vvEPA
           United States
           Environmental Protection
           Agency
           Effluent Guidelines Division
           WH-552
           Washington, DC 20460
EPA 440/1-79/007
           Water and Waste Management
Development
Document for
Effluent Limitations
Guidelines and
Standards for the

Inorganic Chemicals
Manufacturing
Proposed
                            *
           Point Source Category

-------

-------
TABLE OF CONTENTS

LIST
LIST

OF FIGURES
OF TABLES
ACKNOWLEDGEMENTS
1.0 CONCLUSIONS AND SUMMARY
l.L
1.2
1.3
1.4
1.5
TOXIC POLLUTANTS
CONTROL AND TREATMENT TECHNOLOGY
COSTS OF ADDITIONAL IN-PLANT TREATMENT
SUB CATEGORIZATION
RESTUDY OF REMANDED REGULATIONS
2 . 0 RECOMMENDATIONS
3.0 INTRODUCTION
3.1


3.2
3.3








AUTHORITY
3.1.1 The Federal Water Pollution
Control Act Amendments
3.1.2 Court Remand Regulations
3.1.3 The Settlement Agreement
GENERAL APPROACH AND METHODOLOGY
3.2.1 Industry Data Based Development
and Subcategorization Review
3.2.2 The Screening and Verification
Sampling Programs
3.2.3 Engineering Evaluations
3.2.4 Treatment System Cost Estimates
GENERAL CRITERIA FOR EFFLUENT LIMITATIONS
3.3.1 BPT Effluent Limitations
3.3.2 BAT Effluent Limitations
3.3.3 BCT Effluent Limitations
3.3.4 New Source Performance
Standards
3.3.5 Pretreatment Standards for
Existing Sources
3.3.6 Pretreatment Standards for
New Sources
Page
xxiii
xxxiii
li
1
1
2
2
2
3
5
23
23
23

26
28
36
37
37
37
38
38
38
39
39
42

42
42

         111

-------
                 TABLE OF CONTENTS - Continued
4.0   SUBCATEGORIZATION REVIEW                               45

      4.1     BASIS FOR SUBCATEGORIZATION                    45
              4.1.1     Factors Considered                   45
              4.1.2     General Conclusions                  49
      4.2     SECONDARY SUBCATEGQRIZATION                    49
              4.2.1     Chlor-Alkali                         49
              4.2.2     Titanium Dioxide                     49
              4.2.3     Hydrogen Cyanide                     51
      4.3     REVIEW OF POSSIBLE INTEGRATION OF              51
               SUBCATEGORIES
              4.^.1     Hydrofluoric Acid and Aluminum       51
                         Fluoride
      4.4     SUMMARY                                        52

5.0   SCREENING AND VERIFICATION SAMPLING PROGRAMS           53

      5.1     SCOPE AND METHODOLOGY                          53
              5.1.1     Selecting Plants and Making          54
                         Preliminary Contacts
              5.1.2     Screening and Verification           55
                         Sampling
              5.1.3     Analytical Methodology for           57
                         Toxic Pollutants
              5.1.4     Quality Assurance Provisions         64
      5.2     SUMMARY OF ANALYTICAL RESULTS                  65

6.0   PROCESS AND WASTE TREATMMENT INFORMATION               71
      DEVELOPMENT AND EVALUATION

      6.1     INDUSTRY DATA      DESCRIPTION                 71
              6.1.1     Literature Review               '     71
              6.1.2     Plant Visits                         71
              6.1.3     Telephone and Direct Contact         72
              6.1.4     308 Questionnaire Responses          72
      6.2     PROCESS WASTE SOURCES AND CURRENT              74
              TREATMENT PRACTICES
              6.2.1     Data Acquisition                     74
              6.2.2     Evaluation of Data                   74
              6.2.3     Model Plant and BPT Treatment        75
                         System Specification
              6.2.4     Dissolved Solids in Waste Water      76
                         Effluents

-------
                 TABLE OF CONTENTS - Continued
                                                           Page

7.0   ASSESSMENT OF TECHNOLOGY FOR ADVANCED TREATMENT        79
      AND CONTROL
      7.1     INTRODUCTION                                   79
      7.2     HYDROXIDE PRECIPITATION                        80
      7.3     FERRITE COPRECIPITATION                        85
      7.4     SULFIDE PRECIPITATION                          85
      7.5     THE XANTHATE PROCESS                .           87
      7.6     ION EXCHANGE                                   89
      7.7     REDUCTION PROCESSES                            91
      7.8     OXIDATION PROCESSES                            93
      7.9     MEMBRANE PROCESSES                             96
      7.10    ADSORPTION                                     98
      7.11    FLUORIDE REMOVAL                              101
      7.12    CHLORINE REMOVAL                              102

8.0   TREATABILITY ESTIMATES AND LONG-T1RM DATA             103
       ANALYSIS

      8.1     THE DEVELOPMENT OF TREATABILITY ESTIMATES     103
      8.2     THE USE OF HISTORICAL POLLUTANT DATA          117
              8.2.1     Determination of Limitation         117
                         Guidelines Based Upon
                         Historical Performance
              8.2.2     Assumptions Concerning Daily        118
                         Pollutant Level Measurements
              8.2.3     Assumptions Concerning 30-Day       123
                         Average Pollutant Level
                         Observation

9.0   TREATMENT TECHNOLOGY APPLICATIONS FOR TOXIC           131
       POLLUTANT REMOVAL

      9.1     SELECTION OF POLLUTANTS TO BE CONTROLLED      131
      9.2     APPLICATION OF ADVANCE LEVEL TREATMENT        131
              AND CONTROL ALTERNATIVES
              9.2.1     General Design Objectives           131
              9.2.2     Pretreatment Technology             134
              9.2.3     New Source Performance              134
                         Standards
      9.3     ESTIMATED ACHIEVABLE PERFORMANCE          '    134
              CHARACTERISTICS FOR ADVANCED LEVEL
              APPLICATIONS
              9.3.1     Advanced Level Removal of BPT       135
                         Pollutants
              9.3.2     Advanced Level Removal of Toxic     135
                         Pollutants
                               v

-------
                 TABLE OP CONTENTS - Continued
      9.4     POLLUTION CONTROL PARAMETERS TO BE
              REGULATED
              9.4.1     Conventional Pollutants
              9.4.2     Nonconventional Pollutants
              9.4.3     Toxic Pollutants

10.0  COST OP TREATMENT AND CONTROL SYSTEMS                 139

      10.1    INTRODUCTION                                  139
              10.1.1    Purpose of Cost Data                139
              10.1.2    General Approach                    140
              10.1.3    Cost References and Rationale       340
              10.1.4    Definition of Levels of             141
                         Treatment and Control Cost
                         Development
              10.1.5    Treatment and Disposal              141
                         Rationale Applied to Cost
                         Development
              10.1.6    Expression of Costs                 142
      10.2    COST ESTIMATES FOR EACH SUBCATEGORY           149

11.0  CHLOR-ALKALI INDUSTRY                                 351

      11.1    MERCURY CELL PROCESS INDUSTRY PROFILE         151
              11.1.1    General Description                 151
              11.1.2    General Process Description and     151
                         Raw Materials
      13.2    WATER USE AND WASTE WATER SOURCE              155
               CHARACTERISTICS
              11.2.1    Water Use                           155
              11.2.2    Waste Sources                       155
      11.3    DESCRIPTION OF SPECIFIC PLANTS                158
              11.3.1    Screening Program                   158
              11.3.2    Verification                        160
              11.3.3    Descriptions of Plants Not          166
                         Sampled
              11.3.4    Summary of the Toxic Pollutant      369
                         Data
      11.4    POLLUTION ABATEMENT OPTIONS                   174
              11.4.1    Toxic Pollutants of Concern         174
              11.4.2    Prevailing Control and              174
                         Treatment Practices
              11.4.3    Process Modifications and           174
                         Technology Transfer Options
              11.4.4    Best Management Practices           176
              11.4.5    Advanced Treatment Technologies     176

-------
           TABLE OF CONTENTS - Continued
11.5    SELECTION OF APPROPRIATE TECHNOLOGY AND
        EQUIPMENT
        11.5.1    Technologies for Different
                   Treatment Levels
        11.5.2    Equipment for Different             177
                   Treatment Levels
11.6    TREATMENT COST ESTIMATES                      180
        11.6.1    General Discussion                  180
        11.6.2    Chlorine Bearing Wastes             183
        11.6.3    Model Plant Treatment Costs         183
11.7    BASIS FOR REGULATIONS                         182
        11.7.1    Basis for BPT Limitations           382
        11.7.2    Basis for Proposed BAT Effluent     3.94
                   Limitations
        11.7.3    Basis for Proposed BCT Effluent     205
                   Limitations
        11.7.4    Basis for New Source                205
                   Performance Standards
        11.7.5    Basis for Proposed Pretreatment     205
                   Standards
11.8    DIAPHRAGM CELL PROCESS INDUSTRY PROFILE       206
        11.8.1    General Description                 206
        11.8.2    General Process Description         206
11.9    WATER USE AND WASTE WATER SOURCES             209
        11.9.1    Water Use              "             209
        11.9.2    Waste Sources                       211
11.10   DESCRIPTIONS OF SPECIFIC PLANTS               214
        11.10.1   Screening                           214
        11.10.2   Verification                        214
        11.10.3   Descriptions of Plants Not          221
                   Sampled
        11.30.4   Toxic Pollutant Concentrations      224
11.11   POLLUTION ABATEMENT OPTIONS                   231
        11.11.1   Toxic Pollutants of Concern         231
        11.11.2   Prevailing Control and              236
                   Treatment Practices
        11.13.3   Process Modifications and           237
                   Technology Transfer Options
        11.11.4   Best Management Practices           238
        11.11.5   Advanced Treatment Technologies     239
11.12   SELECTION OF APPROPRIATE TECHNOLOGY AND       239
        EQUIPMENT
        11.12.1   Technologies for Different          239
                   Treatment Levels
        11.12.2   Equipment for Different             244
                   Treatment Levels
                        Vll

-------
                 TABLE OF CONTENTS - Continued
      11.13   TREATMENT COST ESTIMATES
              11.13.1   General Discussion
              11.13.2   Model Plant Treatment Costs
      11.14   BASIS FOR REGULATIONS
              11.14.1   Basis for BPT Limitations
              11.14.2   Basis for BAT Effluent
                         Limitations
              11.14.3   Basis for BCT Limitations           268
              11.14.4   Basis for New Source                270
                         Performance Standards
              11.14.5   Basis for Pretreatment              273
                         Standards

12.0  HYDROFLUORIC ACID INDUSTRY                            275

      12.1    INDUSTRY PROFILE                              275
              12.1.1    General Description                 275
              12.1.2    General Process Description and     275
                         Raw Materials
      12.2    WATER USE AND WASTE SOURCE                    279
              CHARACTERISTICS
              12.2.1    Water Use                           279
              12.2.2    Waste Sources                       279
      12.3    DESCRIPTION OF PLANTS VISITED AND SAMPLED     285
              12.3.1    Screening                           285
              12.3.2    Verification                        291
              12.3.3    Summary of the Toxic Pollutant      291
                         Data
      12.4    POLLUTION ABATEMENT OPTIONS                   295
              12.4.1    Toxic Pollutants of Concern         295
              12.4.2    Process Modifications and           298
                         Technology Transfer Options
              12.4.3    Best Management Practices           299
              12.4.4    Prevailing Control and              299
                         Treatment Practices
              12.4.5    Advanced Treatment Technologies     300
      12.5    SELECTION OF APPROPRIATE TECHNOLOGY AND       30J
              EQUIPMENT
              12.5.1    Technologies for Different          301
                         Treatment Levels
              12.5.2    Equipment for Different             306
                         Treatment Levels
      12.6    TREATMENT COST ESTIMATES                      308
              12.6.1    General Discussion                  308
              12.6.2    Model Plant Control Costs for       313
                         Existing Sources
                             Vlll

-------
                 TABLE OF CONTENTS - Continued
              12.6.3    Model Plant Control Costs for       317
                         New Sources
      12.7    BASIS FOR REGULATIONS                         324
              12.7.1    Evaluation of BPT Treatment         324
                         Practices
              12.7.2    Basis for Proposed BPT Effluent     329
                         Limitations
              32.7.3    Basis for Proposed BCT Effluent     340
                         Limitations
              12,7.4    Basis for Proposed BAT Effluent     341
                         Limitations
              12.7.5    Basis for Proposed New Source       349
                         Performance Standards
              12.7.6    Basis for Proposed Pretreatment     354
                         Standards

13.0  HYDROGEN PEROXIDE INDUSTRY                            357

      13.1    SUMMARY OF DETERMINATIONS                     357
      13.2    ASSESSMENT OF THE WATER POLLUTATION           357
               POTENTIAL
              13.2.1    Production Processes and            357
                         Effluents
              13.2.2    Plants                              358
              13.2.3    Toxic Pollutants                    358
      13.3    STATUS OF REGULATIONS                         358

J4.0  TITANIUM DIOXIDE INDUSTRY                             361
      (RUTILE/UPGRADED ILMENITE—CHLORIDE PROCESS)

      14.1    INDUSTRY PROFILE                              361
              14.1.1    General Description                 361
              14.1.2    General Process Description and     361
                         Raw Materials
      14.2    WATER USE AND WASTE SOURCE                    365
               CHARACTERISTICS
              14.2.1    Water Use                           365
              14.2.2    Waste Sources                       365
      14.3    DESCRIPTION OF PLANTS VISITED AND SAMPLED     368
              14.3.1    Screening                           368
              14.3.2    Verification                        371
              14.3.3    Toxic Pollutant Concentrations      371
      14.4    POLLUTION ABATEMENT OPTIONS                   375
              14.4.1    Toxic Pollutants of Concern         375
              14.4.2    Process Modification and            377
                         Technology Transfer Options

-------
           TABLE OF CONTENTS - Continued
        14.4.3    Best Management Practices           377
        14.4.4    Prevailing Control and              377
                   Treatment Practices
        14.4,5    Advanced Treatment Technologies     378
14.5    SELECTION OP APPROPRIATE TECHNOLOGY AND       379
        EQUIPMENT
        14.5.1    Technologies for Different          379
                   Treatment Levels
        14.5.2    Equipment for Different             379
                   Treatment Levels
14.6    TREATMENT COST ESTIMATES                      383
        14.6.1    General Discussion                  383
14.7    BASIS FOR REGULATIONS                         385
        14.7.1    Evaluation of BPT Treatment         385
                   Practices
        14.7.2    Basis for Proposed BPT Effluent     392
                   Limitations
        14.7.3    Basis for Proposed BCT Effluent     400
                   Limitations
        14.7.4    Basis for Proposed BAT Effluent     400
                   Limitations
        14.7.5    Basis for Proposed New Source       403
                   Performance Standards
        14.7.6    Basis for Pretreatment              406
                   Standards
14.8    TITANIUM DIOXIDE - SULFATE PROCESS            408
        INDUSTRY PROFILE
        14.8.1    General Description                 408
        14.8.2    General Process Description and     408
                   Raw Materials
14.9    WATER USE AND WASTE SOURCE                    411
         CHARACTERISTICS
        14.9.1    Water Use                           411
        14.9.2    Waste Sources                       411
14.10   DESCRIPTION OF PLANTS                         415
        14.10.1   Screening                           415
        14.10.2   Verification                        415
        14.10.3   Other Plant Descriptions            417
        14.10.4   Toxic Pollutant Concentrations      420
14.11   POLLUTION ABATEMENT OPTIONS                   425
        14.11.1   Toxic Pollutants of Concern         425
        14.11.2   Process Modifications and           426
                   Technology Transfer Options
        14.11.3   Best Management Practices           426
        14.11.4   Prevailing Control and              426
                   Treatment Practices

-------
           TABLE OF CONTENTS - Continued
        14.11.5   Advanced Treatment Technologies     426
14.12   SELECTION OP APPROPRIATE TECHNOLOGY AND       427
         EQUIPMENT
        14.12.1   Technologies for Different          427
                   Treatment Levels
        14.12.2   Equipment for Different             430
                   Treatment Levels
14.13   TREATMENT COST ESTIMATES                      431
        14.13.1   General Discussion                  431
        14.13.2   Model Plant Control and             432
                   Treatment Costs
14.14   BASIS FOR REGULATIONS                         433
        14.14.1   Evaluation of BPT Practices    -     433
        14.14.2   Basis for Proposed BPT Effluent     444
                   Limitations Technology Basis
        14.14,3   Basis for Proposed BCT Effluent     452
                   Limitations
        14.14.4   Basis for Proposed BAT Effluent     452
                   Limitations
        14.14.5   Basis for Proposed N'ew Source       452
                   Performance Standards
        14.14.6   Basis for Proposed Pretreatment     452
                   Standards
14.15   TITANIUM DIOXIDE - CHLORIDE ILMENITE          455
         PROCESS INDUSTRY PROFILE
        14.15.1   General Description                 455
        14.15.2   General Process Description and     455
                   Raw Materials
14.16   WATER USE AND WASTE SOURCE                    458
         CHARACTERISTICS
        14.16.1   Water Use                           458
        14.16.2   Waste Sources                       458
14.17   DESCRIPTION OF PLANTS VTSTED AND SAMPLED      461
        14.17.1   Screening                           461
        14.17.2   Verification Program                462
        14.17.3   Toxic Pollutant Concentration       462
14.18   POLLUTION ABATEMENT OPTIONS                   466
        14.18.1   Toxic Pollutants of Concern         466
        14.18.2   Process Modifications and           466
                   Technology Transfer Options
        14.18.3   Best Management Practices           466
        14.18.4   Prevailing Control and              467
                   Treatment Practices
        14.18.5   Advanced Treatment Technology       467
14.19   SELECTION OF APPROPRIATE TECHNOLOGY AND       468
         EQUIPMENT

-------
                 TABLE OF CONTENTS - Continued
              14.19.1   Technologies for Different
                         Treatment Levels
              14.19.2   Equipment for Different
                         Treatment Levels
      14.20   TREATMENT COST ESTIMATES
              14.20.1   General Discussion
              14.20.2   Model Plant Control and
                         Treatment Costs
      14.21   BASIS FOR REGULATIONS                         479
              14.21.1   Evaluation of BPT Treatment         479
                         Practices
              14.21.2   Basis for Proposed BPT Effluent     479
                         Limitation
              14.21.3   Basis for Proposed BCT Effluent     485
                         Limitations
              14.21.4   Basis for Proposed BAT Effluent     485
                         Limitations
              14.21.5   Basis for the Proposed New          485
                         Source Performance Standards
              14.21.6   Basis for Proposed Pretreatment     492
                         Standards

15.0  ALUMINUM FLUORIDE INDUSTRY                 .           493

      15.1    INDUSTRY PROFILE                              493
              15.1.1    General Description                 493
              15.1.2    General Process Description and     493
                         Raw Materials
      15.2    WATER USE AND WASTE SOURCE                    493
               CHARACTERISTICS
              15.2.1    Water Use                           493
              15.2.2    Waste Sources                       497
      15.3    DESCRIPTION OF PLANTS VISITED AND SAMPLED     500
              15.3.1    Screening                           500
              15.3.2    Verification                        500
              15.3.3    Summary of the Toxic Pollutant      500
                         Data
      15.4    POLLUTION ABATEMENT OPTIONS                   510
              15.4.1    Toxic Pollutants of Concern         510
              15.4.2    Process Modifications and           510
                         Technology Transfer Options
              15.4.3    Best Management Practices           510
              15.4.4    Prevailing Control and              511
                         Treatment Practices
              15.4.5    Advanced Treatment Technologies     511
                              Xll

-------
                 TABUE OF CONTENTS - Continued
                                                           Page

      15.5    SELECTION OF APPROPRIATE TECHNOLOGY AND       512
               EQUIPMENT
              15.5.1    Technologies for Different          512
                         Treatment Levels
              15.5.2    Equipment for Different             512
                         Treatment Levels
      15.6    TREATMENT COST ESTIMATES                      518
              15.6.1    General Discussion                  518
      15o7    BASIS FOR REGULATIONS                         536
              15.7.1    Evaluation of BPT Treatment         536
                         Practices
              15.7.2    BPT Effluent Limitations            536
              15.7.3    Basis for Proposed BCT Effluent     542
                         Limitations
              15.7.4    Basis for Proposed BAT Effluent     542
                         Limitations
              15.7.5    Basis for Proposed New Source       548
                         Performance Standards
              15.7.6    Basis for Proposed Pretreatment     549
                         Standards

16.0  CHROME PIGMENTS INDUSTRY                              55.1

      16.1    INDUSTRY PROFILE                              551
              16.1.1    General Description                 551
              16.1.2    General Process Description and     551
                         Raw, Materials
      16.2    WATER USE AND WASTE SOURCE                    560
               CHARACTERISTICS
              16.2.1    Water Use                           560
              16.2.2    Waste Sources                       560
      16.3    DESCRIPTION OF PLANTS                         564
              16.3.1    Screening                           564
              16.3.2    Verification                        567
              16.3.3    Toxic Pollutant Concentrations      570
      16.4    POLLUTION ABATEMENT OPTIONS                   577
              16.4.1    Toxic Pollutants of Concern         577
              16.4.2    Process Modifications and           578
                         Technology Transfer Options
              16.4.3    Best Management Practices           579
              16.4.4    Prevailing Control and              579
                         Treatment Practices
              16.4.5    Advanced Treatment Technologies     580
      16.5    SELECTION OF APPROPRIATE TECHNOLOGY AND       581
               EQUIPMENT
                              Xlll

-------
                 TABUS OF CONTENTS - Continued
              16.5.1    Technologies for Different          581
                         Treatment Levels
              16.5.2    Equipment for Different             581
                         Treatment Levels
      16.6    TREATMENT COST ESTIMATES                      584
              16.6.1    General Discussion                  584
              16.6.2    Model Plant Costs                   592
      16.7    BASIS FOR REGULATIONS                         592
              16.7,1    Evaluation of BPT Treatment         592
                         Practices
              16.7.2    Basis for Proposed BPT Effluent     595
                         Limitations
              16.7.3    Basis for Proposed BCT              602
                         Limitations
              16.7.4    Basis for Proposed BAT Effluent     602
                         Limitations
              16.7.5    Basis for Proposed New Source       604
                         Performance Standards
              16.7.6    Basis for Proposed Pretreatment     609
                         Standards

17.0  HYDROGEN CYANIDE INDUSTRY                             611

      17.1    INDUSTRY PROFILE        '                      611
              17.1.1    General Description                 611
              17.1.2    General Process Description and     611
                         Raw Materials
      17.2    WATER USE AND WASTE SOURCE                    614
               CHARACTERISTICS
              17.2.1    Water Use                           614
              17.2.2    Waste Source                        614
      17.3    DESCRIPTION OF PLANTS VISITED AND SAMPLED     637
              17.3.1    Screening                           617
              17.3.2    Verification                        619
              17.3.3    Toxic Pollutant Concentrations      623
      17.4    POLLUTION ABATEMENT OPTIONS                   628
              17.4.1    Toxic Pollutants of Concern         628
              17.4.2    Process Modifications and           628
                         Technology Transfer Options
              17.4.3    Best Management Practices           630
              17.4.4    Prevailing Control and              630
                         Treatment Practices
              17.4.5    Advanced Treatment Technologies     630
      17.5    SELECTION OF APPROPRIATE TECHNOLOGY AND       631
              " EQUIPMENT
                              xiv

-------
                 TABLE OF CONTENTS - Continued
              17.5.1    Technologies for Different          631
                         Treatment Levels ''
              17.5.2    Equipment for Different             633
                         Treatment Levels
      17.6    TREATMENT COST ESTIMATES                      635
              17.6.1    General Discussion                  635
      17.7.   BASIS FOR REGULATIONS                         636
              17.7.1    Evaluation of BPT Treatment         636
                         Practices
              17.7.2    Basis for Proposed BPT              636
                         Limitations
              17.7.3    Basis for Proposed BCT              649
                         Limitations
              17.7.4    Basis for Proposed BAT              649
                         Limitations
              17.7.5    Basis for Proposed New Source       652
                         Performance Standards
              17.7.6    Basis for Proposed Pretreatment     652
                         Standards

18.0  SODIUM DICHROMATE INDUSTRY                            655

      18.1    INDUSTRY PROFILE                              655
              18.1.1    General Description                 655
              18.1.2    General Process Description and     655
                        Raw Materials
      18.2    WATER USE AND WASTE SOURCE                    658
               CHARACTERISTICS
              18.2.1    Water Use                           658
              18.2.2    Waste Sources                       658
      18.3    DESCRIPTION OF PLANTS VISITED AND SAMPLED     661
              18.3.1    Screeninq                           661
              18.3.2    Verification                        664
              18.3.3    Toxic Pollutant Concentrations      664
                         and Loadings
      18.4    POLLUTION ABATEMENT OPTIONS                   668
              18.4.1    Toxic Pollutants of Concern         668
              18.4.2    Process Modifications and           671
                         Technology Transfer Options
              18.4.3    Best Management Practices           671
              18.4.4    Prevailing Control and              671
                         Treatment Practices
              18.4.5    Advanced Treatment Technoloqies     672
      18.5    SELECTION OF APPROPRIATE TECHNOLOGY AND       672
               EQUIPMENT
              18.5.1    Technology fo'r Different            672
                         Treatment Levels


                              xv

-------
                 TABLE OF CONTENTS - Continued
      18.6    TREATMENT COST ESTIMATES                      676
              18.6.1    General Discussion                  676
              18.6.2    Model Plant Control Costs           677
      18.7    BASIS FOR REGULATIONS                         681
              18.7.1    BPT Effluent Limitations            681
              18.7.2    BCT Effluent Limitations            688
              18.7.3    BAT Effluent Limitations            690
              18.7.4    NSPS Effluent Limitations           693
              18.7.5    Pretreatment Standards              694

19.0  CARBON DIOXIDE INDUSTRY                               697

      19.1    SUMMARY OF DETERMINATIONS                     697
      19.2    ASSESSMENT OF THE WATER POLLUTION             697
               POTENTIAL
              19.2.1    Production Processes and            697
                         Effluents
      19.3    STATUS OF REGULATIONS                         699

20.0  CARBON MONOXIDE AND BY-PRODUCT HYDROGEN INDUSTRY      701

      20.1    SUMMARY OF DETERMINATIONS                     701
      20.2    ASSESSMENT OF THE WATER POLLUTION             701
               POTENTIAL
              20.2.1    Production Processes and            701
                         Effluents
      20.3    STATUS OF REGULATIONS                         703

21.0  COPPER SULFATE INDUSTRY                               705

      21.1    INDUSTRIAL PROFILE                            705
              21.1.1    General Description                 705
              21.1.2    General Process Description and     705
                         Raw Materials
      21.2    WATER USE AND WASTE SOURCE                    708
               CHARACTERISTICS
              21.2.1    Water Use                           708
              21.2.2    Waste Sources                       708
      21.3    DESCRIPTION OF PLANTS ¥ISITSD AND SAMPLED     711
              21.3.1    Screening                           711
              21.3.2    Verification                        713
              21.3.3    Toxic-Pollutant Concentrations      716
      21.4    POLLUTION ABATEMENT OPTIONS                   719
              21.4.1    Toxic Pollutants of Concern         719
              21.4.2    Process Modifications and           719
                         Technology Transfer Options
                              xya.

