPS). The model plant specifications
given below were utilized for cost estimating and for development of
the regulations.
A. 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 kg/year, 70,000 kkg/year, 113,7570
kkg/year, and 157,500 kkg/year.
B. Wastewater Flows
Wastewater is typically segregated into two streams; strong
acidic wastewater flow from beneficiation-chlorination of
ilmenite ore and air emission scrubbing facilities, and the other
wastewater from process reactions, washings, product transport,
cooling tower blowdown, water treatment blowdown, and other
operations. For the model plants, a unit flow of 6 mVkkg of
product for the concentrated acidic wastewater and 114 mVkkg of
product for the dilute wastes is used. The treatment system is
designed to handle a total flow of 120 mVkkg of product (Table
14-39).
For the NSPS model plant, a unit flow of 6 mVkkg of product for
the concentrated acidic wastewater is used. Because of improved
design which allows for recycle systems and more efficient
process water utilization, dilute wastewater is considerably
reduced (62). The total combined wastewater flow of 60 mVkkg of
-------
product is used (Table 14-39). The treatment system for Level 2
(NSPS) is BPT plus dual-media filtration.
C. Pollutant Load
The principal pollutants occurring in the wastewaters are TSS,
iron, chromium, nickel, zinc, and hydrochloric acid. For the
model plants, the following unit pollutant loads have been
considered:
TSS 175 kg/kkg of Ti02
HC1 230 kg/kkg of TiOz
Iron 375 kg/kkg of Ti02
Chromium 1.4 kg/kkg of TiO2
Nickel 0.1 kg/kkg of Ti02
Zinc 0.5 kg/kkg of Ti02
The loading values for TSS, HC1, and iron are based on data
submitted by industry on the chloride-ilmenite process. The
chromium loading is an estimated average derived from a wide
range of ilmenite ore qualities and the nickel and zinc loadings
are taken from the screening and verification data on the Ti02
sulfate process.
D. Chemical Usage
In the model BPT system, powdered limestone is used for first
stage neutralization of strong acidic waste flow at the unit rate
of 302 kg/kkg of Ti02. Pebble lime (CaO) is used for second
stage neutralization of the total* combined flows. Lime is used
at the unit rate of 42 kg/kkg of Ti02.
E. 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 890 kg/kkg of Ti02. Assuming an average
production rate and 50 percent moisture content, the volume of
sludge generated is 1.8 mVkkg.
Model Plant Control and Treatment costs
The estimated costs for four models having different production levels
are given in Tables 14-43, 14-44, 14-45, an<3 14-46.
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.
442
-------
TABLE 14-43. MODEL PLANT TREATMENT COSTS
Subcategory Titanium dioxide - Chloride process (Ilmenite ore)
Production 35,000 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal
Contractor's 0 & Pb....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
($)
BPT
292,200
851,500
20,000
1,163,700
174,555
1,338,255
267,651
1,605,906
160,591
1,766,497
252,000
2,018,497
TOTAL OPERATION AND
MAINTENANCE COST
336,000
31,000
260,000
176,650
60,555
105,000
15,000
984,205
BAT
0
0
0
0
0
0
0
0
0
0
0
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
287,409
1,271,614
aRepresents the incremental cost above that for BPT treatment
Overhead and Profit
443
-------
TABLE 14-44. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Titanium dioxide - Chloride process (Ilmenite ore)
70,000 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal .........
Contractor's O & Pb....
Subtotal
Eng ineer ing
Subtotal
Contingencies
Subtotal
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
BPT
520,000
1,023,500
20,000
1,563,500
234,525
1,798,025
359,605
2,157,630
215,763
2,373,393
492,000
2,865,393
TOTAL OPERATION AND
MAINTENANCE COST
504,000
43,000
510,000
237,339
85,962
105,000
15,000
1,500,301
BAT
0
0
0
0
0
0
0
0
0
0
0
C. AMORTIZATION OP
INVESTMENT COST
TOTAL ANNUAL COST
386,151
1,886,452
Represents the incremental cost above that for BPT treatment
Overhead and Profit
444
-------
TABLE 14-45. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Titanium dioxide - Chloride process (Ilmenite ore)
113,750 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal
Contractor's 0 & P ....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
BPT
784,000
1,377,000
20,000
2,181,000
327,150
2,508,150
501,630
3,009,780
300,978
3,310,758
780,000
4,090,758
TOTAL OPERATION AND
MAINTENANCE COST
588,000
62,000
823,000
331,076
122,723
210,000
15,000
2,151,799
BAT
0
0
0
0
0
0
0
0
0
0
0
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
538,660
2,690,459
Represents the incremental cost above that for BPT treatment
Overhead and Profit
445
-------
TABLE 14-46. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Titanium dioxide - Chloride process (Ilmenite ore)
157,500 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment
Subtotal ....v*........
Contractor's 0 & P .
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
BPT
1,053,600
1,590,000
20,000
2,663,600
399,540
3,063,140
612,628
3,675,768
367,577
4,043,345
1,080,000
5,123,345
588,000
71,000
1,141,000
404,334
153,700
210,000
15,000
2,583,035
($)
BAT
0
0
0
0
0
0
0
0
0
0
0
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
657,852
3,240,887
Represents the incremental cost above that for BPT treatment
Overhead and Profit
446
-------
TABLE 14-47. MODEL PLANT TREATMENT COSTS
Subcategory Titanium dioxide - Chloride (Ilmenite ore)
Production 35,000 metric tons per year
_
A. INVESTMENT COST NSPS
Site development 280,000
Equipment 1,046,500
Monitoring equipment 20,000
Subtotal 1,346,500
Contractor's 0 & P a. 201,975
Subtotal 1,548,475
Engineering 309,695
Subtotal 1,858,170
Contingencies 185,817
Subtotal 2,043,987
Land 252,000
TOTAL INVESTMENT COST 2,295,987
B. OPERATION AND
MAINTENANCE COST
Labor and supervision 350,000
Energy 34,200
Chemicals 260,000
Maintenance 204,399
Taxes and insurance 68,880
Residual waste disposal .... 105,000
Monitoring, analysis
and reporting 15,000
TOTAL OPERATION AND
MAINTENANCE COST 1,037,478
C. AMORTIZATION OF
INVESTMENT COST 332,557
TOTAL ANNUAL COST 1,370,035
Overhead and Profit
447
-------
TABLE 14-48. MODEL PLANT TREATMENT COSTS
Subcategory Titanium dioxide - Chloride (Ilmenite ore)
Production 70,000 metric tons per year
($)
A. INVESTMENT COST NSPS
Site development 496,500
Equipment 1,323,500
Monitoring equipment 20,000
Subtotal 1,840,000
Contractors 0 & P a. 276,000
Subtotal 2,116,000
Engineering 423,200
Subtotal 2,539,200
Contingencies 253,920
Subtotal 2,793,120
Land 492,000
TOTAL INVESTMENT COST 3,285,120
B. OPERATION AND
MAINTENANCE COST
Labor and supervision 518,000
Energy 49,500
Chemicals 510,000
Maintenance 279,312
Taxes and insurance 98,554
Residual waste disposal .... 105,000
Monitoring, analysis
and reporting 15,000
TOTAL OPERATION AND
MAINTENANCE COST 1,575,366
C. AMORTIZATION OF
INVESTMENT COST 454,441
TOTAL ANNUAL COST 2,029,806
Overhead and Profit
448
-------
TABLE 14-49. MODEL PLANT TREATMENT COSTS
Subcategory Titanium dioxide - Chloride (Ilmenite ore)
Production 113,750 metric tons per year
—
A. INVESTMENT COST NSPS
Site development 752,000
Equipment 1,822,000
Monitoring equipment 20,000
Subtotal 2,594,000
Contractor's 0 & Pa 389,100
Subtotal 2,983,100
Engineering 59*5,620
Subtotal 3,579,720
Contingencies 357,972
Subtotal 3,937,692
Land 780,000
TOTAL INVESTMENT COST 4,717,692
B. OPERATION AND
MAINTENANCE COST
Labor and supervision 602,000
Energy 71,700
Chemicals 823,000
Maintenance 393,769
Taxes and insurance 141,531
Residual waste disposal .... 210,000
Monitoring, analysis
and reporting 15,000
TOTAL OPERATION AND
MAINTENANCE COST 2,257,000
C. AMORTIZATION OF
INVESTMENT COST 640,662
TOTAL ANNUAL COST 2,897,662
Overhead and Profit
449
-------
TABLE 14-50. MODEL PLANT TREATMENT COSTS
Subcategory Titanium dioxide - Chloride (Ilmenite ore)
Production 157,500 metric tons per year
($)
A. INVESTMENT COST NSPS
Site development 1,015,000
Equipment 2,175,000
Monitoring equipment 20,000
Subtotal 3,210,000
Contractor's 0 & Pa 481,500
Subtotal 3,691,500
Engineering 738,300
Subtotal 4,429,800
Contingencies 442,980
Subtotal 4,872,780
Land 1,080,000
TOTAL INVESTMENT COST 5,952,780
B. OPERATION AND
MAINTENANCE COST
Labor and supervision 602,000
Energy 85,000
Chemicals 1,141,000
Maintenance 487,278
Taxes and insurance 178,583
Residual waste disposal .... 210,000
Monitoring, analysis
and reporting 15,000
TOTAL OPERATION AND
MAINTENANCE COST 2,718,861
C. AMORTIZATION OF
INVESTMENT COST 792,801
TOTAL ANNUAL COST 3,511,663
Overhead and Profit
450
-------
TABLE 14-51. MODEL PLANT UNIT TREATMENT COSTS
Subcategory Titanium dioxide - Chloride (Ilmenite ore)
Annual Treatment Costs ($/kkg)
COST ITEM
PRODUCTION
(kkg/yO
LEVEL OF TREATMENT
BPT BAT* NSPS
Annual Operation
and Maintenance
Annual
Amortization
Total Annual
35,000
70,000
113,750
157,500
35,000
70,000
113,750
157,500
28.12
21.43
18.92
16.40
8.21
5.52
4.74
4.18
NA
NA
NA
NA
NA
NA
NA
NA
29.64
22.51
19.84
17.26
9.50
6.49
5.63
5.03
Cost
35,000
70,000
113,750
157,500
36.33
26.95
23.65
20.58
NA
NA
NA
NA
39.14
29.00
25.47
22.30
*Represents the incremental cost above BPT
451
-------
The unit waste flow of 6 m3/kkg of product for the concentrated acidic
wastewater 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 wastewater flow for the NSPS system is much less than for
the BPT model; however, the reduced flow has negligible impact on
costs because the unit waste loads are the same. Table 14-47, 14-48,
14-49, and 14-50 present the estimated NSPS treatment costs for four
models having different production levels. 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.
Table 14-51 presents a summary of the unit cost distribution between
amortization and operation and maintenance cost components at
different productions at the BPT and NSPS level of treatment.
Basis for Regulations
Evaluation of BPT Treatment Practices
The prevailing control and treatment practices in the TiOz
Chloride-Ilmenite industry have been reviewed in this section. 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.
Basis for BPT Effluent Limitation
A. Technology Basis
The BPT limitations are based on technology 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 above.
B. Flow Basis
The BPT model plant flow rate is based on the reported average
process contact and clean up wastewater flow at Plant #237 of 120
mVkkg as indicated in Table 14-39.
C. Selection of Pollutants to be Regulated
Specific information concerning the characteristics of the
combined strong and dilute acid waste streams in the
Chloride-Ilmenite process is not available. In view of the lack
of such information which would indicate the concentration of
pollutants in this industry, it is necessary to use data from the
Sulfate Process as a guide in the selection of pollutants to be
limited. Considering the many similarities in the chloride and
sulfate processes, including use of the same ore with known
impurities, treatment technology, and generation of strong and
452
-------
dilute acid waste streams, it is extremely likely that TSS,
chromium, and nickel are present at treatable concentrations, and
are, therefore, limited for this industry. It is also likely
that iron, lead, zinc, cadmium, copper, antimony, and arsenic may
also be present on occasion at treatable levels and are,
therefore, considered as potential candidates for limitation.
Details on the selection of these pollutants for limitation are
presented in this section for the Titanium Dioxide-Sulfate
Process.
1. Conventional and Nonconventional Parameters
a. TSS and Iron
The removal characteristics of total suspended solids
(TSS) is influenced by the removal of iron. For BAT,
the long-term average concentration of 50 mg/1 for TSS
and 330 mg/1 for iron were achieved (Table 14-33a) and
were used as the performance values. The variability
factors for daily measurements and monthly averages
estimated from Plant #559 (Table 4-33a) were used in
determining the concentration basis and effluent
limitations for TSS as follows:
The maximum monthly average TSS concentration is:
(50 mg/1) (1.6) = 80 mg/1
The 24-hour maximum TSS concentration is:
(50 mg/1) (5.8) = 290 mg/1
The maximum monthly average limitation is:
(80 mg/1) (120 mVkkg) ( kq/m^ ) =9.6 kg/kkg
(1000 mg/1)
The 24-hour maximum limitation is:
(290 mg/1) (120 mVkkg) ( kq/rn^ ) = 35 kg/kkg
(1000 mg/1)
For iron, the concentration basis is determined from
data in Table 14-33a as follows:
The maximum 30-day average concentration is:
(330 mg/1) (1.2) = 400 mg/1
The maximum 24-hour concentration is:
(330 mg/1) (1.5) = 500 mg/1
453
-------
TABLE 14-52. EFFLUENT LIMITATIONS
TITANIUM DIOXIDE CHLORIDE PROCESS USING ILMENITE
Best Practicable Control Technology Currently Available
Wastewater Flow: 120 m3/kkg
Pollutant
Estimated
Treatability
Concentration
Basis (mg/1)
Max.
30-day 24-hr.
Avg. Max.
Effluent Limit
(kq/kkq)
Max.
30-day 24-hr.
Avg. Max.
Conventional/
Nonconvent iona 1
Pollutants
Total Suspended
Solids SO*2)
Iron 330*2)
Toxic Pollutants
Chromium*5^ 0.23
Lead 0.15
Nickel*5* 0.14
Zinc 0.50
Cadmium 0.10
Copper 0.40
Antimony 0.80
Arsenic 0.50
5.8/1.6 80 290 9.6
1.5/1.2 400 500 —(4)
4.4/1.9 0.44 1.0 0.053
3.2/2.1 0.32 0.48 — *4)
4.3/2.1 0.29 0.60 0.035
5.8/1.7 0.85 2.9 —(4)
3.9/2.4 0.24 0.39 —(4)
3.8/2.0 0.80 1.5 —(4)
3.8/2.0 1.6 3.0 —(4)
3.8/2.0 1.0 1.9 —(4)
35
— (4)
0.12
..(4)
0.072
— (4)
__(4)
— (4)
— (4)
..(4,
(1) VFR: The ratio of the variability factor for daily measurements to the
variability factor for monthly averages.
(2) Long-term average from plant #559 monitoring data (Table 14-33a)
(3) Long-term average and variability factors are based on the same rationale as
the TiO2 Sulfate Process Table 14-35.
(4) No effluent limitation.
(5_) Applicable to BAT limitations.
454
-------
The concentration basis and appropriate limitations
are summarized in Table 14-52.
2. Toxic Pollutants
The toxic pollutants that are limited include chromium and
nickel. Lead, zinc, cadmium, copper, antimony, and arsenic
are considered as potential candidates for limitation.
Details concerning the selection and treatment performance
of these pollutants for the purpose of setting numerical
limitations is identical to discussions in this section for
the Titanium Dioxide-Sulfate Process. Only those factors
considered specific to the Chloride-Ilmenite Process for the
purpose of setting the limitations are considered. The BPT
effluent limitations are presented in Table 14-52.
a. Chromium
Chromium limitations for BPT treatment were based on
Table 14-35. The maximum 30-day average concentration
of 0.44 mg/1 and daily maximum concentration of 1.0
mg/1 were used to determine the limitations as follows:
The maximum 30-day average limitation is:
(0.44 mg/1) (120 mVkkg ( kq/m3 ) = 0.053 kg/kkg
(1000 mg/1)
The 24-hour maximum limitation is:
(1.0 mg/1) (120 mVkkg} ( kq/m^ ) = 0.12 kg/kkg
(1000 mg/1)
The limitations are presented in Table 14-52 for BPT
treatment.
b. Nickel
Nickel limitations for BPT treatment were based on
Table 14-35. The maximum 30-day average concentration
of 0.29 mg/1 and daily maximum concentration of 0.60
mg/1 were used to determine the limitations as follows:
The maximum 30-day average limitation is:
(0.29 mg/1) (120 mVkkg) ( kq/m3 ) = 0.033 kg/kkg
(1000 mg/1)
The maximum 24-hour limitations is:
(0.60 mg/1) (120 mVkkg) ( kq/m3 ) = 0.72 kg/kkg
(1000 mg/1)
455
-------
The limitations are presented in Table 14-52 for BPT
treatment
c. Other Metals
Potential candidates for limitation include lead, zinc,
cadmium, copper, antimony, and arsenic which are
included in the limitations on a concentration basis in
the event they are identified as a concern. The
concentration basis values are presented in Table 14-52
for BPT and are identical to values developed in Table
14-35 for the Titanium Dioxide -Sulfate Process.
Basis for BCT Effluent Limitations
While EPA has not yet proposed or promulgated a revised BCT
methodology in response to the American Paper Institute v. EPA
decision mentioned earlier, EPA is promulgating BCT limitations for
this subcategory. These limits are identical to those for BPT because
the only technology option that removes significant amounts of
conventional pollutants is not economically achievable. See the
discussion of iron under BPT in the Sulfate Process above. Removal of
significant additional amounts of conventional pollutants can be
achieved in this subcategory only if iron is also removed.
As BPT is the minimal level of control required by law, no possible
application of the BCT cost tests could result in BCT limitations
lower than those promulgated today. Accordingly, there is no need to
wait until EPA revises the BCT methodology before promulgating BCT
limitations.
Basis for BAT Effluent Limitations
For BAT, the Agency is basing limitations on the application of Level
1 technology which is- equivalent to BPT. The model plant flow basis
of 120 m3/kkg used for BPT is also used for BAT. The BAT limitations
are presented in Table 14-52. A more advanced technology using soda
ash precipitation and recycle of wastewater was considered for the
similar sulfate process but was rejected on the basis of cost and
because its performance has not been demonstrated.
Basis for the New Source Performance Standards
A. Technology Basis
For NSPS the Agency is basing limitations on the application of
Level 2 treatment technology which adds dual-media filtration and
iron removal steps to the BPT system for greater efficiency in
the removal of toxic metals, iron, and suspended solids. This
treatment greatly increases the coprecipitation of toxic metals
and prevents large quantities of dissolved and suspended iron
from entering receiving water bodies and interfering with other
water uses.
456
-------
B. Flow Basis
The reported data on process contact and clean-up wastewater flow
at Plant #173 is selected as the basis of a model plant for new
sources. Process modifications resulting in a greatly increased
efficiency of water use reduce the average flow rate to 60 mVkkg
as shown in Table 14-39.
C. Basis for Pollutant Limitations
1. 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 the proposed
Development Document (60) and the JRB Study (52).
b. TSS
The concentration basis for the TSS maximum 30-day
average limitation is obtained by applying an average
filtration efficiency of 38 percent removal (41,60) to
the corresponding concentration of 64 mg/1 with iron
removal (Table 14-37). That is:
(1.00-0.38) (64 mg/1) = 40 mg/1
The maximum 30-day average limitation is obtained by
applying the NSPS model plant flow rate of 60 mVkkg:
(40 mg/1) (60 mVkkg) ( kq/rn^ ) =2.4 kg/kkg
(1000 mg/1)
The TSS daily maximum limitation is determined as
follows:
(1.00-0.38) (230 mg/1) = 140 mg/1
The daily maximum limitation is:
(140 mg/1) (60 mVkkg) ( kq/m^ ) =8.4 kg/kkg
(1000 mg/1)
The NSPS limitations are presented in Table 14-53.
2. Nonconventional Pollutants
The only nonconventional pollutant of concern is iron. For
NSPS, the Agency is basing a maximum 30-day average
limitation on an average filtration efficiency of 38 percent
removal (41,60). Thus the appropriate concentration basis
457
-------
TABLE 14-53. EFFLUENT LIMITATIONS
Titanium Dioxide - Chloride Process Using Ilmenite
New Source Performance Standards*
Waste Water Flow: 60 m3/kkg
Pollutant
Estimated
Treatability
(mg/D \i
Concentration Basis
(mcr/1)
Max.
m 30-day 24^ir.
rFRu' Avg. Max,
Effluent Limit
Ocq/kka)
Max.
30-day 24-hr.
Avg. flax.
Conventional and
Ndnconventional~
Pollutants
Total suspended 13
Solids
Iron
0.38
11.0/3.0 40 140
13.7/4.0 1.6 5.3
2.4
8.4
0.096 0.32
Toxic Pollutants
Chromiurrr J
I«d<3>
Nickel (3)
Zinc (3)
Cadmium1 '
Copper1 J
Antimony
Arsenic
0.060
0.030
0.086
0.19
0.075
0.23
0.80
0.50
3.8/2.0
3.2/2.1
3.8/2.0
6.4/3.1
3.9/2.4
3.8/2.0
3.8/2.0
3.8/2.0
0.12
0.064
0.17
0.58
0.18
0.46
1.6
1.0
0.23
0.096
0.33
1.2
0.29
0.87
3.0
1.9
0.0072 0.014
_(4) _(4)
0.010 0.020
(4) (4)
(4) f4)
(4) (41
J4) _(4)
_{4) _(4)
(1) VFR: Ratio of the variability factor for the daily measurements to
the variability factor of the monthly averages.
(2) Based on the application of pollutant - specific removal efficiencies
for dual-media filtration to adjust the BPT performance on treatability
estimates shown in Table 14-52.
* Including pretreatment standards for new sources (PSNS) covering iron
and toxic metals which are expressed as concentrations.
(3) Applicable to PSNS limitations.
(4) No effluent limitations.
458
-------
is derived from the corresponding concentration basis of 2.5
mg/1 of iron used for the maximum 30-day average (Table
14-37). That is:
(1.00-0.38) (2.5 mg/1) =1.6 mg/1
The limitation for NSPS is:
(1.6 mg/1) (60 mVkkg) ( kq/m3 ) = 0.096 kg/kkg
(1000 mg/1)
Similarly the 24-hour maximum iron concentration is:
(1.00-0.38) (8.5 mg/1) =5.3 mg/1,
and the limitation is:
(5.3 mg/1) (60 mVkkg) ( kq/m3 ) - 0.32 kg/kkg
(1000 mg/1)
Toxic Pollutants
Addition of iron removal and a dual-media filtration system
as a polishing step in NSPS is expected to remove additional
toxic pollutants. This removal is estimated based on
literature treatability (41) and the treatability study
(61). The following removal efficiencies for the toxic
pollutants are used for the limitations after careful
consideration of available information:
Toxic Pollutant Percent Removed Percent Remaining
Chromium
Lead ( 1 )
Nickel
ZincO )
Cadmium( 1 )
Copper ( 1 )
Antimony( 1
Arsenic( 1 )
60
80
14
6
25
42
0
0
40
20
86
94
75
58
100
100
(1)No limitation given; concentration basis presented in
Table 14-53.
a. Chromium
The long-term average concentration is determined for
chromium by application of a 60 percent removal to 0.15
mg/1 in Table 14-37 as follows:
(0.15 mg/1) (1.00-0.60) = 0.060 mg/1
459
-------
The maximum 30-day concentration for chromium is:
(0.060 mg/1) (2.0) = 0.12 mg/1,
and the 24-hour maximum concentration is:
(0.060 mg/1) (3.8) = 0.23 mg/1
Therefore, the maximum 30-day average concentration and
24-hour maximum limitation are determined,
respectively:
(0.12 mg/1) (60 m3/kkg) ( kq/m 3 ) = 0.0072 kg/kkg
(1000 mg/1)
(0~. 23 mg/1) (60 m'/kkg) ( kq/m3 ) = 0.014 kg/kkg
(1000 mg/1)
b. Nickel
The long-term average concentration is determined for
nickel by application of a 14 percent removal to 0.10
mg/1 in Table 14-37 as follows:
(0.10 mg/1) (1.0-0.14) = 0.086 mg/1
The maximum 30-day concentration for nickel is:
(0.086 mg/1) (2.0) = 0.17 mg/1,
and the 24-hour maximum concentration is:
(0.086 mg/1) (3.8) = 0.33 mg/1
Therefore, the maximum 30-day average concentration and
24-hour maximum limitation are determined,
respectively:
(0.17 mg/1) (60 m3/kkg) ( kq/m3 ) = 0.010 kg/kkg
(1000 mg/1)
CO. 33 mg/1) (60 mVkkg) ( kq/rn^ ) = 0.020 kg/kkg
(1000 mg/1)
Other Metals
The concentration basis for lead, zinc, cadmium,
copper, antimony, and arsenic are determined in a
similar manner to the primary pollutants. The
following equation is used in conjunction with the
concentration information in Table 74-37 and the
percent removal presented in this section:
-460
-------
BPT long-term x Percent remaining = NSPS long-term
average 100 average
concentration concentration
The maximum 30-day average concentration and 24-hour
maximum concentration are determined by using the
appropriate variability factors in Table 14-37 and
multiplying by the NSPS long-term average. For
example, the long-term average for zinc is:
(0.20 mg/1) (0.94) = 0.19 mg/1
The maximum 30-day average concentration is:
(0.19 mg/1) (3.1) = 0.59 mg/1
The 24-hour maximum concentration is:
(0.19 mg/1) (6.4) = 1.2 mg/1
Similarly, the concentration bases are determined for
the remaining toxic pollutants and are presented in
Table 14-53.
Basis for Pretreatment Standards
Pretreatment is necessary because it provides better removal of
pollutants than is achievable by a well operated POTW with secondary
treatment installed, and thereby prevents pass-through that would
occur in a POTW in the absence of pretreatment.
Existing Sources
Since there are no indirect dischargers in this subcategory, the
Agency is excluding this subcategory from categorical PSES under the
provisions of paragraph 8(b) of the Settlement Agreement.
New Sources
The Agency is promulgating PSNS that are equal to NSPS because these
standards provide better removal of iron, chromium and nickel than is
achieved by a well-operated POTW with secondary treatment installed
and therefore iron, chromium and nickel would pass through a POTW in
the absence of pretreatment. Pollutants regulated under PSNS are
iron, chromium, and nickel. See Table 14-53.
460-A
-------
-------
SECTION 15
ALUMINUM FLUORIDE INDUSTRY
Industry Profile
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 prior to promulgation of this
new regulation is given in Table 15-2.
General Process Description and Raw Materials
In the dry process for the manufacture of aluminum fluoride, partially
dehyrdated alumina hydrate is reacted with hydrofluoric acid gas. The
reaction is given as:
Al£03 + 6HF = 2A1F3 + 3H20
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.
Water Use and Waste Source Characteristics
Water Use
Water is used in noncontact cooling of the product, for seals on
vacuum pumps and for scrubbing the reacted gases before 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.
Waste Sources
A. Noncontact Cooling Water
Noncontact cooling water is used to cool the product coming out
of the reactor. In some cases it is recirculated and the
blowdown treated separately from other process contact wastewater
or it is discharged without treatment. The water can be
monitored for fluoride and if process contamination occurs, it
461
-------
TRBI£ 15-1.
SOBCHBSQHJf
DMA SttMHBT
ALUKQKM PUXMOEE
Total subcategory capacity rate
Total subcategory production rat*
Nunber of plants in this subcategory
308 Data on £11* for
With total capacity of
Hitfa total production of
Representing capacity
Representing prortrrflnn
Plant production ranges
Average production
capacity
Plant age range:
Wasts water flow range:
Minium
Ttaiume per unit product;
f*i TI 1 n« ••
HA
134,700 kkg/year
5*
6
204,800 Idog/year
120,000 kkg/year
NA
NA
1,000 kkg/year
45,600 kkg/year
24,300 kkg/year
35,500 kkg/year
59 percent
5 years
21 years
539 cubic meters/day
2,200 cubic meters/day
5 cubic meters/kkg
12 cubic meters/kkg
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, D.SJV., 1977, U.S. Department of Connerce, CLu-muL Industrial
Reports, Deceini w 1977; Energy and Eimiiuuuental Analysis, Inc.; Draft
Report, "Preliminary Goancmic AaaeMi**"1: of Effluent T.
-------
HATER
VENT
_
SCRUBBER
HASTE HATER
CYCLONE
HYDRATED
ALUMINA
HYDROGEN
FLUORIDE
REACTOR
NONCONTACT
COOLING WATER
J L
COOLER
PRODUCT
COLLECTION
AND STORAGE
ALUMINUM
'FLUORIDE
PRODUCT
Figure 15-1,
General process flow diagram for production
of aluminum fluoride.
-------
TAELE 15-2 ._ STfiTOS OF REGULATIONS - EFFLUENT IjMTIMlON GUlCOJKgS
SUBCATEGOR* Aluminum Fluoride
SUEPAKT W (40 CFR 415.230, 5/22/75)
STA.KDARDS
BPCTCA* BATEA* NSPS *
Max. Avg. Max. Avg. Max. Avg.
Product Fara- Xg/l&g
Process rretexs (mg/D
A1F- Fluoride 0.68 0.34
(40)3 (20)
TSS O.S6 0.43
(51) (25)
0.34 0.17
(20) (10)
Sections 415.230, 415.231, and 415.232 were revoked ty the Agency
.1*1 FR 51601, November 23, 1376),
"Vax. = Miximin of any one day,
Avg. - !--^:'Jjttm average of daily -values for thirty consecutive days.
basis 17,000
15-3. VJ?^£R USSGE IN THE AUMTNW FLUORIDE SUBCATEGQRY
Source Water use per unit of production
(m /Wtg of AIFjJ
Plant
i 837
Plant
I 705 U)
Plant
i 188
Plant
1 251 (2)
Tto-contact cooling 14.5 NA (1) 6.95 HA
Indirect process 12.2 wJl'lS" f' NA NA
contact (pumps, sea Is, J*
leaks, spills)
Maintenance, e.g. 1.13 2139 NA V'l!.02
cleaning and work area
washdown
Scrutber 3.45 V:92 3.46 18j7
(1) KR. = Not Available
(2) Currently not rr-anufocturing alvminum fluoride.
464
-------
can be diverted to the wastewater treatment facility for fluoride
removal.
B. Floor and Equipment Washings
The quantity and quality of wastewater generated from these
operations varies and depend largely on the housekeeping
practices at the individual plants.
C. Scrubber Wastewater
This is the major source of wastewater requiring treatment before
discharge or recycle to the scrubber. It is contaminated with
hydrofluoric acid, aluminum fluoride and aluminum oxide, and, in
some cases, suIfuric acid and silicontetrafluoride have been
detected. These originate as impurities in the hydrofluoric acid
used in the process. Table 15-4 presents the wastewater flows at
different facilities in the subcategory. Noncontact cooling
water is excluded from consideration since it normally does not
contain pollutants.
D. Solid Wastes
In the aluminum fluoride production, hydrofluoric acid 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 wastewater 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 a 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 wastewater in the aluminum
fluoride subcategory. If the production of aluminum fluoride is
integrated with hydrofluoric acid, then the wastewaters from both
plants are combined and treated.
Description of Plants Visited and Sampled
Screening
Plant #705 was visited in the screening phase of the program. Both
hydrofluoric acid and aluminun fluoride are produced at this facility
by the general processes described earlier. The wastewater from the
hydrofluoric acid and aluminum fluoride plants is mixed and sent to
the treatment facility. At the treatment facility the combined
wastewater 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.
465
-------
TABLE 15-4. WASTE WRTER FLOW AT PLANTS #837, #705 AND #251
FOR ALUMINUM FLUORIDE SUBCATEJGORy
Source
Flow rate per unit of production
( ra3/kkg of A1F3)
(1)
Scrubber water
Maintenance equipment
cleaning and work area
washdown
Total raw waste flew
Plant #837
3.45
1.13
4.58
Plant #705(4) Plant #251(4)
&•&
w^
19.7
Average of above
three flows
11.9
(1) All flow information is from 308 Questionnaires and plant visits. Unit
flew is calculated ty dividing waste water flow in mVday by production
in kkg/day.
(2) From Table 15-6 (see footnotes which describe basis of information).
(3) From Table 15-7 (see footnotes which describe basis of information).
(4) Currently not manufacturing aluminum fluoride.
TABLE 15-5. SOLIDS GENERATED AT PLANT #705 AND #251 PRODUCING
AUKENUM FLUORIDE
Plant
Total Solids Generated(kg/fckg of AlF3)
#705
#251
(1)
(1)
54
69
(1) Currently not manufacturing aluminum fluoride.
466
-------
Figure 15-2 shows a simplified block diagram of the process including
the wastewater 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.
Verification
Plant #705 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 wastewater 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.
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.
Maximum Raw Waste Concentrations Observed
Pollutant Screening Verification
_ Plant#7Q5 _ Plant #705 and #251
Arsenic 200 480
Selenium 68 97
Chromium 70 1100
Copper 1 20 250
Lead 25 91
Mercury 1.6 11
Nickel 150 290
Zinc 450 450
Cadmium 0.70 33
Antimony 0 3.0
Beryllium 0.80 0.80
Section 5 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 #251 . 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
467
-------
HMO*
CO
,
FIUORSPAR
[»»
h»
H2S04 «—
I
— — •• ' -1 ' — i vtwr
1 f
| 1
WVVQ4
DRIP ACID 'P ^ PflOOUCT
HP
i ' " "
TO MHOSniERB 1 1
4 ^wmn f
_ * — Allan,
4" i
A1P3 WIDOUCT-^ : OOOUS
i i
. ^^ ,i SLWKB DRAIK9
[ ^^ " COCCItC TOWH* BUWOHH, BIC.
' MBffMUWnai HEPHJW HW -» fiETlUIB HID ^fc "**"* i*1 £^-^> TUF
*"* TANK ^ BBraaw. KW -• M.TIWI*. i«« » MJuaneff ^J| »* JJJ
Huta atroana wnupled.
Figure 15- 2. General process flew diagram at Plant 1705 showing the sanpling points
(aluminum fluoride manufacture).
-------
TABLE 15-6. FLOW AND POIlilT^^CONCEETORATiai DATA OF THE SAMPLED
WASTE STREAMS FOR PLAN^I^O^PODUCI^ ALUMINUM FLUORIDE
Sampling
Phase
Screening
Verifica-
tion Sanpling
Sampled
Stream
No.
3
4
-------
-J
O
VENT
1
DUST
COLLECTOR
WET
SPAR"
SPAR DRYING
HANDLING
LOSSES
HOSE DOWN
WATER
r
HF KILN
WASTE
AIR
DRIP
ACID
•WATER
SLURRY
TRANSFER
LEGEND
SAMPLING POINTS.
PLANT
REACTOR
T
AIF. PRODUCT
C
VENT
_L
S02 SCRUBBER
WATER
TUFj
SCRUBBER
LIQUEFACTION
AHF
PURIFICATION
JAH
PRC
AHF
PRODUCT
DILUTION
WATER
A1F3 PLANT
HOSE DOWN
HOSE DOWN WATER
AHF PLANT
#2
GYPSUM
POND
13
NEUTRALIZATION
SYSTEM
h
WATER
LA JL*6
Lt
EFFLUENT
TO RIVER
ALKALINE STREAMS
AND ACID FROM OTHER PLANTS
Figure 15-3. General process flow diagram at Plant #251 showing the sampling points.
(aluminum fluoride manufacture).
-------
TAE&E 15-7. FI£W AND POmZTAJJT QOtOMKATION DATA OP THE SAMPLED STREAMS
FOR PLANTff251 PRODUCING AUMCMJM FLUORIDE
Stream
No.
Sampled
Stream
Description
Unit
Flow
(m3Akg
of AlF3)
Verification
Sailing
4
6
4&6
2
3
AlF- scrubber
water
SO™ scrubber
water W
Total raw waste
load
Gypsum pond
influent*2'
Gypsum pond
effluent*25
12.6
6.10
18.7
25.1
25.1
•total
Suspended
Solids
(mg/1) (kg/kkg)
1200
0.0
1200
19,000
9.0
16
0.0
16
470
0.23
Fluoride
(mg/1) (kg/kkg)
470 5.90
20 0.14
320 6,0
660 17
320 8.0
Aluminum
(mg/1) (kg/kkg)
50 0.60
0.20 0.0010
50 0.60
26 0.65
22 0.55
#•>
(1) One half flow of SO2 scrubber water is assumed to contribute to the A1F3 process since the
total flow is connon to the A1F3 and HP process.
(2) Consists of hydrofluoric acid and aluminum fluoride waste water. Plant currently not
manufacturing AlF.,.
-------
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 polutant per day) = (C) (Q)
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)
Similarly, the unit loadings were calculated from the reported
aluminum fluoride production rate, the waste stream flow rate, and the
measured pollutant concentration.
Unit loading (as kg of pollutant
per kkg of aluminum fluoride) * (C) (Q)
1000P
where C and Q are the same as described above, and P is the
aluminum fluoride production rate expressed as kkg/day. (kkg is
1000 kg, a metric ton, which is equal to 2205 Ibs.)
The P and Q factors are for the Aluminum Fluoride Process and
thereby the Agency has segregated that portion of the effluent
attributable only to the Aluminum Fluoride Process.
Tables 15-8 and 15-9 are tabulations 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
sampling during screening. These unit loads were used to determine
the minimum, average, and maximum unit loading values 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:
472
-------
TABLE 15-8. TOXIC POLLUTMSTr AVERAGE RAW WASTE LOADS AND CONCENTRATIONS
SUBCATEGORY
ALUMINUM FLUORIDE
Pollutant
Arsenic
Selenium
Chromium
Copper
Lead
Mercury
Nickel
zinc
Cadmium
Antiitony
Beryllium
Screening
Plant #705
(11 (21
(mg/1) IJJ (kg/kkg) v"
0.18 0^0020
0.050 0.0010
0.030 0.00030
0.10 0.0010
0.0050 0.00010
0.00040 0.0000040
0.11 0.0010
0.16 0.0020
0,00020 0.0000020
0.00020 0.0000020
Verification
Plant
(mg/1)
0.18
— t3*'
0.44
0.070
0.020
0.00040
0.22
0.080
0,010
, 0.00040**
#705
(kg/kkg) "
0.0020
— <3>,
0.0050
0.0010
0.00020
0.0000050
0.0030
0.0010
0.00020
0.0000050
Plant
i '
(mg/1)
0.020
0.050
— <3>
0.010
0.010
0.0030
0.010
0.020
__<3)
__<3)
#251
(kg/kkg)
0.00030
0.0010
— (3)
0.00010
0.00010
0.000050
0.00020
O.OQD30
_(3)
•':",'
Average ^v^ •
Concentration i.oac(?
(mg/1) -V
0.13
0.050
0.24
0.060
0.012
0.0013
0.11
0.090
0.0050
0.00040
0.00020
(1) Concentrations based on average raw waste loads shown and total process production and waste
flows.
(2) kg/kkg of product.
(3) — belcw analytical detection limit.
-------
TABLE 15-9. TOXIC POLLUTANT EFFLUENT CONCENTRATIONS DURING SAMPLING
SUBCATEGORY
ALUMINUM FLUORIDE
Pollutant
Arsenic
Selenium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Cadmium
Antimony
Beryllium
Plant and Sampling Phase
#705
Screening
(mg/1)
ND(1)
ND
0.0070
0.10
0.0020
ND
0.050
0.0020
0.0020
ND
0.0020
#705
Verification
(mg/1)
ND
ND
0.040
0.0010
0.020
ND
ND
0.0010
0.0010
ND
ND
#251
Verification
(mg/D
0.0050
0.070
0.22
0.070
0.030
ND
0.050
ND
ND
ND
ND
Average
(mg/1)
< 0.0050
< 0.070
0.090
0.060
0.020
ND
0.050
0.0020
< 0.0020
ND
< 0.0020
(1) ND — Not Detected
474
-------
TABLE 15-10. SUMMARY OF RAW WASTE LOADINGS FOUND IN SCREENING AND VERIFICATION SAMPLING
SUBCATEGORY
ALUMINUM FLUORIDE
Pollutant
Toxic
Arsenic
Selenium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Cadmium
Antimony
Beryllium
Conventional and
Nonconventional
TSS
Fluorine
Aluminum
loading Range,
kg/day
Minimum Maximum
0.050
0.030
.0.020
0.020
0.0030
0.026
0.026
0.040
0.00010
NA<2>
NA
600
250
100
0.080
0.16
0.22
0.050
0.020
0.0080
0,12
0.080
0.0070
0.00020
0.00010
5400
980
320
Minimum
0.00030
0.0010
0.00030
0.00010
0.00010
0.0000040
0.00020
0.00030
0.0000020
NA
NA
13
5.5
0.60
Unit loading.
Average1 '
0,0013
0.0010
0.0030
0.00070
0.00015
0.000020
0.0013
0.0010
0.000080
o.oooooso'"
0.0000020
50
8.1
3.9
kg/kkg
Maximum
0.0020
0.0010
0.0050
0.0010
0.00020
0.000050
0.0030
0.0020
0.00020
NA
NA
119.0
13.0
7.0
No. of
Plants
Averaged
3
2
2
3
3
3
3
3
2
1
1
3
3
3
(1) Average of unit loadings from Table 15-8,
(2) Not Applicable
-------
Pollutant Waste Load (kg/year)
Arsenic
Selenium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Cadmium
Antimony
Beryllium
180
140
400
94
20
3.
180
140
11
0.
0.
0
70
30
Pollution Abatement Options
Toxic Pollutants of Concern
The toxic pollutants found in actual plant wastewaters 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 in some cases. The toxic pollutants 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.
Process Modification and Technology Transfer Options
A. Total recycle of wastewater to the scrubbers is feasible if final
neutralization is with soda ash. The calcium in the waste is
precipitated in the treatment system as calcium carbonate and
therefore scaling problems in pipes and scrubbers are reduced.
B. 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 wastewater treatment facility.
476
-------
Best Management Practices
A. Rainfall runoff in plant areas, treatment facilities and other
places susceptible to fluoride contamination can be collected and
sent to the wastewater treatment facility.
B. 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.
C. Settling ponds in the wastewater 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.
Prevailing Control and Treatment Practices
Plant #705 practices lime neutralization and settling of the
wastewaters. Since aluminum fluoride production is integrated with
hydrofluoric acid production, the wastewaters from the two processes
are combined before treatment. The plant does not treat noncontact
cooling water.
At Plant ftB37 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 is also
sent to an adjacent facility for use. The wastewaters from area
washdown are combined with other product wastewater, treated with
hydrated lime and sent to a settling lagoon before discharge.
Plant #188 produces aluminum fluoride in small quantities and in
batches. The wastewater 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 #251 mixes the aluminum fluoride waste with hydrofluoric acid
plant waste. The combined wastewater 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.
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
477
-------
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.
Selection of Appropriate Technology and Equipment
Technologies for Different Treatment Levels
A. Level 1 (BPT, BAT and NSPS)
Neutralization with lime is widely used in the industry to remove
the primary nonconventional pollutant as calcium fluoride.
Because lime neutralization to pH 10 results in significant
incidental removal of toxic pollutants, alkaline precipitation
was chosen as the Level 1 technology. The flow diagram is shown
in Figure 15-4.
B. Level 2
A higher removal of suspended metal hydroxides, TSS, and CaF2 can
be achieved by adding dual-media filtration to the Level 1
system. The flow diagram is shown in Figure 15-5.
C. Level 3
Sulfide precipitation is added to the Level 2 treatment to attain
a higher level of heavy metal removal. Chromium and selenium
levels are not appreciably reduced although other toxic
pollutants levels are reduced. The flow diagram is shown in
Figure 15-6.
D. 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.
A detailed cost comparison and performance evaluation of
alternative levels of treatment were utilized in the development
of the proposed regulations and were presented in the proposed
Development Document (60).
478
-------
/""^L
1 LDC
U-»Q>
'
RAH
QfH JID1
a
TN 1 L.^
WfiBTP VfWFF! """ x *
BQUAUZM1CN
^«
1
1
[
,t'1 to
— **- tai
^ t
I '
*
I ^
HOONS
r-A. MOOCN A»,
L J
^
.
-A "DOM JL*
L _
©
Tr
Vi
EEFUIBfr
•Includes flow mnltoring, pH nnitoring «nd «n|>ler.
Figure 15- 4. Level 1 waste water treatment for the aluminum fluoride subcategory.
-------
UOWflH
CD
O
^
IftSTE
1
warn
— -(g>
I
1
i J
/
-** A-**
BQUftLUKnCM 1
Lk^Vi— ^
._
— fl»
r*.
1
1
1
•jj
i!
J£
1
P^X, UOOdf A^l
1
• *
LI
UGOCH /-f»
i
Bi
#
1 ^!
•— Q*-*1
X
PXUIER
f
1
\HNxMnwwr
/w\
{&•}
IT
' H
*ZnoludM flow Monitoring, ftt ROnltorlng and Hn|)ler.
Figure 15-5. Level 2 waste water treatment for the aluminum fluoride subcategory.
-------
Qa
CO
KMf
WBTB—
mmt
^ 1 Z
L*o^
T JL
I w
" I
I
suv rtUBi
-4p~
*lnctude« now monitoring, pH mpnllortn| end
Figure 15-6. Level 3 waste water treatment for the aluminum fluoride subcategory.
-------
Equipment for Different Treatment Levels
A. 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.
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.
B. Chemicals and Handling
In Level 1 (BPT) 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 sulfat.e 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.
C. 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.
D. Monitoring Requirements
482
-------
00
U)
RNf
WSTE-
wren
r
HMUUnoi 1
(low monllotlng, pll monltorlnc and
~l
w
nun •
•mum
-Q '
Figure 15-7. Level 4 waste water treatment for the aluminum fluoride subcategory.
-------
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,
Treatment Cost Estimates
General Discussion
A model plant concept was developed for the subcategory for treatment
cost estimation purposes. The BPT treatment system specifications are
outlined below.
A. Wastewater Flow
The range of wastewater data on file shows flow variations from
4.58 mVkkg of A1F3 to 19.7 mVkkg of A1F3 (see Table 15-4).
Based on these values, a unit flow of 11.9 mVkkg of A.1F3 was
taken as the average for the wastewater treatment model plant for
cost estimating purposes.
B. Production
Five 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 wastewater
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 15,900 kkg/yr, 35,600 kg/yr and
45,800 kkg/yr.
C. Pollutant Loadings
Observed pollutant loadings varied from 14 to 27 kg/kkg of AlF3
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 (kq/kkq-AlF-.) F (kg/kkq-AlF, )
EPA-440/1-75-037 16-20 15-20
Screening and
Verification 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.
D. Treatment Chemicals
484
-------
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 stoichimetric
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.
E. Generation of Solids
From the pollutant loadings and treatment chemicals 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.
F. Cost Estimates
The estimated treatment costs (BPT) for three different
production models 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. A summary of the annual treatment costs is presented
in Table 15-14. At this level of treatment, chemicals, labor,
and amortization have significant impact on the annual costs.
Basis For Regulations
Evaluation of BPT Treatment Practices
EPA is setting BPT limitations based on Level 1 treatment. All plants
in this subcategory have installed BPT technology. Pollutants limited
by the 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 selected 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.
BPT Effluent Limitations
A. Technology Basis
The Agency is setting 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
485
-------
TABLE 15-11. MODEL PLANT TREATMENT COSTS
Subcategory Aluminum Fluoride
Production 15,900 metric tons per year
(S)
A, INVESTMENT COST BPT
Site development 30,300
Equipment 198, 000
Monitoring equipment 20,000
Subtotal 248/300
Contractor's 0 & P a. 37,245
Subtotal 285,545
Engineering 57,109
Subtotal 342,654
Contingencies 34, 265
Subtotal 376,919
Land 24,000
TOTAL INVESTMENT COST 400,919
B. OPERATION AND
MAINTENANCE COST
Labor and supervision 56,000
Energy 3,500
Chemicals 35,000
Maintenance 37,692
Taxes and insurance 12,028
Residual waste disposal .... 5,500
Monitoring, analysis
and reporting 15,000
TOTAL OPERATION AND
MAINTENANCE COST 164,720
C. AMORTIZATION OF
INVESTMENT COST 61,325
TOTAL ANNUAL COST 226,044
Overhead and Profit
486
-------
TABLE 15-12. MODEL PLANT TREATMENT COSTS
Subcategory Aluminum Fluoride
Production 35,500 metric tons per year
($)
A. INVESTMENT COST BPT
Site development 45,900
Equipment 245,000
Monitoring equipment 20,000
Subtotal 310,900
Contractor's 0 & Pa 46,635
Subtotal 357,535
Engineering 71,507
Subtotal 429,042
Contingencies *... 42,904
Subtotal 471,946
Land 42,000
TOTAL INVESTMENT COST 513,946
B. OPERATION AND
MAINTENANCE COST
Labor and supervision 56,000
Energy 5,500
Chemicals 80,000
Maintenance 47,195
Taxes and insurance 15,418
Residual waste disposal .... 12,500
Monitoring, analysis
and reporting 15,000
TOTAL OPERATION AND
MAINTENANCE COST 231,613
C. AMORTIZATION OF
INVESTMENT COST 76,786
TOTAL ANNUAL COST 308,399
Overhead and Profit
487
-------
TABLE 15-13. MODEL PLANT TREATMENT COSTS
Subcategory Aluminum Fluoride
Production 45,800 metric tons per year
($)
A. INVESTMENT COST BPT
Site development 53,300
Equipment 283,000
Monitoring equipment 20,000
Subtotal 356,300
Contractor's 0 & P a. 53,445
Subtotal 409,745
Engineering 81,949
Subtotal 491,694
Contingencies 49,169
Subtotal 540,863
Land 60,000
TOTAL INVESTMENT COST 600,863
B. OPERATION AND
MAINTENANCE COST
Labor and supervision 56,000
Energy 7,500
Chemicals 100,000
Maintenance 54,086
Taxes and insurance 18,026
Residual waste disposal .... 16,000
Monitoring , analysis
and reporting 15,000
TOTAL OPERATION AND
MAINTENANCE COST 266,612
C. AMORTIZATION OF
INVESTMENT COST 87,998
TOTAL ANNUAL COST 354,611
Overhead and Profit
4SS
-------
TABLE 15-14.MODEL PLANT UNIT TREATMENT COSTS
Subcategory Aluminum Fluoride
Annual Treatment Costs ($/kkg)
COST ITEM
PRODUCTION
(kkg/yr)
LEVEL OF TREATMENT
BPT
Annual Operation
and Maintenance 15,900
35,600
45,800
Annual
Amortization
Total Annual
Cost
15,900
35,600
45,800
15,900
35,600
45,800
10.36
6.51
5.82
3.86
2.16
1.92
14.22
8.66
7.74
489
-------
clarified effluent. This technology represents current practice
in the Aluminum Fluoride industry and was therefore selected as
the basis for the BPT effluent limitations.
B. Flow Basis
The basis of flow for BPT limitations is estimated from data
provided in the 308-Questionnaires for three of the four complete
plant responses received, including Plant #837, #251, and # 705.
Plant #188 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 wastewater sources contributing to the
total plant flow estimates include scrubber and work area
washdown. These wastewater sources are summarized in Table 15-4
for the three plants considered. The model plant flow for the
AlFj industry is estimated as the average total raw wastewater
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 mVkkg of product which is largely dependant on the
scrubber design and water utilization. The A1F3 process in Plant
# 251 shares an S02 scrubber with the anhydrous hydrofluoric acid
process. Wastewater 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 }.02 and 2.39 mVkkg of
product.
The average total flow for the three plants is 11.9 mVkkg of
product. This flow is used for the model plant in the aluminum
fluoride subcategory.
C. Selection of Pollutants to be RegulatecJ
The selection of pollutants for which specific numerical effluent
limitations are set was based on an evsiuation 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 Plants £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. Fluorid'
490
-------
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 set 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 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 lifted 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 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 regulated.
Specific numerical effluent loading limitations are set 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 set 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.
D. Basis of Pollutant Limitations
1. 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 the proposed
Development Document (60) 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 under "Basis of Pollutant Limitations." There is no
data available from plants where the BPT treatment
491
-------
performance can be evaluated for the treatment of raw
aluminum fluoride process wastewater alone. The
aluminum fluoride plants for which data is available
integrate raw process wastewater with wastewaters
generated from the hydrofluoric acid process.
In view of the similar wastewater 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, based on limitations established for the HF
industry (Table 12-21}, a maximum 30-day average
concentration of 97 mg/1 TSS and 53 mg/1 fluoride are
set for the A1F3 industry. These are relatively high
values that are unique to this industry. The
variability factor ratio of 2.1 was selected based on
the evaluation in the HF subcategory (Table 12-23).
The unit effluent load limitation is determined as
follows:
L (as kg/kkg) = (Q? (C)
1000
where C is the maximum 30-day average concentration in
mg/1, Q is the unit flow in mVkkg, and 1000 is the
conversion factor for kg to grams. (Note: kg/m3 =
1000 mg/1.)
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 30-day average effluent is:
(97 mg/1) (11.9 mVkkg) (kg/in3) = 1.2 kg/kkg
(1000 mg/1)
The 24-hour maximum effluent is:
{200 mg/1) (11.9 m3/kkg) (kq/m3) =2.4 kg/kkg
(1000 mg/1)
In the same manner the concentration basis for
fluorides is:
2.1 x 53 mg/1 =110 mg/1.
492
-------
The 30-day average effluent is:
(53 mg/1) (11.9 mVkkg) (kq/m3) = 0.63 kg/kkg
(1000 mg/1)
The 24-hour maximum effluent is:
(110 mg/1) (11.9 mVkkq) (kq/m3) = 1.3 kg/kkg
(1000 mg/1)
Toxic Pollutants
The effluent limitations set for the selected toxic
pollutant control parameters are derived from three sources
of information including 1) literature based treatability
estimates (Section 8), 2) screening and verification
sampling data, and 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 wastewater from the HF and A1F3 processes combined,
since no plant is available \*hich treats &1F3 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 for lime settling (BPT). Removal of toxic
pollutants from one wastewater 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 have been used as the basis for
determining specific numerical limitations for toxic
pollutants.
a. Chromium
The literature treatability value of 0.32 mg/1 from
Table 8-13 for lime settling is considered to represent
a long-term average concentration value for chromium in
view of plant performance data in the HF and combined
HF/AlFj industries.
Since long-term monitoring data on chromium is not
available, the variability factor ratio (VFR) of 3.3
was derived from the data in the nickel sulfate
493
-------
industry. This is justified by the fact that nickel is
one of the dominant metal pollutants selected for
control in the A1F3 subcategory and the wastewater
characteristics from the A1F3 and NiS04 industries can
be assumed to be similar. Therefore,
VFR = VF of daily measurements = 3.9
VF of 30-day averages 1.2
VFR =3.3
The effluent limitations for chromium are determined as
follows:
The maximum 30-day average is:
(0.32 mg/1) (1.2) = 0.38 mg/1
The 24-hour maximum concentration is:
{0.32 mg/1) (3.9) = 1.3 mg/1
The maximum 30-day average effluent limit is:
(0.38 mg/1) (11.9 mVkkq) (kg/m3) = 0.0045 kg/kkg
{1000 mg/1)
The 24-hour maximum effluent limit is:
(1.3 mg/1) (11.9 m3/kkq)(kq/m3) = 0.015 kg/kkg
(1000 mg/1)
The effluent limitations on chromium are presented in
Table 15-15 for BPT treatment.
b. Nickel
For the purpose of regulation, a value of 0.17 mg/1 was
derived from the HF industry as the subcategory
performance for nickel. The limitations are determined
as follows:
The maximum 30-day average is:
(0.17 mg/1) (1.2) = 0.20 mg/1
The 24-hour maximum concentration is:
(0.20 mg/1) (3.3) = 0.66 mg/1
where 3.3 is the VFR as discussed for chromium.
The maximum 30-day average effluent limit is:
494
-------
{0.20 mg/1) (11.9 m3/kkq)(kq/m*) = 0.0024 kg/kkg
(1000 mg/1)
The 24-hour maximum effluent limit is:
(0.66 mg/1) (11.9 mVkkg) (kq/m^) = 0.0079 kg/kkg .
(1000 mg/1)
c. Other Metals
The concentration bases for arsenic, copper, selenium,
and zinc are also presented in Table 15-15. These
pollutants are listed to serve as guidance in cases
where these pollutants are found to be of water quality
concern. The 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.
Basis for the BCT Effluent Limitations
While EPA has not yet proposed or promulgated a revised BCT
methodology in response to the American Paper Institute v. EPA
decision mentioned in Section 3, EPA is promulgating BCT limitations
for this subcategory. These limits are identical to those for BPT.
EPA is not promulgating any more stringent limitations since we have
identified no technology option which would remove significant
additional amounts of conventional pollutants. As BPT is the minimal
level of control required by law, no possible application of the BCT
cost tests could result in BCT limitations lower than those
promulgated in this regulation. Accordingly, there is no need to wait
until EPA revises the BCT methodology before promulgating BCT
limitations.
Basis for BAT Effluent Limitations
A. 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 BAT regulations.
For BAT, EPA is setting limitations based on Level 1 treatment.
Pollutants limited in the BAT regulations are fluoride, chromium,
and nickel.
495
-------
The Agency considered the use of treatment Level 2 (addition of
dual-media filter) but did not adopt it because the installation
of the filter is not very effective in the removal of TSS,
fluorides and toxic metals. EPA also 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.
B. Technology Basis
For BAT, the Agency is setting the effluent limitation on
fluoride and the toxic metals based on the BPT treatment system.
Similar to the Hydrofluoric Acid Subcategory, the addition of
dual-media filtration after alkaline precipitation and settling
is not considered cost effective based on the conclusions made in
Section 7 "Hydrofluoric Acid Subcategory" of EPA's "Treatability
Studies for the Inorganic Chemicals Manufacturing Point Source
Subcategory" {61) as referred to in Section 12 of this
Development Document.
C. Flow Basis
The same flow established for BPT above is used in the
development of the BAT effluent limitations. The flow used is
11.9 mVkkg of product (Table 15-4).
D. Selection of Pollutants to be Regulated
The Agency has selected fluoride and the same two toxic
pollutants identified in the BPT regulations for the BAT
regulations. The rationale for their selection is discussed
above.
E. Basis of Pollutant Limitations
The BAT limitations were set equal to BPT and are given in Table
15-15.
Basis for New Source Performance Standards
A. Technology Basis
For NSPS, the Agency set the same treatment technology as for
BAT.
B. Flow Basis
The same flow established for BPT and BAT is used in the
development of the NSPS effluent limitations.
496
-------
TABLE 15-15. EFFLUENT LIMITATIONS
Aluminum Fluoride
Best Practicable Control Technology Currently Available*
Waste water Flow; 11,9
Subcategory
Pollutant Performance
Conventional and
(mg/1)
Concentration Basis
(1) (iKf/l)
VFR
Max.
30-day 24-hr.
avg. max.
Effluent Limit
(kg/kkg)
Max.
30-day 24 hr.
avg. max.
ttonoonventiohal Pollutants:
Total Suspended
Solids, TSS
Fluoride
Tbxic Pollutants:
Arsenic
Chromium
Copper
Nickel
Selenium
Zinc
57 u'
33 (2)
0,50
0.32<3>
0.32<3>
0.17<3)
0. 20
0.55C3>
3
3
3
3
3
3
3
3
.5/1.7
.4/1
-9/1
.9/1
.9/1
-9/1
.9/1
.9/1
.6
.2
.2
.2
.2
.2
.2
(5)
(5)
(5)
(5)
(5)
(5)
97
53
0.
0.
0.
200
110
60
38
38
0.20
0.
24
0.66
2.
1.
1.
0.
0.
2.
0
3
3
66
78
2
1.2 2.4
0.63 1.3
_(6) _(6)
0.0045 0.015
_(6) __<6)
0.0024 0.0079
_(6) _<5>
_(6) _{6)
(1) VFR: Ratio of the 24-hour variability factor to the 30-day variability
factor
(2) Long term average construction based on the HF si&categary regulation
(Section 12.7.2).
(3) The lower limit of treatability estimate (Table 8-11) and the industrial
waste water treatment system performance (Table 8-12) are used as the
basis for the long term average limitation and subcategory performance,
since no plant is available where BPT treatment can be evaluated for
the AlF 3 waste water alone.
(4) VFR based on HF subcategory evaluation.
(5} VFR based on limited long term data.
(6) tto effluent limitation proposed.
* The BPT effluent limitations are also applicable to BAT except for the TSS
limitations. The BPT limitations are also applicable for NSPS.
497
-------
C. 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 above. The NSPS limitations are
presented in Table 15-15.
Basis for Pretreatment Standards
The Agency is not promulgating PSES and PSNs because a well-operated
POTW provides equal or better removal of chromium and nickel (the only
toxic pollutants regulated under BAT for this subcategory) than is
achieved by BAT for this subcategory, and therefore there is no
pass-through of these toxic pollutants in the Aluminum Fluoride
Subcategory. At present, there are no indirect dischargers in the
Aluminum Fluoride subcategory.
498
-------
SECTION 16
CHROME PIGMENTS INDUSTRY
Industry Profile
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 green, chrome orange, molybdenum
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 prior to promulgation of this new regulation is given in
Table 16-2.