-------
                 TABLE OF CONTENTS - Continued
              21.4.3    Best Management Practices           719
              21.4.4    Prevailing Control and              720
                         Treatment Practices
              21.4.5    Advanced Treatment Technologies     720
      21.5    SELECTION OP APPROPRIATE TECHNOLOGY AND       721
               EQUIPMENT
              21.5.1    Technologies for Different          721
                         Treatment Levels
              21,5.2    Equipment for Different             721
                         Treatment Levels
      21.6    TREATMENT COST ESTIMATES                      724
              21.6.1    General Discussion                  724
              21.6.2    Model Plant Cost Estimates          726
      21.7    BASIS FOR REGULATIONS                         726
              21.7.1    Evaluation cf BPT Treatment         726
                         Practices
              21.7.2    Basis for Proposed BPT Effluent     731
                         Limitations
              21.7.3    Basis for Proposed BCT Effluent     733
                         Limitations
              21.7.4    Basis for Proposed BAT Effluent     733
                         Limitations
              21.7.5    Basis for Proposed New Source       738
                         Performance Standards
              21.7.6    Basis for Proposed Pretreatment     740
                         Standards

22.0  NICKEL SDLFATE INDUSTRY                               741

      22.1    INDUSTRIAL PROFILE                            741
              22.1.1    General Description                 741
              22.1.2    General Process Description and     741
                         Raw Materials
      22.2    WATER USE AND WASTE SOURCE                    744
               CHARACTERISTICS
              22.2.1    Water Use                           744
              22.2.2    Waste Sources                       744
      22.3    DESCRIPTION OF PLANTS VISITED AND SAMPLED     747
              22.3.1    Screening                           747
              22.3.2    Verification                        747
              22.3.3    Summary of Toxic Pollutant Data     752
      22.4    POLLUTION ABATEMENT OPTIONS                   755
              22.4.1    Toxic Pollutants of Concern         755
              22.4.2    Process Modifications and           755
                         Technology Transfer Options
              22.4.3    Best Management Practices           756
                              xvn

-------
                 TABLE OF CONTENTS - Continued
                                                           Paqe
              22.4.4    Prevailing Control and              756
                         Treatment Practices
              22.4.5    Advanced Treatment Technologies     757
      22.5    SELECTION OF APPROPRIATE TECHNOLOGY AND       757
               EQUIPMENT
              22.5.1    Technologies for Different          757
                         Treatment Levels
              22.5.2    Equipment for Different             757
                         Treatment Levels
      22.6    TREATMENT COST ESTIMATES                      760
              22.6.1    General Discussion                  760
              22.6.2    Model Plant Control Costs           761
      22.7    BASIS FOR REGULATIONS                         767
              22.7.1    Evaluation of BPT Treatment         767
                         Practices
              22.7.2    Basis for Proposed BPT Effluent     770
                         Limitations
              22.7.3    Basis for Proposed BCT Effluent     770
                         Limitations
              22.7.4    Basis for Proposed BAT Effluent     770
                         Limitations
              22.7.5    New Source Performance              777
                         Standards
              22.7.6    Basis for Proposed Pretreatment     777
                         Standards

23.0  SILVER NITRATE INDUSTRY                               779

      23.1    SUMMARY OF DETERMINATIONS                     779
      23.2    ASSESSMENT OF THE WATER POLLDTATION           779
               POTENTIAL
              23.2.1    Production Processes and            779
                         Effluents
      23.3    STATUS OF REGULATIONS                         781

24.0  SODIUM BISULFITE INDUSTRY                             783

      24.1    INDUSTRY PROFILE                              783
              24.1.1    General Description                 783
              24.1.2    General Process Description and     783
                         Raw Materials
      24.2    WATER USE AND WASTE SOURCE                    783
               CHARACTERISTICS
              24.2.1    Water Use                           783
              24.2.2    Water Sources                       787
                             XVlll

-------
                 TABLE OF CONTENTS - Continued
      24.3    DESCRIPTION OF PLANTS VISITED AND SAMPLED
              24.3.1    Screening
              24.3.2    Verification
              24.3.3    Toxic Pollutant Analytical
                         Results
      24.4    POLLUTION ABATEMENT OPTIONS                   800
              24.4.1    Toxic Pollutants of Concern         800
              24.4.2    Prevailing Control and              800
                         Treatment Practices
              24.4.3    Advanced Treatment Technologies     800
      24.5    SELECTION OF APPROPRIATE TECHNOLOGY AND       801
               EQUIPMENT
              24.5.1    Technologies for Different          801
                         Treatment Levels
              24.5.2    Equipment for Different             801
                         Treatment Levels
      24.6    TREATMENT COST ESTIMATES                      805
              24.6.1    General Discussion                  805
              24.6.2    Cost Estimates                      806
      24.7    BASIS FOR REGULATIONS                         813
              24.7.1    Evaluation of BPT Treatment         833
                         Practices
              24.7.2    Basis for Proposed BPT Effluent     813
                         Limitations
              24.7.3    Basis for Proposed BCT Effluent     822
                         Limitations
              24.7.4    Basis for Proposed BAT Effluent     822
                         Limitations
              24.7.5    Basis for Proposed New Source       823
                         Performance Standards
              24.7.6    Basis for Proposed Pretreatment     823
                         Standards

25.0  SODIUM HYDROSDLFITE INDUSTRY                          827

      25.1    INDUSTRY PROFILE                              827
              25.1.1    General Description                 827
              25.1.2    General Process Description and     827
                         Raw Materials
      25.2    WATER USE AND WASTE SOURCE                    830
               CHARACTERISTICS
              25.2.1    Water Use                           830
              25.2.2    Waste Sources                       830
      25.3    DESCRIPTION OF PLANTS VISITED AND SAMPLED     832
              25.3.1    Screening and Verification          832
              25.3.2    Toxic Pollutant Concentrations      834
                               xix

-------
                 TABLE OF CONTENTS - Continued
      25.4    POLLUTION ABATEMENT OPTIONS                   841
              25.4.1    Toxic Pollutants of Concern         841
              25.4.2    Prevailing Control and              841
                         Treatment Practices
              25.4.3    Advanced Treatment Technologies     842
      25.5    SELECTION OP APPROPRIATE TECHNOLOGGY AND      842
               EQUIPMENT
              25.5.1    Technologies for Different          842
                         Treatment Levels
              25.5.2    Equipment for Different             842
                         Treatment Levels
      25.6    TREATMENT COST ESTIMATES                      846
              25.6.1    General Discussion                  846
              25.6.2    Cost Estimates                      846
      25.7    BASIS FOR REGULATIONS                         846
              25.7.1    Evaluation of BPT Treatment         846
                         Practices
              25.7.2    Basis for Proposed BPT Effluent     853
                         Limitations
              25.7.3    Basis for Proposed BCT Effluent     855
                         Limitations
              25.7.4    Basis for Proposed BAT Effluent     855
                         Limitations
              25.7.,5    Basis for Proposed New Source       860
                         Performance Standards
              25.7.6    Basis for Proposed Pretreatment     861
                         Standards

26.0  EXCLUDED SDBCATEGORIES                                863

      26.1    ALUMINUM SULFATE                              863
      26.2    AMMONIUM CHLORIDE                             864
      26.3    AMMONIUM HYDROXIDE "                           866
      26.4    BARIUM CARBONATE                              868
      26.5    BORAX                                         870
      26.6    BORIC ACID                                    871
      26.7    BROMINE                                       873
      26.8    CALCIUM CARBIDE                               874
      26.9    CALCIUM CARBONATE                             874
      26.10   CALCIUM CHLORIDE                              877
      26.11   CALCIUM HYDROXIDE                             878
      26.12   CHROMIC ACID                                  879
      26.13   CUPROUS OXIDE                                 879
      26.14   FERRIC CHLORIDE                               880
      26.15   FERROUS SULFATE                               881
      26.16   FLUORINE                                      882

-------
                 TABLE OF CONTENTS -  Continued
      26.17
      26.18
      26.19
      26.20
      26.21
      26.22
      26.23
      26.24
      26.25
      26.26
      26.27
      26.28
      26.29
      26.30
      26.31
      26.32
      26.33
      26,34
      26.35
      26.36
      26.37
      26.38
      26.39
      26.40
      26.41
      26.42
      26.43
      26.44
      26.45
      26.46

REFERENCES

BIBLIOGRAPHY

APPENDIX A
APPENDIX B
HYDROCHLORIC ACID
HYDROGEN
IODINE
LEAD MONOXIDE
LITHIUM CARBONATE
MANGANESE SULFATE
NITRIC ACID
OXYGEN AND NITROGEN
POTASSIUM CHLORIDE
POTASSIUM DICHROMATE
POTASSIUM IODIDE
POTASSIUM METAL
POTASSIUM PERMANGANATE
POTASSIUM SULFATE
SODIUM BICARBONATE
SODIUM CARBONATE
SODIUM CHLORIDE
SODIUM FLUORIDE
SODIUM HYDROSULFIDE
SODIUM METAL
SODIUM SILICATE
SODIUM SILICOFLUORIDE
SODIUM SULFITE
SODIUM THIOSULFATE
STANNIC OXIDE
STRONG NITRIC ACID
SULFUR DIOXIDE
SULFURIC ACID INDUSTRY
ZINC OXIDE
ZINC SULFATE
Analysis of Long-Term Effluent Monitoring
Data for the Inorganic Chemicals Industry

pH Control of Industrial Was;fce Waters  in
the Inorganic Chemicals Industry
882
885
886
887
887
888
889
890
894
895
896
899
899
900
901
903
902
904
905
906
909
912
912
914
914
916
918
919
922
923

925

931

A-l
B-J
                               xxi

-------

-------
                        LIST OF FIGURES


                                                           Page

5-1    Sample flow sheet for metal analysis                  59

7-1    Solubility of metal hydroxides and sulfides           82
       as a function of pH.

7-2    Electrodialysis process                               97

8-1    Cumulative distribution of daily concentrations      120
       of mercury in treated effluent from plant
       #251.

8-2    Cumulative distribution of daily concentrations      121
       of cyanide in treated effluent from plant
       1765.

8-3    Statistical distribution for daily pollution         124
       measurements.

8-4    Cumulative distribution of 30-day averages           126
       of total cyanide in treated effluent from
       plant 1782.

8-5    Cumulative distribution of 30-day averages           127
       of ammonia in treated effluent from plant
       1782.

8-6    Statistical distributions for 30-day average         129
       pollution measurements.

11-1   General process diagram for production of            156
       chlorine/caustic by mercury cells

11-2   General process flow diagram at plant f299           161
       showing the sampling points.  Chlorine/caustic
       (mercury cell) manufacture

11-3   General process flow diagram at plant f747           165
       showing the sampling points.  Chlorine/caustic
       (mercury cell) manufacture
                             XXlll

-------
                  I.IST OF FIGURES - Continued

                                                           Page
11-4   General process flow diagram at plant fl67           165
       showing the sampling points.  Chlorine/caustic
       (mercury cell) manufacture

11-5   General process flow diagram at Plant f317           167
       showing the sampling points.  Chlorine/caustic
       (mercury cell) manufacture

11-6   Level 1 waste water treatment for                    178
       chlorine-mercury cell subcategory

11-7   Level 2 waste water treatment for                    179
       chlorine-mercury cell subcategory

11-8   Annual treatment cost versus production for          186
       the chlorine subcategory  (mercury cell process)

11-9   Annual unit treatment cost versus production         187
       for the chlorine subcategory (mercury cell
       process)

11-10  Annual treatment cost versus production for          189
       the chlorine subcategory  (mercury cell process)

11-11  Annual unit treatment cost versus production         190
       for the chlorine subcategory (mercury cell
       process)

11-12  Annual treatment cost versus production for          191
       the chlorine subcategory  (mercury cell process)

11-13  Annual unit treatment cost versus production         192
       for the chlorine subcategory (mercury cell
       process)

11-14  General process flow diagram for production          210
       of chlorine/caustic by diaphragm cells

11-15  General process flow diagram at plant f034           215
       showing the sampling points.  Chlorine/caustic
       (diaphragm cell) manufacture

11-16  General process flow diagram at plant £261           218
       showing the sampling points.  Chlorine/caustic
       (diaphragm cell) manufacture
                              xxxv

-------
                  LIST OF FIGURES - Continued

                                                           Page

11-17  General process flowsheet at Plant 1738-A            219
       showing the sampling points.  Chlorine/caustic
       (diaphragm cell) manufacture

11-28  General process flow diagram at Plant                220
       f738-B showing the sampling points.
       Chlorine/caustic  (diaphragm cell) manufacture

11-19  General process flow diagram at Plant f736           222
       showing the sampling points.  Chlorine/caustic
       (diaphragm cell) manufacture

11-20  General process flow diagram at" Plant $967           223
       showing the sampling points.  Chlorine/caustic
       (diaphragm cell) manufacture

11-21  Level 1 waste water treatment for                    241
       chlorine-diaphragm cell subcategory

11-22  Level 2 waste water treatment for                    242
       chlorine-diaphragm cell subcategory

11-23  Level 3 waste water treatment for                    243
       chlorine-diaphragm cell subcategory

11-24  Annual treatment cost versus production for          253
       the chlorine subcategory (diaphragm cell process)

11-25  Annual unit treatment cost versus production         254
       for the chlorine subcategory (diaphragm cell
       process)

12-1   General process flow diagram for production          280
       of hydrofluoric acid

12-2   Production vs. waste flow data for HF plants         284

12-3   General process flow diagram at plant f705           289
       showing the sampling points.  Hydrofluoric
       acid manufacture

12-4   General process flow diagram at Plant i'251           293
       showing the sampling points.  Hydrofluoric
       acid manufacture

12-5   Level 1 waste water treatment for hydrofluoric       302
       acid subcategory
                              xxv

-------
                  LIST OF FIGURES - Continued

                                                           Page
12-6   Level 2 waste water treatment for hydrofluoric       303
       acid subcategory

12-7   Level 3 waste water treatment for hydrofluoric       304
       acid subcategory

12-8   Level 4 waste water treatment for hydrofluoric       305
       acid subcategory

12-9   Waste water treatment new source performance         307
       standard for hydrofluoric acid subcategory

12-10  Annual treatment cost versus production for          314
       the hydrofluoric acid subcategory

12-11  Annual unit treatment cost versus production         315
       for the hydrofluoric acid subcategory

12—12  Annual treatment cost versus production for          321
       the hydrofluoric acid subcategory (NSPS)

12-13  Annual unit treatment cost versus production         322
       for the hydrofluoric acid subcategory (NSPS)

12-14  Fluoride loads and concentrations discharged         330
       at selected hydrofluoric acid plants

14-1   General process diagram for production of            366
       titanium dioxide (chloride process)  from high
       grade ores

14-2   General flow diagram at Plant f559 showing           370
       the sampling points.  (Titanium dioxide -chloride
       process manufacture)

14-3   General flow diagram at Plant ^172 showing           373
       the sampling points.  Titanium dioxide  (chloride
       process) manufacture

14-4   Level 1 waste water treatment for titanium           380
       dixoide - chloride process

14-5   Level 2 waste water treatment for titanium           381
       dioxide - chloride process

14-6   Level 3 waste water treatment for titanium           382
       dioxide - chloride process
                              xxvx

-------
                  LIST OF FIGURES - Continued

                                                           gage

14-7   Annual treatment cost versus production for          389
       the titanium dioxide subcategory, chloride
       process

14-8   Annual unit treatment cost versus production         390
       for the titanium .dioxide subcategory, chloride
       process

14-9   General process flow diagram for production          412
       of titanium dioxide by sulfate process

14-10  General flow diagram at Plant |559 showing           418
       the sampling points.  (Titanium dioxide -sulfate
       process)

14-11  Level 1 waste water treatment for titanium           428
       dixoide - sulfate process

14-12  Level 2 waste water treatment for titanium           429
       dioxide - sulfate process

14-13  Annual treatment cost versus production for          437
       the titanium dioxide subcategory, sulfate
       process

14-14  Annual unit treatment cost versus production         438
       for the titanium dioxide subcategory, sulfate
       process

14-15  General process flow diagram of the titanium         457
       tetrachloride portion of a titanium dioxide
       plant using the chloride-ilmenite process.

14-3.6  Level 1 waste water treatment for titanium           469
       dioxide - chloride  (ilmenite ore) process

14-17  Level 2 waste water treatment for titanium           470
       dioxide - chloride  {ilmenite ore) process

15-1   General process flow diagram for production          496
       of aluminum fluoride

15-2   General process flow diagram at Plant £705           501
       showing the sampling points.  (Aluminum fluoride
       manufacture)
                             xxvii

-------
                  LIST OF FIGURES - Continued

                                                           Page
15-3   General process flow diagram at Plant f251           503
       showing the sampling points.  Aluminum fluoride
       manufacture

15-4   Level 1 waste water treatment for aluminum           513
       fluoride subcategory

15-5   Level 2 waste water treatment for aluminum           514
       fluoride subcategory

15-6   Level 3 waste water treatment for aluminum           515
       fluoride subcategory

15-7   Level 4 waste water treatment for aluminum           516
       fluoride subcategory

15-8   Annual treatment cost versus production for          523
       the aluminum fluoride subcategory

15-9   Annual unit treatment cost versus production         524
       for the aluminum fluoride subcategory

15-10  Effect of variation of pollutant load on             527
       treatment cost at level 1 technology

15-11  Effect of variation of pollutant load on             528
       treatment cost at level 4 technology

15-12  Effect of variation of hydraulic load on             532
       treatment cost at level 2 technology

15-13  Effect of variation of hydraulic load on             533
       treatment cost at level 3 technology

15-14  Effect of variation of hydraulic load on             534
       treatment cost at level 4 technology

16—1   General process diagram for production of            555
       anhydrous chrome oxide

16-2   General process diagram for production of            556
       hydrated chromic oxide

16-3   General process diagram for production of            558
       chrome yellow

16-4   General process diagram for production of            559
       molybdenum orange


                            xxviii

-------
                  LIST OF FIGUBES - Continued
16-5   General process diagram for production of            561
       chrome green

16-6   General process diagram for production of            562
       zinc yellow

16-7   General process diagram for production of            565
       chrome pigment complexes

16-8   General waste water treatment process flow           568
       diagram at Plant §002 showing the sampling
       points.  (Chrome pigment manufacture)

16-9   General waste water treatment process flow           571
       diagram at Plant |894 showing the sampling
       points.  (Chrome pigment manufacture)

16-10  Level 1 waste water treatment for chrome pigments  ,  582

16-11  Level 2 waste water treatment for chrome pig'ments    583

16-12  Annual -treatment cost versus production for          590
       the chrome pigments subcategory

16-13  Annual unit treatment cost versus production         591
       for the chrome pigments subcategory

17-1   General process flow diagram for production          615
       of hydrogen cyanide by the Andrussow Process

17-2   General waste water treatment process flow           620
       diagram at Plant t765 showing the sampling
       points.  (Hydrogen cyanide manufacture)

17-3   General waste water treatment process flow           624
       diagram at Plant 1782 showing sampling points.
       (Hydrogen cyanide manufacture)

17-4   Level 1 waste water treatment for hydrogen           632
       cyanide subcategory

17-5   Level 2 waste water treatment for hydrogen           634
       cyanide subcategory

17-6   Annual treatment cost as a function of production    640
       for the hydrogen cyanide subcategory
                              xxix

-------
                  LIST OP FIGURES - Continued
17-7   Annual unit treatment cost as a function of          641
       production for the hydrogen cyanide subcategory

18-1   General process diagram for production of            659
       sodium dichromate

18-2   General waste water treatment process flow  .         662
       diagram at Plant f493 showing the sampling
       points.  (Sodium dichromate manufacture)

18-3   General waste water treatment process flow           665
       diagram at Plant f376 showing the sampling
       points.  (Sodium dichromate manufacture)

18-4   Level 1 waste water treatment for sodium dichromate  674
       subcategory

18-5   Level 2 waste water treatment for sodium dichromate  675
       subcategory
             »
18-6   Relationship of annual treatment cost to production  682
       for the sodium dichromate subcategory

18-7   Relationship of annual unit treatment cost           683
       to production for the sodium dichromate subcategory

21-1   General block diagram of the manufacture of          709
       copper sulfate

21-2   General process flow diagram at plant |034           714
       showing the sampling points. (Copper Sulfate
       manufacture)

21-3   Level 1 waste water treatment for copper sulfate     722
       subcategory - batch process

21-4   Level 2 waste water treatment for copper sulfate     723
       subcategory - batch process

22-1   General process flow diagram for nickel sulfate      745
       manufacture

22-2   General waste water treatment process flow           748
       diagram showing sampling points at Plant f369.
       (Nickel sulfate subcategory,)
                               xxx

-------
                  LIST OF FIGURES - Continued

                                                           Page

22-3   General process flow diagram at Plant |572           750
       showing the sampling points.  {Nickel sulfate
       manufacture,)

22-4   General waste water treatment process flow           751
       diagram at Plant £120 showing the sampling
       points.  (Nickel sulfate manufacture)

22-5   Level 1 waste water treatment for nickel sulfate     758
       subcategory - batch process

22-6   Level 2 waste water treatment for nickel sulfate     759
       subcategory - batch process

22-7   Relationship of annual treatment cost to production  765
       for the nickel sulfate subcategory

22-8   Relationship of annual unit treatment cost           766
       to production for the nickel sulfate subcategory

24-1   General process flow diagram at Plant f282           790
       showing the sampling points.  Sodium bisulfite
       manufacture

24-2   General flow diagram at Plant f586 showing           791
       the sampling points.  Sodium bisulfite manufacture

24-3   General process flow diagram at Plant 1987           794
       showing the sampling points.  Sodium bisulfite
       manufacture

24-4   Level 1 waste water treatment for sodium bisulfite   802
       subcategory - batch process

24-5   Level 2 waste water treatment for sodium bisulfite   803
       subcategory - batch process

24-6   Level 3 waste water treatment for sodium bisulfite   804
       subcategory

24-7   Variation of annual treatment cost with production   810
       for the sodium bisulfite subcategory

24-8   Variation of annual unit treatment cost with         811
       production  (sodium bisulfite subcategory)
                              xxxi

-------
                  LIST OF FIGUBES - Continued
                                                           Page
25-1   General process flow diagram at Plant f672.          831
       (Sodium hydrosulfite manufacture)
25-2   General process flow diagram at Plant $672           835
       showing the sampling points.  (Sodium hydrosulfite
       manufacture)
25-3   Level 1 waste water treatment for sodium             843
       hydrosulfite subcategory
25-4   Level 2 waste water treatment for sodium             844
       hydrosulfite subcategory
                              XX2O.X

-------
                        LIST OF TABLES

                                                           Page

2-1       Summary of Proposed Regulations - Best              6
          Practicable Control Technology Currently
          Available (BPT)

2-2       Summary of Proposed Regulations - Best              9
          Available Technology (BAT)

2-3       Summary of Proposed Regulations - Pretreatment      12
          Standards for Existing Sources (PSES)

2-4       Summary of Proposed Regulations - New               15
          Source Performance Standards (NSPS)

2-5       Summary of Proposed Regulations - Pretreatment      19
          Standards for New Sources (PSNS)

2-6       Summary of Proposed Regulations - Best             22
          Conventional Pollutant Control Technology
          (BCT)

3-1       Recommended List of Toxic Pollutants                28

3-2       Scope of Industry Coverage within the               34
          Inorganic Chemicals Manufacturing Point
          Source Category

5-1       Analytical Detection Limits for Metals             62

5-2       Pollutant Frequency Based on Sampling               66
          Program Results Including Raw Waste

5-3       Distribution of Pollutants According                67
          to Subcategory

6-1       308 Questionnaire Response Data                    73

7-1       Solubility Products of Trace Metals                 83

7-2       Comparison of Reverse Osmosis Concepts             99
                             XXXill

-------
              LIST OF TABI^ES - Continued

                                                          Page

8-1       Waste Water Treatment Options and Performance      104
          Data Summary - Antimony and Arsenic  Removal

8-2       Waste Water Treatment Options and Performance      105
          Data Summary - Beryllium and Cadmium Removal

8-3       Waste Water Treatment Options and Performance      106
          Data Summary - Copper Removal

8-4       Waste Water Treatment Options and Performance      107
          Data Summary - Chromium III and Chromium
          VI Removal

8-5       Waste Water Treatment Options and Performance      108
          Data Summary - Lead Removal

8-6       Waste Water Treatment Options and Performance      109
          Data Summary - Mercury II Removal

8-7       Waste Water Treatment Options and Performance      110
          Data Summary - Nickel Removal

8-8       Waste Water Treatment Options and Performance      111
          Data Summary - Silver Removal

8-9       Waste Water Treatment Options and Performance      112
          Data Summary - Selenium and Thallium Removal

8-10      Waste Water Treatment Options and Performance      113
          Data Summary - Zinc Removal

8-11      Estimated Achievable Maximum 30-Day  Averages       115
          for the Applied Technologies

9-1       Prioritization of Toxic Metals Found in           132
          Each Subcategory

11-1      Subcategory Profile Data Summary                  152

11-2      Status of Regulations - Effluent Limitation        153
          Guidelines

11-3      Summary of Waste Water Plow Data for Chlorine      159
          Mercury Cell Plants

11-4      Pollutant Concentration and Loads at Plant         162
          £299
                             XXXIV

-------
              LIST OF TABLES - Continued

                                                           Paqe
11-5      Pollutant Concentrations and Loads at             163
          Verification Plants (1)

11-6      Toxic Pollutant Raw Waste Concentrations           172
          and Loads at Verification Plants   mg/1
                                           kg/kkg

11-7      Summary of Raw Waste Loadings at Verification     173
          Plants

11-8      Model Plant Treatment Costs                       183

11-9      Model Plant Treatment Costs                       184

11-10     Model Plant Treatment Costs                       185

11-11     Model Plant Dnit Treatment Costs                  188

11-12     Estimated Chemical Dechlorination Costs            193
          for the Chlor-Alkali Industry

11-13     Mercury Discharges from Selected                  195
          Chlor-Alkali Mercury Cell Plants*

11-14     Residual Chlorine Discharges at Selected           196
          Chlor-Alkali Plants*

11-15     Comparison of Raw Waste Concentrations             198
          of Toxic Pollutants with Treatability

11-16     Proposed Limitations, BAT                         199

11-17     Effluent Concentrations of Toxic Pollutants        202
          from Verification Sampling

11-18     Subcategory Profile Data Summary                  207

11-19     Status of Regulations - Effluent Limitation        208
          Guidelines

11-20     Waste Water Plows at Diaphragm Cell Chlorine       213
          Plants

11-21     Pollutant Concentrations and Loads at             216
          Screening and Verification Plants
                             xxxv

-------
              LIST OF TABLES - Continued

                                                           Page
11-22     Results of Asbestos Sampling at Diaphragm         225
          Cell Plants

11-23     Maximum Raw Waste Concentrations of Toxic         226
          Metals Observed at Diaphragm Cell Chlorine
          Plants (mg/1)

11-24     Toxic Metal Concentrations and Loads at           229
          Screening and Verification Plants/'  mg/1
\(  mg/1  \
 V kg/kkg /
11-25     Summary of Raw Waste Loadings at Screening        230
          and Verification Metal Anode Plants

11-26     Toxic Metal Concentrations and Loads in           232
          Cell Room Waste Waters at Screening and
          Verification Plants/ mg/1\
                             \kg/kkg/

13-27     Raw Waste Toxic Metals Concentration and          233
          Loads in Process Streams Other Than Cell
          Room Wastes from Screening and Verification
          Plants

11-28     Raw Waste Toxic Organics at  a Graphite            234
          Anode Plant

11-29     Raw Waste Toxic Organics by  Waste Water           235
          Source at a Graphite Anode Plant

11-30     Model Plant Treatment Costs                        249

11-31     Model Plant Treatment Costs                        250

11-32     Model Plant Treatment Costs                        251

11-33     Model Plant Treatment Costs                        252

11-34     Summary of Unit Flows at Diaphragm Cell           255
          Plants

11-35     Comparison of Toxic Metals Treatability           257
          with Screening and Verification Sampling
          Data

11-36     Proposed Limitations, BPCTCA                      258
                             XXXVI

-------
              LIST OP TABLES - Continued

11-37
11-38
11-39
11-40
11-41
11-42
12-1
12-2
12-3
12-4
12-5
12-6
12-7
12-8
!
Lead and TSS Discharges from Selected Diaphragm
Cell Chlorine Plants (1)
Toxic Pollutants in Diaphragm Cell Plant
Effluents
Proposed Limitations, BAT
Proposed Limitations, BCT
i
Proposed Limitations, NSPS
Comparison of Raw Waste Ch-a-racteristics
at a New Metal Anode Plant with Treatability
of Toxic Metals
Subcategory Profile Data Summary-
Status of Regulations - Effluent Limitation
Guidelines
Water Usage in the Hydrofluoric Acid
Subcategory
Waste Water Flow and Reuse Data for the
Hydrofluoric Acid Subcategory
Waste Flow from Hydrofluoric Acid
Manufacturing Plants
Solid Waste Generated at the Hydrofluoric
Acid Plants Sampled
Gypsum Solids Production in the Hydrofluoric
Acid Subcategory
Plow and Pollutant Concentration Data of
Page
261
262
265
269
271
272
276
277
281
282
286
287
288
290
          the Sampled Waste Streams for  Plant  f705
          Producing Hydrofluoric Acid

12-9      Plow and Pollutant Concentration Data  of           292
          the Sampled Waste Streams for  Plants f705,
          $251,  and f!67 Producing Hydrofluoric  Acid

12-10     Toxic Pollutant Raw Waste Data                    296

12-11     Summary of Raw Waste Loadings  Found  in            297
          Screening and Verification Sampling

                             xxxvii

-------
     LIST OF TABLES - Continued

12-12
12-13
12-14
12-15
12-16
12-17
12-18
12-19
12-20
12^21
12-22
12-23
12-24
12-25
12-26
12-27
12-28

12-29
13-1
14-1
14-2

Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Treatment Costs
Model Plant Treatment Costs
Summary of Waste Water Control and Treatment
Technology Employed at Hydrofluoric Acid
Plants
Summary of Long-Term Monitoring Data from
Pour Hydrofluoric Acid Plants
Toxic Pollutant Treated Effluent Data
Development of TSS and Fluoride Limitations
Proposed Limitations, BPCTCA
Proposed Limitations, BAT
Performance of Alternative Technology,
Level 3 Treatment
Performance of Alternative Technology,
Level 4 Treatment
Toxic Pollutant Raw Waste Data Used to
Represent New Sources*
Proposed Limitations, NSPS
Subcategory Profile Data Summary
Subcategory Profile Data Summary
Status of Regulations - Effluent Limitation
Page
310
311
312
316
318
319
320
323
325
327
328
334
335
343
345
346
351

353
359
362
363
Guidelines
                xxxvxxi

-------
              LIST OF TABLES - Continued

                                                           Page

14-3      Water  Osage in Titanium Dioxide-Chloride          367
          Process/High Grade Ores Subcategory

14-4      Waste  Water Flow for  Titanium Dioxide-Chloride     369
          Process Subcategory

1*4-5      Flow and Pollutant Concentration Data of          372
          the Sampled Waste Streams of Plant f!72
          Producing Titanium Dioxide by Chloride-
          Rutile Process

14-6      Flow and Pollutant Concentration Data of          374
          the Sampled Waste Streams for Plant f!72
          Producing Titanium Dioxide (Chloride Process)

14-7      Raw Waste Pollutant Data Summary of the            376
          Sampled Streams

14-8      Model  Plant Treatment Costs                       386

14-9      Model  Plant Treatment Costs                       387

14-10     Model  Plant Treatment Costs                  "    388

14-11     Model  Plant Treatment Costs                       391

14-12     Historical Effluent Monitoring Data Summary       393
          with Variability Factor

14-13     Historical Effluent Monitoring Data Summary       394
          with Variability Factors Daily Measurements

14-14     Treatment Performance Data of Sampled Plants       395
          1599 and f!72

14-15     Proposed Limitations, BPCTCA                      401

14-16     Proposed Limitations, BAT  .                       402

14-17     Proposed Limitations, NSPS                        407

14-18     Subcategory Profile Data Summary                  409

14-19     Analysis of Ilmenite  Ores                         410

14-20     Water  Usage in Titanium Dioxide - Sulfate          413
          Process Subcategory
                             XXXLX

-------
              LIST OF TABLES - Continued

                                                          Page

14-21     Raw Waste Characteristics (Industry Data)          416
          for Plant 1555 (Production of Ti02  by Sulfate
          Process)

14-22     Flows and Pollutant Concentrations  for            419
          the Waste Streams Sampled for Plant $559
          Producing Titanium Dioxide

14-23     Process Waste Water Flow at Plants  f555r           421
          1694 and £559 Titanium Dioxide (Sulfate
          Process)

14-24     Summary of Raw Waste Loadings Found in            423
          Screening and Verification Sampling

14-25     Toxic Pollutants: Average Raw Waste Loads          424
          and Concentrations

14-26     Model Plant Treatment Costs                       434

14-27     Model Plant Treatment Costs      '                435

14-28     Model Plant Treatment Costs                       436

14-29     Model Plant Treatment Costs                       439

14-30     Historical Effluent Monitoring Data Summary       440

14-31     Verification Results from - Sulfate Process       443
          Titanium Dioxide Plant f559

14-32     Proposed Limitations, BPTCTA                      451

14-33     Proposed Limitations, BAT                         453

14-34     Proposed Limitations, NSPS                        454

14-35     Subategory Profile Data Summary                   456

14-36     Average Water Usage for TiO~ Production           459
          by the Chloride - Ilmenite process

14-37     Average Raw Waste Loads for Ti02 Production       460
          by the Chloride - Ilmenite Process

14-38     Summary of Raw Waste Loadings Found in            464
          Screening and Verification Sampling

-------
              LIST OF TABLES - Continued

                                                          Page
14-39     Toxic Pollutant Average Raw Waste Loads            465
          and Concentrations

14-40     Model Plant Treatment Costs                       474

14-41     Model Plant Treatment Costs                       475

14-42     Model Plant Treatment Costs                       476

14-43     Model Plant Treatment Costs                       477

14-44     Model Plant Treatment Costs                       478

14-45     Proposed Limitations, BPCTCA                      482

14-46     Proposed Limitations, NSPS                         487

15-1      Subcategory Profile Data Summary                  494

15-2      Status of Regulations - Effluent Limitation        495
          Guidelines

15-3      Water Usage in the Aluminum Fluoride              498
          Subcategory

15-4      Waste Water Plow at Plants  |837, f705  and          499
          1251 for Aluminum Fluoride  Subcategory

15-5      Solids Generated at Plant f705  and f251            499
          Producing Aluminum Fluoride

15-6      Flow and Pollutant Concentration Data  of           502
          the Sampled Waste Streams for Plant f705
          Producing Aluminum Fluoride

15-,7      Flow and Pollutant Concentration Data  of           504
          the Sampled Streams for Plant f251 Producing
          Aluminum Fluoride

15-8      Toxic Pollutant Average Raw Waste Loads            507
          and Concentrations

15-9      Toxic Pollutant Effluent Concentrations            508
          During Sampling

15-10     Summary of Raw Waste Loadings Found in            509
          Screening and Verification  Sampling
                              xli

-------
              LIST OF TABLES - Continued

                                                          Page
15-11     Model Plant Treatment Costs                       520

15-12     Model Plant Treatment Costs                       521

15-13     Model Plant Treatment Costs                       522

15-14     Model Plant Treatment Costs                       525

15-15     Model Plant Treatment Costs                       526

15-36     Model Plant Treatment Costs                       530

15-17     Model Plant Treatment Costs                       531

15-18     Model Plant Treatment Costs                       535

15-19     Proposed Limitations, BPCTCA                      543

15-20     Proposed Limitations, BAT                         544

15-21     Performance of Alternative Technology Level        545
          3 Treatment

15-22     Performance of Alternative Technology Level        546
          4 Treatment

15-23     Proposed Limitations, NSPS                        550

16-1      Subcategory Profile Data Summary                  552

16-2      Status of Regulations - Effluent Limitation        553
          Guidelines
                   *>

16-3      Water Usage in the Chrome Pigments  Subcategory     563

16-4      Summary of Waste Water Plow                       566

16-5      Plow, Pollutant, Concentration and  Load            569
          Data of the Sampled Waste Streams for Plant
          1002

16-6      Flow, Pollutant, Concentration and  Load            572
          Data for the Sampled Streams at Plant £894

16-7      Toxic Pollutant Raw Waste Data                    574
                              xlii