Subcategorization
Several factors were originally considered in the Subcategorization
process, such as raw materials, products, manufacturing process, size
and age of equipment, and water pollution control technology. It was
concluded that if effluent limitations were to be tied to units of
production, only subdivision by dominant product was viable as a
method of primary Subcategorization. Further subdivision was not
warranted in the chrome pigment industry. A more detailed discussion
of the Subcategorization process may be found in Section 4.
General Process Description and Raw Materials
The general manufacturing process for each of the above
given below.
compounds is
A.
Chromium Oxide
This pigment consists of two compounds: anhydrous and hydrated
chrome oxide (Guigets Green). The amount of the anhydrous
chromic 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.
The anhydrous oxide is almost pure chromium oxide and the
commercial grade consists of a minimum of 98.5 percent Crz03. It
is prepared by calcination of sodium dichromate with sulfur or
carbon according to the reactions given below:
S = Cr0
23
Na2SO4
(1)
499
-------
TRBLK 16-1. SUBCAIEGOBy PROFILE .DMA SUMMARY
(1)
SUBGAOEGORy
CHROME PIOENIS
Total subcategory capacity rate
Ttotal 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: (2)
Minimum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
K&stewater flow range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
73,500 kkg/year
73,500 kkg/year
12
39,800 kkg/year
62 percent
100 kkg/year
18,000 kkg/year
6,300 kkg/year
6,400 kkg/year
78 percent
38 years
60 years
800 cubic meters/day
11,363 cubic meters/day
32 cubic meters/kkg
170 cubic meters/kkg
(1) Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S, Department of Commerce, Current Industrial
Reports, Decanber 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chenical Industry," June, 1978 and "Ecoronic Analysis of Proposed
Revised Effluent Guidelines and Standards for the Inorganic Chemicals Industry,
March, 1980.
(2) Based, on production at 11 plants, all other figures are based on 308
Ques tionnaires.
500
-------
TABLE 16-2. STATUS OF REGOIAHCNS - EEFUJENT-UMZTKnO? GUIDEU2ES
SDFTflTOaOKSr
SOBPAET
Chrome Pigments
AH (40CFR 415.340, 5/22/75)
STANDARDS
Product
Process
Chrome
Pigment
Para-
meters
TSS
Cr(T)
Cr*6
Pb
Zn
CN
CN(A)
Fe
BEdCA* BSOEA NSPS
Max.1 Avg.2 Max. Avg, Max. Avg.
iCQfi£tCO Tfrr /\r\rrt l^<"r /Irlrt-i T^rr /\r\ff^r If t*r J]F\rr* T^rr /I^l/*/^
«y/ JWg rj-j/ ™s-y i^y/ "-"M K*"y/ ^'iSJ ^y/ '^Ss J^^' ^^3
Cmg/1) (ing/1) (mgA) fin^l) &ng/l) Tirag/1)
5,1
(76.1)*
0.10
(1.5)
0.010
(0.2)
0.42
(6,3)
0.72
(10.8)
0.010
(1.5)
0.10
(0.2)
0.72
(10.8)
1.7 Reserved Reserved
(25.4)
0.034
(0.5)
0.0034
(0.1)
0.14
(2.1)
0.27
(4.0)
0.0034
(0.5)
0.034
(0.1)
0.27
(4.0)
415.340, 415.341, and 415.342 were revoked by the Agency
£41 FR 51601, November 23, 19761.
jMax. » Maximum of any one day,
2
Avg. m Average of daily values for thirty consecutive days.
*£Low basis 67,000 1/kkg,
501
-------
Na2Cr2 07 + 2C =
NazC03 + CO
(2)
The use of sulfur as the reducing agent eliminates C02 and CO
emissions but increases the sulfates in the raw waste as well as
producing 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
chrome oxide is given in Figure 16-1.
preparat ion of anyhdrous
Hydrated chromium oxide, CrzO3»2H20 or Cr20(OH)4/ also known as
chromium hydrate and Guigets Green, is a brilliant bluish green.
It is made by reacting sodium dichromate with boric acid as
follows:
2Na2Cr2O7 + 8H3B03 = 2Cr203»2H20
30
(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:
2NazB407 * H20 = 4H3BO3
(4)
A waste stream containing some boric acid and sodium sulfate is
discharge from the boric acid unit. Figure 16-2 is a generalized
flow diagram of the process.
B. Chrome Yellow and Chrome Orange
Chrome yellow is one of the more important synthetic pigments.
The chrome yellows cover the range of hues from light greenish
yellow to reddish medium yellow and consist mainly of lead
chromate. They are made by reacting sodium dichromate, caustic
soda, and lead nitrate. The reactions are given as:
2HN03 + PbO = Pb(N03)z + H20
Na2Cr207 + 2NaOH + 2Pb(N03)2 = 2PbCr04 + 4NaN03 + H20
(5)
(6)
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 wastewater
502
-------
VHff
— Of
SCHUBBER *• "•
IJQUID SCW-
i
in WATER
" m . 1
>2
f
^S «?" "T
f •
SOLFTm ^ PACXflGING OF
CHROME OXIDE
raceucT
VRSH HATER
Figure 16-1. General process diagram for production of anhydrous chrome oxide.
-------
SCOIUM
PICHKCHATE
BORIC ACID
ui
o
».
[ I*"
HIXEK
^
CVIK
WiTER
SUWRTf
1ANK
WKBER
PILTTO
1 — •P'1
HtVER
HVDRA.1Q) CHRCMB
C«II« TO CRIKflNG,
SCREEN! rC AtC
VRSTE HVTER
Figure 1G-2. General process diagram for production of hydrated chromic oxide.
-------
treatment facility. A flow diagram of the chrome
manufacturing process is shown in Figure 16-3.
Molybdenum Orange
yellow
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 reactions are given as follows:
Mo03 + 2NaOH = Na2Mo04 + H20
PbO + 2HN03 = Pb(N03}2 + H20
NazMo04 + Pb (N03)2 = PbMo04 + 2NaN03
Na2Cr04 + Pb (N03)2 = PbCr04 + 2NaN03
PbMo04 + PbCr04 = PbCr04»PbMo04
(7)
(8)
(9)
(10)
(11)
A simplified flow diagram for the manufacture of molybdenum
orange is given in Figure 16-4.
D. 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 solutions of iron sulfate and ammonium sulfate with
sodium ferrocyanide. 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 = PbCr04Fe(NH4)«Fe(CN)6
(12)
Figure 16-5 gives a process flow diagram for the manufacture of
chrome green.
Zinc Yellow
Zinc yellow, also called zinc chromate, is a greenish yellow
pigment. It is a complex compound of zinc, potassium, and
chromium which has the approximate composition
4ZnO»K20«4Cr207*3H20. It is made by the reaction of zinc oxide,
505
-------
UftDQBOCE
l/i
O
CHOB VEU£H
TO DWfINQ, HIUJNC
MD BAOCAGJNQ
WTER
FigiJre 16-3. General process diagram for production of chrome yellow.
-------
VENT
ui
o
-J
CMDE
WMER
CAUSTIC SOCft
DISSOLVE*
SODIUM
CHKMUE
11
MIXER
OXIDE __
jc ACID
VENT
DISSOSJVER
^^ PILTOftTICN
mSHINQ
fASTE HA.TER
DRYING
MILLING
AND
PACKAGING
CF
MDLYBCENIM ORANGE
J4'
PRCCUCT
Figure 16-4. General process diagram for production of molybdenum orange.
-------
hydrochloric acid, sodium di chroma te, 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:
2HC1 + 2Na2Cr207»H20 = K2Cr4013 + 4NaCl + 3H20 (13)
K2Cr4013 = 3H20 + 4ZnO«K20»4Cr04»3H20 (14)
2KC1
4ZnO
A general flow diagram of the manufacturing process is given in
Figure 1 6-6.
Water Use And Waste Source Characteristics
Water Use
In the chrome pigments industry, water is used primarily for
noncontract 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.
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 wastewaters combined and treated at
a single facility. A generalized flow diagram applicable to all
chrome pigment plants is given in Figure 16-7. The wastewater sources
are similar for all pigment products except that at chrome oxide
plants, an additional scrubber waste is generated. Table 16-4 gives
the wastewater flow data summary for several plants. The quantity of
wastewater and the pollutants are dependent of the raw materials used.
The figures in Table 16-4 represent actual plant discharges.
The data sources for the plants used in the determination of unit flow
values presented in Table 16-4 are outlined below:
Plant fIBpl Data based on 308-Questionnaire submission.
pigment production and flows were included.
Only chrome
Plant WifBKW Oo&a-* .based on 308-Questionnaire submission.
pigment and iron blue production and flows were included.
Chrome
508
-------
HATER
1
IRON BLUE
RESLURRY
tn
O
HATER
ij^-jj W.ITKA-IT: ^
SODIUM CHfcOMATE ^
SODIUM SOIFATE _
PEACTIOH
»
f 1
FILTER
AND
HASH
1
SHADE
f
TANK
DRYER
1
•» GRINDING
BUDDING
AND
PACKING
aiROME GREEN
PRODUCT
HASTE HATER
HASTE HATER
Figxore 16-5. General process diagram for production of chrome green.
-------
uo
Ul
H
O
^
KM fc
— T
use-new TMK
piuwncw
ma tun
HWINQ
>NIUJNSf PACKAGDC
OP 1KB UN? YQiCH
Figure 16-6. General process diagram for production of zinc yellow.
-------
TABLE 16-3. WATER USAGE IN THE CHRCME PIGMENTS SUBCAGEGOPY
(1)
USE
#464
UNIT FLOW (m3/kkg)
Plant Designation
#436 #214
Noncontact cooling
Direct process contact
Indirect process contact
Maintenance
Scrubbers
Boiler Feed
Total
9.50
18.6
7.18
12.0
3.30
2.52
53.1
6.45
147
NA<2>
1.78
9.56(3>
11.1
176
NA
32.6
NA
0.152
NA.
0.152
32.9
(1) Includes all chrome pigment product mixes. Values indicated only
for those plants that reported complete information.
(2) NA - Not applicable.
(3) Iron blue pigment process.
511
-------
HATER
HASH HATER
i
IMREACTED MATERIALS,
ETC.)
WASH DOWN HATER
H RAW MATERIALS ^
to ^
REttTOR
FILTER
COOfER
HILLING AID
SCREENING;
*» PIGMEHT
PRODUCTS
TO
1 11 1 PACKAGING
PIGMENT
WASTE WATER HCN-OOWiaCT KOfflCDlATE
(BY-PPDDUCT SAL'l-S, fflVAM WASTE
Figure 16-7. General process diagram for production of chrome pigment complexes.
-------
Plant 1436. Data based on 308-Questionnaire submission. Chrome
pigment production and flows were included.
Plant #002. Data based on three days of sampHng. Chrome pigment and
organic pigment (20 percent) productions and flows were included.
Plant #894. Data based on three, davs^of^sampling. Chrome pigment,
iron blue, and organ i c pi gmen t (15 percent) 'IpFocTucTlTDns 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.
Description Of Plants
Screening
Plant #894 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 hexavalent chromium 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 cycled 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.
Verification
Two plants were visited during the verification phase of the program.
The first plant, #002, 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 wastewaters 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 f i1ters, 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
513
-------
TSBIE 16-4. SU&MAEY OF WBSTE WaiTER FLOW
SUBCftTEQORY: CHRCME PIGMENTS
Plant Designation Waste Water Flovr '
(m3Akg)
#464 41.1
#214 32.8
#436 149
#002 78.4(2>
#894 170(2)
(3)
Average Flow 105v '
(1) includes waste water from 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 [ (unit flew) (production) ]
I (production)
.e. = Q) + Q2(P2) + Q3(P3)+...-K3n(Pn)
Where Q = Unit flew and P = production (which is considered
confidential information),
514
-------
loadings. At sample point #2, 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 #002. 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, and
2) inadequate clarification which in turn caused blinding of the
filter and the subsequent need for filter bypass.
Plant #894 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 was 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 solids, and toxic metals.
No organics were analyzed during verification.
Figure 16-9 shows the treatment system flow diagram with the sampling
points indicated. Table 16-6 gives waste flows and pollutant
loadings.
Toxic Pollutant Concentrations
The toxic pollutants found above treatable concentrations in the raw
wastes during sampling are given in the table below.
Maximum Concentration (j/g/1)
Pollutant Screening Verfication
2 Plants
Cadmium
Cyanide
Chromium
Copper
Lead
Zinc
Antimony
Selenium
Silver
Nickel
Phenol*
Bis (2 ethylhexyl)
Phthalate*
79
360
55,000
7,500
36,000
4,100
7,700
> 10
7
160
73
>0. 1
1,250
8,200
349,000
4,700
69,000
273,000
1,475
28
20
740
* from organic pigment process
515
-------
RAW WASTE SO,
ACID
CHROME TREATMENT
TANK
pH 3.0
CAUSTIC
CAUSTIC ADDITION
THROUGH
pH 8.5
~l
FILTER FEED
TANK
LAB FILTERED
1
FILTER AID
BACKWASH
OUTFALL
TO SEWER
I J
(FILTERS NOT WORKING SO
WERE BEING BYPASSED.
THIS WOULD BE THE FLOW
PATTERN IF FILTERS WERE
OPERATING.)
LEGEND
SAMPLING POINTS
Figure 16-8.
General waste water treatment process flow diagram at plant #002
showing the sampling points. (Chrome pigment manufacture.)
516
-------
TABLE 16-5. FLOW, POLLUTANT CONCENTRATION AND LOAD
DATA OF THE SAMPLED WASTE STREAMS FOR PLANT # 002
SUBCATEGORY: CHROME PIGMENTS
Conventional and Nonconventional Pollutants
mg/1
(kg/kkg of chrome pigments)
Stream
#
1
2-U
2-F
Stream
Description
Raw Waste
Unfiltered
Treated
Waste
Filtered
Treated
Waste
Flow
(m3/kkg) TSS
78.4 700
(55)
78.4 970
(76)
78.4 NA '
Fe
1.6
(0.13)
2.3
(0.18)
0.06
(0.0047 )
Cr(VI)
300
(24)
120
(9.4)
NA(1)
(1) NA - Not available
517
-------
SLAKED
LIME
WASTE
H2S04 SO.
LJL
WASTE / £fr
WATER '~
#1
(EPA SAMPLE
POINT AISO)
BLEND
TANK
PH 2.S-3.0
Ul
M
03
EQUALIZATION
TANK
NEUTRALIZATION
TANK
pH 6.2-6.5
NEUTRALIZATION
TANK
pH 8.0-8.3
BACKWASFT
HOLDING
TANK
BACKWASH
(2)
SAND
FILTERS
CLARIFIER
EFFLUENT
HOLDING
TANK
(3) CLARIFIERS
#2 (EPA SAMPLE
POINT ALSO)
FINAL
DISCHARGE
TO RIVER
#3
e
SLUDGE
LEACHATE
GRAB SAMPLE
SLUDGE
HOLDING
TANK
FILTRATE
LEGEND
SAMPLE POINTS
Figure 16-9.
General waste water treatment process flow diagram at plant #894
shewing the sampling points. (Chrome pigment manufacture.)
-------
TABLE 16-6. FLOW, POLLUTANT, CONCENTRATION AND LOAD DATA FOR THE SAMPLED
WASTE STREAMS AT PLANT # 894
SUBCATEGORY: CHROME PIGMENTS
Conventional and Nonconventional Pollutants
mg/1
(kg/kkg of chrome pigments)
Stream Stream Flow TSS Fe Cr (VT)
# Description (m3/kkg)
1
2
3
5
Raw Waste
Final
Discharge
Leachate
Sand Filter
Influent
170
170
NA*1'
170
770
(130)
3.9
(0.66)
ND<2>
11
(1-9)
48
(8.2)
0.30
(0.051)
0.04
(NA)
1.0
(0.17)
ND<2>
0.023
(0.0039)
ND(2>
ND(2)
(1) Not Applicable
(2) Not Detected
(3) Verification sampling which involves three 24-hour composite samples.
519
-------
Screening data was obtained at Plant 1894. Verification was completed
at Plants #894 and #002. The only organic pollutant not listed above
found in the raw waste above the protocol detectable limit (10 *g/l)
was naphthalene at 14 i*g/l. It should be noted, however, that some
nitrobenzene (56 »/g/l} and phthalates at levels up to 220 ^g/1 were
found in ttte treated effluent and one raw water intake. Since they
were not present in the raw wastes, it is presumed they are present as
a result of sample contamination; i.e., plasticizer in Tygon Tubing.
No organic pollutant sampling was made during verification.
Section 5 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 #002. This
involved five 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
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 chrome
pigments production rate, the waste stream flow rate, and the measured
pollutant concentration.
Unit loading (as kg of pollutant = (C) (Q)
per kKg of chrome pigments) (1000HP)
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 Ibs.)
The minimum, average, and maximum values are based on data from those
plants where the particular pollutant was found at concentrations
greater than the analytical detection limits and significant in that
it could conceivably be treated by an available treatment technology
regardless of economic considerations.
In Table 16-7, the toxic pollutant raw waste data are presented as the
average daily concentrations and the unit loading found at the
520
-------
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 ( 1 )
Lead
Nickel (1)
Zinc
Mercury
Cyanide ( 1 )
Phenol (1)
Phenolics ( 1 )
42,000
8,000
1,160,000
54,000
252,000
1,400
104,000
260
34,000
900
1 ,500,000
(1) Probable Source is organic pigment process or other processes.
The waste load for zinc and lead in the above have been adjusted to
reflect the fact that not all plants produce a chrome pigment
containing lead, and most do not produce zinc yellow. For example, a
plant that does not produce zinc yellow will have a very low level of
zinc in the raw waste load. See the data for plant #894 in Table 16-
7.
521
-------
16-7. TOXIC POLLOTRNT PAH HASTE DMA
CHHCMS
Average Daily Pollutant Concentrations and loadings at Plants Saiplad (1)
JB/1.
(kg/kkg of Chrome Pigments)
Pollutant
Antimony
Cadnimx
Chrcjnuin
Copper
Lead
Nickel
Zinc
Mercury
Cyanide, CM
Cyanide, OJ(A)
#894 (S) (2
7.7
U.5)
0.79
Co-is)
55
(10)
7.5
(1.4)
36
(6-8)
0.16
(0.030)
4.1
(0.78)
*
*
3.6
(0.68)
*
p1ap4- nM-i^n*
!) t894
-------
TABLE 16-8. SU44ARY OF RAH VRSTE LADINGS PQUM) IN SCREENING AND VESUFIOVTIGN SAMPLING
W
SUBCAIBGORY
Pollutant
Itoxic
Antimony
Cadmium
ChrcmW1'
OOpper
load
Nickel
Kino
Meccuty
Cyanide, aj
CHROME PIGMENTS
leading tenge,
(kg/day)
HinimiiQ fbxinun
6.
0.
700
6.
55
0.
48
0.
3.
Cyanide, CN(A)
Phenol
Phenol ica
Conventional
Total
Suspended
Solids, TSS
Fe
Hexavalent*1*
Chroniun
Cr+fi
and
0 98
87 10
. 1300
1 96
459
19 2.0
714
0019 0,48
1 56
9.8
0.93
8.8
Nonoonvaitional
3100 8800
7.
1 550
1300
Unit loading,
(kg/kkg)
Mininuro Average
0.11
0.016
10
0.11
0.82
0.0028
0.71
0.000034
0.056
55
0.13
0.58
0.11
16
0.74
3.9
0.019
4.8
0.0036
0.53
0,15
0.014
0,13
93
4.2
24
Maxinun
1.5
0.15
24
1.4
6.8
0.030
1313
0.0072
0*84
130
8.2
K,. of'2'
Plants
Averaged
3
3
3
3
3
3
3
2
3
1
1
1
2
2
1
(1) Hexavalent chromium is only one valent form of chromiun.
(2) Oily those plants where the pollutant vas observed at significant levels were included.
-------
TABLE 16-9. TOXIC POLLUTANT TREATED V&STE DATA
Pollutant
Cmg/1)
Antimony
Cacbuim
Chromium
Copper
Lead
Nickel
Zinc
Mercury
Cyanide, CN
Cyanide, OJ(A)
SUBCATEGOIGf:
CHEDME PIQYENTS
Plant Designation
#894. #002
0.30
0.0084
0.33
0.035
0.11
0.021
0.058
ND(3>
0.065
0.0067
0.43
0.12
130
0.077
1.5
0.083
117
ND
*
*
Overall C2)
Average Concentration
0.37
0.064
65
0.056
0.81
0.052
59
ND
0.065
0.0067
(1) Verification sampling concentration data, average of three 24-hour
composite samples.
(2) Average of two plants shown during verification sampling.
(3) Not detected.
* No data
524
-------
Pollution Abatement Options
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, and 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 wastewaters 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.
A. Chrome Yellow and Chrome Orange
The raw wastewater contains sodium acetate, sodium chloride,
sodium nitrate, sodium sulfate, and lead salts.
B. Chrome Oxide
The aqueous process effluent contains sodium sulfate. If boric
acid is used in the preparation of hydrated chromic oxide then
the wastewater will contain sodium borate and boric acid.
C. Chrome Yellow and Chrome Orange
Additional pollutants present in the raw wastewater from chrome
yellow and chrome orange manufacture include sodium acetate,
sodium chloride, sodium nitrate, sodium sulfate, and lead salts.
D. Molybdenum Orange
Process waste effluents from the manufacture of molybdenum orange
contain sodium chloride, sodium nitrate, sodium sulfate, chromium
hydroxide, lead salts, and silica.
525
-------
E. Chrome Green
The raw wastewater contains sodium nitrate. If iron blue is
manufactured on-site as part of the process for chrome green
manufacture, the wastewater also contains sodium chloride,
ammonium sulfate, ferrous sulfate, sulfuric acid and iron blue
pigment particulates.
F. Zinc Yellow
The raw wastes contain hydrochloric acid, sodium chloride,
potassium chloride, and soluble zinc salts.
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.
A. Close attention to product quality in conjunction with reduction
of product rinses.
B
Reduction in equipment cleaning rinses
methodologies:
by the following
1 .
2.
Recycle of rinse waters.
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 (#002).
C. 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.
D. The use of ion exchange and/or reverse osmosis on isolated
wastewaters. This will allow total recovery of product as well
as total reuse of wastewater. This system is in use on one line
at Plant 1409.
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.
Best Management Practices
A. 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.
526
-------
B. 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.
Prevailing Control and Treatment Practices
A description of the individual treatment facilities for those plants
visited is given above. In addition, the following information was
obtained for the remaining plants.
Plant #214 manufactures pigments and other chemicals. The plant does
not have a wastewater treatment facility. After pH adjustment, waste
is discharged to a POTW. Part of the process waste is recycled.
Plant #593 manufactures organic and inorganic chemicals. The existing
combined wastewater treatment plant consists of lagoon, aeration,
clarifiers, and filters. The sludge disposal is on-site landfill.
Plant #464 manufactures both organic and inorganic pigments. After pH
adjustment, wastewater is discharged to a POTW.
Plant #101 manufactures inorganic ceramic pigments, color and
porcelain. The existing combined wastewater treatment facility
consists of a series of settling basins. Sludge disposal is to an
off-site landfill. After pH adjustment, the final discharge is to a
POTW.
Plant #502 manufactures both organic and inorganic pigments, of which
chrome pigments are a small part. Treatment consists of pH adjustment
prior to discharge.
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 #409 manufactures specialty chemicals and inorganic pigments.
The existing wastewater 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 wastewater treatment facility
consists of pH adjustment, S02 reduction and lagoons.
Plant #962 manufactures inorganic pigments (chrome yellow). The
existing waste treatment plant consists of flocculation,
clarification, and filters. After pH adjustment, the effluent is
discharged to a POTW. Sludge is recycled to process.
527
-------
Plant #200 manufactures 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 #894. 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.
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
technologies were reviewed for model plant development: sulfide
precipitation, ion exchange, reverse osmosis, and the xanthate
process.
Selection o£ Appropriate Technology and Equipment
Technologies for Different Treatment Levels
A careful review of the end-of-pipe treatment methods available to
industry was made. As a result, the following methodology was set
forth as the only treatment level for this subcategory. Originally,
sulfide precipitation was considered as a level 2 addition to the
level 1 technology. However, an Agency treatability study (61)
indicated no significant improvement in pollutant reduction was
achieved by this addition. The result of eliminating level 2
technologies from consideration is to set NSPS standards equal to
those for BPT/BAT. Considerations made in establishing the level 1
model included:
Effective reduction of pollutants.
Established treatment practices in the industry.
The cost of technology.
The adaptability of the model to different situations.
A. Level 1 (BPT/BAT/NSPS/PSES/PSNS)
Consists of equalization, S02 reduction, alkaline precipitation,
clarification, and filtration. A flow diagram of the level 1
system is shown in Figure 16-10.
528
-------
in
to
I
SULFURIC
I ACID
BACKWASH
CAUSTIC SOD A
RAW '
1 .
WASTE WATER 1 '
L
•
j
\
SULFUR
DIOXIDE
.1 .
i
PO
I HOLDING TANK REACTION MIX
TANK TANK
I ^^
w inmiiiK-t
I SLUDGE
f TANK
TO LANDFILL
*
Includes flow monitoring, pH monitoring and sampler.
ADJUSTMENT
D - »-
^EFFLU
EFFLUENT
FILTER
Figure 16-10. Level 1 waste water treatment for chrcme pigments.
-------
B. Level 2
Sulfide treatment addition to BPT system has been dropped from
consideration as a level 2 treatment. For a discussion of the
proposed system and the cost evaluation see the Proposed
Development Document (60). Figure 16-11 is a diagram of the
proposed system.
Equipment for Treatment
A. Equipment Functions
In the model treatment system, 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, lime, or other neutralizing agent is
then added to neutralize the acid and precipitate heavy metal
hydroxides and a polymeric coagulant is added to help settle the
heavy metal hydroxides in a clarifier. The settled effluent is
then filtered in a dual-media filter and discharged after pH
adjustment to the range 6 to 9.
B. 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 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.
C. Separation and Disposal of Solids
Solids from the clarifier, including recirculated filter backwash
solids are dewatered in a filter press and hauled to a chemical
landfill. Sludge filtrate is returned to the influent holding
tank.
D. 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.
530
-------
FERROUS SULFATE SODIUM BISULFIDE
0
BACKWASH
CAUSTIC
I SODA
RAW
i
WASTE WATE
r
SULFUR
DIOXIDE
| HOLDING TANK
I
1
1
1
POLYMER
J
REACTION MIX
TANK TANK
U -
T
SLUDGE
TO LANDFILL
Includes flow monitoring, pH monitoring and sampler.
EFFLUENT
Figure 16-11. Level 2 waste water treatment for chrome pigments.
-------
Treatment Cost Estimates
General Discussion
To prepare cost estimates, a model plant concept was developed and
plant criteria developed for BPT (Level 1).
A. Wastewater Flow
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 mVkkg (25,200 gal/ton). This
value was used for sizing the model plants.
B. Chromium Pigment Production
Production in the chrome pigment subcategory ranges from a low of
500 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 many
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.
C. Wastewater 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 estimated costs of the BPT system for the four models are
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.
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:
532
-------
Variations in land costs.
Variations in hydraulic loading.
Varying costs of solid waste disposal.
Modifications to the existing wastewater treatment facility and
other related costs
The following overall results were obtained:
Model Plant
Plant #002
Plant #894
Annual Costs
($/kkg)
89.75
85.38
91 .03
The above data indicate a fair correlation between the model
plant and site specific cost estimates.
Table 16-14 presents a summary of the unit cost distribution
between amorization, operation, and maintenance cost components
at various production levels.
For the model plant, the primary sources of wastewater are from
product washings, slurrying of reaction products, scrubbing of
reactor vent gases, and washing of equipment due to product
changes.
Model Plant Costs
The major costs for the Level 1 (BPT) model plant are equipment,
labor, and chemical costs. Engineering design and equipment
maintenance are also fairly large. The majority of the annual cost is
tied up in operation and maintenance. This cost can approach 50
percent of the total capital cost.
Basis for Regulations
Evaluation of BPT Treatment Practices
A number of factors are anticipated to contribute to a wide variation
in the effluent quality at chrome pigment plant treatment facilities.
Consideration of these variations is included in establishing
limitations in that the performance of the plant on which limitations
are based is a large complex plant that encounters all of these
factors. These include the following:
533
-------
TABLE 16-10- MODEL PLANT TREATMENT COSTS
Subcategory
Production
Chrome Pigments
1,500 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment
Subtotal ..............
Contractor's 0 & P b.
Subtotal
Engineering ,
Subtotal
Contingencies
Subtotal
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
($)
BPT
600
317,000
20,000
337,600
50,640
388,240
77,648
465,888
46,589
512,477
6,000
518,477
112,000
7,500
53,000
51,248
15,554
5,000
15,000
259,302
BAT
0
0
0
0
0
0
0
0
0
0
0
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
83,380
342,682
a Represents the incremental cost above that for BPT treatment
Overhead and Profit
534
-------
TABLE 16-11. MODEL PLANT TREATMENT COSTS
Subcategory Chrome Pigments
Production 4,000 metric tons per year
A. INVESTMENT COST BPT BAT a
Site development 1,200 0
Equipment 564,000 0
Monitoring equipment 20,000 0
Subtotal 585,200 0
Contractor's 0 & p b. 87,780 0
Subtotal 672,980 0
Engineering 134,596 0
Subtotal 807,576 0
Contingencies 80,758 0
Subtotal 888,334 0
Land 12,000 0
TOTAL INVESTMENT COST 900,334 0
B. OPERATION AND
MAINTENANCE COST
Labor and supervision 112,000 0
Energy 15,000 0
Chemicals 141,500 0
Maintenance 88,833 0
Taxes and insurance 27,010 0
Residual waste disposal .... 15,000 0
Monitoring , analysis
and reporting 15,000 0
TOTAL OPERATION AND
MAINTENANCE COST 414,343 0
C. AMORTIZATION OF
INVESTMENT COST 144,532 0
TOTAL ANNUAL COST 558,875 0
f* Represents the incremental cost above that for BPT treatment
Overhead and Profit
535
-------
TABLE 16-12. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Chrome Pigments
6,000 metric tons per year
A. INVESTMENT COST
BPT
($)
BATa
Site development
Equipment
Monitoring equipment
Subtotal
Contractor's 0 & P .
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
1,200
738,000
20,000
759,200
113,880
873,080
174,61*5
1,047,696
104,770
1,152,466
12,000
1,164,466
112,000
20,000
211,500
115,247
34,934
20,000
15,000
528,681
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
187,506
716,187
a Represents the incremental cost above that for BPT treatment
Overhead and Profit
536
-------
TABLE 16-13. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Chrome Pigments
18,000 metric tons per year
A. INVESTMENT COST
Site development ,
Equipment ,
Monitoring equipment ..,
Subtotal ,
Contractor's O & P ^...,
Subtotal ,
Engineering
Subtotal
Contingencies
Subtotal
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
($)
BPT
1,800
1,700,000
20,000
1,721,800
258,270
1,980,070
396,014
2,376,084
237,608
2,613,692
18,000
2,631,692
TOTAL OPERATION AND
MAINTENANCE COST
112,000
28,000
635,000
261,369
78,951
60,000
15,000
1,190,320
BAT
0
0
0
0
0
0
0
0
0
0
0
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
425,248
1,615,568
a Represents the incremental cost above that for BPT treatment
b
Overhead and Profit
537
-------
A. Product Changes
Changes in products require that equipment be thoroughly cleaned
prior to reuse. Therefore, frequent product changes will result
in higher waste flows.
B. Product Application
The final disposition of the product will affect the quality
required. The higher the quality, the more water required for
rinsing.
C. 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.
D. Other Related Products
Many plants manufacture other types of pigments including iron
blues and organic pigments. These products generate significant
quantities of wastewater which tend to dilute chrome pigment
wastes. However, these wastewaters 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
regulations:
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 wastewater 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, 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
538
-------
TABLE 16-14 MODEL PLANT UNIT TREATMENT COSTS
Subcategory Chrome Pigments
Annual Treatment Costs ($/kkg)
LEVEL OF TREATMENT
COST ITEM PRODUCTION BPT
(kkg/yr)
Annual Operation
and Maintenance 1,500 172.87
4,000 103.59
6,000 88.11
18,000 66.13
Annual
Amortization 1,500 55.59
4,000 36.13
6,000 31.25
18,000 23.62
Total Annual
Cost 1,500 228.45
4,000 139.72
6,000 119.36
18fOOO 89.75
539
-------
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 #894 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-1la sets forth means,
variability factors, and the maximum 30-day average performance.
Maximum daily performance 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 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 20-30 percent of total production. One
other plant, with a different treatment system, meets the
promulgated BPT/BAT limitations. Most other plants have some
type of treatment installed, but none of these appear to be
adequate. This technology is expected to remove 3,450,000 pounds
per year of toxic metals.
Basis for BPT Effluent Limitations
A. Technology Basis
For BPT, the Agency is promulgating limitations based on
equalization, reduction of hexavalent chromium followed by
alkaline precipitation, and dual-media filtration. Reduction of
flow by the methods given in this section 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.
B. Flow Basis
The basis of flow for the 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. Plants producing a greater quantity
540
-------
TABLE 16-15. SUMMARY OF LONG TERM AND VERIFICATION EFFLUENT SAMPLING
RESULTS AT PLANT #894
SUBCATEGORY:
Pollutant
Total Suspended
Solids, TSS
Iron
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Cyanide (CN-A)
Cyanide (Total)
Chromium (VI)
(1) ND, Not Detected.
(2) NAr Not Available.
CHROME PIGMENTS
Verification Sampling
(mg/1) (kg/kkg)
3.9
0.30
0.30
ND(1)
0.0084
0.33
0.035
0.11
ND
0.021
0.058
0.065
0.0067
0.023
0.66
0.051
0.051
ND
0.0014
0.056
0.0060
0.019
ND
0.0036
0.0099
0.011
0.0011
0.0039
Achievable Performance ^
Max 30-day Avg
(mg/1) (kg/kkg)
23 3.8
NA NA
NA NA
0.16 0.027
0.12 0.020
0.73 0.12
0.25 0.42
0.87 0.15
0.0016 0.00027
NA NA
0.074 0.013
0.068 0.012
0.31 0.053
0.30 0.051
(3) From Tables 16-6 and 16-9.
(4) Fran Table A-lla,
"Historical
Effluent MDnitt
Dring Data Summary."
541
-------
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
wastewaters from the various pigment processes for treatment, the
basis of flow for the purpose of regulation includes all process
related wastewater combined. The flow basis is 105 mVkkg from
Table 16-4. This flow does not include any recycle or reuse of
wastewaters other than some incidental recycle being done at five
plants included in the data base.
C. Selection Basis for Pollutants to be Regulated
The selection of pollutants for which specific numerical effluent
limitations are set was based on an evaluation of raw waste data
from the screening and verification sampling program. Pollutant
data from the plants sampled during screening was used to
determine the need for verification sampling. Verification
sampling at Plants 1002 and #894 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 this section 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 at
significant concentration levels during sampling. Pollutants
from this list were considered 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.
The relative significance of the toxic pollutants was estimated
based on the total annual raw waste load for each pollutant which
appears in a Table in this section. The total annual load is
based on the average concentration observed during screening and
verification which is tabulated in Table 16-8, in addition to the
estimated annual production of 73,500 kkg of product for the
industry.
Specific numerical effluent loading limitations had been proposed
for those 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
542
-------
chrome pigment wastes at treatable levels (mercury is treatable
by sulfide precipitation which was rejected as the basis for
BAT). These metals were all considered for regulation, however,
final regulations include limitations on chrome (total), zinc,
and lead only. Organic pollutants and cyanide are not included,
since they are considered products of iron blue, ' organic
pigments, or HCN production as discussed above. In addition,
these parameters will be covered by future regulations in other
subcategories.
The Agency's decision to restrict metals regulation to chrome,
lead, and zinc was made in light of the performance
characteristics of the treatment technology for this subcategory,
and the nature of the waste involved. The performance of this
technology is such that control of certain "key" metals will
result in sufficient control of the remaining metals. This is
discussed in detail in Section 8.
The treatment technology designed for removal of the more
prominent metals in this subcategory, namely chrome, zinc and
lead, will ensure control of all the toxic metals originally
considered for regulation. Performance data from this
subcategory substantiates this (Tables 16-15 and 16-16).
The Agency is also aware that specific plants may have unusually
high loadings for one or more of the unregulated metals. In
these instances, limitations should be set on a case by case
basis. Table 16-16 includes guidance for copper, antimony,
nickel, cadmium, and mercury concentrations for those cases where
additional control is deemed necessary.
Hexavalent chromium was excluded from consideration in the final
regulations. The complexity and subsequent accuracy of the
analysis may cause misleading conclusions if hexavalent chromium
is used as an effluent monitoring parameter. Sulfur dioxide
reduction under acidic conditions will convert . hexavalent
chromium to its trivalent form which can be conveniently verified
by analysis of total chromium in the treated effluent. Chromium
cannot be removed by alkaline precipitation unless it is in the
trivalent 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 Limitations
A. Conventional and Nonconventional Parameters
1 . 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 the proposed Development Document (60) and
the JRB Study (52).
543
-------
2. Total Suspended Solids (TSS)
Review of the long-term monitoring and verification sampling
data in Table 16-15 indicates a maximum 30-day average TSS
discharge of 3.8 kg/kkg for the purpose of the limitation
determination. The 30-day average concentration basis is
then determined as follows:
(3.8 kq/kkq) (1000 mq/1) = 36 mg/1
(105 mVkkg) (kg/m*)
The 24-hour maximum loading limitation is determined by the
following relationship:
Maximum 30-day X VFR = 24-hour maximum
average loading loading or
or concentration 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 $894 (Table 16-15). The
VFR of 2.1 for zinc in the Ti02 Sulfate Process reflects the
overall metal removal performance of alkaline precipitation
followed by settling and discharge without filtration. The
choice of VFR is also supported by a statistical evaluation
of data developed in the Agency Treatability Study (61).
For the chrome pigments subcategory, the study data
indicated a VFR of 2.3 for chrome concentrations in treated,
but unfiltered wastes from a level 1 system. Therefore, a
VFR of 2.4 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,
(3.8 kg/kkg) (2.-4) = 9.1 kg/kkg
3. 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.
B. 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
544
-------
TABLE 16-16. EFFLUENT LIMITATIONS
Chrome Pigments A
Best Practicable Control Technology Currently Available
Waste Water Flow: 105 m3/kkg
Concentration Basis Ef f luent Limit
(mg/1) (kg/tog)
Subcategory Max.
Performance / - , 30-day 24-hr
Pollutant (mg/1) VFRU' Avg. Max.
Conventional and Nonconventional Pollutants
Total Suspended ,-.
Solids, TSS 18.1 4.8/2.0UJ 36 87
Iron 0.30(4) 4.4/1.8(6) 0.54 1.3
Max.
30-day 24-hr
Avg, Max.
3.8 9.1
Toxic Pollutants
Chromium
Zinc
Lead
Antimony
Cadmium
Copper
Nickel
0.71
0.71(5)
0.66
0.40 (7>
0.13
0.21
0.30<8>
4.1/1.7
4.1/1.7
5.0/2.1
4.4/1.8(6)
3.8/1.6
4.6/1.9
4.4/1.8(6)
1.2
1.2
1.4
0.72
0.20
0.41
0.54
2.9
2.9
3.4
1.7
0.49
0.99
1.3
0.13
0.13
0.15
—
—
—
—
0.31
0.31
0.36
—
—
—
—
(1) Subcategory long-term average concentrations adjusted for model plant flow,
unless otherwise indicated.
(2) VFR: Ratio of the 24-hour variability factor to the 30-day average
variability factor.
(3) VFR of 2.4 from long-term evaluation in Titanium Dioxide Subcategory,
verified by analysis of chrome pigments subcategory results from an
Agency Treatability Study (61). 30-day variability factors from Table
A-11A in Appendix A unless otherwise specified.
(4) Verification sampling results based on three 24-hour composite effluent
samples, adjusted for model plant flow.
(5) long-term averages for zinc, which include zinc chromate production are
not available; therefore, limitations are set equal to those for chrome.
(6) Variability factors based on average of those for other parameters since
no long-term data available.
(7) Lower limit of treatability estimate (Table 8-11).
(8) Estimated achievable long-term average (Table 8-13).
(*) Conventional and nonconventional pollutant limitations apply also to NSPS.
Toxic pollutant limitations apply also to BAT, PSES, and PSNS.
545
-------
data, 2} literature based treatability estimates (Section 8), 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 #894 appears to demonstrate that in
some cases the effluent quality for metal pollutants are
considerably better for BPT treatment than indicated by
literature treatability in Section 8. This high degree of
incidental removal supports the contention that applying effluent
limitations just to the dominant metal pollutants, assures
effective control of the other metals.
The VFR used to determine the proposed 24-hour maximum
limitations is based on long-term data for zinc in the Titanium
Dioxide subcategory, which is supported by analysis of
treatability results for this subcategory (61).
1. 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 0.13 kg/kkg which is the
limitation basis. The concentration' basis then becomes:
(0.13 kq/kkq) (1000 mq/1) =1.2 mg/1
(105 mVkkg) (kg/m3)
The 24-hour maximum is determined as follows:
(0.13 kg/kkg) (2.4) = 0.31 kg/kkg
where the VFR is set equal to 2.4 based on data from the
Ti02 subcategory and the Treatability Study (61).
2. Zinc
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 #002 where zinc is found at very high raw
waste concentrations.
3. 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
546
-------
discharge of 0.15 kg/kkg which is used as the 30-day average
limitation. The concentration basis then becomes:
(0.15 kg/kkq)(100 mq/1) =1.4 mg/1
(105 mVkkg) (kg/m3)
The 24-hour maximum limitation then becomes:
(0.15 kg/kkg) (2.4) = 0.36 kg/kkg.
4. 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. Therefore, the
concentration basis becomes:
(0.043 kq/kkq) (1000 mq/1) = 0.41 mg/1
(105 mVkkg) (kg/in3)
5. Antimony
The raw waste concentration for antimony was observed as
high as 7.7 mg/1 and averaged 3.3 mg/1 during sampling.
Since no long-term data is available, the lower limit of
treatability (Table 8-11) for antimony is used to determine
the concentration basis. This then becomes:
(0.4 mg/1) (1.8) = 0.72 mg/1
where 1.8 is an estimated 30-day average variability factor
for antimony, obtained by averaging known factors for
pollutants in the subcategory (Table 16-16).
6. 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 0.021 kg/kkg. The concentration basis
then becomes:
(0.021 kq/kkq) (1000 mq/1) =0.20 mg/1
(105 mVkkg) (kg/m3)
7. 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 an industry
547
-------
long-term average of 0.30 mg/1 (Table 8-13). Therefore, the
concentration basis becomes:
(0.30 mg/1) (1.8) = 0.54 mg/1
where, again, 1.8 is an estimated 30-day average variability
factor for nickel in this subcategory.
The limitations are summarized in Table 16-16 for BPT.
Basis for BCT Limitations
While EPA has not yet proposed or promulgated a revised BCT
methodology in response to the American Paper Institute v. EPA
decision mentioned in Section 3, EPA is promulgating BCT limitations
for this subcategory. These limits are identical to those for BPT.
EPA is not promulgating any more stringent limitations since we have
identified no technology option which would remove significant
additional amounts of conventional pollutants. As BPT is the minimal
level of control required by law, no possible application of the BCT
cost tests could result in BCT limitations lower than those
promulgated in this regulation. Accordingly, there is no need to wait
until EPA revises the BCT methodology before promulgating BCT
1 imitations.
Basis for BAT Effluent Limitations
A. 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 evaluated in
detail and taken into consideration in the selection of the
technology basis for the BAT regulations.
The Agency is promulgating BAT limitations based on treatment
consisting of Level 1 technology which is equivalent to BPT. The
implementation of BPT/BAT will remove 2,400,000 pounds of toxic
metals annually. The two plants currently meeting the
promulgated BPT/BAT limitations aro removing an additional
900,000 pounds per year of toxic metals.
B. Technology Basis
For BAT, the Agency is utilizing the identical technology basis
discussed for BPT in this section. BAT includes no additional
treatment because the Agency has concluded, based on the
Treatability Study (61) results, that sulfide treatment after
alkaline precipitation does not significantly increase treatment
performance, and, therefore, does not justify costs associated
with such additional treatment technology.
548
-------
C. Flow Basis
The unit flow of 105 mVkkg is also used for BAT.
D. Selection of Pollutants to be Regulated
The basis of pollutant selection is discussed for BPT above. For
BAT, the toxic metals shown in Table 16-16 are selected for
regulation. These include chromium, zinc and lead.
E. Basis of Pollutant Limitations
The basis of the limitations are discussed in detail under BPT
above.
Basis for New Source Performance Standards
A. 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 wastewater from the chrome pigment process with
wastewater from unrelated processes. The Agency proposes that
for new sources, the wastewater from the chrome pigments process
be segregated from wastewater from other processes unless the
other wastewater contains toxic metal pollutants. Segregation
and separate treatment of the wastewaters can conceivably reduce
treatment costs, and simplify the treatment of metals without
complications from unrelated wastewater constituents not amenable
to metals treatment.
B,
D
Technology Basis
For New Source Performance Standards (NSPS), the agency is
setting limitations based on the identical technology basis
discussed for BPT above. The Agency also recommends that all
unrelated wastewater sources which are not amenable to metals
treatment be segregated before treatment as previously discussed.
setting
Flow
The basis for the unit flow used for the purpose of
limitations is 105 mVkkg and does not differ from BPT.
Selection of Pollutants to be Regulated
The same conventional, nonconventional, and toxic pollutants
selected for BPT are also considered here for the NSPS
limitations. These include TSS, pH, iron, and the same eight
toxic metal pollutants.
Basis of Pollutant Limitations
549
-------
The basis of the limitations are discussed in detail under BPT
above.
Basis for Pretreatment Standards
The Agency is promulgating PSES and PSNS that are equal to BPT
limitations because BPT provides better removal of chromium, lead, and
zinc than is achieved by a POTW and,, therefore, these toxic pollutants
would pass through a POTW in the absence of pretreatment. In
particular, 1.4 million pounds per year of hexavalent chromium from
existing sources would pass through in the absence of pretreatment.
The promulgated pretreatment regulations will remove over 1.3 million
pounds of hexavalent chromium per year. Pollutants regulated under
PSES and PSNS are chromium, lead, and zinc.
Existing Sources
There are currently eight indirect discharge chrome pigment plants in
the subcategory. For Pretreatment Standards for Existing Sources
(PSES), the Agency is setting limitations based on BPT described
above. The pollutants to be limited are chromium, zinc, and lead as
presented in Table 16-16.
We are excluding small plants discharging less than 210,000 m3 process
wastewater per year to POTW from compliance with these PSES. They
will be subject only to the general pretreatment standards in 40 CFR
Part 403. The exclusion is intended to apply to plants producing less
than 2000 kkg per year, but we have established a flow basis for the
convenience of POTW's, since water use is much easier to monitor than
production. The annual flow basis for the exclusion was calculated by
multiplying the unit flow, 105 mVkkg, by 2000 kkg per year: (105
mVkkg) (2000 kkg/year) = 210,000 mVyear. Plants discharging less
than 210,00 m3 process wastewater per year produce less than 2000 kkg
per year chrome pigments. There would be very significant economic
impacts on this segment of the industry if they were required to
comply with these PSES. See the Preamble to the Regulation and the
Economic Impact Analysis of Pollution Control Technologies for
Segments of the Inorganic Chemicals Manufacturing Industry, EPA 440/2-
81-023.
New Sources
For Pretreatment Standards for New Sources (PSNS), the Agency is
setting limitations based on NSPS. The pollutants are indicated in
Table 16-16.
550
-------
SECTION 17
HYDROGEN CYANIDE INDUSTRY
Industry Profile
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, plexiglas 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
prior to promulgation of this new regulation is given in Table 17-2.
Subcategorization
The method of primary Subcategorization chosen for the inorganic
chemicals point source category was subdivision by dominant product.
Other factors taken into consideration for Subcategorization included:
raw materials used, manufacturing process employed, geographical
location, size and age of equipment and facility involved,
non-water-quality aspects of waste characteristics, water pollution
control technology, treatment costs, energy requirements and solid
waste disposal. A detailed discussion of these factors is given in
Section 4.
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.
The Hydrogen Cyanide Subcategory in this regulation is confined to the
Andrussow process. Since hydrogen cyanide is a by-product of the
acrylonitrile manufacturing process, this process will be covered in
the organic chemicals manufacturing category with the primary product.
General Process Description and Raw Materials
The raw materials are reacted at elevated temperature {900-1000
degrees C) over a platinum catalyst. The reaction is given as:
2CH4 -f 2NH3 4 30Z = 2HCN -f 6HZ0 (1)
551
-------
TABLE 17-1. SUBCATEGORY PROFILE DATA SUMMARY
SDBCATEGORY HYDRO2N CYANIDE*
Total subcategory capacity rate 289,000 kkg/year
Total subcategory production rate 166,000 kkg/year
No. of plants in this subcategory 7
Plant age range:
Minimum 5 years
Maximum 30 years
308 Data** on file for 2
With total capacity of 179,000 kkg/year
With total production of 116,000 kkg/year
Representing capacity 62 percent
Representing production 70 percent
Average production 57,800 kkg/year
Average capacity utilization 65 percent
Waste water flow per unit product
Minimum 10 m3/kkg of HCN
Maximum 57 m3/kkg of HCN
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 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.
* Includes data from plants using Andrussow Process and from plants
recovering HCN as a byproduct from the manufacture of acrylonitrile.
**Includes data from plants using Andrussow Process.
552
-------
TABLE 17-2. STATUS OF REGULflTICNS - EFFLUENT LIMITATION GUIDELINES
SUBPART
HYDROC2N CYANIDE
AP (40 CFR 415.420, 5/22/75)
STANDARDS
BPCTCA*
Max.
Avg.
Product
Process
Parana
eters
BATEA
Max. Avg.
kg/kkg kg/kkg
(mg/1) (mg/D
NSPS
Max. Avg.
kg/kkg kg/kkg
(mg/1) (mj/1)
Andrussow
Process
TSS
CN
CN(A)
BGDC
2.4 1.2
(48.0)** (24.0)
0.005
(1.0)
0.005
(0.1)
3.6
(72.0)
0.36
(7.2)
0.025
(0.5)
0.005
(0.05)
1.8
(36.0)
0.18
(3.6)
Sections 415.420, 415.421, and 415.422 were revoked by the Agency
( 41 FR 10681, February 23, 1977).
jyiax. = Maximum of any one day.
2
Avg. = Average of daily values for thirty consecutive days shall not exceed.
**
flow basis 50,000 1/kkg.
553
-------
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 non-methane 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
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 atomsphere. 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.
Water Use and Waste Source Characteristics
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 #765 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.
Waste Sources
The following are sources of wastewater produced from the manufacture
of hydrogen cyanide by the Andrussow process:
554
-------
HBIIWB
T j
OQU) VQIT
HRTOR
Ul
Ul
Ul
1__ -
REMHOR
MtmuA
ABSORDGR
I1CH ABSORPTICH
oxanuATKH
1
HCN PflOOKT
USB) POM TIE
MftHUFACTlffiE Of
oniot pnxucis
OR RKVCISJ,
MENRBCKXQ),
A BLEHJ IS SBTf
TOfllE WSTB
OTEMMNT IUKT.
ne DisnuATMH BOITDH
IS EIHER RBCVCUD (A
PURGE 18 DISQIMOED) OR
SEHT
FACILIW.
Figure 17-1. General process flow diagram for production of hydrogen cyanide
by the Andrussow process.
-------
TfiBLE 17-3. WKTER USM3 IN HYDROGEN CXKNIDE - SNDBI3SSOW PROCESS
Plant
Water Usage, (m3/kkg of HCN)
Total Consurrption Noncontact Cooling
#782
#765
(1)
29.5
58.3
18.9
8.00
(1)
Detail water usage (m3/kkg) at Plant #782 is:
Noncontact cooling
Direct process contact
Indirect process contact
(pumps, seals, leaks/
spills, etc.)
Maintenance, e.g. cleaning
and work area washdown
Noncontact ancillary uses
(boilers, utilities, etc.)
Exported steam
18.9
7.45
0.71
0.31
0.67
1.44
556
-------
A. Distillation Bottoms
The wastewater 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
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.
B. Scrubber Streams
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 be used for the
manufacture of other products in which case nothing is discharged
to the treatment plant.
C. Other Wastewater
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 wastewater
treatment facility. During shutdown, the equipment is drained to
avoid freeze-up and the resulting wastewater is discharged to the
treatment facility.
The quantity of wastewater 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 #765 is not recylced. 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 mVkkg of HCN.
Description of_ Plants Visited and Sampled
Screening
Plant #765 was visited and the wastewater sampled during the screening
phase of the program. The combined wastes consist of distillation
bottoms, ammonia recovery purge liquor, tank car washings, leaks,
557
-------
TABLE 17-4. WASTE FLOW DATA FOR HCN PRODUCTION BY THE ANDHUSSOW
PROCESS
Plant Total waste going to the treatment facility (m3/kkg)
#765 57
#782 9.9*
*
The breakdown and flow of the different waste streams comprising the total
is given below:
Source Chit Flow(m /kkg)
Recovery and purification 6.3
Pump seal quenches 0.58
Flare stack flushes 0.09
Sample hoods 0.02
NH3 stripper caustic 0.24
Steam condensate from NH3 stripper 0.90
Freeze protection 0.06
Washdowns and cleanup 0.25
Boiler blowdown and condensate 1.48
558
-------
spills and equipment clean out, purge from the noncontact cooling
water system, and stormwater runoff. These combined wastes are
commingled with the other cyanide product wastewaters and sent to the
alkaline chlorination treatment facility. The first unit of the
treatment facility is a trench where the pH of the wastewater is
raised to tne 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
wastewater is sent to two 8-hour retention ponds. Chlorination is
accomplished by adding sodium hypochlorite at the pond entrance. The
chlorinated wastewater 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 use<5 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 BOD5, 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. We believe that thallium is not
contributed by the hydrogen cyanide manufacturing process.
Verification
Plant 1765 was sampled again in the verification phase. One
additional stream of hydrogen cyanide wastewater was sampled in the
verification phase at a point upstream of mixing with other cyanide
produce wastewater. 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 |782. The wastewater from the hydrogen cyanide plant mainly
consists of blowdown from the distillation column which is combined
with a portion of the other product wastewater and sent to an ammonia
stripper. Effluent from the ammonia stripper is mixed with the rest
of the process wastewater from other products and sent to a single
stage biological system. The primary treatment facility consists of
grit removal, oil skimmers, and pH adjustment. The effluent from
primary treatment goes through an oil 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
559
-------
HYDROGEN CYANIDE
WASTE WATER
OTHER CYANIDE PRODUCT
WASTE HATER
DILUTE CAUSTIC
±
POND
SODIUM
HYPOCHLORITE
CAUSTIC
—i
t ''
e
LEGEND
Waste streams sampled
Waste stream was
sampled in the
verification program
since it is free from
other cyanide wastes.
FINAL TREATED
EFFLUENT
CHLORINE
Figure 17-2. General waste water treatment process flow diagram at plant #765
showing the sampling points. (Hydrogen cyanide manufacture.)
-------
TSBLE 17-5. FLOW AND POLLUTANT DATA OF THE PAW AND TREATED WASTE
STREAMS OF PLANT #765 PRODUCING HYDROGEN CYANIDF BY
ANDRDSSOW PROCESS
Stream
Description
Unit
NH3-N
Total
Cyanide .
Thallium
V 1
#2
Influent
to
Treatment
57(2) 7.8 4.4 <3) 107
6.1
(3)
,'028 0.0016
Treatment
(Alkaline
Chlorination)
Effluent
57
(2)
3.5 2.0 t°' 0.36
0.020
(3)
.010 0.00057
(1)
Unit Load *
in kg/kkg
Unit Flow
(57 ,i)
\ kkg/
pollutant
concentration
in mg/1
x
( fcg/m \
I 1000 mg/y
(2) The stream is a coraraingled waste water. The flow 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.
561
-------
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 wastewaters 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 wastewater 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
wastewater from the hydrogen cyanide plant is the waste from the
recovery and purification operation and represents a unit flow of 6.3
m3/kkg of HCN. The total wastewater flow going to the treatment
facility from the hydrogen cyanide plant is 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.
Toxic Pollutant Concentrations
Total cyanide and thallium were the toxic pollutants detected in the
raw waste from Plant #765 which was sampled in the screening phase.
We believe that thallium in the wastewater is not contributed from the
hydrogen cyanide process.
The HCN wastewater at Plant #765 is mixed with other product
wastewater and the combined flow was sampled upstream of the treatment
system. It is probable that thallium is contributed from these other
product wastewaters.
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 wastewater 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 wastewater in the screening and
verification were:
562
-------
TABLE 17-6. FLOW AND POLLUTANT CONCENTRATION DATA OF THE SAMPLED WASTE
STREAMS FOR PLANT #765 PRODUCING HYDROGEN CYANIDE
Stream
Description
#1 Raw HCN
Unit Flow TSS Load
(m3/kkg of HCN) {kgA*g of HCN)
waste 57 1.1
#2 Influent to 57 f 1 > NA
NH3-N Load
(kg/kkg of HCN)
27
11(2)
Ctf(A)
(kg/kkg of HCN)
0.82
0.39(2)
CN(T)
(kg/kkg of
1.6
1.6
HCN)
the pond
#3 Treated
effluent from
the final pond
57(2),(3)
1.9(2)
7.1(2)
NA
0.00015
en
cn
CO
(1* The stream is comingled wastewater. The flow given is the amount contributed
by the HCN process.
(2) The pollutant load was calculated by apportioning the mass emitted between
the waste streams on the basis of measured flows. This is clearly a very
approximate process and the results must be used with caution.
(3) The addition or loss of water from rainfall, addition of chemicals and
evaporation has not been estimated.
= Not Available
-------
DISTILLATION
BOTTOM PURGE
OTHER PRODUCT
'WASTE WATERS
OTHER PRODUCT WASTE WATER
PRIMARY
OIL SEPARATOR
pH
ADJUSTMENT
COMPOSITING
POND
SECONDARY
OIL SEPARATOR
AERATED
LAGOON
FLOCCULATOR
CLARIFIER
DISCHARGE
SETTLING
POND
SURFACE DRAINS
EQUALIZATION
POND
1
TRICKLING
FILTER
LEGEND
SAMPLING POINTS
Figure i?~3.
Genera! waste water treatment process flow diagram at plant #782
showing sampling points. (Hydrogen cyanide manufacture.)
564
-------
Table 17-7. FLOW AND POLLUTANT CONCENTRATION DATA OF THE SAMPLED WASTE
STREAMS FOR PLANT #782 PRODUCING HYDROGEN CYANIDE
Stream
No.
Waste
Stream
Description
Flow CMC]
m3/day
C) CN(A) NH3-N
(mg/1)
TSS
1 Distillation(1>
bottom purge
2 Ammonia stripper'3^
influent
3 Ammonia stripper^3'
effluent
4 Influent tot3J
primary treatment
facility
5 Final treated^3)
effluent
(6.3)(2> 71
5400
5400
6400
NA
167
51
31
2.2
62
145
41
7.0
1.7
886
410
1380
5.6
24
76
162
110
74
(1) - The total waste is composed of the blowdown from 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 samples for the three waste stream sources.
(2) - The value given is the total unit flow in m3/kkg of HCN for the three
purge streams.
(3) - The stream is combined wastewater- It includes the waste effluents
from hydrogen cyanide and other products.
565
-------
TABLE 17-8. UNIT FLOW AND UNIT POLLUTANT LOADING FOR RAW AND
TREATED WASTE EFFLUENTS AT PLANT #782
Unit
Stream Unit
Flow
(m3Akg)
Pollutant Loading
Total
Cyanide
CN(T)
(kg/kkg) <1)
Free Ammonia-
Cyanide N
CN/B, NH,-N
(A; 3
Total
Suspended
Solids
TSS
Process raw
wastewater
(distillation
botton purge
6.3
0.45
0.39
5.6
(1) Unit pollutant load = unit flow
(m3/kkg)
0.15
Process 6.3^J 0.014 0.011
wastewater
treated
effluent
Total HCN 9.9<3> 0.022 0.017
wastewater
treated
effluent (2>
0.035 0.47
0.055 0.74
pollutant concentration x kg/3
(in mg/1 from Table 17-7) 1000 mg/1
(2) The pollutant load was calculated by apportioning the mass emitted from
the total treated effluent (which includes other product wastewater} on
the basis of measured flow contributed by the HCN process. This is clearly
an approximate process and the results must be used with caution.
(3) The wastewater flow consists of direct process contract and noncontact
effluent from the HCN plant going to the treatment system.
566
-------
Maximum Raw Waste Concentration Observed
Ug/1)
Screening Verification
Pollutant Plant #765 Plants #765, #782
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. A total of nine days
of sampling was conducted at Plants #765 (sampled twice) and #782.
Thirteen wastewater sampling points were involved which included the
raw wastewater, combined wastewater, and combined treated effluent
streams. The evaluation of the toxic metal and toxic organic
pollutant content of these process steams 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 mVday (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.
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 Ibs).
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.
567
-------
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 verification phases. The overall average
polluant 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
Toxic Pollutants of Concern
The toxic pollutants of concern in the HCN raw waste are free (or
oxidizable) cyanide and total cyanide. Free (or oxidizable) cyanide
is cyanide amenable to chlorination, and is designated in the
regulation as Cyanide A. No organic toxic pollutants of significance
were found in the raw waste of the sampled plants.