-------
              LIST OF TABLES - Continued

                                                          Page

16-8      Summary of Raw Waste Loadings Found  in            575
          Screening  and Verification Sampling

16-9      Toxic Pollutant Treated Waste Data                576

16-10     Model Plant Treatment Costs                       586

16-11     Model Plant Treatment Costs                       587

16-12     Model Plant Treatment Costs                       588

16-13     Model Plant Treatment Costs                       589

16-14     Model Plant Treatment Costs                       593

16-15     Summary of Long Term and Verification Effluent     596
          Sampling Results at Plant f894

16-16     Proposed Limitations, BPCTCA                      603

16-17     Proposed Limitations, NSPS                        606

17-1      Subcategory Profile Data Summary                  612

17-2      Status of  Regulations - Effluent Limitation        613
          Guidelines

17-3      Water Usage in Hydrogen Cyanide - Andrussow        636
          Process Subcategory

17-4      Waste Flow Data for HCN Production by the          618
          Andrussow  Process

17-5      Flow and Pollutant Data for the Raw  and            621
          Treated Waste Streams of Plant f765  Producing
          Hydrogen Cyanide by Andrussow Process

17-6      Flow and Pollutant Concentration Data of           622
          the Sampled Waste Streams for Plant  |765
          Producing  Hydrogen Cyanide

17-7      Flow and Pollutant Concentration Data of           625
          the Sampled Waste Streams for Plant  #782
          Producing  Hydrogen Cyanide

17-8      Unit Flow  and Unit Pollutant Loading for           626
          Raw and Treated Waste Effluents at Plant
          £782


                              xliii

-------
              LIST OF TABLES - Continued

                                                          Page

17-9      Summary of Pollutant Raw Waste Loading            629
          Pound in Screening and Verification Sampling

17-10     Model Plant Treatment Costs                      637

17-11     Model Plant Treatment Costs                      638

17-12     Model Plant Treatment Costs                      639

17-13     Model Plant Treatment Costs                      642

17-14     Statistical Analysis of the 28-Day Effluent       645
          Sampling Results on Total Cyanide and Ammonia
          from Plant f765

17-15     Statistical Analysis of Historical Effluent       648
          Monitoring Data on Free Cyanide from Plant
          f765

17-16     Proposed Limitations, BPCTCA                     650

17-17     Proposed Limitations, BAT                        653

17-18     Control Parameter Limitations, NSPS              '654

18-1      Subcategory Profile Data Summary                 656

18-2      Status of Regulations - Effluent Limitation       657
          Guidelines

18-3      Water Usage in Sodium Dichromate Subcategory      660

18-4      Flow and Pollutant Concentration Data of          663
          the Sampled Waste Streams for  Plant f493
          Producing Sodium Dichromate

18-5      Flow and Pollutant Loading Data of the            666
          Sampled Waste Streams for Plant |376 Producing
          Sodium Dichromate

18-6      Flow and Pollutant Loading Data of the            667
          Sampled Waste Streams for Plant f398 Producing
          Sodium Dichromate

18-7      Toxic Pollutant Raw Waste Data                   669
                              xLiv

-------
              LIST OF TABLES - Continued

                                                          Page

18-8      Summary of Raw Waste Loadings  Pound  in            670
          Screening  and Verification Sampling

18-9      Model Plant Treatment Costs                       678

18-10     Model Plant Treatment Costs                       679

18-11     Model Plant Treatment Costs                       680

18-12     Model Plant Treatment Costs                       684

18-13     Effluent Sampling Data from Sodium Bichromate      687
          Plants

18-14     Proposed Limitations, BPCTCA                      689

18-15     Proposed Limitations, BAT                          695

18-16     Control Parameter Limitations, NSPS                696

19-1      Subcategory Profile Data Summary                   698

20-1      Subcategory Profile Data Summary                   702

21-1      Subcategory Profile Data Summary                   706

21-2      Status of  Regulations - Effluent Limitation        707
          Guidelines

21-3      Water Usage in Copper Sulfate  Subcategory          710

21-4      Waste Water Flow for the Copper Sulfate            712
          Subcategory

21-5      Flow and Pollutant Concentration Data  of           715
          the Sampled Waste Streams  for  Plant  f034
          Producing  Copper Sulfate

21-6      Raw Waste  Data                                    718

21-7      Model Plant Treatment Costs                       727

21-8      Model Plant Treatment Costs                       728

21-9      Summary of Long-Term Monitoring Data from          729
          Plant 1034
                              xlv

-------
              LIST OF TABLES - Continued

                                                          Paqe
21-10     Treated Effluent Data                            730

21-11     Average Pollutant Levels and  Removal Efficiency   732
          for Plant 1034

21-12     Proposed Limitations, BAT                        736

21-13     Performance of Alternative Technology,            739
          Treatment Level 2

22-1      Subcategory Profile Data Summary                 742

22-2      Status of Regulations - Effluent Limitation       743
          Guidelines

22-3      Water Use in the Nickel Sulfate  Subcategory       746

22-4      Flow and Pollutant Concentration Data of          749
          the Sampled Waste Streams for Plants Producing
          Nickel Sulfate

22-5      Toxic Pollutant Raw Waste Data                   754

22-6      Model Plant Treatment Costs                      762

22-7      Model Plant Treatment Costs                      763

22-8      Model Plant Treatment Costs                      764

22-9      Model Plant Treatment Costs                      768

22-10     Toxic Pollutant Treated Effluent Data             769

22-11     Proposed Effluent Limitations, BAT                773

22-12     Proposed Effluent Limitations, Treatment          776
          Level 2

23-1      Subcategory Profile Data Summary                 780

24-1      Subcategory Profile Data Summary                 784

24-2      Status of Regulations - Effluent Limitation       785
          Guidelines

24~3      Water Usage in the Sodium Bisulfate Subcategory   786
                             xlvi

-------
              LIST OF TABLES - Continued

                                                          Page
24-4      Waste Water  Flow at Plant  |987  and  f282           788
          for Sodium Bisulfite Subcategory

24-5      Flow and Pollutant Load  Data of the Sampled       789
          Waste Streams for Plant  f282 Producing
          Sodium Bisulfite

24-6      Flow and Pollutant Load  Data of the Sampled       792
          Waste Streams for Plant  1586

24-7      Flow and Pollutant Load  Data of the Sampled       793
          Waste Streams for Plant  1987

24-8      Toxic Pollutant Raw Waste  Loads                  797

24-9      Summary of Raw Waste Loadings Found in            798
          Screening and Verification Sampling

24-10     Toxic Pollutant Concentrations  Observed           799
          in Treated Effluent During Verification
          Sampling

24-11     Model Plant  Treatment Co'sts       '               807

24-12     Model Plant  Treatment Costs                      808

24-13     Model Plant  Treatment Costs                      809

24-14     Model Plant  Treatment Costs                      812

24-15     Plant Performance Evaluation Summary  for          814
          Conventional and Nonconventional  Pollutants

24-16     Proposed Limitations, BPCTCA                     818

24-17     Performance  of Alterantive Technology,            824
          Treatment Level 2

24-18     Performance  of Alternative Technology,            825
          Treatment Level 3

25-1      Subcategory  Profile Data Summary                  828

25-2      Status of Regulations -  Effluent  Limitation       829
          Guidelines

25-3      Waste Source Data at Plant |672                  833
                             xlvii

-------
              LIST OF TABLES - Continued
25-4      Flowr Pollutant,  Concentration,  and Load          836
          Data of the Sampled Waste Streams  for Plant
          §672 Producing Sodium Hydrosulfite

25-5      Sampling Results  and Treatment System             839
          Performance for Toxic Pollutants Plant
          #672

25-6      Summary of Raw Waste Loadings and  Concentration   840
          Found at a Sodium Hydrosulfite Plant  (Formate
          Process)

25-7      Model Plant Treatment Costs                       847

25-8      Model Plant Treatment Costs                       848
                                          \

25-9      Subcategory Performance Evaluation Summary        850
          at Plant #672 for Conventional and
          Nonconventional Pollutants in the  Effluent

25-10     Proposed Limitations, BPCTCA                     854

25-11     Proposed Limitations, BAT                        858

25-12     Proposed Limitations, NSPS                       862

26.2-1    Subcategory Profile Data Summary                 865

26.3-1    Subcategory Profile Data Summary                 867

26,4-1    Subcategory Profile Data Summary                 869

26.6-1    Subcategory Profile Data Summary                 872

26.9-1    Subcategory Profile Data Summary                 876

26.17-1   Subcategory Profile Data Summary                 884

26.23-1   Subcategory Profile Data Summary                 891

26.24-1   Subcategory Profile Data Summary                 893

26.27-1   Subcategory Profile Data Summary>                 898

26.32-1   Subcategory Profile Data Summary                 903

26.35-1   Subcategory Profile Data Summary                 907
                              xlviii

-------
              LIST OF TABLES - Continued
                                                          Page
26.36-1   Subcategory Profile Data Summary                  908
26.37-1   Subcategory Profile Data Summary                  911
26.38-1   Subcategory Profile Data Summary'                 913
26.40-1   Subcategory Profile Data Summary                  915
26.42-1   Subcategory Profile Data Summary                  917
26.43-1   Subcategory Profile Data Summary           ~       920
26.44-1   Subcategory Profile Data Summary                  921
                              xlix

-------

-------
                       ACKNOWLEDGEMENTS
     The technical study supporting the proposed regulation was
conducted by Jacobs Environmental Division of Jacobs Engineering
Group Inc. of  Pasadena,  California under the direction  of Mr.
Henry Cruse,  Vice  President  and  Mr.  Michael Warner,  Project
Manager.   Major contributors were  Ms.  Bonnie J.  Parrott,  Mr.
Mazhar Mohiuddin,  Mr.  Mahendra  L.  Shah, Mr.  Carl  B.  Johnston,
Mr. Dennis Merklin, Mr. Mark A. Jackson, Mr.  Dale  Newkirk, Mr.
William  E.  Rorie, Mr.  John H.  Taylor,  Jr.,   Dr.  Nelson  F.
Phelan,  Ms.  Maureen  Smith,  Mr.   Dale Rushneck,  Dr.  Ben  C.
Edmondson and Dr. Martin D, Schwartz.

     Cooperation  and  assistance of  the Chemical Manufacturers
Association and  the Chlorine  Institute and  numerous individual
corporations is appreciated for  their review and comment on the
draft development document.   The  numerous  company and  plant
personnel who submitted information, opened their plants to the
program  staff,  and otherwise cooperated  are acknowledged and
thanked for their patience and help.

     The guidance  and  assistance  of Mr. Swep Davis,  Associate
Assistant  Admimistrator  for  Water  and  Waste  Management  is
greatly appreciated.

     Mr.  Steve  Schatzow,  Deputy  Assistant  Administrator  for
Water Planning and Standards is  gratefully acknowledged for his
contributions to the project, both as an administrator and as a
member of the Office of General  Counsel.

     Mr. Walter J. Hunt, former  Chief of the Inorganic Chemicals
and   Services   Industries  Branch,   is   acknowledged   for   a
significant contribution in the  initial stages of the project.

     Ms. Susan Lepow and Staff of the Office of  General Counsel
are specially  acknowledged  for  their extensive  contribution to
the drafting of the regulations  and this development document.

     Ms. Emily  Hartnell and Mr. Richard Catz of the  Office of
Analysis  and Evaluation,  Mr..Mark  Segal,  Monitoring  and  Data
Support  Division,  and  Fred  Talcott,  Office of Planning  and
Evaluation are acknowledged for  their assistance.

     Word  processing   for  this  project was  performed  by  Ms.
Nancy Zrubech with assistance from Ms. Kaye Starr.   Their  work
is especially appreciated.

                              li

-------

-------
                           SECTION 1
                    CONCLUSIONS AND SUMMARY
1.1  TOXIC POLLUTANTS
     For the  purpose  of  establishing   waste  water effluent
limitation  guidelines  for  existing sources  and standards  of
performance  for  new sources  in  this   study, the  following 35
inorganic chemical product subcategories were screened:
      1.  Chlor-Alkali
      2.  Hydrofluoric Acid
      3.  Titanium Dioxide
      4.  Aluminum Fluoride
      5.  Chrome Pigments
      6.  Hydrogen Cyanide
      7.  Sodium Bichromate
      8.  Copper Sulfate
      9.  Nickel Sulfate
     10.  Sodium Bisulfite
     11.  Sodium Hydrosulfite
     12.  Hydrogen Peroxide
     13.  Hydrochloric Acid
     14.  Nitric Acid
     15.  Sodium Carbonate
     16.  Sodium Metal
     17.  Sodium Silicate
     18.  Sulfuric Acid
19.  Carbon Dioxide
20.  Carbon Monoxide and
      by-product Hydrogen
21.  Silver Nitrate
22.  Ammonium Chloride
23.  Ammonium Hydroxide
24.  Barium Carbonate
25.  Boric Acid
26.  Calcium Carbonate
27.  Cuprous Oxide
28.  Manganese Sulfate
29.  Strong Nitric Acid
30.  Oxygen and Nitrogen
31.  Potassium Iodide
32.  Sodium Hydrosulfide
33.  Sodium Silicofluoride
34.  Sodium Thiosulfate
35.  Sulfur Dioxide
     The screening  studies  showed  that the plant process waste
waters from subcategories  1  through 11 contain the toxic  metals
(see Table 3-1)f   cyanide  and asbestos.  Very few of the organic
toxic pollutants   were  found in process waste streams and those
that were  identified, in  most cases, were  present  in low level
concentrations.

     The screening results which  indicated  the presence of toxic
pollutants in significant  amounts were  largely confirmed by  the
results  of the  verification  program.    Verification  sampling

-------
accounted for 50 to 75 percent of. the current  inorganic chemical
production rate in the subcategories covered.

     The sources  of  most of  the  toxic  pollutants found in the
raw  wastes  and  treated  effluents  can be  traced  to  specific
process-related   raw   materials   and  chemicals  used  in  the
manufacturing operations.   In  the  case of   certain pollutants
found in widely  varying amounts or with erratic  frequencies of
occurrence, the precise identities of  the sources  remain unknown
at this timer but are  suspected to be process-related.


1.2  CONTROL AND TREATMENT TECHNOLOGY

     A  considerable   amount  of   toxic  pollutant  removal  is
presently achieved in  the industry by the  existing control and
treatment practices.  Additional removal can  be accomplished by
the application of available and demonstrated  technologies which
would add to or modify existing treatment systems.  Recovery of
the heavy metals for value or  reuse in a process does not  appear
to be  an attractive alternative  in  those  industries  where the
product  recovery  practices  now  in  effect  do  not  already
accomplish this.

     The treatment of  toxic metal-bearing waste streams results
in the production of  sludges or  residues  which are potentially
hazardous  and  may  require  special  means   for   handling  and
disposal under the Resource  Conservation and Recovery Act (RCRA)
regulations.


1.3  COSTS OF ADDITIONAL IN-PLANT TREATMENT

     The estimated incremental costs  of  applying the candidate
BAT treatment options represent a relatively small proportion of
the  investment  and  operating  and  maintenance  costs  already
committed to  the  existing BPT level treatment systems.    These
costs, however, vary widely  from   industry  to industry  and are
highly dependent on site-specific  factors.


1.4  SUBCATEGORIZATION

     A review of the product/process  basis for  subcategorization
of the inorganic chemical product subcategories   designated for
study revealed  that certain  modifications  may   be appropriate
in the interest of developing effective regulations.  The toxic
pollutant problem per se impacts  subcategorization directly only
in the Chlor-Alkali Industry where the use of graphite anodes
contributes to  the generation of chlorinated hydrocarbons.  In
the Titanium Dioxide   Industry, major  process and raw material

-------
differences justify the creation  of a separate  segment for the
sulfate process, the   chloride  process,   and for   the chloride
process  using   ilmenite  ore.    Consideration  was  given   to
creating    a  subcategory  for   the  combined   production  of
hydrofluoric acid and aluminum fluoride in view of their similar
waste  characteristics  and  the  current  practice  of  combined
treatment at several plants.  However, combining these products
into  a  single  subcategory  does  not  appear  to  offer  any
regulatory  advantages.

     Hydrogen cyanide is produced by  the  Andrussow process and
as a by-product in  the manufacture of  acrylonitrile.  By-product
hydrogen  cyanide will  be covered  under  its primary product,
acrylonitrile,   in   the   Inorganic   Chemicals   Manufacturing
Category.    The  hydrogen  cyanide   subcategory  includes  only
manufacture by the  Andrussow process.


1.5  RESTUDY OF REMANDED REGULATIONS

     The Fourth Circuit, U.S. Court  of Appeals remanded effluent
limitations  guidelines  promulgated   for  11  major  inorganic
chemical products.  E.I.  du Pont de Nemours  v.  Train, 541 F.2d
1018 (4th. Cir. 1976) revised in Part  430 U.S. 112 (1977).  The
factors  affecting  the  control  and  treatment  of  pollutant
discharges in those  industries  have been studied in response to
the  remanded  issues.   It h'as  been  concluded  that  alternative
control   and   treatment   technologies  to   those   originally
considered for BAT and NSPS may be appropriate.

-------

-------
                           SECTION 2


                        RECOMMENDATIONS
     •On  the  basis   of   the  toxic  pollutant  screening  and
verification   results  and   the   evaluation   of   applicable
technologies  for  discharge   control   and   treatment,   it  is
recommended that  effluent limitation  guidelines,   new  source
performance standards  and pfetreatment standards  for new  and
existing sources  be  proposed  for  the  following   11  inorganic
chemical manufacturing subcategories:


          Chlor-Alkali             Sodium Dichromate
          Hydrofluoric Acid        Copper Sulfate
          Titanium Dioxide         Nickel Sulfate
          Aluminum Fluoride        Sodium Bisulfite
          Chrome Pigments          Sodium Hydrosulfite
          Hydrogen Cyanide


     Table 2-1 summarizes  the  proposed   regulations   for  Best
Practicable  Control   Technology Currently    Available  (BPT).
Summaries of proposed regulations for Best Available Technology
(BAT)r  Pretreatment  Standards,  and  New  Source  Performance
Standards  are  given  in  Tables  2-2,  2-3,  2-4,  2-5,  and  2-6.
These tables indicate that Chlor-Alkali has been  divided  into
two  segments  and  Titanium  Dioxide in  three  segments  before
listing  the  numerical  effluent limitations  for  the  proposed
regulations.

     In addition, in the  following  subcategories, although toxic
pollutant  discharges  have  not   been   found   in  significant
quantities,  discharge   of  conventional  and   nonconventional
pollutants should be controlled by the permitting authority.

          Hydrogen Peroxide       Sodium Metal
          Hydrochloric Acid       Sodium Silicate
         . Nitric Acid             Sulfuric Acid
          Sodium Carbonate

-------
         TABLE 2-1.   SUMMARY OP PROPOSED REGULATIONS -
                           KRUCTICAHiE CONTROL TEOTSDLOGY
                      CURRENTLY AVAILABLE (BPT)
                                    Effluent Limitations
QLUJUO, wsyu iy

Chlor-alkali,
Mercury Cells
jrcticuiRSUtiL

TSS
Mercury
pH
Max
30-day Avg
kg/kkg (or lb/1000
0.32-
0.00014
24-hr
Max
Ib. ) of product
0.64
0.00028
pH Range

6.0 to 9.0
CKLor-alteli,    TSS
Diaphragm Cells  Chromium  (T)
                 Copper  (T)
                 Lead  (T)
                 Nickel  (T)
                 Zinc  (T)
Hydrofluoric
 Acid
Sodium
 Bichromate
                 pH
TSS
Plouride  (T)
Antimony  (T)
Chromium  (T)
Copper  (T)
Lead  (T)
Nickel (T)
Zinc  (T)
PH
TSS
Hexavalent
                  0.51
                  0.00088
                  0.0044
                  0.010
                  0.0044
                  0.0044
5.3
2.9
0.044
0.0055
0.027
0.016
0.0093
0.030
0.22
                   1.1
                   0.0023
                   0.011
                   0.026
                   0.011
                   0.011
11.0
 6.1
 0.088
 0.011
 0.054
 0.033
 0.019
 0.060
 0.44
                                                     6.0 to  9.0
                                                                      6.0 to 9.0



Titanium
Dioxide
(sulfate
process)




Chromium
Chromium (T)
pH


TSS
Icon (T)
Arsenic (T)
Antimony (T)
Cadmium (T)
Chromium (T)
0.0044
0.00050



30
1.2
0.24
0.38
0.070
0.070
0.0088
0.0009
6.0 to 9.1


110
4.1
0.46
0.71
0.11
0.13
 (contajnued)

-------
                      2-1.  Qontinued
Subcategory
                                   Effluent Limitations
Parameter
                                   Max"
                               30-c3ay Avg
                                      24 -hr
                                       Max
pH Range
..-r,. _ -1 i .. _. ni, -- -
Titanium
Dioxide
(sulfate
process)


Titanium
Dioxide
(Chloride
Process)

Titanium Diox-
ide (Chloride
Ilmenite Ero-
cess)








Aluminum
Fluoride




Copper Sulfate




kg/kkg (or lb/1000 Ib.) of product
-•,!••

Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
PH

TSS
Iron (T)
Chromium (T)
pH

TSS
Iron (T)
JtatinDny (T)
Arsenic (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
pH

TSS
Fluoride (T)
Chromium (T)
Nickel (T)
pH
TSS
Copper (T)
Nickel (T)
Selenium (T)
pH

0.24
0.14
0.10
0.24


6.4
0.25
0.14


7.7
0.30
0.096
0.060
0.012
0.012
0.060
0.036
0.024
0.060


1.2
0.63
0.0012
0.0024

0.023
0.0010
0.0020
0.00050


0.46
0.21
0.18
0.50


23
0.84
0.027


28
1.0
0.18
0.11
0.019
0.023
0.11
0.054
0.046
0.013


2.4
1.3
0.0024
0.0048

0.069
0.0030
0.0060
0.0015






6.0 to 9.0



6.0 to 9.0











6.0 to 9.0





6.0 to 9.0




6.0 to 9.0

-------
                        2-1.   Continued
Subcategory
                                    Effluent Limitations
 I&rametex
                                    Max,                24-hr
                                30-day Avg               Max
                                kg/kkg  (or lb/1000 Ib.)  of product
                                                      pH Range
Hydrogen
Cyanide




TSS
Jtoraonia-N
Cyanide (Free)
Cyanide (T)

2.0
4.3
0.016
0.23

5.4
12
0.043
0.65
Nickel Sulfate
Chrome Pigments
Sodium Bisul-
 fite
Sodium Hydro-
 sulfite
pH

TSS
Nickel  (T)
pH

TSS
Antimony  (T)
Cadmium (T)
Chrcanium  (T)
Copper  (T)
Lead  {T)
Nickel  (T)
Zinc  (T)
PH
TSS
COD
Chrcmium  (T)
Copper  (T)
Lead  (T)
Nickel  (T)
Zinc•(T)
pH
TSS
COD
  0.032
  0.0020
 3.9
 0.051
 0.020
 0.12
 0.042
 0.15
 0.018
 0.12
 0.033
 1.2
 0.00017
 0.00075
 0.00045
 0.00030
 0.00075
 0.12
13
 0.096
 0.0060
 9.4
 0.12
 0.048
 0.29
 0.10
 0.36
 0.043
 0.29
 0.12
 3.6
 0.00032
 0.0014
 0.00086
 0.00057
 0.0014
 0.44
46
                                                                       6.0 to 10.5
                                                                       6.0 to 9.0
                                                                       6.0 to 9.0
                                                                       6.0 to 9.0
                                                                       6.0 to 9.0

-------
                 TABLE 2-2,          OP          REOjILSTIONS -
                             BEST &WEL&BEE TEaWOLOGf (BRT)
                                   Effluent Limitations
ka/kkg (or lb/1000 Ib.) of product
Chlor-alkali
Mercury Cells








Oiler-alkali
Diaphragm
Cells





Hydrofluoric
£cid






Sodium
Dichromate




Arsenic (T)
Cadmium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Silver (T)
Zinc (T)
Total Residual
Chlorine
Chromium (T)

Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Total Residual
Chlorine

Fluoride (T)
Jtotimony (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)

Chromium (T)
Hexavalent
Chromium
Nickel (T)
Zinc (T)
0.00021
0.00011
0.00011
0.00034
0.00010
0.00021
0.00015
0.00042
0.00042

0.00044

0.0035
0.0019
0.00088
0.0035
0.0018


1.0
0.023
0.0013
0.0097
0.0020
0.0050
0.017

0.0022
"
0.00035
0.0012
0.0033
0.00046
0.00024
0.00024
0.00074
0.00022
0.00046
0.00032
0.00092
0.00071

0.00097

0.00077
0.0042
0.0019
0.0077
0.0030


2.2
0.047
0.0027
0.019
0.0040
0.010
0.035

0.0045

0.00070
0.0024
0.0066
(continued)

-------
TABLE 2-2.  Continued

Subcategory
Titanium
Dioxide
StxLfate
Process







Titanim
Dioxide
Chloride
Process
Titanium
Dioxide
Chloride
Ilmenite
Process





Aluminum
Fluoride

Chrome Pigments






Parameter

Iron (T)

Arsenic (T)
Antimony (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)

Iron (T)
Chrcmium (T)


Iron (T)
Antimony (T)
Arsenic (T)
Cadmium (T)
Chromiurti (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Fluoride (T)
Chromium (T)
Nickel (T)
Antimony (T)
Cadmium (T)
Chrcmium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Effluent Limitations
«lfax
30-day Av§
kg/kkg (or lb/1000

1.2

0.24
0.38
0.070
0.070
0.24
0.14
0.10
0.24

0.25
0.014


0.30
0.096
0.060
0.012
0.012
0.060
0.036
0.024
0.060
0.036
0.00048
0.0020
0.051
0.020
0.12
0.042
0.15
0.018
0.12-
24 -hr
Max
Ib.) of product

4.1

0.46
0.71
0.11
0.13
0.46
0.21
0.18
0.52

0.84
0.027


1.0
0.18
0.11
0.019
0.023
0.11
0.054
0.046
0.013
0.75
0.00096
0.0040
0.12
0.048
0.29
0.10
0.36
0.043
0.29
                      10

-------
            TABLE 2-2.  Continued
                                   Effluent Limitations
Subcategory
Parameter
                                      Max             24 -hr
                                  30-day Avg           Max
kg/kkg (or Ib/lOQQ Ib.) of product
Copper Sulfate Antimony (T)
Arsenic (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel ,(T)
Selenium (T)
Zinc (T)
Hydrogen .Ammonia - N
Cyanide Cyanide (Free)
Cyanide (T)
Total Residual
Chlorine
Nickel Sulfate AntLomony (T)
Chromum (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Sodium COD
Bisulfite Chrcraium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
Sodium COD
Hydrosulfite Zinc (T)
Nickel (T)
Lead (T)
Chromium (T)
0.00038
0.00047
0.000047
0.000047
0.00038
0.000047
O.OOOQ94
0.000094
0.00038
4.3
0.16
0.23

0.011
0.00027
0.000010
0.00027
0.000034
0.00014
0.00027
1.2
0.00017
0.00075
0.00045
0.00030
0.00075
13
, 0.0024
0.00094
0.0014
0.00047
0.00072
0.00089
0.000089
0,000089
0.00072
0.000089
0.00018
0.00018
0'. 00072
12
0.043
0.65

0.031
0.00081
0.000034
0.00081
0.00010
0.00042
0.00080
3.6
0.00032
0.0014
0.00086
0.00057
0.0014
46
0.0046
0.0018
0.0027
0.00087
                                       11

-------
        oanr,E 2-3.   SUMMARY OF PROPOSED EHSJIATIONS -
                     PREHCKEMMENT SESMBBDS FOR EXISTING
                     SOURCES (PSES)

Subcategory
CKLor-alkali
Mercury Cells






CMor-alkali
Diaphragm
Cells



Hydrofluoric
Acid






Sodium Dichro-
mate




Titanium
Dioxide
Sulfate
Brocess




Parameter
Arsenic (T)
Cadmium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Silver
Zinc

Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)

ELuoride (T)
Antimony (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)

Chromium (T)
Hexavalent
Chromium
Nickel (T)
Zinc (T)


Iron (T)
Arsenic (T)
Antimony (T)
Cadmium (T)
Chromium (T)
Copper (T)
Effluent Limitations
Max
j
(mg/1)
0.10
0.050
0.050
0.16
0.048
0.10
0.070
0.20

0.050
0.40
0.22
0.10
0.40

30 •
0.70
0.040
0.29
0.060
0.15
0.52

0.32

0.050
0.17
0.47


2.5
0.50
0.80
0.15
0.14
0.50
30-day
^vg
or (kg/kkg)
0.00021
0.00011
0.00011
0.00034
0.00010
0.00021
0.00015
0.00042

0.00044
0.0035
0.0019
0.00088
0.0035

1.0
0.023
0.0013
0.0097
0.0020
•> 0.0050
0.017

0.0022

0.00035
0.0012
0.0033


1.2
0.24
0.38
0.07
0.07
0.24
24-hr
Max
(mg/1) or
0.22
0.11
0.11
0.35
0.10
0.22
0.15
0.44

0.11
0.88
0.48
0.22
0.88

66
1.4
0.080
0.58
0.12
0.30
1.0

0.64

0.10
0.34
0.94,


8.5
0.95
1.5
0.24
0.27
0.95
(kg/kkg)
0.00046
0.00024
0.00024
0.00074
0.00022
0.00046
0.00032*
0.00092

0.00097
0.0077
0.0042
0.0019
0.0077

2.2
0.047
0.0027
0.019
0.0040
0.010
0.035

0.0045

0.0070
0.0024
0.0066
"

4.1
0.46
0.71
0.11
0.13
0.46
(continued)
                                     12

-------
                 TlffiLE 2-3.  Continued
Subcategory
                                  Effluent Limitations
Parameter
                                   -Max 30-day
                                      Avg
                                (mg/L)  or (kg/kkg)
                                              24-hr
                                              Max
                                        (mg/1) or  (kg/kkg)



Titanium
Dioxide
Chloride
Process
Titanium
Dioxide ,
Chloride
Ilmenite
Process






Aluminum
Fluoride


Chrome
Pigments






Copper
Sulfate





Lead (T)
Nickel (T)
Zinc (T)

Iron (T)
Chromiun (T)


Iron (T)

Antimony (T)
Arsenic (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)

Fluoride (T}
dsrcmium (T)
Nickel (T)

Antimony (T)
Cadmium (T)
Ohrcsnium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)

Copper (T)
Nickel (T)
Arsenic (T)
Selenium (T)
Cadmium (T)
Zinc (T)
0.30
0.20
0.50

2.5
0.14


2.5

0.80
0.50
0.10
0.10
0.50
0.30
0.20
0.50

30
0.040
0.17

0.48
0.19
1.1
0.40
1.4
0.17
1.1

0.40
0.10
0.50
0.10
0.050
0.40
0.14
0.10
0.24

0.25
0.014


0.30

0.096
0.060
0.012
0.012
0.060
0.036
0.024
0.060

0.36
0.00048
0.0020

0.051
0.020
0.12
0.042
0.15
0.018
0.12

0.00038
0.000094
0.00047
0.000094
0.000047
0.00038
0.45
0.37
1.1

8.4
0.27


8.5

1.5
0.95
0.16
0.19
0.95
0.45
0.38
1.1

63
0.080
0.34

1.2
0.46
2.6
0.96
3.4
0.41
2.6

0.76
0.19
0.95
0.19
0.095
0.76
0.21
0.18
0.52

0.84
0.027


1.0

0.18
0.11
0.019
0.023
0.11
0.054
0.046
0.013

0.75
0. 00096
0.0040

0.12
0.048
0.29
0.10
0.36
0.043
0.29

0.00072
0.00018
0.00089
0.00018
0.000089
0.00072
 (continued)
                      13

-------
                 3&ELE 2-3.  Continued
                                 Effluent Limitations
Subcategory
I&rameter
                                   Max 30-day
                                      Avg
                                (mg/1) or  (kg/kkg)
                                             24-hr
                                              Max
                                        (mg/1) or (kg/kkg)