Process Modifications and Technology Transfer Options
Process modifications have not been identified for the subcategory.
Best Management Practices
No best management practices have been identified for the subcategory.
Prevailing Control and Treatment Practices
Out of a total of seven plants currently producing hydrogen cyanide by
the Andrussow Process, 308 data are available for only two. The
production at these two plants constitutes more than 70 percent of the
total subcategory production. Since the two plants produce a
significant amount of the total subcategory production, their
wastewater 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 recylced since low cost cold water is readily
available at the site. The wastewater consisting of scrubber purge,
absorption water, and plant runoff is mixed with other cyanide product
wastewaters and sent to an alkaline chlorination system. The pH of
the wastewater is raised to about 10 with dilute caustic in an inlet
trench and then it is discharged to one of two 8-hour ponds as sodium
hypochlorite is added to oxidize the cyanide to cyanate. The
chlorinated wastewater is transferred to a small pond equipped with
568
-------
TABLE 17-9. SUMMARY OF POLLUTANT RAW WASTE LOADING FOUND IN SCREENING AND
VERIFICATION SAMPLING
SUBCATEGORY
HYDROOT3 CYANIDE
Average Daily Pollutant Loading and Concentrations at Plants Sampled
kgAkg of HCN
(mg/D
Pollutant
TOXIC
Free Cyanide
Total Cyanide
#765 (s)
NA
6.1
(110)
t 765 (v)
0.82
(14)
1.6
(29)
# 782 (v)
0.39
(62)
0.45
(71)
Overall
Average
0.61
2.7
Conventional
and Nonconventional
TSS
NH3-N
(S) =
(V) =
NA
4.4
(78)
Sampled in
Sampled in
2.0
(35)
37
(480)
screening phase
verification phase
0.15 1.1
(24)
5.6 12
(890)
569
-------
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
wastewater from the HCN plant consists mainly of distillation column
blowdown and is combined with other cyanide product wastewater and
sent to an ammonia stripper. The effluent from the stripper combines
with other product wastewaters and is treated by means of a grit
chamber, an oil separator, a compositing pond, a second oil separator,
an aerated lagoon, a flocculator and a final clarifier. The overflow
from the clarifier is sent to a final settling basin before discharge.
The runoff 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
wastewater A general block diagram of the treatment system is shown in
Figure 17-3.
Advanced Treatment Technologies
The three pollutants of concern in hydrogen cyanide plant effluents
are cyanide, ammonia, and chlorine. The treatment technologies for
cyanide removal include alkaline chlorination, biological treatment,
ozonation, wet air oxidation, electrolytic decomposition,
acidification, activated carbon, permanganate oxidation, lime reaction
with sulfur, radiation, evaporative recovery/ catalytic oxidation and
ion exchange. Except for alkaline chlorination and biological
treatment, the remaining treatment technologies are not effective or
advantageous for one or more of the following reasons:
A. The technology has low cyanide removal efficiency.
B. The technology cannot treat wastewater 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.
570
-------
Selection of Appropriate Technology and Equipment
Technologies for Different Treatment Levels
A. 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.
B. 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 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 related. The general flow
diagram of the treatment process is given in Figure 17-5.
Addition of chlorine to remove ammonia ("break-point
chlorination") is not intended for either Level 1 or Level 2.
Break-point chlorination for ammonia removal generally is very
expensive. In this industry, ammonia control, where necessary,
should be accomplished by steam or air stripping and recovery and
reuse of the ammonia. The achievable concentrations of ammonia
in the final effluent from a hydrogen cyanide plant with cyanide
removal after ammonia recovery are presented below for guidance.
Equipment for Different Treatment Levels
A. Equipment Functions
In level 17 the raw wastewater 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 wastewater is chlorinated further in a second
tank which is equipped with automatic pH control. The final
effluent is neutralized to pH 6-10.5 before discharge. In Level
2, cyanide control is accomplished, using the same equipment as in
Level 1. To remove excess chlorine before discharge, sulfur
dioxide is fed into the chlorinated wastewater by a modified gas
chlorinator, with oxidation-reduction potential control.
Solutions of sulfur dioxide in water are acidic, so the addition
of sulfur dioxide to react with the residual chlorine will tend
to neutralize the alkalinity of the wastewater, thus reducing the
amount of suIfuric acid that must be added for final pH
adjustment. As in Level 1, the effluent is then adjusted to pH
6-10.5 before discharge.
571
-------
m
-4
to
RSW
WASTE-
WWER
CAUSTIC 9DDA
CHLORINE
KJUJJNG AND 13F
STAGE ALKALINE
CHLQRBATIDN
SBOCMD 9EAGE
ALKALINE CHDORIHATICH
pH
~ 1
Includes How monitoring, pH monitoring and sampler,
Figure 17-4. Level 1 waste water treatment for hydrogen cyanide subcategory.
-------
B. 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
specific hazards when handled by conventional corrosion-resistant
feeding equipment. Chlorine and sulfur dioxide are received in
one-ton containers as liquified 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.
C. Separation and Removal of Solids
Since few solids are produced in the treatment process, there is
no significant sludge disposal problem.
D. 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
permits.
Treatment Cost Estimates
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 wastewater 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.
A. Wastewater Flow
The unit process wastewater flows for the two plants visited in
this study are 6.3 mVkkg of HCN (Plant $782) and 57 mVkkg of
HCN (Plant 1765). The difference results from the different
absorption water discharge practices at the two plants. 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
573
-------
ui
-J
CAUSTIC
SODA
HAW
WASTE WATEH
-a
HOLDING AND 1ST STAGE
ALKALINE CHLORINATION
SULFUR
DIOXIDE
-----9,
SECOND STAGE
ALKALINE CHLORINATION
II ADJUSTMENT
IFFLUENT
Includes flow monitoring, pH monitoring and sampler,
ORP = Oxidation Reduction Potential Control
Figure 17-5. Level 2 waste water treatment for hydrogen cyanide subcategory.
-------
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.
B. Production
For wastewater treatment cost estimates, three production levels
were selected for the model plant. These are 31,800, 50,900 and
63,600 kkg/yr.
C. Wastewater 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 product HCN and 12 kg of NH3/kkg of product HCN
(Table 17-9) developed from the screening and verification
results were used for the model plant raw waste loads.
D. Chemicals Used
At the BPT level of treatment, alkaline chlorination requires 33
kg of chlorine (most to react with ammonia) and 5.0 kg of caustic
per kkg of product HCN. For BAT treatment, 9.0 kg of SOZ per kkg
of product HCN is used for dechlorination in addition to the
chemicals used for BPT treatment.
E. Solids Generated
Few, if any, solids are produced in treating HCN produciton
wastes.
The estimated costs, for the three model plants at different
production levels are given in Tables 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.
The costs shown in Tables 17-10, 17-11, and 17-12 at each level
of treatment correspond to BPT (Level 1} with incremental costs
to meet the more stringent BAT requirements.
Basis for Regulations
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 are combined with the waste from an organic cyanide process and
sent to a biological treatment system to reduce organic and cyanide
pollutants. Four of the other six HCN producers (using the Andrussow
Process) use alkaline chlorination for treatment of raw waste
575
-------
effluents. There is no available information concerning the treatment
practices at the other two plants.
Basis for BPT Limitations
A. Technology Basis
The predominant treatment practice for raw waste effluent in the
HCN subcategory is alkaline chlorination. The Agency is
therefore promulgating BPT effluent limitations based on
alkaliine chlorination to destroy cyanide amenable to treatment
by chlorination (free cyanide or Cyanide A).
B, Flow Basis
The effluent limitations are based on the high flow (57 mVkkg of
HCN) model, that is, with no recycle of absorber water. A low
flow basis (7 m3/kkg of HCN based on tne flow of Plant #782) 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.
C. Selection ol Pollutants to be Regulated
The selection of pollutants on which specific limitations are
promulgated is based on the evaluation of raw waste composition
as determined during the screening and verification programs.
Plant #765 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 promulgated for this parameter. Thallium is best controlled
by management practices developed by the permitting authority on
a case-by-case basis.
D. Basis of Pollutant Limitations
1. Conventional and nonconventional parameters
a. pH
The treated effluent is to be controlled within the pH
range of 6.0 to 10.5. This limitation is based on the
576
-------
TABLE 17-10. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Hydrogen Cyanide
31,800 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal
Contractors 0 & P b....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
($)
BPT
6,000
850,000
20,000
876,000
131,400
1,007,400
201,480
1,208,880
120,888
1,329,768
6,000
1,335,768
TOTAL OPERATION AND
MAINTENANCE COST
84,000
9,000
296,000
132,977
40,073
0
15,000
577,050
BAT
0
125,000
0
125,000
18,750
143,750
28,750
172,500
17,250
189,750
0
189,750
14,000
3,100
97,000
18,975
5f693
0
7,500
146,268
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
216,353
793,403
30,872
177,140
Represents the incremental cost above that for BPT treatment
Overhead and Profit
577
-------
TABLE 17-11. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Hydrogen Cyanide
50,900 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal
Contractor's 0 & P b....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
BPT
6,000
1,340,000
20,000
1,366,000
204,900
1,570,900
314,180
1,885,080
188,508
2,073,588
6,000
2,079,588
TOTAL OPERATION AND
MAINTENANCE COST
84,000
9,800
476,000
207,359
62,388
0
15,000
854,546
(S)
BAT
0
135,000
0
135,000
20,250
155,250
31,050
186,300
18,630
204,930
0
204,930
14,000
3,100
154,000
20,493
6,148
0
7,500
205,241
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
337,373
1,191,919
33,342
238,583
Represents the incremental cost above that for BPT treatment
Overhead and Profit
578
-------
TABLE 17-12
MODEL PLANT TREATMENT COSTS
Subcategory
Production
Hydrogen Cyanide
63,600 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal
Contractor's 0 & P b....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
($)
BPT
9,000
1,600,000
20,000
1,629,000
244,350
1,873,350
374,670
2,248,020
224,802
2,472,822
9,000
2,481,822
TOTAL OPERATION AND
MAINTENANCE COST
84,000
11,500
592,000
247,282
74,455
0
15,000
1,024,237
BAT
0
190,000
0
190,000
28,500
218,500
43,700
262,200
26,220
288,420
0
288,420
14,000
4,600
191,000
28,842
8,653
0
7,500
254,595
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
402,328
1,426,565
46,926
301,521
Represents the incremental cost above that for BPT treatment
Overhead and Profit
579
-------
TABLE 17-13 MODEL PLANT UNIT TREATMENT COSTS
Subcategory Hydrogen Cyanide
Annual Treatment Costs ($/kkg)
LEVEL OF TREATMENT
COST ITEM PRODUCTION
(kkg/yr)
Annual Operation
and Maintenance
Annual
Amortization
Total Annual
Cost
31
50
63
31
50
63
31
50
63
,800
,900
,600
,800
,900
,600
,800
,900
,600
BPT
18.
16.
16.
6.
6.
6.
24.
23.
22.
15
79
10
80
63
33
95
42
43
BAT*
4.
4.
4.
0.
0.
0.
5.
4.
4.
60
03
00
97
66
74
57
69
74
*Represents the incremental cost above BPT
580
-------
data presented in Appendix B of the proposed
Development Document (60) and the JRB Study (52).
TSS
The concentration of suspended solids found during
sampling of the raw wastewater 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 sol ids. The
maximum concentration of 35 mg/1 of TSS was found in
the raw waste during screening and verification
sampling (Table 17-9). However, 35 mg/1 was accepted
as the long term average concentration on the basis of
historical data shown in Table A-13a. The variability
factors estimated for ammonia (Table 17-14) are used to
calculated the concentration bases and effluent limit.
The TSS maximum 30-day average concentration is:
(35 mg/1) (1.6) = 56 mg/1
and the TSS maximum 30-day average effluent limit
is given by:
(56 mg/1 (57 mVkkg) (kq/m3) =3.2 kg/kkg
(1000 mg/1)
The TSS 24-hour maximum concentration is given by:
(35 mg/1) (4.2) = 150 mg/1
and the TSS 24-hour maximum effluent limit is given
by:
(150 mg/1) (57 mVkkg) (kg/in*) =8.6 kg/kkg
(1000 mg/1)
Ammon i a
The Agency is not promulgating effluent limitations for
ammonia in the discharge for the reasons described
above. However, we are providing guidance for use by
permit writers and POTW's in cases where the discharge
of ammonia is of concern. This guidance is based on
performance data from plant #765. Plant #765 has
recently submitted one and one-half years of monitoring
data on the treated effluent for ammonia. Plant #765
uses a proprietary process for removal of ammonia,
however, the same performance can be achieved by steam
stripping. The variability factors for the daily data
and 30-day averages were calculated from the long-term
data as shown in Table 17-14. The long-term average
581
-------
TABLE 17-14. STATISTICAL ANALYSIS OF HISTORICAL EFFLUENT
MONITORING DATA ON FREE CYANIDE AND AMM3NIA
FROM PLANT #765
Pollutant
Free
Cyanide
Ammonia
Time
Period
9-79
to
7-80
1-79
to
7-80
Monitoring
Frequency
Daily
30-day
average
Daily
30-day
average
N
No.
512
11
540
11
X
Mean
(mg/1)
0.20
0.21
45
52
S(1)
Std
Dev
(mg/1)
0.36
0.087
42
19
cv(2)
Coeff . of
Variation
1.8
0.40
0.93
0.36
VF
Variability
Factor
8.0<3)
1.7(4)
4.2 (3)
1.6(4)
For free cyanide and ammonia, the long-term monitoring data were screened of
outliers. In the first place, values recorded as zero were interpreted to
mean "inability to measure pollutant" and were rejected prior to the statisti-
cal 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-3anin)/s)
for extreme values as outliers. Screening was performed on a month-by-month
basis, and any datum with a calculated t value exceeding the 99% confidence
limits from the t distribution was concluded to be an outlier. Given rejec-
tion of a value, recomputation of statistical measures for that month was
performed.
(1) Arithmetic standard deviation, S
where S2 = Z (X-X)2/(N-1)
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) = S1 (2.33 - O.SS'l
T-_TJ_ _ ._ /f* * \ S T«» / n i /firi\ &
Where (S1) = In (1 + (CV)
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.
582
-------
concentration of 52 mg/1 was used as the basis for
guidance. The estimated variability factors and model
plant flow rate were used in calculating the
concentration bases as follows:
The maximum 30-day average concentration basis for
ammonia is given by:
(52mg/l) (1.6) = 83 mg/1
The 24-hour maximum concentration is given by:
(52 mg/1) (4.2) = 220 n\g/l
Toxic Pollutants
The toxic pollutants promulgated for regulation are free
cyanide and total cyanide.
a. Free Cyanide
Plant #765 practices alkaline ch1orination and has
recently submitted almost one year 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 17-14. The long-term average
concentration of 0.21 mg/1 (Table 17-14) was used as
the basis for the limitations. The estimated
variability factors and model plant flow rate were used
in calculating the concentration bases and effluent
limitations.
The maximum 30-day average concentration basis for free
cyanide is given by:
(0.21 mg/1) (1.7) = 0.36 mg/1
The 24-hour maximum concentration is given by:
(0.21 mg/1) (8.0) = 1.7 mg/1
The maximum 30-day average effluent limitation is
calculated by:
(0.36 mg/1 (57 mVkkg) (kq/m3) * 0.021 kg/kkg
(1000 mg/1)
583
-------
The 24-hour maximum effluent limitation is given by:
(1.7 mg/1) (57 mVkkg) (kg/m3) = 0.10 kg/kkg
(1000 mg/1)
Total Cyanide
The variability factors for total cyanide for daily
data and 30-day averages were estimated from a 28-day
study conducted by Plant #765 and are given in Table
17-15. In the case of the 30-day average variability
factor, it was necessary to apply a different approach
which requires estimation of the 30-day average
arithmetic standard deviation. This approach, also
used in the Treatability Study (61), is necessary when
30-day average data is not available. The approach for
determining the estimated 30-day average standard
deviation is as follows:
Estimated 30-day Average Standard Deviation =
24-hour Measurement Standard Deviation
30
The limitations for total cyanide are derived from the
average unit effluent load (0.19 kg/kkg given in Table
17-15), variability . factors estimated from the 28-day
test, and model plant flow of 57 m3/kkg.
The maximum 30-day average effluent for total cyanide
limitation is calculated by:
(0.19 kg/kkg) (1,2) = 0.23 kg/kkg
The total cyanide 24-hour maximum effluent limitation
is given by:
(0.19 kg/kkg) (3.4) = 0.65 kg/kkg
The total cyanide maximum 30-day average concentration
basis is:
(0.23 kg/kkg) 1 (1000 mg/1) = 4.0 mg/1
(57 mVkkg) (kg/m3)
The total cyanide 24-hour maximum concentration basis
is:
(0.65 kg/kkg) 1 (1000 mg/1) = 11 mg/1
(57 mVkkg) (kg/m3)
The final effluent limitations for Hydrogen Cyanide
produced by the Andrussow Process are summarized in
584.
-------
TABLE 17-15. STATISTICAL ANALYSIS OF THE 28-DAY EFFLUENT
SAMPLING RESULTS CN TOTAL CYANIDE
FBQM PLANT #765
Total Cyanide
Daily Data
No. of points 25
Average Unit load 0.192
kg/kkg of HCN
Std. Deviation(S) 0.128
Std. Deviation (Sf) 0.61
Variability Factor 3,44
30-Day Average Data
The Standard error
of the mean (A) 0.023
Coefficient of
variation for the mean(CV) 0.119
Variability factor 1.19
Variability Factor Ratio
V.F.R. 2.9
585
-------
Table 17-16 for toxic, conventional, and
nonconventional pollutants.
Basis for BCT Limitations
While EPA has not yet proposed or promulgated a revised BCT
methodology in response to the American Paper Institute v. EPA
decision mentioned in Section 3, EPA is promulgating BCT limitations
for this subcategory. These limits are identical to those for BPT.
EPA is not promulgating any more stringent limitations since we have
identified no technology option which would remove significant
additional amounts of conventional pollutants. The dechlorination
technology added to BPT for BAT does not remove additional
conventional pollutants. As BPT is the minimal level of control
required by law, no possible application of the BCT cost tests could
result in BCT limitations lower than those promulgated in this
regulation. Accordingly, there is no need to wait until EPA revises
the BCT methodology before promulgating BCT limitations.
Basis for BAT Limitations
The Agency considered different advanced level technologies and their
cost effectiveness relative to the base level system (BPT) for the
removal of toxic, conventional, and nonconventional polluants. For
BAT, the Agency is utilizing Level 2 technology which includes
dechlorination before final discharge.
The Agency also considered break point chlorination for essentially
complete destruction of cyanide and ammonia removal. 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.
A. Technology Basis
For BAT, the Agency is promulgating limitations based on BPT with
the addition of dechlorination (Figure 17-5, Level 2). Control
of chlorine in the discharge is 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
chlor-alkali industry (Appendix A). 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.
B. Flow Basis
The BPT effluent discharge rate of 57 mVkkg of HCN has been used
as the basis for the BAT model plant.
586
-------
TABLE 17-16. EFFLUENT LIMITATIONS
HYDROGEN CYANIDE (ANDRUSSOW PROCESS)
Best Practicable Control Technology Currently Available
Wastewater Flow: 57 m3/kkg of HCN
Subcategory
Performance
Pollutant (mg/1)
Convent iona I/
Nonconventional
Pollutants
Total Suspended
Solids 35(1>
Ammonia-N 5 2 ( '
Toxic pollutants:
Free Cyanide*4' 0.2l(2)
Total Cyanide <4> 3.3<3>
Daily
Variability
Factor
30-day
Variability
Factor
4.2/1.6
4.2/1.6
8.0/1.7
3.4/1.2
Concentration
Basis (mg/1)
Max.
30-day 24-hr.
Avg. Max.
56 150
83 220
0.36 1.7
4.0 11
Effluent Limit
(kgAkg)
Max.
30-day 24-hr .
Avg. Max.
3.2 9.6
— (5) __(5)
0.021 0.10
0.23 0.65
(1) Average effluent concentration from monitoring data (Table A-13a).
(2) Average based on recently received long-term monitoring data submitted by
Plant #765 (Table 17-14).
(3) Average based on a 28-day sampling data submitted by Plant #765
(Table 17-15).
(4) Also applicable for PSNS limitations.
(5) No effluent limitation has been established.
587
-------
C. Selection of Polluants to be Regulated
For the BAT regulation, the Agency has selected chlorine in
addition to the pollutants identified in BPT.
D. Basis of Pollutant Limitations
1, Nonconventional Pollutants
The nonconventional pollutant promulgated for regulation is
total residual chlorine. For total residual chlorine the
BAT regulation is based on long-term monitoring data from
the chlor-alkali industry given in Appendix A. The
long-term average concentration is 0.64 mg/1. The daily and
30-day variability factors are 2.3 and 1.4, respectively.
The 24-hour maximum concentration is:
{0.64 mg/1) (2.3) = 1.5 mg/1
The maximum 30-day average concentration is:
{0.64 mg/1) (1.4) = 0.90 mg/1
The determination of load limitations for total residual
chlorine (kg/kkg) was calculated based on the unit flow
rate of 57 m3/kkg, thus the 24-hour maximum limit is given by
(1.5 mg/1) (57 mVkkg) (kg/in*) = 0.086 kg/kkg
(1000 mg/1)
The maximum 30-day average was calculated similarly,
(0.9 mg/1) (57 mVkkg) (kg/in*) = 0.051 kg/kkg
(1000 mg/1)
2. Toxic Pollutants
The Agency has selected the same limitations for free
cyanide and total cyanide as those promulgated 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.
Basis for New Source Performance Standards
Level 2 treatment technology (also promulgated for BAT) was selected
as the basis for NSPS limitations. The pollutants to be controller
for NSPS are pH, total suspended solids, total residual chlorine, free
cyanide, and total cyanide. The NSPS limitations are given in Table
17-18. For pH, the NSPS limitation is 6.0 to 10.5
588
-------
TABLE 17-17. EFFLUENT LIMITATIONS
HYDROGEN CYANIDE (ANDRUSSOW PROCESS)
Best Available Technology
Waste Water Flow: 57 m3/kkg of HCN
Daily Concentration Effluent Limit
Variability Basis (mg/1) (kg/kkg)
^ Var^lity 3^ 2«* \°^
Factor Avg Max Avg
Nonconventional
pollutants :
24-hr
Max
(4)
Total Residual
Chlorine(2) Q.64 2.3/1.4 0.90 1.5 0.051 0.086
Toxic pollutants:
Free
Total
Cyanide
Cyanide
-U
(2)
0.21
3.3 <3>
8.0/1.
3.4/1.
7
2
0
4
.36
.0
1.7
11
0
0
.021
.23
0.10
0.65
(1) Average based on recently received long-term monitoring data submitted by
Plant #765 (Table 17-14).
(2) Average based on long-term monitoring data from, the chlor-alkali industry
given in Appendix A.
(3) Average based on a 28-day sampling data submitted by Plant #765
(Table 17-15).
(4) No effluent limitation has been established.
589
-------
Basis for Pretreatment Standards
Pretreatment is required because NSPS provides better removal of free
and total cyanide than is achieved by a well-operated POTW with
secondary treatment installed and therefore these polutants may pass
through a POTW in the absence of pretreatment.
A. Existing Sources
The Agency is excluding this subcategory from Pretreatment
Standards for Existing Sources (PSES) under the provisions of
paragraph 8(b)(ii) of the Settlement Agreement because the
concentrations of toxic pollutants in the effluent to POTW from
the one existing indirect discharger are below treatable levels.
B. New Sources
For Pretreatment Standards for New Sources (PSNS), the Agency is
promulgating limitations based on .NSPS excluding dechlorination
of the final plant effluent. Dechlorination is not required
because influent to a POTW is often chlorinated. The pollutants
to be regulated are free cyanide and total cyanide as summarized
in Table 17-16.
590
-------
TABLE 17-18 . EFFLUENT LIMITATIONS
HXDFOGEN CYANIDE (ANDRUSSOW PROCESS)
New Source Performance Standards
Waste Water Flew: 57 m3/kkg of HCN
Daily Concentration Effluent Limit
Variability Basis £mg/l) (kg/kkg)
Pollutant Subcategory -Factor
Performance 30-day Max Max
(mg/1) Variability 30-day 24-hr 30-day 24-hr
Factor Av9 Max Av9 Max
Conventional and
nonconventional pollutants:
Total Suspended
Solids, TSS 35 4.2/1-6 56 150 3.2 8.6
Total Residual
Chlorine ,
Anmonia-N
Toxic pollutants:
Free Cyanide
Total Cyanide
0.64
52
0.21
3.3
2.3/1.4
4.2/1.6
8.0/1.7
3.4/1.2
0.90
83
0.36
4.0
1.5
220
1.7
11
0.051
„ (*)
0.021
0.23
0.086
__ (*)
0.10
0.65
* No effluent limitation has been established.
591
-------
-------
SECTION 18
SODIUM DICHROMATE INDUSTRY
Industry Profile
General Description
Most of the sodium dichromate produced is used in the chromic acid and
chrome 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 prior to promulgation of this new
regulation is given in Table 18-2.
Subcategorization
Subcategoration in the sodium dichromate industry is established on
the basis of the manufactured product. This follows from the
requirement that the effluent limitations are to be tied to units of
production. Furthermore, for the two plants discharging process
wastewater in the dichromate subcategory, the characteristics of the
wastewater are similar and therefore the saine treatment technology can
be applied. Subcategorization is discussed in more detail in Section
4 of this report.
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 + 8NaaC03 + 702 = 8Na2Cr04 + 2Fe203 + 8C02 (1)
2Na2Cr04 + H2S04 = Na2Cr207 + HZ0 + Na2S04 (2)
Chromite ore is a chromium iron oxide containing ferrous chromite
T~FeCr204 or FeOCr203). Small amounts of alumina, 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 alumina present in the thickener overflow is hydrolyzed
593
-------
TAHtZ 18-L
SDBC3OD33R5r .EHCPIIZ DMA
SCOIUM DKSBOMWE
Total sabcategory 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:
Average production
Median production
Average capacity utilization
Plant age range:
Maxbnin
Haste water flow range:
Vblune per unit product:
Minium
Maxinun
140,000 kkg/year
136,500 kkg/year
3
3
NA
112,000 khg/year
NA,
82 percent
20,700 kkg/year
66,800 Jdcg/year
37,300 kkg/year
24,800 kkg/year
77 percent
7 years
28 years
455 cubic maters/day
720 cubic meters/day
4 cubic metersAkg
8 cubic meters/Wag
Sources of data ara Stanford Besearch Institute, Directory of Chemical
Producers, U.SJU* 1977, U.S. Deportment of Qatmerce, Current Industrial
Deports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Beport, "Preliminary Economic Assessment of Effluent Limitations in the
Tpr"n3anic Oimijcal Industry," June, 1578 and "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards for ttie Inorganic Chsnicals Industry,1
, 1980.
Mot Available
594
-------
TABLE 18-2. STATUS OF REGULATIONS - EFFLUENT LIMITATION GUIDELINES
SUBCATEGORY SODIUM DICHROMATE
SUBPART p_ (40 CFR 415.170, 3/12/74)
STANDARDS
BPCTCA
Max. f 1 )
Product kg/kkg
Process Parameters (mg/1)
NaaC^Oy TSS 0.44
(52)
Avg. ( 2 )
kg/kkg
(mg/1)
0.22
(26)
BATEA* NSPS
Max. Avg. Max.
kg/kkg kg/kkg kg/kkg
(mg/1) (mg/1) (mg/1)
No discharge 0.30
of pwwp
Avg.
kg/kkg
(mg/1)
0. 15
Cr
+6
0.0090t4) 0.00050
(0.11) (0.060)
Cr(T)
0.0088
(1.0)
0.0044
(0.50)
No discharge
of pwwp3
No discharge
of pwwp-*
0.0090<4> 0.00050
0.0088 0.0044
Section 415.173 was remanded and reserved (41 CFR 51606, November 23, 1976).
1 Max. = Maximum of any one day.
2 Avg. = Maximum, average of daiLy values for thirty consecutive days.
3 pwwp = Process wastewater pollutants.
4 The published value in 40 CFR 415.172 and 415.175 is incorrect and should be
0.0009 kgAkg.
595
-------
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 chromite,
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.
Water Use and Waste Source Characteristics
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.
Waste Sources
A. 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.
B. Noncontact Cooling Water and Cooling Tower Slowdown
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, the blowdown may contain chromates.
C. 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 wastewater treatment facility for
treatment.
596
-------
M.
.TCR
Ul
U3
-J
CRUSHER
wo
QUNDER
1
SODA ASH M BLEHCR
'
•
_„___„_„ KI1DS
1 i
1
flSCKCLE-^ ITMJBBER fc LBOI
FBOM f
PROCESS 1
1
f
1 SOUBBEH
1
1
10 tftSTO
RESI
BBC*
J-WIER
( nun |
ai
DUB
CLE
COOCUCTOMffl
MD BOILER
BUMXM)
1
1 TO WSTC
f 4
CRUDE 80OIUH
HCHW* soumc
OOOLD
.P™™ BYVtmn
i
PH
<
CBHTl
i
M
B
CT
ID SALES
OR USB
I
NCNocrracr
A COOJMJ |
T wa
un»
i
i
** f
UJZER
'
EVAPORATION ^Jffl&M CBWWHJGB
i
i
f OJBfUUC
f ACID
JCIDIFIER MWER
i
SCOIUH
PBCCUCT
x saoiuH *
RDCVAL HdOGIMS
1 |
CUdUH SUJDCZ 1
GKHBRAMD *H X
Hf>T|Jip f f
BH£MXW TOHV9IB TO SALES
g^WEN
HCMXK
^^- ooou
^ win
K
CKT
tNG
31 ^
LIQU1
pica
1
D
CT
•
PACKWnC
ID SALES
Figure 18-1. General process diagram for production of sodion dichromate
-------
TABIE 18-3.
WATER USAGE IN SODIUM DICHRCMATE SUBCATEX3QRY
Source Water usage at plants , fa3/kkg of Na2Qr207)
Plant #398 Plant #376 <3> Plant #493
Nonoontaet cooling 277
Nbnoantact ancillary 0.5
uses
Direct
process contact ^ 5.7^
Indirect process contact 0.9^ -^
(pumps, seals, leaks and
spills)
Maintenance, e.g. 0.5^
cleaning and work area
11.39 5.7
NA 3.12
7.83(4) 2.85
0.2
» 4.16
0.2
washdown
Air pollution control
2.5
(2)
NA
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 from the pcdmary
pond during 9 to 10 months of the year. There was no primary pond
effluent at the time of sampling and only 4,16 nrVkkg of the indirect
contact sources were being treated and discharged.
598
-------
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
wastewater, consisting of boiler and noncontact cooling tower
blowdown, is used to slurry the spent ore residue to the
wastewater treatment facility. At one plant, the only wastewater
resulting from process operations is the noncontact cooling
water, which is used on a once-through basis.
Description of Plants Visited and Sampled
Screening
Three sodium dichromate plants were visited and the wastewater streams
sampled. Plant I4JL3. was sampled ifn the screening ghase and Plants
#376 and #398 were sampled "in the verification phase.
At Plant #493, the wastewater 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
wastewaters 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.
Verification
At Plant #376, sodium sulfide is used for simultaneous chromate
reduction and precipitation. The wastewaters at this plant are
segrated 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
storaiwater runoff from both the solids disposal areas and the
production areas. The first wastewater 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
599
-------
RAW WASTE WATER
WATER
SLUDGE TO
LAND DISPOSAL
IMPORTED ACID
INDUSTRIAL WASTE
«. TREATED EFFLUENT
LEGEND
SAMPLING POINTS.
Figure 13-2. Genera? waste water treatment process flow diagram at Plant #493 showing
the sampling points. (Sodium dichromate manufacture).
600
-------
TABLE 18-4. FLCW 2ND POLLUTMTT CONCEWTRMTICN DATA OF THE SAMPLED
STREAMS FCR PIANT #493 PRODUCING SODIUM DICHRCMATE
Stream No.
1
2
Haste Stream
Disoription
Raw Waste
Water
Treated
Effluent
Unit Flow TSS Load
CntVkkg (fcg/kkg
of 1^20:207) of ift^C^OJ
4.25 183
28.91* 0.018
Cr+6 Load
(kg/kkg
of Na2Cr2a7>
3.5
0.0001
Chromium
Load
(fcg/kkg
of m&^rflj^
}
3.30
0.072
This value includes the flow Ii.au the sodiun dichrcrnate plant, imported
acid used for neutralization, and the water used for vashing the solids.
601
-------
settling pond where the suspended solids are settled and the overflow
discharged. A simplified flow diagram of the wastewater treatment
process is given in Figure 18-3. Table 18-5 gives the flow data and
pollutant emissions for the streams sampled.
Plant 1376 has implemented some changes in the process technology and
treatment system since the time of sampling. The dichromate
production was converted from a "high-lime" process to a "no-lime"
process, using only chromite ore and soda ash as the raw materials.
This requires additional treatment facilities to be installed for
removal of impurities from the product solution, which is treated with
acid and lime to remove alumium and vanadium/ respectively. At the
time of sampling, the data obtained from this plant was considered a
valid part of the data base for assesssing 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 the HZS gas production. This
has been confirmed in treatability studies 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.
Toxic Pollutant Concentrations and Loadings
Toxic pollutants detected in the raw wastes during sampling were as
follows:
Maximum Concentrations Observed (
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 <0.50 230*
Arsenic <10 <5
Selenium <5 !40**
* Found at one plant only
602
-------
COOLING TOWER
BLOWDOWN
SODIUM SULFIDE
WASTE
MUD
SLURRY
SURFACE
RUNOFF
TREATED
WATER
RECYCLED
SETTLING POND
TREATED EFFLUENT
LEGEND
WASTE STREAMS SAMPLED.
AT THE TIME OF SAMPLING,
ONLY SURFACE RUNOFF v
(STREAM #3) WAS BEING
TREATED 111 THE REACTOR.
Figure 18-3.
General waste water treatment process flow diagram at Plant #376
showing the sampling points. (Sodium dichromate manufacture)
-------
TABLE 18-5. FLOW AND POLLUTANT LOADING DATA OF THE SAMPLED VRSTE
STREAMS FCR PLANT #376 PRODUCING SODUM DICHPCMATE
Stream
No.
Average
Waste Stream Unit Flow TSS load
7]
0.407
NA
0.057
NA
< 0.00004
Chromium
load
(kg/kkg
of Na2Cr2O<7)
1.041
0.808
0.55
0.77
0.0034
* Due to a high evaporation rate, there is normally no discharge fron the
primary pond for 9 or 10 months of the year.
NA = Itot available
604
-------
TABLE 18-6. FLOW AND POLLUTANT LOADING DATA OF THE SAMPLED V&STE
STREAMS FOR PLANT #398 PRODUCING SODIUM DICHRCMATE
Average Observed Loadings
Stream
No.
1
2
Waste Stream
Description
Noncontact
cooling water
Noncontact
cooling water
Unit Flow TSS load Cr*6 load
(m3/kkg (kg/kkg (kg/kkg of
of Na2Cr2O_) of Na2Cr2O7) Na2Cr20^)
71 0.426 NNI*
206 0.55 NNI*
Chronium
load
(kg/kkg of
Na2Cr207)
NNI*
NNI*
* NNI= No net increase of the pollutant load, compared to the intake
source.
605
-------
** 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:
Pollutant
Total Subcategory Raw Waste Load Generation
Waste Load (kg/year)
Chromium (Total) 290,000
Cr (Hexavalent) 210,000
Nickel 3,700
Zinc 330
Copper 5 5
Silver 20
Lead < 8.2
Selenium 4.0
Arsenic < 5.0
Pollution Abatement Options
Toxic Pollutants of Concern
The most significant toxic pollutants found are the primary pollutant,
chromium, and the common heavy metals often present 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 also occurs.
The existing BPT regulations control pH, TSS, and chromium (Table
18-2). In the new promulgated regulations, effluent limitations on
nickel are added to the BPT-based regulations. Based on the
discussion in Section 8, there is no BAT effluent limitation set for
zinc, which is also removed in the hydroxide precipitation by
controlling chromium. Although copper, silver, selenium, lead, and
arsenic were detected in trace quantities (Section 18 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 set.
Process Modifications and Technology Transfer Options
Appropiate process modifications can be made where opportunities exist
for recycle of chrome-bearing wastewaters for recovery and reuse in
the process or for use in other product manufacturing operations.
Plant #398 currently practices extensive recovery of chromium values
606
-------
TfcHLE 18-7.
TOXIC POLLUTANT .RAW WASTE DATA
SUBCATEGORY
POIiLUTANT
Chromium, Cr
Copper, Cu
Lead, Pb
Nickel, Ni
Zinc, Zn
Silver, Ag
Selenium, Se
Arsenic, As
SODIUM
DICHRCMATE
AVERAGE RAW WS
\, p
PLANT #493 ,
(mg/i) (fcg/kkg) :
250.0
0.035
0.0090
1.25
0.580
< 0.005 <
< 0.005 <
< 0.010 <
0.94
0.00013
0.000030 :
0.0047 '•-
0.0022
0,00002 :
0.00002
0.00004
A ':
TE INFLUENT
PLANT
420.0
0.085
0.011
0.64
0.318
0.036
< 0.005
< 0.005
#376 (V ^
(kg/kkg)
3.30
0.00067
0.000090
0.0050
0.0025
0.00028
< 0.00004
< 0.00004
607
-------
TABLE 18-8.
OF RAW msTE LOADINGS FOUND IN
SCREENING AND VERIFICATION SAMELIN3
SUBCATEGORY
SODILM D3SCHFCMA1E
Ibllutant
Unit loading, (fcg/kkg)
Minimum
Average
Maximum
ND. of
Plants
Toxic
Chromium, total
Chromium,
Hexavalent
Copper
Nickel
Silver
Zinc
Selenium
Arsenic
Conventional
TSS
0.94 2.12
0.47 1.6
0.00013 0.0004
0.0047 0.027
0.000020 0.00015
0.0022 0.0024
* < 0.00003
* < 0.00004
140 2100
3.30
2.6
0.00067
0.050
0.00028
0.0025
*
*
4000
2
3
2
2
2
2
2
2
2
* Concentrations ware at or below the detection limits
60S
-------
for use in other processes
contact wastewaters.
Best Management Practices
and has no discharge of direct process
Extensive recycle and reuse of process contact wastewater limits
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.
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.
At Plant #493, 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 precipitation,
settling, and filtration. Overall, this technology is roughly
equivalent to the sulfide reduction/alkaline precipitation technique
used by Plant #376 and has the advantage of not risking operator
exposure to hydrogen sulfide gas.
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 precitation.
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.
609
-------
Selection o£ Appropriate Technology and Equipment
Technology for Different Treatment Levels
Alkaline precipitation or reaction with hydroxide 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.
A. Level 1 (BPT, BAT, NSPS)
The system utilizes pickle liquor containing ferrous iron and
hydrochloric acid added to the raw wastes to reduce hexavalent
chromium to its trivalent form 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. Other
reducing agents may be utilized instead of ferrous iron for the
reduction of hexavalent chromium such as sodium sulfide or sulfur
dioxide. Using either of these reagents, chrornate reduction
under acid conditions would be followed by pH adjustment with
lime or caustic to obtain alkaline precipitation of the metal
hydroxides. A number of operational difficulties were associated
with the treatment with sodium sulfide. Based on the
Treatability Studies on the dichromate subcategory (61), it was
found that at pH levels above 8, the addition of sodium sulfide
required excessive reaction times to reduce chromate to the
trivalent form. At pH levels below 8, the evolution of H2S gas
poses a potential safety hazard for the plant operator. The flow
diagram for the ferrous iron-based option for Level 1 is shown in
Figure 18-4.
B. Level 2
Dual-media filtration is added to achieve a higher level of
suspended solids removal, including metallic hydroxides which may
have passed through the clarifier. The effluent is adjusted to a
pH range of 6 to 9 as in Level 1. At proposal, Level 2 was
selected as a possible BAT and NSPS treatment because it was
being practiced by one plant in the industry and it could provide
a method of removing additional quantities of toxic metals from
the wastewater. However, the incremental cost of dual-media
filtration applied to Level 1 is measurably more costly for the
Sodium Dichromate Subcategory than other subcategories and does
not significantly improve effluent quality. The Treatability
Studies on the dichromate subcategory (61) showed that dual-media
filtration is only marginally effective in reducing toxic metals
and TSS beyond BPT. Therefore, this level has not been selected
as the basis for BAT or NSPS. The flow diagram for the ferrous
iron-based option for Level 2 is shown in Figure 18-5.
610
-------
C. Equipment Functions
The raw waste flows into an equalizing lagoon, where the influent
flows are measured by a magnetic flow meter which controls
application of pickle liquor solution into the influent pipeline.
Hexavalent chromium is converted to the less toxic trivalent form
and together with inert solids passes to the first-stage lagoon.
A second application of ferrous iron 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.
D. Chemicals and Handling
"Ferrous chloride, hydrochloric acid and lime are used in the
treatment process. The ferrous chloride and hydrochloric acid
solution can be applied at the influent pipeline in proportion to
flow. 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.
E. Separation and Disposal of Solids
As a basis for estimating model plant costs, influent suspended
solids, metallic hydroxide precipitates, and filter backwash are
returned to, or left in, the influent lagoon(s). As each lagoon
becomes filled with solids it is decanted from each filled lagoon
and the solid material must be periodically removed to a chemical
landfill.
F. 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 hexalent 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.
611
-------
FERROUi
JRCH
rOl
i :
RAW |
g WASTE i i
MAGNETIC
METER
T
I
1
1
1
1
1
-«X LAGOON /-M
•
-«\ LAGOON /— •»
FEB
IK
!
•
MWHH
BOOS I
CM ^
!
W?-
MPC
TANK
1LIME
QpH ADJUSTMENT
-_ . «„„
9 1 ta)
1 u' u Y
Y^LARIFIER
1
1
Includes flow monitoring, pH monitoring and sampler.
Figure 18-4. Level 1 waste water treatment for sodium dichrcmate subcategory.
-------
~1
FERROUS
IRON |
FERROUS
LIMB
RAW
MAGNETIC
MBTER
LfiOOON
•A LAGOON /-»
I
MIX
i
-U-o^E
SIMP Fn/ralR
*
BEFUJENT
OARIFIER
\
-t
^Includes flow monitoring, pH monitoring and sampler.
Figure 18-5. Level 2 waste water treatment for sodium dichromate subcategory.
-------
Treatment Cost Estimates
General Discussion
Model plant specifications were selected for
estimation. The rationale for the selection
characteristics is as follows:
A. Production
the purpose of cost
of model plant
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 wastewater 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 wastewater treatment cost
estimates, three production levels were selected. These are
20,000 kkg/year, 50,000 kkg/year and 70,000 kkg/year.
B. Wastewater Flow
Unit waste flows for three plants either treating or recycling
their wastewaters are approximately 9.6, 11.59, and 4.25 mVkkg
of product. For the model plant, 8.5 mVkkg of sodium dichromate
was used as the wastewater flow.
C. 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 #969 is 290 kg/kkg NazCr207.
The hexavalent chromium loading in the wastewater 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
NazCr207 produced, and hexavalent chromium at 5.0 kg/kkg.
D. Chemicals Required
To reduce hexavalent chromium to trivalent chromium a ferrous
chloride dosage of 4.3 g/1 is needed, but to allow for reaction
with other metals, a model dosage of 5.0 g/1 was used. This is
equivalent to 42.5 kg/kkg of product in a unit flow of 8.5
mVkkg. To raise the pH to 9.5, 1.7 g/1 of lime is needed,
equivalent to 15 kg/kkg of product. For final neutralization,
HC1 is used in the amount of 10 percent of the lime dosage.
E. Solids Generated
Tota] dry solids produced from treatment are 260 kg/kkg of sodium
dichromate.
614
-------
F. 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.
Table 18-12 gives a summary of the unit cost distribution between
amorization, 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.
Basis for Regulations
BPT Effluent Limitations
A. 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 ferrous reduction of hexavalent
chromium, followed by alkaline precipitation of metals and
clarification. As an alternative to the use of ferrous iron, the
reduction of hexavalent chromium may be accomplished by reaction
with sodium sulfide or sulfur dioxide under acidic conditions.
All three plants in this subcategory have installed BPT
technology and are meeting the limits.
The control of suspended solids is necessary to the achievement
of good effluent quality after precipitation of heavy metals. 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 considered.
B. Response to Remand Issues
The zero discharge requirements originally promulgated as BAT for
sodium dichromate production were 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 wastewater
discharge, has been identified and the performance levels
achievable have been demonstrated.
615
-------
TABLE 18-9.
MODEL PLANT TREATMENT COSTS
Subcategory
Production
Sodium dichromate
20,000 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment
Subtotal
Contractor's 0 & Pb
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
($)
Level 1
608,700
185,000
20,000
813,700
122,055
935,755
187,151
1,122,906
112,291
1,235,197
156,000
1,391,197
56,000
2,500
17,000
123,520
41,736
0
15,000
255,756
Level 2'
0
38,000
0
38,000
5,700
43,700
8,740
52,440
5,244
57,684
0
57,684
14,000
600
0
5,768
1,731
0
7,500
29,599
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
200,966
456,722
9,385
38,984
a Represents the incremental cost above that for BPT treatment
Overhead and Profit
616
-------
TABLE 18-10. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Sodium dichromate
50,000 metric tons per year
A. INVESTMENT COST
Site development .-
Equipment
Monitoring equipment
Subtotal
Contractor's 0 & P b.
Subtotal ....
Engineering ,
Subtotal ,
Contingencies
Subtotal
Land ,
TOTAL INVESTMENT COST
B. OPERATION AMD
MAINTENANCE COST
Labor and supervision ......
Energy
Chemicals
Maintenance ..
Taxes and insurance
Residual waste disposal ....
Monitoring , analysis
and reporting
TOTAL OPERATION AND
MAINTENANCE COST
($)
Level 1
1,300,000
315,000
20,000
1,635,000
245,250
1,880,250
376,050
2,256,300
225,630
2,481,930
252,000
2,733,930
56,000
2,800
42,000
248,193
82,018
0
15,000
446,011
Level 2
0
89,000
0
89,000
13,350
102,350
20,470
122,820
12,282
135,102
0
135,102
14,000
1,000
0
13,510
4,053
0
7,500
40,063
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
403,810
849,821
21,981
62,044
a Represents the incremental cost above that for BPT treatment
Overhead and Profit
617
-------
TABLE 18-11. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Sodium dichromate
70,000 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment
Subtotal ...**
Contractor's 0 & Pb.
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
Level 1
1,766,000
405,000
20,000
2,191,000
328,650
2,519,650
503,930
3,023,580
302,358
3,325,938
324,000
3,649,938
($)
56,000
2,800
58,000
332,594
109,498
0
15,000
573,892
Level 2'
0
105,000
0
105,000
15,750
120,750
24,150
144,900
14,490
159,390
0
159,390
14,000
1,000
0
15,939
4,782
0
7,500
43,221
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
541,130
1,115,022
25,933
69,153
Represents the incremental cost above that for BPT treatment
Overhead and Profit
618
-------
TABLE 18-12 MODEL PLANT UNIT TREATMENT COSTS
Subcategory Sodium Dichromate
Annual Treatment Costs ($/kkg)
COST ITEM
PRODUCTION
(kkg/yr)
LEVEL OF TREATMENT
Level 1 Level 2
Annual Operation
and Maintenance
Annual
Amortization
Total Annual
Cost
20,000
50,000
70,000
20,000
50,000
70,000
20,000
50,000
70,000
12.79
8.92
8.20
10.05
8.08
7.73
22.84
17.00
15.93
1.48
0.80
0.62
0.47
0.44
0.37
1.95
1.24
0.99
*Represents the incremental cost above BPT
619
-------
C. Flow Basis
The model plant waste 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 mVkkg, 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.
D. Selection of Pollutants to be Regulated
For BPT regulations the Agency is retaining the pollutants that
are presently limited under 40 CFR 415.172, which are pH, total
suspended solids (TSS), hexavalent chromium (CrVI), and total
chromium (Cr) and is adding nickel. The significance of these
pollutants is substantiated by the screening and verification
data presented in Section 18.
The available treatment technology for the removal of chromium
from wastewater 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, 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.
Nickel is included in the limitations since it provides a means
of controlling the group of toxic metals represented by removal
at slightly higher pH values than for chromium. A detailed
discussion of the two toxic metal groups represented by nickel
and chromium is in Section 8. Control of nickel and chromium
will ensure that toxic metals that may occasionally occur at
treatable levels will be adequately controlled.
E. Basis for Pollutant Limitations
1.
Conventional Parameters -
a.
PH
620
-------
2.
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 the proposed Development Document
(60) 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
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/ 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 achievabililty 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 mq/1) 8.5 m /kkq)
(1000 mg/1)
(kg/m*)
=0.22 kg/kkg
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). For hexavalent chromium, the observed
performance level of 0.004 mg/1 was below the accepted lower
621
-------
TABLE 18-13. EFFLUENT SAMPLING DATA FPCM
SODHM BICHROMATE PLANTS
Pollutant
Ttotal Suspended
Solids, TSS
Hexavalent
Chrcmium, Cr (VI)
Tbtal
Chrcmium, Cr (T)
Copper, Cu
Nickel, Ni
Selenium, Se
Silver, Ag
Zinc, Zn
Flow (m3/kkg)
TSS
Cr (VT)
Cr (T)
Screening & Verification Data
Plant #376 Plant #493
(mg/1)
11
< 0.01
0.81
0.012
0.20
< 0.005
0.015
0.008
Long
(mg/1)
25
0.023
0.072
(kg/kkg) (mg/1) (to^Akg)
0.046 2.0 0.0085
< 0.00004 0.004 0.00002
0.0034 2.5 0.011
0.00005 0.016 0.00007
0.00083 0.090 0.00038
< 0.00002 0.10 0,00043
0.00006 < 0.007 < 0.00003
0.0003 0.11 0.00047
4,16 4.25
Term Monitoring Data-Maximum 30-Day Averages
Plant #493(1)(2)
(kg/kkg)
0.11
0.00010
0.00031
(1) Filtered effluent data reported in response to SOS-Questionnaire
(12-22-76)
(2) The nunber of samples is unknown.
622
-------
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).
The VFR of 4.0/2.0 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 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 mq/1) (8.5 m /kkq) (kq/m3) = 0.0044 kg/kkg
(1000 mg/1)
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 chromium, the maximum 30-day average
limitation is:
(0.60 mq/) (8.5 m /kkq) (kq/m3) = 0.00050 kg/kkg
(1000 mg/1)
and the 24-hour maximum is obtained using the VFR value of
1.8, that is:
(1.8) (0.00050 kg/kkg) = 0.00090 kg/kkg
b. Nickel
The long-term average nickel concentration was selected from
Table 8-12 which presents the industrial wastewater system
performance. The variability factors are based on the
primary pollutant chromium.
The maximum 30-day average concentration basis for nickel
is,
(0.20 mg/1) (2.0) = 0.40 mg/1
The 24-hour maximum concentration basis for nickel is,
(0.20 mg/1) (4.0) = 0.80 mg/1
623
-------
TABLE 18-14. EFFLUENT LIMITATIONS*
SODIUM DICHROMATE
Best Practicable Control Technology Currently Available
Wastewater Flow: 8.5 m3Akg
Pollutant
Subcategory
Performance
(mg/l)
VFR(
Concentration Effluent Limit
Basis (mg/l) (kg/kkg)
Max. Max.
30-day 24-hr. 30-day 24-hr.
Avg. Max. Avg. Max.
Conventional
Pollutants:
Total Suspended
Solids
Toxic Pollutants:
2.0
26
52
0.22
0.44(2)
Total Chromium(8)
Hexavalent
Chromium*8)
Nickel*8*
Zinc
0.
0.
0.
0.
25U)
050(5)
20(4)
50(4)
4.
1.
4.
2.
0/2. 0(5)
8(6)
0/2.0<5)
0
0
0
0
0
.50
.060
.40
.50
1.
0.
0.
1.
0
11
80
0
0.
0.
0.
0044
00050
0034
--(7)
O.OOSS*2)
0.00090(2)
0.0068
— (7)
(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) Long-term monitoring 30-day averages from Plant #493 (Table 18-13).
(4) Industrial Wastewater System Performance (Table 8-12).
(5) The VFR used in original regulation is confirmed by long-term data on alkaline
precipitation of chromium in another subcategory (Tables A-9a-1, etc.)
(6) VFR used in original regulation.
(7) No limitations.
(8) Applicable to BAT, and PSNS.
* - Also applicable to New Source Performance Standards (NSPS).
624
-------
The maximum 30-day average limitation for nickel is,
(0.40 mq/1) (8.5 mg/kkq) (kq/m^) = 0.0034 kg/kkg
(1000 mg/1)
The 24-hour maximum limitation for nickel is,
(0.80 mq/1) (8.5 mVkkg) (kg/m^) = 0.0068 kg/kkg
(1000 mg/1)
c.
Other Metals
The concentration basis for zinc is also given in Table
18-14. This 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
nickel and chromium.
BCT Effluent Limitations
While EPA has not yet proposed or promulgated a revised BCT
methodology in response to the American Paper Institute v. EPA
decision mentioned in Section 3, EPA is promulgating BCT limitations
for this subcategory. These limits are identical to those for BPT.
EPA is not promulgating any more stringent limitations since we have
identified no technology option which would remove significant
additional amounts of conventional pollutants. As BPT is the minimal
level of control required by law, no possible application of the BCT
cost tests could result in BCT limitations lower than those
promulgated in this regulation. Accordingly, there is no need to wait
unitl EPA revises the BCT methodology before promulgating BCT
limitations.
BAT Effluent Limitations
For BAT, the Agency is setting limitations equal to BPT, since the
cost of dual-media filtration is too high to justify the relatively
small additional toxic metal removal. The pollutants limited include
hexavalent chromium, total chromium and nickel which are the same as
presented in Table 18-14.
NSPS Effluent Limitations
For NSPS, the Agency is setting limitations equal to BPT, since the
cost of dual-media filtration is too high to justify the relatively
small additional toxic metal removal. The pollutants limited include
TSS, total chromium, hexavalent chromium, and nickel which are the
same as presented in Table 18-14.
625
-------
Pretreatment Standards
The Agency is promulgating pretreatment regulations that are equal to
NSPS in order to provide better removal of toxic metals than is
achieved by a well-operated POTW with secondary treatment installed.
These pollutants would pass through a POTW in the absence of
pretreatment.
A, Existing Sources
The Agency is excluding this subcategory from pretreatment
standard for existing sources (PSES) under the provisions of
paragraph 8(b) of the Settlement Agreement because there are no
existing indirect dischargers.
B. 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 retaining this BPT treatment for NSPS and is limiting
total chromium, hexavalent chromium, and nickel as presented in
Table 18-14.
626
-------
SECTION 19
CARBON DIOXIDE INDUSTRY
Summary of Determinations
We have determined that no further effort need 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
wastewater during the screening of one plant. The subcategory is
excluded under Paragraph 8 of the Settlement Agreement.
Assessment of the Water Pollution Potential
Production Processes and Effluent
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 refigeration, in the food industry for the carbonation of
beverages, in fire extinguishing equipment, and oil well stimulation.
The process wastewater is derived from gas scrubbing and condensation.
The only toxic pollutant found at a significant concentration in the
raw waste during screening at one plant was zinc (910 pg/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
plant were:
Pollutant
Zinc
Copper
Chromium
Status of Regulations
Subpart AF has been reserved for this subcategory.
screening at one
627
-------
TABLE 19-1- -
SUBCKEBQOKf PROFILE DATA SUMMARY
SUBCAIEQ3R*
CARBON DIOXIDE
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
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 commerce, Current Industrial
Reports, Decenber 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chsnical Industry," June, 1978 and "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards for the Inorganic Chemicals Industry,"
March, 1980.
NA = Not Available
628
-------
SECTION 20
CARBON MONOXIDE AND BY-PRODUCT HYDROGEN INDUSTRY
Summary of Determinations
We have determined that no further effort need 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 wastewater during the screening of one plant. The
subcategory is excluded under Paragraph 8 of the Settlement Agreement.
Assessment of the Water Pollution Potential
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:
Concentration (
Pollutant
Chromium
Zinc
Silver
Mercury
2590
820
1 .4
1 .2
The only pollutants of significance in terms of waste loads are chrome
and zinc. However, those result from the additives 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 permit ing
authority,
Status gf_ Regulations
Subpart AG has been reserved for this subcategory.
629
-------
TAHtE 20-1
SUBCATEGOBY PRCFILE DATA
SUBCATEQORY
CARBON MONOXIDE AND B^-PRODUCT HYDROGEN
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 *yt* 1 i
Plant age range:
Mininun
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
MA
NA
8 years
19 years
NA
NA
NA
NA
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Oomnerce, Current Industrial
Reports/ December 1977;"Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assesanent of Effluent Limitations in the
Inorganic Chemical Industry, " June, 1978 and "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards for the Inorganic Chenicals
Industry," March, 1980.
NA = Not Available
630
-------
SECTION 21
COPPER SULFATE INDUSTRY
Industrial Profile
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 percent of the total U.S. production. Two of these facilities
account for over 50 percent.
The industrial profile data for this subcategory are given in Table
21-1. The status of regulations prior to promulgation of this
regulation is summarized in Table 21-2.
Subcategorization
Primary subcategorization originally chosen for the inorganic
chemicals manufacturing category was based on the dominant chemical
produced. Other factors that were considered for subcategorization
were: raw materials, manufacturing processes, size and age of plants
and equipment, geographical location, water pollution control
technology and solid waste handling. A detailed discussion of these
factors is given in Section 4 of this document.
In the original study in 1974, the Agency had further subcategorized
the copper sulfate process by raw material, promulgating regulations
for pure copper and one for copper slag and copper refinery waste.
The conclusion made in this study was that there was no need to
subcategorize the copper sulfate industry beyond the dominant product.
This is because pure raw materials make complete recycle possible, and
using them will allow a plant to comply with effluent limitations
without operation of a treatment system, if the production process is
properly operated and maintained. Both types of raw material will be
adequately covered under one regulation.
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
= CuS04 + H20
(1)
631
-------
TABLE 21-1. SUBCATEGOFY PROFILE DATA SUMMARY
SUBCATEGOKf
COPPER SULFAIE
Total subcategory capacity rate
•total subcategory production rate
Number of plants in this subcategory
308 Data on file for
With total capacity of
With total production of
Representing capacity
Representing production
Plant production ranges
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Vfeiste water flow range:
Minimum
Maximum
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 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.
632
-------
TABLE 21-2. STATUS OP REGOtATICNS - EFTTZ3ENT LIMITATION GUIDELINES
SDBCATEGORy
SUHPART
Copper Sulf ate
AJ (40 CFR 415.360, 5/22/75)
STANDAHDS
BFCTCA
BfiTER
NSPS
Product
Process
Para-
meters
Max.*1) Avg. (2) Max. Avg.
kg/kkg kg,fckg kg/Wcg kg/kkg
(mg/1) ftng/1) (mg/1) (mg/1?
Max.
Avg,
Pure Paw
Materials Cu
Process
Recovery
Process
TSS
Cu
Ni
Se
0.0006 0.0002
0.069 0.023
0.003
0.001
0.006 0.002
0.0015 0.0005
(1)- Max. = Maximum of any one day.
(2)- Avg. = Maximum average of daily values for thirty consecutive days.
633
-------
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.
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 approximately 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.
Water Use and Waste Source Characteristics
Water Use
Water is used in direct contact with copper sulfate production as the
reaction medium. A portion of it is evaporated to the atmosphere
during crystallization, while the remainder becomes part of the dry
product as its water of crystallization (hydration). Noncontact
cooling water, including steam condensate, constitutes the major water
use. This is used to cool the reactor and crystallizers. 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.
Waste Sources
A. Noncontact Cooling Water
Noncontant 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.
634
-------
WATER
F1ECTROLYTE
FRCM COPPER.
REFINER OR~
SHOT COPPER
SULFURIC.
ACID
REACTOR
FILTER
I
U)
WASH FILTER
WATER CAKE
EVAPORATOR
CRYSTAL-
LIZER
1
CENTRIFUGE
DRYER
I
MOTHER
LIQUOR
RT.FFD
CuS04.5H2O
Figure 21-1.
General process flow diagram of the manufacture of copper sulfate.
-------
TABLE 21-3. WATER USAGE IN COPPER SULFATE SUBCATEGOKY
Water Usage at Plants (m3/kkg)
Source #034 #284 #313 (1) #069
#571
Process * 1.21(2) 24.8 3.30
Contact
Noncontact 19 . 6 0 37 . 3 105
Cooling
0.075
0
Maintenance 1.25
Cleaning and
Washdcwn, Pumps
Seals and Leaks
Steam 38.6
Air Pollution 0
Control
0.35
0.28
0
0.52
3.77
0.017
(1) Includes uses for other processes
(2) Maxiumum - includes groundwater infiltration
* Utilizes feed solution from another industry
636
-------
B. Washclowns, Leaks, and Spills
Washdown, pump seal leaks, and spills are sources of contact
wastewater. These flows, however, are relatively small and
intermittent, and do not represent a major waste source.
Wastewaters emanating from this source are either combined with
the mother liquor, or treated and discharged.