Hydrogen
Cyanide


Nickel Sol-
fate





Sodium Bi~
sulfate





Sodium H^3ro-
sulfite




Ctoamium (T)
Lead (T)
Antimony (T)

Cyanide (Eree)
Cyanide (T)
Ararc>nia-$I "

Antimony (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)

COD
Chromium (T)
Zinc (T)
Copper (T)
Lead (T)
Nickel (T)

COD
Zinc (T)
Nickel (T)
Lead ,{T)
Chromium (T)
0.050
0.05
0.40

0.27
4.0
75

0.40
0.050
0.40
0.050
0.20
0.40

680
0.11
0.50
0.50
0.30
0.20

2700
0.50
0.20
0.30
0.10
0.000047
0.000047
0.00038

0.016
0.23
4.3

0.00027
0.00010
0.00027
0.000034
0.00014
0.00027

1.2
0.00017
, 0.00075
0.00075
0.00045
0.003

13
0.0024
0.00094
0.0014
0.00047
0.095
0.095
0.76

0.74
11
210

1.2
0.15
1.2
0.15
0.60
1.2

2400
0.22
d.O
1.0
0.57
0.38

9700
0.95
0.38
0.57
0.19
0.00089
0.000089
0.00072

0.043
0.65
12

0.00081
0.000034
0.00081
0.00010
0.00042
0.00080

3.6
0.00032
0.0014
0.0014
0.00086
0. 00057

46
0.0046
0.0018
0.0027
0.00089

-------
                TABLE 2-4.   SUMmRY OP PROPOSED BEGUIATIONS -
                             NEW SOURCE PERFORMANCE STANDARDS
                             (NSPS)
Subcategory
Chlor-alkali
Mercury Cells










Chlor-alkali
Diaphragm Cells




Hydrofluoric
Acid





Sodium Dichro-
mate

Parameter
TSS
Arsenic (T)
Cadmium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Silver
Zinc
Total Residual
Chlorine
pH
TSS
Chromium (T)
Lead (T)
Total Residual
Chlorine
PH

'TSS
Fluoride (T)
Chromium (T)
Nickel (T)
Zinc (T)
PH

TSS
Chromium (T)
Effluent Limitations
Max
30-day Avg
kg/kkg (or lb/1000
0.32
0.00021
o.oooii
1 0.00011
0.00034
0.00010
0.00021
0.00015
0.00042
0.00042


0.10
0.00047
0.00044
0.0018

*

0.41
0.18 •
0.00024
0.0009
0.0030


0.18
0.0022
24 -hr
Max
Ib. ) of product
0.64
0.00046
0.00024
0.00024
0.00074
0.00022
0.00046
0.00032
0.00092
0.00071


0.20
0.00097
0.00097
0.0030



0.86
0.38
0.00048
0.0018
0.0060


0.35
0.0045
pH Range











6.0 to 9.0





6.0 to 9.0






6.0 to 9.0



Hexavalent Chrom- ,




ium
Nickel (T)
Zinc (T)
pa
0.00035
0.0012
0.0033

0.00070
0.0024
0.0066




6.0 to 9.0
(continued)
                                       15

-------
2-4.  Continued
             Effluent Limitations
kg/Wcg (or lb/1000 Ib.) of product
Titanitm , Diox-
ide (Sulfate
toocess)









Titanitm Diox-
ide ' (Chloride
Srocess)


Titanium Diox-
ide (Chloride
Hmenite pro-
cess)








Muminum
Fluoride





TSS
Iron (T)
Arsenic (T)
Antimony (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
pH

TSS
Iron (T)
Chromium (T)
PH
*
TSS
Icon (T)
Antimony (T)
Arsenic (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
FH

TSS
ELuoride (T)
Chromium (T)
Nickel (T)
pH-

30
1.2
0.24
0.38
0.07
0.07
0.24
0.14
0.10
0.24


4.5
0.18
0.005


1.2
0.050
0.025
0.016
0.0023
0.0012
0.0090
0.0019
O.Q053
0.015


0.81
0.36
0.00050
0.0020


110
4.1
0.46
0.71
0.11
0.13
0.46
0.21
0.18
0.52


16
0.59
0.01


4.3
0.17
0.048
0.030
0.0037
0.0023
0.017
0.0029
0.010
0.032


1.7
0.75
0.0010
0,0040












6.0 to 9.0




6,0 to 9.0











6,0 to 9.0





6.0 to 9.0
                16

-------
                       2-4.  Continued
Subcategory
                                   Effluent Limitations
Parameter
                                     Max
                                  30-day Avg
                                      24 -hr
                                       Wax
pH Range
kg/kkg (or lb/1000 Ib.) of product
Chrome Pigments









Copper Sulfate










Hydrogen Cyan-
ide






Nickel Sulfate







TSS
• Antimony (T) •
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Zinc (T)
PH ;
TSS
Copper (T)
Nickel (T)
Arsenic (T)
Selenium (T)
Cadmium (T)
Zinc (T)
Chromium (T)
Lead (T)
Antimony (T)
PH

TSS
Cyanide (Free)
Cyanide (T)
Ammonia-N
total Residual
Chlorine
pH
TSS
Antimony (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)
PH
3.9
0.042
0.0011
0.0053
0.0053
0.0053'
0.0011
0.0053
0.0021

0.023
0.00038
0.000094
0.00047
0.000094
0.000047
0.00038
0.000047
0.000047
0.00038


2.0
0,016
0.23
4.3

0.011

0.032
0.00027
0.000010
0.00027
0.000034
0.00014
0.00027

9.4
0.10
0.0026
0.013
0.013
0.013
0.0026
0.013
0.0050
6.0 to 9.0
0.069
0.00072
0.00018
0.00089
0.00018
0.000089
0.00072
0.000089
0.000089
0.00072
6.0 to 9.0

5.4
0.043
0.65
12

0.031
6.0 to 10.5
0.096
0.00081
0.000034
0.00081
0.00010
0.00042
0. 00080
6.0 to 9.0
 (continued)
                                     17

-------
                    TftBUS 2-4.  Continued
Subcategory
 Parameter
                                    Effluent Limitations
                                      Max
                                  30-day Avg
                                       24-hr
                                        Max
                                kg/kkg  (or lb/1000 lb.)  of product
                                    pH Range
Sodium Bisul-
 fite
Sodium Hydro-
 sulfite
                 TSS
                 ODD
                 Qiranium  (T)
                 Zinc (T)
                 Copper  (T)
                 L^d (T)
                 Nickel  (T)
                 PH
TSS
K30D
Chranium (T)
Lead  (T)
Nickel  (T)
Zinc
pH
                   0.033
                   1.2
                   0.00017
                   0.00075
                   0.00075
                   0.00045
                   0.00030
 0.18
13
 0.00047
 0.0014
 0.00094
 0.0024
                     0.12
                     3.6
                     0.00032
                     0.0014
                     0.0014
                     0.00086
                     0.00057
 0.44
46
 0.00089
 0.0027
 0.0018
 0.0046
                                                       6.0 to  9.0
                                                                       6.0 to 9.0
                                      18

-------
              TABLE 2-5.  SUM^EY OF PEDPOSED REGULATIONS -
                                       STANDARDS FOR NEW SOURCES
                           (PSNS)
                                Effluent Limitations
buDcacegory jraiamecer 	 • — 	
, 30-day Avg „ , t
(rng/L) or (Rg/kkg)
Chlor-alkali
Mercury tells






Chlor-alkali
Diaphragm
Cells
Hydrofluoric
Acid



Sodium Di-
chromate




Titanium Di-
oxide (sul-
fate pro-
cess)






Titanium
Dioxide
(chloride
process)
Arsenic (T)
Cadmium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Silver
Zinc
Chromium (T)

Lead (T)

Fluoride (T)
Chromium (T)
Nickel (T)
Zinc (T)

Chromium (T)
Hexavalent
Chromium
Nickel (T)
Zinc (T)

Iron (T)
Arsenic; (T)
Antimony (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Zinc (T)

Iron (T)
Chromium (T)

0.10
0.050
0.050
0.16
0.048
0.10
0.070
, 0.20
0.050

0.050

30
0.040
0.15
0.50

- 0.32

0.050
0.17
0.47

2.5
0.50
0.80
0.15
0.14
0.50
0.30
0.20
0.50

1.8
0.05

0.00021
0. 00011
0.00011
0.00034
0.00010
0.00021
0.00015
0.00042
• 0.00044

0.00044

0.18
0.00024
0.00090
0.0030

0.0022

0,00035
0.0012
0.0033

1.2
0.24
0.38
0.07
0.07
0.24
0-.14
0.10
0.24

0.18
0.005

24 -hr
Max
(mg/L) or
0.22
0.11
0.11
0.35
0.10
0.22
0.15
0.44
0.11

0.11

63
0.080
0.30
1.0

0.64

0.10
0.34
0.94

8.5
0.95
1.5
0.24
0.27
0.95
0.45
0.37
1.1

5.9
0.10

(kg/kkg)
0.00046
0.00024
0.00024
0.00074
0.00022
0.00046
0.00032
0.00092
0.00097

0.00097

0.38
0. 00048
0.0018
0.0060

0.0045

0.00070
0.0024
0.0066

4.1
0.46
0.71
0.11
0.13
0.46
0.21
0.18
0.52

0,59
0.01

(continued)
                                    19

-------
                ISELE 2-5.  Continued

Subcategory
Titanium
Dioxide
Chloride
Ilmenite
Process






Aluminum
Fluoride
t

Chrome
Pigments


*




Copper
Sulfate








Hydrogeij
Cyanide


Parameter

Iron (T).

Antimony (T).
Arsenic Crl
Cadmium (3?)
Chromium (T)
Copper (T) •
Lead (T)
Nickel (T)
Zinc (T)

Fluoride (T)
Chromium (T)
Nickel (T)

Antimony (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Mercury (T)
Nickel (T)
Zinc (T)

Antimony (T)
Arsenic (T)
Cadmium (T)
Chromium (T)
Copper (T)
Lead (T)
Nickel (T)
Selenium (T)
Zinc' (T)

Atimonia - N
Cyanide (Free)
Cyanide (T)
Effluent Limitations
30-
(mg/1)

1.6

0.80
0.50
0.075
0.040
0.029
0.060
0.17
0.47

30
0.04
0.17

0.40
0.010
0.050
0.050
0.050
0.010
0.050
0.020

0.40
0.50
0.050
0.050
0.40
0.050
0.10
0.10
0.40

75 ,
0.27
4.0
Max
•day Avg
or (kg/kkg)

0,050

Q..025
0*016
0.0023
0.0012
0.0090
0.0019
0.0053
0.015

0.36
0.00050
0.0020

0.042
0.0011
0.0053
0.0053
0.0053
0.0011
0.0053
0.0021

0.00038
0.00047
0.000047
0.000047
0.00038
0.000047
0.000094
0.000094
0.00038

4.3
0.016
0.23
24-hr
Max
(mg/L) or (kg/kkg)

5,4

1.5
0.95
0.12
0.076
0.055
0.090
0.32
0.99

63
0.08
0.34

0.96
0.024
0.12
' 0.12
0.12
0.024
0.12
0.048

0.76
0.95
0.095
0.095
0.76
0.095
0.19
0.19
0.76

210
0.74
11

tt, 017

-0.048
Q,C130
0,0037
0.0023
0.017
0.0029
0.010
0.032

0.75
0.0010
0.0040

0.10
0.0026
0.013
0.013
0.013
0.0026
0.013
0.0050

0.00072
0.00089
0.000089
0.000089
0.00072
0.000089
0.00018
0.00018
0.00072

12
0.043
0.65
(continued)
                                       20

-------
               TABLE 2-5.  Continued
                                 Effluent Limitations
Subcategory    Parameter
                                      Max
                                  30-day Avg
                                (mg/1) or  (kg/kkg)
      24-hr
       Max
(rog/1)  or (kg/kkg)
Nickel
Sulfate





Sodium
Bisulfite





Sodium
Hydrosulfite





Antimony (T)
Chromium (T).
Copper (,Tl
Lead (T)
Nickel (T)
Zinc (T)

COD
Chromum (T)
Zinc (T)
Copper (T)
Lead (T)
Nickel (T)

COD
Zinc (T)
Nickel (T)
Lead (T)
Chromium (T)

0.40
0.05
0.40-
0.05
0.20
0.40

680
0.11
0.5
0.5
0.3
0.2

2700
0.50
0.20
0.30
0,10

Q. 00027

1,2
0,000010 6,15
Q.QQQ27
0.000034
0.00014
0.00027

, 1.2
0.00017
0.00075
0.00075
0.00045
0.00030

13
0.0024
0.00094
0.0014
0.00047
1,2
0.15
0.60
1.2

2400
0.22
1.0
1.0
0.57
0.38

9700
0.95
0.38
0.57
0.19

0.00081
0,00,0034
O..QQ081
0.00010
0.00042
0.00080

3.6
0.00032
0.0014
0,0014
0.00086
0.00057

46
0.0046
0.0018
0.0027
0.00089
                                      21

-------
                 TKW..E 2-6.  SDMgaKY OF PROPOSED REGDLATJDNS -
                             EEST OTNVENTICNMj PCUinSNT COKHTOL
                             TESCHNOLOGY  (BCT)
Subcategory
                                    Effluent Limitations
 Parameter
                                   Max
                                30-day Avg
                                       24 -far
                                        Max
                                kg/Wcg (or lb/1000 Ib.) of poxiduct
                                   pH Range
Chlor-alkali
 Diaphragm Cell  TSS
                 pH
TSS
pH
 Jteid
                   0.36
2.3
                    0.72
                                                       4.8
                                                     6.0 to 9.0
                                                                      6.0 to 9.0
                                       22

-------
                           SECTION 3


                          INTRODUCTION



3.1  AUTHORITY


3.1.1  The Federal Water Pollution Control Act Amendments

     The  Federal   Water  Pollution  Control  Act   (the  Act)
Amendments of  1972, 33  USC  1251 et  seq.,  stated  the .national
goal  of  attaining  by  July  1,  1983,  a   water quality  which
provides  for  the  protection  and propagation   of    fish  and
shellfish, for recreation in or on the nation's waters, and the
goal of  eliminating the discharge of pollutants into navigable
waters by  1985.

Purpose and Authority

     The Federal Water pollution Control Act Amendments of 1972
established a comprehensive program to "restore and maintain the
chemical,  physical, and biological  integrity of the  Nation's
waters,"  Section 101(a).  By July  1,  1977, existing industrial
dischargers  were  required  to   achieve   "effluent  limitations
requiring  the  application  of   the   best  practicable  control
technology currently available"  ("BPT"),  Section 301(b) (1) (A);
and by July 1, 1983, these dischargers were,required to achieve
"effluent  limitations   requiring  the  application  of  the  best
available  technology   economically   achievable...which   will
result in  reasonable further progress toward the national goal
of eliminating the discharge  of all pollutants" ("BAT"), Section
301(b)(2)(A).  New industrial  direct dischargers were required
to comply  with  Section 306  new source  performance  standards
("NSPS"), based  on best available demonstrated technology?  and
new and existing dischargers to  publicly  owned treatment works
("POTW")  were  subject  to pretreatment standards under Sections
307 (b) and (c)  of  the  Act.  While  the requirements  for direct
dischargers  were to be incorporated  into  National  Pollutants
Discharge  Elimination  System   (NPDES)  permits  issued  under
Section  402   of  the  Act,   pretreatment  standards  were  made
enforceable  directly  against dischargers  to  POTW  (indirect
dischargers).

                               23

-------
     Although Section  402(a)(1)  of the 1972 Act authorized the
setting of requirements for direct  dischargers on a case-by-case
basis,  Congress  intended  that   for  the  most  part  control
requirements would  be  based on  regulations  promulgated by the
Administrator of  EPA.   Section  304 (b) of  the  Act required the
Administrator  to promulgate  regulations  providing  guidelines
for  effluent  limitations  setting  forth  the  degree of  effluent
reduction  attainable  through the  application  of BPT  and BAT.
Moreover,  Sections   304 (c)  and  306   of  the  Act   required
promulgation  of  regulations  for  NSPS,   and Sections  304(f),
307(b),  and  307 (c)  required promulgation of  regulations for
pretreatment standards.   In addition  to these  regulations for
designated  industry  categories,  Section  307(a)  of   the  Act
required  the Administrator to  promulgate  effluent standards
applicable to  all  dischargers  of toxic  pollutants.   Finally,
Section  501(a)   of  the  Act  authorized  the Administrator  to
prescribe any additional regulations "-necessary to carry out his
functions" under  the Act.

     The EPA was  unable to  promulgate many of these  regulations
by  the  dates  contained in  the Act.   In 1976, EPA was  sued by
several environmental groups, and  in settlement of this lawsuit
EPA  and the plaintiffs  executed  a "Settlement  Agreement" which
was  approved  by  the  Court.   This  Agreement  required EPA to
develop a program and adhere to  a  schedule for promulgating BAT
effluent  limitations   guidelines,  pretreatment  standards,  and
new  source performance standards  for  65 "priority"  pollutants
and  classes of pollutants  for 21 major industries.   See Natural
Resources Defense Council, Inc.  v. Train, 8 ERG   2120  (D.D.C.
1976), modified March 9, 1979.

     On December  27,  1977,  the  President signed  into  law the
Clean  Water  Act  of 1977.  Although  this  law makes  several
important  changes  in  the Federal  water  pollution  control
program,  its most significant feature is  its  incorporation of
several  of  the  basic  elements  of  the  Settlement Agreement
program for toxic pollution control.  Sections 301 (b) (2) (A) and
301(b) (2) (C)  of  the Act now require the achievement by July 1,
1984  of  effluent limitations requiring  application  of  BAT for
"toxic" pollutants,  including the 65 "priority" pollutants and
classes of  pollutants  which   Congress  declared "toxic" under
Section 307(a)  of the Act.   Likewise,  EPA's  programs for new
source performance standards and pretreatment standards are now
aimed  principally  at  toxic pollutant  controls.  Moreover, to
strengthen the toxics control program Section 304 (e)  of the Act
authorizes  the  Administrator  to  prescribe "best  management
practices"  ("BMPs")  to   prevent  the   release  of  toxic  and
hazardous pollutants  from  plant site runoff,spillage or  leaks,
sludge  or waste   disposal, and  drainage  from  raw   material
storage associated  with,  or ancillary to, the manufacturing or
treatment process.
                               24

-------
     In keeping with its emphasis on toxic pollutants,  the Clean
Water Act of 1977 also revises  the control program for  non-toxic
pollutants.    Instead  of  BAT  for  "conventional"  pollutants
identified  under   Section   304(a)(4)  (including  biochemical
oxygen demand, suspended solids, fecal coliform and  pH),  the new
Section  301(b)(2)(E)  requires  achievement by  July  1,  1984,  of
"effluent  limitations  requiring the  application of   the  best
conventional    pollutant  control  technology"  ("BCT").    The
factors considered  in assessing BCT for an industry include the
cost  of  attaining  a  reduction  in effluents  and  the  effluent
reduction  benefits  derived  compared  to the  costs  and effluent
reduction  benefits   from   the  discharge  of  publicly  owned
treatment   works   (Section   304(b) (4) (B).     For   non-toxic,
nonconventional pollutants, Sections  301(b)(2)(A) and  (b) (2) (F)
require  achievement of  BAT effluent limitations  within three
years after  their  establishment or July  1,  1984,  whichever  is
later, but not later than July lr  1987.

     The  purpose of these  proposed regulations is  to provide
effluent  limitations  guidelines for BPTr  BAT, and  BCT, and to
establish  NSPS,  pretreatment  standards  for  existing sources
(PSES), and pretreatment standards for new sources  (PSNS), under
Sections  301, 304,  306,  307, and 501  of the Clean Water Act.

     The  United States  Environmental  Protection  Agency  (the
Agency) was entrusted with  the responsibility to carry out the
requirements  of  the Act, and  initiated  an intensive effort to
develop the necessary regulatory means  which would achieve the
stepwise   reduction  and  elimination  of  pollutant   discharge
practices  in all  major  U.S.  Industries.   For  the   Inorganic
Chemicals  Manufacturing  Point  Source   Category,  the  Agency
designed  a comprehensive,  two phase  program  to  identify  the
control  parameters and  establish the  technological  basis  for
regulations  development.  Phase  I covered  22 Major   Inorganic
Chemical  Products  (1) ,  and  the  fjinal   regulations  for these
industrial subcategories were published in the Federal Register
on March  12,  1974.  The  regulations included specific  numerical
effluent  limitations  and •standards  of  performance   for  both
existing   and  new  sources.       Zero-discharge  requirements
specified  for many  of   the  subcategories were to be applied
either at  the 1977 BPT step or later.  Phase II of the Agency's
effort  resulted  in  the promulgation  of BPT  based  effluent
limitations for an additional group of 27  subcategories referred
to as Significant Inorganic Chemical  Products  (2).  The  interim
final  regulations  were  published on  May  22,  1975.   Taken
together,  the two  groups  of  regulations  cover 49   inorganic
chemical  subcategories  many  of  which include more  than  one
specific  chemical  product.    Although  some   toxic   pollutant
parameters were covered  in  cases where a  direct relationship to
the process was obvious  (e.g.,  mercury and/or lead in the Chlor-
Alkali Industry),  the'main  thrust of the regulations was  the
                               25

-------
control of  the bulk  pollutant parameters which  accounted,  in
terms  of   quantity,  .for  most of  fche  pollution  loading  of
navigable waters  attributable to the  manufacture  of inorganic
chemicals.

3.1,2  Court Remand of Regulations

     On March  10,  1976, the  United  States  Court of Appeals for
the Fourth Circuit decided in E.I.  duPont de Nemours & Company,
et al. v.  Train, 541 F.2d  1018 (4th  Cir. 1976),  to set aside and
remand for reconsideration a  number of general  definitions and
specific  -discharge  regulations  promulgated  in  1974.    These
regulations are all  within Title  40,  Parts 401 and  415  of the
Code of Federal Regulations and are listed below:

     General Provisions
         401,11 (i)  -  Definition of effluent limitations
         401.11 (g)  -  Definition of process waste water
         401.11 (r)  ~  Definition of process waste water
                           pollutant


     Chlor-Alkali
         415.63      - BATEA


     Hydrochloric Acid
         415.72      -  BPCTCA
         415.73      -  BATEA
         415.75      -  New sources
     Hydrofluoric Acid
         415.82      -  BPCTCA
         415.83      -  BATEA
         415.85      -  New sources
     Hydrogen Peroxide
         415.93      -  BATEA
         415.95      -  New sources
     Nitric Acid
         415.102     -  BPCTCA
         415.103     -  BATEA
         415.105     -  New sources
                               26

-------
     Sodium Carbonate
         415.152    -  BPCTCA
         415.153    -  BATEA
         415.155    -  New sources
     Sodium Bichromate
         415.173    -  BATEA
     Sodium Metal
         415.182    -  BPCTCA
         415.183    -  BATEA
         415.185    -  New sources
     Sodium Silicate
         415.192    -  BPCTCA
         415.193    -  BATEA
         415.195    -  New sources
     Sulfuric Acid
         415.210    -  Applicability
         415.212    -  BPCTCA
         415.213    -  BATEA
         415.215    -  New sources


     Titanium Dioxide
         415.220    -  Applicability
         415.222    -  BPCTCA
         415.223    -  BATEA
         415.225    -  New sources
                                     , -,,

     For the  most  part,  the main target of  the  remand  was the
zero discharge regulations from  which  the  industry petitioners
sought relief on grounds  of technological infeasibility.  During
1975, the Agency funded a special study of the remand issues (3)
and  was  prepared  to  propose  amended  regulations.    Where
appropriate,  the  results  of  that   study  are  included in  an
Addendum  to   the   present  report  covering   those  remanded
regulations for subcategories which have been excluded from the
present study.

     Following   the   court  remand  of   the  Phase  I  final
regulations, the Agency  revoked  the  Phase II  interim final and
proposed  regulations  published  in  May,  1975,  for Aluminum
Fluoride,  Chrome   Pigments,   Hydrogen  Cyanide,   and   Sodium
Silicofluoride.  -In  €his instance,  the Agency's  intent was to
reconsider  the specific  effluent limitations established  for

                               27

-------
these  industries  (1977  step) in  the  light of information made
available on process  differences  between plants and additional
data  on  the   actual  concentrations  and  treatability of  the
regulated discharge constituents.  The information was presented
to  the Agency  in the  form of various documents  prepared  by
members  of  the industries  concerned.   These  sources  are also
cited  in the appropriate sections of this  report.

3.1.3  The Settlement Agreement

     A  consent  decree  was issued  in  a  suit  filed  by fdur
environmental  groups  in Natural  Resources Defense  Council  v.
Train, 8 EEC ,2120  (June 8,  1976) modified  12 IRC 1833  (December
15, 1978) .  The consent decree  contained a Settlement Agreement
wherein the Agency agreed to regulate 65 toxic pollutants under
Sections 301,  304,  306, and 307 of the Act  in accordance with
the schedule and provisions stipulated.  The  original list of 65
chemicals and  classes of chemicals attached  to the Settlement
Agreement  was  redefined  to  cover   129   chemical   substances,
including  specific  organic  compounds, pesticides  and  their
metabolites,  polychlorinated'  biphenyls  (PBC's),  cyanide,  13
heavy  metals   and  asbestos.   Table  3-1  lists  the  129  toxic
pollutants (sometimes re'ferred to in the literature as "priority
pollutants11) .

        TABU!  3-1.   RECOMMENDED IiIST OF TOXIC POLLUTANTS

Compound Name

   1.  *Acenaphthene
   2.  *Acrolein
   3.  *Acrylonitrlle
   4,  *Benzene
   5.  *Benzidine
   6.  *Carbon  tetrachloride (tetrachloromethane)


       *Chlorinated benzenes (other than dichlorobenzenes)


   7.       *Chlorobenzene
   8.       1,2,4-Trichlorobenzene
   9.       Hexachlorobenzene

       *Chlorinated ethanes  (including  1,2-dichloroethane,
            1,1,1,-trichloroethane and  hexachloroethane)


  10.       1.2-Dichloroethane
  11.       1,1,1-Trichloroethane
                               28

-------
12.       Hexachloroethane
13.       1,1-Diehloroethane
14.       1,1,2-Trichloroethane
15.       1,1,2,2-Tetrachloroethane
16.       Chloroethane
    *Chloroalkyl ethers (ehloromethylr chloroethyl and
          mixed ethers)
17.       Bis(chloromethyl) ether
18.       Bis(2-chloroethyl) ether
19.       2-Chloroethyl vinyl ether  (mixed)
    *Chlorinated naphthalene


20.       2-Chloronaphthalene
    *Chlorinated phenols  (other than those listed
          elsewhere? includes trichlorophenols
          and chlorinated cresols)
21.       2,4,6-Trichloro.phenol
22.       Parachlorometa cresol
23.       *Chloroform  (trichloromethane)
24.       *2-Chlorophenol
     *Dichlorobenzenes
25.       Ir2-Dichlorobenzene
26.       Ir3-Dichlorobenzene
27.       1,4~Dichlorobenzene
      *Dichlorobenzidine
28.       3r3'-Dichlorobenzidine
                             29

-------
    *Dichloroethylenes (lrl-<3ichloroethylene and
          1,2-dichloroethylene)
29.       1,1-Dichloroethylene
30.       Ir2-Trans~dichloroethylene
31.       *2r4-Diehlorophenol
     *Dichloropropane and dichloropropene
32.       1,2-Dichloropropane
33.       1,2-Dichloropropylene (1,3-dichloropropene)
34.       *2,4-Dimethylphenol
     *Dinitrotoluene
35.       2,4-Dinitrotoluene
36.       2r6-Dinitrotoluene
37.       *lr2-Diphenylhydrazine
38.       *Ethylbenzene
39.       *Fluoranthene
     *Haloethers (others than those listed elsewhere)
40.       4-Chlorophenyl phenyl ether
41.       4-Bromophenyl phenyl ether
42.       Bis(2-chloroisopropyl)  ether
43.       Bis(2-chloroethoxy) methane
     *Halomethanes (other than those listed elsewhere)
44.       Methylene chloride (dichloromethane)
45.       Methyl chloride (chloromethane)
46.       Methyl bromide (bromomethane)
47.       Bromoform (tribromomethane)
48.       Dichlorobromomethane
49.       Trichlorofluoromethane
50.       Dichlorodifluoromethane
51.       Chlorodibromomethane
52.       *Hexachlorobutadiene
53.       *Hexachlorocyclopentadiene
54.       *Isophorone
55.       *Naphthalene
56.       *Nitrobenzene

                             30

-------
      *Nitrophenols (including 2,4-dinitrophenol and
           and dinitrocresol)
57.        2-Nitrophenol
58.        4-Nitrophenol
59.        2r4-Dinitrophenol
60.        4,6-Dinitro-o-cresol
       *Nitrosamines
61.        N-nitrosodimethylamine
62.        N-nitrosodiphenylamine
63.        N-nitrosodi-n-propylamine
64.        *Pentachlorophenol
65.        *Phenol
       *Phthalate esters
66.        Bis(2-ethylhexyl)  phthalate
67.        Butyl benzyl phthalate
68.        Di-n-butyl phthalate
69.        Di-n-octyl phthalate
70.        Diethyl phthalate
71.        Dimethyl phthalate


       *P_plynu clear aromatic hydrocarbons


72.        Benzo(a)anthracene (1,2-benzanthracene)
73.        Benzo (a)  pyrene (3,4-bejizopyrene)
74.        3,4-Benzofluoranthene
75.        Benzo(k)fluoranthane (11,12-benzofluoranthene)
76.        Chrysene
77.        Acenaphthylene
78.        Anthracene
79.        Benzo(ghi)perylene (1,12-benzoperylene)
80.        Fluorene
81.        Phenanthrene
82.        Dibenzo (a,h)anthracene (I,2r5r6-dibenzanthracene)
83.        Indeno  (lr2,3-cd)pyrene (2,3,-o-phenylenepyrene)
84.        Pyrene
85.        *Tetrachloroethylene
86.        *Toluene
87.        *Trichloroethylene
88.        *Vinyl chloride (chlorethylene)


                             31

-------
     *Pesticides and metabolites
 89.        *Aldrin
 90.        *Dieldrin
 91.        *Chlordane (technical mixture & metabolites)
     DDT and metabolites
 92.        4,4'-DDT
 93.        4f4'-DDE (pfp'-DDX)
 94.        4,4'ODD (p,p'-TDE)
     *Endosulfan and metabolites
 95.        A-endosulfan-Alpha
 96.        B-endosulfan-Beta
 97.          Endosulfan sulfate
     *Endrin and metabolites
 98.        Endrin
 99.        Endrin aldehyde
     *Heptachlor and metabolites
100.        Heptachlor
101.        Heptachlor epoxide
     *Hexac'hlorocyclohexane (all isomers)
102.        A-BHC-Alpha
10'3.        B-BHC-Beta
104.        R-BHC (lindane)-Gamma
105.        G-BHC-Delta
     *Polychlorinated biphenyls (PCB's)
106.        PCB-1242 (Arochlor 1242)
107.        PCB-1254 (Arochlor 1254)

                             32

-------
 108.        PCB-1221  (Arochlor 1221)
 109.        PCB-1232  (Arochlor 1232)  '
 110.        PCB-1248  {Arochlor 1248)
 111.        PCB-1260  (Arochlor 1260)
 112.        PCB-1016  (Arochlor 1016)
 113.        *Toxaphene
 114.        *Antimony  (Total)
 115.        *Arsenic  (Total)
 116.        *Asbestos  (Fibrous)
 117.        *Beryllium (Total)
 118.        *Cadmium  (Total)
 119.        *Chromium  (Total)
 120.        *Copper (Total)
 121.        *Cyanide  (Total)
 122.        *Lead  (Total)
 123.        *Mercury  (Total)
 124.        *Nickel (Total)
 125.        *Selenium  (Total)
 126.        *Silver (Total)
 127.        *Thallium  (total)
 128.        *Zinc  (Total)
 129.        **2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)

 *Specific  compounds  and  chemicals  classes  as  listed  in  the
Consent Decree.

 **This compound was specifically listed in the Consent Decree.
Because of the extreme toxicity of TCDD, the Agency recommended
that  laboratories   not  acquire analytical  standards  for  this
compound.  Categories and specified the scope of application of
effluent  limitations,   new  source performance standards,  and
pretreatment  standards within  each  category  in  terms  of  the
Standard Industrial Classification (SIC)  code numbers.  For the
Inorganic Chemicals Manufacturing Point  Source Category,  the
major industries included are:

           SIC 2812  - Alkalies and Chlorine
           SIC 2813  - Industrial Gases
           SIC 2816  - Inorganic Pigments
           SIC 2819  - Industrial Inorganic Chemicals,
                         Not Elsewhere Classified

     Within  these   industries,  the  Agency  has  identified  63
subcategories listed in Table 3-2 for  the  initial  study  of  the
toxic pollutant  problem.   Most of  these subcategories, 49  in
all,  had  already  been  covered  by  BPT  and  BAT  discharge
regulations promulgated  in 1974  and 1975.   Those regulations
established  point   of  discharge   control   levels   for   the
conventional parameters such  as pH,  TSS, TOC,  BOD, and oil  and
grease.   In  many  cases,  specific  chemical  parameters  were
regulated,  particularly  Arsenic,  Chromium,  Copper,  Mercury,
                               33

-------
   TABLE  3-2.   SCOPE  OF  INDUSTRY  COVERAGE  WITHIN  THE  INORGANIC
          CHEMICALS MANUFACTURING POINT  SOURCE  CATEGORY

            Subcategories Designated  for  Initial Study
  1.  Chlor-Alkali                  33.
  2.  Hydrofluoric  Acid             34.
  3.  Hydrogen  Peroxide             35.
  4.  Titanium  Dioxide              36.
  5.  Aluminum  Fluoride             37.
  6.  Chrome  Pigments               38.
  7.  Hydrogen  Cyanide              39.
  8,  Sodium  Bichromate             40.
  9.  Carbon  Dioxide                41.
 10.  Carbon  Monoxide/Hydrogen      42.
 11.  Copper  Sulfate                43.
 12.  Nickel  Sulfate                44.
 13.  Silver  Nitrate                45.
 14.  Sodium  Bisulfite              46.
 15.  Sodium  Hydrosulfite           47.
 16.  Hydrochloric  Acid             48.
 17.  Nitric  Acid                   49.
 18.  Sodium  Carbonate              50.
 19.  Sodium  Metal                  51.
 20.  Sodium  Silicate               52.
 21.  Sulfuric  Acid                53.
 22.  Ammonium  Chloride             54.
 23.  Ammonium  Hydroxide            55.
 24.  Barium  Carbonate              56.
 25.  Boric Acid                    57.
 26.  Calcium Carbonate             58.
 27.  Copper  Oxide                  59.
 28.  Manganese Sulfate             60.
•29.  Strong  Nitric Acid            61.
 30.  Oxygen  and Nitrogen           62.
 31.  Potassium Iodide              63.
 32.  Sodium  Hydrosulfide
Sodium Silicofluoride
Sodium Thiosulfate
Sulfur Dioxide
Bromine
Calcium Hydroxide
Chromic Acid
Fluorine
Hydrogen
Iodine
Potassium Chloride
Stannic Oxide
Zinc Sulfate
Calcium Carbide
Calcium Oxide
Potassium Metal
Potassium Sulfate
Sodium Bicarbonate
Borax
Ferric Chloride
Lead Monoxide
Sodium Fluoride
Aluminum Chloride
Aluminum Sulfate
Potassium Dichromate
Calcium Chloride
Sodium Chloride'
Sodium Sulfite
Pdtassium Permanganate
Zinc Oxide
Lithium Carbonate
Ferrous Sulfate
                               34

-------
Nickelr  Lead,  Selenium,  Zinc,  and  Cyanide,  which  are  now
included  in  the  list  of  toxic pollutants.   Other  regulated
parameters such as Al,  Bar  Fe,  ammonia, fluoride and sulfide are
not presently listed as toxic chemicals but  are to be treated as
nonconventional  pollutants under  future  discharge limitations
and standards of performance.1

     Nearly  half  of  the  initial 63  subcategories have  been
recommended  for  exclusion from  this  study  on  the  basis  of
specific provisions for such exclusion under Paragraph 8 of the
Settlement Agreement.   The  bases  for these exclusions  are  as
follows:

     No. 63,  Ferrous Sulfate, is already covered by the Titanium
     Dioxide - Sulfate Process subcategory and does not require
     separate consideration.

     No's. 60,  61,  and 62  (Potassium Permanganate, Zinc Oxide,
     and Lithium Carbonate)  have  only one  plant  each  (or  one
     plant  with   a wet  process  discharge),   and  represent
     nonsignificant discharges  of  toxic pollutants.  No's.  27
     and 28 (Copper Oxide and Manganese Sulfate) are also single
     plants,  but were covered in screening.

     No's. 36  through  59  have  existing  BPT or  BAT regulations
     requiring  zero  discharge  of   process  waste  water   to
     navigable  water  and  there are  no  known discharges to a
     POTW.  Continued  enforcement of  the existing regulations
     will provide adequate control of  toxic pollutants.

     The  remaining 35  nonexcluded  subcategories  (Table  3-2,
No's. 1 through 35) are covered in this report.  This group also
includes  the   11  subcategories  whose  final  regulations  were
remanded for  restudy in E.I. duPont  de 'Nemours  and Company,  et
al. v. Train,  supra, and the  four additional subcategories whose
interim, final or proposed regulations were  revoked and reserved
by the Agency.

     It was anticipated by the Agency  that a substantial number
of  the 35 industries  to  be screened  would  also  qualify  for
exclusion  under Paragraph 8  on  the  basis of the  analytical
results obtained  from  the process waste  water  toxic  pollutant
screening program.  A preliminary prioritization indicated that
the  initial  detailed  study and  regulation development  would
focus on the first  15 subcategories.

     This  judgment has been  substantially  supported  by  the
analytical results of   the  screening programs  and a  number  of
additional exclusions  are  being   recommended  for  subcategories
in  which   nonsignificant  •toxic  pollutant discharges  have
been   determined.   A  detailed presentation  of  the analytical
                              35

-------
results is given  under the individual  subcategory  sections  of
this report.   The additional recommended exclusions include the
following:
           No.
Subcategory
            1.    Hydrogen Peroxide
            2.    Carbon Dioxide
            3.    Carbon Monoxide/Hydrogen
            4.    Hydrochloric Acid
           17.    Nitric Acid
           18.    Sodium Carbonate
           19.    Sodium Metal
           21.    Sulfuric Acid
           22.    Ammonium Chloride
           23. '   Ammonium Hydroxide
           24.    Barium Carbonate
           25.    Boric Acid
           26.    Calcium Carbonate
           27.    Copper Oxide (one plant)
           28.    Manganese Sulfate (one plant)
           29.    Strong Nitric Acid
           30.    Oxygen and Nitrogen
           31.    Potassium Iodide
           32.    Sodium Hydrosulfide
           34.    Sodium Thiosulfate
           35.    Sulfur Dioxide

Silver Nitrate, No. 13, and Sodium Silicofluoride, No.  33, -are
being  deferred for  future study  under  Phase  II  of the  BAT
regulation development  program for Inorganic Chemicals.   This
deferrment was caused by problems  with  plant  access during the
course of the present study.


3.2  GENERAL APPROACH AND METHODOLOGY

     Initiating  and  undertaking a comprehensive study of  the
toxic pollutant problem in the Inorganic Chemicals Industry was
preceded by an intensive  evaluation by  the Agency of the  kinds
of data and supporting information  that  should be assembled as a
basis for  the  development  of  regulations.  All  major decisions
on the identity of pollutants and the establishment of effluent
limitations and  standards of performance for each subcategory
had  to  be  supportable  by  documented  evidence   collected  from
operating  production  facilities.    Similarly,  the  necessary
information on production  rates, processes, raw materials,  water
use, waste sources,  and treatment  technologies  in practice had
to be acquired with sufficient detail and  breadth of coverage to
permit  an  analysis of  the engineering and  economic variables
that are characteristic of  each subcategory.    Toxic pollutant
                              36

-------
control regulations  would be based on  the application-of best
available  technology for  treatment  and  reliable  performance
evaluations for the removal of specific waste substances.

     The following paragraphs briefly.describe  the major study
tasks and their results as they are presented in this -report.

3.2.1   Industry  Data Base  Development  and Subcategorization
Review

     Information  from  individual  manufacturers  and  previous
study documents were reviewed in detail  and an evaluation of the
appropriateness of  subcategorization was performed.  Section 4
presents   a   discussion   of  the  factors   considered   in
subcategorization and presents the rationale for maintaining the
present scheme of subcategorization for the industries studied.

3.2.2  The Screening and Verification Sampling Programs
 *
     The collection of detailed analytical  data on conventional,
nonconventional and  toxic pollutant concentrations  in raw and
treated  process waste  streams  was completed  in  a  two-phase
sampling program.   The  first phase, screening,  was designed to
provide a representative, one-time  72-hour sampling of a plant
in each subcategory in order to determine  the presence of toxic
pollutantp  and   to   evaluate  their   potential  environmental
significance.    The  sampling  and  analytical  methodology  is
described  in Section  5, along  with  the basis  for  making  a
decision  on   the   need   for  verification  sampling  in  each
subcategory.

3.2.3  Engineering Evaluations

     Section  6  describes  the procedures  and  sources  used  in
developing the  industry  productions and waste  water generation
characteristics that form the basis of the model plant concept.
The  sources of detailed process and waste  treatment information
are  also  presented.   Section  7  contains an evaluation  of
treatment  technology presently  applied  in BPT  systems  and
advanced  technologies that  may  be recommended  for BAT and NSPS
applications.  Section 8 provides estimates of the treatability
of  selected  toxic  and nonconventional  pollutants to be applied
in the development of achievable performance characteristics for
specific technologies.   Section 8 also presents a discussion of,
the  approach  taken  in  the  statistical analysis  of long-term
monitoring  data.     The   statistically  derived  parameters,
including variability factors for  24-hour maxima and maximum 30-
day  averages  are  presented  in Appendix A.   Section 9 lays the
groundwork for the estimation of pollutant removal performances
for   each  nonexcluded   subcategory.     The  candidate  toxic
pollutants  to be controlled  in  each subcategory are identified
                              37

-------
on  the  basis of  the screening  and  verification data  and the
rationale for the application of advanced level technologies is
presented.

3.2.4  Treatment System Cost Estimates

     Section  10   presents  the  general  approach   to  cost
estimating,  discusses  the  assumptions  made,   and  gives  the
detailed cost estimates for alternative levels of treatment and
control.   For  each  subcategory verified,  the  total estimated
installed cost of a typical BPT  treatment  system is developed on
the basis of the model plant design specifications and estimated
incremental  costs are  given for  each of  the  advanced level
treatment alternatives.
3.3  GENERAL CRITERIA FOR EFFLUENT LIMITATIONS

                                                          »

3.3.1  BPT Effluent Limitations

     The factors considered in defining best practicable control
technology currently available  (BPT)  include  the total cost of
applying such technology in relation to the effluent reductions
derived  from  such  application,  the  age  of  equipment  and
facilities  involved,  the  process employed,  non-water quality
environmental  impacts   (including  energy  requirements),  and
other factors the  Administrator considers appropriate (Section
304(b)(1)(B)).  In general, the BPT technology level represents
the  average of  the best  existing  performances of  plants of
various ages, sizes, processes,  or other common characteristics.
Where existing performance is uniformly inadequate, BPT may be
transferred  from  a  different  subcategory  or  category.   BPT
focuses on end-of-pipe treatment  rather than process changes or
internal  controls,  except  where  such  are  common  industry
practice.    The  cost/benefit  inquiry for  BPT  is  a limited
balancing, committed to EPA's  discretion, which does not require
the Agency  to quantify benefits in monetary terms.   See, e.g.,
American Iron and  Steel  Institute v.  EPA,  526  F.2d  1027  (3rd
Cir.  1975).   -In  balancing   costs   in  relation   to  effluent
reduction  benefits, EPA  considers  the volume  and  nature of
existing  discharges,  the  volume  and  nature  of  discharges
expected after  application of  BPT,  the  general environmental
effects of the pollutants, and  the cost and economic  impacts of
the required pollution control  level.  The Act does not require
or permit  consideration  of water quality problems attributable
to  particular  point sources  or  industries,  or  water quality
improvements in particular water bodies.  Therefore, EPA  has not
considered these factors.  See  Weyerhaeuser  Company v. Costie,
590 F.2d 1011 (B.C. Cir.  1978).
                              38

-------
3.3.2  BAT Effluent Limitations        -

     The   factors   considered  in   assessing   best  available
technology  economically achievable  (BAT)   include  the  age^ of
equipment and facilities involved,  the  process employed, process
changes,  non-water  quality  environmental  impacts  (including
energy requirements), (Section 304(b) (2) (B)) .  At a minimum, the
BAT technology level represents the best economically achievable
performance  of plants  of  various  ages,  sizes,  processes,  or
other shared characteristics.   As with BPT, uniformly inadequate
performance  may  require'  transfer  of BAT from a  different
subcategory  or  category.    BAT may  include process  changes or
internal controls,  even when  these technologies are not common
industry practice.  The statutory assessment of BAT "considers"
costs,  but  does  not  require a  balancing of  costs  against
effluent reduction benefits (see Weyerhaeuser v. Costle, supra).
In  developing   the  proposed  BAT,   however,   EPA  has  given
substantial  weight  to the  reasonableness of costs.   The Agency
has considered  the  volume  and nature of discharges, the volume
and nature of discharges expected after application of BAT, the
general environmental effects of  the pollutants, and the costs
and economic  impacts of the required pollution  control levels.
Despite  this  expanded  consideration  of   costs,  the  primary
determinant  of BAT  is  effluent  reduction capability.   As  a
result of the Clean Water Act  of 1977, 33 OSC 1251 et seq., the
achievement  of BAT  has  become the principal national  means of
controlling water pollution due to toxic pollutants.

3.3.3  BCT Effluent Limitations

     The 1977 amendments added Section 301(b)(2)(E) to the Act,
establishing  "best  conventional pollutant  control technology"
(BCT)   for  discharges of conventional  pollutants from existing
industrial  point sources.   Conventional  pollutants  are those
defined in Section  304(b)(4) - BOD, TSS, fecal coliform, and pH
and  oil  and  grease,  designated   by  the  Administrator  as
"conventional" on  July  30, 1979,  44 PR 44501.   BCT  is  not an
additional  limitation,  but  replaces BAT  for  the control  of
conventional  pollutants.    BCT requires  that  limitations  for
conventional  pollutants be assessed  in light  of a  new "cost
reasonableness"  test, which involves a comparison  of  the cost
and level  of  reduction of conventional  pollutants  from  the
discharge  of publicly  owned   treatment works  to the  cost  and
level of reduction  of such pollutants  from a  class or category
of industrial sources.  In  its review of BAT for industries not
covered by the NRDC Consent Decree,  the Agency  promulgated BCT
levels based  on  a  methodology described at 44  FR 50732  (August
26, 1979).  This methodology compares subcategory removal costs
(dollars per pound of pollutant, measuring  from  BPT  to BAT) with
costs experienced by POTWs. EPA applied this methodology to the
costs  of   removal   of   conventional  pollutants   in   the  11


                              39

-------
 subcategories  of  the inorganic chemicals  manufacturing  industry
 affected  by these proposed regulations.  Models were chosen  to
 represent the  average size plant in each subcategory. The total
 annualized cost of each  control technology and  the  total  pounds
 per  year   of TSS  removed were  then  computed  for each of  these
 model  plants.  The Agency is  proposing, based on this analysis,
 that BCT  should  be  equal  to  BPT  except  in  the  case  of the
 diaphragm cell portion of the Chlor-Alkali subcategory.  In this
 subcategory,  EPA  is proposing  limitations based on dual  media
 filtration.  The  costs of this technology in the diaphragm cell
 portion of the Chlor-Alkali  subcategory  are explained below  in
 the  discussion of  the treatment options  for that  subcategory.
 For  all other  subcategories,  EPA is proposing that BCT equal BPT
 either because  additional  removal  failed the  cost  test  or
 because  EPA  is proposing BAT equal  to BPT.

     For  Aluminum Fluoride and Sodium Dichromate,  the cost  in
 dollars  per   pound  for  removal   of  additional  conventional
 pollutants is  $2.06 and $13.40, respectively. For Chlor-Alkali -
 mercury cell process, Hydrogen Cyanide, and Sodium Hydrosulfite,
 BCT  is being proposed equal  to BPT  because the  technology added
 for  BAT does  not  impact  the  removal  of conventional pollutants.
 EPA  is proposing  a distinct BCT standard  for the Chlor-Alkali -
 diaphragm cell process and Hydrofluoric Acid as  discussed  below.
 In the remaining  subcategories,  BPT  is  proposed equal to BAT
 which  automatically makes BCT equal  to BPT.

     In the Chlor-Alkali - mercury  cell process  segment and the
 Hydrogen  Cyanide  and Sodium  Hydrosulfite  Subcategories,  BCT  is
 being  proposed equal to BPT because the technology added for BAT
 does not  impact the removal  of conventional pollutants.

     In the Aluminum Fluoride subcategory, the  cost for removal
 of additional  conventional pollutants  is $2.06 per pound.   Thus,
 BCT  is proposed equal to BPT because the cost  is  greater  than
 the  $1.15 per  pound cost for  removal of conventional pollutants
 from a publically owned treatment works (POTW).   The calculation
 is as  follows:

        $1.77  (kg/2.2 Ib.)        =  $2.06 per pound
      (1.2 kg/kkg  - 0.81  kg/kkg)       of TSS removed

     Where $1.77  is the' increased  cost for BAT  treatment  over
 BPT  treatment  cost in dollars per kkg of  production from  Table
 15-9,   1.2  kg/kkg  is  proposed  for  the  BPT  suspended  solids
•limitation from Table 15-19,  0.81 kg/kkg is achievable by  use  of
 BAT  technology applied to suspended solids removal developed for
 Table  15-23.   A conversion factor of 2.2  pounds  per kilogram  is
 used.
                              40

-------
     In the Sodium Bichromate Subcategory, the cost for removal
of  additional  conventional pollutants  is  $13.40 per  pound.
Thusf BCT is proposed  equal to  BPT because the cost is greater
then  the $1.15  per  pound  cost   for  removal  of  conventional
pollutants  from  a  publically  owned  treatment   works.    The
calculation is as follows:

        $1.18  (kg/2.2 Ib.)         =  $13.40 per pound
     (0.22 kg/kkg - 0.18 kg/kkg)      of TSS removed

     Where $1.18  is  the increased  cost  for  BAT treatment over
BPT treatment  cost in  dollars per  kkg  of production from Table
18-12,  0.22  kg/kkg  is proposed for  the BPT  suspended  solids
limitation from Table  18-15, and  0.18 kg/kkg  is  achievable by
use of  BAT  technology applied  to  suspended  solids removal for
Table 18-16.

     In the Hydrofluoric Acid Subcategory, the cost for removal
of additional conventional pollutants  is  $0.37 per  pound.  Thus,
a BCT regulation is established because  the cost for removal of
additional conventional  pollutants is less  than  the  $1.15 per
pound  cost  for  removal  of conventional  pollutants  from  a
publically  owned  treatment works.   The  calculation  is  as
follows:

        $2.42  (kg/2.2 Ib.)          =  $0.37 per pound
     (5.3 kg/kkg - 2.3 kg/kkg)         of TSS removed

     Where $2.42  is  the increased  cost  for  BAT treatment over
BPT treatment  cost in  dollars per  kkg  of production from Table
12-15, where 5.3 kg/kkg  is  proposed for  the BPT total suspended
solids  limitation  from  Table  12-24,  where  2.3  kg/kkg  is
reduction of TSS achievable  by application of filtration to the
waste  waters.    Because  additional  removal  of  conventional
pollutants passes the cost  test, the regulation for BCT for TSS
is  set  at 2.3 kg/kkg  as  a 30-day maximum average and  using a
variability  factor ratio   (VFR) of 2.1  to  establish  a  daily
maximum of 4.8 kg/kkg.

     In   the   diaphragm  cell  segment   of   the  Chlor-Alkali
Subcategory,  the  cost for  removal of  additional  conventional
pollutants is  $1.09 per pound.  This is  less than  the $1.15 per
pound cost  of conventional  pollutant  removal  in  a publically
owned   treatment   works.     This  determination  was  made  by
estimating the BAT cost  (Table 11-33) of a 30 percent reduction
in  the  BPT  maximum 30-day  average TSS  effluent loading (Table
11-36) as follows:

     	($0.36/kkg) (kg/2.2 Ib.)	  =   $1.09/
     0.51 kg/kkg - (1.00 -  0.30) (0.51 kg/kkg)      Ib.
                              41

-------
     Thus, on  the  basis  of adding dual media filtration to the
BPT treatment, the Agency  is proposing a BCT regulation of TSS.
The proposed maximum 30-day average effluent limitation is:

     (1.00 - 0.30)  (0.51 kg/kkg)  =  0.36 kg/kkg

and  the proposed  daily maximum  is  obtained  by  applying the
variability factor ratio (VFR) value of 2.0 as follows:

     (2.0)(0.36 kg/kkg)  = 0.72 kg/kkg

     In the  remaining  subcategories,  BPT  is proposed  equal to
BAT which automatically makes BCT equal to BPT.

 3«3-4  New Source Performance Standards

     The basis for  new source performance  standards (NSPS) under
Section  306 of  the  Act  is  the  best available  demonstrated
technology.  New plants have the opportunity to design the best
and most  efficient inorganic chemicals manufacturing processes
and waste water  treatment  technologies, and Congress therefore
directed EPA to consider the best demonstrated process changes,
in-plant controls,  and end-of-pipe treatment technologies which
reduce pollution to the maximum extent feasible.

3.3.5  Pretreatment Standards fear Existing Sources

     Section  307(b)   of  the  Act  requires  EPA to  promulgate
pretreatment standards for existing sources (PSES)  which must be
achieved within three years of promulgation.  PSES are designed
to  prevent  the  discharge  of pollutants  which pass  through,
interfere with, or  are  otherwise  incompatible with  the operation
of POTWs.  The Clean Water Act of 1977 adds a new dimension by
requiring pretreatment for pollutants,  such as heavy metals,
that limit  POTW sludge management  alternatives,  including the
beneficial  use  of   sludges  on  agricultural  lands.     The
legislative history of the 1977 Act indicates that pretreatment
standards  are  to  be  technology-based,  analogous  to  the  best
available  technology  for  removal  of  toxic  pollutants.   The
general pretreatment regulations  which served  as  the framework
for these proposed pretreatraent  regulations can be found at 40
CWR Part  403,  43 PR 27736  (June  26,  1978).   In some instances
PSES  regulations  have been  established  for  subcategories not
presently discharging  to a POTW.  This establishes regulation
for plants that may choose to change their discharge to a POTW.

3.3.6  Pretreatment Standards for New Sources

     Section  307 (c)   of  the  Act  requires  EPA to  promulgate
pretreatment standards for new sources (PSNS)  at  the same time
that it  promulgates  NSPS.   New  indirect  dischargers,  like new
                              42

-------
direct dischargers, have the opportunity to incorporate the best
available demonstrated  technologies  including  process changes,
in-plant controls,  and  end-of-pipe  treatment technologies, and
to use plant site selection to ensure adequate treatment system
installation.
                              43

-------

-------
                           SECTION 4


                    SUBCATEGORIZATION  REVIEW
4.1  BASIS FOR SDBCATEGORIZATION


4.1.1  Factors Considered

     The inorganic   chemicals  industry  is   very  large   and
diversified and   has been   segmented   into  subcategories for
the  purpose  of   establishing effluent  guidelines.   Factors
taken  into   consideration   for   subcategorization   includes
raw   materials used,  product produced,  manufacturing process
employed,  geographical  location,   size and  age   of equipment
and  facility  involved,   nonwater   quality aspects  of  waste
characteristics,    water   pollution   control    technology,
treatment    costs,    energy   requirements  and  solid  waste
disposal.  Following is a discussion  of each of  the general
factors considered for this industry.

Raw Materials

     Different raw  materials  are  used  to manufacture  a wide
variety  of products, and vary  from raw brines and  ores to pure
reagent  chemicals.   Some  proceses  use  waste  or  by-product
streams  from  other  plants or from other  processes  within the
same  plant.                       *

     Because of     this    diversification,     raw    material
characteristics  generally do not constitute a   logical basis
for  subcategorization.  Variations   in raw material quality or
purity  are  not normally sufficient to cause  a great difference
in waste   water   treatment  needs,   except   in   the  case  "of
trace  toxic  materials which may occur in some sources but not
in others.

Dominant Product

     Subcategorization   by   chemical  name   of   the  dominant
inorganic  chemical   produced   involves   the   least  ambiguity
in   applying   standards to  a given point   source.  This  is
critical   because  of   the  great   variety of  product  mix,

                              45

-------
manufacturing  processes,  waste   water    constituents,    and
other  factors   at  existing   plants.    Subcategorization  by
product  becomes  less  useful as  product   mix   increases   in
complexity  because   multi-product waste  water  also   becomes
more   complex   and    less    susceptible   to  simple  uniform
treatment.

     h subcategory   established  on   the    basis   of  product
manufactured  might  have two or  more  different  processes but,
in   the  majority  of  cases, the  characteristic of  the  waste
waters  is  similar  and the  same   treatment technology  can  be
applied  for   different  process   waste  waters.  If   two  or
more   dissimilar  processes produce waste water  of  different
quality, and different  treatment technologies have to be used,
then  the subcategory has   to  be  further   classified     or
segmented,   for  example,   the  Chlor-alkali Industry.

Manufacturing Process

     Typically,  inorganic    chemicals  are  manufactured  for
captive  or merchant use in four or more steps starting from raw
material   to final product.   Two  or  more different products
might use the  same   process but  then  the raw materials  used,
process  sequence,   control,   recycle  potential, handling, and
quality  control    will   vary,     producing     wastes     of
different   quality.    Primary  subcategorization,  therefore,
by  process  is  unlikely   to  be  useful.  However,  secondary
subcategorization by process  has been   necessary in some cases.

Geographical Location

     Inorganic chemical   plants   exist  in all  parts   of the
United   States   but  subcategorization   on   this  basis  is not
appropriate.   Geographical location is  important  in analyzing
the  feasibility    of    various     treatment    alternatives.
Evaporation   ponds   are   functional    only   in  areas.  where
evaporation exceeds   rainfall.     Ocean dumping and  deep well
disposal  are  possible only *in  certain   areas,      and must be
consistent  with     local,   State      and   Federal      laws.
Theppossibility of ground water contamination may preclude  the
use of unlined holding and settling ponds in many locations.

     In   the  northern  regions,   climatic  conditions   may
necessitate  the  inclusion   of  special  provisions  to  prevent
freezing  of     treatment   system  components,    particularly
biological    oxidation   units,  clarifiers,   ponds,  and  open
collection  systems.    The costs    of    utilizing   waste heat
sources   from   the  process or  providing   various   types  of
thermal protection, such as insulation  or burial  of  pipes  and
tanks  and  .building  structural  shelters, may add  considerably
to the capital   and   O  &   M cost associated with a  treatment
technology.


                              46

-------
     Thus, the   influence of    geography,   climate,  geology,
etc.    is  reflected    in    waste   treatment   modifications
and   is  primarily manifested in  the cost of  treatment.  This,
of itself,  is  not a good basis for subcategorization.

Plant Size

     Plant  size and  production  capacity  were  not found   to
affect   the  characteristics  of the  waste  produced.   Although
plant size  can affect  treatment  cost, this  variability  can
be expressed  graphically or  mathematically  without   the  need
for  further  segmentation of the category.

Plant Age

     Plant age  can  have an  important  bearing  on waste  water
volume  and quality and is, therefore,  a significant factor  to
consider   in   evaluating  the   applicability   of   treatment
technologies  and   assessing  the  relative   costs of   treatment
for  plants  of   widely differing age producing   the   same  or
similar   products.  A  particular  problem with  older  plants is
that their  present patterns of water use may have  evolved over
a  long  period    of time with  little  consideration  for  the
principles of efficient   waste  segregation,  collection,  and
treatment.   To   a  limited   degree,  plant  modernization can
correct  or  at  least  mitigate some   of these  shortcomings in
older  facilities,  however,  only a   small   proportion   of the
cost  of  revamping  collection systems or of  converting  from
contact to  noncontact  cooling systems  can  be   offset by  the
resulting lower cost  of  treatment.  In  general,   older plants,
even   after  considerable modernization,   normally   have   a
higher  volume   of waste   water   flow   and   higher   waste
loadings  (although  pollutant  concentrations   may be   lower
due    to   poor  segregation   from  noncontact  sources)    in
comparison  to     relatively  new  plants.    The   present  and
forthcoming   requirements  for pollution  control may  impose a
severe treatment   cost penalty  on older plants due to the need
for backfitting and replumbing of  outdated collection  systems.
Land''  availability  and    land   use    restrictions   are also
factors  which  may translate into higher  treatment costs  for
older  facilities  which   find  themselves  surrounded  by highly
developed industrial and  residential areas.

     Unfortunately, plant  age  does not  readily  lend  itself
to   an   unambiguous    definition    where   a    series    of
plant    modifications  has  taken  place.    The   extent    of
modifications also  varies  greatly  among  plants   within  the
same product   industry.     For  those  plants  that have  been
enlarged  or  modified  from their  original status, plant  age  is
not  unambiguously calculable   and   therefore not a reasonable
basis for subcategorization.
                              47

-------
Nonwater Quality Characteristics

     Airborne  emissions from  manufacturing operations  can be
kept  within   air quality control limits   through  the  use of
cyclones,  wet  scrubbers  and other methods.  The nature of the
air pollution   is   related   to  the  products(s)  manufactured
and/or  the  raw  material used.  Since  both of these elements
vary widely within  the   inorganic  chemicals  industry,  there
is   no   logic   in  subcategorization   on   the    basis   of
nonwater    quality  characteristics.

Treatment Cost

     From a  technical  viewpoint,  subcategorization  by common
technological  requirements    for   treatment  processes  could
provide   a  logical   basis   for selecting one or   more unit
processes    to    accomplish    the  same   treatment  "function,
regardless of the  source  of  the waste  water.  For  example,
residuals  of  dissolved  heavy   metals will   respond  to lime
precipitation    and sedimentation at  high pH  without respect
to the  specific origin  of   the  metals.   This "building block"
concept    could   conceivably    result  in   selecting various
combinations     of     unit      processes   to     meet    the
treatment  requirements.  However,   if  the  treatment  cost must
be expressed in  terms  of  dollars per unit   production,   this
method   of subcategorization     crosses  product  lines    and
interferes   with comparison  of   treatment  costs based on  the
production   of a  specific   chemical.    Even   if  the   unit^
operation  is  commonly applicable for   treating  waste  flows
of different  products, the cost  of   treatment will   fluctuate
because of  variations   in   quality, loading  and  flow  rates
and  subcategorization on  the basis of  treatment cost  is not
recommended.

Energy Cost

     Manufacturing  processes  in    the  Inorganic  Chemicals
Industry   typically   have   large   energy requirements.   In
contrast,   waste   water   treatment processes  consume a small
fraction of the  total   energy  used.    There  appears  to be no
major energy   requirements   for  the waste   water    treatment
facility  and subcategorization   on  the basis of energy cost is
not justified.