C. 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.
D. Steam Condensate
A few plants use evaporators to concentrate the production
solution. Steam condensate is an additional noncontact
wastewater formed in the process. This can also be discharged
without treatment.
E. 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 direct
contact waste streams from the process. Plants utilizing pure
copper feedstock are able to recycle most contact wastewaters and
generally have no discharge of contact wastes. Table 21-4
summarizes the quantities of wastewater that go to the treatment
facility, their sources, and the handling practices for plants
which do not discharge wastewaters. The data was taken from
308-Questionnaire responses, previous development documents, and
industry visits.
Description of Plants Visited and Sampled
Screening
Plant #034 was visited and process wastewater and effluent samples
were collected and analyzed for conventional and toxic pollutants.
The process used at this plant is similar 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
637
-------
TABLE 21-4. WASTE WATER FLOW FOR THE COPPER SULFATE SUBCATEQORY
Plant
Avg. Waste
Water Flow to
Treatment
(m3/Wcg of CuSO4)
Waste Water
Handling
Practice
#034
#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
CuSOj waste (lime treatment)
Waste streams and treatment
are combined with other
mining, milling and man-
ufacturing process wastes.
Waste streams and treatment
are combined with other
metal process wastes.
Waste streams are combined
with waste from other re-
agent grade processes and
discharged to sewer.
No discharge of waste from the
process (recycle)
No discharge of waste from the
process (recycle)
No discharge of waste from the
process (recycle)
No discharge of waste from the
process (recycle)
No discharge of waste from the
process (recycle)
No discharge of waste from the
process (recycle)
* Flow is for the combined waste from all process per kkg of CuSOfl.
Actual amount of flow contributed by CuSO4 process is unavailable.
638
-------
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 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 groundwater, and is then pumped to
holding tanks. About one quarter of this wastewater volume is
comprised of contaminated groundwater 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.
Verification
Plant #034 was sampled again during the verification phase. Prior to
this, the system was changed so that only the efluent 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 se'wer.
Figure 21-2 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 wastewaters
(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.
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.
639
-------
aor
Cu
HlgMC
JCID
cu tanaaei
, *
g.T
^ OUD1
q -
J,
ZEB
TCB
L*
ID a
»1 w
IAX
croeot
H-lflc
MSH
|
SEfflJ
ID
point*
nj «4
•Uan
MIOI
1
9
pllnn
w«>Ii
ng
1
4
1
"
Nan
)|3">
am*
4
i
MB
CON
M
U
4
1
IV
not
OKL1
nm
«" <
carnm
UDK
CM
cuso
SrtE
"
nxz
E9t
,-SBjO
1
anua.
f
*
'
wunu.
nuc
12)
won ra-
nt
»
^ OUWm - OKLK
T aH. STI * OH
! i
J
^*\^
^
^
•
DISC
*
SB
UMS
laiiv
A
*»
•
I2OI
S
IQI
ESS
»«
rtoH
K
-
m
i ,
*'
•
UkHX
3
*a
LU
Figure 21-2. General process flow diagram at plant #034 showirig the sarrpling points.
(Copper sulfate manufacture.)
-------
TABLE 21-5. FLOW AND POLLUTANT CONCENTRATION DATA OF THE SAMPLED
STREAMS FOR PLANT #034 PRODUCING COPPER SULFATE
Stream
No.
1
2
Sampled Unit Flow
Stream ( m3/kkg of Cu9C
Description
Screening (1)
CuS04 waste * 1.25
Effluent from 1.25
TSS
ty (all
0.087
0.078
Cu
in kg/kkg of
4.2
0.010
Ni
CuSO4)
0.25
0.00053
lime treatment
Steam Condensate 0.209 0.00021
Verification (2)
CuS04 waste * 1.25 1.8
Effluent from
0.00016
5.0
0.000025
0.20
3
lime treatment 1.25 0.030
Noncontact 14.2 0.11
Cooling Water
and Steam
Condensate
0.0042 0.00038
0.024 0.0020
(1) From grab samples composited 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 tine of sarrpling.
641
-------
Maximum Raw Waste Concentration Observed
Ug/1)
Pollutant Screening
Antimony 330
Arsenic 3,500
Cadmium 870
Chromium 140
Copper 1,850,000
Lead 180
Nickel 112,000
Zinc 11,000
1,1,1-trichloroethane 240
NA = Not analyzed
Verification
1,300
127,000
2,500
940
3,940,000
2,200
136,000
17,000
NA
A large portion of the raw wastewater at this plant consists of
groundwater which seeps and collects in the basement, along with leaks
and washdown water from the process. The groundwater 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 of this report describes the methodology of the screening
and verification sampling program. In the copper sulfate industry, a
total of six days of sampling were conducted at Plant t034. 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 1034 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.
That is.
Unit loading (as kg of pollutant per
kkg of copper sulfate)
(C) (Q)
1000 (P)
where:
is the concentration of the pollutant expressed in units
mg/1 (Note: kg/m3 = mg/1),
of
Q
is the waste stream flow rate expressed in units of mVday.
(m3, a cubic meter, is equal to 264.2 U.S. gallons), and
642
-------
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 Ibs).
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:
Waste Load (kg/year)
1
26
400
74
15
124,000
30
6,200
700
Pollutant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Nickel
Zinc
Pollution Abatement Options
Toxic Pollutants of Concern
The principal pollutant of concern is copper. The other toxic
pollutants found in plant wastewaters 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 impure 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 containing these impurities. Arsenic was also found in
fairly high concentrations in the raw wastewater. A possible source
of arsenic, and other copper ore trace metals, is the use of sulfuric
acid made from sulfur dioxide produced in the roasting of copper
sulfide ore. 1,1rl-trichloroethane was found and several other trace
organic toxic pollutants were found in the raw waste at Plant #034
which contains infiltrated groundwater. The general area around Plant
#034 is heavily industrialized. The local groundwater is known to be
contaminated with various organic compounds. Since there are no known
organic compounds used in the feedstock, or copper sulfate process
643
-------
TABLE 21-6. RAW WASTE OAIA
Subcategory: Copper Sulfate
Average Daily Pollutant Concentrations and.loadinga found during Sampling of
Plant #034 U)
(mcrA)
(kg/Meg of CuSC^.SH^)
Pollutant
TOXK
Antimony, Sb
Arsenic, As
Cadmium, Cd
Copper, Cu
Lead, Rs
Nickel, Ni
Zinc, Zn
Chromium, Cr
Selenium, Se
CCHVEHTIOHRL
TSS
(1) The meth
Screening {2)
0.31
0.00069
3.5
0.0078
0.87
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
odology of the sanf
Verification
0.54
0.0012
44.0
0.097
1.6
0.0035
2200
5.0
0.78
0.0018
91.0
0.20
L2.0
0.027
0.36
0.000080
< 0.0050
< 0.000011
790
1.80
jling program is
t3) Overall Average (4)
0.44
0.00095
24.0
0.052
1.2
0.0027
2000
4.S
0.48
0.0011
102
0.23
12.0
0.026
0.25
0.00055
< 0.008
< 0.000018
410
0.92
described in Section 5.1,;
and Section 21.1.2 presents the scope of sampling in the Copper
Sulfate industry.
(2) Screening data from one 72-hour grab composite sample of individual or
combined raw waste streams.
(3) Verification data from three 24-hour grab composite samples, averaged.
(4) taen averaging values indicated as "less than" (<), the absolute value
was used and the resulting average was indicated as a "less than" value.
644
-------
itself, the organic toxic pollutants found at Plant #034 are atypical
and are related to the contaminated groundwater. Selenium was not
detected in the raw waste at Plant #034. However, the average
concentration of selenium was 0.1 mg/1 in the treated effluent for all
sampling trips. This phenomenon was also observed in previous studies
at this plant. The increase in selenium occurs in the treatment
operation and the source is presently unknown. It is apparent that
copper, arsenic, cadmium, chromium, lead, antimony, nickel and zinc
are typical pollutants encountered in copper sulfate wastewaters, and
that selenium appears only in the effluent sfter lime treatment.
Process Modifications and Technology Transfer Options
Mechanical scrapers could be installed on filters in plants using
impure raw materials. This would eliminate the need for backwashing
and the wastewater from this source would b§ eliminated. Installation
of these scrapers would constitute a small capital cost.
Best Management Practices
The best technology available for the treatment of copper sulfate
waste, where pure copper is used as the raw material, is total recycle
of process waste. This would require floor dikes, plumbing and sumps
to segregate the wastes, and pumps and piping for recycle.
The best technology for waste treatment where copper sulfate is
prepared from impure copper sources or copper refinery by-product is
collection of waste mother liquor and process spills, washdowns, etc.,
followed by lime precipitation of metals, settling of suspended solids
and filtration. This would require installing dikes, sewers, a
treatment tank, a settling tank, filter presses, and associated piping
and pumps (2).
Prevailing Control and Treatment Practices
Plant #034 collects leaks, spills and washdown water in a basement
sump and pumps it to holding tanks having a combined volume of 6000
gallons. The batch is treated using lime neutralization and
precipitation and is filtered by a filter press. The filtrate, after
mixing with other streams, is polished further by passing through a
cloth filter and is finally discharged to a .sewer. The filter cake is
hauled to a landfill.
Plant #284 sends mother liquor purges and filter sludges to other
processes. Wastewaters from maintenance an<3 dust control are combined
with a multitude of other process wastes and treated by lime
neutralization with aeration, followed by clarification before
discharge.
Plant #069, which produces a reagent grade product, sends periodic
purges and washdown water to a combined collection system with waste
water from various other products. Treatment consists of
neutralization and equalization of the wastes and discharge to a POTW.
64!
-------
Plant #313 also combines its wastewaters from copper sulfate
production with wastes from various other metal processes and
presently discharges the combined waste, after settling, to a pond. A
treatment system is being designed which uses lime precipitation at pH
10 followed by gravity separation and centrifugation to thicken the
sludge. The waste will then be neutralized to pH 6.5-7.5 and
discharged.
, #885, and #571 have no discharge of
o nrnroesc; _
Plants #100, #969, #050, #458, ffooD, m
wastewater from the copper sulfate process.
Advanced Treatment Technologies
Copper, nickel, cadmium and zinc can be separated from solution by
alkaline precipitation at pH values from 7.2 (copper) to 9.7
(cadmium). Alternatively, sulfide precipitation can be used. These
metals can also be removed from clarified solutions by ion exchange,
but the metal ions remaining on the exchange resins or in the
regenerant solutions may create additional disposal problems. Removal
of trace metal concentrations by the xanthate process, although
possible, has not been used widely. Some reduction of arsenic
concentrations at high pH levels has been reported, although the
removal mechanism is not clear. More effective arsenic removal would
require the addition of ferric chloride during alkaline or sulfide
precipitation of the process wastes.
Selection of Appropriate Technology and Equipment
Technologies for Different Treatment Levels
A. Level 1 (BPT, BAT, NSPS)
Alkaline precipitation using caustic soda in a batch process was
considered as the most effective technology for removal of heavy
metals and arsenic. The Agency selected Level 1 technology as
the basis for BPT because it represents the prevailing treatment
practice in this industry. All direct dischargers have BPT
installed. To accommodate a 40-hour, five-day production
schedule, the wastes are received in daily batches, and are
raised to pH 10, mixed, and settled. At the end of the workweek,
the batch is filtered and the pH adjusted to a range of 6 to 9.
Figure 21-3 shows the schematic flow diagram for Level 1
treatment.
B. Level 2
In the Level 2 treatment, the Level
the addition of ferrous sulfide in
alkaline precipitation, to increase
metals. Metallic hydroxides are
the bottom of the reaction vessel.
in the reactor with residual metals
sulfide precipitation, filter aid
1 system is supplemented by
the reaction vessel following
the precipitation of trace
allowed to form and settle in
Then ferrous sulfide is mixed
Following completion of
is added while the mixture is
646
-------
being filtered through a filter press. As in Level 1, the pH is
adjusted and the filter effluent is discharged until the weekly
batch is exhausted.
This technology was not recommended, however, because Level 1
afforded adequate control and the additional reduction was not
sufficient to offset the additional cost. Figure 21-4 shows the
schematic flow diagram for Level 2 treatment. Further
information on estimated treatment system operation, costs and
effectiveness, may be found in the proposed Development Document
(60).
Equipment for Different Treatment Levels
A. Equipment Functions
At both levels the models are designed for batch operation. Each
day's wastes are transferred from holding sumps to a reaction
vessel for storage. At the end of a workweek the BPT treatment
of the accumulated waste consists of raising the pH to 10 with
caustic soda, mixing, and applying filter aid while filtering in
a filter press. After pH adjustment to the 6 to 9 range, the
filter effluent is discharged. In the Level 2 model the
equipment remains the same but precipitation is accomplished in
two steps.
B. Chemicals Used and Handling Precautions
Caustic soda solution is added manually to each batch until the
proper pH level is reached. In Level 2, batches of ferrous
sulfide are prepared by mixing ferrous sulfate and sodium
bisulfide in a well-ventilated area. Inert filter aid is applied
as a filter precoat and is added continuously during the
filtering process. With normal precautions there are no special
chemical handling problems in the treatment of copper sulfate
wastes.
C. Separation and Removal of Solids
All solids in both levels are collected as filter cake in the
filter press. At both levels 'the dewatered cake containing
metallic hydroxides, metallic sulfides, and spent filter aid is
hauled to an off-site chemical landfill.
D. Monitoring Requirements
Alkaline precipitation of the heavy metals is assured by bringing
the reaction vessel contents to the proper pH, as determined by
the operator, using field pH equipment. Periodic specific
analyses of the final effluent for toxic pollutants can be made
by atomic absorption methods through a commercial laboratory.
647
-------
so*
WMVS1E
KUttO
INK
tKIER
i
mcrfOH
u«
V / -
f
1
<
_~*_ Jim — '
T I* MUUBBOir
.
I
^
I
1_ .,_»ru..,imi)air^
CO
uwrm.
Include* Oow nxinllorlnf, pH manUorlng and umpler
Figure 21-3. Level 1 waste water treatment for copper sulfate subcategory
batch process.
-------
Treatment Cost Estimates
General Discussion
To prepare treatment cost estimates, a model plant concept was
developed for Level 1 technology. The BPT model treatment consists
of:
Collection of wastewaters in a batch according to the production
mode.
Hydroxide treatment to precipitate metals, followed by settling
and filtration.
pH adjustment before discharge.
A. Production
Copper sulfate production ranges from 45 kkg/yr to 9100 kkg/yr in
ten plants for which 308-Questionnaires were available. The
average of the ten plants is 2100 kkg/yr and the median
production is 790 kkg/yr. The operational mode for all these
plants is assumed to be batch and to run 250 days per year.
For wastewater treatment cost estimates, one production level was
chosen as the model plant. This is the average production of
2100 kkg/year. One production level is sufficient because the
wastewaters will be collected in batches and treated as necessary
when the batch tanks are full. The amount of wastewater to be
treated at any one time is then independent of the 'production
rate, although it will determine the frequency of treatment. All
known plants with production rates below the model plant rate
have no discharge of wastewaters from the copper sulfate process.
B. Wastewater Flow
The data on Table 21-4 for plants with a wastewater discharge
shows a unit flow range from 0.52 mVkkg of CuS04 to over 23
mVkkg of CuSO^. One plant flow is for reagent grade CuS04 and
so cannot be considered a normal waste flow. Only Plant #034 has
separate treatment for CuS04 wastewater, and the flow is the
median of those normal processes sending wastewater for
treatment. The wastewater unit flow used for the model plant is
0.94 mVkkg of CuS04. All the other plants except #034 have
either no discharge of wastewater, combine their wastes with
other process wastes or produce reagent grade products.
C. Solid Waste Generation
Copper hydroxide from filtration is the only solid waste that
required disposal. This waste must be disposed of in a chemical
landfill since the solids may contain other contaminants or
become oxidized and begin to migrate into the soil or
649
-------
FIUBRUD —
Q
now monitoring. pH monitoring Bad u
Figure 21-4. Level 2 waste water treatment for copper sulfate subcategory -
batch process.
-------
groundwater. Slimes from the mother liquor and copper sulfate
solid wastes are all recycled or sent to another facility for
metal recovery. There is no solid waste generation from
processes using pure copper raw material.
Based on sludge production of 5 Ibs/day for 250 days/yr in the
model plant, the annual solids production is 558 kg, equivalent
to unit solids generation of 0.27 kg/kkg of product.
D. Treatment Chemicals
Caustic soda is required to precipitate metals and for pH
adjustment. For the model plant, the assumed caustic soda dosage
was 0.33 kg/kkg of copper sulfate.
Model Plant Cost Estimates
The cost estimate of the model plant having one level of treatment
(BPT) and one level of production is presented in Table 21-7. Table
21-.8 gives a summary of the unit cost distribution between
amortization and operation and maintenance cost components for Level 1
treatment.
Cost estimates developed for the first level of treatment (BPT, BAT,
NSPS) indicate that amortization and labor constitute a major portion
of the annual costs.
Basis for Regulations
Evaluation of BPT Treatment Practices
Copper sulfate can be manufactured using pure copper as the raw
material or an impure copper'raw material. Waste loads emanating from
the two sources differ greatly in that total recycle of process wastes
can be accomplished at plants using a pure copper source, while at
plants using an impure raw material, waste streams need to be removed
to some extent to avoid buildup of contaminants in the process.
Based on the process technology of total recycle at plants in this
study using pure raw material, the industry practices indicate that
the degree of waste control attainable is zero discharge of process
wastes.
A. Pollutant Removal with BPT Treatment
BPT technology for copper sulfate plants utilizing impure raw
materials is equivalent to Treatment Level 1. Table 21-9
presents a summary of long-term effluent monitoring data for
Plant 1034 -on total suspended solids (TSS), copper, nickel, zinc,
arsenic and selenium. Means, standard deviations, and
variability factors are given where sufficient data are
available. These performance characteristics are later utilized
for the development of the regulations.
651
-------
Table 21-10 presents the toxic and conventional pollutant data
for effluent from the two samplings at Plant #034 in the same
manner as Table 21-6 did for raw waste data. The ability of BPT
treatment to remove toxic pollutants can be estimated by
comparing the overall averages from Table 21-6 and 21-10. This
comparison is presented in Table 21-11 which also expresses the
removal efficiency as the calculated average percent removal
observed at this plant.
Table 21-11 shows that the treatment efficiency for removal of
copper, nickel, arsenic, cadmium and zinc is above 99.5 percent,
while removal for antimony is just slightly over 80 percent. The
toxic pollution concentrations were at or below the lower limit
of concentration achieved by alkaline precipitation with the
exception of copper and nickel. These toxic metal pollutants
comprised the majority of the treatment loading which suggests
that the optimum conditions for metal hydroxide formation were
not being attained at the time of sampling. The thirteenfold
increase in selenium concentrations in the treated effluent
should be noted. This phenomenon was observed in previous
studies at this and other plants. The observed concentrations
appear to remain at the lowest achievable concentration for
alkaline precipitation. The source is presently unknown, but it
is suggested that the selenium may be introduced in the treatment
chemicals.
Treatment system performance data was unavailable for other
facilities generating a waste discharge because they combine
their wastes with other process wastes for treatment.
Basis for BPT Effluent Limitations
The BPT regulations for the Copper Sulfate Subcategory were
promulgated in 40 CFR 415.363.and are presently in effect (see Table
21-2), The technology basis for the existing BPT is alkaline
precipitation plus filtration and final pH adjustment before
discharge. Of the 16 plants in this subcategory, 10 are reported to
have no discharge, five are direct dischargers, and one is an indirect
discharger. All direct dischargers have BPT technology installed.
In the existing BPT regulations, the Agency has different limitations
for pure and impure raw materials processes. The Agency has eliminatd
this distinction for BPT and is not establishing different limits for
these processes in the BAT, BCT, NSPS, PSES and PSNS regulations.
This is because both processes are adequately covered by one
regulation, since the pure raw material process should, with proper
operation and design, comply without end of pipe treatment.
A. Flow Basis
The model plant BPT treatment system is based on an inflow rate
of 0.94 mVkkg. This is derived from the average flow of Plant
#034, and was the median of plants with a wastewater discharge
652
-------
TABLE 21-7. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Copper Sulfate
2,100 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal
Contractor's 0 & Pb....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
BPT
1,800
61,500
20,000
83,300
12,495
95,795
19,159
114,954
11,495
126,449
1,800
128,249
TOTAL OPERATION AND
MAINTENANCE COST
8,000
30
1,000
12,645
3,847
100
5,000
30,622
($)
BAT'
0
0
0
0
0
0
0
0
0
0
0
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
20,573
51,196
a Represents the incremental cost above that for BPT treatment
b Overhead and Profit
653
-------
TABLE 21-8 MODEL PLANT UNIT TREATMENT COSTS
Subcategory Copper Sulfate
COST ITEM
PRODUCTION
(kkg/yr)
Annual Treatment Costs ($/kkg)
LEVEL OF TREATMENT
BPT BAT* NSPS
Annual Operation
and Maintenance 2,100 14.58
Annual
Amortization 2,100 9.80
Total Annual
Cost 2,100 24.38
NA NA
NA NA
NA NA
•Represents the incremental cost above BPT. All plants paresently
meet BPT which is equal to BAT/NSPS in this subcategory.
654
-------
TABLE 21-9. SUMMARY OF LONG TERM MONITORING DATA FROM PLANT #034
(1)
Pollutant
Total Suspended
Solids (TSS)
Copper
Nickel
Zinc
Arsenic
Selenium
Lead
Number
of Months
16
16
16
16
16
15
16
Long Term
Averages
(mg/1) (kg/kkg)
25.0 0.093
4.4 0.016
0.36 0.0013
0.12 0.00044
0.0012 0.0000044
0.0073 0.000027
0.033 0.00012
w«
2.4
1.6
2.2
2.4
3.4
6.2
2.5
(1) Values are for monthly measurements of the treated effluent combined
with nonoontract cooling water and steam oondensate discharges.
(2) For 30-day average measurements, a normal distribution is obtained
and the variability factor is found by the expression, VF = 1.0 +Z / S \
where X is the arithmetic mean and S is the arithmetic standard \X/
deviation. Hhen the value of Z is 1.64, the variability factor
is for -95 percentile which is used to set the proposed maximum 30-day
average effluent limitation. Refer to Section 8.2 for detailed discussion.
655
-------
from industrial grade CuS04 manufacturing processes. Four other
plants with wastewater discharges combine their waste with other
processes for treatment. All other plants either produce reagent
grade product, or have no discharge.
B. Selection Basis for Pollutants to be Regulated
The selection of pollutants for which numerical effluent
limitations were established was based on an evaluation of raw
waste data from the screening and verification sampling program.
The two major factors considered were: 1) individual raw waste
concentrations, and 2) the total subcategory raw waste loadings.
1. Raw Waste Pollutant Concentrations
A tabular summary of maximum raw waste concentrations found
in sampling is presented above. Data from screening
sampling was used to determine the need for verification
sampling. The maximum concentrations found during both
screening and verification are shown for comparison. As
previously discussed, selenium was not found in the raw
waste although it was present in the effluent. 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 includes copper, nickel, zinc,
arsenic, cadmium, antimony, 1-1-1-trichloroethane, and lead
in decreasing order of their apparent pollution potential.
These pollutants were observed at least once during
screening at concentrations considered treatable in the
industry using one of the available treatment technology
options. The source of trichloroethane is presumed to be
groundwater contamination. It is not process related, and
was not considered for verification. In verification, the
same metals found during screening appeared along with the
addition of chromium. The other metals found exhibited
maximum concentrations that were considerably lower than
those treatable by available technologies.
2. 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 plant sampled are summarized in
Table 21-6. This information, coupled with the estimated
total copper sulfate production rate of 27,300 kkg/year,
yielded the approximate total annual pollutant loading rates
for the subcategory shown above. This method of ranking the
pollution potential of the observed toxic metals confirms
the dominance of the eight toxic metals and ranks them as
copper, nickel, arsenic, zinc, cadmium, lead, antimony and
chromium in terms of both total mass loading and treatable
raw waste concentrations. The existing BPT regulations
656
-------
21-10. TRSKCED EFFLUENT Dfllft
Subcategory: Copper Sulfate
Average Daily Pollutant Concentrations and Loadings Found During Sampling of
Plant
{Jeg/kkg of O.iS04 . 5H20)
Toxic
PoU-utanta
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Nickel
Zinc
Selenium
Conventional
TSS
Screening (2)
0.036
0.00008
< 0.02
< 0.00004
0.001
0.000002
0.005
0.00001
4.6
0.010
0.005
0.00001
0.24
0.00053
0.016
0.000036
0.10
0.00022
Pollutant
35.0
0.078
Verification (3)
0.12
0.00027
0.057
0.00013
0.0042
0.0000089
0.017
0.000038
1.9
0.0042
< 0.031
< 0.000069
0.17
0.00038
0.02
0.000044
0.11
0.00024
13.7
0.03
(41
Overall Average
0.08
0.00018
0.038
0.000085
0.0026
0.0000054
0.011
0.000024
3.3
0.0072
< 0.018
0.00004
0.20
0.00046
0.018
0.00004
0.10
0.00023
24.0
0.054
(1) The effluent data presented here corresponds to the raw waste
data shown in Table 21-6. The methodology of the sampling program
is described in Section 5.1.2, and the scope of sampling in the
industry is described in Section 21.3.3 Copper Sulfate.
(2) Screening data from one 72-hour grab conposita sample of treated
effluent.
(3) Verificatim data from three 24-hour grab composite samples, averaged.
(4) Vhen averaging values indicated as "less than" ( < ) ,the absolute
value was used and the resulting average was indicated as a "less
than" value.
657
-------
TABLE 21-11.
AVERAGE POLLUTANT LEVELS AND REMOVAL EFFICIENCY FOR PLANT #034
Subcategory: Copper Sulfate
Waste Water Flow = 1.25 m3Akg
Pollutant
TSS
Copper
Nickel
Antimony
Arsenic
Cadmium
Chromium
Lead
Selenium
Zinc
Haw
(mg/1)
410
2000
102
0.44
24.0
1.2
0.25
0.48
< 0.008
12.0
Waste
(kg/kkg)
0.92
4.5
0.23
0.00095
0.052
0.0027
0.00055
0.0011
0.000018
0.026
Treated
(mg/1)
24.0
3.3
0.20
0.08
0.038
0.0026
0.011
< 0.018
0.10
0.018
Effluent *
(kg/kkg)
0.054
0.0072
0.00046
0.00018
0.000085
0.0000054
0.000024
0.00004
0.00023
0.00004
Percent
Removal
94
99 +
99 +
81.6
99 +
99 +
96
96
Effluent
> Inf luent
99 +
* Before cotbining with noncontact cooling and steam condensate streams.
658
-------
included selenium limitations, although selenium was not
found to be a significant pollutant in raw wastes at Plant
#034. However, its continued presence in the effluent from
alkaline treatment in significant concentrations indicates
that selenium should continue to be included in the
pollutants to be regulated.
3. Final Selection of Pollutants to be Regulated
Originally, 1 imitations were proposed for all the previously
mentioned significant toxic pollutants. The Agency has now
decided to promulgate regulations on copper, nickel and
selenium. Copper and nickel are the most predominant toxic
metals in the wastewater of this subcategory. The nature of
the treatment technology used by this industry is such that
the control of the dominant metals will ensure control of
all the toxic pollutants of concern. Further elaboration of
this subject may be found in Section 8 of this document. It
was also decided that selenium should continue to be
regulated where it is shown to be present in the effluent
and posing a potential problem. At plants where selenium is
not found at significant concentrations in the effluent, the
limitations would not apply. Because selenium was found
only in the effluent, and not in the raw waste at treatable
levels, it may not be directly related to the manufacturing
process.
Section 8 indicates that toxic metals can be divided into
two groups for pH optimization of alkaline precipitation,
and that control of one or more of the metals in each of
these groups, will ensure control of others in that group
(see Table 8-14). By controllng both copper and nickel, the
two different metal groups are covered. Therefore, because
these two metals (copper and nickel) are the predominant
toxic pollutants in the raw waste, control of these
parameters will ensure control of all the metals previously
considered for limitations. The data from all the samplings
at Plant #034 supports this.
Optimization of treatment conditions for nickel and copper
removal may cause a lack in performance of chromium control.
However, because of the relatively low incoming chromium
concentrations, the control of chromium should be adequate.
Although aresenic is not in either of the two groups, and
the incoming concentrations were the highest after copper
and nickel, the results of the sampling data show excellent
arsenic removal and very low effluent concentrations. Table
8-14 and Figure 7-1 indicate a pH around 10 would be optimum
for the control of pollutants in this subcategory.
655
-------
C. Basis of Pollutant Limitations
The BPT Effluent Limitations are presented in Table 21-12. Since
the existing BPT limitations are being retained, the
concentration bases for 4the maximum 30-day average and the daily
maximum are back calculated from the pollutant mass limitations
by applying the model plant flow rate of 0.94 mVkkg. The
variability factors used are consistent with those estimated from
the statistical analysis of treatability data (61).
1
Conventional Pollutant Limitations
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 the proposed
Development Document (60) and the JRB study (52).
TSS
The existing BPT maximum 30-day average TSS limitation
is 0.023 kg/kkg. The corresponding concentration basis
is:
(0.23 kq/kkq) (1000 mg/1 ) = 24 mg/1
(0.94 mVkkg ) ( kg/m* )
The existing BPT daily maximum TSS limitation is 0.069
kg/kkg and the concentration basis is:
(0.023 kq/kkq) (1000 mq/1) = 73 mg/1
(0.94 mVkkg
kg/m* )
from the
estimated
follows:
The long-term average for TSS is calculated
daily maximum concentration and the
variability factor for daily measurements as
(73 mq/1) = 20 mg/1
(3.6)
2. Toxic Pollutants
The effluent limitations for the selected toxic pollutants
are based on the existing BPT regulations for the Copper
Sulfate Subcategory. However, the separate limitation on
copper for the pure raw material process is being dropped
and the present limitations on TSS, copper, nickel, and
selenium for the recovery process (Table 21-2) will apply to
both processes.
660
-------
Copper
The existing BPT maximum 30-day average effluent
limitation on copper is 0.0010 kg/kkg. Applying the
model plant unit flow rate of 0.94 m3/kkg/ the
concentration basis is calculated as follows:
(0.0010 kq/kkq) (1000 mq/1) =1.1 mg/1
( 0.94 mVkkg ) ( kg/m3 )
Similarly, the existing BPT daily maximum limitation is
0.0030 kg/kkg, and the concentration basis becomes:
0.0030 kq/kkq
0.94 mVkkg)
(1000 mq/1)
( kg/m3 )
=3.2 mg/1
The long-term average for copper is then determined
from the estimated variability factor for daily
measurements:
(3.2 mg/1? = 0.89 mg/1
(3.6)
Nickel
The existing BPT maximum 30-day average effluent
limitation on nickel is 0.0020 kg/kkg. Using the model
plant unit flow rate of 0.94 mVkkg, the concentration
basis is calculated as:
(0.0020 kq/kkg) (1000 mq/1) =2.1 mg/1
( 0.94 mVkkg ) ( kg/m3)
The existing BPT daily maximum limitation is 0.0060
kg/kkg and the corresponding basis is:
(0.0060 kq/kkq) (1000 mq/1) =6.3 mg/1
( 0.94 mVkkg ) ( kg/m3)
The expected long-term average concentration is
determined from the estimated- variability factor for
daily measurement:
(6.4 mq/1) =1.8 mg/1
(3.6)
Selenium
The existing BPT maximum 30/day average limitation on
selenium is 0.00050 kg/kkg. The concentration basis is
calculated as:
€61
-------
TABLE 21-12. EFFLUENT LIMITATIONS
Copper Sulfate ...
Best Practicable Control Technology Currently Available
Waste water Flow; 0.94 n?/kkg of CuSO4
en
to
Subcategory -2\ Concentration Basis
Pollutant Performance VFR J (mg/1)
(rog/1)
Conventional Pollutant:
TSS 20
Toxic Pollutants:
Copper 0.89
Nickel 1.8
Selenium 0.44
Max.
30-day
Avg.
3.6/1.2 24
3.6/1.2 1.1
3.6/1.2 2.1
3.6/1.2 0.53
24-hr.
Max.
73
3.2
6.4
1.6
Effluent Limit
(kg/kkg)
Max.
30-day
Avg.
0.023
0.0010
0.0020
0.00050
24-hr.
Max.
0.069
0.0030
0.0060
0.0015
(1) Based on the existing BPT limitations, 40 CFR 415.360. Also applicable to NSPS.
(2) Ratio of the variability factor for daily measurements to the variablility factor for
30-day average.
-------
(0.00050 kq/kkq) (1000 mq/1) = 0.53 mg/1
( 0.94 mVkkg ) < kg/m* )
The existing BPT daily maximum limitation is 0.0015
kg/kkg and the concentration basis is:
(0.0015 kq/kkq) HOOP mq/1) =1.6 mg/1
(0.94 mVkkg ) ( kg/m* )
The expected long-term average is then determined from
the estimated variability factor for daily
measurements:
(1.6 mg/1) =0.44 mg/1
(3.6)
Basis for BCT Effluent Limitations
While SPA has not yet proposed or promulgated a revised BCT
methodology in response to the American Paper Institute v. EPA
decision mentioned in Section 3, EPA is promulgating BCT limitations
for this subcategory. These limits are identical to those for BPT.
promulgating any more stringent limitations since we have
no technology option which would remove significant
amounts of conventional pollutants. As BPT is the minimal
control required by law, no possible application of the BCT
could result in #CT limitations lower than those
EPA is not
identified
additional
level of
cost tests
promulgated in this regulation. Accordingly, there is no need to wait
until EPA revises the BCT methodology before promulgating BCT
limitations.
Basis for BAT Effluent Limitations
The Agency is promulgating BAT limitations that are equal to the
existing BPT limitations for the selected toxic pollutants (copper,
nickel, and selenium) as indicated in Table 21-13. Also shown in
Table 21-13 are concentration guidance values for the other toxic
pollutants (arsenic, cadmium, zinc, chromium, lead, and antimony)
which do not have effluent limitations. The basis for these guidance
values are developed below where treatability based performance
estimates for Level 1 treatment are presented.
When the BAT regulations were recently proposed (60), the Agency had
considered requiring more stringent limitations for the toxic
pollutants because the results of treatability studies (61) indicated
that Level 1 treatment system performance was better than the previous
BPT data base suggested. However", the actual achievability of this
improved performance is not sufficiently substantiated at present to
warrant the promulgation of more stringent limitations and, therefore,
the existing BPT limitations, with the Level 1 treatment technology
basis, are being promulgated for BAT.
663
-------
The Agency also considered control treatment Level 2 (sulfide
precipitation), but rejected this treatment because it removes only a
small additional amount of toxic metals which did not justify the
additional cost.
Basis for New Source Performance Standards
The Agency is promulgating New Source Performance Standards (NSPS)
based on treatment technology equivalent to BPT/BAT for the Copper
Sulfate Subcategory. The conventional pollutant limitations for TSS
are the same as shown for the existing BPT (BPCTCA) regulations in
Tables 21-2 and 21-12. Also, pH is limited to the range 6 to 9 as
based on data presented in Appendix B of the proposed Development
Document (60) and the JRB study (52). The toxic pollutant parameters
to be regulated are those identified in the development of the BPT/BAT
regulations as shown in Table 21-12 and the specific numerical
limitations proposed for NSPS are identical to those indicated for
BPT/BAT.
Basis for Pretreatment Standards
There is an existing PSES regulation, 40 CFR 415.364, which is based
on BPT. The Agency is amending that section in these final
regulations to establish the same effluent concentrations as were used
for the basis on BPT. In the original regulation, the concentrations
for copper were 0.5 mg/1 as the maximum 30-day average and 1.0 mg/1 as
the daily maximum. For nickel, the concentrations were 1.0 mg/1 as
the maximum 30-day average and 2.0 mg/1 as the daily maximum. Under
the amended PSES regulation, the concentrations for copper become 1.2
mg/1 and 3.2 mg/1 for the maximum 30-day average and daily maximum,
respectively, and for nickel, 2.1 mg/1 and 6.4 mg/1 as shown in Table
21-13. Selenium is also regulated under the amended PSES regulation.
For new sources, the Agency is setting PSNS equal to NSPS as indicated
in Table 21-13. Copper, nickel, and selenium are regulated.
Pretreatment standards are necessary because BAT provides better
removal of toxic metals than is achieved by a well-operated POTW with
secondary treatment installed and, therefore, these pollutants would
pass through a POTW in the absence of pretreatment.
Basis for Level 1 Treatment Performance
A. Technology Basis
The screening and verification data were collected when the
filter in the treatment system was not operating properly; hence
the treated effluent data were not representative of the
performance achievable with precipitation/filtration technology
and, therefore, the basis for Level 1 performance was derived
from other sources of performance data. The treatability study
results (61) are summarized and incorporated with other data
submitted by industry, as well as the sampling data, and
664
-------
Ul
TABLE 21-13. EFFLUENT LIMITATIONS
Copper Sulfate m
Best Available Technology
Waste Water Flow: 0.94 m3/kkg of CuSO4
Pollutant
Copper
Nickel
Selenium
Arsenic
Cadmium
Zinc
Chromium
Lead
Antimony
Subcategory
Performance
fog/1)
0.89
1.8
0.44
0.20
0.10
0.20
0.16
0.050
0.40
,„. Concentration Basis
VFIT J (mg/1)
Max.
30-day 24-hr.
3.6/1.2
3.6/1.2
3.6/1.2
4.0/1.2
4.0A-2
3.5/1.2
3.5/1.2
3.5/1.2
4.0/1.2
Avg.
1.2
2.1
0.53
0.24
0.12
0.24
0.20
0.060
0.48
Max.
3.2
6.4
1.6
0.80
0.40
0.70
0.56
0.18
1.6
Effluent Limit
(kg/kkg)
Max.
30-day 24-hr.
Avg.
0.0010
0.0020
0.0005Qv
(3)
(3)
(3)
(3)
(3)
(3)
Max.
0.0030
0.0060
0.0015
(*)
(3)
(3)
(3)
(3)
(3)
(1) Also applicable to PSES and PSNS.
(2) Ratio of the variablility factor for daily measurements to the variablility factor for
30-day averages.
(3) No effluent limitation has been established.
-------
presented in Tables 8-12 and 8-13. Where there was insufficient
data on a particular pollutant from these sources, the published
literature treatability data from Table 8-11 was used to estimate
performance.
The flow basis and the selection of pollutant parameters is the
same as that presented for the BPT limitations in this section.
B. Basis of Pollutant Control Guidance
1. Toxic Pollutants
The effluent control guidance for the selected toxic
pollutants are derived from estimated industry achievable
long-term averages (Tables 8-12 and 8-13), and literature
based treatability estimates (Table 8-11). This is
necessary because plant performance data from long-term
monitoring {Table 21-9) and the screening and verification
sampling (Table 21-10) do not reflect optimum operation of a
BPT system for removal of copper and nickel. In addition,
these sources show effluent concentrations below the lower
limit of treatability estimates for all the other toxic
metals.
Estimated effluent mass loadings for copper, nickel and
selenium are shown in Table 21-14 along with the
corresponding concentration basis. The performance levels
estimated on the basis of treatability data are better than
are presently achieved by dischargers in the Copper Sulfate
industry and the Agency does not have sufficiant evidence
that the improved performance can actually be achieved in
practice. The estimated performance is for guidance
purposes only. Also shown are the recommended concentration
bases for the other toxic metals. These are included as a
guidance reference, should the control of one or more of
these metals be found necessary in specific cases.
The concentration bases for copper, cadmium, zinc and
chromium are derived from estimated achievable long-term
industrial averages (Tables 8-12 and 8-13). Concentration
bases for nickel and arsenic are derived from the observed
concentrations from sampling of several Level 1 treatment
systems, supported by industry averages, while those for the
remaining metals of concern were derived from the lower
limit of the literature treatability estimates (Table 8-11).
This approach is used to estimate achievable performance and
provides for wider variations in the influent quality that
may be associated with different purities of copper feed
material or other process variables.
Variability factors used to calculate the maximum 30-day
average and the daily maximum concentrations from long-term
666
-------
averages were determined by
presented in the Treatability
Sulfate Subcategory.
statistical
Study (61)
analysis
for the
of data
Copper
a.
Copper
The estimated achievable long-term average for copper
indicated in Table 8-13 is 0.30 mg/1. This reflects
industry performance using Level 1 technology. This is
below the effluent level found during sampling of Plant
#034, indicating that the system was not being run at
its optimum performance. Therefore, using this
concentration value and a 30-day variability factor of
1.2 from the copper sulfate treatability results for
copper (61), the maximum 30-day concentration basis is:
(0.30 mg/1) (1.2) = 0.36 mg/1
The 24-hour maximum variability factor derived from the
copper sulfate Treatability Study for copper (61) was
found to be 3.5. Therefore, using this variability
factor and the long-term average concentration of 0.30
mg/1, the daily maximum concentration is:
(Q.3Q
U.5) = \ .\ mg/1
The effluent loadings for copper were determined using
a model plant flow of 0.94 mVkkg, with the maximum
30-day average concentration calculated as follows:
(0.36 mg/1) (0.94 mVkkg) (kq/rn^) = 0.00034 kg/kkg
(1000 mg/1)
The variability factor ratio is the ratio of the daily
maximum variability factor and the 30-day average
variability factor or:
VFR = VF of daily measurements
VF of 30-day averages
For copper the VFR is equal to:
VFR = 3.5 =2.9
1 .2
Therefore, the 24-hour maximum is calculated to be:
(2.9) (0.00034 kg/kkg) = 0.0010 kg/kkg
667
-------
TABLE 21-14. GUIDANCE FOR EFFLUENT CONTROL
Copper Sulfate
Level 1 Technology
Waste Water Flow: 0.94 m3Akg of CuSO.
Long-Term
Pollutant Average ,.»
(mg/1) VFRW
Copper 0.30(1> 3.5/1.2
Nickel 0.20(2) 4.0/1.2
C*}
Selenium 0 . 10 u } 4 . 0/1 . 2
Arsenic 0.20(2) 4.0/1.2
Cadmium 0.10(2) 4.0/1.2
Zinc 0.20(1) 3.5/1.2
Chromium 0.16(11 3.5/1.2
Lead 0.050(3) 3.5/1.2
Antimony 0.40(3J 4.0/1.2
Concentration
Basis
Effluent Loading
( kg/kkg )
(mg/D
Max
30-day
Avg
0.36
0.24
0.12
0.24
0.12
0.24
0.20
0.060
0.48
24-hr
Max
1.1
0.80
0.40
0.80
0.40
0.70
0.56
0.18
1.6
Max
30-day
Avg
-J5)
-J5>
(5)
v-* i
_J5)
_J5)
_J5)
-J5>
-J5>
_<5>
24-hr
Max
_J5>
__<5>
f 5)
—
..(5)
-J5)
_(5>
— <5)
— <5>
-J5)
(1) Industry Long-Term Average (Table 8-13) .
(2) Industry and Treatability Data
(Table 8-12) ,
(3) Lower limit of literature treatability data
(4) Variability Factor Patio: the
the monthly variability factor
study data for copper sulf ate
ratio of the
•
(Table
8-11) .
daily variability factor to
based on statistical
(61).
analysis of
treatability
(5) Indicates no control specification is needed.
668
-------
Nickel
The observed concentration of nickel in the effluent
samples from Level 1 treatment for both copper sulfate
and nickel sulfate was 0.20 mg/1. This supported by
other industry data using Level 1 technology (Table
8-12) and is used as the estimated achievable long-term
average. The 30-day average variability factor for
nickel is also 1.2. and the daily maximum variability
was 4.0. Therefore, the calculation of the maximum
30-day average concentration basis is:
(0.20 mg/1) (1.2) = 0.24 mg/1
and the 24-hour maximum concentration is:
(0.20 mg/1) (4.0) = 0.80 mg/1
The VFR for nickel is calculated as:
4.0 = 3.3
1 .2
Therefore, the maximum 30-day average based on a flow
of 0.94 mVkkg is:
(0.24 mg/1) (0.94 mVkkg) (kq/m*) = 0.00023 kg/kkg
(1000 mg/1)
and the 24-hour maximum is:
(3.3) (0.00023 kg/kkg) = 0.00076 kg/kkg
Selenium
Long-term monitoring and sampling data indicate
effluent quality either at or below the lower limit of
the estimated literature treatability data (Table
8-11). This lower limit of 0.1 mg/1 was used as the
long-term average concentration because there was
insufficient industry data in Tables 8-12 and 8-13.
Using the same variability factors and rationale
applied to nickel, the maximum 30-day average
concentration basis is:
(0.10 mg/1) (1.2) = 0.12 mg/1
and the 24-hour maximum concentration is:
(0.10 mg/1) (4.0) = 0.40 mg/1
The effluent loading for the maximum 30-day average is:
669
-------
(0.12 mg/1) (0.94 mVkkg) (kg/ri^) = 0.00011 kg/kkg
(1000 rng/1)
and for the 24-hour maximum:
(3.3V (0.00011 kg/kkg) = 0.00038 kg/kkg
Arsenic
From the data in Table 8-12, the observed concentration
of arsenic in the effluent of Level 1 treatment for
copper indicated a value of 0.20 mg/1. Using the
variability factors from nickel, the maximum 30-day
average concentration basis is:
(0.20 mg/1) (1.2) = 0.24 mg/1
and the 24-hour maximum is:
(0.20 mg/1) (4.0) = 0.80 mg/1
Cadmium
Industry data from metal finishing indicates an
effluent value of 0.10 mg/1 using Level 1 technology.
This was used as the long-term average concentration
basis rather than the lower 1imit of 1iterature
treatability estimates to account for more variation in
influent and treatment operation. Again, using
variability factors for nickel, the maximum 30-day
average concentration basis is:
(0.10 mg/1) (1.2) = 0.12 mg/1
and the 24-hour maximum basis is:
(0.10 mg/1) (4.0) = 0.40 mg/1
Zinc
Industry averages from Tables 8-12 and 8-13 indicate a
long-term average concentration of 0.20 mg/1. Again,
using the variability factors for copper, the maximum
30-day average concentration basis is:
(0.20 mg/1) (1.2) = 0.24 mg/1
and the 24-hour maximum concentration basis is:
(0.20 mg/1) (3,5) = 0.70 mg/1
670
-------
Chromium
Industry averages from Tables 8-12 and 8-13 indicate a
long-term average concentration of 0.16 mg/1. Again,
using the variability factors for copper, the maximum
30-day average concentration basis is:
(0.16 mg/1) (1.2) = 0.20 mg/1
and the 24-hour maximum concentration is:
(0.16 mg/1) (3.5) = 0.56 mg/1
Lead
From Table 8-11, the lower limit of treatability from
literature data shows a value of 0.05 mg/1. This was
used as the achievable long-term average concentration
for the same reasons as selenium. Using the
variability factors for copper, the maximum 30-day
average concentration basis is:
(0.05 mg/1) (1.2) = 0.060 mg/1
and the 24-hour maximum concentration basis is:
(0.05 mg/1) (3.5) = 0.18 mg/1
Antimony
The long-term average concentration value of 0.40 mg/1
was taken from the lower limit of literature
treatability data (Table 8-11) for the same reasons as
selenium. Using the variability factors for nickel,
the maximum 30-day average concentration basis is:
(0.40 mg/1) (1.2) = 0.48 mg/1
and the 24-hour maximum concentration basis is:
(0.40 mg/1) (4.0) = 1.6 mg/1
671
-------
-------
SECTION 22
NICKEL SULFATE INDUSTRY
Industrial Profile
General Description
Most of the nickel sulfate produced is sold in the merchant market.
The major use of nickel sulfate is in the metal plating industry, but
it is also used in the dyeing and printing of fabrics, and for
producing a patina on zinc and brass.
The industry profile data summary is given in Table 22-1, while the
status of regulations is summarized in Table 22-2.
Subcategorization
Several factors were originally considered in the subcategorization
process, such as raw materials, products, manufacturing process, size
and age of equipment, and water pollution control technology. The
conclusion was made that if effluent limitations were to be tied to
units of production, only subdivision by dominant product was viable
as a method of primary subcategorization. Option was left for further
subdivision, if necessary; however, this was not warranted in the
nickel sulfate industry.
General Process Description and Raw Materials
Nickel sulfate is produced by reacting various forms
sulfuric acid. The general reaction is:
of nickel with
NiO
= NiS04 + H20
Two different types of raw materials are used to produce nickel
sulfate. Pure nickel or nickel oxide powder may be used as a pure
material source, while spent nickel catalysts, nickel plating
solutions or residues are impure sources.
The nickel sulfate produced when pure raw materials are used is
filtered and sold or processed further. This is done by heating the
solution to 300°C in a crystallizer to produce a solid nickel sulfate
product. This must be classified, dried, and screened before it is
ready for sale.
The use of impure raw materials produces a nickel sulfate solution
which must be treated sequentially with oxidizers, lime, and sulfides
to precipitate impurities which are then removed by filtration. The
nickel sulfate solution can be sold or it may be crystallized, and the
crystals classified, dried, and screened to produce sol id nickel
sulfate for sale. Figure 22-1 shows a general process flow diagram
for the manufacture of nickel sulfate.
673
-------
TABLE 22-1. SIBCKTEGORY PROFILE DATA SUMMARY
SUBCATEGQRY NICKEL SULEATE
Total subcategory capacity rate (1)
Total subcategory production rate (1)
Number of pi ants in this subcategory (2)
308 Data on file for
With total capacity of
With total production of
Representing capacity
Representing production
Plant production range:
Minimum
Maximum
Average production
Medium production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
Indeterminant
6,350 kkg/year
11
6
17,700 kkg/year
12,650 kkg/year
NA
NA
NA
45 kkg/year
5,900 kkg/year
2,100 kkg/year
1,600 kkg/year
71.5
3
48
1.5 cubic meters/day
17.0 cubic meters/day
0.42 cubic meters/kkg
0.72 cubic meters/kkg
(1)= "Economic Analysis of Proposed Revised Effluent Guidelines and
Standards for the Inorganic Chemicals industry/' March, 1980.
(2) = Sources of data are Stanford Research institute, Directory of Chemical
Producers, U.S.A., 1977
NA = Not Available
674
-------
TABLE 22-2. STATUS OF RBGOIATICNS - EETXOENT UMITKnaN GUIDELEffiS
SCBCKEEGORX" Nickel Sulf ate
SUBPAKT
AU (40 CF-R 415.470,
5/22/75)
STANDARDS
Product
Process
Pure
Raw
Materials
Bnpure
Raw
Materials
BfCICA,
Max*1 Avg.2
Para- kg/kkg kg/kkg
meters (mg/1) (mg/1)
... Mo discharge
Ni of pwwp3
-,- Nb discharge
ras ofpwwp
m 0.006 0,002
.j^ 0.096^ 0.032
BAIEA
Max. Avg.
kg/kkg kg/kkg
(nq/1) Ong/1)
No discharge
of pwwp
Nb discharge
of pwwp
NSPS
Max* Aver.
kg/kkg kg/kkg
(rng/1) (mg/1)
Nb discharge
of pwwp
No discharge
of pwwp
Max, « Maxiiman of any one day*
2
Avg, « Average of daily valiies for thirty consecutive days.
> Process wastewater pollutants.
675
-------
Water Use and Waste Source Characteristics
Water Use
Noncontact cooling water is used for nickel sulfate production in the
reactor and crystallizers, and constitutes the major water use. Water
is used in direct process contact as a reaction component. A portion
of this water goes into the dry product as its water of
crystallization and the remainder is evaporated into the atmosphere.
Small amounts of water are used for maintenance purposes, washdowns,
cleanups, etc., and several plants use water in scrubbers for dust
control. Table 22-3 gives a summary of water usage for plants where
information was available from 308-Questionnaires and previous
documents.
Waste Sources
This stream
A. Noncontact Cooling Water
Noncontact cooling is the main source of wastewater.
is usually not treated before discharge.
B. Direct Process Contact
Plants which use impure nickel raw materials generate a filter
sludge which is treated as a solid waste. They also generate a
small filter backwash waste stream with high impurity levels
which must be treated before discharge. The filter sludges from
processes using pure nickel can be recycled back to the process.
Mother liquor, and wastewater streams from dust control are also
recycled back to the process,
C. Maintenance
Washdowns, cleanups, spills, and pump leaks are perodic streams
and account for the remaining wastes produced by nickel sulfate
plants.
Table 22-3 also shows the unit flow of total wastewater generated
from the nickel sulfate process at each plant where this
information was available.
Description of_ Plants Visited and Sampled
Screening
Plant #369 was visited and process wastewater and effluent samples
were collected and analyzed for conventional and toxic pollutants.
The process used at Plant #369 is similar to that described earlier
and utilizes nickel oxide powder feedstock. Mother liquor is recycled
back to the reactor. Sources of wastewater consist of small
quantities of mother liquor from the filter press, washdown water,
leaks, and spills. Wastewater from the process area is collected in a
676
-------
NICKEL
OIGESTOR
OXIDE
SOLUTION
PRODUCT
-^
LIOU
OIGESTOR
t - *
|— FILTER — | f—
FILTER
WENT NICKEL T 1 ...»* „„..„,
^ju, ^ OIGESTOR
^^— K3IOUC3
AGO I
^K— OXIDIZES
j FILTER
i
auffmi H TREATING TANK
t
| FILTER
CONCENTRATOR |
^
1 FILTER
|
1 CRYSTALJUZER ]
j
f
^ 3LU08C
^h">— •-•» SPENT PL ATI NO SOLUTION
«
ACID
^ SUIOM
tZ^—
QC LAS
UENT
VAPORATION TANKf*~ STEAM
^ COOL0M
^ WATER
1
OUSTS
* :-si
( COOL, SCREEN,
I PACKAGE
SOLID MWOOCT
OUSTS
i
HOLDING TANK
t ..
SCRUBBER •• w«ra»
Figure 22-1. General process flow diagram for nickel sulfate manufacture,
677
-------
TABLE 22-3- MITER USE IN THE NICKEL SULFATE SUBCATEGORY
Source
#313*
Water Uses at Plants (m3Akg)
#069 #572 #369 #120
#603*
Direct Process
Contact
Noncontact
Cooling
Water
Maintenance
Cleaning and
Hashdowns,
Pumps, Seals
and Leaks
Air Pollution
Control
Vfeste pfeter
Haste
Water
Flow to
Treatment
24.8
37.3
0.0098 0,35
1.67
4.98
0.278 0
0.751 4.01 814
0.417 13.6 2035
0.278 0.00196 0.896 0.094 Nil
Nil
0.498 0.094 1.28 0
23.4(1) .0196(1) 20.3(1)* 0.42<2) 0.72(2> NA
* = Flow data includes uses for other products.
(1)= Data source: 308-Questionnaires
(2)= Data source: Plant visits
NA = Not Available
678
-------
tank and treated as a batch by adjusting the pH to about 12.5 using
sodium hydroxide. The precipitated metal hydroxides are allowed to
settle, and the supernatant is decanted to another tank, checked for
quality and discharged to a POTW. The sludge is hauled away to an
approved landfill. Figure 22-2 shows a general treatment system flow
diagram with the location of the sampling points. Table 22-4 gives
data on flow, total suspended solids (TSS), and nickel and copper
emissions for the waste streams sampled during screening.
Verification
Plants #572 and #120 were visited and sampled during the verification
phase of the program. At Plant #572, pure nickel oxide is used as the
raw feedstock. The wastewater streams discharge on a batch basis and
are collected together in a floor drain. The wastes consist of
washdowns, leaks, and air scrubber water which are collected in an
equalization tank. In the equalization tank, alkaline wastes from
another process are mixed in and the pH is raised to 10. Solids are
allowed to settle and the clear supernatant is discharged to a POTW.
Plant #120 uses nickel oxide powder and impure nickel as raw
materials. Wastewaters from the nickel sulfate process emanate from
the filter wash, air scrubber, washdowns, and leaks, and are sent to
the treatment system. The raw wastes are mixed with other plant
nickel raw wastes prior to treatment. This consists of pH adjustment
to precipitate nickel and other trace metals followed by sand
filtration.
Figures 22-3 and 22-4 show the general treatment system flow diagram
with the waste streams sampled for Plants #572 and #120, respectively.
Table 22-4 also shows the waste stream flow and waste characteristics
for both plants. The data for Plant #572 are presented on a
concentration basis only, because a representative flow value for the
sampling point was unavailable.
Summary of Toxic Pollutant Data
Seven toxic pollutants were found at detectable concentrations in the
raw waste sample from nickel sulfate at Plant #369. Six of these
toxic metals were verified in the raw waste at two other nickel
sulfate plants. In addition, two more toxic pollutants were observed
at detectable concentrations in the raw waste during verification
sampling. No toxic organics were found at detectable concentrations
in the raw waste at Plant #369. Consequently, organic toxic
pollutants were not sought in the verification phase. The results
were:
679
-------
MttCU
00
o
Figure 22-2. General waste water treatment process flow diagram showing
sampling points at plant 1369. (Nickel sulfate sdxategory.)
-------
TABLE 22-4 . FLOW AND POIIOTANT CONCENTRATION DATA OF THE SAMPLED
STREAMS FOR PLANTS PRODUCING NICKEL SULFATE
SUBCATEGQRY NICKEL SULFATE
Stream
No.
Sampled
Stream
Description
Flow
/m3/kkg\
•°f
\NiS04/
TSS
/kg/kkg\
( -°f j
\NiS04/
Ni
/kg/kkg\
V of )
XNiS)/
Screening Data ' Plant
1
2
Raw untreated waste
Treated waste
0.42
0.42
0.093
0.045
0.073
0.00058
Cu
/kg/kkg\
( -°f )
\NiSCy
#369
0.030
0.0076
Verification Data
1
2
Raw NiSO4 waste
All Nickel raw
0.72
0.72
Plant
0.031
0.05
#120 (2)
0.035
0.0089
0.00016
<0. 0000036
vastes*
Treated effluent*
Scrubber waste
0.72
0.0031
0.00014
Plant #572
(3)
3.2
(mg/1)
1100
0.000031
.04
(1) = One grab sample of each waste water stream representing a
composited batch sample of that day's nickel sulfate production.
(2) = Average of three 24-hour composite samples of each waste water
stream.
(3) = Flow data was unavailable. Only waste water quality is presented
here.
* = The stream is a commingled waste water. The flow given is the amount
contributed by the nickel sulfate plant.
681
-------
"
00
10
MUNICIPAL
SUPPLY WATER
LEAKS. SPILLS. ETC.. FROM
OTHER PROCESSES
UASTt WATER
COLLECTION
SunP
LEAKS
WASHDOUM
SOLID wsu
TO HSPOSAL
EQUALIZATIOH
TAHK
pN - 10
ALKALINE WASTES
PCINIS.
DISCHARGE TO
SEWCR
-------
OSHER SXCKEL WASTES
NtSO. PEOCESS
HUOT
WXCSSS
DISCHARGE
Figure 22-4. General waste water treatment process flow diagram at plant #120
showing the sampling points. (Nickel sulfate manufacture.)
683
-------
Pollutant
Maximum
Concentration
Ug/1)
Screening
Plant #369
Observed
Verification (2 Plants)
Plants #572 and #120
Nickel
Copper
Chromium
Antimony
Lead
Mercury
Cadmium
Selenium
Zinc
175,500
73,300
1 ,300
476
55
1 .0
9.0
10
430
1 ,115,000
355
20
18
120
10
160
141
382
Section 5 of this report describes the methodology of the screening
and verification sampling program. In the nickel sulfate industry, a
total of seven days of sampling were conducted at Plants #369, #572
and #120, Nine 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 195 analytical data
points. The screening for toxic organic pollutants at Plant #369
generated an additional 342 analytical data points. The unit loadings
were calculated from the reported nickel sulfate production rate, the
waste stream flow rate measured or estimated at the time of sampling,
and the measured pollutant concentration.
That is,
Unit loading (as kg of pollutant per
kkg of nickel sulfate}
where:
(Q)
1000 (P)
C is the concentration of the pollutant expressed in limits of
mg/1 (Note: kg/m3, - 1000 mg/1}, and
Q is the waste stream flow rate expressed in units of mVday,
(m3, a cubic meter, is equal to 264.2 U.S. gallons) and P is the
nickel sulfate production rate expressed in units of kkg/day (kkg
is 1000 kg, a metric ton, which is equal to 2205 pounds).
The average values are based on data from those plants where
particular pollutant was found at detectable concentrations.
the
In Table 22-5, the toxic pollutant raw waste data are presented as the
average daily concentrations and the unit loading found at th ..•
584
-------
individual plants, with the exception of Plant #572 which presents
only the concentrations. The overall averages are also shown and are
calculated only for Plants #369 and #120, because they represent total
composited wastewater from the entire NiS04 process, while Plant #572
data are for one of several sources.
Based on the total annual production rate of this subcategory (See
Table 22-1) 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
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Nickel
Selenium
Zinc
Waste Load (kg/year
1 .27
0.22
0.018
1 .72
95.2
0. 19
343
< 0.17
0.48
Mercury is not included in this list as it was found at a detectable
concentration only in the one stream at Plant #572. This stream is
shared with another nickel compound process using different source
materials. Cross contamination is suspected. Since mercury was found
below detectable concentrations in all other nickel sulfate waste
water, using Plant #572 would yield an erroneously high average. This
would subsequently show an unrepresentative yearly waste load in the
previous table. Based on the reliable data, mercury is not a
pollutant of concern in the nickel sulf.ate industry.