Solid Waste

     Not  all    inorganic   manufacturing   processes   produce
solid  wastes.  Solid waste producers practice various disposal
methods,   such as  on-site   landfills,  contract  'hauling to
approved   dump   sites   or   incineration.     Solid     waste
disposal  becomes,  very  site specific  and exhibits a wide range
                              48

-------
of costs.  Because of the  lack   of  uniformity   within   the
industry,   solid  waste, generation  and disposal practices are
not  a  staisfactory basis  for subcategorization.
                                   B
4.1.2  General Conclusions '

      If effluent limitations  are  to  be  tied  to   units  of
production,  only one method  of  primary  subcategorization  is
broadly  applicable  to  the  inorganic  chemicals point  source
category;   viz., subdivision  by  dominant  product.   However,
there    are  three    subcategories,    Chlor-Alkalir    Titanium
Dioxide,    and    Hydrogen    Cyanide  which  require  further
subdivision  based on    the   difference   in the quantity and
quality of the waste  water  from  the  processes, and two others,
Hydrofluoric  Acid  and   Aluminum  Fluoride,   have  been reviewed
for  possible  integration  (see  Section 4.3).


4.2  SECONDARY SUBCATEGORIZATIOW


4.2.1  Chlor-Alkali

     Mercury and  .diaphragm  cells  are the two distinct types
of   electrolytic  cells that  are   used  in  the  production of
chlorine   and caustic   soda.     Major   process   differences
between  mercury cell   and  diaphragm  cell   plants  produce
corresponding   differences in  the volume and  nature of waste
water generated.  A   principal difference is the   presence pf
mercury as a  contaminant  in  the waste waters from the mercury
cell process  and asbestos in   the    diaphragm   cell   plant
wastes..   The  TSS   discharges from  diaphragm cell plants are
generally larger^ than from mercury cell  plants,   due  to the
higher volumes of contact and noncontact water used.  Also,  in
diaphragm   cells a  large  amount  of  water   is used and  an
appreciable   quantity   of  waste water   is   produced   in the
caustic evaporation  process.   Such  water is not   produced in
mercury  cell plants.   The  quantity  of waste water generated
from the  diaphragm  cell plants  is  almost double that of   the
mercury   cell    plants   for   the   same   chlorine production
capacity.  Based   on  the  quantity and characteristics of   the
waste water, further subcategorization is justified.

4.2.2  Titanium Dioxide

     Two major  ores,  rutile   and  ilmenite,  are used  for the
manufacture of titanium dioxide.  The ilmenite ore contains 40-
70 percent titanium  dioxide   (Ti02),  up   to 35  percent ferrous
oxide  (PeO),  and 25  percent  ferric oxide (Fe203).   Rutile ore
contains more than 90 percent Ti02.  Two processing techniques,
the  sulfate  process and  the  chloride process,   are used   to
extract titanium dioxide from the ores.

                              49

-------
     The sulfate process uses   ilmenite ore and sulfuric  acid
as  raw  materials.   The  chloride  process  uses  rutile  ores
and  chlorine.   The high  grade  rutile ore  is  expensive  and
its    availability    is declining.    In    recent  years,  new
technological    advances    have  alleviated  the raw  material
shortage problem.   By   upgrading  the  ilmenite ore quality,  the
chloride   process  can  be  used to produce  titanium  dioxide of
high purity.  Because of the  difference  in quality and quantity
of  waste  waters generated  from   the  sulfate and  chloride
processes using the two different  ores,  the titanium  dioxide
industry may  be  further  subdivided   into   three  segments as
follows:

      a.  Sulfate process

      b.  Chloride process using  rutile ore

      c.  Chloride process using  ilmenite ore (one step).

     The sulfate process generates large amounts of strong and
weak   sulfuric   acid   water-borne   wastes.   Application of
pollution  control technology to the acid wastes  generates about
five times   as much gypsum as  product.   The chloride  process
generates large  amounts  of  dissolved metal  chlorides  and the
treatment  technology   is  expensive.    Solid waste  from both
processes present difficult disposal  problems.   These   solids
include  ferrous   sulfate   (FeSO4) and   a hydrated  by-product
from  the  sulfate   process  and heavy  metal sludges  from  the
chloride process.   Ilmenite ore  has  to  be upgraded   before
it   is   used  to    extract  titanium dioxide  by the chloride
process,  and  this   beneficiation    process step    generates
additional wastes.

     The application of the  chloride  process to  ilmenite ore
may  proceed  in either  one or  two  steps.  A patented one-step
process   accomplishes  both "beneficiation and  chlorination of
the ore  in   a  single  fluidized  bed  reactor   and generates
raw    waste   loadings which   are  similar to  those  from the
sulfate process  in   terms  of   acidity and metals,  and similar
to  wastes from  the  chloride-rutile process in terms of spent
coke  solids  and still  residues.   In  the two-step process, ore
beneficiation  resulting   in either a synthetic rutile  or an
enriched  titanium oxide slag   is carried   out separately  at
the   mine  or  the   plant.     The   discharge  of waste  water
generated  by  the beneficiation  step  would be  regulated under
the Ore  Mining  and  Dressing Point Source Category and will not
be considered in this document.   The  second   step of the  two-
step   process   generates   wastes   that are very   similar in
quantity and  quality to those from the chloride-rutile  process
and  will be  governed  by  the  discharge  regulations for  that
segment of the TiO2 subcategory.
                              50

-------
     Therefore,  further subclassification  based  on the  amount
and  characteristics  of  the    waste  water  appears   to   be
justified,   and    the  three   process  subdivisions  indicated
above   are  appropriate   for  this purpose.

4.2.3  Hydrogen Cyanide

     Hydrogen cyanide    (HCN)   is   made  from   two   different
processes,  the Andrussow   process    and   as   a  by-product   of
acrylonitrile  manufacture.   In  the  Andrussow  process, air,
ammonia,  and  natural gas  are reacted to  produce  the  dominant
product hydrogen cyanide.

     Water-borne    wastes      from   the   process     consist
" principally of  ammonia and sulfates in addition to cyanide  and
nitriles.

     The   primary   product     in   the     other     process   is
acrylonitrile   (CH2 = CHCN)  and the  hydrogen cyanide  is   a  by-
product.  Because   the hydrogen cyanide  is a by-product it will
be covered  in  the   organic  chemicals   manufacturing  category
with   the   primary product.


4.3  REVIEW OP POSSIBLE  INTEGRATION  OP SOBCATEGORIES
4.3.1  Hydrofluoric Acid and Aluminum Fluoride

     Aluminum   fluoride   (A1F3)  usually   is  produced   by   the
reaction   of  hydrated   alumina    (A12O3.3H20)   with  hydrogen
fluoride   (HF),   although one plant produces  aluminum   fluoride
from fluorosilicic   acid  (H2SiF6), a  by-product  of   phosphoric
acid  (H3FO4).    With  one   exception,    all   the    aluminum
fluoride   plants    are  integrated   with  hydrogen   fluoride
 (or  hydrofluoric  acid)  production.

     The  two  major   uses   of    hydrogen   fluoride   are   in
the    fluorocarbon    industry  and  as raw  material  in    the
manufacture  of   aluminum  fluoride.  A  ban on the   fluorocarbon
propellants  has  curtailed   the  use of hydrogen  fluoride   in
that industry  and  it   was   completely  stopped  in 1978.    The
selling    of   hydrogen   fluoride in  the merchant market  has
declined  and  the   primary  use    is   limited    to     the
production   of    aluminum  fluoride   and   fluorocarbon  plastics
until  some other  major use is found.

     For  both    products  (HF     and  A1F3) ,   process waste
waters are  generated by the various  gas  scrubbers and  by leaks
and  spills.     In both  cases,  air pollution control  scrubber
effluents   contain  mainly fluoride, acidity and sulfate.   The
fluoride is present as  the free ion  as well as various   complex

                              51

-------
fluoro anions.  Calcium  fluoride   (CaP2),  generated as  a  solid
waste, is  a   disposal  problem   for  both  the  subcategories
because of  its moderate  toxicity.   Only one additional solid
waste, gypsum  (CaSO4.2H2O),    is  generated from  the   hydrogen
fluoride manufacture alone,  and  it can be treated and handled
independently.

     Combining  hydrofluoric   acid  and    aluminum   fluoride
into a  single  subcategory  does  not   appear   to  offer  any
regulatory  advantages when  the   two  products are manufactured
at the same  plant  location.   The   waste  waters  associated
with   the   two   products   are similar  and a common treatment
facility is normally  utilized.    In  addition,  the  combined
manufacture  of   these  products   does  not  create a unique or
unusual situation, either  with   regard  to  the  waste  water
treatment    requirements   or    compliance   with   discharge
regulations.   Although  the   waste  gypsum produced  at  an HF
plant  supplies   enough  calcium for  adequate  fluoride removal
from neutralized  scrubber  waste waters   generated  by   both HF
and A1F3  production,  the  applied treatment   technology    is
essentially   the .  same   as  that  applied  by  manufacturers
of  either  product  alone.    However,    the    effluent   water
quality   and   the  toxic   pollutant  loadings   would   not be
expected  to be the same. Further,  the  opportunities for  drip
acid recycle  (or  the hydrolysis of  complex fluorides  prior  to
treatment)   and scrubber  water  recycle are a function of  plant
design and age, rather than product mix.

     In  view   of  these   considerations,  a   recommendation
for the  creation  of an  HF/A1F3 combined product subcategory is
not being  made at this  time,


4.4  SUMMARY


     The    recommended    subcategorization    with    process
subdivisions  include the following:


     Subcategory                Process Subdivisions

     Chlor-Alkali               Mercury Cell
                                Diaphragm Cell
     Titanium Dioxide           Sulfate
                                Chloride-Rutile
                                Chloride-Ilmenite
     Hydrogen Cyanide           Andrussow Process
                                Acrylonitrile By-Product
                               52

-------
                           SECTION 5


          SCREENING AND VERIFICATION  SAMPLING PROGRAMS
5.1  SCOPE AND METHODOLOGY


     The specific  objective  of the  sampling  programs was  to
establish  the  extent  of   the   required  regulation  of  toxic
pollutant  discharges  in  the inorganic  chemicals  industry in
terms of factual information derived  from the chemical analysis
and flow measurement  of representative process raw waste water
streams and treated effluents.  Prior  to this study, most of the
information  available on  toxic pollutants has  been concerned
with  a  relatively   small   number   of  known  process-related
substances -contaminating  a  variety  of  direct  and  indirect
contact process  waters  discharged from a  production facility.
There  had  been  no  previous requirement  for a  comprehensive
survey of waste water chemistry addressing the possibility that
a large  number  of other potentially  toxic  substances could be
present, albeit at extremely low concentrations.

     The screening phase of the  sampling program was designed to
ascertain the presence  in each  subcategory  of any of  the 129
listed  toxic pollutants at  raw waste * concentrations  or  daily
loadings  which,   if   untreated,   could   be   environmentally
significiant.  Screening is based on the sampling  of one or more
typical manufacturing operations in  each subcategory.   Where
significant  pollutant  concentrations were found,  additional
plants  were  sampled  during   the  verification   phase   for
confirmation  and   further  quantification  of   data  on  the
particular toxic pollutants  in  question.    A,  goal was  set for
screening and verification   sampling   of  a  sufficient  number
of plants to account  for at  least 75  percent  of  the total-U.S.
production,    in   each    .subcategory   having    significant
concentrations  of priority pollutants.

     A detailed description of  the  screening  and verification
programs is presented in the'paragraphs below.
                              53

-------
5.1.1  Selecting giants and Making Preliminary Contacts

     In each subcategory, plants were selected for screening on
the basis of the following general criteria:

     1.  Minimal product mix and no organic product lines which
         could 'increase the potential  for  interprocess  cross
         contamination of waste waters.


     2.  Presence  of  a physical  chemical  treatment  facility
         rather than a  biological  one,  or  no treatment system.
         (Biological  systems   are  neither  widely  used  nor
         generally   applicable   in  the   inorganic  chemicals
         industries.)

     3.  Manufacture  of industrial  grade  products  in  volume,
         rather than low volume reagent grade products.

     4.  Median production capacity within the subcategory.

     5.  Segregated waste streams  to facilitate sampling.

     6.  NPDES  discharges  rather  than  POTW  discharges,  since
         treatment  for  a  NPDES  discharge  is  usually  more
         extensive.

     7.  Geographical   clustering   of  selected  plants   to
         facilitate field logistics, but only extent that other
         factors are equal.

     Preliminary   phone   contacts  were    made    with   plant
representatives of  those facilities which  satisfied  the above
criteria.   If  requested,1 a letter  was  written  to describe the
objectives  of  the  sampling  program  and   to cite the  legal
authority  of  the  Agency  and  its  sampling contractor  under
Section  308  of   the   Federal  Water   Pollution  Control  Act
Amendments  of  1972.   Secrecy agreements,  when   required, were
executed  at  this   time for  the protection of   any  company
proprietary information disclosed  to the sampling contractor.

     Prior to the actual sampling  of  waste  streams,  a lead visit
to the selected plant was made  to  gather background  information,
confirm and update any 308 Questionaire  responses, and to obtain
additional technical information regarding  processes  and  waste
treatment practices.  Sampling sites  were selected and described
relative  to a  detailed waste  source  inventory and  a   flow
diagram of the process and waste treatment  system.  Arrangements
were made for the  subsequent .sampling  visit  and the details of
the lead visit  and  sampling point descriptions  were documented
in an interim report to the Agency.


                              54

-------
5.1.2 Screen!rig and Verification Sampling

Collection of Samples for Screening

     In  the  screening  phase  of  the  sampling program,  the
specific  objective was  the detection  and quantification   of
water-borne waste constituents included on the list  of  129 toxic
pollutants  (Table 3-1).  Each sample of an individual  raw waste
stream,  a  combined  waste  stream,  or  a  treated  effluent was
collected   where   possible  by  an   automatic,   time  series,
compositor  over  a  single  72-hour  sampling period.   Where
automatic compositing was not possible,  grab   samples were taken
at  intervals  during the same   sampling  period  and composited
manually.

     Each   sample   was  divided  into   several   portions  and
preserved,  as  required  for  different  types  of analysis,  in
accordance  with the procedure  established by EPA  (4) for the
measurement of  toxic pollutants.
" \
     Samples  were  also  taken  from   the  composites,  or  as
individual  grabs,   for  the  analysis  of  the  conventional  and
nonconventional pollutants.

Collection of Samples for Verification

     The  objective  of  verification  sampling was  to confirm the
first  observations from screening  and  further quantify  the
concentrations  and  waste loadings of   the toxic pollutants and
conventional and nonconventional pollutants.   Where any  toxic
pollutant  metals  were  found  during  screening  sampling  of  a
particular  plant^,  analyses were made  for all toxic  pollutant
metals during the verification sampling.

     The   established  protocol   for  verification  sampling
required  the  collection of  three  24-hour composites  at   each
sampling point.  Again,  where composites could not be taken with
automatic  samplers,  grab samples were  taken periodically over
the same  time period and composited manually.

Sample Shipping

     All  samples,  individually  labeled,  were placed  in large
plastic bags, which were  then placed  in a waterproof  insulated
shipping  container. -  Enough ice  was  included  to  maintain  a
temperature of approximately 4 degrees C.  during  shipment to the
laboratory.

     Containers were  shipped  by   the  best   available  route,
usually air freight, usually  arriving  at  the laboratory on the
same day,* but occasionally taking overnight.    Upon  receipt, all
samples  were   immediately  placed   in   a  walk-in  refrigerator
maintained  at 4 degrees C.

                               55

-------
     In order to maintain the chain of  custody and to keep track
of samples, sampling personnel kept logs of  samples  taken in ink
in page numbered hard-bound books.   The data recorded included:
date, time, plant code, number, sample  type,  and sampler.  This
information  was  also  included on the label  of  individual
samples.   Prior to their  arrival  at the laboratory,  a list of
samples shipped, including number,  type of samples, and analysis
to be performed, was sent to each department supervisor to alert
him  of incoming work.

     A  master  analytical  control   chart was maintained  which
included: date sample was received, date due, number and type of
each sample, and the analysis  required.

     At  the  time   of  analysis,  the  individual  samples  were
distributed  to  the analytical chemists along  with a list which
included:  I.D.  number  of sample,  type of  sample,  analysis
required, date  samples received, and due dates.

     Upon completion  of  analysis,  the  sample was sent back to
the  refrigerator  and placed  in  identified  bins.    All samples
were kept  in the refrigerator at   4 degrees C.  when  not being
analyzed.  A list  of completed samples  was then sent to the EPA
Sample Control  Center.

Verification Sampling Plant Selection

     After the decision was made to verify the presence of toxic
pollutants found in the screening of a  subcategory, verification
plants were selected.  The basis   for  selection was essentially
the same as that used  in selecting  screening plants.

     The screening program  results  were evaluated  to identify
those   toxic   pollutants  that  were  present  at  significant
concentration or significant daily   loadings.  Concentrations or
loadings which  could be reduaed by  the  highest quality treatment
systems were considered significant.  Two situations occurred:

     1.  A  subcategory  which  had  a   significant  raw  waste
concentration   of  any  toxic  pollutant would  be  subject  to
verification sampling, and BAT-based regulations would likely be
proposed by' the Agency for  the treatment and control  of  that
toxic pollutant,

     2.  A  subcategory  which  had  no  significant  raw  waste
concentration  of  any  toxic  pollutant  would not be  subject to
verification   sampling  and   would  likely   be  excluded  from
regulatory  coverage   at   this  time   in  accordance  with  the
provisions  for exclusion  under  Paragraph 8 of  the Settlement
Agreement.
                               56

-------
     In   analyzing   screening  data,  only   those  pollutants
attributable  to  process sources  were considered.   Pollutants
which  result  from  cooling  tower  operations,  corrosion  or
corrosion control,  control  of biological growth,  or  any other
operation not directly  tied to the production process were not
used as a basis for verification.

     The  number  of  plants  selected  for  verification  in  each
subcategory was roughly proportional  to the  number of existing
plants  in  that  subcategory  with a  maximum of  five   plants
selected.  In  small subcategories  (relatively few  production
facilities), an effort was made to select  a sufficient number of
plants  to  account   for   the  majority  of   the  total  U.S.
Production.

     When the verification  phase of the  program was initiated,
an important decision was made with regard to  metals analysis.
First, in view of  the frequent presence  of metal contamination
in the wastes  screened, and the  inability in some cases to show
a  direct  relationship  between certain metals  found    and  the
known  process chemicals  or  the materials  of construction, it
was decided  that  all 13  of the toxic metals should be determined
again during verification, regardless  of whether they were found
in screening.  This was  intended  to provide a much more complete
data  base  than  would  be  obtained  by  running  verification
analyses for only those metals  found  in screening to exceed the
verification criteria levels at the time of sampling.

5.1.3  Analytical Methodology for Toxic Pollutants

     The analytical protocol for the screening and verification
of  toxic  pollutants was  established  in  Sampling  and Analysis
Procedures  for  Priority   Pollutants by  U.S.  Environmental
Protection   Agency,   Environmental   Monitoring   and   Support
Laboratory,  Cincinnati, Ohio, April 1977.

     The  specified   analytical   methodologies  were   employed
without modification except  where noted below in connection with
toxic  metals analysis during verification.

     Implementation  of  the methodology  and quality  assurance
provisions required the establishment  of special sample handling
and  control procedures specifically  suited  to  each type  of
analysis.   These procedures,  together with  a discussion of  the
achievable  detection limits  for  each  parameter  or  group  of
similar parameters are presented in the following paragraphs.
                               57

-------
Trace Metal Analysis

     Figure 5-1  shows  a data flow diagram for metals analysis.
Atomic absorption methods  described in 40 CFR  136  per Section
304 (h)  were  used.    A  set   procedure  was  followed  in  the
laboratory  to generate the  analytical  values and  the quality
control data.  The data flow diagram shows  the actual sequence
employed in verification analysis and the following notes, which
are keyed to  the diagram, provide additional information on the
procedures:

     1.  Blanks — two for each  set,of analyses digested.
         Duplicates — one every seventh sample.

     2.  Quality Control at Operator Level (Atomic Absorption):
         Blanks —
         Standards —
         Spikes —
         Duplicates —
These were run at the beginning and the
end  of  every  set   analyzed  for  each
metal.   Also,  air  blanks were  run  on
furnace,  or  heated  graphite  atomizer,
(EGA) ,  after  any sample  with  a large
positive value.

Three different concentrations were run
at the  beginning  and end of  every set
analyzed  for  each   metal.    Standards
were also run every   tenth sample during
the analysis of a set.

These  were run  every  seventh  sample,
and  were  made  by  taking  a mixture  of
equal  parts   of a   sample and standard
and comparing the  resulting absorbance
with  individual  sample  and  standard
absorbances.

For  furnace   analysis,  the sample was
run twice when the absorbance was low to
identify errors. The average of  the two
values  was   used  as  the  determinate
value.
         UTD = "Unable
         interferences.
  To
Determine"
due
to
matrix
         Criteria Employed  in Spike Selection:

         a.   Samples  were  chosen to be  spiked  based upon the
              following criteria;

              —   those which were not subject to interference
                   effects.
                               58

-------
FIELD
SAMPLING
                                                               PRESERVATIVE ADDED
                                                              ICED , AND AIR SNIPED
                                                             RECEIPT . LOS IN  SAMPLES

                                                               AND REFRIGERATE
U1
                                        STANDARDS
                                        PREPARED
UTD
O)

DETERMINATE
VALUE
t


                                   FLAME  ANALYSIS
                                                           QUALITY CONTROL BLANKS AND
                                                           DUPLICATES CR1AT1O (1)
ATOMIC ADSORPTION ANALYSIS (2)

I
NACE
b,Sb,T1 )




FLAME
- t Aa3»,Cr,Cu,NI,Zn !

~»
*•
VAPOR OENERTION
(Ho)

                                                                                                                      HYDRIDE  GENERATION
                                                                                                                          ( Aa,S» )
                                                                                                                    OFF-SCALE
                                                                                                                     RESULT
                                                                     DETERMINATE
                                                                       VALUi
                                        Figure 5-1.   Sample flow  sheet for metal analysis.

-------
              —   those that had a measurable concentration of
                   the metal being determined.
              —   those whose concentration  was  in the linear
                   range of the instrument.
              —   approximately every seventh sample.

     b.   The  level  of spike  chosen  was controlled  by  the
following factors:

              —   it should be  approximately 40-60 percent of
                   the determinate value.
              —   the  determinate  value  absorbance 4-  spike
                   absorbance must  give  total  absorbance" that
                   was within the linear  range.

     c.  A reagent blank was run  with each set of spiked samples
         prepared.


     During the  screening  phase  of  the  sampling   program, the
standard protocol followed for metals analysis was:

     1.  Twelve     elements     were     determined     by    AA
         spectrophotometry in the furnace  (HGA) mode.

     2.  If subject to matrix interference (UTD), they were then
         determined in the flame mode.

     3.  Mercury  was  determined by the  standard  cold  vapor
         method.

     Certain  changes  in  analytical  protocol  were  instituted
during verification analysis  in  order   to avoid the excessive
matrix interference experienced during screening when the heated
graphite atomizer   (HGA)  was the primary method applied to the
analysis of 12 of  the  metals. The modified protocol for metals
was:

     1.  Six elements were determined by flame only, namely, Ag,
         Be, Cu, Cr, Ni and Zn.

     2.  Pour elements were determined by furnace  (HGA), namely,
         Cd, Pb, Tl  and Sb.   If  interference  occurred,  Cd, Pb,
         Tl and Sb were determined by flame.

     3.  Hg was still analyzed by the cold vapor method.

     This modification reduced the  number of   preparations per
sample from three to two and^achieved  adequate detection limits
which were still well below the verification criteria levels.
                               60

-------
     Additional modifications were made during the verification
program to improve the reproducibility and precision for Hg, As
and Se.  These wer,e:

     1.  The  cold  vapor  procedure  for  Hg was  modified  to
         eliminate the pump and allow dilution  and  rerun from
         the  same  sample.    This  saved   time  and  increased
         reproducibility.

     2.  Selenium  and   arsenic   were  determined   by  hydride
         generation  using  sodium  borohydride  (NaBH4).    This
         greatly  minimized  problems  associated  with  matrix
         interference.  The method is very reproducible and the
         detection  limits   were   at  levels  well  below  the
         verification criteria for thes_e two elements.

     After  the  above modifications  were  adopted,  screening
samples which  originally were  unable to be  analyzed,  or which
were recorded as below excessively high  detection  limits due to
the effects of  matrix interferences,  were rerun.  Satisfactory
results  were  then  obtained  in nearly  all cases  due  to  the
greatly  improved sensitivity and reproducibility.

     It  should  be  noted  that  these  modifications  of  the
analytical protocol were  in the direction  of improved precision
and reproducibility'and not towards lower detection limits.  The
original screening procedures  generally had a  lower detection
limit when it   was  achievable.   However, the methods were  too
susceptible to  giving no r.esult  at  all with complex industrial
matrices,  and  so  the   revised   protocols  sacrificed  some
sensitivity  for precision  and  reproducibility.     The   final
detection limits -were still below levels  that would  be regarded
as significant.

     Table 5-1 presents  a summary of  the  analytical detection
limits  for each  of  the 13  toxic metals  using  the  original
protocol  and   the  two   subsequent  modifications  which  were
applied.

Organic Compound Analysis

     The  organic  toxic  pollutants  were  determined  by  the
standard protocol (40  CFR 136 proposed December 3, 1979) which
includes   sample  preparation,   extraction,  and  analytical
methodologies.   Extractions were  carried out using methylene
chloride  in  the  case  of  the  acid  and  base/neutral  organic
fractions  and  with hexane/methylene  chloride  to    obtain  the
pesticide-containing  fractions.   The  acid  and  base/neutral
fractions  were  reduced  in  volume  and  analyzed   by  gas
chromatog,raphy-mass  spectrometry (GC/MS) .  The  pesticides were
analyzed  by  electron  capture gas  chromatography followed  by
                               61

-------
                 5-1.  ANALYTICAL DETECTION LIMITS FOR METALS
                                                             (1)
Element
Antimony, Sb
Arsenic, As
Beryllium, Be
Cadmium, Cd
Chromium, Cr
Copper, Cu
Lead, Kb
Mercury, Hg
Nickel, Ni
Selenitm, Se
Silver, Ag
Thallium, Tl
Zinc, Zn
Original Screening
Protocol t2)
Method (pg/D
EGA*
EGA
EGA
EGA.
EGA
EGA.
HG&
Cold Vapor
EGA
EGA
EGA
EGA.
EGA
10
3
0.2
1
1
1
10
0.5
1
9
0.5
2
1
First Modification Second Modification
of Protocol ^ of Protocol ^
Method (jag/1) Method (pg/1)
EGA
EGA
Flame
HG&
Flame
Flame
EGA
Cold Vapor
Flame
EGA
Flame
EGA
Flame
10
3
15
1
25
20
10
0.5
25
9
15
2
25
EGA
Eydride
Flame
EGA
Flame
Flame
EGA
New Cold
Vapor
Flame
Eydride
Flame
EGA
Flame
- 10
10
15
1
25
20
10
0.5
25
10
15
2
1
   Heated Graphite Atomizer

(1) Assuming no matrix interferences requiring dilution of sample.

(2) EPA Contract  No.  68-01-4492 (September 29,  1977), Exhibit C,
   "Protocol for the Measurement of Toxic Substances",  Environmental
   MDnitoring and Support Laboratory,  Cincinnati,  Ohio

(3) June,  1978

(4) August, 1978
                                      62

-------
GC/MS  confirmation  of  positive  results.   Volatile organics
were analyzed  by  the purge and trap method  of introducing the
material  into the GC/MS inlet system.

Cyanide Analysis

     The standard methods for the wet chemical analysis of  total
cyanide and cyanide  amenable to chlorination'(Cyanide, A). were
utilized  (40  CRF  136) Cyanide analysis  is subject to   several
sources of interference including:

     Metals -  The presence of Fer  Cd,  Ca, Ni, Ag,  and  Zn may
cause low results due to the formation of  stable complexes with
cyanide.    The iron complexes may form insoluble precipitates
which are particularly difficult  to break  up  both at the time
of  alkaline chlorination  of the  sampled waste water  and during
the,chemical analysis for cyanide.
    \
     Oxidizing agents  - The   presence of  free chlorine in the
waste water sample will destroy cyanide  and cause low  analytical
results.   The addition  of ascorbic acid to destroy chlorine at
the  time  of  sampling is  intended  to  mitigate  this problem.
Other oxidizing agents such as peroxides and  chromates may also
react with cyanides  over a period of time  and cause low  results.

     Sulfides  -    Sulfide  or bisulfide will  interfere  in the
analysis of cyanide  by reacting with the colorometric  reagents.

     The presence of sulfur dioxide or  bisulfite in  the  waste
water sample should  have   no   appreciable   effect  on   cyanide
results.   Detection  limits  on  the order  of  1-4  ug/1  can be
achieved by  the   analytical method   employed,  but the  results
have to be interpreted with regard to  the  possible interfering
components of the sample.

Hexavalent Chromium  (Cr VI) Analysis

     The determination of Cr VI  in  waste water samples  is also
subject  to a  number  of  interferences  which can  take  effect
either during sampling and storage or during analysis.

     Acids  -   Samples taken  and held  at a  very low  pH can
experience the conversion of other forms of chromium  into  Cr VI
causing a positive interference.

     Reducing  agents  -    Samples  containing  sulfur dioxide,
bisulfite, bisulfide, sulfide, ferrous  iron,   and other reducing
agents will result in low values of Cr  VI by converting  it to
trivalent  chromium  (Cr   III) .    Under  these  conditions  the
chromates  originally present would be  included   in  the  total
chromium determination but the analytical results for  hexavalent
chromium would be proportionately low.   (See Reference 52.)

                               63

-------
     The detection limits for Cr VI  using the diphenylcarbazide
colorometric  method  are on  the order of 1-3 ug/1 in the absence
of substances which interfere with color development.

Asbestos Fiber Analysis

     The analysis  of   selected  samples   for  asbestos  fiber
(chrysotile) was conducted by the recommended method utilizing
transmission electron microscopy  with  selected area  electron
diffraction as described by Dr. Charles Anderson   (EPA, Athens,
Georgia) at the Analytical Protocol Meeting  in Denver (November,
1977)  (56).

Conventional and Nonconventional Pollutants

     All  techniques used  for  the  analysis of   BPT   control
parameters  (conventional and  nonconventional  pollutants) were
those  recommended  by  the Agency.  The  list of  approved test
procedures was published in the Federal Register   on October 16,
1973 (38 FR 28758)  and may be also found in Title 40 of the Cod'e
of Federal Regulations  (40 CFR  136).

5.1.4  Quality assurance Provisions

     The Agency  and the  contractor's   analytical laboratories
maintain consistently high standards for accuracy  and  quality
control.  As an in-house requirement,  a minium of  ten percent of
all  samples  are  routinely run  in  duplicate.   Quantitation is
based  on standards  that are prepared  in the  same matrix  as the
samples.  The standards are also checked by participation  in the
EPA  Reference   Sample Program  that   utilizes  a double blind
technique.   (EMSL, Cincinnati,  Ohio,   Office  of Research and
Development.)

     Additionally, outside laboratories are retained  for checks
on  quality by analyzing  split samples  and  running submitted
standards.  Accuracy is also insured by analysis  of a minimum of
fifteen percent  of all samples with   spikes  by the  method of
standard  additions.   The spikes   are  added prior   to sample
preparation and  are carried  through the  entire sample analysis
procedure.

     The contractor's laboratories have consistently maintained
the standards for  laboratory certification which are imposed by
the  State of California.  Certification is dependent upon  the
accurate  performance  of   routine  analyses  on  check  samples
submitted by  the State, as well as on-site  inspections  by  the
State  of   California's Sanitation and  Radiation  Laboratory,
Department  of  Fish and Game,  and the  U.   S.  Environmental
Protection Agency, NEIC, Denver, Colorado.
                               64

-------
     The  quality  assurance  provisions  outlined  in   the  EPA
Protocol for GC/MS Analysis of Toxic  Pollutants  are rigorously
adhered  to  with  one  added  precaution,  namely,  the use  of
internal standards as a means   of  measuring recovery.  Although
not  required  by  the  protocol  for  pesticide  analysis,  this
technique is utilized as  an  in-house quality control requirement
to insure  the accuracy of  results in this analysis.