Pollution Abatement Options
Toxic Pollutants of Concern
The toxic pollutants present in a nickel sulfate process wastewater
depend upon the purity of the sources and the nature of the raw
materials being used, which vary with time.
If impure raw materials are used, most of the heavy metal impurities
will be removed in the purification process, handled and disposed of
as solid sludge. These impurities build up in the mother liquor and
subsequently appear in purges, leaks, and washdowns. The toxic metals
such as nickel and copper, and to a lesser extent antimony, arsenic,
cadmium, chromium, lead, selenium, and zinc found in the wastewaters
during sampling originate in the raw material source. Pure raw
materials make complete recycle possible, allowing plants using these
materials to comply with the effluent limitations without operation of
a treatment system provided the manufacturing process used is properly
designed, operated, and maintained.
685
-------
TABLE 22-5. TOXIC POLLOTBtfr
WASTE DATA
: Nickel Suifatft
Average Daily Pollutant Concentrations and Loadings at
Plants Sampled Cl)
(nn/1)
(KKg Of
Pollutant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
#369 IS)
0.48
0.00020
<±>
0.009
0.0000038
1.3
0.00054
73.3
0.030
0.055
0.000022
4120 (V)
* ^- : ,
0.049
0.000035
0.0027
0.0000019
0.012
0. 00 00086
0.22
0.00016
0.052
0.000038
.,*s$%[ j
0.018
,•' *
0.16
*
0-04
0.097
.';/ Overall1^
Average
0.48
0.00020
0.049
0.0000035
0.0058
0.0000028
0.66
0.00027
36.8
0.015
0.054
0.000030
Mercury
Nickel
Selenium
Thai linn
Zinc
*
176
0.073
< 0.010
< 0.0000041
0.021
0.0000088
0,27
O.OOOU
*
49.2
0.035
0.069
0.00005
*
0.055
0.00004
0.01
1100
0.009
*
0-38
*
112
0.054
< 0.04
< 0.000027
0.021
0.0000088
0.16
0.000075
(S) - Screening data from one grab composite sample of die batch process
ccnbined raw waste streams.
(V) - Verification data frcci three 24-hour ocnpoaite aanples, averaged, from
each raw waste sampling point,
* - Concentration below significant level.
(1) - The methodology of the sanpling program is described in Section 5.1,2,
and Section 22.3.3 presents tin scope of sanpling in the 'Nickel Sulfate
industry.
(2) - Data for Plant #572 is presented in concentration basis only.
(3) - Average of Plants #369 and #120 only.
+ - Vtven averaging values indicated as "less than" (<}, the absolute value
•was used, and the resulting average waa indicated, as a "less than" valve.
686
-------
Wastewater quality for an air pollution control scrubber at Plant #572
is shown in Table 22-5. However, this source is not used to evaluate
raw waste data, since it is only one of several sources and does not
represent a total wastewater stream. This scrubber also serves in
another nickel compound manufacturing process alternately with nickel
sulfate so it cannot be considered totally representative of the
process of interest.
No toxic organic pollutants were found in
streams at significant concentrations.
the process-related waste
Process Modifications and Technology Transfer Options
Mechanical scrapers should be installed on filters at plants which use
impure raw materials. This would eliminate the backwash and reduce
the amount of wastewater produced. Solids would need to be disposed
of in a secure landfill. Installation of the scrapers would incur
only a small capital cost.
Best Management Practices
The best technology for the treatment of wastewater from processes
using pure raw materials is recycle of all process waters. To
implement this treatment, recycle piping and pumping are needed.
The best technology available where nickel sulfate is manufactured
from impure plating solutions is caustic soda addition to precipitate
nickel and other metallic hydroxides, followed by sand filtration to
remove the suspended solids. This requires installing treatment
tanks, filters, pH control equipment, and related piping and pumps.
Prevailing Control and Treatment Practices
Plant #369 sends filter leaks and wash water to a collection tank.
When the batch manufacturing process is complete, the collected waste
is treated with caustic soda to pH 12.5, The metals are precipitated
as hydroxides, settled, and the sludge disposed of at an approved
landfill. The supernatant is sampled and analyzed before discharge to
a POTW.
Plant #120 wastewaters are generated from .leaks, washdowns, filter
wash, and air scrubbers. These are combined with other nickel process
wastes and treated with caustic soda to precipitate trace metals. The
waste is then treated by filtration followed by pH adjustment prior to
final discharge.
Plant #572 also combines wastes from the air scrubbers, leaks, and
washdowns. These wastewaters are sent to an equalization tank where
they are mixed -with alkaline wastes to raise the pH to 10. After
settling, the wastewaters are discharged to a POTW.
Plant #069, which produces a reagent grade product, -sends periodic
purges and washdown water to a combined collection system with
687
-------
wastewater from numerous other products. Treatment consists of
neutralization and equalization of the wastes prior to discharge to a
POTW.
Plant #313 also combines its wastewaters from nickel sulfate
production with wastes from various other metal processes and
presently discharges the combined waste after a period of settling in
a pond. A treatment system is being designed which uses lime
precipitation at pH 10 followed by gravity separation. Centrifugation
is to be used to thicken the sludge. The clarified wastewater will
then be neutralized to pH 6.5 - 7.5 and discharged.
Plant #603
process.
has no discharge of wastewaters from the nickel sulfate
Advanced Treatment Technologies
Alkaline precipitation will remove nickel and most other heavy metals
from solution, allowing them to be settled and filtered in successive
steps. Nickel and the common heavy metals (except chromium) can also
be precipitated as metallic sulfides, for later separation by settling
and filtration. Sulfide precipitation generally yields lower
concentrations of the metals in the final effluent.
Selection o|_ Appropriate Technology and Equipment
Technologies for Different Treatment Levels
A. Level 1 (BPT, BAT, NSPS)
Level 1 consists of alkaline precipitation with caustic soda,
followed by filtration. The system is operated as a batch
process. This technology is generally the treatment practice in
place within this industry. For evaluating Level 1 treatment,
the Agency has developed two models—one for high production
plants (7,000 kkg/year), and another for low (900 kkg/year) and
medium production (4,000 kkg/year) plants—the difference being
in the method of filtration. Flow diagrams for these models are
shown in Figures 22-5 and 22-6.
B. Level 2
In proposed Level 2 treatment, the Level l system was to be
supplemented by the additional step of sulfide precipitation with
ferrous sulfide after alkaline precipitation. However, this
technology was not employed, because Level 1 affords adequate
control and the additional pollutant reduction was not sufficient
to offset the additional cost. Proposed models for this
technology are shown in Figures 22-7 and 22-8. Further
information on estimated treatment system operation costs and
effectiveness may be found in the proposed Development Document
(60).
688
-------
Equipment for Different Production Levels
A. Equipment Functions
In both Level 1 models, wastes are received in a one-day holding
tank or wastewater collection sump which is drained each day to a
reaction vessel. At the end of a normal work week, the contents
of the reaction vessel are raised to about pH 10 with caustic
soda and thoroughly mixed. In the high production model, the
contents of the reaction tank are allowed to settle and the
supernatant is filtered through a dual-media filter, while the
precipitates are filtered through a high pressure filter.
Filtrate from the dual-media filter is adjusted to a pH from 6-9
and is discharged as a final effluent. Filtrate from the high
pressure filter is recycled to the holding tank.
In the low and medium production models, the entire contents of
the reaction tank are sent through the high pressure filter after
mixing. The filtrate is adjusted for pH and discharged, while
sludge is sent to a landfill. Filter backwash is recycled to the
holding tank.
B. Chemicals and Handling
Caustic soda in solution form is used for alkaline precipitation
to form insoluble metallic hydroxides. The choice of caustic
soda avoids precipitating calcium sulfate, as would occur with
lime application. Caustic soda solution is handled in
conventional equipment, or is drawn in batches from shipping
containers when small volumes are needed.
C. Separation and Removal of Solids
In the low and midrange production models at both levels,
essentially all solids are collected in a filter press, which is
cleaned periodically. The dewatered sludge is hauled to a
chemical landfill. In the larger model plant, backwash from
cleaning the dual-media filter returns to the influent holding
tank, from which the suspended solids pass via the reaction tank
to the sludge filter press.
D. Monitoring Requirements
Satisfactory separation of heavy metals can be assured by
maintaining the proper reaction pH, which can be determined
manually on each batch, using simple field equipment. For
reporting purposes, occasional monitoring of nickel and copper in
the effluent should be done by atomic absorption methods.
689
-------
oajsnc
SOCft
OXBING
TANK
RAWVftSTE W.TER
u
53WK
1
SLUCK2
AH)
ECftL
MEDIA
FILTER
FILTES PRESS
D
Incltrles flOM mwiitoring, j« nonitoring,
and sanpler.
IWCFILL
Figure 22-5. Level 1 waste water treatment for the nickel sulfate subcategory —
high production model — batch process.
-------
CAUSTIC
SODA
RAH
WASTE WATER
REACTION
TANK
FILTER
•fi*-*
O4-
SLUDGE:
HOLDING
TANK
BACKWASH
IANDPILL
ADJUSTMBir
TER
1
1
1 <
•
PRESS
t
®
\
1
1
1
*
fTT tH V
-------
CAUSTIC 9COV
RAH
wist
vO
K>
rcumc
•BWK
r
BKKJASH
FEDBCOS
SUUXEE
SODIUM
EtSOffTEE
REACTHM
TANK
"fl*
SOPEPHKTMTT
FJlTHt MD
HMHMGT3UK
•D
KQSSS
H-1
I
J
* Inclutfes flow nemitorinj, f« nonitoring,
and sanpler.
UVMFUl,
Figure 22-7. Level 2 waste water treatment for nickel sulfate subcategory — high production
model — batch process.
-------
RAW WASTE
WATER
to
MXTUSTMENT
BACKHftSH
o
EFFLUENT
IflNtFILL
LEGEND
Includes Elow monitoring, pH monitoring
and sampler.
Figure 22-8. Level 2 waste water treatanent for nickel sulfate — low and medium
production models — batch process.
-------
Treatment Cost Estimates
General Discussion
To prepare treatment cost estimates, a model plant concept was
developed for Level 1 technology as follows:
A. Wastewater Flow
Table 22-3 shows the wastewater discharged to treatment for five
plants. The unit wastewater flow for the two single waste source
plants ranged from 0.42 mVkkg of NiS04 to 0.72 mVkkg of NiS04.
For the model plant cost estimates, a production-weighted average
of 0.68 mVkkg for the two plants was used. This was
accomplished by multiplying the unit flow of each plant by its
daily production, adding the resultant values and dividing by the
total production of the two plants, which results in these values
being representative of the different production level plants.
B. Production
Nickel sulfate production ranges from 45 kkg/yr to 5,900 kkg/yr
in the plants for which the 308-Questionnaires were available.
The average production for these six plants was 2,100 kkg/yr, the
median was 1,600 kkg/yr. For wastewater treatment cost
estimates, three production levels were selected as model plants.
These are 900 kkg/yr, 4,000 kkg/yr, and 7,000 kkg/yr. The mode
of operation at all nickel sulfate plants is the batch process
and, for the model plant, is assumed to operate for 250
days/year.
C. Solid Waste Generation
Solid wastes are generated from the filtration and settling of
metals from the nickel sulfate solution. The solids can be
recycled to the process for reuse when pure raw materials are
used. If the solids are not recycled, they are disposed of in an
industrial landfill. The quantity of solids generated is 0.39
kg/kkg of nickel sulfate.
D. Treatment Chemicals
Caustic is required for neutralization to precipitate the metals
as their hydroxides. Acid is needed for pH adjustment before
final discharge. For the model plant, these practices were
estimated to use 0.016 kg/kkg and 0.00010 kg/kkg, respectively.
Model Plant Control Costs
The cost estimates for three models having different production levels
are presented in Tables 22-6, 22-7, and 22-8.
694
-------
Table 22-9 gives a summary of the unit cost distribution between
amortization and operation and maintenance cost components at various
production rates and levels of treatment.
Cost estimates developed for the first level of treatment (BPT, BAT,
NSPS) indicate that at low production levels, labor cost has a
significant impact on the total annual costs. At a medium production
level, amortization, operation, and maintenance costs are the
important factors in the annual costs. At a high production level,
amortization cost is the significant factor in the annual costs.
Basis for Regulations
Evaluation of BPT Treatment Practices
Nickel sulfate can be manufactured using pure nickel as the raw
material or an impure nickel raw material. Waste loads emanating from
the two sources differ in that total recycle of process wastes can be
accomplished at plants using a pure nickel source, while at plants
using an impure raw material, waste streams need to be purged
periodically to avoid build-up of contaminants in the process.
A. Pollutant Removal with BPT Treatment
BPT technology for nickel sulfate plants utilizing impure raw
materials is equivalent to treatment Level 1. Table 22-10
presents the toxic pollutant treated effluent data for both
Plants #369 and #120 in a similar manner as Table 22-5 presented
the raw waste data. In evaluating BPT treatment the data from
Plant #120 was used, rather than Plant #369, or overall average
data. This is because the treatment at Plant #120 represents a
typical BPT system, while Plant #369 has no filtration before
discharge to a POTW. Long-term effluent monitoring data for
Plant #120 can be found in Tables A-15a through A-15d. The data
is for nickel only and is presented in concentration and daily
loading units for both daily and monthly measurements.
In comparing raw waste and effluent data (Tables 22-4, 22-5, and
22-10), BPT treatment gave a suspended solids removal of over 93
percent, while the toxic metals nickel and copper had over 98
percent removal. All of the toxic pollutant concentration were
below the lower limit of treatability-based achievable
concentrations. Many of the toxic metals from the effluent of
Plant #369 were below BPT based achievable levels, with only
hydroxide precipitation and settling.
Basis for BPT Effluent Limitations
BPT regulations for the Nickel Sulfate Subcategory are presently in
effect, 40 CFR 415.472 (See Table 22-2). The technology basis for the
existing BPT is alkaline precipitation plus dual-media filtration and
final pH adjustment before discharge. Most direct dischargers in this
subcategory have installed BPT technology or equivalent.
695
-------
TABLE 22-6. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Nickel Sulfate
900 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment
Subtotal
Contractor's 0 & P°
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
(S)
BPT
1,200
35,000
20,000
56,200
8,430
64,630
12,926
77,556
7,756
85,312
1,200
86,512
8,000
30
200
8,531
2,595
100
5,000
24,457
BATa
0
0
0
0
0
0
0
0
0
0
0
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
13,880
38,337
a Represents the incremental cost above that for BPT treatment
Overhead and Profit
696
-------
TABLE 22-7. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Nickel Sulfate
4,000 metric tons per year
A. INVESTMENT COST
Site development
Equipment ..
Monitoring equipment
Subtotal
Contractor's 0 & Pb
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
{$)
BPT
1,800
70,500
20,000
92,300
13,845
106,145
21,229
127,374
12,737
140,111
1,800
141,911
TOTAL OPERATION AND
MAINTENANCE COST
8,000
40
900
14,011
4,257
100
5,000
32,308
BAT
0
0
0
0
0
0
0
0
0
0
0
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
22,796
55,105
Represents the incremental cost above that for BPT treatment
Overhead and Profit
697
-------
TABLE 22-8. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Nickel Sulfate
7,000 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment
Subtotal
Contractor's 0 & Pb.........
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
($)
BPT
3,000
106,500
20,000
129,500
19,425
148,925
29,785
178,710
17,871
196,581
3,000
199,581
8,000
50
1,500
19,658
5,987
200
5,000
40,396
BATa
0
0
0
0
0
0
0
0
0
0
0
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
31,984
72,379
a Represents the incremental cost above that for BPT treatment
Overhead and Profit
698
-------
TABLE 22-9 MODEL PLANT UNIT TREATMENT COSTS
Subcategory Nickel Sulfate
COST ITEM
PRODUCTION
(kkg/yr)
Annual Treatment Costs ($/kkg)
LEVEL OF TREATMENT
BPT BAT* NSPS
Annual Operation
and Maintenance
Annual
Amortization
Total Annual
Cost
900
4,000
7,000
900
4,000
7,000
900
4,000
7,000
27.17
8.08
5.77
15.42
5.70
4.57
42.60
13.78
10.34
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
*Represents the incremental cost above BPT
699
-------
A. Conventional Parameters
For 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 the proposed Development Document (60) and the JRB
Study (52).
In the existing BPT regulations, EPA has different limitations
for pure and impure raw materials processes. EPA is not
establishing different limits for these processes in the BPT,
BAT, BCT, NSPS, PSES, and PSNS regulations. This is because both
processes are adequately covered by the one regulation since the
pure raw materials process can, with proper design, comply
without end of the pipe treatment. Only nickel, TSS and pH are
regulated because these are the only three parameters limited in
the existing BPT regulation.
Basis for BCT Effluent Limitations
While EPA has not yet proposed or promulgated a revised BCT
methodology in response to the American Paper Institute v. EPA
decision mentioned in Section 3, EPA is promulgating BCT limitations
for this subcategory. These limits are identical to those for BPT.
EPA is not promulgating any more stringent limitations since we have
identified no technology option which would remove significant
additional amounts of conventional pollutants. As BPT is the minimal
level of control required by law, no possible application of the BCT
cost tests could result in BCT limitations lower than those
promulgated in this regulation. Accordingly, there is no need to wait
until EPA revises the BCT methodology before promulgating BCT
limitations.
Basis for BAT Effluent Limitations
A. Technology Basis
For BAT, the Agency is establishing limitations based on BPT
technology which is alkaline precipitation followed by dual-media
filtration. The Agency considered treatment Level 2 (sulfide
precipitation), but rejected it because the treatment removed
only small additional amounts of toxic metals in this
subcategory.
B. Flow Basis
The model plant BAT treatment system is based on an inflow rate
of 0.68 mVkkg for effluent limitation purposes. The rationale
for the flow is the same as that used for the model plant for
cost estimating as described in this section.
700
-------
TABLE 22-10. TOXIC POLLUTANT TREATED EFFLUENT DATA
StBCATEGQRY: Nickel Sulfate
Average Daily Pollutant Concentrations and Loadings at Plants Sampled
(mg/1)
(1)
Pollutant
#369(S)
(kkg of NiS04.7H20)
#120 (V)
Overall
Average
(2)
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Nickel
Selenium
Thallium
Zinc
0.2
0.000083
0.26
0.00011
s 0.001
: 0.00000042
0.45
0.00019
18.0
0.0075
0.001
0.00000042
1.4
0.00058
0.012
0.000005
0.029
0.000012
0.17
0.000071
<0.010
<0.0000072
0.00013
0.000000094
0.057
0.000041
<0.043
<0.000031
0.003
0.0000022
0.20
0.00014
<0.008
<0.0000058
0.00033
0.00000024
0.058
0.000042
0.2
0.000083
0.13
0.000059
0.00056
0.00000026
0.25
0.00012
9.02
0.0038
0.002
0.0000013
0.8
0.00036
0.01
0.0000054
0.015
0.0000061
0.11
0.000056
(S)
(V)
(1)
(2) -
* _
Screening data from one grab composite sample of treated effluent.
Verification data from three 24-hour composite samples, averaged,
The effluent data presented here corresponds to the raw waste data shown
in Table 12-5 excluding Plant #572. The methodology of the sampling pro-
gram is described in Section 5.1.2, and the scope of sampling in the
Nickel Sulfate industry is described in Section 22.3.3.
When averaging values indicated as "less than" (<), the absolute value
was used and the resulting average was indicated as a "less than" value.
Concentration below significant level.
701
-------
C. Selection Basis for Pollutants to be Regulated
The selection of pollutants for which numerical effluent
limitations were proposed was based on an evaluation of raw waste
data from the screening and verification sampling program. The
two major factors considered were: 1) individual sampling raw
waste concentrations, and 2) the total subcategory raw waste
loadings.
1. Raw Waste Pollutant Concentrations
A tabular summary of maximum raw waste concentrations found
in sampling is presented above. Data from the plants
sampled in screening and verification are 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 nickel, copper, chromium, and to a
lesser extent, lead, antimony, and zinc were included as
candidates for regulation. These pollutants were observed
at least once during screening at concentrations considered
treatable in this industry using one of the available
treatment ^technology options. The other toxic metals
(mercury, cadmium, and selenium) were found to have maximum
concentrations that were lower than the minimum levels
achievable by treatment.
2. Total Subcategory Raw Waste Pollutant Loadings
Pollutant raw waste loading data were used to evaluate the
overall magnitude of the pollution potential. Data from the
plants sampled are summarized in Table 22-5. This
information, coupled with the estimated total nickel sulfate
production rate of 6,350 kkg/year found in Table 22-1,
yielded the approximate total annual pollutant loading rates
for the subcategory shown in this section. This method of
ranking the pollution potential of the observed toxic metals
confirms the maximum concentration-based ranking and
indicated that nickel, copper, chromium, antimony, zinc, and
lead were the six dominant toxic metals in terms of both
total mass loading and treatable raw waste concentrations.
3. Final Selection of Pollutants to be Regulated
Originally, limitations were proposed for all the
aforementioned dominant toxic pollutants. The Agency has
decided, however, to promulgate regulations on copper and
nickel only. Copper and nickel are by far the most
prominant toxic metals in this subcategory's waste. The
nature of the treatment technology employed in this industry
is such that control of certain key parameters ensures
control of all the toxic metals of concern. Elaboration on
this topic may be found in Section 8 of this document.
702
-------
Section 8 (See Table 8-14) indicates that toxic metals can
be divided into two groups for pH optimization of alkaline
precipitation, and that control of one or more of the metals
in each of these groups will ensure control of others in
that group. In controlling copper and nickel, both of these
groups are covered. Therefore, since these two metals are
the major toxic pollutants in the raw waste, control of
these parameters will ensure control of all metals
previously considered for limitations. This is supported by
sampling data from Plant #120.
Optimization of treatment conditions for nickel and copper
removal may cause less efficient performance in chromium
control, but in light of the relatively low incoming
chromium concentrations, control should be adequate. Table
8-14 and Figure 7-1 indicate a pH in the neighborhood of 10
would be optimum for control of pollutants in this
subcategory.
D. Basis of Pollutant Limitations
1. Toxic Pollutants
The effluent limitations for the selected toxic pollutant
control parameters are derived from estimated industry
achievable long-term averages (Table 8-13), and
literature-based treatability estimates {Table 8-11). This
is because plant performance data from sampling at Plant
#120 (Table 22-10) show effluent concentrations below the
lower limit of treatability estimates for all of the toxic
metals except nickel.
BAT effluent limitations for copper and nickel are shown in
Table 22-11 along with corresponding concentration bases.
Also indicated are recommended concentration bases for other
toxic metals detected at treatable levels. Those are
included for those cases where control of one or more of the
unregulated metals is deemed necessary.
Concentration bases for copper, nickel, and chromium are
derived from estimated achievable long-term averages (Table
8-13), while those for the other metals of concern were
derived from the lowest applicable treatability level (Table
8-11). This approach results in the setting of achievable
limitations, and provides for variations in the influent
quality that may be associated with different nickel or
nickel solution impurity levels or other process variables.
The basis for proposed BAT limitations on each of the
selected metals is given below. Calculations of maximum
30-day average and daily maximum concentrations from
long-term averages are based on variability factors of 1.2
703
-------
and 3.6, respectively. This gives a variability factor
ratio of:
3.6 = 3,0
1 .2
These values were determined by statistical analysis of data
presented in the Treatability Study (61) for the nickel and
copper sulfate subcategories.
a. Nickel
Table 8-13 indicates an estimated achievable long-term
average for nickel of 0.30 mg/1 which is supported by
the Treatability Study (61). This reflects industry
performance using BAT technology and is above effluent
levels indicated during sampling at Plant #120.
Therefore, using this value and the 30-day variability
factor of 1.2 the 30-day concentration basis becomes:
the daily
maximum
(0.30 mg/1) (1.2) = 0.36 mg/1
Multiplying by a VFR of 3.0,
concentration is:
(0.36 mg/1) (3.0) = 1.1 mg/1
The effluent limitations for nickel are determined
using a model flow of 0.68 mVkkg NiS04, with the
maximum 30-day average limitations calculated as
follows:
(0.36 mg/1) (0.68 mVkkg) (kg/m*) = 0.00024 kg/kkg
(1000 mg/1)
The 24-hour maximum limitation is:
(3.6H0.3 mg/1) (0.68 mVkkg) (kg/m*) = 0.00074 kg/kkg
(1000 mg/1)
These values are considerably lower than the BPT limits
presented in Table 22-2. This is because the
information on treatment system performance based on
information collected to support this regulation showed
better performance than that expected when the existing
BPT regulation was developed.
b. Copper
Because the effluent concentration of copper observed
at Plant #120 was below treatability estimates, the
estimated long-term average (Table 8-13) of 0.30 mg/1
704
-------
TABLE 22-11. EFFLUENT LIMITATIONS
Nickel Sulfate ,..,
Best Available Technology
Waste Water Flow: 0.68 m3/kkg of NiSO.
Concentration
Pollutant
Antimony
Cadmiun
Chromium
Copper
Lead
Nickel
Seleniun
Zinc
Subcategory
Performance
(mg/1)
0.40(3)
0.05(3)
0.20(4)
0.30(4)
0.050(3)
0.30(6)
0.10C3)
0.20(4)
(?)
VFR^'
3.6/1.2
3.6/1.2
3.6/1.2
3.6/1.2
3.6/1.2
3.6/1.2
3.6/1.2
3.6/1.2
Basis
Max
30-day
Avg
0.48
0.060
0.24
0.36
0.060
0.36
0.12
0.24
(mg/D
24-hr
Max
1.4
0.18
0.72
1.1
0.18
1.1
0.36
0.72
Effluent Limit
(kg/kkg
Max
30-day
Avg
_J5)
— (5)
_J5)
0.00024
— <5>
0.00024
_<5>
_<5>
of NiS04)
24-hr
Max
_J5>
__(5)
__(5>
0.00074
_J5)
0.00074
-J5)
_J5>
(1) Also applies to NSPS.
(2) Variability factor ratio: Ratio of the daily variability factor to monthly
based on statistical analyses of treatability study data for nickel and
copper sulfate (61).
(3) Lowsr limit of treatability (Table 8-11) used as long-term average.
(4) Industry long-term average (Table 8-13).
(5) Indicates no effluent limitation established.
(6) Treatability study estimate.
705
-------
was chosen to determine the concentration bases for
copper. Again, the 30-day average variability factor
of 1.2 and the VFR of 3.0 apply- Calculation of the
maximum 30-day average concentration basis is:
(0.30 mg/1) {1,2} = 0.36 mg/1
and the 24-hour maximum is:
(3.0) (0.36 mg/1) =1.1 mg/1
The maximum 30-day average effluent limitation based on
a 0.68 mVkkg flow is:
{0.36 mg/1) (0.68 mVkkg) (kq/m3) = 0.00024 kg/kkg
(1000 mg/1)
and the 24-hour maximum is:
(3.6X0.30 mg/1 X0.68 mVkkg) (kq/m3) = 0.000074 kg/kkg
1000 mg/1)
Chromium
Effluent concentration for chromium at Plant #120 was
near the lower limit of treatability, however, to
account for variations in raw waste loading the
long-term average estimate (Table 8-13) is used. The
30-day average variability factor of 1.2 and VFR of 3.0
apply. The maximum 30Lday average concentration basis
is:
(0.20 mg/1) (1.2) = 0.24 mg/1
and the 24-hour maximum is:
(3.0) (0.24 mg/1) =0.72 mg/1
Antimony
The lower limit of treatability (Table 8-11) of 0.40
mg/1 is chosen as the long-term average. Effluent
concentrations at Plant #120 were well below this
value. The 30-day variability factor of 1.2 and VFR of
3.0 are used. The maximum 30-day average concentration
basis is:
(0.40 mg/1). (1.2) = 0.48 mg/1
and^the 24-hour maximum is:
(3.0) (0.48 mg/1) =1.4 mg/1
706
-------
Lead
By the same rationale applied to antimony, the
long-term average for lead is set to the lower limit of
treatability which is 0.050 mg/1. The same variability
factors apply. The maximum 30-day average
concentrations basis is:
(0.050 mg/1) (1.2) * 0.60 mg/1
and the 24-hour maximum is:
(3.0) (0.060 mg/1) =0.18 mg/1
Zinc
Again, the lower limit of treatability is used as a
long-term average by the rationale mentioned above.
This value is 0.20 mg/1, therefore, by applying the
same variability factors, the maximum 30-day average
concentration basis is:
(0.20 mg/1) (1.2) = 0.24 mg/1
and the 24-hour maximum is:
(3.0) {0.2.4 mg/1) =0.72 mg/1
Cadmium
Using the lower limit of treatability, which is 0.05
mg/1/ as a long-term average and applying the same
variability factors, the maximum 30-day average
concentration basis for cadmium is:
(0.05 mg/1) (1.2) = 0.060 mg/1
and the 24-hour maximum is:
(3.0) (0.060 mg/1) =0.18 mg/1
Selenium
The lower limit of treatability applies for selenium.
This value is 0.10 mg/1, therefore, by applying the
same variability factors, the maximum 30-day average
concentration basis is:
(0.10 mg/1) (1.2) = 0.12 mg/1
and the 24-hour maximum is:
(3.0 (0.12 mg/1) =0.36 mg/1
707
-------
New Source Performance Standards
After examination of the effectiveness of the two treatment
technologies applicable to nickel sulfate wastes, it has been
determined that BAT technology be the basis for NSPS. The effluent
limits for toxic metals are the same as for BAT shown in Table 22-11,
and the TSS is being limited at the same effluent level as in the
existing BPT regulation presented in Table 22-2. Also, pH is limited
to the range 6 to 9 as based on data presented in Appendix B of the
proposed Development Document (60) and the JRB Study (52).
Basis for Pretreatment Standards
All five of the direct dischargers and four of the six indirect
dischargers in the Nickel Sulfate Subcategory have BPT/BAT treatment
installed.
There is an existing PSES regulation, 40 CFR 415.474. The Agency is
amending that section of these regulations based on new treatment
system performance data and the PSES limitations are the same as those
presented for BAT in Table 22-11. EPA is also setting PSNS
limitations equal to the BAT limits presented in Table 22-11. The
pollutants limited by the final PSES and PSNS regulations are nickel
and copper. Pretreatment is necessary because BAT provides better
removal of toxic metals than is achieved by a well-operated POTW with
secondary treatment installed, and therefore these pollutants would
pass through a POTW in the absence of pretreatment.
708
-------
SECTION 23
SILVER NITRATE INDUSTRY
Summary of Determinations
Action on this subcategory has been deferred, and a new subcategory
including all silver compounds will be reviewed under Phase II BAT
review, because the logical sequence of guideline promulgation was to
start the guideline process with nonferrous metals to be followed
later by a regulation on silver compounds.
Assessment of the Water Pollution Potential
Production Processes and Effluents
Most of the silver nitrate produced is for captive use in the
photographic industry. It is also used in the manufacture of silver
salts, mirrors, for silver plating, coloring porcelain and as a
chemical reagent.
The industry profile data is given in Table 23-1.
Toxic pollutants found at significant levels during sampling at one
plant were:
Pollutant
Silver
Cyanide
Concentration (
Screening Verification
164
580
65
470
Silver was not found at a significant concentration during
verification sampling of the same plant. However, a significant level
of cyanide was found again. The source of cyanide was found to be
from a soaking solution which is used to remove silver nitrate stains
from workers' clothes. The solution is sent to the silver recovery
treatment system. When plant personnel discontinued this practice,
cyanides disappeared from the effluent.
Status of Regulations
BPT limitations for this subcategory (Subpart BA) were promulgated on
May 22, 1975, (40 FR 22421) and PSES were promulgated on July 20,
1977, (42 FR 37301). Both the BPT limitations and PSES are still in
effect (40 CFR Part 415.530, 415.531, and 415.534).
709
-------
TABLE 23-1
SUBCKTBGORY PRCFILE :DAIA
SUBCATEGORy
SILVER NITRAOE
Tbtal subcategory capacity rate
Tbtal 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
Maxinum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
35,000 kkg/year
7
2
6,507 kkg/year
3,256 kkg/year
NA
9 percent
50 kkg/year
3,206 kkg/year
NA
NA
NA
20 years
64 years
<1 cubic meters/day
38 cubic meters/day
1 cubic meter/kkg
4 cubic meters/kkg
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A,, 1977, U.S. Department of Commerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.;. Draft
Report, "Preliminary 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 Chonicals
Industry," March, 1980.
NA = Not Available
710
-------
SECTION 24
SODIUM BISULFITE INDUSTRY
Industry Profile
General Description
Sodium bisulfite (NaHS03) is manufactured both in liquid and powdered
form. Captive use is very small. Sodium bisulfite is used in the
of photographic chemicals, textiles, and in food
It is also used in the tanning industry and in the
manufacture
processing
sulfite process for the manufacture of paper products.
The industry profile data are given in Table 24-1, while the status of
regulations prior to promulgation of this regulation is summarized in
Table 24-2.
Subcategorization
The method of primary Subcategorization chosen for the inorganic
chemicals point source category was subdivision by dominant product.
Other factors taken into consideration for Subcategorization included:
raw materials used, manufacturing process employed, geographical
location, size and age of equipment and facility involved, non-water-
quality aspects of waste characteristics, water pollution control
technology, treatment costs, energy requirements and solid waste
disposal. A detailed discussion of these factors is given in Section
4. No further Subcategorization of the sodium bisulfite industry,
besides dominant product, was required.
General Process Description and Raw Materials
Sodium,, bisulfite .is .produced by reacting sodium carbonate (soda ash)
with sulfur dioxide and water. The reaction is:
NaC0
2SO
H20 = 2NaHS03 + C02
(1)
This reaction produces a slurry of sodium bisulfite crystals which can
be sold, but which is usually processed to form anhydrous sodium
metabisulfite. This requires thickening, centrifuging, drying, and
packaging operations.
Water Use and Waste Source Characteristics
Water Use
Direct process contact water is used to slurry the sodium carbonate
for the reaction. Noncontact cooling water is another water use at
one plant. Water is also used for pump seals, maintenance and
washdowns. Table 24-3 gives a summary of water usage at the plants
for which 308-Questionnaires were available.
711
-------
TABLE 24-1. SUBCATEGORY PROFILE DATA SUMMARY
SUBCATEGQRY
Sodium Bisulfite
Total subcategory capacity rate
Total subcategory production rate
Number of plants in this subcategory
308 Date on file for
With total capacity of
With total production of
Representing production
Plant production range:
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Miniinum
Maximum
Volume per unit product:
Minimum
Maximum
98,000 kkg/year£D
7
2
46,000 kkg/year
28,300 kkg/year
4,700 kkg/year
23,600 kkg/year
17,800 kkg/year
16,900 kkg/year
62 percent
4 years
19 years
3 cubic meters/day
100 cubic meters/day
< 1 cubic meters/kkg
< 1 cubic meters/kkg
Sources of data are Stanford .Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of ccmnerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chemical Industry."
(1)"Energy and Environmental Analysis, Inc.;
Economic Analysis of Proposed Revised Effluent
Guidelines And Standards for the Inorganic
Chemicals Industry," March 1980.
712
-------
TABLE 24-2.
STATUS OF REGULATIONS - EFFLUENT LIMITATION GUIDELINES
SUBCATEGORY Sodium Bisulfite
SUBPART BB (40 CPR 415.540, 5/22/75)
STANDARDS
BPCTCA
Max. Avg.
Product Para- kg/kkg kg/kkg
Process meters (mg/1) (mg/1)
BATEA
Max. Avg.
kg/kkg kg/kkg
(mg/1) Cmg/1)
NSPS
Max. Avg.
kg/kkg kg/kkg
(mg/1) (mg/1)
Sodium Reserved Reserved
Bisulfite
Reserved
Reserved
= Maximum of any one day.
2Avg = Average of daily values for thirty consecutive days shall not exceed.
713
-------
Waste Sources
Noncontact cooling water for the centrifuge is a source of waste at
one plant. However, direct process contact water is the main source
of wastewater which must be treated, together with miscellaneous
wastes such as water used for maintenance purposes, washdowns, and
spill cleanup.
Table 24-4 summarizes the wastewater unit flows from the major waste
source for plants #987 and #282. Plant #987 has two facilities that
produce sodium bisulfite which are designated A and B.
There is little solid waste generation in the production of sodium
bisulfite and process waste treatment. There are minor quantities
which are precipitated as metal hydroxides resulting in insignificant
amounts of filter cake requiring disposal. Generation of solid waste
is therefore assumed negligible.
Description of. Plants Visited and Sampled
Screening
Plant #282 was visited in the screening phase of the program. The
bisulfite waste is treated on a batch basis every two or three days.
Sodium hypochlorite is added to the waste to oxidize the sulfite
waste, which is then mixed with wastes from an organic chemical plant
and neutralized. The combined wastes are then discharged to a sewer.
Table 24-5 shows the flow data and pollutant discharges, while Figure
24-1 gives the process flow diagram and shows the sampling points used
in screening.
Verification
In verification, two plants were visited, Plants #586 and #987. At
Plant #586 the sodium bisulfite wastes are combined with many other
process wastes and all treated together. Figure 24-2: shows the
flowsheet and the points sampled. Table 24-6 gives the pollutant
emissions and flow data for the waste streams. The filter wash is the
main process waste at Plant #987. This waste is neutralized with
caustic soda to pH 9 - 10 to convert the bisulfite waste to sulfite.
The sulfite is then oxidized with air to sulfate. The treated waste,
including solids, is discharged to a river. Table 24-7 shows the
pollutant emissions and flow data for the waste streams sampled.
Figure 24-i shows the process flow diagram and sampling points at
Plant #987.
Toxic Pollutant Analytical Results
The following table is a tabulation, of the toxic pollutants identified
at detectable concentrations in the raw process waste during screening
and verification. The concentrations presented under verification
represent the highest observed in the raw process waste durinp
7H
-------
TABLE 24-3,
WATER USAGE IN THE SODIUM BISULFITE SUBCATEGORY
Plant Direct Contact Process Noncontact Cooling Maintenance
, ., Washdoims, etc.
(in/kkg) (mVkkg) (nT/kkg)
# 282
I 586
# 987
0.14
NA
1.15
3.85
NA
0
1.00
NA
0.38
NA = Not Available
715
-------
TABLE 24-4. WASTE WATER FLCW AT PLANTS #987 AND #282
FOR SODIUM BISULFITE SUBCATEQOFK
SUBCATEGQRy
SODIUM BISULFITE
Source
Flow Rate Per Unit of Production (m/kkg)
Direct Process * '
Contact
Indirect process
Contact
Miscellaneous
Washdown
Total
#987A
0.018
1.50
0.31
1.83
#987B
0.018
1.17
0.42
1.61
#282
0.14
0.030
1.00<2)
1.17
Average
1.50
Plant #987 contains tWD separate facilities labeled A and B for
the purpose of comparison.
(1)
(2)
(3) - tother liquor filter wash.
Includes steam condensate which is currently treated prior to
discharge.
716
-------
TABLE 24-5. FLCW AND POLLUTANT LOAD DATA OF THE SAMPLED WASTE STREAMS FOR
PLANT #282 PRODUCING SODIUM BISULFITE1)
Waste Stream
Flow
(m3/kkg)
TSS
(kg/kkg)
COD
(kg/kkg)
Untreated waste
Treated waste
2.67
2.67
HUT •
0.424
4.04
2.61
(1) - Data based on screening sampling
which involves one 72 hour composite
sample.
(2) - Unable to determine.
717
-------
SUBLDCD
SULFUR
H
00
TO SEWER
COOLING
WSTE
Sampling points.
Figure 24-1. General process flew diagram at plant #282 showing the sampling points.
Sodium bisulfite iranufacture.
-------
SQDXIH
Bifiuu
PROCESS
II WD 12
IJBGBO
Ibsta •traana sampled.
ATR
Of If MA
Figure 24-2. General flow diagram at plant #586 showing the sampling points.
Sodium bisulfite manufacture.
-------
TABLE 24-6. FLOW AND POLLUTANT LOAD DATA OF THE SAMPLED WASTE STREAMS FOR
PLANT #586
Stream
Number
1
2
3
4
5
6
7
8
Waste Stream
Description
MBS Sump #1
MBS Sump #2
Total loads (1,2)
Amine Oxidation Pond
2nSO4 Pond Effluent
Lime Treatment Effluent
Truck Washdown
SO- Wastes
Treated Effluent
Total loads (1,2,3,4,6,7)
Flow
9.68<2>
9.68<2)
19.4
2.77(35
78.5 <3)
110 (3)
0.134(3)
85.9 (3)
188 <4>
187
TSS
(kg/kkg)
0.19
0.051
0.24
2.4
12
11
0.012
2.0
4.3
17
COD
(kg/kkg)
1.1
0.46
1.6
2.3
0.76
29
0.098
53
22
57
(1) - Data based on verification sampling
which involves three 24 hour composite
samples.
(2) - Includes noncontact process water that
does not contribute to the pollutant
load.
(3) - Raw process waste flews that are not
directly related to the sodium bisulfite
industry, but are currently treated
in combination with raw process waste
that is related.
(4) - Treated effluent from contained treatment
of a nuntoer of different raw process
waste streams not all related to sodium
sulfite production.
720
-------
TABLE 24-7. FLCW AND POLLUTANT LOAD DATA OF THE SAMPLED WASTE STREAMS FOR
PLANT #987 (D
Stream
Number
1
2
3
4
5
6
Waste Stream
Description
No. 1 Filter Wash
Floor wash, spill,
No. 2 Filter Wash
Raw Process Waste
(Streams 1+2+3)
54 Hour Aeration
Treated Effluent
Flow
(m3/kkg)
0.055
etc. 0.013
0.041
0.11
0.14
0.14
TSS
(kg/kkg)
0.11
0.046
0.0052
0.32
0.33
0.0031
COD
(kg/kkg)
1.4
0.30
0.91
3.5
1.2
1.0
(1) - Data based on verification sampling
which involves three 24-hour composite
samples.
721
-------
TO A3JCSFHERE
ALKALINE SLURRY
111 AW) 13
DRAINS, DRIPS,
SPILLS, SftSHDCWS
OUTFALL TO RIVER
12
|6
LEGQJD
Waste streams sailed.
t T
NaCH AIR
Figure 24-3. General process flow diagram at plant 1987 shewing the sampling points.
Sodium bisulfite manufacture.
-------
sampling.
levels.
No organic toxic pollutants were found at detectable
Maximum Raw Waste Concentration Observed
Pollutant
Arsenic
Copper
Zinc
Cadmium
Chromium
Antimony
Lead
Mercury
Nickel
Silver
Thallium
Screening
Plant #282
12
380
2500
6
0
30
8
3
250
2
8
Verification
Plant #586 and #987
67
930
3600
41
3400
650
1100
17
460
15
8
Section 5 of this report describes the methodology of the screening
and verification sampling program. In the Sodium Bisulfite
subcategory a total of seven days of sampling were conducted at Plants
#282, #586, and #987. Sixteen different sampling points were
identified for the various waste streams at these three plants. The
evaluation of toxic metal content of these waste streams was based on
429 analytical data points and an additional 516 points for the toxic
organic pollutants sampled during screening.
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 units of
.mg/1 (Note: kg/m3 = 1000 mg/1), and
Q is the waste steam flow rate expressed in units
(m3, a cubic meter, is equal to 264.2 U.S. gallons).
of mVday.
Similarly, the unit loadings were calculated from the reported
sodium bisulfite production rate, the waste stream flow rate, and
the measured pollutant concentration.
723
-------
Unit loading (as kg of pollutant per kkg of NaHS03
1000P
where C
sodium
(kkg is
and Q are
bisulfite
1000 kg,
the same as
production
a metric ton
described above, and
rate expressed in units
which is equal to 2205
P is the
of kkg/day.
Ibs.)
Table 24-8 presents the toxic pollutant unit loading and concentration
at the three plants sampled. Each concentration represents the
average of three composite samples for verification and a single
composite sample during screening.
In Table 24-9, the toxic pollutant raw waste data are presented as the
minimum, average, and maximum unit loadings based on the results
summarized for each plant in Table 24-9. The average unit loading is
based on the average obtained at the three plants sampled.
Based on the total annual production of this subcategory and the
average waste load generated per unit product, the estimated toxic
pollutant raw waste loads generated each year for this subcategory are
as follows:
Total Annual Pollutant Load
Pollutant
Antimony
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Silver
Arsenic
Thallium
Raw Waste Load (kg/year
5.0
1 .0
1100.0
45
9.0
0.60
30
520
6.2
2.3
22
Table 24-10 presents the average toxic pollutant concentration
observed in the treated effluent during verification sampling.
Pollution Abatement Options
Toxic Pollutants of Concern
It is reported that some sources of sodium carbonate contain zinc and
other trace metals in measurable amounts. The screening and
verification sampling program revealed zinc, chromium, copper, lead,
nickel, and antimony were at significant concentration levels which
may require regulation. Zinc may enter the waste stream by corrosion
of galvanized metals by coproduct operations or from nonprocess zinc
compounds used by the industry. Cadmium, arsenic, thallium, mercury,
and silver though detected in the raw waste are not at treatable
724
-------
TABLE 24-8. TOXIC POLLUTANT RAW WASTE LOADS
SUBCATEGORY
SODIUM BISULFITE
POLLUTANT
Arsenic
Copper
Zinc
Cadmium
Chromium
Lead
Mercury
Nickel
Antimony
Thallium
Silver
PLANT AND SAMPLING PHASE
Screening
#282
(mg/1) (kg/kkg)
0.012 0.000030
0.38 0.0010
2.5 0.0070
0.0060 0.000017
0.0025 0.000007
0.0030 0.000007
0.25 0.00070
0.030 0.000070
0.0080 0.000020
0.0060 0.000017
(2)
Verification v '
#987
(mg/1) (kg/kkg)
0.067 0.000010
0.74 0.000070
2.4 0.00020
0.040 0.000004
2.6 0.00030
0.60 0.000070
0.012 0.000001
0.46 0.000050
0.65 0.000070
<0.050 <0. 000004
<0.030 <0. 000003
#586
(mg/1)
0.0020
0.018
0.52
0.00050
1.3
0.012
0.00060
0.010
0.0050
0.025
0.010
(kg/kkg)
0.000030
0.00030
0.0088
0.000010
0.022
0.00020
0.000010
0.00017
0.000080
0.00042
0.00017
(1) - One 72-hour composite sample
(2) - Average of three 24-hour composite samples
725
-------
TABLE 24-9. SUMMARY OP RAH HASTE LOADINGS FOUND IN SCFUMIHG AM> VERIFICATION SAMPLING
•U
to
SUBCATBQORY
Pollutant
Priority
Antimony
Cadmium
Chromium
Copper
Lsafl
Mercury
Nickel
Zinc
Silver
Arsenic
Thallium
Conventional
Total Suspended
SODIUM BISUIfTFE
Loading Range
(kg/day)
Mini"?**
0.00045
0.00023
0.018
0.0050
0.000091
0.000091
0.0032
0.016
<0. 00020
0.00040
<0. 000052
Solids (TSS)
3.20
Maximum
0.0041
0.00041
1.1
0.015
0.0095
0.00045
0.0091
0.42
0.0080
0.0014
0.020
25.4
Unit loading
OcgAkg)
Minima
0.000007
0.000004
0.00030
0.000070
0.000007
0.000001
0.000050
0.00020
<0. 000003
0.000010
-------
TABLE 24-10. TOXIC POLLUTANT CONCENTRATIONS OBSERVED IN TREATED EFFLUENT DURING
VERIFICATICN SAMPLING
Pollutant
Arsenic
Copper
Zinc
Cadmium
Chromium
Lead
Mercury
Nickel
Antimony
Thallium
Silver
#987 P]
(nn/1) .
ND
0.27
0.010
ND
0.11
0.15
ND
ND
ND
ND
ND
Lant #586
(mcr/1)
ND
ND
ND
ND
ND
ND
0.010
0.050
0.020
ND
ND
ND - Not Detected
727
-------
concentrations and thus
concern.
are not considered toxic pollutants of
Prevailing Control and Treatment Practices
Plant #987 adds 50 percent caustic solution to the oxidation tank to
raise the pH to approximately 9.5 and blows air through while
mechanically agitating. The waste is discharged to a river following
a 17-hour retention period.
Plant #282 uses caustic soda or sodium carbonate for pH control
followed by sodium hypochlorite addition to oxidize sulfite and other
reduced sulfur species. The waste is then neutralized and discharged
to a County sewer.
Plant #586 mixes the bisulfite waste, amine plant waste, and truck
wash waste. Lime is added to the combined wastes which are then
passed through an aeration tank with eight-hour's retention time where
zinc sulfate production wastes are added. The treated waste goes
through primary and secondary settling before final discharge.
Advanced Treatment Technologies
Toxic metals may be precipitated at alkaline pH values, when reacted
with sulfides in various forms, in some cases by ion exchange resins,
and the Xanthate process. Sulfide precipitation from cleared
solutions could be used to provide additional removal of zinc, lead,
nickel, copper, and to a lesser extent, antimony.
Selection of Appropriate Technology and Equipment
Technologies for Different Treatment Levels
A. Level 1 (BPT/BAT)
Neutralization with caustic soda to a pH of 9.5 followed by
aeration and settling was chosen as the most cost-effective
method of lowering the COD associated with the primary pollutant,
sodium bisulfite. Toxic metals in the waste form metal hydroxide
precipitates which are removed from the wastewater by settling.
The flow Diagram is shown in Figure 24-4.
B. Level 2
Aerated effluent from the BPT system is chlorinated to complete
COD removal, and is then filtered to remove finely divided
suspended matter carried through or produced in the BPT system,
particularly if toxic metals are present in the incoming wastes.
The flow diagram is shown in Figure 24-5. Cost estimates for
this level of treatment are given in the proposed Development
Document (60).
728
-------
C. Level 3
Ferrous sulfide is applied ahead of the Level 2 filter, to
precipitate any residual metals by the more effective sulfide
process. The flow diagram is shown in Figure 24-6. Cost
estimates for this level of treatment are given in the proposed
Development Document (60).
Equipment for Different Treatment Levels
A. Equipment Functions
In Level 1, the raw wastes are received in one of two similar
holding/reaction ranks (sized for one day's flow), recirculated
and automatically adjusted to pH 9.5. During the second working
day, while the contents of the first tank are jet aerated and
allowed to settle, the daily raw wastes are received,
recirculated and pH adjusted in the second tank. At Level 1, the
aerated and settled effluent is discharged directly from the
alternate holding/reaction tanks. Sludge is drawn off the tanks
at suitable intervals and disposed of. At Levels 2 and 3, the
aerated and settled effluent is pumped through a multi-media
filter on a schedule which will leave its tank empty and ready
for the next working day's influent. At Levels 2 and 3,
supplemental ch1orination equipment is provided. At Level 3,
residual heavy metals which may have escaped prior treatment, or
may have been released by pH change during chlorination, will be
converted to metallic sulfides by addition of ferrous sulfide
ahead of the Level 2 filter. Should COD and toxic metals be very
low in specific instances, the chlorination and ferrous sulfide
steps could be eliminated.
B. Chemicals and Handling
Caustic soda solution, chlorine, and ferrous sulfide are used in
the treatment processes. Caustic soda and chlorine are common
industrial chemicals which are fed by conventional equipment
designed to minimize leaks, spills, and hazards to personnel.
Ferrous sulfide is prepared by mixing ferrous sulfate with sodium
bisulfide under well ventilated conditions. When the usual
precautions are taken in the proper handling of corrosive and
toxic chemicals, there should be no special problems in applying
the proposed technologies.
C. Separation and Removal of Solids
No solids are formed in the proposed treatment, with the possible
exception of small amounts of metal hydroxides and sulfides if
metals should be present in the raw wastes. In that event, the
precipitated solids, which are formed in the holding/reaction
tank or which are returned to the holding/reaction tank during
backwashing, will settle in the hopper bottom of the reaction
729
-------
RAW_
WASTE
WVTHl
-J
Ul
o
CRUSTIC
SCOft
^
1
eta !
w 1 1 ,
\
_rV-UM
» 1
i
KJLDINCy
REBCTICW
TBW
-MR
SUKXX
1— @
-L. !
! 1
+ f
U>*—
'
HctDiwy
REACTION
TBJK
]
^r^O-1
LEEBC
* Includes flow nDtiiboring, ffl
nonitoring, and
SLUDCZ
Figure 24-4. Level 1 vraste treatment for the soditm bisiiLfite subcategory
batch process.
-------
BftCKWASH
FILTER
r-(S)
t^ '
p^ i |
O-—
^ 1
1
•
HCtDINcy
FEflCTlCW
TANK
1
=+Q>-
f*-AIR
i
-OttORlNKTICW
t EFFLUEHT
* Includes flow mcnitaring, pH monitorinj,
and sanpler.
SLUDGE
Figure 24-5, Level 2 waste water treatment for the sodium bisulfite subcategory —
batch process.
-------
-J
CJ
10
RPH
WASTE
HATER
CMETTC
scm
n
r^>
.L. '
<*-• 1
_^ 1
^ T
«3UJ»
BERCT1
TANF
1
•1
C/
X»
'
— »*Oj-
SUKGB
--/S^j
i \ry
exj ,
Bou>n>
FEKTl
TAM
iq/
CN
J
'
— *-&-
)
FILTEH
Q
EtfUUHW
* Includes flow monitoring, ffl monitoring,
and sawpler.
Figure 24-6. I^wel 3 waste water treatment for the sodium bisulfite subcategory
batch process.
-------
tank. As necessary, these solids can be drawn off and disposed
of in an appropriate manner.
D. Monitoring Requirements
Internal process monitoring will be done with standard field
equipment measuring pH, dissolved oxygen, and chlorine. If
metals are present- in the raw materials a periodic laboratory
analysis for metals should be made on the final effluent.
Monitoring for dissolved sulfide should not be necessary, since
the ferrous sulfide applied in Level 3 is rather insoluble.
Treatment Cost Estimates
General Discussion
To prepare treatment cost estimates, a model plant
developed for a single level of technology as follows:
A. Wastewater Flow
concept was
The sources of wastewater include wet air scrubbers, filter
backwash, floor washings, leaks, and spills. The unit flow rates
ranged from 1.8 mVkkg to 1.2 m3/kkg of product at the three
facilities for which information was available. The average was
approximately 1.5 mVkkg and this was used for the model plant
(Table 24-4).
B. Production
Sodium bisulfite production ranges from 4770 kkg/yr to 31,800
kkg/yr at the three plants for which data was available. The
average production is 17,800 kkg/yr. The production rates at the
three plants were used as the model plant production rates. The
operational mode is continuous and is assumed to run 350 days per
year.
C. Solid Wastes
In the production of sodium bisulfite and process waste treatment
there is little solid waste generation, although precipitation of
metal hydroxides may result in small quantities of solids
requiring disposal. The model plants assumed no significant
solid waste production.
D. Treatment Chemicals
Caustic soda is needed to adjust the pH to 9.5. The only other
requirement is air to oxidize the waste. For the model plant,
the caustic soda dosage was assumed to be 0.195 kg/kkg.
733
-------
Cost Estimates
The cost estimates of three models having different production
are presented in Tables 24-11, 24-12 and 24-13.
level
Table 24-14 gives a summary of the unit cost distribution between
amortization, operation and maintenance. Cost components at various
production levels are also shown.
Basis for Regulations
Evaluation of BPT Treatment Practices
All seven plants in this subcategory have installed BPT or equivalent
technology. Plant peformance was estimated on the basis of
verification sampling results for Plant #987. Plant #282 was excluded
from the evaluation since the treatment technology applied at the
particular point of treated effluent sampling does not represent the
appropriate level of treatment. Plant #586 was excluded from
consideration, since the combined treatment of the sodium bisulfite
process waste with wastes from other unrelated processes, has made an
evaluation of the plant performance data too speculative.
Table 24-15 is a comparison of the maximum raw waste concentrations
from screening and verification sampling with treatability data from
Plant #987, literature-based estimates, and industrial wastewater
treatment system performance. The data used in this table are from
the screening and verification results, Table 8-11, Table 8-12, Table
8-13 and the Treatability Study (61).
Basis for BPT Effluent Limitations
A. Technology Basis
The Agency is setting limitations based upon hydroxide
precipitation of toxic metals with caustic soda plus batch
aeration and settling for BPT. The flow schematic for BPT is
shown in Figure 24-4 in Section 24 as Level 1 treatment. The
Agency has selected Level 1 treatment as the basis for BPT
because it reflects current industry practice.
B. Flow Basis
The basis of flow for BPT limitations is estimated from data
provided in the Section 308-Questionnaires for two of the three
complete plant responses, including Plant #987 and #282. Plant
#586 was omitted in view of the lack of adequate information to
identify the wastewater streams contributed by the sodium
bisulfite process alone.
The three major raw process wastewater streams include direct and
indirect process contact waste and miscellaneous floor and tank
washdown wastewater.
734
-------
TABLE 24-11. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Sodium Bisulfite
4,770 metric tons per year
A. INVESTMENT COST
BPT
($)
BATa
Site development
Equipment
Monitoring equipment ..
Subtotal
Contractor's 0 & Pb....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
Land
TOTAL INVESTMENT COST ,
B. OPERATION AND
MAINTENANCE COST
Labor and supervision .
Energy
Chemicals
Maintenance
Taxes and insurance ...
Residual waste disposal
Monitoring, analysis
and reporting
3,000
70,000
20,000
93,000
13,950
105,950
21,390
128,340
12,834
141,174
3,000
144,174
TOTAL OPERATION AND
MAINTENANCE COST
56,000
2,100
2,900
14,117
4,325
0
15,000
94,443
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
22,969
117,412
a Represents the incremental cost above that for BPT treatment
Overhead and Profit
735
-------
TABLE 24-12. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Sodium Bisulfite
16,900 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal
Contractor's 0 & Pb....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
Land
TOTAL INVESTMENT COST .
B. OPERATION AND
MAINTENANCE COST
Labor and supervision .
Energy
Chemicals
Maintenance
Taxes and insurance ...
Residual waste disposal
Monitoring, analysis
and reporting
($)
BPT
3,000
117,000
20,000
140,000
21,000
161,000
32,200
193,200
19,320
212,520
3,000
215,520
TOTAL OPERATION AND
MAINTENANCE COST
56,000
2,900
9,800
21,252
6,466
0
15,000
111,418
BAT
0
0
0
0
0
0
0
0
0
0
0
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
34,577
145,995
Represents the incremental cost above that for BPT treatment
Overhead and Profit
736
-------
TABLE 24-13. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Sodium Bisulfite
31,800 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal .........
Contractor's 0 & Pb....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
($)
BPT
6,000
195,000
20,000
221,000
33,150
254,150
50,830
304,980
30,498
335,478
6,000
341,478
TOTAL OPERATION AND
MAINTENANCE COST
56,000
3,800
19,000
33,548
10,244
0
15,000
137,592
BATa
0
0
0
0
0
0
0
0
0
0
0
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
54,582
192,174
Represents the incremental cost above that for BPT treatment
Overhead and Profit
737
-------
TABLE 24-14 MODEL PLANT UNIT TREATMENT COSTS
Subcategory Sodium Bisulfite
Annual Treatment Costs ($/kkg)
COST ITEM
PRODUCTION
(kkg/yr)
LEVEL OF TREATMENT
BPT
Annual Operation
and Maintenance
Annual
Amortization
Total Annual
Cost
4,770
16,900
31,800
4,770
16,900
31,800
4,770
16,900
31,800
19.80
6.59
4.33
4.82
2.05
1.72
24.61
8.64
6.04
738
-------
TABLE 24-15.
COMPARISON OF MAXIMUM RAW WASTE
CONCENTRATIONS WHH TREftTABILITY
Pollutant Maximum Avg. Raw Treated Effluent Literature-Based Industrial Waste
Waste
Concentration
During Screening
and Verification
(mg/1)
Concentration
Fran Plant #987
Verification
Sampling
(mg/D
Treatability
Estimates from
Table 8-11 for
Lime/Settling
(mg/1)
Water Treatment
System Performance
for Lime/Settling
(mg/1)
TSS
COD
Arsenic
Copper
Zinc
Cadmium
Chromium
Lead
Mercury
Nickel
Antimony
Thallium
Silver
2,250
24,700
0.067
0.74
2.5
0.0060
2.6
0.60
0.012
0.46
0.65
0.050
0.030
22
7,300
ND
0.27
0.010
ND
0.11
0.15
ND
ND
ND
ND
ND
—
—
0.50
0.50
0.50
0.10
0.10
0.30
0.20
0.80
0.20
0.40
38(1)
450 (2)
0.080(3J
0.40(4)
Q.80(4)
0;32(4>
0.15(4)
0.40(4)
0.18(3)
—
—
ND = Not Detected
(1) Data from Treatability Study (61) on lime/settling in nickel sulfate and
sodium dichromate industries.
(2) Average of extended aeration tests from Treatability Study (61).
(3) Data from Table 8-12.
(4) Estimated achievable long-term average concentrations from Table 8-13.
739
-------
Table 24-4 summarizes the unit flows reported for each of the
three sources at each facility. Plant #987 has two facilities
which manufacture sodium bisulfite and are designated as A and B.
The basis of model plant flow for the sodium bisulfite industry
is estimated as the average total raw wastewater flow for the
three plants, and is used to estimate pollutant discharge
loadings for the purpose of regulation. The average total flow
for the three facilities considered is 1 .5 mVkkg of product.