     The  high  sensitivity  of instrumentation  used in   trace
organic chemical  analysis  dictates  that  contamination of  the
samples from any  possible   source   must be  diligently guarded
against.  Accordingly, only glass  sample containers  with Teflon-
lined lids were   used  and  these were  subjected to a three step
cleaning procedure prior  to use, even  though only new liners and
glass  containers were used.   All  glassware  used for sample
preparation  and  analysis   was  subjected  to  a  dual  cleaning
system.

     The sample extraction  and preparation rooms are dedicated
solely  to  toxic' pollutant   analysis,   and   have  their  own
ventilation systems  that   are isolated  from  the  other  sample
preparation and receipt areas of the laboratories.

     A  documented  system  of  existing  practices,  including
calibrations and  operational  checks  is  maintained to   assure
uniformity of performance and  to serve as a basis for alteration
of standardization intervals.  A  chemist is  assigned  full time
to maintain this  system,   assure strict record  formating  and
controls,  and  to direct  the  quality   control  program  of  the
laboratories.  The primary vehicle of  this system is the quality
assurance  manual containing  the  detailed  procedures  used  in
sample preparation and analysis, and the complete records of all
quality control standards,  blanks, spikes and duplicates.


5.2  SUMMARY OF ANALYTICM.  HESDLTS

                   f
     The results obtained during the screening   and verification
sampling  program are  summarized  in Table 5-2  and Table  5-3.
These  tables  show  the  frequency  and  distribution  of  the
pollutants according to selected plant groupings, concentration
ranges, and subcategories in which the pollutants occur.

     Pollutant frequencies  as shown in   columns 5,  6,  7,  and 8
of Table  5-2  are based  on  the  highest  individual  pollutant
concentration  found  for  each plant's  raw waste during   the
screening and verification  sampling program.


     The  toxic pollutant  asbestos  has  not  been  included  in
either of the two tables mentioned above.  Asbestos

                               65

-------
TABLE 5-2.
            POLLUTANT FBEQUENClf BASED ON SAMPLING PROGRAM RESULTS
            INCLUDING RAW WASTE
Pollutants Detected
Antimony
Arsenic
Beryllium
Cadmium
Chrcrniun
Copper
Cyanide
Lead
Mercury
Nickel
Selenium
Silver
Thallium
Zinc
Benzene
Carbon Tetrachloride
Chlorobenzene
1, 2-Dichloroethane
1/1, 1-Trichloroethane
Hexachloroe thane
1/1, 2-Trichloroethane
1,1,2, 2-Tetrachloroethane
Chloroform
1,2-Dichlorobenzene
1, 1-Dichloroethylene
1, 2-Dichlorcpropylene
2, 6-Dinitrotoluene
Ethylbenzene
Fluoranthene
Bis(2-Chloroisopropyl) ether
Hsthylene chloride
Dichlorobromore thane
Trichlorofluorcroethane
f'^riyrv'HVu tn»"4pA^fraTtf»
Naphthalene
Nitrophenol
Pentachlorophenol
' Phenol
Bi3(2-Etnylhexyl) phthalate
Butyl benzyl phthalate
Di-n-butyl phthalate
tdethyl phthalate
Dimethyl phthalate
Benzo(a) anthracene
, Benzo(a) pyrene
3, 4-Benzof luoroethane
Chrysene
Anthracene
Fluorene
Phenanthrene
Pyrene
Tetrachloroethylene
Toluene
Trichloroethylene
Nitrobenzene
2,4-Dinitrophenol
Pollutant Occurrence Based
on Plant Grouping
5 or <5
Plants






X














X

X
X
X
X

X
X

X
X
X
X
X
X
X
X
X

X
X
X
X
X
X
X
X
X
X
X

X
X
X
>5 but slO
Plants



























• X
























X



>10 Plants
X
X
X
X
X
X

X
X
X
X
X
X
X








X







X









X















Pollutant Occurrence Based on
Concentration Classification (ug/1)
£50
28
38
49
45
20
21

25
46
17
46
45
41
9
6
2
1
2
4
1
2
3
15
1
3

1
7
1
1
11
5
2
2
1
1
2
2
20
3
15
5 ,
2 "
1
1
1
1
1
1
1
1
4
7
3


>50 but
<500
19
12
4
4
13
16

15
2
20
7
7
11
18
1



1



2


1

1
_

3



1

1
3


















>500 but
12,500
4
3

4
9
9

7

8

1
1
14








1







1

1



1

1













1

2
2
>2,500
1



10
7
2
6
5
8



12






















1

1

















                     66

-------
              TKSLE 5-3.  DISTRIBUTION OF lOLHJffiNTS ACCORDING
                          TO SUBCMEGORY 1
Pollutants Detected              Subcategory Numbers Where Pollutants Found

Antimony                      '    AlAut 7, 23,  27, 28,  33
Arsenic                            "    "    "     "   "     "
Beryllium                          "    "    "
Cadmium                            "    "    "     "   "     "
Chromium                           "    "    "     "   "
Copper                             "    "    "     "   "     "
Cyanide                           7
lead                             , Air but 7, 23,.27, 28,  33
Mercury                            "    "    "     "„„  "     "
Nickel                             "    "    "     "   "     »
Selenium                           "    "    "     »   "     »
Silver                             "    "    "     »   "     »
Thallium                           "    "    "     «   "
Zinc                               "    "    "     "   "     "
Benzene                           1,'  3, 4,  10,  11, 25,  32
Carbon Tetrachloride              1,  2
Chlorobenzene                     1,  35
1,2-Dichloroethane                1,  11, 13, 22,  35
1,1,1-Trichloroetfaane             1
Hexachlorcethane                 - 4,11
1,1,2-Trichloroethane             1,  10, 35
1,1,2,2-Tetxachloroethane         1,  3, 4,  10,  13, 15,  19,  21,  22,  25, 32,  35
Chloroform                       24
1,2-Dichlorobenzene               1,  11, 13
1,1-Dichloroethylene              26
1,2-Dichloropropylene             1
2,6-Dinitrotoluene                1
Ethylbenzene                      1,  3, 4,  9, 11, 21, 25, 32
Fluoranthene                      8
Bis{2-€hloroisopropyl) ether      22
Methylene  chloride                1,  4, 8,  9, 12, 13, 19, 21,  22,  25,26,  32, 35
Dichlorobromoroethane              1,  4, 19, 32
Trichlorofluoromethane            1,  4, 25
Clilorodibrornomethane              19, 32
Naphthalene                      1,  32
4-Nitrophenol                     17
Pentachlorophenol                 2,  3, 4,  8, 15
Phenol                            2,  15, 26, 31,  32
Bis(2-Ethylhexyl) phthalate       1,  4, 7,  8, 10, 11, 12,13,15,18,24,25,26,30,31,35
Butylbenzyl phthalate             1,  2, 12
Di-n-butyl phthalate              1,  4, 8,  11,  17, 18,  19,  21,  22,  30, 31,  34,  35
Diethyl phthalate                ' 8,  10, 11, 19,  31
Dimethyl phthalate                12, 31
Benzo(a) anthracene               8
Benzo  (a)  pyrene                  8

    For name of subcategory,  refer to Table 3-2.                (Continued)
2 "All" means subcategory numbers 1 through. 35  of Table 3-2.

                                       67

-------
                            TABLE 5-3.  Continued
Pollutants 'Detected               Subcategory Numbers Where Pollutants Found


3,4-Benzofluoranthane            8
Chr^sene                          8
Jtatihraeene                        8
Fluorene                          8, 12
Phenanthrene                      8
Pyrene                            8
ffetradiloroetfiylene               1, 4, 10, 22
toluene                           1, 3, 4, 10, 11, 15, 18, 32
Trichloroetbylene                 lf 4, 25
                                      6P

-------
concentration is  reported  in  million  fibers per liter   (MFL)
which is not compatible with  the  concentration units  in which
the  other  pollutants have been  reported.  Asbestos was found
in three  plants at  concentration levels of  2.1E8, 2.QE7, and
9.4E4  MFL,  respectively, where E  is  exponential on  base 10.
All  three plants belong to the Chlor-Alkali subcategory.
                               69

-------

-------
                           SECTION 6
      PROCESS AND WASTE TREATMENT? INFORMATION DBVU.QPM1OT
                        AND EVALUATION
6.1  INDUSTRY DATA BASE DESCRIPTION
     Information  and  data on  the  Inorganic chemicals industry
were obtained from a number  of  sources.   These sources included
literature  reviews,  plant  visits,  telephone  contacts,  and
industry responses to the Section 308 Questionnaires,  The type
material gathered from these sources is discussed below.

6.1.1  Literature Review

     A review -of the literature  has been  conducted  to identify
and collect information related  to manufacturing processes, raw
materials, water use, waste water sources, waste water treatment
technology,  raw  waste  characteristics,  and  economic  data.
Relevant information  from   reports,, books,  papers,  conference
presentations and periodicals were identified by computer search
and are presented in the reference  section of this  report.  This
information was  incorporated into a broad  based  assessment of
process and  technology practices aimed  at selecting the best
available treatment technology   and best demonstrated technology
for the various  industry subcategories.  It also provided  the
background reguired for  evaluating  the subcategorization of the
industries.

6.1.2  Plant Visits

     During  the  screening  and  verification  phase  of  this
project, much  information  was  gathered from  individual  plants
relating to production capacity, manufacturing processes, waste
flows,  water   reuse,   waste   water  treatment   systems   and
performance, and  best management practices  (BMP).  The   lead
visits also provided  an opportunity to update and clarify some
of the information given  in the 308 responses.
                              71

-------
6.1.3  Telephone and Direct Contact

     Numerous contacts were  made  with knowledgeable persons in
both industry and government to   gather  and exchange information
concerning all phases  of  this  study.    These sources are cited
in  the text as personal communications.

6.1.4  308 Questionnaire Responses

     The  basis  for  much  of  the work  in  this  study  is  the
responses from industrial  inorganic   chemical  firms  to the 308
data requests.

     Data from 284 manufacturers' responses were utilized by the
project team for  the development  of   appropriate guidelines for
the inorganic chemicals subcategory.  Industrial firms, through
their  compliance  with the needs  of  the  308  Questionnaire,
provided a valuable  industry-wide data base  used extensively in
this analysis.

     Essential  data  elements   from  each  questionnaire  were
extracted for  the purpose of creating  a working data  base for
this report.  Specific elements selected for  this smaller, more-
manageable data base are given in Table 6-1.

     These data  provided  the basis for the  subcategory review
through a profile of  each  industry.    After  compilation of the
questionnaire data, industry totals for capacity and production
(for the respondents) were  available.   In addition,  derivative
quantities  such   as  percent utilization,  effluent  per  ton  of
product,  and  conversion  to  metric   units   were   compiled.
                               72

-------
          TABLE 6-1.   308 QUSTIONNAIRE RESPONSE DATA
                         DATA ELEMENTS
              INORGANIC CHEMICALS GUIDELINES STUDY

Datum Reference            Description           Comments
Manufacturer
Product
Plant
Process
Effluent Treatment
Name
Location
EPA Region

Name
Subcategory

Number of other
Products
Capacity
Production
Age

Name
Volume of Process
Effluent
Volume of Noncontact
Effluent

Type
Permit
Major Pollutants
Confidential
                                                 Inorganic
                                                 Chemicals
Fiscal year
1976
1976
1976
                               73

-------
6.2  PROCESS WASTE SOURCES AND CURRENT TREATMENT PRACTICES


6.2.1  Data Acquisition

     The information presented in this section was obtained from
a  variety  of  published  sources  and  the  available  industry
responses to the 308 Questionnaires as well  as from plant visits
and interviews with industry personnel conducted by  the Agency
and  its  contractor during  the  toxic   pollutant  screening  and
verification program.  The results of visits and interviews are
documented in field notebooks,  interim plant visit reports, and
telephone  communication records  which are  part  of the  rule
making record.

     Plant visits were particularly  useful  for  confirming and
updating the detailed technical  information contained in the 308
Questionnaire responses.  The cooperative  attitude displayed by
industry  greatly   facilitated   the  acquisition  of  reliable
operating data and meaningful sampling results.

6.2.2  Evaluation of Data

     Each  of  the   various  industrial  subcategories in  which
verification  sampling was  conducted was  the  subject  of   an
extensive   evaluation   to   provide  the  technical  basis  for
selecting   candidate   advanced   treatment  technologies   and
developing the related base and  incremental  cost estimations.
In the subsections  which follow,   individual plant descriptions
are  presented   according   to  the  general  format  for  each
subcategory:

      General Process Description
        Description of process reactions and unit operations.
        Inventory of raw materials used.
        Typical process flow diagram.

      Water Use and Waste Source Inventory
        Description of individual plants visited, sampled
         and plant  information from other sources.
        Inventory of water uses for contact and noncontact
         purposes.
        Inventory of raw process waste water sources and
         identification of sampling points.
        Process waste water quality and flow data.
        Solid waste generation and disposal.
                              74

-------
      Control and Treatment Practices
        Description of specific treatment technologies
         and operating facilities.
        Description of the total input to the treatment system
         including sources attributed to other production
         operations and noncontact water (e.g., cooling
         water, etc.).

      Evaluation of Production and Waste Flow Data
        Tabular summary of plant-specific data.
        Waste flows per unit of production  (unit waste flows)
         with the range and average values.
        Solid waste quantities.
        Treatment chemical requirements.

      Process Modifications and Technology Transfer Options

      Best Management Practices (BMP)
        Plant area operations and housekeeping.
        Runoff control.
        Solid waste handling  (e.g., fugitive dust and
         leachate control, etc.).

6.2.3  Model Plant and BPT Treatment System Specification


     The model plant  concept  plays a central  role  in both the
development of alternative treatment system designs  for priority
pollutant removal and for   estimating  the related internal costs
of  such  treatment  in  each  sufecategory.    In  order  to  be
representative  of  a  subcategory,   each   set  of  model  plant
specifications  was  composited  from  a profile  data  summary
derived from the  available information  on  production and waste
flow.

     Based on the typically  achievable waste flow rate per unit
of production,  the model plant was  used as a starting point for
an  appropriately designed and sized BPT  level  waste   water
treatment system.    Certain assumptions   were made  regarding
"the  possible process  variations  and  the    specific  raw waste
sources incorporated  into each model.  In   most  cases,  ,it was
appropriate to assume that the waste flow per unit of production
did not vary over the particular range of production capacities
covered.  Production rates  were selected  in most subcategories
to  represent  the  small,  mid-range  and   large  size  plants
presently in operation.  Small subcategories were represented by
single mid-range production rates   for  the model  plants.   Cost
estimates were  developed  for each  set  of  bas,e level (BPT) and
advanced    level    (BAT/NSPS)     treatment   system    design
specifications.
                               75

-------
     Beginning with  Section  11, the  model  plant  and BPT level
treatment  system  descriptions  and  specifications  for  each
subcategory include the following information:

         Production rates and mode of operation.
         Specific process type and waste sources.
         Waste flow per unit of production.
         Solid waste generation and handling.
         Treatment chemical requirements.


     If applicable, the new source model plant is also described
and the design  specifications given  for   its waste treatment
system.

     The model  plants  do  not  represent  exemplary or specific
existing  plants,   but  are  typical plants  of adequate  design
derived from the range  of  plants  and  treatment facilities found
in the entire subcategory.   For the.purpose  of cost estimating,
it is necessary to   specify  cost rationale, define a set   of
initial assumptions,  and consider  the variability  of factors
such  as  waste  water  flows,  pollutant  load, unit  treatment
process, plant ager etc.  General  assumptions have been detailed
under Section  10  of this report  and  are employed  as the basis
for developing baseline model plant cost estimates presented in
the  subsequent sections dealing with individual industries.

6.2.4  Dissolved Solids in Waste Water Effluents

     Many waste treatment plants discharge  final   effluent into
watercourses which feed  fresh water streams   used  as sources of
water supply by  downstream agencies or  industries.  Groundwater
aquifers which underlie large  portions of the country are tapped
to  supply   fresh   water   through  wells  serving  public  and
industrial water needs.  Saline  wastes discharged into streams
or into unlined lagoons can significantly alter the salt content
(total dissolved solids) of the  fresh water.  Although Federal
regulations  seldom limit  the total  dissolved  solids  or  the
various  ions  such  as  chloride,  sulfate,  bicarbonate,  and
nitrate, these constituents can be of  serious 'concern to local
water  users.

     To protect the mineral quality of ground and surface waters
State  and  local   water pollution  control  agencies typically
establish limits on the discharge  of substances which contribute
sodium,    potassium,    hardness,    chloride,   sulfate,    or
conductivity, which  is  a measure of  total solids  in solution.
This restriction  can  affect the   chemicals chosen for   waste
treatment.     For  example,   alkaline   precipitation   can   be
accomplished by using  lime,  which forms   an insoluble calcium
sludge, or by  adding  caustic  soda,  forming  a  soluble  sodium
salt.

                               76

-------
     In choosing an acid  for  neutralization of alkaline wastes,
it is important to  weigh the overall  effects of chloride   (from
hydrochloric   acid)    and  sulfate   (from  sulfuric   acid),
particularly with  respect to  irrigational  use  of  the receiving
water.

     Chemicals used  in  the  model plant processes  were selected
on the  basis of best  performance, including  consideration of
scaling problems,  which can  be severe when calcium  and sulfate
are  at  saturation  levels.  It  may be  necessary  to alter the
nature of  chemicals used at  a  specific plant, in order to meet
local water  quality requirements.
                               77

-------

-------
                           SECTION 7


  ASSESSMENT OF TECHNOLOGY FOR ADVANCED TREATMENT AND CONTROL


7 .1  INTRODUCTION


    In  the inorganic  chemicals  industry,  pollution abatement
practices vary and a wide range of treatment technologies can be
found,  ranging  from no  treatment  to the  application of highly
advanced technologies for the removal of specific pollutants.

    Until  the  NRDC  Consent  Decree,  industry  attention  was
primarily directed  towards  general pollution problems including
removal of  trace  metals, but not  towards  treatment of over 100
individual  specific  organic  compounds   now  listed  as  toxic
pollutants.     Even  with   the   classical   (conventional  and
nonconventional)  pollutants,   treatment   technology  has  been
directed to removal  down to the part per million level, whereas
now the thrust  is towards part per  billion  level  requirements.
For both  these  reasons, higher level  technologies  are  not in
place in the inorganic chemicals industry,  and it  is necessary to
look into technologies that have been applied in other  industries
or developed at the laboratory or pilot plant scale specifically
for the removal  of  these toxic substances from industrial waste
water,  and determine  whether  they  can   be  adopted  as  viable
technological options.

    A  list of  candidate  technologies  was  compiled from  the
literature, in-house  expertise, and  industry contacts.   These
were,evaluated with respect to:

    1.  Treatment effectiveness

    2.  Cost

    3.  Nonwater pollution environmental effects

    4.  Applications in  the inorganic  chemicals  industry or on
        other   industrial   wastes   with   similar  waste  water
        characteristics.
                               79

-------
    The  anticipation that  few  of the  organic toxic pollutants
would be found in inorganic chemical wastes was justified by the
results  of  the  analytical  programs.    Only  one  industrial
subcategory,  namely,  Chlor-Alkali  production  using  graphite
anodes  had  potentially  significant  levels  of   organic   toxic
pollutants.  As  a result, the initial  search  for  candidate BAT
technologies  became limited  to  treatment technolbgies  for the
thirteen metals, cyanide, and asbestos.

    The  technologies finally  adopted  were  not new  or  untried
technologies since  it was found that most treatment requirements
could  be met  by taking conventional  techniques—for  example,
chemical precipitation—and developing them to  a higher degree of
engineering  and  design  sophistication,  so  that optimum removal
efficiencies could  be achieved.

    The  following  pages  describe  the  theoretical  basis for
treatment systems adopted for BAT application.


7.2  HYDROXIDE PKECIPITATION
    Hydroxide  precipitation is the  most  widely used  technology
for removing trace metals from waste  waters,  with lime  or caustic
soda commonly  used to supply the hydroxide ions.  Under suitable
conditions the metals form  insoluble metal hydroxides which can
be separated from solution.

    The  chemistry of  the  process  is not  simple,  and  must be
understood  for each  metal.   Many  metals  are  amphoteric,  the
optimum  pH  for precipitation varies,  and organic complexes can
interfere.  A  simple form of the reaction may be written as:

    M-H- + 20H- = M(OH)2                                      (1)

    Metal ion  + two hydroxyl ion = insoluble metal hydroxide

    If the pH  is below the  optimum for hydroxide precipitation
    soluble complexes form:

    M-H- + OH-  = M(OH)+                                       ,(2)

    Metal ion  -t- hydroxyl ion = soluble metal complex

    Since most metals have the capability  of coordinating with
other ions or  molecules,  these  simple equations assume that the
hydroxonium ion is the coordinated species.  However,  if organic
radicals are present,  they can form chelates  and mask the typical
precipitation  reactions:
                               80

-------
    M++ +OH- +nR = M (R)nOH+                                  (3)

    Metal ion 4- hydroxyl ion + organic ions = soluble
    metal chelate

    Such  complexes  may require  unusual  treatment  to hydrolyze
them,  and  their  presence  often  explains  why  some  treatment
practices yield relatively poor results.

    Assuming  the  absence  of  organic  complexing  agents,   the
treatment  levels  attainable by  hydroxide precipitation  can be
forecast  from a  knowledge of the pH of the  system.   Figure  7-1
shows  the  theoretical  solubility  of  those  metals  which form
insoluble  hydroxides,   while  Table  7-1   shows   the  solubility
product constants.   For comparison, the values for sulfides  are
also given.

    It  is .clear  from the  range of optimum pH's illustrated that
for  waste  waters  containing  more  than   one  metal,  no   single
optimum pH  exists,  and problems  arise at the  threshold  of  the
alkaline  range , (circa  pH  10)  where  some   metals  have  least
solubility, while others  are at  the point of redissolving as an
anionic species.   For  successful  application as  a  waste water
treatment technology, careful control of pH must  be practiced if
the best removals are to be achieved.

    In  practice  the  solubility of  metallic  hydroxides;  and  the
tendency  for fine  insolubles to remain in suspension may yield
effluents  which  will  not  meet  ug/1  standards,  and  hydroxide
precipitation is  often  supplemented  by  the  use  of  coagulating
agents  to improve solids removal, or sulfide  co-precipitation to
reduce  ultimate solubilities.

    In practice, ,the technology uses unit  process steps which  are
simple, well-established, and well-understood by  the industry.

    Depending on  the quantity of waste  flow, the treatment  can
either  be a batch or continuous  operation, with batch treatment
being favored when waste flows  are  small.   In  batch treatment  the
equipment usually consists of two tanks, each with a capacity to
treat the total waste water volume  expected during the treatment
period.  These systems can be economically designed for flows up
to 50,000 gallons per day (5).

    The   treatment   tanks  serve   the  multiple  functions  of
equalizing  the  flow,  acting  as  a  reactor   and  as  a  settler.
During  operation  the waste water is  stirred, and a homogeneous
sample  is  taken  and analyzed  to determine  the  chemical  dosage
requirements.  The  chemicals are then  added, mixed  and stirred
for about 10 minutes. After  the  reaction  is complete, the  solids
                               81

-------
                                                             Pb(OH),
10
10
  -12
        '0123
            Figure '7-1.
 45    67    8    9   10   11  12   13
Solubility of metal hydroxides and sulf ides
as a function of pH.
14
                                      82

-------
TABLE 7-1.  SOLUBILITY PRODUCTS OF TMCE METALS
       Solubility Product Constant (log K^}
                                            Ethyl Xanthate
Metal	Hydroxide	SoLrLde
Cadmium, Cd
Copper, Cu
4-2
Ferrous, Fe
Lead, Pb
Mercury, Hg
Nickel, Ni
Zinc, En
Omcmium (VT) ,Cr+6
13.6
18.6

15.3
16.1
25.4
14,8
15.7
8.9
26.1
35,2

16.9
26.6
52.2
25,7
25.2
_
13.6
-

7.1
16.9
37.8
11.9
8.3
-.
                      83

-------
are allowed to settle for a few hours.  The clear liquid is then
decanted and discharged.  Settled sludge is retained to serve as
a seed for crystal growth for  the next batch, but must be drawn
off periodically and disposed of, usually in  a chemical landfill.

    For layer  daily  flows  a  typical  continuous  flow  treatment
scheme consists of a flash mixer,  flocculator, settling unit with
sludge storage tank, and, in some cases, a filtration system.

    The ability  to  separate the  solids  from  the  waste water is
important.    Metallic  hydroxides  tend  to   be  gelatinous  and
separ-ate poorly  in gravity  separators.  Finely suspended solids
tend to pass out with  the effluent  and increase the total metal
content.  Thus,  improvements  in precipitation applications have
been directed toward fine solids  removal,  and this is  reflected
in  the  addition of  various filtration  systems and the  use of
flocculant aids as improved levels of  treatment.

    Hydrated lime suspensions are more commonly  used than caustic
soda as the hydroxide source because they are cheaper.   However,
if  there  is  sulfate  ion present in the waste water, gypsum will
be formed:

    Ca  (OH) 2 +  (S04)— =  CaS04 + 20H-                      (4)

    Hydrated lime -f sulfate ion = calcium sulfate  (gypsum) +
    hydroxyl ions

    This  increases  the  sludge  produced,  may  cause  scaling
problems in pipelines,  and may  clog a dual media filter.   Using
caustic soda  is  more expensive^' but it generally feliminates the
scaling problem.  Total dissolved solids in  the  form  of sodium
salts are  increased  in  the  caustic soda  tre'ated  waste waters.
Although  low  concentrations  of  sodium ,are  not  regarded  as
polluting, high levels can make drinking  water unpalatable, limit
the use of water for agriculture, and promote degradation of the
structure  of arable  soils.   Thus,  where high  total  dissolved
solids are of concern,  lime would be the preferred neutralizing
agent.

    This  treatment  technology is  wide'ly  applied in  treating
industrial waste waters.   Industries  that are  using  hydroxide
precipitation include:

        Inorganic Chemicals
        Plating and Metal Finishing
        Mining
        Textiles
        Steel and Iron
        Non-Ferrous Metal Processing and
        Electronics
                               84

-------
    Better  than  99 percent  removal of  trace metals  have been
reported  in the  literature  with  final  concentrations  in  the
treated effluents ranging from  sub ppm  to low  ppm  (see Tables 8-1
through 8-10) .


7.3  PERRITE COPKECIPITATION
    An   interesting   variation  on   the  theme   of   hydroxide
precipitation is a process developed in Japan for the removal of
heavy metals  from acidic  waste water.   The process,  known as
ferrite  coprecipitation,  has  the  potential  for  producing  a
marketable residual by converting the  metal ions  in solution into
insoluble ferromagnetic oxides  or ferrites which can  be removed
magnetically or by filtration (5).  The treatment is  applied by
adding  a ferrous salt  to the  metal-bearing waste water,  then
neutralizing  and  oxidizing  the  complex  heavy  metal-ferrous
hydroxide precipitate by  aeration  to  form  the  stable ferrite
coprecipitate.   Particle  sizes are  reported  to  be  relatively
large   and  sludges   formed   can  be   safely  disposed,  of  by
landfilling.

    Although extensive performance data have not been developed,
the  information  available  indicates  that  very  hiqh  removal
efficiencies can be achieved for most  of the  common heavy metals,
including mercury and hexavalent chromium.  The  method has not
been considered here  as  an available  technology  due to the lack
of sufficient information on  chemical dosing requirements, energy
requirements,  and performance  in situations  similar  to those
found in  the  inorganic chemicals industry.  In  connection with
waste water treatment in the Titanium  Dioxide Subcategory for the
sulfate  process,  the wastes  contain  considerable  amounts  of
ferrous iron from the processing of ilmenite ore and the current
practice  of  neutralization  and aeration  may  involve  the  same
chemistry as the ferrite coprecipitation process.


7.4  SDLFIDE PRECIPITATION


    The  basic  principle  of  sulfide  treatment  technology  is
similar to that of hydroxide precipitation.  Sulfide is added to
precipitate the metals as metal sulfides and  the  sludge  formed is
separated  from  solution  by gravity   settling  or  filtration.
Sodium  sulfide  and  sodium  bisulfide  are  the  two  chemicals
commonly  used,  with  the choice  between these  two precipitation
agents  being strictly an economic consideration.

    Metal sulfides form according to the following equation:

    M-H- -f Na2S = MS + 2Na+                                   (5)


                               85

-------
    Metal ion + sodium sulfide = insoluble metal sulfide +
    sodium ions

    Figur.e 7-1  shows the  theoretical solubility  of  the metals
that  form  insoluble  sulfides,  while   Table  7-1  shows  the
corresponding solubility product constant.

    The   major   problem   in  applying   sulfide  precipitation
techniques is  associated with the  toxicity of  sulfides.   This
warrants both care  in application and post  treatment  systems to
remove excess sulfide.   Pretreatment involves raising the pH of
the waste stream to minimize evolution of hydrogen sulfide gas.

    A recently  developed and patented process  to eliminate the
potential  hazard of  excess  sulfide  in  the  effluent  and  the
formation of gaseous hydrogen sulfide uses ferrous sulfide  as the
sulfide source  (6) .   The fresh  ferrous  sulfide  is prepared by
adding sodium sulfide  to ferrous sulfate.   The ferrous sulfide
slurry formed  is added  to  a waste  water to  supply  sufficient
sulfide  ions  to  precipitate  metal   sulfides ' which  have  lower
solubilities than ferrous sulfide.  Typical reactions are:

    FeS + Cu++ = CuS + Fe++                                  (6)

    Ferrous sulfide + copper ion = insoluble copper sulfide 4-
    iron ion

    FeS + Ni  (OH) 2 - Fe(OH)2 + NiS       ^                   (7)

    Ferrous sulfide + nickel hydroxide = ferrous hydroxide +
    insoluble nickel sulfide

    A detention time of 10-15 minutes is sufficient to allow the
reaction to go to completion (7) .  Ferrous sulfide  itself is also
a  relatively   insoluble  compound.     Thus   the  sulfide  ion
concentration is  limited by the solubility  of ferrous sulfide,
which  amounts  to  about 0,02  mg/lf   and  the  inherent problems
associated with conventional sulfide precipitation are minimized
(8).

    One other advantage of this process is that  if chromium  (VI)
is present, it will  also be  reduced at the pH of  normal operation
(8 to 9)  and precipitate as the trivalent hydroxide (Cr III).

    Treatment systems for  sulfide precipitation  are  similar to
those used for  hydroxide precipitation.   A continuous treatment
scheme generally consists of a pH adjustment tank, flash mixer,
flocculator, settling units with sludge storage,  and a dual media
filter..

    Before the addition of sodium sulfide or bisulfide the pH of
the incoming  wasteflow  is  adjusted  to pH of  7-8  in  the first


                               86

-------
reaction  tank  to  reduce  the  formation  of  obnoxious hydrogen
sulfide gas.   The chemicals  are  then added to  the flash mixer
where they are thoroughly mixed with  the waste water.

    After  the  flash mix,  the  waste  water  passes  through  a
flocculating basin where the floe agglomerates and settles in the
settling  unit.   The overflow from the settling unit generally
passes  through  a filter  to remove any fine precipitates.  Any
excess  sulfide  will  need to  be  removed before final  discharge.
This  can  be  achieved   either  by  aeration  or   other chemical
oxidation techniques.

    Sulfide  precipitation  is  being  practiced in the inorganic
chemicals  industry,  mining  industry,  textile   industry,  and
nonferrous metal processing industry.  Most of  the  Chlor-Alkali
industry  is  applying this technology to remove  lead or  mercury
from its waste streams.