C. Selection Basis for Pollutants to be Regulated
The selection of pollutants for which specific numerical effluent
limitations are promulgated 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 #586 and #987 provided additional pollutant
raw waste concentration data needed to assess the performance of
treatment technology.
Results of the screening and verification sampling are
in Table 24-15 for the raw process waste streams.
tabulated
Toxic pollutants are listed based on their presence, during
sampling, at detectable concentration levels. Pollutants from
this list were considered as candidates for regulation if their
concentration appeared at least to equal or exceed the lowest
level estimated as treatable using any available technology
appropriate for their removal (Table 8-11).
The relative significance of the candidate toxic pollutants was
estimated based on the total annual raw waste load for each
pollutant which appears in a Table above. The total annual load
is based on the average concentration observed during screening
and verification, which is tabulated in Table 24-9, in addition
to the estimated annual production of 98,000 kkg of product for
the industry.
On the basis of concentration and total annual raw waste loads
determined during sampling, COD, TSS, chromium, zinc, copper,
nickel, lead, and antimony have been identified at significant
concentration levels in the raw waste stream and are candidates
for regulation. These pollutants are listed in order of their
relative significance with regard to decreasing. raw waste
concentration.
In view of the treatment technology currently practiced and the
related nature of the candidate toxic pollutants (see Section 8),
control of the more significant toxic pollutants will ensure
adequate control of those metals which may occasionally appear at
treatable levels. Therefore, only chromium and zinc will have
effluent limitations. Concentration values for the other metals
are intended for guidance only.
740
-------
The Agency conducted treatability studies (61) using Level 1
(BPT) technology (without settling) on typical raw wastewater
from the sodium bisulfite industry. In the tests the average COD
level was reduced from 950 mg/1 to 450 mg/1. Use of the standard
iodide-iodate test for sulfite indicated the tests reduced the
average iodate demand as oxygen from 800 mg/1 to 25 mg/1. The
Agency is evaluating the standard iodide-iodate test for sulfite
for possible application in effluent monitoring but will continue
to set effluent limitations using the conventional COD test.
D. Basis of Pollutant Limitations
BPT limitations are presented in Table 24-16.
1. 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 the proposed
Development Document (60) and the JRB Study (52).
b. TSS
The BPT limitations for TSS are based on data from the
Treatability Study (61) using lime/settling from the
nickel sulfate and sodium dichromate industries. The
long-term average is 38 mg/1. The daily and 30-day
variability factors are 5.6 and 1.4, respectively.
The 24-hour maximum concentration is:
(38 mg/1) (5.6) = 210 mg/1
The maximum 30-day average concentration is:
(38 mg/1) (1.4) = 53 mg/1
The load limitations for TSS (kg/kkg) are calculated
based on the unit flow rate of 1.5 mVkkg, thus:
(210 mg/1) (1.5 mVkkg) ( kg/m3 ) = 0.32 kg/kkg
(1000 mg/1)
for the 24-hour maximum limit. The maximum 30-day
average is calculated similarly, i.e.:
(53 mg/1) (1.5 mVkkg) ( kg/m3 ) = 0.080 kg/kkg
(1000 mg/1)
741
-------
TABLE 24-16. EFFLUENT LIMITATIONS
SODIUM BISULFITE
Best Practicable Control Technology Currently Available
Wastewater Flow: 1.5 m^/kkg
Pollutant
Subcategory
Performance
(mg/1) ___
Daily
Variability
Factor
30-day
Variability
Factor
Concentration Effluent Limit
Basis (mg/1) (kg/kkg)
Max. Max.
30-day 24-hr. 30-day 24-hr.
Avg. Max. Ayg, Max.
Conventional/
Nonconventional
Pollutants:
Total Suspended
Solids, TSs(6>
Chemical Oxygen
Demand, COD*7)
Toxic Pollutants :
Chromium*7)'*9)
Zinc*7)
Copper
Lead
Nickel
Antimony
38(D
450<2)
0.32*3)
0.80(3)
0.40*3)
0.15(3)
0.40(3)
0.80(5)
5.6/1.4
5.6/1.4
4.2/1.3
4.2/1.3
4.2/1.3
4.2/1.3
4.2/1.3
4.2/1.3
53
630
0.42
1.0
0.52
0.20
0.52
1.0
210
2500
1.3
3.4
1,7
0.63
1.7
3.4
0.080
0.95
0.00063
0.0015
— (8)
— (8)
— (8)
~(8)
0.32
3.8
0.0020
0.0051
— (8)
-.(8)
— (8)
~(8)
(1) Data from Treatability Study (61) on lime/settling in nickel sulfate and
sodium dichromate industries.
(2) Based on data from Treatability Study (61).
(3) Based on estimated achievable long-term average concentrations from
Table 8-11.
(4) Maximum untreated effluent concentration from screening and verification
sampling data.
(5) Based on lower limit of literature-based treatability estimate from
Table 8-11.
(6) Also applicable to NSPS regulations.
(7) Also applicable to NSPS/ and BAT regulations.
(9) No effluent limitation.
(9) Also applicable to PSNS regulations*
742
-------
c. COD
The BPT limitations for COD are based on data from the
Treatability Study (61) on the aeration of sodium
bisulfite waste. The average concentration of the
treated effluent from the tests, 450 mg/1, is used as a
long-term average. The daily and 30-day average
variability factors used for TSS above are employed
again.
The 24-hour maximum concentration is:
(450 mg/1) (5.6) = 2500 mg/1
The maximum 30-day average concentration is:
(450 mg/1) (1.4) = 630 mg/1
The load limitations for COD (kg/kkg) are calculated
based on the unit flow rate of 1.5 m3/kkg, thus:
(2500 mg/1) (1.5 mVkkg) ( kq/m^ ) =3.8 kg/kkg
(1000 mg/1)
for the 24- hour maximum limit. The 30-day average
limit is calculated similarly, i.e.:
2.
(630 mg/1) (1 .5 mVkkg)
Toxic Pollutants
( kg/m3 )
(1000 mg/1)
=0.95 kg/kkg
No long-term sodium bisulfite industry data is available to
establish limitations and guidelines for the selected toxic
pollutants. The effluent limitations and guidelines are
based on four information sources including: (1) estimated
achievable long-term average concentrations from Table 8-13,
(2) industrial wastewater treatment system performance data
from Table 8-12, (3) screening and verification data, and
(4) literature-based treatability estimates.
a. Chromium
The limitations for chromium are based on the estimated
achievable long-term average concentration of 0.32 mg/1
from Table 8-13. The daily and 30-day average
variability factors of 4.2 and 1.3, respectively, are
obtained from data in the Treatability Study (61) on
chromium removal using lime/settling from the
chlor-alkali, titanium dioxide, sodium dichromate and
chrome pigments industries.
743
-------
b.
The 24-hour maximum concentration is:
(0.32 mg/1) (4.2) = 1.3 mg/1
The maximum 30-day average concentration is:
(0.32 mg/1) (1.3) = 0.42 mg/1
The load limitations for chromium (kg/kkg) are
calculated based on the unit flow rate of 1.5 mVkkg,
thus:
(1.3 mg/1) (1.5 mVkkg)
kq/m3 ) - 0.0020 kg/kkg
1000 mg/1
for the 24-hour maximum limit. The maximum 30-day
average is calculated similarly, i.e.:
(0.42 mg/1) (1.5 mVkkg)
Zinc
kq/m3 ) = 0.00063 kg/kkg
(1000 mg/1)
The limitations for zinc are based on the estimated
achievable long-term average concentration of 0.80 mg/1
from Table 8-13. The daily and 30-day average
variability factors used for chromium are employed
again.
The 24-hour maximum concentration is:
(0.80 mg/1) (4.2) = 3.4 mg/1
The maximum 30-day average concentration is:
(0.80 mg/1) (1.3) = 1.0 mg/1
The load limitations for zinc (kg/kkg) are calculated
based on the unit flow rate of 1.5 mVkkg, thus:
(3.4 mg/1) (1.5 mVkkg) ( kq/m^ ) = 0.0051 kg/kkg
(1000 mg/1)
for the 24-hour maximum limit. The maximum 30-day
average is calculated similarly, i.e.:
(1,0 mg/1) (1.5 mVkkg) ( kg/m3 ) = 0.0015 kg/kkg
(1000 mg/1)
744
-------
Copper
The guidance for copper is based on the estimated
achievable long-term average concentration of 0.40 mg/1
from Table 8-13. The daily and 30-day average
variability factors used for chromium are employed
again.
The 24-hour maximum concentration is:
(0.40 mg/1) (4.2) = 1.7 mg/1
The maximum 30-day average concentration is:
(0.40 mg/1) (1.3) = 0.52 mg/1
Lead
The guidance for lead is based on the estimated
achievable long-term average concentration of 0.15 mg/1
from Table 8-13. The daily and 30-day average
variability factors used for chromium are employed
again.
The 24-hour maximum concentration is:
(0.15 mg/1) (4.2) = 0.63 mg/1
The maximum 30-day average concentration is:
(0.15 mg/1) (1.3) = 0.20 mg/1
Nickel
The guidance for nickel is based on the estimated
achievable long-term average concentration of 0.40 mg/1
from Table 8-13. The variability factors used for
chromium are employed again.
The 24-hour maximum concentration is:
(0.40 mg/1) (4.2) = 1.7 mg/1
The maximum 30-day average concentration is:
(0.40 mg/1) (1.3) = 5.2 mg/1
Antimony
The guidance for antimony is based on the
literature-based treatability estimate in Table 8-11
which is higher than the industrial wastewater
treatment system performance data from Table 8-12. The
745
-------
value of 0.80 mg/1 for the lime/settling technology is
used as a long-term average. The variability factors
used for chromium are employed again.
The 24-hour maximum concentration is:
(0.80 mg/1) (4,2) = 3.4 mg/1
The maximum 30-day average concentration is:
(0.80 mg/1) (1.3) = 1.0 mg/1
Basis for BCT Effluent Limitations
While EPA has not yet proposed or promulgated a revised BCT
methodology in response to the American Paper Institute v. EPA
decision mentioned in Section 3, EPA is promulgating BCT limitations
for this subcategory. These limits are identical to those for BPT.
EPA is not promulgating any more stringent limitations since we have
identified no technology option which would remove significant
additional amounts of conventional pollutants. As BPT is the minimal
level of control required by law, no possible application of the BCT
cost tests could result in BCT limitations lower than those
promulgated in this regulation. Accordingly, there is no need to wait
until EPA revises the BCT methodology before promulgating BCT
limitations.
Basis for BAT Effluent Limitations
A. 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 and toxic pollutant removal based on the cost
estimates presented in this report and the proposed Development
Document (60). The regulations being promulgated for BAT consist
of Level 1 or BPT treatment. The removal of additional
pollutants by Levels 2 and 3 treatment systems is not sufficient
to offset the additional cost to install and operate these
advanced treatment systems.
B. Technology Basis
The BAT treatment system is the same as that described above for
BPT.
C. Flow Basis
The model plant flow developed for BPT treatment applies also to
BAT in the development of the regulations. Therefore the value
of 1.5 mVkkg of product is used for the unit flow.
746
-------
D, Selection of Pollutants to be Regulated
For the BAT regulations, the Agency is setting the regulation of
COD and the same two toxic metals considered for BPT limitations
listed in Table 24-16. Concentration values for the other metals
are intended for guidance only.
E. Basis of Pollutant Limitations
1. Nonconventional Pollutants
The only nonconventional pollutant is COD in the sodium
bisulfite subcategory. Since BAT has been set equal to BPT
by the Agency, the limitation is then identical to BPT for
COD. Refer to Table 24-16 for the BAT regulation.
2. Toxic Pollutants
The Agency is setting limitations on chromium and zinc which
equal those for BPT. See above for the development of these
limitations.
Basis for New Source Performance Standards
The NSPS limitations {applicable to. pH, TSS, COD and two toxic metals)
are set equal to BAT for toxic and nonconventional pollutants and BPT
for conventional pollutants. Table 24-16 for the BPT and -BAT
limitations would be identical in all respects with NSPS limitations.
See above for the development of the regulations.
Basis for Pretreatment Standards
Pretreatment standards are necessary because NSPS provides better
removal of chromium than is achieved by a well-operated POTW with
secondary treatment installed and therefore chromium would pass
through a POTW in the absence of pretreatment. Based on the average
raw waste loads given in Table 24-9 and the long-term average
subcategory performance given in Table 24-16, NSPS treatment reduces
COD by 76% and zinc by 77%. A POTW will reduce COD by 80% and zinc by
76% (See Fate of Priority Pollutants in Publicly Owned Treatment Works
- Interim Report, EPA-440/1-80-301, October 1980). A POTW therefore
achieves a percent removal equal to or greater than that achieved by
NSPS for COD and Zinc, and therefore there is no pass through of those
two pollutants in this subcategory.
A. Existing Sources
The Agency is excluding this subcategory from Pretreatment
Standards for Existing Sources (PSES) under the provisions of
paragraph 8{b)(ii) of the Settlement Agreement. The total toxic
metal discharge from the one existing indirect discharger to POTW
is 120 pounds per year, which is so insignificant as not to
justify developing a national standard.
747
-------
B, New Sources
For Pretreatment Standards for New Sources (PSNS), the Agency is
promulgating limitations based on NSPS. The pollutant limited is
chromium.
748
-------
SECTION 25
SODIUM HYDROSULFITE (FORMATE PROCESS) INDUSTRY (Excluded)
Summary of Determinations
We proposed BPT, BCT, and BAT limitations and NSPS, PSES, and PSNS for
this subcategory. The proposed regulation basically added control of
selected toxic metal pollutants to existing treatment practiced in the
industry. We have reviewed the basis for the proposed regulation and
we concluded that the total current treated discharge load of only
0.42 pounds per day total toxic metals from all plants in the
subcategory is too insignificant to justify developing a national
regulation. Accordingly, we have excluded this subcategory from
national regulation development under paragraph 8(a)(iv) of the
Settlement Agreement.
The information presented in the remainder of this Section is provided
as guidance for use by permit writers.
Industry Profile
General Description
Most of the sodium hydrosulfite produced in the U.S. is sold in the
merchant market. Sodium hydrosulfite is used extensively in dyeing
cotton and in the printing industry. It is a powerful reducing agent
and is used in wood pulp bleaching, and stripping operations in the
food, vegetable oil, and soap industries.
The industry profile data are presented in Table 25-1, while status of
regulations are summarized in Table 25-2.
Subcategorization
A detailed summary of factors considered in Subcategorization is
presented in Section 4. Sodium hydrosulfite is produced by three
processes including the Werbs, formate, and zinc processes. The zinc
and Werbs processes are deferred to Phase II of the inorganic
chemicals regulation development effort. This section concerns the
formate process only.
General Process Description and Raw Materials
In the formate process, sodium hydrosulfite is produced by reacting
sodium formate solution, sodium hydroxide solution, and liquid sulfur
dioxide in the presence of a recycled stream of methanol solvent. The
general reaction is:
HC02Na + 3NaOH + 3S02 = Na2S204 + NaHC03 •*- Na2S03 + CO + 2H20 (1)
749
-------
TABIE 25-1. SOECKSECOfS PRCFILE DATA SIMIRRf
SUBCKBSGQEV
SODIUM HQSROSUCFETE (FQEWKEE PBCCESS)
Total subcategory capacity rate
Total subcatagory production rate
Number of plants in this subcategory
308 Data on file for
With total capacity of
With total production of
Representing capacity
nepmnni hij ptt xlucti on
Plant production range t
Maxinun
Avaraga prcduction
Median production
Avwrag* capacity utilization
Plant ag« ranges
Mixiinua
Maxinun
Wutetater flow range:
Maxinua
Vblun per unit product:
Hininun
Maximum
40,340 kkg/y«ar
39,940 Wcg/year
2
X
20,450 kfcg/year
20,450 kkg/year
50 percent
51 percent
NA
100 percent
NA
273 cubic raters/day
NA
KA
4.68 cubic rosters/kkg
NA
NA
Sources of data are Stanford Research Institute, Directory of
Producers, U.S.A., 1977, U.S. Department of Qcrnnarce, 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,'
torch, 1980.
NA • Not Available
750
-------
TABLE 25-2. STATUS OF REGUTAUCNS - EFFLUEKP TJMTTWPTrW GUIDELINES
SOPCATFXjQKg
SUBPAET
Sodium Hydrosulf ite
BE (40 CFR 415.570, 5/22/75}
STANDARDS
Product
Process
BPCTCA
Max.1 Avg,
EATEA
Max. Avg.
Para-
meters
(mg/1) (mg/1)
kg/kkg
(ng/1) Ong/1)
NSPS
Max* Avg,
kg/kkg kg/ldcg
(mg/1) (mg/1)
Sodium
Hydro-
Sulfite
Reserved
Reserved
Reserved
Reserved
= Maximum of any one day.
Avg. » Average of daily values for thirty consecutive days,
751
-------
The operation occurs in several steps:
An aqueous solution of sodium formate is prepared and introduced
into the reactor.
The recycled stream of methanol
introduced into the reactor.
containing sulfur dioxide is
The sodium hydroxide and sodium formate solutions, liquid sulfur
dioxide, and recycled methanol are then contacted under pressure
at slightly elevated temperatures.
Sodium hydrosulf ite then precipitates and forms a slurry in the
reactor. The by-product, sodium sulfite, and sodium bicarbonate and
carbon monoxide gas are formed.
There is a small amount of methyl formate produced in the reactor as a
side reaction between the sodium formate and methanol:
HC02Na + CH3OH = HCOZCH3
NaOH
(2)
This side reaction product remains in the recycling methanol during
the entire process. As a result, some of the methanol must be
periodically purged from the recycle system to avoid excessive buildup
of this impurity,
The resulting slurry of sodium hydrosulfite in the solution of
methanol, methyl formate, and by-products is sent to a pressurized
filter operation which recovers the crystals of sodium hydrosulfite.
The crystals are dried in a steam heated rotary drier, recovered and
packaged. The filtrate and backwash liquors from the filter operation
are sent to the solvent recovery system as is the vaporized methanol
from the drying operation. The drying of the sodium hydrosulfite
filter cake must be done very carefully as it is heat sensitive and
tends to decompose to sulfite.
A general process flow diagram for Plant 1672 can be found in Figure
25-1, as it is typical for this subcategory,
Water Use and Waste Source Characteristics
Water Use
Water is used in the process as make up for the reaction solutions and
for steam generation in the rotary dryers. Water is also used for
noncontact cooling in the reactor gas vent scrubbers and dryers, as
well as pump seals and washdowns, and as dilution water in the
wastewater treatment system to assist in biological oxidation of
organic materials.
752
-------
GASEOUS
BY-PRCCUCTS
VENT GAS
SCRUBBER
LIQUID
DICMIEC
SODIUM FORMRTE SOLUTION
-4
U1
W
SODIUM HYDPDXIDE SCUTHON
REC*ClH>"MErmNCI,, SO. AND
RECTOR
FORMATE
BftCKNftSH
LIQUOR
80DHM HYDROSULFITB
VtWT SVS
8CRJBBER
DISnUJBR
OOUMN
RBCOVERY)
VAPORIZED
MEIHANX
DRYER
OOUfW BOTTCMS
TO K*D)
PHCOUCT
Figure 25-1. General process flow diagram at plant #672.
(Sodium hydrosulfite manufacture.)
-------
Waste Sources
The strongest process waste is the aqueous residue from the
distillation column bottoms (solvent recovery system). This waste
contains concentrated reaction by-products and is purged from the
system at a rate of approximately 14,000 gallons per day. At Plant
#672, this waste is sent to a by-product pond where it is held and
either sold to the pulp and paper industry or bled into the treatment
system.
The dilute wastes from the process are contributed by leaks, spills,
washdowns, and tank car washing. At Plant #672, this is collected in
a sump and then sent to the biological treatment system.
Cooling tower and boiler blowdown constitute a noncontaminated
wastewater source. This is sent to the final compartment of the
chlorine contact tank without treatment, for discharge with the
combined effluent of the treatment plant.
The vent gas scrubbers create a wastewater source which is sent to the
methanol recovery stills for recycle. At Plant #672, this waste
eventually goes to the by-product pond with the distillation column
bottoms.
Table 25-3 presents the unit flows for the three primary sources of
process wastewater which contribute to the pollutant load.
Solid wastes currently are generated in the activated sludge waste
treatment system. An estimated 2,400 gallons of biological sludge are
discharged per day to an on-site drying bed. Application of more
stringent waste treatment of toxic pollutants is estimated to generate
an additional 6 kg/kkg of product of solid waste which must be
disposed of at an approved site.
Description of Plants Visited and Sampled
Screening and Verification
The only plant visited during the sampling program was Plant #672,
where verification sampling procedures were used. Plant #672 is one
of two plants that currently utilize the formate process in the sodium
hydrosulfite subcategory. An evaluation of plants that currently
utilize the zinc process for sodium hydrosulfite manufacture has been
deferred to a later phase of regulation development by the Agency.
Data from Plant #672 can be considered representative of this process
for both plants, since the other plant in this subcategory has an
identical, though slightly smaller, production process. However, the
second plant has a different waste treatment system. It also receives
large loadings of waste from several other products. Because of this
the plant is considered nonrepresentative of the hydrosulfite process
and visits were limited to Plant #672 for this reason.
754
-------
TABLE 25-3. WASTE SOURCE DATA AT PLANT #672
WASTE SOURCE
FLCW
(m3/kkg)
Dilute Waste (spills, etc.)
Dilution Water (contact)
By-product Waste
1.95
1.75
0.95
Ttotal 4.65
(Basis of flow for model plant and regulation development)
755
-------
in
KXU1H
mDBuurm
0mtM *
•mum
IWMHi*!,
n
nun
nun
riHlUIU MO tNK
Mawa went
tMRAUDQB
•r-wntcf
J-
CHUBB VM1
MDmURMOBNI
mugMi
u*T
oujam
M cuuorn*
Mci,oaNacr
^
•*•»««.
14
Figure 25-2. General process flew diagram at plant 1672 showing the sanpling points.
(Sodium hydrosulfite manufacture.)
-------
A general flow diagram of Plant # 672 showing process waste sources
and sampling points is shown in Figure 25-2. The sources of
wastewater for each sampling point are as follows:
By-product pond.
Dilute waste from sodium hydrosulfite process area and sumps.
Combined influent to treatment. This point collects waste from
points 1 and 2, plus the sodium bisulfite waste stream.
Treated effluent at the outfall.
At the time screening sampling was conducted at Plant #672, none of
the by-product wastewater was being sent to the biological treatment
system. As a result, the sodium hydrosulfite process waste being
treated was from the dilute waste area only.
Table 25-4 presents the results of the conventional and
nonconventional pollutant concentrations and unit loads for each of
the streams sampled. The results are based on three 24-hour composite
samples. It should be noted that sampling was done during a time when
no by-product waste was entering streams 3 and 4. The unit flow
indicated is the estimated flow observed during sampling.
Toxic Pollutant Concentrations
Toxic pollutants were identified in the raw process waste stream at
Plant #672. The following toxic pollutants were found at detectable
concentration levels.
Maximum Raw Waste Concentrations Observed
d.g/1)
Pollutant
Arsenic
Cadmium
Chromium
Copper
Lead
Nickel
Silver
Zinc
Pentachlorophenol
Phenol
Cyanide
Mercury
Selenium
Verification
Plant #672
79
43
9300
1500
1300
1700
130
27000
580
170
100
28
34
Two toxic organic pollutants, pentachlorophenol and phenol, were
identified at low, but detectable, concentration levels. The raw
process materials, including sodium formate and methanol, are likely
sources. The sodium formate currently used in the process is a
757
-------
TABLE 25-4. FLOW, POLLUTANT CONCENTRATION, AND LOAD DATA OF THE SAMPLED
WASTE STREAMS FOR PLANT #672 PRODUCING SODIUM HYDRDSULFITE
Flow TSS COD
Stream ~
Designation Description (m /kkg) (mgA) (kgA^g) (mg/1) (kg/kkg)
1
2
3
4
By-product 0.95 61 0.058 78,000 74
Dilute Waste 1.95 260 0.51 15,000 29
Dilute Haste and /,,
SBS Waste 2.05 840 1.7 16, 000 ^ 32
Final Discharge 4.87 25 0.12 740°-^ 3.6
(1) Value is that observed during sampling which may differ significantly
if the by-product stream is contributing.
758
-------
by-product from an unrelated organic chemicals process which may
contain the organic impurities. Methanol is also a suspect source of
organic impurities in view of the difficulty involved with its
purification and high degree of solubility with pentachlorophenol.
Also possible is coincidental formation of pentachlorophenol in the
process due to the presence of specific chlorinated hydrocarbons under
conditions conducive to its development.
Section 5 of this report describes the methodology of the screening
and verification sampling program. In the sodium hydrosulifte
industry, a total of three days of sampling were conducted at Plant
#672. Three 24-hour composite samples were taken at four different
sampling points. The sampling involved 169 analytical data points for
the toxic inorganic pollutants and 387 additional points for the toxic
organic pollutants. 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
mg/1 (Note: kg/m3 = 1000 mg/1), and
in units of
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 sodium
hydrosulfite production rate, the waste stream flow rate, and the
measured pollutant concentration.
Unit loading (as kg of pollutant = (C)(Q)
per kkg of sodium hydrosulfite) 1000P
where C and Q are the same as described above, and P is the
sodium hydrosulfite production rate expressed in units of
kkg/day. (kkg is 1000 kg, a metric ton, which is equal to 2205
Ibs.)
Table 25-5 presents the average toxic pollutant concentrations
observed during sampling for the raw and treated wastewaters at Plant
A;b72. The concentration indicated is based on three'24-hour composite
samples. Table 25-6 is a tabulation of the unit loadings for each of
the toxic pollutants found at detectable levels in the raw waste
water.
The estimated total toxic pollutant raw waste loads generated each
year for this subcategory were based on the total estimated annual
production of sodium hydrosulfite. The loads are as follows:
759
-------
Pollutant
Waste Load (kg/year
Arsenic 4.8
Cadmium 1 .3
Chromium 22
Copper 7.6
Lead ' 40
Nickel 64
Silver 6.4
Zinc 960
Pentachlorophenol 33
Phenol 6.0
Cyanide 1.56
Mercury 0.80
Selenium 1.2
Pollution Abatement Options
Toxic Pollutants of Concern
Although sodium hydrosulfite is being manufactured by both the zinc
process and the formate process, the trend is away from the zinc
process for environmental reasons. This discussion concerns only the
formate process, using a sodium formate feed stock from a source which
appears to contain heavy metal impurities (chromium, zinc, nickel,
lead, and copper) as well as trace amounts of cyanide. A predominant
characteristic of sodium hydrosulfite wastes is their high chemical
oxygen demand resulting from various forms of sulfite, from methyl
formate,, and from residual methanol after a solvent recovery process.
Low levels of phenolic compounds are also found in the raw wastes.
Prevailing Control and Treatment Practices
Due to the nature of the two primary raw waste streams, each one is
handled differently. The dilute waste is first sent to a holding pond
where the flow is equalized and the waste mechanically aerated. This
pond also contains approximately 1500 gallons per day of waste from a
sodium bisulfite process. The pond effluent is pH adjusted with
sulfuric acid and sent to an aeration basin. A nitrogen-phosphate
fertilizer and urea are added to provide nutrients. Approximately
3500 gallons per day of sanitary waste and up to 25,900 gallons per
day of clean dilution water are also added to the aeration basin.
This basin formerly had mechanical aerators, but now has air diffusers
which allow better temperature control for biological oxidtion. The
effluent from aeration goes to a clariifer. Approximately 14,000
gallons per day of settled sludge is returned to the aeration basin
and 2,400 gallons per day is sent to drying piles on site. More
dilution water is added to the clarifier when needed for Total
Dissolved Solids control. The overflow from the clarifier goes to a
chlorine contact tank because of the sanitary waste. The blowdown
water from the'cooling tower and boilers is added to the final chamber
of the chlorine contact tank. The effluent from this unit is sent to
a final polishing pond for settling and equalization before discharge.
762
-------
The by-product waste from the distillation column bottoms is sent to a
lined by-product pond at a rate of 14,000 gallons per day and held for
one of two possible disposal methods. When there is a market for the
by-products, the waste is concentrated and sold to the pulp and paper
industry. At times when this is not possible, and the pond nears
capacity, the waste is bled into the treatment system described above
through the dilute waste holding pond.
Advanced Treatment Technologies
Practical technologies for controlling COD include various forms of
mechanical and biological oxidation. For the relatively simple
chemical oxidation of hydrosulfite to sulfate, intimate contact with
atmospheric oxygen is effective, using submerged air diffusers,
induced air in a circulating system, or mechanical surface aeration.
For biochemical oxidation of resistant organics such as formates,
phenols, chlorinated hydrocarbons, and methanol, the use of trickling
filtration, rotating biological discs, or variations of the activated
sludge process can provide intimate contact between organic pollutants
and the microbiological organisms which use them as food.
Technologies for controlling heavy metals include alkaline
precipitation, which is effective for the common heavy metals, and
sulfide treatment, which precipitates nickel, zinc, and copper, but
does not increase control of chromium. Other less appropriate metal
removal techniques have been discussed in Section 8.
Selection of Appropriate Technology and Equipment
Technologies for Different Treatment Levels
A. Level 1
Treatment system pH adjustment, biological oxidation, settling,
and chlorination are used to reduce COD and coliform organisms in
the combined wastes, in accordance with existing plant practice.
The flow diagram is shown in Figure 25-3.
B. Level 2
The by-product wastes are treated separately by alkaline
precipitation to remove the toxic metals and then are combined
with the product wastes for biological oxidation treatment and
chlorination, as in Level 1.
If an acutal formate process plant employs metal-free sodium
formate in its process there is no reason to expect heavy metals
in the process wastes and Level 2 treatment should not be
necessary. The flow diagram is shown in Figure 25-4.
763
-------
Equipment for Different Treatment Levels
A. Equipment Functions
Product waste and by-product wastes are received in a mixed and
aerated equalizing basin, adjusted to a neutral pH, and aerated
in a four-day aeration lagoon, including 50 percent return of
underflow to the influent. Plant sewage, nutrients, and diluting
water are added to the lagoon to promote biological oxidation of
organics and other COD. Lagoon effluent is clarified,
chlorinated, and sent to a polishing pond before discharge
through effluent monitoring facilities. Cooling tower and boiler
blowdown wastes enter the system after chlorination, since they
require no treatment except settling of scale and inert debris in
the polishing pond. Floating aerators are used in the
equalization basin and compressed air is diffused in the aerated
lagoon, for mixing and introduction of dissolved oxygen into the
mixed liquor.
In Level 2 treatment, by-product wastes are received in a
separate 18-hour aerated and recirculated holding tank, which is
pumped at average daily flow to a gravity clarifier, adding
sufficient lime to reach a pH of 10.5. The clarifier overflow
joins the product waste stream in the equalization basin of the
Level 1 system. All features of the Level 1 system remain the
same, since it was originally sized to handle the combined
wastes.
B. Chemicals and Handling
SuIfuric acid, lime, filter aid, and chlorine are chemicals
commonly used in waste treatment. When handled in corrosion
resistant equipment designed for their use, no unusual hazards
are expected. Raw sewage and 10-10-10 liquid fertilizer
introduced into the aerated lagoon become thoroughly mixed and
are eventually consumed in the biological oxidation process,
constituting no threat to operating personnel. Chlorine, used
for control of coliform bacteria, is received in ton containers
and applied as a chlorine water solution using standard solution
feed chlorination equipment. There are no unusual chemical
handling problems in treating these wastes, provided the waste
streams are kept at a neutral or alkaline pH.
C. Separation and Disposal of Solids
In the Level 1 system, waste activated sludge solids are assumed
to be dried in sludge beds at the site, to be'used as fertilizer
for plant landscaping. Clarifier underflow from alkaline
precipitation of by-product waste in Level 2 is assumed to be
sent to a sludge holding tank and dewatered at suitable intervals
in a filter press, followed by hauling of solids to a chemical
landfill. Filter press filtrate is returned to the holding tank
for retreatment.
764
-------
a
M-ncnicr
tAnB WlTHI
I MnuMf uuinKH Nvm
1 All L
JL / ' ' j—\ •** /—i
-J"
OMtPIM
CMJIB MM)
MUMMMXMI
RtWDOIOD
•vumr
ui
IneludM Elow Monitoring* ffl Monitoring atd
Figure 25-3. Level 1 waste water treatment for sodium hydrosulfite subcategory.
-------
-j
a*
o*
EitnuuLT
HV'llK
1
ute
a
I K-
UB I •
*1 I '
I~?H
• i
W-HCOET WSfE MUER
uoumenMc
saxx
UCUtDG
USIUJUTION
fnclutkfa flaw iDniL^ring, (41 wutoring and ia^>l^r.
"t
/ 1
rtmatwr UILJT
JjJ
CM VKHK
AJ*
AB*TS1 IMXKM
j i
T
OBUWMMICM
—a-A
HiJSJIItC li»ii
y^
•W9W SUWGC
Figxnre 25-4. Level 2 waste water treatment for sodiun hydrosulfite subcategory.
-------
D. Monitoring Requirements
Internal monitoring should include simple field tests for pH,
chlorine residual, and settleable solids. Maintenance of the
by-product stream clarifier at a pH of 10.5 is expected to
provide control of toxic metals without need for routine metal
analyses. Periodically, effluent samples should be analyzed for
chromium, zinc, copper, nickel, and lead by atomic absorption in
addition to routine COD tests for general evaluation of the
treatment.
Treatment Cost Estimates
General Discussion
A model plant concept was prepared for the purpose of the cost
estimates. The specifications of the waste input parameters and the
design of the model plant Level 1 treatment system are based on the
information presented for Plant #672.
In this subcategory, commercial fertilizer and urea are added to
stimulate growth of the biomass employed in biological treatment, and
not for direct reaction with a residual pollutant. Therefore, the
chemicals used do not bear a fixed relationship to the plant
production in units of sodium hydrosulfite.
Organic solids generated in the model treatment system are assumed .to
be disposed of on land at the site, without a separate cost for sludge
disposal.
Cost Estimates
The model plant cost estimate for two levels of treatment applied to
•the same level of production is presented in Table 25-7. Table 25-8
gives a summary of the unit cost distribution between amortization,
operation and maintenance cost components at two levels of treatment.
Cost estimates developed for the first and the second level of
treatment indicate that labor and supervision costs constitute a major
portion of the annual cost. This reflects the manpower requirements
for operating the treatment systems on a 24-hour basis.
Basis for Guidance
Evaluation of Level 1 Treatment Practices
There are two plants producing sodium hydrosulfite by the formate
process, both of which have Level 1 equipment in place and both are in
compliance with their NPDES permits.
Level 1 technology has been specified as the technology presently in
use at Plant #672. Design and cost estimates are based on inclusion
of by-product wastes.
767
-------
TABLE 25-7. MODEL PLANT TREATMENT COSTS
Subcategory
Production
Sodium Hydrosulfite - Formate process
20,450 metric tons per year
A. INVESTMENT COST
Site development
Equipment
Monitoring equipment ..
Subtotal ,.....
Contractor's 0 & Pb....
Subtotal
Engineering
Subtotal
Contingencies
Subtotal
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
Level 1
23,000
138,500
20,000
181,500
27,225
208,725
41,745
250,470
25,047
275,517
12,000
287,517
TOTAL OPERATION AND
MAINTENANCE COST
168,000
12,000
3,500
27,552
8,626
0
15,000
234,677
($)
Level 2
0
121,000
0
121,000
18,150
139,150
27,830
166,980
16,698
183,678
3,000
186,678
84,000
1,200
18,500
18,368
5,600
0
7,500
135,168
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
44,827
279,504
29,884
165,053
a Represents the incremental cost above that for Level 1 treatment
Overhead and profit
768
-------
TABLE 25-8 MODEL PLANT UNIT TREATMENT COSTS
Subcategory Sodium Hydrosulfite -Formate process
Annual Treatment Costs ($/kkg)
COST ITEM
PRODUCTION
(kkg/yr)
LEVEL OF TREATMENT
Level 1 Level 2
Annual Operation
and Maintenance 20/450
Annual
Amortization 20,450
Total Annual
Cost 20,450
11.48 6.61
2.19 1.46
13.67 8.07
*Represents the incremental cost above Level 1.
769
-------
An evaluation of treatment practices was performed at Plant #672 based
on the pollutant sampling, since long-term monitoring data was not
available for the pollutants of concern. Details concerning the
performance evaluation calculations and assumptions are discussed
below for^the pollutants of concern.
A. Conventional and Nonconventional Pollutants
1. Chemical Oxygen Demand (COD)
At the time of sampling, the by-product waste (stream #1,
Figure 25-2) was not flowing into the wastewater treatment
system (Stream #3, Figure 25-2).
Review of Table 25-4 indicates that a majority of the COD
load is contributed by the by-product waste stream. The
other major source of COD is contributed by the dilute waste
stream #2. Estimates of subcategory performance are made
for COD based on the following assumptions:
a. Assumption 1
The COD load for the by-product stream must be included
in the evaluation of the treatment system performance,
since its contribution to the final COD load will have
a significant influence. Therefore, it is assumed that
the percent COD removed in the treatment system would
be the same percent COD removed for the by-product
waste stream as if it had received treatment. This
assumption is necessary since plant performance
information is available only for the dilute waste
stream.
b. Assumption 2
In order to estimate the COD removed during treatment
in the dilute waste stream, two minor assumptions must
also be made to account for COD contributions from the
sodium bisulfite (SBS) and sanitary streams which are
not considered sodium hydrosulfite process-related. It
is assumed that the final COD concentration for the
treated sodium bisulfite waste stream is 680 mg/1 (from
Table 24-16) and 60 mg/1 which is a conservative
estimate for treated sanitary wastes. These
assumptions are not critical since the total combined
waste flow from these two waste sources is only 0.30
mVkkg compared with 2.9 mVkkg of other process
related wastes.
Table 25-9 is a summary of the subcategory performance
evaluation of TSS and COD for Plant #672. The COD
evaluation is developed in the table on the bases of
770
-------
the assumptions above and sampling information in Table
25-4.
A determination of achievable effluent COD load is
shown in the following steps beginning with an
estimation of the COD removal efficiency in the waste
treatment system.
The observed effluent COD load is 3.6 kg/kkg from Table
25-4 which includes contributions from the sodium
bisulfite (SBS) and sanitary waste streams which are
not process related. These loads are determined as
follows:
SBS load = (680 mg/1) (0.10 mVkkg) (kg/m3) = 0.068 kg/kkg
(1000 mg/1)
(0.10 m3/kkg from Table 25-9; 680 mg/1 from Assumption
2 above)
Sanitary waste COD load =
(60 mg/1) (0.24 mVkkg) { kq/m3) = 0.014 kg/kkg
(1000 mg/1)
(0.24 mVkkg from Table 25-9; 60 mg/1 from Assumption 2
above)
The effluent COD load contributed by the SBS and
sanitary waste streams are substracted from the
observed load of 3.6 kg/kkg to obtain the actual COD
load contributed by the process related dilute waste as
follows:
3.6 kg/kkg - (0.068 kg/kkg +. 0.014 kg/kkg) =3.5 kg/kkg
The effluent COD load is 3.5 kg/kkg which when
expressed as a ratio with the raw COD waste load
(Assumption 1) can be used to estimate the additional
COD contributed by the by-product waste as follows:
Raw COD load contributed = 29 kg/kkg from Table 25-4
by dilute waste
Raw COD load contributed = 74 kg/kkg from Table 25-4
by by-product waste
Effluent COD load of = (74 kq/kkq) (3.5 kg/kkg)
(29 kg/kkg)
=8.9 kg/kkg
771
-------
TABLE 25-9. SOBCA3EGORY PERFORMANCE EVALUATICN SUM4ABY AT PLANT #672 FOR
CONVENTIONAL AND tCNOCNVENTXQNAL PCLLUTANTS IN THE EFFLUENTS
Effluent Haste Flow TSS
Description (m3/kkg) fno/ii
A - Dilute Haste
B - Sodium Bisulfite
Haste
C - Sanitary Haste
D - Dilution Hater
E - Boiler Slowdown
F - By-product
Total load (A+O+F)
Effluent Concentration
Model Plant Flow
(A4O+-F)
Concentration At
Model Plant Flow
BASIS OF UMTTATION
(1) - NA Not applicable
1.95
0.10
0.24
1.75
0.83
0.95
4.87
4.65
4.65
4.65
260
NA(1)
NA
NA
NA
61
NA
25
NA
26
25
(2)
(kq/kkg)
0.51
NA
NA
NA
NA
0.058
NA
NA
NA
NA
NA
OOD(2>
(rog/1) (kg/kkg)
1800 3.5
680 (3* 0.068
. .
60 0.014
0 0
0 0
9600 9.2
NA 13
NA NA
NA NA
2700 NA
NA 12
to evaluation.
(2) - Data based on average of
three
(3) -
24- hour composite sanples.
Assumed value discussed in Section 25.7.1 under Chemical Oxygen Demand,
772
-------
Total effluent COD load contributed by both the dilute
and by-product waste =
3.5 kg/kkg +8.9 kg/kkg = 12 kg/kkg
The effluent load for COD is 12 kg/kkg based on the
plant performance evaluation and sampling data.
2. Total Suspended Solids (TSS)
The removal of TSS from the raw wastewater is much more
complex on a load basis. TSS removal must therefore be
estimated on a concentration basis as indicated in Table
25-9. A TSS concentration of 25 mg/1 was observed in the
treated effluent during sampling (Table 25-4) which is used
as the long-term average in developing the TSS guidance.
B. Toxic Pollutants
The removal of toxic pollutants in the treatment system at Plant
#672 during sampling is indicated in Table 25-5 for the purpose
of evaluating plant performance.
Basis for Level 1 Treatment Performance
A. Technology Basis
The technology basis is, or is equivalent to, equalization, pH
adjustment, aeration in a biological oxidation. system,
clarification, and chlorination before discharge of the treated
effluent.
B. Flow Basis
The basis of flow used for the cost estimates, and as a basis to
estimate pollutant discharge loadings for the purpose of guidance
development, was derived from plant information received at Plant
#672. Table 25-3 presents the unit flows from the three primary
waste sources identified in the industry. The dilute and
by-product wastewaters are primarily process related, whereas the
dilution water is required for proper operation of the biological
waste treatment system.
There are only two plants which currently use the formate process
for the manufacture of sodium hydrosulfite. The model plant flow
is 4.7 m3/kkg of product for the sodium hydrosulfite subcategory
as presented in Table 25-3 and is based on Plant #672 data.
Plant #672 was chosen for evaluation because it is not
complicated by other unrelated manufacturing processes.
773
-------
Selection of Pollutants to be Controlled
The Level 1 treatment technology is directed primarily toward
removal of TSS and COD. In addition to these conventional and
nonconventional pollutants, toxic organic pollutants were
identified . in small quantities. These toxic organic polluants
include pentachlorophenol, phenol, and other trace organics. The
presence of these toxic pollutants is currently under
investigation by the plant to: 1) determine the source of the
pollutants and identify whether they are process related, and 2)
determine whether process modifications or best management
practices might be available to eliminate their presence if they
are discovered to be process related.
Basis of Pollutant Control Guidance
1 . Conventional and Nonconventional Parameters
PH
The treated effluent is to be controlled
range of 6.0 to 9.0. This is based
presented in Appendix B of the proposed
Document (60) and the JRB Study (52).
TSS and COD
within the
on the data
Development
The data presented in Table 25-9 was used for the
development of TSS and COD control guidance. The data
presented is for Plant #672 which is the only plant in
the subcategory where the treatment performance can be
observed clearly.
No long-term monitoring data is currently available for
a statistical estimation of the variability factors in
the Sodium Hydrosulfite Subcategory.
Therefore, the 30-day average and daily variability
factors are estimated from the treatability study (61)
for COD removal. The study indicates a 30-day average
variability factor of 1.2 and daily variability factor
of 4.0.
The maximum 30-day average COD level is estimated from
plant performance data in Table 25-9 to be 12 kg/kkg.
The maximum 30-day average concentration is determined
as follows:
(12 kq/kkg) (1000 mq/1) * 2600 mg/1
(4.7 mVkkg) (kg/m3)
774
-------
The long-term average COD concentration is determined
using the 30-day average variability factor of 1.2 as
follows!
2600 mq/1 = 2200 mg/1
1 .2
The 24-hour maximum concentration is determined using
the daily variability factor as follows:
(2200 mg/1) (4.0) = 8800 mg/1
The 24-hour maximum loading for COD is then:
(8800 mg/1)) (4. 7 mVkkg) (kg/in*) = 41 kg/kkg
(1000 mg/1)
The long term average TSS concentration is 25 mg/1 from
plant performance information in Table 25-9. The TSS
control guidance is based on this long-term average and
the variability factors used for COD. The 30-day
average variability factor of 1.2 is used to determine
the TSS maximum 30-day average concentration as
follows:
concentration is determined
(25 mg/1) (1.2) = 30 mg/1
The 24-hour max imum
similarly as follows:
(25 mg/1) (4.0) = 100 mg/1
The maximum 30-day average loading is determined as
follows:
(30 mg/1) (4.7 mVkkg) (kq/m3) =0.14 kg/kkg
(1000 mg/1)
The 24-hour maximum loading is:
(100 mg/1) (4.7 mVkkg}) ( kq/m3) =0.47 kg/kkg
(1000 mg/1)
Guidance for effluent control is presented in Table 25-
10.
2. Toxic Organic Pollutants
The verification sampling results presented in Table 25-5
for pentachlorophenol and phenol indicate that both of these
toxic organic pollutants are currently removed by the
existing treatment system to the analytical detection limit
and are therefore excluded from further consideration.
775
-------
T&BLE 25-10. GUIDANCE FOR EFFLUENT CONTROL
Sodium Hydrosulf ite
Level 1 Performance
Waste Water Flow: 4.7 m3/kkg
Pollutant
Conventional and
Subcategory ,,*
Performance VFR
(mg/1)
nonconventional
Concentration
Basis (mg/1)
Max
30-day 24-hr
Avg Max
Effluent Loading
(kg/Kkg)
Max
30-day 24-hr
Avg Max
pollutants:
Total Suspended
Solids, TSS
Chemical Oxygen
Demand, COD
25(2) 4.0/1.2(3) 30
2200 4.0/1.2(3) 2600
100
8800
0.14
12
(2)
0.47
41
(1) - VFR: ratio of the 24-hour variability factor to the 30-day average
variability factor.
(2) - Based on subcategory performance estimates utilizing three 24-hour
composite samples.
(3) - Based on treatability study (61) for the removal of COD.
(4) - No effluent limitation has been established.
776
-------
3. Toxic Metal Pollutants
The Level 1 treatment technology is not amenable to the
removal of toxic metal pollutants although, as shown in
Table 25-5, considerable incidental removal can occur.
Level 1 technology can not reasonably ensure the removal of
toxic metals on a consistent basis.
Basis for Level 2 Treatment Performance
A. The Application of Advanced Level Treatment
The Agency has analyzed the cost effectiveness of the base level
systems and the various advanced level options for the removal of
pollutants based on cost estimates presented in this report. No
plant has this additional technology installed.
B. Technology Basis
Level 2 is based on treatment that provides more stringent
removal of toxic pollutants in the by-product waste stream by
introducing alkaline precipitation with lime and settling prior
to base level treatment. The by-product waste stream was the
primary source of toxic metal pollutants observed during
sampling.
C. Flow Basis
The unit flow used for the treatment performance estimates is
based on 4.7 m3/kkg of product. The estimated flow does not
change for Level 1 and Level 2 treatment.
D. Selection of Pollutants to be Controlled
Selection of pollutants is based on an evaluation of verification
sampling data at Plant #672. Results of the sampling program are
presented in this section. On the basis of concentration and
total annual raw waste loads, zinc, nickel, lead, chromium, and
copper have been identified at treatable levels and were
considered as candidates for control in addition to TSS and COD.
Review of Section 8 indicates that not all toxic pollutants
present in raw waste effluents need to be specified for alkaline
precipitation. In fact, two primary groups of heavy metals
appear to reach a minimum solubility at slightly different pH
levels. Zinc is the major toxic pollutant determined at Plant
#672 which represents one of these primary groups of metals and
nickel, also identified at treatable levels, represents the other
group that achieves optimum removal at slightly higher pH values.
Adequate control of zinc and nickel can be achieved in a single
step alkaline precipitation system operated in an intermediate pH
range of approximately 9.5 to 10.5. The basis for this is
presented in Section 8 of this report. Control of zinc and
777
-------
nickel will also ensure removal of lead, chromium, and copper
which may occasionally occur at treatable concentrations in the
raw waste.
E. Basis of Pollutant Control Guidance
1. Nonconventional Pollutants
The only nonconventional pollutant selected for limitation
is COD. In view of the proposed technology for Level 2, no
additional removal of COD is anticipated beyond what is
already estimated for Level 1. That section discusses the
development of the COD control guidance.
The maximum 30-day average COD effluent loading is 12 kg/kkg
and the 24-hour maximum is 41 kg/kkg presented in Table
25-11.
2. Toxic Pollutants
Alkaline precipitation and settling of the by-product waste
is expected to remove the five candidate toxic metal
pollutants to within the limits of treatability. Review of
Table 25-5 indicates that the existing Level 1 treatment
system is providing incidental removal of the toxic metals.
Table 8-12 presents the estimated levels achievable for the
toxic metal .pollutants based on literature treatability
which was used for the purpose of establishing the long-term
average concentration used for guidance control.
No long-term pollutant monitoring data were available to
develop the variability factors. However, treatability
studies (61) were performed that estimate the 30-day average
and 24-hour maximum variability factors for a number of
toxic pollutants. Variability factors for zinc were
estimated as the average of values obtained in the nickel
and copper sulfate industries. The 30-day average and daily
variability factors are 1.2 and 4.1, respectively, and are
used for guidance development.
a. Zinc
Review of the zinc concentration in the raw by-product
waste stream indicates levels as high as 27 mg/1 and an
average of 24 mg/1 from three 24-hour composite samples
(Table 25-6). Literature treatability presented in
Table 8-12 indicates an achievable long-term average
concentration of 0.80 mg/1 for alkaline precipitation
and settling of zinc. The maximum 30-day average
concentration is developed as follows:
(0.80 mg/1) (1.2) = 0.96 mg/1
778
-------
TABLE 25-11. GUIDANCE FOR EFFLUENT CONTROL
Sodium Hydrosulf ite
Level 2 Performance
Waste Water Plow: 4.7 m3/kkg
Conventional and
pollutants ;
Chemical Oxygen
Demand, COD
Subcategory f 1 ^
(mg/1)
nonconventional
2200 4.0/1.2(5)
Concentration
Basis Cmg/1)
Max
30-day 24-hr
Avg Max
2600 8800
Effluent Loading
(kg/kkg)
Max
30-day
Avg
12(2)
24-hr
Max
41
Toxic pollutants;
Zinc
Nickel (6)
Lead(6>
^ . (6)
Chromium
Copper(6)
0.8010' 4.1/
0.40(3) 4. 1/
0.15(3) 4. I/
0.32(3) 4. I/
0.40(3) 4.1/
'1.2v:j' 0.96
1.2(5) 0.48
1.2(5) 0.18
1.2(5) 0.38
1.2(5) 0.48
3.3
1.6
0.62
1.3
1.6
0.0045
0.0023
f 4 )
f 4 1
_J4>
0.016
0.0075
— <4>
_J4)
— W)
(1) - VFR
(2)
ratio of the 24-hour variability factor to the 30-day
average variability factor.
Based on subcategory performance estimates utilizing three 24-hour
composite samples.
(3) - The long-term average concentration estimate from Table 8-12 is used.
(4) - No control specification is needed.
(5) - Based on treatability study (61) for nickel'and copper sulfate
subcategor ies.
(6) - Also applicable for pretreatment.
779
-------
The 24-hour maximum
developed as follows:
concentration is similarly
(0.80 mg/1) (4.1) = 3.3 mg/1
The maximum 30-day average effluent loading for zinc
is:
(0.96 mg/1) (4.7 mVkkg) (kq/m3) = 0.0045 kg/kkg
{1000 mg/1)
The 24-hour maximum loading for zinc is:
(3.3 mg/l)(4.7 mVkkg) (kq/m») = 0.016 kg/kkg
(1000 mg/1)
Guidance for effluent control is presented in Table
25-11.
Nickel
The concentration of nickel was observed as high as 1.7
mg/1 in the raw by-product waste stream and averaged
1.1 mg/1 in the three 24-hour composite samples (Table
25-6). Literature treatability presented in Table 8-12
indicates an estimated achievable long-term average
concentration of 0.40 mg/1. The maximum 30-day average
concentration is as follows:
(0.40 mg/1) (1.2) * 0.48 mg/1
The 24-hour maximum concentration is similarly
developed as follows:
(0.40 mg/1) (4.1) = 1.6 mg/1
The maximum 30-day average loading for nickel is:
(0.48 mg/1) (4.7 mVkkg) (kq/m3) = 0.0075 kg/kkg
(1000 mg/1)
The 24-hour maximum loading for nickel is:
(1.6 mg/1) (4.7 mVkkg) (kq/m3) = 0.0075 kg/kkg
(1000 mg/1)
Guidance for effluent control is summarized in Table
25-11 for Level 2 treatment.
Other Pollutants
The concentration basis for lead, chromium, and copper
are also presented in Table 25-11. These
780
-------
concentrations are intended to serve as guidance in
cases where these toxic pollutants are found to be of
serious concern.
780-A
-------
-------
SECTION 26
EXCLUDED SUBCATEGORIES
Aluminum Sulfate
Summary of Determinations
It has been determined that no further effort will be given to
developing revised BAT or NSPS for this subcategory. The basis for
this recommendation is that there is a zero discharge regulation in
effect for BAT and NSPS and it controls toxic pollutants.
Production Process and Effluents
Aluminum sulfate is produced by the reaction of bauxite ore with
concentrated sulfuric acid. Ground ore and acid are reacted in a
digester yielding aluminum sulfate in solution plus muds and insoluble
wastes. The aluminum sulfate is sold as a solution or evaporated to
produce a solid product. Waste muds are ponded to allow settling and
the clear liquor is returned to the process. Wastes from washing and
leaks are directed to the pond and also returned to the process.
Toxic pollutants in the pond include zinc, copper, chromium and
cadmium.
Plants
There are 82 aluminum sulfate producing facilities in the industry.
BPT Limitations
BPT limitations were promulgated March 12, 1974 (40 CFR 415.20). The
limitations provide for zero discharge of process wastewater except
that if the pond has sufficient volume to hold a 10-year, 24-hour
storm, the amount of water equal to the precipitation less the
evaporation may be discharged. The water must have a pH of 6.0 to 9.0
and average less than 25 mg/1 of suspended solids.
BAT, Pretreatment and NSPS Limitations
BAT and NSPS limitations were promulgated on March 12, 1974 (40 CFR
415.23 and 415,25). The limitations provide for zero discharge of
process wastewater except that if the pond has sufficient volume to
hold a 25-year, 24-hour storm, the volume of precipitation that falls
within the pond in excess of a 25-year, 24-hour storm may be
discharged. These zero discharge limitations adequately control the
toxic pollutants. Development of Pretreatment Standards has been
deferred to Phase II.
781
-------
Ammonium Chloride
Summary of Determinations
It has been determined that no further effort be given to developing
or revising BAT, NSPS, or Pretreatment regulations and the subcategory
is excluded under Paragraph 8 of the Settlement Agreement. The bases
for this determination are: 1) Only one of the major producers of
ammonium chloride uses the Solvay
recovered as a by-product. 2)
Process. Ammonium chloride is
No toxic pollutants were found at
significant concentrations
ammonium chloride plant.
in the waste during screening of one
Production Process and Effluents
Ammonium chloride is used in the manufacture of dry cell batteries,
explosives, dyes, washing powder, soldering flux, chemical reagent,
and as a medicinal additive to livestock feed. It is also used in
pharmaceutical preparations and freezing mixtures.
Ammonium chloride is produced by three methods. A major portion is a
by-product in the manufacture of sodium carbonate by the Solvay
process. The wastes produced are associated with the sodium carbonate
subcategory. A second process prpduces ammonium chloride by the
reaction of hydrogen chloride with ammonia. Discharges from this
process come from crystallization and wet scrubber operations.
The industry profile data for this
26.2-1.
Toxic Pollutants
subcategory are given in Table
Data have been received on about 50 percent of the industry as a
result of Section 308 letters. In additon, a sampling survey for
toxic pollutants was made at one plant. The results of the 308
letters and the sampling survey indicate that no toxic pollutants are
being discharged in significant quantities. Ammonia was found to be
the only pollutant of significance. Since ammonia is adequately
controlled by the existing BPT regulation 40 CFR 415.242, this
subcategory is being excluded under Paragraph 8 of the Settlement
Agreement.
Pollutants found during sampling at one plant are:
Pollutant
Chromium
Nickel
Zinc
Ammon i a
Concentration
29 ng/1 (max.)
25 cg/1 (max.)
29 *ig/l (max.)
104 mg/1 (avg.)
782
-------
TABLE 26.2 - 1
PROFILE DAIA SUMMARY
SUBCA2EGOEY
AMMONIUM CHLORIDE
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
Volume per unit product:
Minimum
Maximum
NA
NA
6
3
52,400 kkg/year
29,800 kkg/year
NA
NA
4,600 kkg/year
13,400 kkg/year
NA
NA
NA
17 years
43 years
NA
NA
NA
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S, Department of Ocmrerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chemical Industry, " June, 1978 and "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards for the Inorganic Chemicals
Industry," March, 1980
Not Available
783
-------
Status of Regulations
BPT regulations (40 CFR 415.242) are in effect for this subcategory.
Ammonium Hydroxide
Summary of Determinations
It has been determined that no further effort be given to developing
BPT, BAT, NSPS, and Pretreatment regulations for the Ammonium
Hydroxide Subcategory. The bases for this determination are: 1) the
process has no toxic pollutants as reactants, and 2) no direct process
waters are discharged from manufacturing operations. The subcategory
is excluded under Paragraph 8 of the Settlement Agreement.
Production Processes and Effluents
Ammonium hydroxide is used predominately as a chemical intermediary
and reagent. It is also used in the dyeing and bleaching of fabrics,
the production of ammonium salts and aniline dyes, and the extraction
of alkaloids from plants.
The most common method of ammonium hydroxide production is the
modified Haber-Bosch process, wherein hydrogen and nitrogen are
reacted directly over a catalyst surface to form ammonia. The
hydroxide is formed by adding water. The only process wastewater
source is derived from equipment washing.
The industry profile for this subcategory is given in Table 26.3-1.
Toxic Pollutants
Data was received on six of seven plants as a result of 308 letters.
In addition, a sampling survey was made at one plant which had a
potential for discharge. However, no process water discharge was
found at the facility. There are low volume discharges as a result of
spills and washdowns. The amount discharged was such that a sample
could not be obtained for analysis.
Status of Regulations
Because no significant quantities of toxic pollutants are present, no
further effort will be given to development of pretreatment
regulations for this subcategory. Subpart Y. has been reserved for
this subcategory.
Barium Carbonate
Summary of Determinations
It has been determined that no further effort be given to developing
or revising BPT, BAT, NSPS, and Pretreatment regulations for the
Barium Carbonate Subcategory. The basis for this determination is
784
-------
TABLE 26.3-1 -
SUBCMEGOH* PRCFUE DKTA SUMMARY
SUBCATEGORY
AMVDNIUM HYDROXIDE
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 flew range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
NA
NA
6
41,800 kkg/year
17,000 kkg/year
NA
NA
206 kkg/year
9,500 kkg/year
NA
NA
NA
10 years
26 years
NA
NA
NA
NA
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Connerce. 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 Qiemicals
Industry," March, 1980
N&= Not Available
785
-------
that the small quantities of toxic pollutants found during screening
are far below accepted treatability levels.
Production Processes and Effluents
Barium carbonate is used in glass manufacturing, as a flux in ceramics
and enamelling, as an intermediate in the production of barium oxide
and hydroxide, and as a coating for photographic paper. It is also
used in the synthetic dyestuff industry and for the removal of soluble
sulfate in brick manufacturing.
Barium sulfide solution is reacted with soda ash to precipitate barium
carbonate. The reacted solution is filtered. The filter cake is
washed, dried, and calcined. Wastewater results from filter cake
washing, leaks and spills. The industry profile data for this
subcategory are given in Table 26,4-1.
Toxic Pollutants
Data has been received on about 70 percent of the industry as a result
of Section 308 letters. In addition, a sampling survey for toxic
pollutants was made at one plant. The results of the 308 letters and
the sampling survey indicate that no toxic pollutants are being
discharged in significant quantities.
The maximum concentration found the raw waste load in sampling for
this subcategory were:
Pollutant
Nickel
Concentration. (»q/l
21
Zinc 68
Status of Regulations
Subpart Z has been reserved for this subcategory,
Borax
Summary of Determinations
It has been determined that no further effort will be given to
developing revised BAT and NSPS regulations for the Borax Subcategory.
The basis for this determination is that existing BPT regulations
specify zero discharge of process wastewater pollutants to navigable
waters. Development of pretreatment regulations is deferred to Phase
II.
Production Processes and Effluents
Borax is produced by dissolving sodium borate ores in recycled mother
liquors and water. The insolubles settle out in ponds or are removed
736
-------
TABLE 26.4-1 -
SUBCKTEGOEY PRCFILE DATA .SUMMARY
SUBCATEGORY
BARIUM CARBONATE
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
Voluna per unit product:
Minimum
Maximum
NA
NA
7
5
57,000 kkg/year
48,745 kkg/year
NA
NA
158 kkg/year
26,190 kkg/year
NA
NA
NA
9 years
24 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, " j^e, 1978 and "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards for the Inorganic Chemicals
Industry," March, 1980
NA= Not Available
787
-------
by thickeners, and the clarified borax solution (mother liquor) is fed
to crystallizers where a slurry of borax crystals is formed. The
borax is separated from the water by centrifugation, dried, screened
and packaged. Process effluents are recycled with excess going to
evaporation ponds or returned to source.
Plants
Three plants produce borax in the United States.
total recycle of wastewater.
BPT Limitations
All three practice
BPT limitations were promulgated on May 22, 1975 (40 CFR 415.272), and
require no discharge of wastewater pollutants to navigable waters.
BAT and NSPS Limitations
BAT and NSPS limitations were proposed on May 22, 1975. They were
never promulgated. Since BPT already requires zero discharge, BAT and
NSPS are being excluded under Paragraph 8 of the Settlement Agreement.
Boric Acid
Summary of Determinations
It has been determined that no further effort be given to developing
or revising BAT, NSPS, or Pretreatment regulations for the boric acid
industry. The basis for this determination is that there is only one
plant which manufactures boric acid *from mined ore. There is an
indication that this plant will discontinue operation. All other
plants manufacture boric acid using the Trona process and have zero
discharge. This subcategory is excluded under Paragraph 8 of the
Settlement Agreement.
Production Processes and Effluents
Boric acid is manufactured by acidification of borax. It is used in
the manufacture of chromic oxide, glazes, enamels, textiles,
fiberglass, and heat resistant glass. It is also used medicinally as
a mild antiseptic and in atomic power plants as a nuclear moderator.
Process wastes may conist of excess boric acid liquor, waste sodium
sulfate by-product liquor and filtration impurities.
The industry profile data is given in Table 26.6-1.
Toxic Pollutants
Toxic pollutants found at significant concentrations during screening
of one plant were:
788
-------
Pollutant
Copper
Thallium
Zinc
Bis(2-ethylhexyl
phthalate
Mercury
Concentration (nq/1
340
140
1200
530
1 .6
There is an indication that this plant will discontinue manufacture of
boric acid. All other plants have zero discharge because of the use
of a different process.
Status of Regulations
BPT limitations were promulgated on May 22, 1975 (40 CFR 415-272) for
this subcategory.
Bromine
Summary of Determinations
It has been determined that no further effort will be given to
developing or revising BAT and NSPS regulations for the Bromine
Subcategory. The basis for this recommendation is that existing BPT
regulations specify zero discharge of process wastewater to navigable
waters. Development of pretreatment is deferred to Phase II.
Production Processes and Effluents
Most bromine is produced from brines pumped from brine wells. A small
amount {1 percent) is produced from brines from Searles Lake near
Trona, California. This is not a navigable water in that it is 35
percent solids. The brine, after appropriate dilution and degassing
is extracted by debromination with chlorine and steam. The steam and
bromine is condensed, separated and distilled to obtain bromine. The
raw waste load from the process is the residual brine and the chloride
salts forced when the chlorine replaces the bromine. The raw wastes
are returned to the brine well or brine source.
Plants
There are nine plants producing bromine in the United States,
which return their wastes to the brine source.
BPT Limitations
all of
Regulations were promulgated on May 22, 1975, (40 CFR 415.292)
requiring zero discharge of process wastewater pollutants to navigable
789
-------
TABLE 26.6-1 -
SUBCATEQORy PROFILE DATA SUMMARY
SUBCATBGOK?
BORIC ACID
Total subcatecpry 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
Volume per unit product:
Minimum
Maximum
NA
122,600 kkg/year
3
2
97,500 kkg/year
93,850 kkg/year
NA
77 percent
30,156 kkg/year
63,694 kkg/year
NA
NA
30
years
83 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
790
-------
waters except that residual brine depleted liquor may be returned to
the body of water from which the brine solution was originally
withdrawn. In no case is the brine source a navigable water. The
source is wells except for a small portion that comes from a "lake"
having 35 percent dissolved solids.
BAT and NSPS Limitations
BAT and NSPS were proposed on May 22, 1975, but never finalized.
Since BPT already requires zero discharge, BAT and NSPS are being
excluded under Paragraph 8 of the Settlement Agreement.
Calcium Carbide
Summary of Recommendations
It has been determined that no additional effort be given to
developing revised BAT and NSPS regulations for this subcategory. The
basis for this recommendation is that BPT, BAT and NSPS regulations
specify zero discharge of process wastewater pollutants. Pretreatment
standards will be developed under Phase II.
Production Processes and Effluents
Calcium carbide is manufactured by reaction of calcium oxide and coke.
Calcium oxide and dried coke are reacted in a furnace and the product
is cooled, crushed, screened, packaged and shipped. There are
generally no process related wastewaters except that one plant had a
wet scrubber discharge.
Plant
There are five plants producing calcium carbide.
BPT, BAT and NSPS Limitations
BPT, BAT and NSPS regulations were promulgated on March 12, 1974 (40
CFR 415.32, 415.33 and 415.35). All subparts require zero discharge
of process wastewater pollutants. It has been determined that the
calcium carbide subcategory will be excluded from development of
revised BAT and NSPS limitations because the operations are now
subject to zero discharge regulations.
Calcium Carbonate
Summary of Determinations
It has been determined that no further effort be given to developing
or revising BAT, NSPS, and Pretreatment regulations for the Calcium
Carbonate Subcategory. The bases for this determination are: 1)
there are only four plants manufacturing calcium carbonate, and 2) the
small quantities of pollutants found during screening were at or very
791
-------
near detectable levels of analysis. This subcategory is excluded
under Paragraph 8 of the Settlement Agreement.
Production Processes and Effluents
Calcium carbonate is manufactured both in pure and impure form and is
used extensively in many industries. In the pure form, it is used in
the rubber, paint, cement, paper and pharmaceutical industries.
In one process, slaked lime is reacted in slurry form with carbon
dioxide. The slurry is then screened and filtered. The recovered
product is dried, milled and packaged for sale. The waste liquor from
the filtration step is recycled or discharged, depending on
requirements. The coarse materials recovered from the screening step
are discharged.
The second process is based on waste streams from the Solvay Process.
A solution of sodium carbonate and sodium bicarbonate from the soda
ash plant is reacted with waste calcium chloride liquor which has been
treated through a settler. The calcium carbonate produced together
with by-product sodium chloride and unreacted calcium chloride is
pumped to a thickener. The overflow from the thickener is collected
with plant drainage streams in a sump to which soda ash finishing
wastewater is added, precipitating calcium carbonate. This mixed
stream then goes to waste collection. The calcium carbonate underflow
is filtered, washed, atomized with steam, dried in a spray dryer,
collected in a particle collector and packaged for sale.
An ultrafine grade of calcium carbonate is produced in a similar
manner to that described above with some additional polish filtering,
tunnel drying and milling. At each plant, the neutralized brine and
process wastewater are returned to the brine cavity. No process
wastewater is discharged.
The industry profile for this subcategory is given in Table 26.9-1.
Toxic Pollutants
There are four plants producing calcium carbonate in the United
States. One discharges to a POTW. Data has been received on three
plants as a result of Section 308 letters. In addition, a sampling
survey for toxic pollutants was made at one plant which represents
approximately 50 percent of total industry capacity. The results of
the 308 letters and the sampling survey indicate that no toxic
pollutants are being discharged. The sampling survey results found
pollutant levels below treatability levels.
Maximum concentration of toxic pollutants found in raw waste were:
792
-------
Pollutant
Nickel
Concentration
21
Zinc
Status of Regulations
68
Interim final regulations (40 CFR 415.302) were promulgated on May 22,
1975. These regulations require conrol of pH and suspended solids for
both the Solvay and lime process. No change in the regulations is
needed.
BAT and NSPS regulations (40 CFR 415.303) were proposed on May 22,
1975. These regulations were never finalized. It has been determined
that the Calcium Carbonate Subcategory be excluded from the
development of BAT and NSPS limitations under Paragraph B of the
Settlement Agreement for the following reasons: There are only four
plants manufacturing calcium carbonate and the 308 letters and
sampling survey indicate that no toxic pollutants are being discharged
in significant quantities.
Because no significant quantities of toxic pollutants are present, no
further effort will be given to development of pretreatment
regulations for this subcategory.
Calcium Chloride
Summary of Determinations
It has been determined that no further effort be given to developing
or revising BAT or NSPS for the Calcium Chloride Subcategory. The
bases for this determination are: 1) there are existing BAT and NSPS
regulations that prohibit discharge of process wastewater pollutants
from the brine extraction process, and 2) there is only one Solvay
process plant in the United States where calcium chloride is recovered
as a by-product.
Production Processes and Effluents
There are • two processes for the manufacture of calcium chloride. In
the first and major production process, calcium chloride is extracted
from natural brines. The salts are solution mined and the resulting
brines first are concentrated to remove sodium chloride by
precipitation and then purified by the addition of other materials to
precipitate sodium, potassium, and magnesium salts. The purified
calcium chloride brine is then evaporated to yield a wet solid which
is flaked and calcined to a dry solid product. The second process is
the Solvay Process, which is primarily used for the manufacture of
soda ash. In the Solvay Process, calcium chloride is recovered as a
by-product. All the wastes from the process are associated with the
sodium carbonate subcategory.
793
-------
TABLE 26.9-1 -
SUBCATEGOE* PROFILE: DAIA SUMMMff
SUBCKEEGOFY
CALCIUM CARBONATE
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:
Mininun
Maxinum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maxiitun
Waste water flew range:
Minimum
Maximum
Volume per unit product:
Minimum
Maxinun
NA
129,600 kkg/year
NA
3
81,300 kkg/year
72,400 kkg/year
NA
56 percent
555 kkg/year
49,800 kkg/year
NA
NA
NA
25 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 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 an
-------
Plants
There are 11 plants producing calcium chloride in the United States,
one of which recovers it as a by-product from the Solvay Process.
Status of Regulations
Existing regulations for calcium chloride (40 CFR 415.4) include
regulations for BAT and NSPS that prohibit discharge of wastewater
pollutants from the brine process.
Calcium Hydroxide
Summary of Determinations
It has been determined that no further effort be given to developing
or revising BAT or NSPS for the Calcium Hydroxide Subcategory. The
basis for this determination is that an existing BPT regulation
provides for zero discharge of process wastewater pollutants (40 CFR
415.312). Pretreatment regulations will be developed in Phase II.
Production Processes and Effluents
Calcium hydroxide is produced by adding water to calcium oxide in a
pug mill premixer. The reacted mixture goes to an agitated hydrator
where more water is added, resulting in an exothermic reaction. No
wastewater is produced and therefore, there is zero discharge to
navigable waters.
Plants
There are approximately fifteen plants producing calcium hydroxide in
the United States.
Chromic Acid
Summary of Recommendations
It has been determined that no additional effort will be given to
developing revised BAT, and NSPS regulations for this subcategory.
The basis for this determination is that the existing interim final
BPT regulation is zero discharge. Pretreatment standards will be
developed in Phase II.
Production Process and Effluents
Sodium dichromate liquor from the dichromate manufacturing operation
is reacted with sulfuric acid and filtered to recover impure chromic
acid as a solid. The mother liquor is returned to the dichromate
process for reuse. The recovered chromic acid is fed to a melter in
which the sodium bisulfate liquifies and is separated from the chromic
acid. The bisulfate is returned to the dichromate operation. The
795
-------
chromic acid is resolidified, flaked and packaged for sale. Wastes
are returned to the dichromate process for reuse.
Plants
There are five plants producing chromic acid.
BPT Limitations
Regulations for BPT were promulgated on May 22, 1975 (40 CFR 415.352).
It has been determined that this subcategory will be excluded from
development of BAT and NSPS regulations under Paragraph 8 of the
Settlement Agreement because the operations are subject to zero
discharge regulations for BPT.
Cuprous Oxide
Summary of Determinations
It has been determined that no further effort be given to developing
BPT, BAT, NSPS, and Pretreatment regulations for the Cuprous Oxide
Subcategory. The basis for this determination is that there is only
one plant manufacturing cuprous oxide. The subcategory is excluded
under Paragraph 8 of the Settlement Agreement.
Production Processes and Effluents
Cuprous oxide is manufactured by reducing cupric oxide by thermal
decomposition in an oxygen-free environment. The reaction occurs at
high temperature aided by a proprietary catalyst. There is no process
related wastewater.
Cuprous oxide is used in the manufacture of glass, ceramics, marine
paints, and photoelectric cells. It is also used in agriculture as a
seed fungicide, as an antiseptic and as a catalyst.
Status of Regulations
Subpart AK has been reserved for this subcategory (Table 26.13-2).
Ferric Chloride
Summary of Determinations
It has been determined that no further effort will be given to
developing revised BAT or NSPS regulations for this industry. The
basis for this determination is that the existing regulation for BPT
is zero discharge. Pretreatment will be developed in Phase II.
Production Processes and Effluents
Ferric chloride is produced from
liquor is reacted with iron,
waste pickle liquor. The pickle
chlorine and hydrochloric acid. The
796
-------
solution is filtered and sold as a solution or evaporated to dryness
to produce a solid product. Wastewater from filter washes, equipment
washing and leaks and spills is returned to the process.
Plants
There are 21 plants producing ferric chloride.
to discharge to POTW.
Toxic Pollutants
Two plants are known
The source of toxic pollutants is the pickle liquor feed. Toxic
pollutants involved are chromium, copper, lead, nickel and zinc.
BPT Limitations
Regulations for BPT were promulgated on May 22, 1975 (40 CFR 415.382),
which require zero discharge of process wastewater pollutants. The
regulations have not been challenged.
BAT and NSPS Limitations
Zero discharge regulations were proposed May 22, 1975 for BAT and
NSPS. Since BPT already requires zero discharge, BAT and NSPS are
being excluded under Paragraph 8 of the Settlement Agreement.
Ferrous Sulfate
Summary of Determinations
It has been determined that no further effort be given to developing
BAT, NSPS, or Pretreatment regulations for the Ferrous Sulfate
Subcategory. The basis for this determination is that ferrous sulfate
is recovered as a by-product and in each of the two processes, the
wastes are attributable to the primary process. Recovery of ferrous
sulfate actually reduces the waste load of both the primary
operations. This subcategory is excluded under Paragraph 8 of the
Settlement Agreement.
Production Processes and Effluents
Ferrous sulfate is made using two processes. In the first case, it is
recovered from the waste su If uric acid pickle liquor containing
ferrous sulfate, ferric sulfate, and unreacted sulfuric acid. The
solution is reacted with iron to reduce ferric ions to ferrous ions.
The process is a by-product recovery from a waste solution rather than
a direct manufacturing process. In the second process, the sulfate
process, ferrous sulfate is obtained as a by-product during the
manufacture of titanium dioxide. In the sulfate process, titanium
dioxide-bearing ores are dissolved in sulfuric acid at a high
temperature to produce iron {ferrous sulfate) and titanium sulfate.
Iron sulfate is removed by crystallization and titanium sulfate is
hydrolyzed and then calcined to produce the final titanium dioxide
797
-------
product. All the wastes from the second process are associated with
the titanium dioxide production.
Process wastewater is derived principally from gas scrubbers.
Plants
There are 13 producers recovering ferrous sulfate from titanium
dioxide manufacture as a by-product or from the waste pickle liquor.
Four of the 13 producers recover ferrous sulfate as a by-product from
the sulfate process. The ferrous sulfate subcategory is excluded
under Paragraph 8 of the Settlement Agreement because there is no
direct method used for its manufacture and it is either recovered from
the waste pickle liquor or as a by-product from the titanium dioxide
manufacture and contributes no wastewater discharge of its own.
Fluorine
Summary of Determinations
It has been determined that no additional effort be given to
developing revised BAT or NSPS regulations for this subcategory. The
basis for this recommendation is that the existing interim final BPT
regulations is zero discharge. Pretreatment standards will be
developed in Phase II.
Production Processes and Effluents
Flourine is produced by electrolysis of liquid hydrogen fluoride.
Fluorine is formed at one electrode and hydrogen at the other. The
fluorine is compressed and packaged in cylinders. There is no process
wastewater from this process.
Plants
There are 3 plants producing fluorine.
BPT Limitations
Regulations for BPT were promulgated on May 22, 1975, (40 CFR 415.402)
and require zero discharge of process wastewater pollutants. The
regulations have not been challenged.
BAT and NSPS Limitations
Zero discharge regulations were proposed for BAT and NSPS but never
promulgated. Since BPT already requires zero discharge, BAT and NSPS
are being excluded under Paragraph 8 of the Settlement Agreement.
798
-------
Hydrochloric Acid
Summary of Determinations
It has been determined that no further effort be given to developing
regulations for BPT, BAT, NSPS, or Pretreatment for the Hydrochloric
Acid Subcategory. The basis for this determinations is: the small
quantities of toxic pollutants found during screening are far below
levels treatable by demonstrated treatment technology. This
subcategory is excluded under Paragraph 8 of the Settlement Agreement.
Production Processes and Effluents
Most of the hydrochloric acid is produced as a by-product in the
manufacture of chlorinated organic compounds. It is used in oil well
activations, pickling of steel, metal cleaning, in monosodium
glutamate manufacture and starch hydrolysis. It is also used as an
acid reagent in several chemical manufacturing processes.
The industry profile data for this subcategory is given in Table
26.17-1. The data given and coverage of this subcategory applies only
to the manufacture of hydrochloric acid by the thermal combination of
chlorine and hydrogen. Wastes from this process come from combustion
chamber condensate and from a fume scrubber.
While most of the hydrochloric acid is produced as a by-product in the
manufacture of chlorinated organic compounds, the wastes are
attributable to the organic compounds involved. This by-product
production is not covered in this subcategory.
Toxic Pollutants
Data has been received on about 25 percent of the industry as a result
of Section 308 letters. In addition, a sampling survey for toxic
pollutants was made at one plant. The results of the 308 letters and
the sampling survey showed concentrations close to the limits of
detectability.
The maximum concentrations of priority pollutants found were:
Pollutant
Lead
Mercury
Nickel
Maximum Concentration
Observed Uq/1)
3.5
2
5.5
799
-------
Status of Regulations
BPT, BAT, NSPS and PSNS regulations (40 CFR 415.72) requiring zero
discharge were promulgated on March 12, 1974. These regulations have
since been remanded by the court and are not in effect.
Hydrogen
Summary of Determinations
It has been determined that no further effort will be given to
developing revised BAT or NSPS regulations for this subcategory. The
basis for this recommendation is that the existing BPT regulation is
zero discharge of process wastewaters to navigable waters.
Preteatment standards will be developed in Phase II.
Production Processes and Effluents
Hydrogen is made chiefly from two sources: purification of petroleum
refinery by-product gases and as a co-product in the manufacture of
carbon monoxide. In the latter case, the wastes are attributed to the
carbon monoxide subcategory. Only the production of hydrogen from
refinery by-product gases will be discussed.
Crude hydrogen as a refinery by-product is passed through a catalytic
bed to remove oxygen and a drier to remove the water formed by the
catalytic reaction. The gas is then cooled, purified and passed
through a converter to change ortho-hydrogen to the para-form.
Hydrogen is usually cooled to a liquid form for storage or shipping.
No contact process water is used during the manufacture.
None are known
Plants
There are approximately 137 plants producing hydrogen.
to have discharges.
BPT Limitations
Regulations were promulgated on May 22, 1975, (40 CFR 415.412)
requiring zero discharge of process wastewater pollutants to navigable
waters. Only contaminated non-process water is allowed. "This
includes rain water, waters which come in contact with accidental
spills and leads, and discharges for personal safety. All reasonable
measures must have been made to prevent, reduce, and control each
contact and to mitigate the effects.
BAT and NSPS Limitations
BAT and NSPS were proposed on May 22, 1975, requiring zero discharge
of process wastewater to navigable waters. Since BPT already requires
zero discharge, BAT and NSPS are being excluded under Paragraph 8 of
the Settlement Agreement.
800
-------
TABLE 26 ..17-1 -
SUBCAIEGORY PROFILE DATA SUMMARY
SUBCAOEGORY
HYDROCHLORIC 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:
Minimum
Maximum
Vfaste water flow range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
MA
NA
NA
6
163,000 kkg/year
119,000 kkg/year
NA
NA
NA
NA
NA
NA
NA
4 years
20 years
NA
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
801
-------
Iodine
Summary of Determinations
It has been determined that no additional effort will be given to
developing revised BAT or NSPS regulations. The basis for this
determination is that the existing regulation for BPT is zero
discharge. Pretreatment standards will be developed in Phase II,
Production Process and Effluents
Iodine is produced from brine solutions containing iodine. The brine
is acidified and chlorinated liberating free iodine. The free iodine
is stripped from the brine and treated again with chlorine yielding
solid iodine. The slurry is filtered, treated with sulfuric acid and
refiltered. The product is then crushed and packaged for sale. The
wastes from this process are spent brine solutions which are returned
to the well from which the brine was initially obtained.
Plants
There are 4 plants producing iodine.
BPT Limitations
Regulations for BPT were promulgated on May 22, 1975, (40 CFR 415.432)
and require zero discharge of process wastewater pollutants. The
regulations have not been challenged.
BAT and NSPS Limitations
Zero discharge regulations were proposed for BAT and NSPS on May 22,
1975. Since BPT already requires zero discharge, BAT and NSPS are
being excluded under Paragraph 8 of the Settlement Agreement.
Lead Monoxide
Summary of Determinations
It has been determined that no further effort will be given to
developing BAT or NSPS regulations for this subcategory. The basis
for this recommendation is that the existing BPT regulation requires
zero discharge. Pretreatment standards will be developed in Phase II.
Production Processes and Effluents
Lead monoxide is produced by the thermal oxidation of lead. There are
no process wastewater streams generated by lead monoxide production.
Dust control is the problem in this subcategory. Use of dry
collection systems rather than a water collection system is the
control technology for meeting the regulation.
802
-------
Plants
There are 15 plants producing lead monoxide in the United States. Ten
plants are known to use dry bag collection systems and have no
discharge of wastewater. Others are subject to existing zero
discharge regulations.
BPT Limitations
BPT limitations were promulgated on May 22, 1975 (40 CFR 415.442).
The limitations require zero discharge of process wastewater
pollutants into navigable waters.
BAT and NSPS Limitations
On May 22, 1975, zero discharge regulations were proposed but never
promulgated for BAT and NSPS. Since BPT already requires zero
discharge, BAT and NSPS are being excluded under Paragraph 8 of the
Settlement Agreement.
Lithium Carbonate
Summary of Determinations
It has been determined that no further efforts be given to developing
or revising regulations for BPT, BAT, NSPS, or Pretreatment for the
Lithium Carbonate Subcategory. The bases for this determination are:
. 1) there is only one plant in this subcategory using the spodumene ore
process and discharging process wastewater, and 2) there is an
existing zero discharge regulation for the brine process. This
subcategory is excluded under Paragraph 8 of the Settlement Agreement.
Production Processes and Effluents
Lithium carbonate is produced by two processes. In one process,
spodumene ore is heated at a high temperature to render it highly
reactive. It is then cooled, ball-milled, and mixed with concentrated
sulfuric acid. The acid-roasted ore is leached with water, and the
excess acid is neutralized with ground limestone. This mixture is
filtered and further treated with lime and soda ash. Further
processing precipitates 1ithium carbonate. Wet scrubbers are the
sources of wastewater. Significant quantities of toxic pollutants are
not found in the wastewater.
In the other process, lithium carbonate is produced by the reaction of
lime with concentrated brine, and lithium carbonate is precipitated by
filtration. Process wastewater consists of spent brines, which are
sent to on-site evaporation ponds. There is no process wastewater
discharge from this process.
803
-------
Status of Regulations
There is an existing BPT regulation for this subcategory (40 CFR
415.452).
Manganese Sulfate
Summary of Determinations
It has been determined that no further effort be given to developing
or revising BPT, BAT, NSPS, or Pretreatment regulations for the
Manganese Sulfate Subcategory. The bases for this determination are:
1) there is only one plant making commercial grade manganese sulfate
that has a wastewater discharge, and 2) the amount of wastewater
produced by that plant is low. The subcategory is excluded under
Paragraph 8 of the Settlement Agreement.
Production Processes and Effluents
There are two processes for the manufacture of manganese sulfate; the
hydroquinone process and the coke and ore process. In the
hydroquinone process, manganese ore, aniline and sulfuric acid are
reacted to produce manganese sulfate, quinone and ammonium sulfate.
The reacted mixture is steam distilled to remove quinone which is
further processed to hydroquinone. The mixture of manganese and
ammonium sulfate is filtered, evaporated, and crystallized. Manganese
sulfate is recovered as crystals, and the spent 1iquor contains
ammonium sulfate. In the second process, manganese ore and coke are
reacted in a kiln and the product is leached with sulfuric acid. The
resulting slurry is evaporated to dryness to recover a 30 percent
product for agricultural purposes. The amount of wastewater produced
from the hydroquinone process is small and the other process produces
no waterborne waste.
Plants
Four plants are manufacturing manganese sulfate. Two of the producers
use it for fertilizer production and they generate no waterborne
wastes. One plant produces reagent grade product and the total
production is very low. Only one other plants manufactures manganese
sulfate (commercial grade) and has a significant wastewater flow.
Status of Regulations
Since only one manganese sulfate plant discharges waste to navigable
waters, the subcategory is excluded from federal discharge regulation
for BPT, BAT, NSPS, and Pretreatment standards under Paragraph 8 of
the Settlement Agreement.
804
-------
Nitric Acid
Summary of Determinations
The existing nitric acid regulation in the fertilizer category (40 CFR
418.5) is applicable to all nitric acid plants captive to a fertilizer
production facility. In addition, sampling has shown that there are
no significant quantities of toxic pollutants in the process
wastewaters from stand alone nitric acid plants. Further BPT, BAT,
NSPS or Pretreatment regulations will not be developed for this
subcategory.
Production Processes and Effluents
Most of the nitric acid produced is used in the manufacture of
ammonium nitrate and other nitrogen fertilizers. On site captive use
is extensively practiced. It is also used in the manufacture of
explosives, plastics and other organic products.
acidic and pickling agent. The source of
equipment washing operations.
the
Other uses are as an
process wastewater is
The industry profile data for this subcategory are given in Table
26.23-1.
Toxic pollutants found in raw wastes during sampling were as follows:
Maximum Concentration Observed
Ug/1)
Pollutant
Chromium
Zinc
Lead
Mercury
Silver
2,4-Dinitrophenol
Nickel
Cyanide
Screening
(2 Plants)
no
120
29
0.
0.
215
170
< 0.
47
5
04
Verification
(1 Plant)
100
791
< 10
4.5
< 15
Not Analyzed
85
< 0.02
The 2,4-Dinitrophenol is caused by contamination from organic products
manufactured at the plant and will be addressed in that guideline.
The chromium and zinc are ingredients of cooling water conditioners
present in the blowdown which is mixed with process streams.
Appropriate control is by best management practice 'not end-of-pipe
treatment via national regulation. Other metals are below the limit
of treatability.
805
-------
Status of Regulations
BPT, BAT, NSPS, and PSNS regulations (40 CFR 415.100) requiring zero
discharge were promulgated on March 12, 1974. These regulations have
since been remanded by the court and are not in effect.
Oxygen and Nitrogen
Summary of Determinations
It has been determined that no further effort be given to developing
or revising regulations for BAT, NSPS, or Pretreatment for the Oxygen
and Nitrogen Subcategory. The bases for this determination are: 1)
the wastewater discharge mainly consists of compressor water, and 2)
the only toxic pollutant detected at or above treatability level was
copper which is at the level of treatability. This subcategory is
excluded under Paragraph 8 of the Settlement Agreement.
Production Processes and Effluents
Oxygen and nitrogen are produced from air by distillation of liquified
air. Oxygen is used in the production of steel, gas welding,
medicine, jet fuel, in sewage treatment plants and in the manufacture
of ethylene and acetylene. In rocket propulsion, liquid oxygen is
often used as a cryogenic liquid oxidizer.
The greatest use of nitrogen is in the manufacture of ammonia by the
Haber process. It is also used in cryosurgery. As an inert gas, it
is used to prevent oxidation by air. In the liquid form, it is used
for low temperature refrigeration.
The wastewater discharge mainly consists of compressor cooling water.
Other wastewaters are small quantities of boiler blowdown, intake air
scrubber waters, and compressor condensate.
The industry profile for this subcategory is given in Table 26.24-1.
Toxic Pollutants
Data has been received on 10 plants as a result of Section 308
letters. There are at least 230 plants in the United States.
However, all operate using the same basic process. One plant was
sampled during the screening program. Toxic pollutants found in the
raw waste loading during sampling were:
806
-------
TABLE 26.23-1 -
SUBCAIEGOHf PRCFILE DATA SUMMARY
SUBCAIBGOKy
NITRIC ACID
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 ranges
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
9,177,000 kkg/year
7,171,000 kkg/year
87
11
1,106,000 kkg/year
774,400 kkg/year
12 percent
11 percent
NA
NA
NA
NA
NA
4 years
83 years
NA
NA
NA
NA
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,1' June,1978 and "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards for the Inorganic Chemicals
Industry," March, 1980
NA = Not Available
807
-------
Pollutant
Chromium
Copper
Lead
Nickel
Zinc
Concentration Ug/1)
26
590
51
79
170
The likely sources of copper are corrosion and bearing wear, and
concentration in boiler and cooling tower blowdowns. The copper
levels are at the accepted levels of treatment, therefore further
reduction is not practical.
Status of Regulations
Interim Final BPT regulations (40 CFR 415.492) were promulgated on May
22, 1975. These regulations require limitations on pH and oil and
grease. These regulations remain in effect and no change is needed.
BAT and NSPS regulations (40 CFR 415.494) were proposed on May 22,
1975. These regulations were never finalized. It has been determined
that the Oxygen and Nitrogen Subcategory be excluded from the
development of. BAT and NSPS limitations under Paragraph 8 of the
Settlement Agreement for the following reasons: The discharge
consists of compressor water wherein the only toxic pollutant found is
copper which is at the level of treatability.
Potassium Chloride
Summary of Determinations
It has been determined that no further effort will be given to
developing revised BAT or NSPS regulations for the potassium chloride
subcategory. The basis for this determination is that existing BPT
regulations specify zero discharge of process waters. Pretreatment
standards will be developed in Phase II.
Production Processes and Effluents
Potassium chloride is produced in the United States by two principal
processes: extraction from sylvite ore and extraction from lake brine
(Trona Process). Sylvite ore is a combination of potassium chloride
and sodium- chloride. The ore is crushed, screened, and wet-ground in
brine. The ore is then separated from clay impurities in a desliming
process. The clay impurities are fed to a gravity separator which
removes some of the sodium chloride precipitated from the leach brin^
and insolubles for disposal as waste. After desliming, the ore is
chemically treated and the potassium chloride is separated from the
sodium chloride in a flotation process. The tailings from flotation
808
-------
TftBLE 26.24-1 -
SUBCRTEGORy PKCFZLE DKEA .SUMMARY
SUBCAIEGORY
OX3OTN AND NITROGEN
•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
Volume per unit product;
Minimum
Maximum
35,526,000 kkg/year
NA
113
9
1,588,000 kkg/year
1,473,000 kkg/year
4.5 percent
NA
2,400 kkg/year
378,000 kkg/year
NA
NA
NA
4 years
36 years
NA
NA
NA
NA
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1979, 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,11 June, 1978 and "Economic Analysis of
Proposed Revised Effluent Guidelines and Standards for the Inorganic
Chemicals Industry," March, 1980
NA = Not Available
809
-------
are wasted and the resulting potassium chloride slurries are
centrifuged to recover the potassium chloride. The product is then
dried, screened, and packaged. The centrifuge liquors are recycled to
the flotation circuit.
The Trona Process uses a cyclic evaporation-crystallization system in
which saline brine is evaporated to nominal dryness. The brine and
recycle liquor is concentrated in triple effect evaporators to produce
a hot liquor high in potassium chloride and borax. Large quantities
of sodium chloride and burkeite (Na2C03.Na2S04) are crystallized and
separated during evaporation. The sodium chloride is returned to the
brine source, and the burkeite is transported to other processes for
separation and refining. The hot liquor is then cooled rapidly in
vacuum crystallizers and the potassium chloride is filtered from the
slurry. The potassium chloride in then dried and packaged. A small
portion may be refined further and/or converted to potassium sulfate.
The cool liquor remaining is then allowed to crystallize the remaining
borax which is then refined further using recrystallization and other
processes. The remaining liquor is recycled back to the
evaporation-crystallization step.
Plant
There are thirteen plants producing potassium chloride, two of which
use the Trona Process.
BPT Regulations
BPT regulations were promulgated on May 22, 1975 (40 CFR 415.502)
requiring zero discharge of process wastewater pollutants to navigable
waters, except that residual brine and depleted liquor may be returned
to the body of water from which the brine solution was withdrawn.
There are no instances where the brine soure is a navigable water.
BAT and NSPS Limitations
BAT and NSPS were proposed on May 22, 1975, requiring zero discharge
of process wastewater to navigable waters. It has been determined
that the potassium chloride subcategory will be excluded from the
development of revised BAT and NSPS limitations under Paragraph 8 of
the Settlement Agreement since a zero discharge regulation is in
effect. In the absence of BAT and NSPS regulations, permits will be
based on BPT.
Potassium Bichromate
Summary of Determinations
It has been determined that no further effort will be given to
developing revised BAT and NSPS regulation for the Potassium
Dichromate Subcategory. The basis for this determination is that
existing BPT, BAT and NSPS regulations specify zero discharge of
810
-------
process wastewater pollutants to navigable waters. Pretreatment
standards will be developed in Phase II.
Production Processes and Effluents
Potassium dichromate is made by reacting a sodium dichromate dihydrate
solution with potassium chloride. The potassium dichromate is
crystallized from solution requiring only removal of water prior to
sizing and packaging. The process water is recycled back to the
initial reaction tank.
BPT, BAT and NSPS Limitations
BPT, BAT and NSPS limitations were promulgated March 12, 1974 (40 CFR
415.122, 415.123 and 415.T25). All subparts require zero discharge of
process wastewater pollutants to navigable waters.
It has been determined that the potassium dichromate subcategory will
be excluded from the development of revised BAT and NSPS limitations
under Paragraph 8 of the Settlement Agreement. The basis for this
determination is that by maintaining existing BPT, BAT and NSPS
limitations, no discharge of wastewater pollutants to navigable waters
will occur.
Potassium Iodide
Summary of Determinations
It has been determined that no further effort be given to developing
or revising BAT, NSPS, and Pretreatment regulations for the Potassium
Iodide Subcategory. The bases for this determination are: 1) because
the wastewater discharge is less than 100 gallons per day, the
quantity of pollutants discharged is very low/ and 2) the
concentrations of the toxic pollutants are at or below accepted
treatment levels. This subcategory is excluded under Paragraph 8 of
the Settlement Agreement,
Production Processes and Effluents
Potassium iodide is used in photographic emulsions, in animal and
poultry feeds, table salts and analytical chemistry. It also has a
number of medical uses.
One manufacturing process is known as the iron carbonate process.
This involves the reaction of iron powder with iodine in aqueous
solution. An intermediate compound, ferrosoferricyanide, is formed
which is subsequently reacted with potassium carbonate to yield
potassium iodide. The raw product is purified, concentrated by
evaporation and cooled to promote crystallization. Water used
directly in the process is lost by evaporation. The only source of
process wastewater is from equipment wash down operations.
The industry profile for this subcategory is given in Table 26.27-1.
811
-------
Toxic Pollutants
Data has been received for approximately 50 percent of the industry as
a result of Section 308 letters. In addition, a sampling survey was
made at one plant. The following toxic pollutants were identified in
the plant wastes:
Pollutant
Antimony
Chromium
Copper
Lead
Silver
Zinc
Concentration
48
22
1040
26
34
30
However, the levels of these pollutants are at or below accepted level
of treatability. In addition, the flows are less than 100 gallons per
day. At the one plant sampled, there was no process wastewater
discharged since the wash water was sent to an evaporation pond.
Status of Regulations
A. BPT Limitations
BPT regulations (40 CFR 415.511) were promulgated on May 22,
1975. These regulations require limitations on pH, TSS, sulfide,
iron and barium. These regulations are adequate for the control
of.conventional and nonconventional pollutants.
B. BAT and NSPS Regulations
NSPS and BAT limitations were proposed in May 22, 1975, but never
finalized. It has now been determined that the Potassium Iodide
Subcategory be excluded from the development of BAT and NSPS
limitations under Paragraph 8 of the Settlement Agreement for the
following reasons: 1) very small quantities of toxic pollutants
are discharged from this industry, and 2) those pollutants
discharged are at or below accepted treatability levels.
C. Pretreatment Limitations
Because no significant quantities of toxic pollutants are
present, no further effort will be given to development of
pretreatment regulations for this subcategory.
Potassium Metal
Summary of Determinations
It has been determined that no further effort will be given to
developing revised BAT or NSPS regulations for the Potassium Metal
Subcategory. These bases for this recommendation are': 1) existing
812
-------
BPT, BAT and NSPS regulations specify zero discharge of process
wastewaters; and 2) there is only one plant producing potassium in the
U.S. and that plant uses a dry process. Pretreatment standards will
be developed in Phase II.
Production Processes and Effluents
Potassium metal is prepared by melting
gas-fired melt pot prior to being fed to an
molten potassium chloride flows down through
is contacted by ascending sodium vapors
reboiler.
chloride,
apparatus.
product.
effluents.
Plant
The reaction yields elemental
which is withdrawn continuously
The elemental potassium is wi
No process water is used so
potassium chloride in a
exchange column. The
a packed column, where it
coming from a gas-fired
potassium and sodium
from the base of the
thdrawn as an overhead
there are no waterborne
Only one plant produces potassium metal in the U.S.
process water and there are no waterborne effluents.
It uses
no
BPT, BAT and NSPS Limitations
BPT, BAT and NSPS limitations were promulgated March
415.112), 415.113 and 415.115). All subparts require
of process wastewater pollutants to navigable waters.
12, 1974 (40 CFR
zero discharge
It has been determined that the potassium metal subcategory will be
excluded from the development of revised BAT and NSPS limitations
under Paragraph 8 of the Settlement Agreement. Maintaining the
existing regulations will eliminate the discharge of toxic pollutants.
Potassium Permanganate
Summary of Determinations
It has been determined that no further effort be given to developing
BPT, BAT, NSPS, and Pretreatment regulations for the Potassium
Permanganate Subcategory. The basis for this determinations is that
there is only one plant manufacturing Potassium Permanganate. The
subcategory is excluded under Paragraph 8 of the Settlement Agreement.
Production Processes and Effluents
Manganese ore is s lurried with potassium hydroxide solution and
treated with oxygen to produce potassium manganate. This intermediate
product and the ore wastes are recovered by centrif ugation and the
solids are then leached to dissolve the manganate. The resulting
slurry is filtered to remove the ore wastes and the manganate
converted in electrolytic cells. The permanganate is crystallized
from the solution to form the product.
813
-------
TABLE 26.27-1 -
SUBCATEQOKSf PRCFUE DATA SUMMARY
SUBCATEGOR*
POTASSIUM IODIDE
Total subcatecpry capacity rate
Total subcatecpry 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
Maxinun
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maxiitun
Volume per unit product:
Minimum
Maximum
NA
NA
9
4
1,985 kkg/year
1,300 kkg/year
NA
50 percent
79 kkg/year
634 kkg/year
NA
NA
NA
27 years
42 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
814
-------
Potassium Sulfate
Summary of Determinations
It has been determined that no further effort
or revising BPT, BAT and NSPS for the Potassium
The bases for .this determination are there is
for BAT and NSPS that requires zero discharge
pollutants (40 CFR 415.133 and 415.135). The
under Paragraph 8 of the Settlement Agreement.
will be developed in Phase II.
Production Processes and Effluents
be given to developing
Sulfate Subcategory.
an existing regulation
of process wastewater
subcategory is excluded
Pretreatment standards
Potassium sulfate is produced by the reaction in solution of potassium
chloride with langbeinite ore. Langbeinite ore is a natural sulfate
of potassium and magnesium (K2S04.MgS04), usually intermixed with
sodium chloride. When the reacted solution is partially evaporated,
potassium sulfate precipitates out, and is recovered by centrifugation
or filtration from the brine liquor, dried, and sold. The remaining
brine liquor containing magnesium chloride is the source of raw waste.
Depending on the sodium content of the ore used, the brine is either
sold (low sodium content) or is ponded. In the latter case, the brine
liquor is recycled or evaporated and the mud is landfilled.
Therefore, no discharge results from the production of potassium
sulfate.
Plants
There are approximately eight producers of commercial grade potassium
sulfate in the United States.
Sodium Bicarbonate
Summary of Determinations
It has been determined that no further effort will be given to
developing revised BAT and NSPS regulations for the Sodium Bicarbonate
Subcategory. The basis for this determination is the existing BPT,
BAT and NSPS regulations specify zero discharge of process wastewater
pollutants to navigable waters. Pretreatment standards will be
developed in Phase II.
Production Processes and Effluents
Sodium bicarbonate is produced by reaction of sodium carbonate with
water and carbon dioxide under pressure. The bicarbonate precipitates
from the solution and is filtered, washed, dried, and packaged.
Wastewater from the filtration is used in product scrubbers and then
returned to the process.
815
-------
Plants
Four plants produce sodium bicarbonate in the United States.
BPT, BAT and NSPS Limitations
BPT, BAT and NSPS limitations were promulgated March 12, 1974 (40 CFR
415.142, 415.143 and 415.145). All subparts require zero discharge of
process wastewater pollutants to navigable waters.
It has been determined that the sodium bicarbonate subcategory will be
excluded under Paragraph 8 of the Settlement Agreement. The basis for
the determination is that maintaining the existing regulations will
eliminate the discharge of toxic pollutants.
Sodium Carbonate
Summary of Determinations
It has been determined that no further effort be given to developing
or revising BAT, NSPS, or Pretreatment regulations for the Sodium
Carbonate Subcategory. The bases for this determination are: 1) no
wastewater is discharged to navigable waters from the plants using
natural deposits to produce sodium carbonate, and 2) only one plant
exists that uses the Solvay Process to produce sodium carbonate. This
subcategory is excluded under Paragraph 8 of the Settlement Agreement.
The Solvay Process does have a discharge but because there is only one
plant, it is inappropriate to write a regulation for the subcategory.
Production Processes and Effluents
Two methods are used for the production of sodium carbonate. One
method is the recovery from natural sodium carbonate deposits and the
other method is the Solvay process. The wastewater resulting from the
use of natural deposits is sent to evaporation ponds and no water is
discharged to navigable streams. In the Solvay process, sodium
chloride (brine) is purified and saturated with ammonia and then
chlorinated. The reacted solution is filtered and sodium bicarbonate
is removed as a filter cake. The filter cake is calcined to produce
sodium carbonate, driving off moisture and carbon dioxide. The
production of sodium carbonate by the Solvay process requires the use
of large volumes of water for non-contact cooling and process contact
purposes and generates large loads of suspended solids, alkalinity,
and ammonia.
Plants
Only one plant uses the Solvay process to produce sodium carbonate.
The Solvay process is energy intensive and generates large pollution
loads. The process is being replaced by production from natural
deposits. It is unlikely that new Solvay process plants will be built
in the future. The industry profile is presented in Table 26.32-1.
816
-------
The other plants using natural deposits have zero discharge.
Status of Regulations
The regulation originally established has been remanded by the court.
The Solvay* process does have a discharge but because there is only one
plant it is inappropriate to write regulations for this subcategory.
Sodium Chloride
Summary of Determinations
It has been determined that no further effort be given to developing
or revising BAT and NSPS regulations for the Sodium Chloride
Subcategory. The basis for this determination is that there are
existing BAT and NSPS regulations that prohibit discharge of process
wastewater (40 CFR 415.163 and 415.165). The subcategory is excluded
under Paragraph 8 of the Settlement Agreement. Pretreatment standards
will be developed in Phase II.
Production Processes and Effluents
Sodium chloride is produced by three methods: 1) solar evaporation of
sea water, 2) solution mining of natural brines, and 3) mining of rock
salt.
In the solar evaporation process, salt water is concentrated by
evaporation in open ponds to yield a saturated brine solution. After
saturation is reached, the brine is fed to a crystallizer, wherein
sodium chloride precipitates, leaving behind a concentrated brine
solution (bittern) consisting of sodium, potassium, and magnesium
salts. The precipitated sodium chloride is recovered for sale and the
brine is recycled from the operation.
The brine in the second process is first aerated to remove hydrogen
sulfide. The brine is then pumped to settling tanks where it is
treated with caustic soda and soda ash to remove most of the calcium,
magnesium, and iron present as insoluble salts. After clarification
to remove these insolubles, the brine is then sent to multiple effect
evaporators. As water is removed, salt crystals form and are removed
as a slurry. The slurry is washed with fresh brine to remove calcium
sulfate. The washed slurry is filtered, the mother liquor is returned
to evaporators, and the crystals from the filter are dried and
screened. Wastes are generated from the multiple effect evaporators
and driers, basic brine purification, and from water treatment. Zero
aqueous discharge can be accomplished by replacing barometric
condensers with non-contact exchangers, eliminating packing station
wastes by more efficient design, and recycling all process water.
Mining of rock salt produces no process wastewater.
817
-------
related wastewater during the screening of one plant. The subcategory
is excluded under Paragraph 8 of the Settlement Agreement.
Production Process and Effluents
Sodium hydrosulfide is produced by reaction of hydrogen sulfide with
sodium hydroxide. Sodium hydrosulfide is used in the manufacture of
sodium sulfide, other chemicals, and paper (Kraft). It is also used
in dehairing of hides and industrial wastewater treatment. Process
wastewater may be derived from filter backwash water.
The subcategory profile data are given in Table 26.35-1.
Toxic Pollutants
Toxic pollutants found in the waste during screening of one plant were
phenol (76 pg/1) and naphthalene (90 ^g/1) which are below
treatability levels. Due to the very small flows and waste loads
generated by this industry, this subcategory is excluded under
Paragraph 8 of the Settlement Agreement
Status of Regulations
Subpart BD has been reserved for this subcategory.
Sodium Metal
Summary of Determinations
It has been determined that no further effort be given to developing
or revising BPT, BAT, NSPS, or Pretreatment regulations for the Sodium
Metal Subcategory. The basis for this determination is that the small
quantities of toxic pollutants found during screening are far below
accepted treatability levels. This subcategory is excluded under
Paragraph 8 of the Settlement Agreement.
Production Processes and Effluents
Sodium metal is manufactured with chlorine by electrolysis of fused
sodium chloride. It is used in the production of tetraethyl lead,
sodium cyanide, sodium peroxide, and titanium and zirconium metals.
In liquid form, it is used as a nuclear reactor coolant; it is also
used as a light, thermally conductive solid in various applications.
The industry profile for this subcategory is given in Table 26.36-1.
Toxic Metals
Data has been received on about 60 percent of the industry as a result
of Section 308 letters. In addition, sampling surveys were made at
two plants representing 38 percent of the industry. Toxic pollutants
found during sampling were as follows:
820
-------
Pollutant
Copper
Zinc
Dichlorobromomethane
Chloroform
Maximum Concentration
Observed
31
13
33
10
These pollutants are at very low concentrations which are far below
accepted treatability levels.
Status of Regulations
BPT regulations (40 CFR 415.182) were promulgated on March
These regulations have since been remanded by the court.
12, 1974
BAT and NSPS regulations requiring zero discharge (40 CFR 415.183)
were promulgated on March 12, 1974. These regulations have been since
remanded by the court. However, it has been determined that the
sodium metal subcategory be excluded from BAT and NSPS regulations
because data from Section 308 letters and sampling surveys indicate
that toxic pollutant concentrations are far below accepted treatable
levels.
Because no significant quantities of toxic pollutants are present, no
further effort will be given to development of pretreatment
regulations for this subcategory.
Sodium Silicate
Summary of Determinations
It has been determined that no further effort be given to developing
BPT, BAT, NSPS, and Pretreatment regulations for the Sodium Silicate
Subcategory. The basis for this determination is that the small
quantities of toxic pollutants found during screening are below
accepted levels of treatability. This subcategory is excluded under
Paragraph 8 of the Settlement Agreement.
Production Processes and Effluents
Sodium silicate is manufactured both in liquid and anhydrous powdered
iorm. It has many industrial uses, such as additives in adhesives,
flocculants, and cleaning agents. It is also used in the production
of soap and household detergents. Sources of process wastewater
include contact cooling water, filter backwash, gas scrubbers and tank
cleaning.
The industry profile for this subcategory is given in Table 26.37-1,
821
-------
Toxic Pollutants
Data has been received on about 63 percent of the industry as a result
of Section 308 letters. In addition, a sampling survey was made at
one plant which represents about 6 percent of the industry. The
following pollutants were detected: nickel, copper, and zinc. These
levels are below accepted treatability levels. In addition, the
sampling data was taken from wastewaters receiving insufficient
treatment. The wastes were ponded to remove suspended sol ids
consisting essentially of sand and other silicates. Normally the pH
of the wastes would be lowered to 9 and receive additional settling.
However the dissolved silicate and high pH are considered beneficial
by sewerage authorities in the removal of solids in primary and
secondary settling systems.
Maximum concentrations of toxic pollutants found during sampling are:
Pollutant
Copper
Nickel
Zinc
Status of Regulations
347
121
181
BPT, BAT, and NSPS regulations (40 CFR 415.192) requiring zero
discharge of pollutants were promulgated on March 12, 1974. These
regulations have since been remanded by the court and are not in
effect.
Because no significant quantities of toxic pollutants are present, no
further effort will be given to development of pretreatment
regulations for this subcategory.
Sodium Silicofluoride
Summary of Determinations
This subcategory has been excluded from the present study but will be
included in the Phase II, Inorganic Chemicals, review.
Production Processes and Effluents
Sodium silicofluoride is used in the manufacture of sodium fluoride
and in the light metal industry as a protective agent. It is also
used as an insecticide, as a fluxing and opacity agent for ceramics
and in detergent products.
The industry profile for this subcategory is given in Table 26.38-1.
822
-------
TABLE 26.35-1 -
SUBCKTECOKT PROFILE .DATA SUMMARY
SUBCftTEGOEY
SODIUM HYDRCSULETDE
Total subcate^ry 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 voter flow range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
NA
NA
12
3
56,900 kkg/year
44,700 kkg/year
NA
NA
3,800 kkg/year
36,500 kkg/year
NA
NA
NA
5 years
14 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
823
-------
Sodium Sulfite
Summary of Determinations
It has been determined that no further efforts be given to developing
or revising regulations for the Sodium Sulfite Subcategory. The basis
for this determination is that there are existing regulations for BAT
and NSPS that require zero discharge of process wastewater pollutants
(40; CFR 415.203 and 415.205). The subcategory is excluded under
Paragraph 8 of the Settlement Agreement. Pretreatment standards will
be developed in Phase II.
Production Processes and Effluents
Sodium sulfite is produced by two processes. One is the direct
reaction of soda ash with sulfur dioxide. There are four plants
manufacturing sodium sulfite using this process. In the other
process, sodium sulfite is produced as a by-product in the manufacture
of phenol. Since this process is used primarily to produce phenol and
its derivatives, it is not considered for this subcategory.
Sodium Thiosulfate
Summary of Determinations
It has been determined that no further effort be given to developing
BPT, BAT, NSPS, or Pretreatment regulations for the Sodium Thiosulfate
Subcategory. The basis for this determination is that no toxic
pollutants were found at significant levels in the raw waste during
screening of one plant. The subcategory is excluded under Paragraph 8
of the Settlement Agreement.
Production Processes and Effluents
Most of the sodium thiosulfate is produced by the sulfur-sodium
sulfite process. It is used extensively in the development of
negatives and prints in the photographic industry. It is also used in
medicine, in the paper and dyeing industries, and as a bleaching agent
for natural products. Process wastewater sources include filter
backwash and the discharge from barometric condensers.
The subcategory profile data age given in Table 26.40-1.
Toxic Pollutants
Data has been received on about 33 percent of the industry as a result
of Section 308 letters. A sampling survey at one plant indicated that
toxic pollutants in the effluent are below treatment levels.
Toxic pollutants identified in the effluent were:
824
-------
TABLE 26.36-1 -
SUBCAOEQORY PROFILE DATA SUMMARY
SUBCATEGORY
SODIUM METAL
Total subcategory capacity rate
Tbtal subcategory production rate
Number of plants in this subcategory
308 Data on file for
With total capacity of
With total production of
Representing capacity
Representing production
Plant production range:
Minimum
Maximum
Average production
Median production
Average capacity utilization
Plant age range:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
381,000 kkg/year
NA
5
3
96,340 kkg/year
78,541 kkg/year
25 percent
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1979, U.S. Department of Conmerce, 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
825
-------
Pollutant
Copper
Zinc
Concentration Uq/1)
91
94
Status of Regulations
Subpart BG has been reserved for this subcategory.
Stannic Oxide
Summary of Determinations
It has been determined that no additional effort be given to
developing revised BAT or NSPS regulations for this subcategory. The
basis for this recommendation is that existing regulation for BPT is
zero discharge. Pretreatment standards will be developed in Phase II.
Production Process and Effluents
Tin is reacted with air and oxygen in a furnace to form stannic oxide.
The product is recovered with dry bag collectors and packaged for
sale. There is no process wastewater from this process.
Plants
There are three plants producing stannic oxide.
BPT Limitations
*
Regulations for BPT were promulgated on May 22, 1975 (CFR 415.602) and
require zero discharge of process wastewater pollutants. The
regulations have not been challenged.
BAT and NSPS Limitations
Zero discharge regulations were proposed for BAT and NSPS on May 22,
1975. Since BPT already requires zero discharge, BAT and NSPS are
being excluded under Paragraph 8 of the Settlement Agreement.
Strong Nitric Acid
Summary of Determinations
It has been determined that no further effort be given to developing
regulations for the Nitric Acid (Strong) Subcategory. The basis for
this determination is that no process related toxic pollutants were
found at significant levels in the process wastewater during screening
of two plants and verification of one plant. The subcategory is
excluded under Paragraph 8 of the Settlement Agreement.
826
-------
TABLE 26.37-1 -
SUBCATBQOEY PRCFHE DATA StMMAKf
SUBCftlEQOK*
SODIUM SILICATE
Tbtal subcategory capacity rate (27 PlantsI
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
Volume per unit product:
Minimum
Maximum
927,300 kkg/year
NA
39
21
NA
431,000 kkg/year
47 percent
NA
12,400 kkg/year
57,300 kkg/year
NA
NA
NA
7 years
43 years
NA
NA
NA
NA
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1979, 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
827
-------
Production Processes and Effluents
Most of the strong nitric acid is produced by dehydration of dilute
nitric acid. Strong nitric acid is used in the manufacture of organic
compounds where nitric acid acts as an oxiding agent instead of an
acid. It is also used in the manufacture of dye intermediates and
explosives. The principal wastewater source is derived from equipment
washing. The industry profile data are given in Table 26.42-.
Toxic Pollutants
Toxic pollutants found in the waste streams during sampling of
nitric acid plants were:
strong
Pollutant
Chromium
Zinc
Lead
Mercury
Silver
Nickel
Cadmium
Cyanide
(2
Maximum
Concentration Observed (
Screening Verification
Plants) (1 Plant) !
40,000
900
70
8.6
0.69
5.0
2.0
0.020
50
120
10
1 .2
15
50
2.0
0.020
In a follow-up, it was found that the chromium and zinc are used as
corrosion inhibitors in the cooling tower, and are not process
related. Control of these pollutants should involve best management
practices instead of end-of-pipe treatment.
Status of Regulations
Subpart AV has been reserved for this subcategory.
Sulfur Dioxide
Summary of Determinations
It has been determined that no further effort be given to developing
BPT, BAT, NSPS, and Pretreatment regulations for the Sulfur Dioxide
Subcategory. The basis for this determination is that no toxic
pollutants were found at significant levels in the raw waste during
screening of one plant. The subcategory is excluded under Paragraph 8
of the Settlement Agreement.
Production Processes and Effluents
Most of the sulfur dioxide is produced by air oxidation of sulfur.
The major portion of sulfur dioxide production is in the gaseous form,
although a small percentage is also produced in liquid form. In the
828
-------
TORT.*! 26.38-1 - SUBCATBQORy PRCFILE DATA SUMMARY"
SUBCAOEGOK*
SODIUM SILICOFLUOKIDE
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 vZa'Ssr flow range-:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
NA
51,800 kkg/year
6
1
7,460 kkg/year
3,970 kkg/year
NA
7.5 percent
NA
NA
NA
NA
NA
NA
NA
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
829
-------
gaseous form, it is predominantly used in on-site manufacture of
sulfuric acid. It is also used in the paper and petroleum industries,
as well as for fermentation control in the wine industry, for
bleaching in the textile and food industries, and in the production of
other chemicals. The wastewater source at one plant was a process
effluent from an extraction operation.
The subcategory profile data are given in Table 26.43-1.
Toxic Pollutants
Data has been received on about 33 percent of the industry as a result
of Section 308 letters. No toxic pollutants were found at significant
levels in wastewaters during the screening of one sulfur dioxide
plant.
has been reserved for this
Status of Regulations
Subpart BI (40 CFR 415.610, 5/22/75
subcategory.
Sulfuric Acid Industry
Summary of Determinations
It has been determined that no further effort be given to developing
BPT, BAT, NSPS, and Pretreatment regulations for the Sulfuric Acid
Subcategory. The basis for this determination is that the small
quantities of toxic pollutants found during screening are far below
accepted treatability levels. This determination applies to the
production of sulfuric acid by the contact process from elemental
sulfur only. This subcategory is excluded under Paragraph 8 of the
Settlement Agreement.
Production Processes and Effluents
Sulfuric acid is one of the most extensivley used of all manufactured
chemicals. The major industrial use is in the fertilizer industry,
with on-site captive use of the product as a dominant practice. It is
also used in the manufacturing of plastics, explosives, detergents,
hydrofluoric acid, nuclear fuel and several other organic and
inorganic products. This industry has no process wastewater, but does
have cooling tower blowdown.
The industry profile data for this
26.44-1.
Toxic Pollutants
subcategory are given in Table
Data has been received on about 21 percent of the industry as a result
of Section 308 letters. In addition a sampling survey was made at one
plant which represents less than 1 percent of the industry. Only
nickel and copper were detected but were at levels far below accepted
830
-------
TABLE 26.40-1 -
SUBCAIEGORT PRCFILE DATA SUMMARY
SUBCATEGOKf
SODIUM THIOSULFATE
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 flew range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
NA
NA
6
5
88,000 kkg/year
70,300 kkg/year
NA
NA
4,400 kkg/year
27,000 kkg/year
NA
NA
NA
3 years
51 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 ," JXine, 1978 and "Econonic Analysis of
Proposed Revised Effluent Guidelines and Standards for the Inorganic
Chemicals Industry," March, 1980
NA = Not Available
831
-------
treatability concentrations. They probably result from corrosion and
are concentrated by recycling the cooling water. Apart from waterside
corrosion some corrosion products results from acid leaks. However,
NPDES permits generally require pH control and this will limit this
problem.
Status of Regulations
BPT regulations (40 CFR 415.212) were promulgated on March
These regulations have been remanded by the court.
12, 1974
NSPS and BAT regulations requiring zero discharge {40 CFR 415.212)
were promulgated on March 12, 1974. These regulations have been
remanded by the court. Because no significant quantities of toxic
pollutants are present no further effort will be given to development
of pretreatment regulations for this subcategory.
Zinc Oxide
Summary of Determinations
It has been determined that no further effort be given to developing
BAT, NSPS, or pretreatment regulations for Zinc Oxide Subcategory.
The bases for this determination are 1) only one plant exists the
generates process liquid effluents from the manufacture of zinc oxide
using the wet chemical process, and 2) processes using oxidation of
zinc produce no waterborne wastes. The subcategory is excluded under
Paragraph 8 of the Settlement Agreement.
Production Processes and Effluents
Two major processes are used for the manufacture of zinc oxide: 1)
those involving oxidation of zinc, and 2) those involving
precipitation from solution followed by calcination. Two methods are
used to prepare zinc oxide using the oxidation process— the American
Process and the French Process.
In the American Process, zinc ore is dried and converted to crude
oxide by roasting at approximately 1000 degrees C. Zinc sinter is
introduced into a reaction kiln with equal amounts of coke. The zinc
vapor and carbon monoxide formed are oxidized to zinc oxide and carbon
dioxide and drawn off through ducts to cyclone and baghouse collection
equipment. No waterborne wastes are generated.
In the French Process, crude zinc oxide sinter and dried coke are
mixed with binder and fed through briquetting rolls. The raw
briquettes are fed into cokers operating at temperatures between 500
and 1000 degrees C. Zinc sinter is converted to zinc metal in vapor
using electrical ore vaporizers or rotary burners. The zinc vapors
are purified to remove lead and cadmium impurities. The purified zinc
is then vaporized and reacted with oxygen to produce zinc oxide, which
is recovered by dry collection methods, cooled, and packaged. No
waterborne wastes are generated.
832
-------
TABLE 26.42-1 -
SUBCKEEOOHlf PRCFUE DATA SUMMARY
SUBCA2EGQKT
STRONG NITRIC ACID
Tbtal subcatecpry 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
Volume per unit product:
Minimum
Maximum
NA
NA
NA
5
155,200 kkg/year
121,000 kkg/year
NA
NA
5,300 kkg/year
60,200 kkg/year
NA
NA
NA
11 years
49 years
NA
NA
NA
NA
Sources of data are Stanford Research Institute, Directory of Chemical
Producers/ U.S.A., 1977, U.S. Department of Ccnroerce, 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 "Eooncmic Analysis of
Proposed Revised Effluent Guidelines and Standards for the Inorganic
Chemicals Industry/" March, 1980
NA = Not Available
833
-------
In the wet chemical process, crude zinc oxide recovered from lead
smelters is used as the raw material. The zinc oxide is leached with
caustic soda solution to remove sulfate and dissolve lead saltsr^ The
undissolved zinc oxide is then recovered from the leaching mixture,
washed, neutralized to remove alkali, dried, calcined, and packaged.
Waterborne 'wastes are generated from the deleading step, desulfating
step, etc.
Plants
There are about 20 zinc oxide producers in the United States. The
producers include both primary and secondary. The final product of
the primary producers is zinc oxide, while the secondary producers
manufacture zinc oxide and use it to make zinc chloride and zinc
nitrate. About 80-90 percent of the total zinc oxide produced is made
by the American and French Processes, and 10 percent is made by one
plant using the wet chemical process.
Zinc Sulfate
Summary of Determinations
It has been determined that no further effort be given to developing
revised BAT or NSPS regulations for the Zinc Sulfate Subcategory. The
basis for this determination is that existing BPT regulations specify
zero discharge of process wastewaters to navigable waters.
Pretreatment standards will be developed in Phase II.
Production Processes and Effluents
Zinc sulfate is produced by reaction of sulfuric acid with various
crude zinc starting materials, such as zinc oxide from brass mill
fumes, zinc metal residues from various sources, and zinc carbonate
by-product from sodium hydrosulfite manufacture. The following basic
steps are followed: reaction of the zinc containing raw material with
refiltration of solids, and either evaporation to dryness or sale as
solution grade. Liquors from the preceding processes are, in some
cases, refined to recover by-products and other wastewaters are
recycled. The only wastes are filter cake residues.
Plants
There are eighteen plants producing zinc sulfate and none are known to
discharge wastes from ,the process system.
BPT Limitations
Regulations were promulgated on May 22, 1975 ( 40 CFR 415.632)
requiring no discharge of process wastewater pollutants to navigable
waters. The discharge of contaminated non-process wastewater is
permitted. This includes rainfall runoffs, accidental spills,
accidental leaks, and discharges related to personal safety equipment.
All reasonable measures must be taken to prevent, reduce and control
834
-------
such contact to the maximum degree feasible, and to mitigate the
effects of such contact once it has occurred.
BAT and NSPS Limitations
BAT and NSPS guidelines were proposed on May 22, 1975, requiring zero
discharge of process wastes to navigable waters. Since BPT already
requires zero discharge, BAT and NSPS are excluded under Paragraph 8
of the Settlement Agreement.
835
-------
TABLE 26.43-1 -
PRCFUE DATA SUMMARY
SUBCATEGOR*
SULFUR DIOXIDE
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
Maxinum
Volume per unit product:
Minimum
Maximum
NA
NA
15
5
453,000 kkg/year
364,000 kkg/year
NA
NA
27,800 kkg/year
170,000 kkg/year
NA
NA
NA
3 years
51 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 Assesanent 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
836
-------
TABLE 26.44-1 -
SUBCAIEGORY PROFUSE DATA SUMMARY
SUBCATEGOEY
SULFURIC 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:
Minimum
Maximum
Waste water flow range:
Minimum
Maximum
Volume per unit product:
Minimum
Maximum
33,619,000 kkg/year
NA
109
52
7,758,000 kkg/year
6,308,000 kkg/year
23 percent
NA
5,300 kkg/year
47,700 kkg/year
NA
NA
NA
3 years
78 years
NA
NA
NA
NA
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1979 , U.S. Department of Ocnmerce, Current Industrial
Reports, Decenber 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chanical Industry ," June, 1978 and "Economic Analysis of
Proposed Revised Effluent Guidelines and Standards for the Inorganic
Chemicals Industry," March, 1980
NA = Not Available
837
-------
838
-------
839
-------
840
-------
REFERENCES
1. U.S. Environmental Protection Agency. Major Inorganic Products,
Development Document. EPA-440/l-74-007a, 1974.
2. U.S Environmental Protection Agency. Development Document for
Interim Final Effluent Limitations Guidelines and Proposed New Source
Performance Standards for the Significant Inorganic Products.
EPA-440/1-75-037, 1975. 358 Pp.
3. Calspan Corp. Addendum to Development Document for Effluent
Limitations Guidelines and New Source Performance Standards. Major
Inorganic Products Segment of Inorganic Chemicals Manufacturing Point
Source Category, Contract No. 68-01-3281, 1978.
4. Sampling Screening Procedure for the Measurement of Priority
Pollutants. U.S, Environmental Protection Agency, 1976. 6 Pp.
5. Coleman, R.T., J.D. Colley, R.F, Klausmeiser, D.A. Malish, N.P.
Meserole,, W.C. Micheletti, and K. Schwitzgebel. Treatment Methods for
Acidic Wastewater Containing Potenially Toxic Metal Compounds. EPA
Contract No. 68-02-2608, U.S. Environmental Protection Agency, 1978.
220 Pp.
6. Kraus, K.A., and H.O. Phillips. Processes for Removal and/or
Separation of Metals from Solutions. U.S. Patent 3,317,312, U.S.
Patent Office, May 2, 1:967. 9 Pp.
7. Scott, M.C. Heavy Metals Removal at Phillips Plating. WWEMA
Industrial Pollution Conference, St. Louis, Missouri, 1978. 16 Pp.
8. Scott, M.C. Sulfext, - A New Process Technology for Removal of
Heavy Metals from Waste .Streams. The 32nd Annual Purdue Industrial
Waste Conference, Lafayette, Indiana, 1977. 17 Pp.
9. Patterson, J.W., and R.A, Minear. Wastewater Treatment
Technology. Illinois Institite of Technology,, 1973.
10. Patterson, J.W. Wastewater Treatment Technology. Ann Arbor
Science Publishers, Inc. Ann Arbor, Michigan, 1975.
11. Schlauch, R.M., and A.C. Epstein. Treatment of Metal Finishing
Wastes by Sulfide Precipitaion. EPA-600/2-75-049, U.S. Environmental
Protection Agency, 1977. 89 Pp.
12. Campbell, H.J., Jr., N.C. Scrivner, K. Batzar, and R.F. White.
Evaluation of Chromium Removal from a Highly Variable Wastewater
Stream. tcThe 32nd Annual Purdue Industrial Waste Conference,
Lafayette, Indiana, 1977. 38 Pp.
-------
13. Wing, R.E., C.L. Swanson, W.M. Doane, and C.R. Russell. Heavy
Metal Removal with Starch Xanthate-Cationic Polymer Complex. J. Water
Pollution Control Federation, 46 (8): 2043-2047, 1974.
14. Wing, R.E. Heavy Metal Removal from Wastewater with Starch
Xanthate, In: Proceedings of the 29th Annual Purdue Industrial Waste
Conference, Lafayette, Indiana, 1974. Pp. 348-356.
15. Wing, R.E. Removal of Heavy Metals from Wastewater with Starch
Xanthate-Cationic Polymer Complex. The 46th Annual Conference of the
Water Pollution Control Federation, Cleveland, Ohio, 1973. 38 Pp.
16. Wing, R.E. Removal of Heavy Metals from Wastewater with Starch
Xanthate. Presented at the Traces of Heavy Metals in Water: Removal
and Monitoring Conference, Princeton, New Jersey, 1973. Pp. 258-273.
17. Swanson, C.L., R.E. Wing, W.M. Doane, and C.R. Russell. Mercury
Removal from Wastewater with Starch Xanthate Cationic Polymer Complex.
Environmental Science & Technology 7(7): 614-619, 1973.
18. Hanway, J.E., Jr., R.G. Mumford, and D.G. Barth. A Promising New
Process for Removing Heavy Metals from Wastewater. Civil
Engineering-ASCE 47(10): 78-79, 1976.
19. Hanway, J.E., Jr., R.G. Mumford, and P.N, Mishra. Treatment of
Industrial Effluents for Heavy Metals Removal Using the Cellulose
Xanthate Process. The 71st Annual Meeting of the American Institute
of Chemical Engineers, Miami, Florida, 1978. 21 Pp.
20. Wing, R.E., L.L. Navickis, B.K. Jasberg, and W.E. Rayford.
Removal of Heavy Metals from Industrial Wastewaters Using Insoluble
Starch Xanthate. EPA-600/2-78-085, U.S. Environmental Protection
Agency, 1978. 116 Pp.
21. DeJong, G.J., and Ir. C.J.N Rekers. The Akzo Process for the
Removal of Mercury from Wastewater. Journal of Chromatography 102:
443-451, 1974.
22. Van der Heem, P. The Removal of Traces of Heavy Metals from
Drinking Water and Industrial Effluent with Ion Exchangers. The
Regional American Chemical Society Meeting, 1977. 16 Pp.
23. Chemical Marketing Recorder, July 24, 1978.
24. Slen, T.T., M. Chem, and J. Lauber. Incineration of Toxic
Chemical Wastes. Pollution Engineering 10(10):42, 1978.
25. TRW Systems Group. Recommended Methods of Reduction,
Neutralization, Recovery or Disposal of Hazardous Waste. NTIS
PB-224589, 1973.
842
-------
26. Ellerbusch, F,, and H.S. Skrovronek. Oxidative Treatment of
Industrial Wastewater. Industrial Water Engineering 14(5):20-29,
1977.
27. Knopp, P.V., and T.L. Randall. Detoxification of Specific Organic
Substances by Wet Oxidation. The 51st Annual Conference of Water
Pollution Control Federation, 1978.
28, Arthur D. Little, Inc. Treatment Technology Handbook.
29. Schell, W.J. Membrane Ultrafiltration for Water Treatment.
Envirogenics Systems Co.
30. Vanderborght, B.M., and R.E. Van Grieken. Enrichment of Trace
Metals in Water Adsorption of Activated Carbon. Analytical Chemistry
49 (2):311-316, 1977.
31. Cheremisinoff, P.N., and F. Ellerbusch. Carbon Adsorption
Handbook. Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan,
1978.
32. Jacobs Engineering Group Inc. Study of the Application of BAC to
Industrial Wastewater. Office of Water Research and Technology, U.S.
Environmental Protection Agency, 1978.
33. Otsubo, K., S. Yamazaki, and Y. Sakuraba. Advanced Water
Treatment for Fluoride-Containing Wastewater. Hitachi Hyoron
58(3):219-, 1976. Trans. For Rockwell Intl.
34. Zabban, W., and H.W. Jewett. The Treatment of Fluoride Wastes.
In: Proceedings of the 22nd Annual Purdue Industrial Waste
Conference, Layfayette, Indiana, 1967. Pp. 706-716.
35. Rubel, F., Jr., and R.D. Woosley. Removal of Excess Fluoride from
Drinking Water. EPA-570/9-78-001. U.S. Environmental Protection
Agency, 1978. 16 Pp.
36. Wu, Y.C. Activated Alumina Removes Fluoride Ions From Water.
Water Sewage Works 125(6):76-82, 1978.
37. Maier, F.J. Partial Defluoridation of Water. Public Works 91
(11), 1960.
38. Maier, F.J. New Fluoride Removal Method Cuts Costs, Engineering
News-Record 148 (24):40, 1952,
39. Kennedy, D.C., M.A. Kimler, and C,A. Hammer. Functional Design of
a Zero-Discharge Wastewater Treatment System for the National Center
for Toxicoligical Research. In: Proceedings of the 31st Annual
Purdue Industrial Waste Conference, Lafayette, Indiana, 1976. Pp.
823-830.
843
-------
40. Hannah, S.A., M. Jelus, and J.M. Cohen. Removal of Uncommon Trace
Metals by Physical and Chemical Treatment Processes. Journal Water
Pollution Control Federation 49( 11 ) .-2297-2309, 1977,
41. Maruyama, T., S.A. Hannah, and J.M. Cohen. Removal of Uncommon
Trace Metals by Physical and Chemical Treatment Processes. Journal
Water Pollution Control Federation 49(11):2297-2309, 1977.
42. Gulledge, H.H., and J.T. O'Connor. Removal of Arsenic (V) from
Water by Adsorption on Aluminum and Ferric Hydroxides. Journal
American Water Works Association 65 (8):548-552, 1973.
43. Gupta, S., and K.Y. Chen. Arsenic Removal by Adsorption. Journal
Water Pollution Control Federation 50 (3}:493, 1978.
44. Larsen, H.P., J.K. Shou, and L.W. Ross. Chemical Treatment of
Metal-Bearing Mine Drainage. Journal Water Pollution Control
Federation 45 (8):1682-1695, 1973
45. Nilsson, R. Removal of Metals by Chemical Treatment of Municipal
Wastewater. Water Research 5:51-60, 1971
46. Sorg, T.J., O.T. Love, and G.S. Logsdon. Manual of Treatment
Techniques for Meeting the Interim Primary Drinking Water Regulations.
EPA-600/8-77-005. U.S. Environmental Protection Agency, 1977. 73 Pp.
47. Colley, J.D., C.A. Muela, M.L. Owen, N.P. Meserole, J.B. Riggs,
and J.C. Terry. Assessment of Technology for Control of Toxic
Effluents from the Electric Utility Industry. EPA-600/7-78-090. U.S.
Environmental Protection Agency, 1978.
48. Smithson, G.R., Jr. An Investigation of Techniques for Removal of
Chromium from Electroplating Wastes. EPA 12010 EIE, U.S.
Environmental Protectin Agency, 1971, 91 Pp.
49. Patterson, J.W., H.E. Allen, and J.J. Scala. Carbonate
Precipitation for Heavy Metals Pollutants. Journal Water Pollution
Control Federation 49{12) : 2397-2410, 1977.
50. Sabadell, J.E. Traces of Heavy Metals in Water Removal Processes
and Monitoring. EPA-902/9-74-001. U.S. Environmental Protection
Agency, 1973.
51. U.S. Environmental Protection Agency. Environmental Multi-Media
Assessment of Selected Industrial Inorganic Chemicals. EPA Contract
No. 68-03-2403, 1977.
52. JRB Associates, Inc, An Assessment of pH Control of Process
Waters in Selected Plants. Draft Report to the Office of Water
Programs, U.S. Environmental Protection Agency, 1979.
53. Energy and Environmental Analysts, Inc. Economic Analysis of
Proposed Revised Effluent Guidelines and Standards for the Inorganic
844
-------
Chemicals Industry. Prepared for Office of Water Planning and
Standards, U.S. Environmental Protection Agency, 1980.
54. Kirsch, E.J., J.E. Etzel J. Water Pollution Control Federation,
Vol. 45, 359, 1973.
55. E.I. Dupont de Nemours & Company, Development Document for
Effluent Limitations Guidelines and New Source Performance Standards
for Simultaneous Beneficiation-Chlorination Process of Ti02
Manufacture, Pigments Department, Edge Moor, Delaware, 1977.
56. U.S. Environmental Protection Agency, Prodeedings: Seminar on
Analytical Methods for Priority Pollutants, Denver. 1977.
57. SRI International, Report No. 61B, Chlorine, Supplement B, by
Yen-Chen Yen, (a private report by the Process Economics Program),
Menlo Park, California, 1978.
58. U.S. Environmental Protection Agency, Development Document for
Proposed Effluent Limitations Guidelines, New Source Performance
Standards and Pretreatment Standards for the Steam Electric Point
Source Category, (Effluent Guidelines Division, Office of Water and
Waste Management, Washington, D.C.), 1979.
59, U.S. Environmental Protection Agency, Quality Criteria for Water,
(EPA-44/9-76-023), Washington, D.C., 1976.
60. U.S. Environmental Protection Agency, Development Document for
Proposed Effluent Limitations Guidelines and Standards for the
Inorganci Chemicals Manufacturing Point Source Category, {Effluent
Guidelines Division, Office of Water and Waste Management, Washington,
D.C.), 1980.
61. U.S. Environmental Protection Agency, Treatability Studies for the
Inorganic Chemicals Manufacturing Point Source Category, (Effluent
Guidelines Division, Office of Water and^Waste Management, Washington,
D.C.), 1980.
62. E. I. duPont de Nemours & Company, Comments on Proposed
Development Document for Effluent Limitations Guidelines and Standards
for the Titanium Dioxide - Chloride Ilmenite Process Subcategory of
the Inorganic Chemicals Point Source Category, Wilmington, Delaware,
1980.
63. U.S. Environmental Protection Agency, Economic Impact Analysis of
Pollution Control Technologies for Segments of the Inorganic Chemicals
Manufacturing Industry. EPA 440/2-81-023, 1982.
84!
-------
846
-------
BIBLIOGRAPHY
Andco Environmental Processes, Inc. Andco Chromate and Heavy Metal
Removal System. 1977
Anderson, J.R., and C.O. Weiss. Method for Precipitation of Heavy
Metal Sulfides. U.S. Patent 3,740,331. U.S. Patent Office. June 19,
1973.
Birkett, J.D. Electrodialysis
Engineering 14 (5):6-9, 1977.
An Overview. Industrial Water
Bowen, L.B., J.H. Mallison, and J.H. Cosgrove. Waste Recovery: Zinc
Recovery from Rayon Plant Sludge. Chemical Engineering Progress
73(5):50-54, 1977.
Case, O.P. Metallic Recovery from Wastewaters Utilizing Cementation.
EPA 670/2-74-008. U.S. Environmental Protection Agency, 1974. 36 Pp.
Cheng, M.H., J.W. Patterson, and R.A. Minear. Heavy Metals Uptake by
Activated Sludge. Journal Water Pollution Control Federation
47(2) :362-376, 1975.
Cheremisinoff, P.N., and Y.H. Habib. Cadmium, Chromium, Lead,
Mercury: A plenary Account for Water Pollution. Part 2 Removal
Techniques. Water and Sewage Works 119(8):46-51, 1972.
Cohen, J.M. Trace Metal Removal by Wastewater Treatment.
Technology Transfer, 1977.
EPA
Current Industrial Reports Inorganic Chemicals. U.S. Department of
Commerce Bureau of the Census, 1977.
Davis, H.J., F.S. Model, and J.R. Leal. PBI Reverse Osmosis Membrane
for Chromium Plating Rinse Water. EPA-600/2-78-040. U.S.
Environmental Protection Agency, 1978. 35 Pp.
Dean, J.G., F.L. Bosqui, and K.H. Lanouette. Removing Heavy Metals
from Wastewater. Environmental Science and Technology 6(6}:518-522,
1972.
Dow Chemical Company. Selecting Ion Exchange Resins for Water
Treatment Applications. Industrial Water Engineering 14(5):11-13,
1977.
Energy and Environmental Analysis, Inc. Draft Preliminary Economic
Assessment of Effluent Limitations . in the Inorganic Chemicals
Industry. 1978.
FMC Corp. Industrial Wastewater Treatment. A guidebook to Hydrogen
Peroxide for Industrial Wastes. Philadelphia, Pennsylvania.
847
-------
FMC Corp. Hydrogen Peroxide.
Philadelphia, Pennsylvania.
Chemical
and Technical Data.
Ford, D.L. Putting Activated Carbon in Perspective to 1983
Guidelines. Industrial Water Engineering 14(3):20-27, 1977.
General Technology Corporation. Draft Contractor Document for Major
Inorganic Chemicals. 1973.
General Technology Corporation. Final Contractor Document for Major
Inorganic Chemicals. 1973.
Greek, B.F., and W.F. Fallwell. Chlorine, Major Alkalies Still in
Doldrums. Chemical and Enginering News 56(6):8-ll, 1978.
Grover, P. A Waste Stream Management System. Chemical Engineering
Progress 73(12):71-73, 1977.
Haggenmacher, J.H. Chromate Removal Decreases Slowdown to Below 0,05
ppm Hexavalent Chrome. Chemical Processing, 1977.
Hoyle, D.L. Designing for pH Control. Chemical Engineering
83(24):121-126, 1976.
Jacobs Engineering Group Inc. Estimating Standards.
Jenkins, S.H., D.G. Keight, and R.E. Humphreys. The Solubility of
Heavy Metal Hydroxides in Water, Sewage and Sewage Sludge - I. The
Solubi1ity of Some Metal Hydroxides. Int. J. Air Wat. Pol1.
8:537-556, 1964.
Jenkins, S.H,, D.G. Keight, and A. Ewins. The Solubility of Heavy
Metal Hydroxides in Water, Sewage, and Sewage Sludge - II The
Precipitation of Metals by Sewage. Int. 3. Air Wat. Poll.
8:679-693, 1964.
Keating, E.J., R.A. Brown, and E.S. Greenberg. Phenolic Problems
Solved with Hydrogen Peroxide Oxidation. The 33rd Annual Industrial
Waste Conference, West Lafayette, Indiana, 1978. 20 Pp.
Kibble, W.H. Hydrogen Peroxide Helps Solve Industrial Wastewater
Problems. Industrial Wastes 26-29, 1978,
Knocke, W.R., T. Clevenger, M.M. Ghosh, and J.T. Novak, Recovery of
Metals from Electroplating Wastes. The 33rd Annual Industrial Waste
Conference, West Lafayette, Indiana. 1978. 37 Pp.
Lanuouette, K.H. Heavy Metals Removal. Chemical Engineering Deskbook
Issue 84{22):73-80, 1977.
Linstedt, K.D., C.P. Houck, and J.T. O'Connor. Trace Element Removals
in Advanced Wastewater Treatment Processes. Journal Water Pollution
Control Federation 43(7):1507-1513, 1971.
848
-------
Mir, L., W. Eykamp, and R.L. Goldsmith. Current and Developing
Applications for Ultrafiltration. Industrial Water Engineering
14(3):14-19, 1977.
Morton, S.D., and E.W. Sawyer. Clay Minerals Remove Organics, Viruses
and Heavy Metals from Water. Water and Sewage Works 123 (Reference
No. 1976):R117-R120, 1976.
Porter, J.W., Brothers, G.W., and W.B. Whitton. Cost Estimating
Guidelines for Wastewater Treatment Systems. EPA 17090 DRU. U.S.
Environmental Protection Agency, 1970. 96 Pp.
Resources Conservation Co.
Washington
The RCC Brine Concentrator. Renton,
Richardsons Estimating Standards. Means Cost Data. 1978.
Rizzo, J.L,, and A.R. Shepard. Treating Industrial Wastewater with
Activated Carbon. Chemical Engineering 84(1):95-100, 1977.
Rosenzweig, M.D. Mercury Cleanup Routes - I Chemical Engineering
82(2):60-61, 1975.
Sabadell, J.E,, ed. Traces of Heavy Metals in Water Removal Processes
and Monitoring. EPA-902/9-74-001. U.S. Environmental Protection
Agency, 1973. 342 Pp.
Shen, Y.S. Study of Arsenic Removal from Drinking Water. Journal
American Water Works Association 65(8): 543-548, 1973.
Sigworth, E.A., and S.B. Smith. Adsorption of Inorganic Compounds by
Activated Carbon. Journal American Water Works Association
64(6):386-390, 1972.
Smith, R. Cost of Conventional and Advanced Treatment of Wastewater.
Journal Water Pollution Control Federation 40(9):1546-1574, 1968.
Sonksen, M.K., P.M. Sittig, E.F. Maziarz. Treatment of Oily Wastes by
Ultrafiltration/Reverse Osmosis. A Case History. The 33rd Annual
Industrial Waste Conference, West Lafayette, Indiana, 1978. 28 Pp.
Stanford Research Institute. Directory of Chemical Producers U.S.A.
1977.
Stinson, G.F. Electrochemical Wastewater Treatment Process. Personal
Communication. 1978.
Theim, L.D. Badorek, and J.T. 0'Connor. Removal of Mercury From
Drinking Water Using Activated Carbon. Journal American Water Works
Association 68 (8):447-451, 1976.
84 9
-------
Thompson, G.S., Jr. First Annual Conference on Advanced Pollution
Control for the Metal Finishing Industry. EPA-600/8-78-010. U.S.
Environmental Protection Agency, 1978. 143 Pp.
S.K. Williams Company. Wastewater Treatment and Reuse in a Metal
Finishing Job Shop. EPA-670/2-74-042. U.S. Environmental Protection
Agency, 1978. 143 Pp.
Wiseman, S., and B.T. Bawden. Removing Refractory Compounds
Wastewater. Chemical Engineering Progress 73(5):60-64, 1977.
From
Zabban, W., and R. Helwick. Defluoridation of Wastewater. The 30th
Annual Industrial Waste Conference, West Lafayette, Indiana, 1975. 38
Pp.
Zievers, J.F., R.E. Wing, S. Grain, and D. Gordon. Practical Use of
Starch Insoluble Xanthates. Industrial Filter and Pump Mfg. Co. 1977.
20 Pp.
850
-------
APPENDIX A
ANALYSIS OF LONG-TERM EFFLUENT
MONITORING DATA FOR
THE INORGANIC CHEMICALS INDUSTRY
Introduction
This appendix contains tabulated summaries of the statistical
parameters derived from the analysis of long-term effluent monitoring
data collected by industry and reported to the EPA or State regulatory
agencies during the last two or three years. The particular sets of
data selected for analysis are taken from plants which apply a well
defined treatment technology to process wastewaters from single
product or product group manufacturing operations associated with a
specific subcategory. Data have been excluded which represent
wastewaters diluted with noncontact cooling water or commingled with
waste sources from unrelated products. Each table in the appendix
indicates the actual number of observations on which the calculated
statistical parameters are based. The derivation of the parameters
was discussed in Section 8.2 of the development document.
The statisitcal performance information presented here was taken into
consideration in the development of limitations for each subcategory
studied in detail in the main report. These were expressed as the
Concentration Bases (mg/1) and Effluent Unit Mass Loadings (kg/kkg)
for each pollutant assuming the model plant flow conditions and
applying the specified pollutant removal technologies at e'ach level of
treatment. The tables on the following pages summarize the available
historical effluent monitoring results and give the individual plant
performance characteristics in concentration and loading units for
both daily and monthly measurements. Variability factors shown on
each table were used to calculate the plant "Performance Standards"
shown in the right hand column of each table. Similarly, the
Variability Factor Ratio (VFR) used in this report is the variability
factor for daily measurements divided by variability factor for 30-day
average data.
In general, the effluent data base acquired included NPDES monitoring
data collected during the period from January 1, 1975 through June 30,
1976. Firms who monitored over this time period provided up to 18
months of 30-day average data and as many as 547 measurements of daily
or 24-hour data. In cases where monitoring was done less frequently
than daily, perhaps omitted on weekends, or only weekly measurements,
the actual number of observations used in the calculation is recorded
for each parameter.
Included in Appendix A are statistical measures appropriate to the
analysis of long-term monitoring data and the historical performance
of inorganic chemical pollutant discharge levels. The statistics
presented include measures of the amount or level of pollutant
A-l
-------
discharge, such as long-term average, minimum level, and maximum level
for both daily, or 24-hour measurements, as well as 30-day average
measurements.
Also given in the table is the coefficient of variation, CV, which
reflects the dispersion of measurements above and below the long-term
average level. Other measures of variability that may be of interest,
such as range or standard deviation are also calculated for any
parameter from any information given herein. In addition to
statistics of pollutant level and variation of pollutant level,
variability factors are given for each parameter. A variability
factor is the ratio of an upper percentile of the distribution of
pollutant measurements to the long-term average pollutant level. The
basis of the particular upper percentile chosen for variability
factors is explained as a footnote to the table.
The historical performance of each firm, using the variability factor,
is given for each parameter and is expressed in the same units as the
long-term average.
For reference, the tables in Appendix A are organized by inorganic
chemical subcategory and the manufacturing process in that
subcategory. For each plant, as many as six tables are included.
These tables appear in the following order.
1. Daily measurements of pollutant concentrations in effluent stream
given in parts per million (ppm).
2. Daily measurements of - total effluent discharge load measured in
kilograms per day.
3. 30-day averages of pollutant concentration (ppm).
4. 30-day averages of total effluent pollutant load (kg/day).
5. Daily measurements of pollutant unit loadings in the effluent
streams given in kilograms of pollutant per thousand kilograms of
product (kg/kkg).
6. 30-day averages of pollutant unit loadings in the effluent stream
given in kilograms of pollutant per thousand kilograms of product
(kg/kkg).
A-2
-------
Table A-la
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Mercury Cell Process
Plant tA
Historical Summary Variability Performance
Parameter Statistics Factors Standards
(mg/1) No Min Avg Max CV * p
Mercury 530 .006 .014 .021 .286 1.88 .026
TSS 530 1.00 7.4 62. .581 3.04 22.5
Chlorine 428 0.08 .638 1.50 .463 2.28 1.46
(Total Residual)
* - 99% of the daily maximum measurements expected to be less
than the performance standard, p.
Table A-lb
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Mercury Cell Process
Plant #A
Historical Summary Variability Performance
Parameter Statistics Factors Standards
(kg/day) No Min Avg Max CV * P
Mercury 530 .015 .031 .047 .129 1.66 .051
Chlorine 420 .156 1.44 3.40 .463 2.54 3.65
(Total Residual)
* _
99% of the daily maximum measurements expected to be less
than the performance standard, P.
A-3
-------
Table A-lc
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Chlorine
Mercury Cell Process
Plant #A
Parameter
(rag/1)
Mercury
TSS
Historical Summary Variability
Statistics Factors
No Min Avg Max CV *
18 .008 .014 .020 ,293 1.47
18 5.1 7.4 12.9 .355 1.58
Chlorine 18 .380 .638 .847 .194 1.38
(Total Residual)
Performance
Standards
P
.021
11.7
0.88
95% of the monthly averages are expected to be within the
performance standard, P.
Table A-ld
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Chlorine
Mercury Cell Process
Plant #A
Historical Summary
Variability Performance
Parameter
(kg/day)
Mercury
Chlorine
No
18
18
St
Min
.020
.91 1
atist
Avg
.031
.44 2
ics
Max CV
.037 .197
.23-
Factors
*
1.33
1.50
Standards
P
.041
2.16
95% of the monthly averages are expected to be within the
performance standard, P.
A-4
-------
Table A-le
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Mercury Cell Process
Plant #A
Parameter
(g/kkg) No Min
Historical Summary
Statistics
Variability Performance
Factors Standards
Avg
Max CV
Mercury
530 .027 .055 .084
Chlorine 420 .00028 .0026 .006
(Total Residual)
.090
.006
* -
99% of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-lf
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Chlorine
Mercury Cell Process
Plant fA
Parameter
Historical Summary
Statistics
No Min Avg Max CV
Variability
Factors
Performance
Standards
Mercury 18 .035 .055 .065
Chlorine 18 1.6 2.52 3.91
.072
3.8
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-5
-------
Table A-2a
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Mercury Cell Process
Plant #B
Parameter
(ug/1)
Historical Summary
Statistics
Variability Performance
Factors Standards
No Min Avg Max CV
Mercury
516 .041 .634 2.87 .910
4.52
2.87
99% of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-2b
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Mercury Cell Process
Plant #B
Parameter
(kg/day)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Mercury
516 .0005 .011 .088 .818
4.35
.046
- 99% of the daily maximum measurements expected to be less
than the performance standard, P.
A-6
-------
Table A-2c
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Chlorine Subcategory
Mercury Cell Process
Plant #B
Parameter
(ug/1)
Historical Summary
Statistics
Variability Performance
Factors Standards
No Min Avg Max CV
Mercury
17 .325 .634 1.15 .293
1.45
0.919
* - 95% of the monthly averages are expected to be within the
performance standardr P.
Table A-2d
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Chlorine
Mercury Cell Process
Plant IB
Parameter
(kg/day)
Historical Summary
Statistics
Variability Performance
Factors Standards
No Min Avg Max CV
Mercury
17 .005 .011 .019
1.45
.015
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-7
-------
Table A-2e
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Mercury Cell Process
Plant fB
Parameter
(9/kkg)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Mercury
516 .0037 .082 .658
.344
- 99% of.the daily maximum measurements expected to be less
than the performance standard, P.
Table A-2f
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Chlorine
Mercury Cell Process
Plant tfB
Parameter
(g/kkg)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Mercury
17 .037 .082 .14
,11
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-8
-------
Table A-3a
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Mercury Cell Process
Plant fC
Parameter
(mg/1)
Historical Summary
Statistics
Variability Performance
Factors Standards
No Min Avg Max CV
Mercury
349 .0005 .014 .136 2.29
9.45
0.132
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-3b
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Mercury Cell Process
Plant fC
Parameter
(kg/day)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Mercury
349 .0001 .003 .088 2.33
10.22
.028
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
A-9
-------
Table A-3c
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Chlorine
Mercury Cell Process
Plant fC
Parameter
(mg/1)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Mercury
17 .0009 .014 .062 1.21
2.99
.042
95% of the monthly averages are expected to be within the
performance standard, P.
Table A-3d
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Chlorine
Mercury Cell Process
Plant *G
Historical Summary
Statistics
Parameter
(kg/day) No Min Avg Max CV
Variability Performance
Factors Standards
Mercury
17 .0002 .003 .014 1.33
3.22
.0088
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-10
-------
Table A-3e
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Mercury Cell Process
Plant fC
Parameter
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Mercury
349 .0006 .017 .485
.154
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-3f
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Chlorine
Mercury Cell Process
Plant #C
Parameter
(g/kkg)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Mercury
17 .0011 .016 .077
.0485
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-ll
-------
Table A-4a
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Mercury Cell Process
Plant #D
Historical Summary
Variability Performance
Parameter Statistics Factors
(mg/1) No Min Avg Max CV *
Mercury
82 .002 .004 .011 .500 2.24
Chlorine 49 2.0 19.1 62 1.01 4.96
(Total Residual)
Standards
P
0.009
94.7
99% of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-4b
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Mercury Cell Process
Plant #D
Parameter
(kg/day)
Historical Summary
Statistics
Variability
Factors
No Min Avg Max CV
Performance
Standards
Mercury
82 .021 .047 .118 .383
Chlorine 49 20.5 203 663 1.03
(Total Residual)
2.20
5.04
.104
1026
99% of the daily maximum measurements expected to be less
than the performance standard, P.
A-12
-------
Table A-4c
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Chlorine
Mercury Cell Process
Plant #D
Historical Summary
Variability Performance
Parameter
(mg/1)
Mercury
Statistics
No
22
Chlorine 14
(Total Residual)
Min
.003
4.0
Avg
.004
19.1
Max
.008
57.8
Factors Standards
CV * p
.250 1.60 0.006
.969 2.91 55.6
95% of the monthly averages are expected to be within the
performance standard, P.
Table A-4d
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Chlorine
Mercury Cell Process
Plant fD
Historical Summary
Variability Performance
Parameter
(kg/day)
Mercury
Statistics Factors Standards
No
22
Chlorine 14
(Total Residual)
Min
.032
39.1
Avg
.047
203
Max cv * P
.098 .340 1.66 .079
616 .945 2.89 588
95% of the monthly averages are expected to be within the
performance standard, P.
A-13
-------
Table A-4e
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Mercury Cell process
Plant |D
Parameter
(gAkg)
Historical Summary
Statistics
Variability Performance
Factors Standards
No Min Avg Max CV
Mercury
82 .386 .864 2.17
1.91
Chlorine
(Total Residual)
99% of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-4f
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Chlorine
Mercury Cell Process
Plant fD
Parameter
Historical Summary
Statistics
No Min Avg Max CV
Variability Performance
Factors Standards
Mercury
22 .588 .864 1.8
Chlorine
(Total Residual)
1.45
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-14
-------
Table A-5a
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Diaphram Cell Process
Plant fE
Parameter
(kg/day)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Lead
153 .045 1.42 5.40
4.12
5.85
99% of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-5b
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Chlorine
Diaphram Cell Process
Plant flE
Historical Summary Variability
Parameter Statistics Factors
(kg/day) No Min Avg Max CV *
Performance
Standards
P
Lead
12 .460 1.42 5.40 .824
1.58
2.25
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-15
-------
Table A-5e
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Chlorine
Diaphram Cell Process
Plant #E
Parameter
(g/kkg)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Lead
153 .205 6.46 24.6
26.6
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-5f
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Chlorine
Diaphram Cell Process
Plant fE
Parameter
(g/kkg)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Lead
12 2.09 6.46 24.6
10.24
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-16
-------
Table A-7a
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Hydrofluoric Acid/
Plant $G
Historical Summary
Variability Performance
Parameter
(kg/day)
Fluoride
TSS
Statistics
No
15
16
Min
4.54
7.26
Avg
16
28
.7
.6
Max
27
52
.2
.2
Factors
CV *
.449 X.74
.441 1.72
Standards
P
29.0
49.2
* - 95% of the monthly averages are expected to be within the
performance standard, P.
Table A-7e
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Hydrofluoric Acid/
Plant #G
Historical Summary
Variability Performance
Parameter
(g/kkg)
Fluoride
TSS
Statistics Factors Standards
No
15
16
Min Avg
99.1 365
158.5 624
Max CV *
594
1140
P
633
1074
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-17
-------
Table A-8b
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Titanium Dioxide
Chloride Process
Plant ffH
Historical Summary
Variability Performance
Parameter
(kg/day)
Chromium
Copper
Zinc
TSS
No
394
394
394
394
Stc
Min
.000
.000
.000
0.40
atist:
Aver
.013
.027
.028
8.34
Lcs
Max
.210
.190
.108
176.
CV
1.69
1.04
.679
1.92
Fa
7
5
3
8
ctors
*
.78
.20
.42
.35
Standards
P
.097
.139
.097
69.7
- 99% Of the daily maximum measurements expected to be less
than the performance standard, P.
A-18
-------
Table A-8c
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Titanium Dioxide
Chloride Process
Plant £H
Historical Summary
Variability Performance
Parameter
(mg/1)
Chromium
Copper
Zinc
TSS
Statistics
No
13
13
13
13
Min
.000
.000
.001
1.20
Aver
.004
.010
.012
3.14
Max
.013
.030
.026
8.60
CV
.750
.700
.500
.599
Factors
2
2
1
1
*
.46
.43
.93
.98
Standards
P
0.010
0.024
0.023
6.22
* - 95% of the monthly averages are expected to be within the
performance standard, p.
A-19
-------
Table A-8d
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Titanium Dioxide
Chloride Process
Plant #H
Parameter
(kg/day)
Chromium
Copper
Zinc
TSS
Historical Summary
Statistics
No
13
13
13
13
Min
.002
.000
.004
2.60
Avg
.013
.027
.028
8.34
Max
.043
.100
.051
24.0
CV
.769
.852
4.29
.695
Var iabili ty Per f ormance
Factors Standards
* P
2.62 .033
2.74 .073
1.80 .051
2.14 17.9
- 95% of the monthly averages are expected to be within the
performance standard, P.
A-20
-------
Table A-8e
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Titanium Dioxide
Chloride Process
Plant fH
Historical Summary
Variability Performance
Parameter
(g/kkg)
Chromium
Copper
Zinc
TSS
Statistics Factors Standards
No
394
394
394
394
Min
.000
.000
.000
5.49
Aver
.178
.37
.384
114
Max CV
2
2
1
.88
.6
.48
2415
*
1
1
1
P
.33
.9
.33
956
* - 99% of the daily maximum measurements expected to be less
than the performance standard, p.
A-21
-------
Table A-8f
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Titanium Dioxide
Chloride Process
Plant fH
Historical Summary
Variability Performance
Parameter
(9/kkg)
Chromium
Copper
Zinc
TSS
Statistics
No
13
13
13
13
Min
.027
.000
.055
35.7
Avg
.178
.37
.384
114
Max CV
.59
1.37
.70
329
Factors Standards
* P
.453
1.0
.70
246
* -* 95% of the monthly averages are expected to be within the
performance standard, P.
A-22
-------
Table A-9a
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
parameter
(mg/1)
Cadmium
Chromium
Iron
(total)
Iron
(diss)
Lead
Nickel
Zinc
TSS
Historical Summary
Statistics
No
26
26
30
153
26
26
26
183
Min
.001
.010
.40
.080
.002
.010
.010
Avg
.009
.021
3.25
.279
.017
.029
.027
35.8
Max
.020
.070
19.1
4.98
.050
.080
.300
CV
.444
.857
1.42
2.01
.765
.690
2.11
1.71
Variability
Factors
2
4
6
8
3
3
9
7
*
.03
.23
.74
.64
.67
.52
.93
.70
= SES = = = 3S=3=S«=ES =
Performance
Standards
P
0.018
0.088
21.9
2,41
0.062
0.102
0.268
276.
* - 99% Of the daily maximum measurements expected to be less
than the performance standard, P.
A-23
-------
Table A-9a-l
Historical Effluent Monitoring Data Summary
with Variability Factors and performance Standards
Daily Measurements
Subcategory Titanium Dioxide
Sulfate Process
Plant *I
April 76 through September 78
Historical Summary
Variability Performance
Parameter
(mg/1)
Cadmium
Chromium
Iron**
Lead
Nickel
Zinc
TSS
Statistics
No
109
128
854
128
128
128
899
Min
,001
,010
.010
.002
.010
.010
.000
Avg
.058
.072
.620
.068
.080
.151
21.2
Max
.100
.400
59.9
.100
.680
1.14
975
CV
.762
.755
5.58
.609
.883
1.35
3.11
Factors
3.
3.
13
3.
4.
6.
11
*
85
81
.5
15
39
41
.0
Standards
P
.224
.275
8.39
.214
.354
.966
233
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
** 04-76 to 08-78
A-24
-------
Table A-9a-2
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
September 78 through February 79
Parameter
(mg/1)
Cadmium
Chromium
Iron
Lead
Nickel
Zinc
TSS **
Historical Summary
Statistics
No
22
22
22
22
22
22
22
Min
.1
.10
212
.1
.1
.1
35.6
Avg
.1
0.27
390
.1
0.15
0.67
257
Max
.1
0.74
605
.1
0.35
1.8
1135
CV
0
0.74
0.31
0
0.47
0.75
1.12
Variability
Factors
*
1
3.73
1.94
1
2.58
3.81
5.46
Performance
Standards
P
.1
1.01
757
.1
0.39
2.55
1403
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
** 09-78 to 01-79
A-25
-------
Table A-9a-3
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
September 79 through April 81
Parameter
(mg/1)
Cadmium
Chromium
Iron
Lead
Nickel
Zinc
TSS**
Historical Summary Variability
Statistics Factors
No
82
81
84
82
82
54
30
Min
0.10
0.05
202
0.10
0.05
0.05
11
Avg
0.10
0.23
330
0.10
0.14
0.50
50
Max CV
0.10 0
0.67 0.87
582 0..22
0.10 0
0.53 0.86
1.18 1.2
398 1.18
1
4
1
1
4
5
5
*
.0
.4
.5
.0
.3
.8
.8
Performance
Standards
0.
1.
495
0.
0.
2.
290
P
10
0
10
60
9
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
** April 81
A-26
-------
Table A-9b-l
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Titanium Dioxide
Sulfate Process
Plant #l
April 76 through September 78
Historical Summary
Variability Performance
Parameter
(kg/day)
Cadmium
Chromium
Iron**
Lead
Nickel
Zinc
TSS
Statistics
No
109
128
854
128
128
128
899
Min
,004
.045
1.00
.008
.047
.049
26 1,
Avg
.432
.526
4.29
.503
.576
1.07
350
Max
.908
2.65
3,854
.908
3.99
55.1
58,820
CV
.782
.707
5.78
.634
.790
1.33
2.79
Factors
*
3.95
3.61
13.6
3.27
3.98
6.32
10.5
Standards
1
1
1
2
6
14,
P
.70
.90
585
.65
.29
.76
120
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
** 04-76 to 08-78
A-27
-------
Table A-9b-2
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
September 78 through February 79
Parameter
(kg/day)
Cadmium
Chromium
Iron
Lead
Nickel
Zinc
TSS**
Historical Summary Variability
Statistics Factors
No
22
22
22 14
22
22
22
22 2
Min
6.95
6.95
,730
6.95
6.95
6.95
,474
Avg Max CV
6.95 6.95 0
18.76 51.41 0.74
27,097 42,035 0.31
6.95 6.95 0
10.42 24.32 0.47
46.55 125.06 0.75
17,856 78,860 1.12
*
1.0
3.73
1.94
1.0
2.56
3.81
5.46
Performance
Standards
P
6.95
69.98
52,568
6.95
26.88
177.36
97,493
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
** 09-78 to 01-79
A-28
-------
Table A-9b-3
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
September 79 through April 81
Parameter
(kg/day)
Cadmium
Chromium
Iron
Lead
Nickel
Zinc
TSS**
Historical Summary Variability
Statistics Factors
No
82
81
84
82
82
54
30
Min
8.
4.
2
1
16,521
8.
4,
4.
2
1
1
900
Avg
8
18
.2
.8
26,990
8
11
40
.2
.4
.9
4,089
Max CV
8.2 0
54.8 0.87
47,599 0.22
8.2 0
43.3 0.86
96.5 1.2
32,551 1.18
*
1.
4.
1.
1.
4.
5.
5.
0
4
5
0
3
8
8
Performance
Standards
8.
81.
P
2
8
40,484
8.
49.
2
0
237
23,718
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
** April 81
A-29
-------
Table A-9c-l
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
April 76 through September 78
Historical Summary
Variability Performance
Parameter
(mg/1)
Cadmium
Chromium
Iron**
Lead
Nickel
Zinc
TSS
Statistics
No
26
30
28
30
30
30
30
Min
.003
.010
.060
.004
.010
.010
1.58
Avg
.058
.072
.620
.068
.080
.151
21.2
Max
.100
.130
3.74
.100
.245
.815
74.8
CV
.722
.524
1.51
.578
.594
1.03
1.03
Factors
2.
2.
4.
2.
4.
3.
3.
*
43
04
00
14
39
05
04
Standards
P
.142
.147
2.47
.147
.354
.406
64.3
* - 95% of the monthly averages are expected to be within the
performance standard, P.
** 04-76 to 08-78
A-30
-------
Table A-9c-2
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
September 78 through February 79
Parameter
(mg/1)
Cadmium
Chromium
Iron
Lead
Nickel
Zinc
TSS**
Historical Summary Variability
Statistics Factors
No
5
5
5
5
5
5
5
Min
.1
.10
244.4
.1
.1
0.30
51.6
Avg
.1
0.27
390
.1
0.15
0.67
257
Max
.1
0.48
502.9
.1
0.25
0.83
595
CV
0
0.48
0.25
0
0.37
0.64
0.750
*
1
1.79
1.41
1
1.61
2.05
2.23
Performance
Standards
P
1
0.48
550
.1
0.24
1,37
573
* - 95% of the monthly averages are expected to be within the
performance standard, P.
** 09-78 to 01-79
A-31
-------
Table A-9c-3
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
September 79 through April 81
Parameter
(mg/1)
Cadmium
Chromium
Iron
Lead
Nickel
Zinc
TSS**
Historical Summary Variability
Statistics Factors
No
20
20
20
20
20
13
20
Min
0.10
0.05
261
0.10
0.05
0.12
48
Avg
0.10
0.23
330
0.10
0.14
0.50
50
Max CV
0.10 0
0.47 0.52
418 0.13
0.10 0
0.41 0.67
0.78 0.45
214 0.35
*
1.
1.
1.
1.
2.
1.
1.
0
9
2
0
1
7
6
Performance
Standards
0.
0.
396
0.
0.
0.
80
P
10
44
10
29
85
* - 95% of the monthly averages are expected to be within the
performance standard, P.
** 09-79 to 04-81
A-32
-------
Table A-9d-l
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
April 76 through September 78
Historical Summary
Variability Performance
Parameter
(kg/day)
Cadmium
Chromium
Iron**
Lead
Nickel
Zinc
TSS
Statistics
No
26
30
Min
.016
.064
28 4.00
30
30
30
30
.021
,065
.074
116 1,
Avg
.432
.526
42.9
.503
.576
1.07
350 4
Max
.780
.862
294
.852
1.49
5.62
,797
CV
.743
.549
1.59
.602
.569
.996
1.01
Factors
2
2
2
2
2
2
*
.47
.09
4.14
.19
.13
.97
.99
Standards
1.07
1.10
P
177
1.10
1.23
3.18
4,041
* - 95% of the monthly averages are expected to be within the
performance standard, P.
** 04-76 to 08-78
A-33
-------
Table A-9d-2
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
September 78 through February 79
Parameter
(kg/day)
Cadmium
Chromium
Iron
Lead
Nickel
Zinc
TSS**
Historical Summary Variability
Statistics Factors
No
5
5
5
5
5
5
5
Min
6.95
6.95
16,981
6.95
6.95
2.4
3,585
Avg
6.95
18.76
27,097
6.95
10.42
46.55
17,856
Max CV
6.95 0
33.4 0.48
34,941 0.25
6.95 0
17.4 0.37
57.7 0.640
41,340 0.750
*
1
1.79
1.41
1
1.61
2.05
2.23
Performance
Standards
P
6.95
33.4
38,214
6.95
16.68
95.19
39,812
* - 95% of the monthly averages are expected to be within the
performance standard, P.
** 09-78 to 01-79
A-34
-------
Table A-9d-3
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
September 79 through April 81
Parameter
(kg/day)
Cadmium
Chromium
Iron
Lead
Nickel
Zinc
TSS**
Historical Summary Variability
Statistics Factors
No
20
20
20
20
20
13
20
Min
8.
4.
2
1
21,350
8.
4.
9.
2
1
8
3,926
Avg
8.2
18.8
26,990
8.2
11.4
40.9
4,089
Max CV
8.
38.
34,
8.
33.
63.
17,
2 0
4 0.52
186 0.13
2 0
5 0.67
8 0.45
502 0.35
*
1.
1.
1.
1.
2.
1.
1.
0
9
2
0
1
7
6
Performance
Standards
P
8.
36.
2
0
32,387
8.
23.
69.
2
7
5
6,543
* - 95% of the monthly averages are expected to be within the
performance standard, P.
** 09-79 to 04-81
A-35
-------
Table A-9e-l
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
April 76 through September 78
Historical Summary
Variability Performance
Parameter
(g/kkg)
Cadmium
Chromium
Iron**
Lead
Nickel
Zinc
TSS
Statistics
No
109
128
854
128
128
128
899
Min
.041
.464
10.3
.082
.484
.505
.268
Avg
4.
5.
44
5.
5.
11
13
45
42
.2
18
93
.02
.91
Factors Standards
Max CV
9.
27
39
9.
41
36
.3
,711
36
.1
568
606
*
17
19
6,0
17
23
69
145
P
.5
.6
28
.0
.6
.7
.5 (kg/kkg)
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
** 04-76 to 08-78
A-36
-------
Table A-9e-2
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
September 78 through February 79
Parameter
(g/kkg)
Cadmium
Chromium
Iron
Lead
Nickel
Zinc
TSS**
Historical Summary Variability
Statistics Factors
No
22
22
22
22
22
22
22
Min
0.67
0.67
152
0.67
0.67
0.67
25.48
Avg
0.67
1.82
279
0.67
1.01
4.51
183.9
Max CV *
0.67
4.99
433
0.67
2.36
12.13
812
Performance
Standards
P
0.67
6.79
542
0.67
2.6
17.2
1004
(kg/kkg)
(kg/kkg)
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
** 09-78 to 01-79
A-37
-------
Table A-9e-3
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
September 79 through April 81
Parameter
(g/kkg)
Cadmium
Chromium
Iron
Lead
Nickel
Zinc
TSS**
Historical Summary Variability Performance
Statistics Factors Standards
No
82
81
84
82
82
54
30
Min
0,80
0.40
170
0.80
0.40
0.40
9.3
Avg
0.80
1.82
278
0.80
1.11
3.97
42.1
Max CV
0.80
5.32
490
0.80
4.20
9.36
335
* P
0.80
7.93
417
0.80
4.75
23.0
244
(kg/kkg)
(kg/kkg)
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
** April 81
A-38
-------
Table A-9f-l
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
April 76 through September 78
Parameter
(g/kkg)
Cadmium
Chromium
Iron**
Lead
Nickel
Zinc
TSS
Historical Summary Variability Performance
Statistics Factors Standards
No
26
30
28
30
30
30
30
Min Avg
.165
.66
41.
.216
.67
.762
1.2
4.45
5.42
2 442
5.18
5.94
11.0
13.9
Max CV
8.
8.
04
88
3209
8.
15
57
49
78
.4
.9
.4
* P
11.0
11.3
1,824
11.3
12.7
32.8
41.6 (kg/kkg)
* - 95% of the monthly averages are expected to be within the
performance standard, P.
** 04-76 to 08-78
A-39
-------
Table A-9f-2
Historical Effluent Monitoring Data Summary
with1Variability Factors and Performance Standards
30 Day Averages
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
September 78 through February 79
Parameter
(g/kkg)
Cadmium
Chromium
Iron
Lead
Nickel
Zinc
TSS**
Historical Summary Variability Performance
Statistics Factors Standards
No
5
5
5
5
5
5
5
Min
0.
0.
174
0.
0.
0.
36
67
67
.9
67
67
23
.9
Avg
0.67
1.81
279
0.67
1.01
4.5
184
Max CV
0.67
3.23
360
0.67
1.69
5.6
426
* P
0.67
3.2
394
0.67
0.17
0.98
410
(kg/kkg)
(kg/kkg)
* - 95% of the monthly averages are expected to be within the
performance standard, P.
** 09-78 to 01-79
A-40
-------
Table A-9f-3
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Titanium Dioxide
Sulfate Process
Plant #1
September 79 through April 81
Historical Summary
Variability Performance
Parameter
(g/kkg)
Cadmium
Chromium
Iron
Lead
Nickel
Zinc
TSS**
Statistics
No
20
20
20
20
20
13
20
0
0
0
0
0
40
Min
.80
.40
220
.80
.40
.95
.44
Avg
0
1
0
1
3
42
.80
.82
278
.80
.11
.97
.11
Factors Standards
Max CV
0.
3.
80
72
352
0.
3.
6.
180
80
25
20
.3
*
0.
3.
334
0.
2.
6.
67.
P
80
49
80
30
74
4
(kg/kkg)
(kg/kkg)
* - 95% of the monthly averages are expected to be within the
performance standard, P.
** 09-79 to 04-81
A-41
-------
Table A-lOa
Historical Effluent Monitoring Data Summary
with Variability Factors and performance Standards
Daily Measurements
Subcategory Aluminum Fluoride
Plant fJ
Parameter
(mg/1)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Lead
152 0.11 2.28 12.8 .601
3.12
7.11
99% of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-lOb
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Aluminum Fluoride
Plant fJ
Parameter
(kg/day)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Lead
152 0.09 2.15 15.3 .753
3.82
8.20
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
A-42
-------
Table A-lOc
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Aluminum Floride
Plant fJ
Historical Summary Variability Performance
Parameter Statistics Factors Standards
(mg/1) No Min Avg Max CV * P
Lead 10 1.51 2.28 3.90 .601 1.55 3.54
* - 95% of the monthly averages are expected to be within the
performance standard, P.
Table A-lOd
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Aluminum Floride
Plant f J
Historical Summary Variability Performance
Parameter Statistics Factors Standards
(kg/day) No Min Avg Max CV * P
Lead 10 1.51 2.15 3.70 .326 1.54 3.30
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-43
-------
Table A-lOe
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Aluminum Fluoride
Plant #J
Parameter
Historical Summary
Statistics
Variability Performance
Factors Standards
No Min Avg Max CV
Lead
152 .89 21.2 150
80.7
* -
99% of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-lOf
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Aluminum Floride
Plant fJ
Parameter
(g/kkg)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Lead
10 14.9 21.2 36.4
32.5
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-44
-------
Table A-lla
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Chrome Pigments Subcategory
Plant #K
Historical Summary
Variability Performance
Parameter
(mg/1)
Arsenic
Cadmium
Chromium
(hexavalent)
Chromium
(Total)
Copper
Lead
Mercury
Zinc
Cyanide
(Available)
Cyanide
(total)
TSS
Statistics
No
23
23
23
23
23
23
23
23
23
23
23
Min
.0096
.050
.028
.197
.038
.217
.0004
.012
.0003
.025
0.27
Factors
Avg Max CV
.079
.079
.112
.442
.134
.412
.001
.04
.019
.118
11.2
.235
.164
.592
.799
.296
1.635
.0018
.087
.076
.316
33.3
.668
.339
1.04
.404
.529
.681
.400
.437
1.57
.995
.662
2
1
2
1
1
2
1
1
3
2
2
*
.02
.56
.70
.66
.87
.12
.66
.72
.58
.63
.01
Standards
P
.156
.123
.302
.733
.250
.873
.0016
.074
.068
.310
22.5
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-45
-------
Table A-12a
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Hydrogen Cyanide
Andrussow Process
Plant #L
Parameter
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Ammonia
35 14. 113. 188. .335
2.02
229
99% Of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-12b
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Hydrogen Cyanide
Andrussow Process
Plant £L
Parameter
(kg/day)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Ammonia
35 112 1533 2419 .365
2.14
3283
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
A-46
-------
Table A-12c
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Hydrogen Cyanide
Andrussow Process
Plant #L
Parameter
(mg/1)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Ammonia
8 80. 113. 134. .335
1.32
150
95% Of the monthly averages are expected to be within the
performance standard, P.
Table A-12d
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Hydrogen Cyanide
Andrussow Process
Plant fL
Parameter
(kg/day)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Ammonia
8 908 1533 1941 .212
1.42
2177
* - 95% of the monthly averages are expected to he within the
performance standard, P.
A-47
-------
Table A-12e
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Hydrogen Cyanide
Andrussow Process
Plant £L
Parameter
(kg/kkg)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Ammonia
35 .606 8.29 1309
17.8
* _
99% of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-12f
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Hydrogen Cyanide
Andrussow Process
Plant fL
Parameter
(kg/kkg)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Ammonia
8 46.9 8.29 10.5
11.8
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-48
-------
Table A-13a
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory
Andrussow Process
Plant £M
Historical Summary
Variability Performance
Parameter
(mg/1)
Cyanide
(Free)
Cyanide
(Total)
Anunon i a
COD
TOC
SS
Statistics
No
534
25
26
25
26
22
Min
.01
.039
.193
2.71
.783
5
Avg
.202
.192
3.63
15.9
8.30
35
Max
3.27
.460
10.2
45.2
25.6
267
CV
1.58
.667
.636
.552
.845
1.57
Factors
7
3
3
2
4
8
*
.26
.42
.51
.90
.22
.16
Standards
P
1.46
.65
12.7
46.1
35.0
286
* - 99% of the daily maximum measurements expected to be less
than the performance standard, P.
A-49
-------
Table A-13c
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Hydrogen Cyanide
Andrussow Process
Plant fM
parameter
(mg/1)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Cyanide
(Free)
19 .082 .202 .351 .391-
1.78
0.359
* - 95% of the monthly averaged are expected to be within the
performance standard, P.
A-50
-------
Table A-13e
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Hydrogen Cyanide Subcategory
Andrussow Process
Plant ffM
Parameter
(g/kkg)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Cyanide
19 .457 1.12 1.95
1.99
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-51
-------
Table A-15a
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Nickel Sulfate
Plant #0
Parameter
(mg/1)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Nickel
88 .080 1.83 8.33 1.21
5.84
10.7
- 99% of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-15b
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance standards
Daily Measurements
Subcategory Nickel Sulfate
Plant $0
Parameter
(kg/day)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Nickel
88 1.02 8.32 44.6 1.31
6.24
51.9
* - 99% of the daily maximum measurements expected to be less
than the performance standard, p.
A-52
-------
Table A-15c
Historical Effluent Monitoring Data Summary
with Variability Factors and performance Standards
30 Day Averages
Subcategory Nickel Sulfate
Plant £0
Parameter
(mg/1)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Nickel
3 1.29 1.83 2.48 1.21
1.54
2.82
* - 95% of the monthly averages are expected to be within the
performance standard, P.
Table A-15d
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Nickel Sulfate
Plant fO
Parameter
(kg/day)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Nickel
3 5.04 8.32 11.1 .302
1.49
12.4
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-53
-------
Table A-15e
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
Daily Measurements
Subcategory Nickel Sulfate
Plant fO
Parameter
(g/kkg)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Nickel
88 112 912 4890
5,691
99% of the daily maximum measurements expected to be less
than the performance standard, P.
Table A-15f
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Nickel Sulfate
Plant #0
Parameter
(9/kkg)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
Nickel
3 553 912 1217
1,360
* - 95% of the monthly averages are expected to be within the
performance standard, P.
A-54
-------
Table A-16a
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Sodium Hydrosulfite
Plant #P
Parameter
(kg/day)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
TSS
36 .91 3.78 41.1 1.69
3.77
14.2
95% of the monthly averages are expected to be within the
performance standard, P.
Table A-16e
Historical Effluent Monitoring Data Summary
with Variability Factors and Performance Standards
30 Day Averages
Subcategory Sodium Hydrosulfite
Plant fp
Parameter
(g/kkg)
Historical Summary
Statistics
Variability
Factors
Performance
Standards
No Min Avg Max CV
TSS
36 16.3 67.5 734
254
* - 95% of the monthly averages are expected to be within the
performance standard, P.
*U.S. GOVERNMENT PRINTING OFFICE: 1982-0-361-085/^62
A-55
-------
-------
-------
------- |