    Literature   citations   on   the    efficiency   of   sulfide
precipitation  (9, 10, 11)  indicate  that most results are  in the
sub  ppm  range,  and  that  sulfide  treatment  is  superior  to
hydroxide treatment  for  the removal of  several trace metals.   A
recent  report  concluded that, with no  complexing agents  in the
waste,  the following effluent quality can be achieved  (11).

                     Meta1s Coneentration

                   Cadmium            0.01 mg/1
                   Copper             0.01 mg/1
                   Zinc               0.01 mg/1
                   Nickel             0.05 mg/1
                   Chrome  (total)     0.05 mg/1

    Adding ferrous "sulfide as a polishing  step to remove residual
metals  appears  to  be  a  promising,   economical  technology.
Although there is no full-scale treatment  system  operating in the
inorganic chemicals  industry,  pilot  studies on chrome  pigment
waste indicate  that  this process  is  superior to sulfur  dioxide
reduction followed by hydroxide precipitation  (12).


7.5  THE XANTHATK PROCESS


    The use  of xanthates  for the removal  of  metals from waste
streams appears  to  be  a new, promising technology  for treating
metal-bearing waste  waters.   Xanthates contain functional groups
capable  of   forming  insoluble  complexes  with  metals,  and  the
sludge  so formed can be  separated by  conventional means.
                               87

-------
    Xanthates can be generated by mixing starch or  cellulose with
carbon disulfide in a  caustic medium.   Three types of xanthates
have been proven in bench pilot scale studies  to be effective in
removing cadmium,  chromium (III), copper,  iron,  lead, mercury,
nickel,  silver  and zinc  from industrial  waste waters  (13-20).
These are:

        Soluble starch xanthate with a cationic polymer,

        Insoluble starch xanthate, and

        Fibrous cellulose xanthate

    The general removal mechanism is as follows:

    2  [ROCS (=S)Na] + M-H- =  [ROCS(=S)2M] +  2Na+                (8)

    Xanthate + metal ion =  insoluble metallic  xanthate +
    sodium  ions

    where R = starch or cellulose

    Unlike  hydroxide precipitation,  this process  is reported to
be effective in removing metals over a wide pH  range of 3 to 11,
with an optimum range  between  7 and 9.

    Brass mill waste waters, lead  battery effluent, circuit board
rinse   waters,  electroless   copper   plating   rinse  waters,
pyrophosphate  electroplating  rinse  waters, and  copper  etching
rinse waters were studied in  a pilot plant  with insoluble starch
xanthate  as  the  complexing  agent   (20).    This pilot  study
demonstrated that the  xanthates can either  be  added to a reactor
to mix  with the waste waters or be applied as a precoat  on  a
pressure filter (20) .  Results of these pilot studies showed that
metals were reduced to below  50 jug/1 (ppb).

    Another study indicated cellulose xanthate is as effective as
starch xanthate  in removing trace metals.   The following table
summarizes  the result  of  the  study  with  a cellulose xanthate
dosage of 90 mg/1 and  a contact time of 30 minutes (18-19):

                      Concentration, mg/1

          Metals            Influent        Effluent
Cadmium
Chromium
Copper
Iron
Lead
Nickel
Zinc
1.35
0.30
1.6
3.1
3.9
2.4
1.0
0.027
0.022
0.06-0.14
0.08-0.36
0.008-0.021
0.077
0.03-0.04
                               88

-------
    This study also concluded that cellulose xanthate is superior
to starch xanthate  in  terms  of sludge settling characteristics,-
filterability,'and handling.

    Xanthate may  also  be used  as  a complexing agent to prevent
the formation of  soluble  anions from insoluble amphoteric metal
hydroxides.

    The xanthate process  is a relatively new technology, and the
reagent compounds are not  yet available in commercial quantities.
More  information  is needed  on dosage rates  in continuous flow
operations.  Potentially  the metals can be recovered by leaching
the xanthate complex with nitric acid,-but metal recovery has not
been demonstrated yet.   Sludge  disposal problems may  arise  if the
sludge complex is unstable and, if xanthates are to be generated
on site,  care  will be  needed  in handling  the hazardous   carbon
bisulfide.
7.6  ION EXCHANGE


    Ion  exchange  is  a  chemical  reaction  between the  ions in
solution and the ionic sites on an exchange resin.  Many natural
solids   (e.g.,  soils,  proteins,  and  zeolites)  exhibit   such
exchange  characteristics.    However, synthetic  resins  are the
predominant  ones  used for  ion exchange  applications  in modern
industrial technology.   These resins contain  functional groups
that  can react with  the  ions  in solution*.  Depending on  these-
functional groups, the resins can be classified into:

        Strongly acidic cation exchanger,
        Weakly acidic cation exchanger,
        Strongly basic anionic exchanger, and
        Weakly basic anionic exchanger.

    Cation exchangers  are  qapable of exchanging with cations in
solution.  Strongly  acidic  cation exchangers contain functional
groups  such   as  sulfonates,  (-SO3H and  -SOSNa), while weakly
acidic exchangers have functional groups derived from carboxylic
acids,  (-COOH and -COONa).

    Anionic  exchangers  are used to  exchange with the anions in
solution.  In general, strongly  basic  exchangers contain  amine
functional   groups   {-R3NOH  and  -R3NC1),  and  weakly   basic
exchangers contain ammonia functional groups (-NH3OH and -NH3C1)

    When  the  functional  groups  are  used up in the reaction, the
resins  can  usually  be  regenerated.   Cationic  resins can be
regenerated by sodium chloride, hydrochloric acid, sulfuric acid
or sodium  hydroxide.   Anionic resins are  regenerated  by sodium
                               89

-------
hydroxide,   ammonium   hydroxide,   sodium   carbonate,   sodium
chloride, or hydrochloric acid.

    The  exchanger  can  either  be added  to the  waste waters  in
batch operations or be packed in a fixed bed or column. Fixed bed
is  by  far  the  more  effective   and  hence more popular.   The
operation  generally   follows  a   four-step  cycle:    exchange
(service), backwash,  regeneration, and rinse.

    During the  exchange step,  the reaction between the ions  in
solution  and  the  ionic sites in the resin  takes  place  as the
waste  water  passes  down the  bed.    The  reaction  is generally
regarded   as   a   result  of  electrostatic   attraction   (20).
Therefore, the size of the hydrated ion and the charge on  the ion
are  the  determining  factors  for  the  exchange  reaction.    A
trivalent  ion  is  attracted  more strongly than a  bivalent ion
which  is  in turn attracted  more  strongly  than a monovalent ion.
For  ions with  the  same charge,  the smaller  hydrated  ion   is
capable  of moving closer  to  the  exchange  site,  and  is  thus
favored.

    Many  synthetic  resins  contain  functional groups  that are
selective to certain metals. For  example,  a resin manufactured  by
a European company reacts preferentially with mercury (Hg+4-) and
mercuric  chloride   (HgCl+)  ions  according  to  the following
equations:

    2RSH + Hg++ = RSHgSR + 2H+                                (9)

    Resin + mercury ion = insoluble resin  complex +
    hydrogen ions
                                                      I
    RSH -f HgCl+ = RSHgCl + E+                               (10)

    Resin •*• mercuric  chloride ion = insoluble resin complex +
    hydrogen ions

    The exchange reaction is governed by the  law of mass  action.
During  the reaction,  the  affinity of  the resin for the two ions
is  so  great  that   essentially  all  the mercury   or  mercury
chloride-resin  complex formation equilibria  are shifted toward
the formation of Hg++ and HgCl+  which are rapidly removed.   A 5
ppb residual mercury concentration in  the effluent  is  achieved  by
this process (22).

    After  all  the exchangeable sites  in  the  resin are used up,
the bed is backwashed by passing  clean water through to loosen  up
the  bed and to  remove  any  fine particulates  that are  trapped
inside  the bed.

    After  the backwash cycle  the resins can be  regenerated with
the appropriate regenerant.


                               90

-------
    RSHgCl + HC1 = RSH + HgC12                               (11)

    Insoluble resin complex + hydrochloric acid = regenerated
    resin + mercuric chloride

    One attractive feature of the ion exchange process is  that it
concentrates  the  metals   in  the  regeneration  step,  and   thus
provides a potential for their recovery.   However,  if  recovery is
not feasible, this creates  a  secondary  stream which needs to be
treated.

    A recent study found that sodium alumino  silicates (zeolites)
might be a low-cost exchanger that can be discarded after a  one-
time use (22).  This would eliminate the  regeneration  step.   On a
batch study with a five-minute contact time,  cadmium  and mercury
were removed to below 10 ppb.  Thermodynamic  considerations  show
this  exchanger  to  have a  high  affinity for  cadmium, copper,
mercury, nickel, silver, zinc, cesium, and barium.

    Ion exchange  is a  proven technology that  can  reduce metal
concentrations  down  to  low levels.   However this technology is
used only in limited industrial pollution abatement applications
because  of   the  high  cost  associated   with   the  process.
Consequently,  ion exchange  has  not been  recommended  in   this
report for BAT  technology.


7.7  REDOCTIOS PROCESSES
          «

    Many  metals  can exist   in  solution  in  several  oxidation
states, and it may be necessary to convert from a higher valency
state to a lower one in  order  to apply a  given chemical  reaction.
The classic example  is  chromium,  which  as the trivalent chromic
ion will precipitate as  the hydroxide in  alkaline solution, while
the hexavalent  chromate or  dichromate ion will not.  The latter
needs to be reduced  if  precipitation  is  to occur.

    Hexavalent  chromium (e.g.,  CrO4= and Cr207=)   is toxic and
soluble.  The most efficient  way of removing this from solution
is a two-step process of reduction followed by precipitation.

    Chromium  (III)  is  much less  toxic  than  chromium  (VI), and
forms an insoluble hydroxide  which  can be removed from solution
by settling and filtration.

    A number of chemicals are used for the reduction of  chromium.
Most common  are sodium  bisulfite,  sodium metabisulfite, sulfur
dioxide and ferrous salts.  The  reduction is  accomplished readily
at low pH with  these reagents.  Typical  reduction reactions  are:
                               91

-------
    3SO2 + Cr207= 4- 2H+ = 2Cr-t--H- 4- 3804= 4- H20               (12)

    Sulfur dioxide + dichromate ion 4- hydrogen  ion = trivalent
    chromium ion 4- sulfates and water

    3303= + Cr207= + 8H+ = 2Cr+++ + 3SO4 = +4H20             (13)

    Sulfite ion + dichromate ion + hydrogen  ion = trivalent
    chromium ion +• water

    6Fe++ 4- Cr207= + 14H+ = 2Cr+++ + 6Fe+++  + 7H20           (14)

    Ferrous ion + dichromate ion 4- hydrogen  ion = trivalent
    chromium ion + ferric ion + water

    The  reduced  chromium and  the ferric  ions produced  in  the
third equation will  exist as the  soluble  sulfate at acid pH's.
If the pH  is above 5,  the reaction rate is  drastically reduced,
and although dithionite will effect reduction at neutral pH's, it
is very costly and its use may be contraindicated.

    After  the  reduction step, lime or  caustic  soda is added to
raise  the  pH  to  8.5-9.0.     Trivalent   chromium   will  be
precipitated.

    Cr-H-+ + 30H- = Cr(OH)3                                   (15)

    Trivalent chromium ion + hydroxide  ion = insoluble
    chromium hydroxide

    The  theoretical  solubility  limit  of chromium  hydroxide is
above 0.02 mg/1 (8).  It is  reported that applying sulfur  dioxide
to a pigment waste consistently reduces Cr.(VI) and Cr(T) to 0.5
mg/1 and 1.5 mg/1 respectively  as  30-day averages (9)   (10) .  By
applying  ferrous sulfide  to a plating  waste with  an  initial
Cr(VI) concentration of  128 mg/1  and  Cr(T)  concentration of  153
mg/lr  an  effluent  quality  of  less  than  0.05 mg/1  of  either
species is achieved (12).

    A  one-step  precipitation-reduction  process  using  sodium
bisulfide  is used in a  dichromate  plant to remove chromium'  from
its  waste  water.   An  effluent quality 'with  less  than  1  mg/1
Cr(VI)r and less than 5 mg/1 Cr(T) was  reported (3)..

    One  other  common  reduction process  is  the  application of
sodium borohydride  to  reduce metals  in waste  streams.  Sodium
borohydride  is a mild  but effective  reducing agent (3),  and is
currently used in some chlor-alkali plants to reduce the  soluble
mercury ion to metallic mercury which  is removed from solution by
carbon adsorption:
                               92

-------
    4Hg++ + BH4- + 80H- = 4Hg +' B (OH) 4- + 4H20               (16)

    Mercury ion + borohydride ion 4- hydroxyl ion = insoluble
    mercury metal + borate ion + water

    A mercury level of 0.01 mg/1  in the final effluent has been
reported (3).

    Sodium  borohydride  is  also  reported  to  be effective  in
removing silver, mercury, gold, lead, and cadmium  (5).  However,
this technology is only being applied in limited cases, the cost
of the  chemical being the major  drawback.   The  cost  of sodium
borohydride was $16.00 per pound in 1978 (23).


7.8  OXIDATION PROCESSES


    The oxidation of organic substances is generally carried out
by thermal processes such as wet  oxidation  and incineration, or
by biological  processes  such  as  the activated sludge process,
trickling filters, biodiscs, and aerated lagoons.

    Incineration  is  actually  a  combination  of  oxidation  and
pyrolysis.  Both  involve chemical  changes  resulting  from heat.
Oxidation involves  actual  reaction  with oxygen, while pyrolysis
refers  to  rearrangement or  breakdown of  molecules  at  high
temperatures in the absence of  oxygen.   There are five types of
incinerators  available commercially.   These  are rotary kiln,
multiple hearth,  liquid  injection,  fluidized bed, and pyrolysis
(24) .   A  minimum  temperature  of 1000 degrees  C and  a residence
time  of two seconds  is  required  for  the  reaction  to proceed.
This  process  has  been  shown  to   be  successful  in  reducing
pesticides to harmless molecules  (25) .

    Wet oxidation is a process  in which an  aqueous waste can be
oxidized in the liquid phase in a closed, high-temperature, high
pressure vessel.  This reduces some of the problems (such as air
pollution  from exhaust  gas),  inherent in  incineration.   Wet
oxidation has been used for  a variety of wastes  including pulping
waste  and  acrylonitrile liquor   (26).   A percent  reduction in
excess  of 99.8 of some of the toxic pollutants  has been reported
(27).

    Thermal  oxidation  processes  are not expected to  have much
application in  the  inorganic  chemicals  industry, mainly because
of the  high energy cost  required and the  low level  of organic
contamination found in the wastes.

    The application of chemical oxidation to industrial wastes is
well  established  for cyanides,  sulfite,   ammonia,  and  other
                               93

-------
harmful  species in  dilute  waste  streams  (phenols, mercaptans,
polysulf ides, etc.).  Common  chemicals  used as oxidizing agents
included  chlorine,  hypochlorite,  hydrogen  peroxide,  potassium
permanganate, ozone,  and  chlorine dioxide.  Air  and oxygen are
also used.

    The most widely used chemical oxidation technology applicable
to the inorganic chemicals industry is the oxidation of cyanide.
The oxidation reaction  between chlorine and cyanide is believed
to proceed in two  steps as follows:

    CN- + C12   = CNC1 + Cl-                                 (17)

    Cyanide + chlorine = cyanogen  chloride + chloride ion

    CNC1 + 2OH- = CNO- + Cl- + H20                           (18)

    Cyanogen chloride + hydroxyl  ion = cyanate ion  + chloride
    ion + water

    The  formation of  cyanogen  chloride  (CNC1)  is essentially
instantaneous.  The second reaction,  the  formation of cyanate,  is
accomplished most  rapidly and completely at a pH  of 10 or higher
(9r 28).  A detention time of 30 minutes to two hours is usually
allowed .

    The  cyanates   can  be further  decomposed into  nitrogen and
carbon dioxide by  excess chlorination or acid hydrolysis:

    2CNO- -I- 40H- + 3C12 « 6C1- +  2CO2 + N2 + 2H2O           (19)

    Cyanate + hydroxyl ion + chlorine = chloride  ion +
    carbon dioxide + nitrogen + water
CNO-
2H2O = CO2 + NH3 + OH-
                                                             (20)
    Cyanate + water = carbon dioxide + ammonia + hydroxyl ion

    The first  reaction  can be accomplished in about one hour if
the pH  is adjusted to  8.0-8.5.   Acid hydrolysis  usually takes
place at pH 2-3 and care must be taken  to avoid the liberation of
the toxic cyanogen chloride as a gas.  Hydrolysis is not usually
the chosen option.

    Other common chemicals used to oxidize cyanide include sodium
hypochlorite,  ozone,  and hydrogen  peroxide.   The  reaction for
sodium hypochlorite is essentially the  same as for chlorine.  For
ozone  and hydrogen  peroxide,  the  oxidation  step  proceeds  as
follows:
                               94

-------
    O3 + CN- = O2 + CNO-                                     (21)

    Ozone + cyanide = oxygen + cyanate ion

    H202 4- CN- = CNO- + H2O                                  (22)

    Hydrogen peroxide + cyanide = cyanate ion + water

    The advantage  of  using  these two oxidizing reagents is  that
no dissolved solids are added  to the waste water.  In addition,
excess chlorine is not discharged.

    A patented process uses hydrogen peroxide and  formaldehyde to
decompose cyanide at about 120  Deg.  F.  This  has the advantage of
precipitating cadmium and zinc simultaneously  (9).

    Alkaline chlorination is  currently  being practiced  in one
hydrogen  cyanide  production plant.   Laboratory  studies  in the
plant indicated that  the  presence of ammonia in the waste water
reduces the efficiency of cyanide removal.   It is  well  known  that
ammonia  reacts  with  chlorine  or   hypochlorous  acid to   form
chloramines:

    NH3 + HOC1 = NH2C1 + H20                                 (23)

    Ammonia -f hypochlorous acid = monochloramine + water, etc.

    NH2C1 + HOC1 = NHC12 + H20                               (24)

    NHC12 + HOC1 = NC13 + H2O                                (25)

    If excess  chlorine is  added,  chloramines can  be converted
into nitrogen oxide(s):

    2NH3 + 4HOC1 - N2O + 4HC1 + 3H20    .                  ,   (26)

    This equation is  not exact  because  the  final form of nitrogen
oxide  is.  believed to  be a  mixture of  nitrous  oxide, nitrogen
dioxide and nitric oxide.

    The treatment  of  cyanide by chemical oxidation is currently
practiced in the following industries:

    Inorganic Chemicals (Hydrogen Cyanide Production)

    Mining

    Plating

    The free cyanide level after treatment  is generally below 0,1
mg/1 (9) .
                              95

-------
7.9  MEMBRANE PROCESSES
    Membrane  processes have  emerged  in the  last decade  as a
promising new  technology  for  the treatment  of  saline water and
waste  waters.    A  membrane  is  a semi-permeable  barrier which
allows the transport of some molecules (ions)  and retains  others.
The  driving force  can either  be  electropotential  differences
(electrodialysis)  or pressure  difference  (reverse  osmosis and
ultrafiltration).   The major  application of these processes has
been  the  desalination  of  brackish  water  and sea  water.   More
recently,  these  have  also  found  application  in  a  number  of
industries, including;

    Mining
    Electroplating
    Metal Finishing
    Printed Circuit Board Manufacturing
    Battery Manufacturing
    Pulp' and Paper
    Food Processing

    In electrodialysis, an even number of alternating anion and
cation selective membranes are  placed  between  two  electrodes.
When  current  is applied the anions  are  attracted  to the anode,
and  cations  are attracted to  the cathode.   In the  process  of
migration, the cations pass through  the cation-permeable membrane
and are blocked  by  the anion-permeable membrane.   Likewise, the
anions pass through the anion-permeable  membrane and  are  blocked
by the cation  membrane.   This  results  in  alternating  paths  of
purified water and concentrated- reject (Figure 7-2) J

    The  electrodialysis membranes  are  made  very  thin  and are
assembled in stacks.  The  flow path is the active portion of the
cells.  Pretreatment  to.remove suspended materials is absolutely
essential.  Other materials in the waste  feed  that may lead to
membrane fouling  include high  organic content,  calcium sulfate,
and  certain complex   ions  such  as   ZnCl-  which  can .partially
convert the anion membrane to  the cation form,  with  significant
loss  in system performance (28).

    As ionic concentration decreases,  the electroconductivity of
the water also decreases, making  it less efficient to remove the
remaining salt.   Most operations  do not  produce a product water
of less than 500 mg/1 total dissolved  solids.

    Reverse osmosis  (RO) and ultrafiltration  (UP) are similar in
basic concepts.   Both are pressure-driven  separation processes
that  employ  high-flux semi-permeable  membranes  operating under
                               96

-------


              PEOOOCT
               WATER
Figure 7-2.  Electzodial^is process.
                  97

-------
dynamic flow  conditions (29).   In  contrast to electrodialysis,
these involve  the  transport of  solvent,  not  solute,  across the
membrane.

    Osmosis is a process in which solvent from a dilute solution
is  transported spontaneously  across a  semi-permeable membrane-
into a  concentrated solution.   By  applying  enough  pressure to
overcome  this  osmotic  pressure,   reverse  osmosis,   i.e.,  the
passage of solvent  from  a concentrated  solution to  a  dilute
solution  through  a   semi-permeable  membrane,   occurs.     The
operating pressure  of  reverse osmosis  units  is  usually between
350  and  600 psi.   Ultrafiltration  usually operates  at  a much
lower  pressure  (5  to  100  psi).    The predominant  transport
mechanism  is   selective  sieving through pores.   The membrane
retains high molecular weight dissolved  solids such as synthetic
resins, colloids,  and  proteins.   The  upper and  lower molecular
weight  limit  is   generally   defined   as  500,000  ~and  500
respectively.

    Membranes  are  usually  fabricated in flat  sheets  or tubular
forms.  The most common material is cellulose acetate but other
polymers  such as  polyamides  are  used.    There  are  four  basic
module  designs:    plate-and-frame,  tubular,   spiral-wound,  and
hollow  fiber.   Table  7-2   is  a comparison between  the various
reverse osmosis  modules.   Membrane processes are  effective in
removing  (concentrating) inorganic  and organic substances from a
wastestream.   Usually extensive  pretreatnient  is required  to
reduce  the   suspended  solids  and  control  pH.     There  are
uncertainties  about" operation  efficiency, membrane  lifetime,
rejection  specificity, and other  factors.   If  recovery  is not
feasible, the concentrated reject must be disposed or treated by
other methods.   The high operation and  capital  cost limits the
widespread application of these  technologies.  For these reasons
membrane  technique  is  not  recommended  as  a BAT  technology for
this industry.


7.10  ADSORPTION


    Adsorption is  a surface phenomenon  in  which a substance is
accumulated on the surface  of  another substance.   Sorption of a
solute on  a solid  surface  is  widely used in pollution abatement
practices.  The  term "adsorbate" refers  to the  substance being
concentrated,  and  the  term "adsorbent" refers  to the material
that provides the surface.

    Activated  carbon  is  the  prevalent  adsorbent  used.    Both
inorganic  and  organic  substances  are  known  to  be  removed
effectively by activated carbon. Certain chlor-alkali plants are
currently using activated  carbon as a polishing  step to remove
mercury.

                               98

-------
                                TABLE 7-2.  COMPARISON OF REVERSE OSMOSIS  CONCEPTS
vo





Plate-and-Frame
Large tubes
Spiral
,
Polyamide hollow
fine fibers
Cellulose acetate
hollow fine
fibers

Packing
Density
2 3
(ftyftj)
150
50
250

5000

2500


Water Flux
at 600 psi
(gal/
day/ft2)
10
10
10

1(400 psi)

3(250 psi)


Water Output
Per Unit
Volume (gal/
day/ft2)
1500
500
2500

" 5000

7500


Parasitic Pressure
Sodium
Chloride
Rejection
Very good
Very good
Very Good

Fair

Good


Losses (psi) Useful
Peed
Channel
30
50
10

10

10


Product
Channel
30
10
50

50

50


pH
Range
2-8
2-8
2-8

0-12

3-7


Ease of
Cleaning
Fair
Very good
Good to
very good '
Fair

Fair


     Source;   Weber,  Physicochemical Processes, 1972,

-------
    Activated carbon  is  made by charring basic substratesr such
as wood, cokef  coal,  shell, husks, etc., at  600  degrees  C in a
controlled atmosphere, where oxygen is kept low by adding carbon
dioxide or steam.  This  process drives  out  volatiles, leaving a
porous carbon lattice in an "activated" state.

    Activated carbon  can be obtained  in powdered  and granular
form.  Powdered carbon is about 50-70 microns  in diameter, and 90
percent should pass through  a  300-mesh  screen.   Granular carbon
is about 0.1-1 mm in diameter,  and because of  this is  three times
more expensive than powdered carbon.

The application involves the passage of the waste waters through
a contact bed.  When  the bed is exhausted,  the carbon is either
regenerated  or  sent  to  landfill.   It  is  economical for large
plants  to  regenerate the carbon.   This  can  be done either  by
thermal regeneration'in a rotary kiln or multihearth incinerator,
or by  chemical  regeneration by using  oxidizing agents  such  as
hydrogen peroxide or acids and  bases.

    The  application of  carbon  adsorption  has  been  mainly  in
organic waste treatment.  Recently, there are studies  indicating
the  effectiveness of  carbon  adsorption  in  removing mercury,
cadmium, cyanide, chromium,  lead, nickel,  zinc,  arsenic,  and
copper  (30, 31).

    An  interesting  development in carbon technology  is  its use
after the waste water is- ozonized.  This combination (known  as
Bacteriologically Activated  Carbon or BAG)  has proved effective
in treating  otherwise biologically inactive  organic  compounds.
The process involves chemical modification of  the organics by the
ozone.  Maintenance of an aerobic region on the carbon allows a
biologically activated film to  develop and the modified organics
are further  treated  by  a mixed  process  of  biological oxidation
and carbon adsorption.   The system has  the advantage of being a
potential  add-on  to  existing  BPT systems,  and  should  be cost
effective  since  it has  been  found  that the carbon  only needs
regeneration at infrequent intervals.

    No  industrial applications of this  technology  are known,
although research is under way  (32).

    Bacteriologically  Activated  Carbon  is  a  very   attractive
potential  BAT  technology  for  the  removal  of  organic  toxic
pollutants from  waste streams, although no  application  to 'the
industry subcategories studied  in  this report was found.
                               100

-------
7.11  FLOORIDE REMOVAL


    The conventional  method of treating fluoride-bearing wastes
is  to  precipitate the  fluoride  as  calcium  fluoride  by  the
addition of lime.  The reaction is:

    Ca(OH)2 + 2F- = CaF2 +  20H-                              (27)

    Hydrated lime + fluoride ion = insoluble  calcium fluoride +
    hydroxyl ion

    Using this process alone,  it is difficult to remove  fluoride
to below  8  mg/1 due  to  the solubility of  calcium fluoride  (9,
33).   Adding  alum with  the lime  generally improves the  removal
efficiency.  Fluoride ions  are removed as follows:

    A1(OH)3  + P- = A1(OH)2F + OH-                           (28)

    Aluminum hydroxide + fluoride ion =
    aluminum monofluorohydroxide + hydroxyl ion, etc.

    A1(OH)2F + F- = Al(OH)F2 + OH-                           (29)

    A1(OH)F2 + F- = A1F3     + OH-        .                  (30)

    Complexed fluorides  are also adsorbed  to  some  extent on  the
aluminum hydroxide surface and removed in the  coagulation process
(33).   Large amounts  of  alum (5000  mg/1)  are required to reduce
the fluoride concentration  to  below 1 ppm.

    Activated alumina has been shown to be effective in  removing
fluoride  and  arsenic in  waste water   (34)  and  fluoride   from
drinking  water  in municipal  water  treatment practice  (35-38).
Typically, the fluoride content of raw water  can be reduced  from
about 8 to 1 ppm (38)  .  Application of activated alumina  to  high
fluoride  industrial wastes  shows  that a low ppm effluent can be
achieved   (39),  although   high   capital  and  operation  costs
generally limit the wide application of this  process.

    Certain  process  operations  used  -in  the  manufacture  of
inorganic fluoride compounds involve the use of sulfuric acid and
starting materials which contain  silicate  or  borate impurities.
This   may  lead   to   the  formation   of   wastes    containing
fluorosulfonate, hexafluorosilicate or tetrafluoroborate  complex
ions.    Although  tetrafluoroborate  is  usually  a  very minor
constituent and  the  hexafluorosilicate is  readily hydolyzed in
treatment systems, the  fluorosulfonate  ion is fairly  stable  and
presents a serious  problem  where low levels of total fluoride  are
required.   The  lime  precipitation method  is  not effective in
removing the fluorosulfonate and the effectiveness of  adsorption
techniques is not known.


                               101

-------
7.12  CHLORINE KEMO¥AL
    The   removal  of   residual  chlorine   (in  the   form  of
hypochlorite) in industrial waste water is normally accomplished
by  the  addition of  sulfur  dioxide or  a  related reducing agent
such  as  sodium bisulfite  or  sodium  metabisulfite.    Typical
reactions are shown  in Equations 31 and 32.

    SO2 4- OC1- 4 H20 = H2SO4  4 Cl-                          (31)

    Sulfur dioxide 4 hypochlorite ion 4 water =  sulfuric acid
    + chloride ion

    Na2SO3 4- OC1-     = Na2SO4 + Cl-                         (32)

    Sodium sulfite + hypochlorite ion = sodium sulfate 4
    chloride ion

    Alternatively,   hydrogen   peroxide,   although   relatively
expensive  may  also  be  used  for  dechlorination according to
Equation 33.

    H202 4 OC1- = H20 4 O2 4 Cl-                             (33)

    Hydrogen peroxide 4 hypochlorite ion = water 4 oxygen 4
    chloride ion

    In the chlor-alkali industry, certain waste  water streams may
have a sufficiently high loading of - chlorine to  warrant recovery
of  the  product  by  air  stripping, steam stripping, or extraction
by carbon tetrachloride.  In some locations, a market exists for
sodium  or  calcium  hypochlorite  solutions which can be generated
by  treating  the  tail gases  with caustic soda or lime.   This may
serve as a means for disposing of waste chlorine which cannot be
economically  recovered.    As  alternatives  for  waste  chlorine
disposal, the streams may be treated to form  the  hypochlorite and
then decomposed  thermally or  catalytically.   These technologies
are  discussed  in  Section  11   dealing  with   the  chlor-alkali
industry.  Chlorine residuals  remaining  after the recovery and/or
decomposition  steps  have  been  taken  would   be  amenable  to
treatment with reducing agents such  as sulfur dioxide, bisulfite,
or hydrogen peroxide as described above.
                               102

-------
                          SECTION  8


      TREATABILITY ESTIMATES AND LONG-TERM DATA ANALYSIS




8.1  THE DEVELOPMENT OF TREATABILITY ESTIMATES
     The review of technological treatment options applicable to
the removal of toxic pollutants  has  led to the conclusion that
the  particular contaminants  found  in  the  raw process  waste
waters of the  subject  industries  can be effectively controlled
by the proper  application of fairly well-known and demonstrated
techniques.  In order  to proceed  from a general discussion and
description  of techniques  to  a  detailed evaluation  for each
subcategory of the levels  of removal that can be  expected,  a
summary is now presented of selected treatability data for the
13 toxic metals.

     The treated  waste concentrations and removal efficiencies
reported in  the literature  are   assumed  to  represent  the best
performance  characteristics that can  be  obtained under  the
specified operating  conditions.    The  treatment  technologies
considered can thus be assigned a set of optimum conditions and
best performance  estimates  for  removal  of the particular toxic
metals that  are   amenable to  treatment.   Taking  each  metal in
turn,   Tables  8-1 through 8-10  give  the  initial and  final
concentrations, the removal efficiencies,  and the pH conditions
for  different  treatment  technologies.    The  best  performance
estimates for metal removal are derived from the tabulated data
and  are  utilized  in  turn  as  the bases  for  making  long-term
achievable performance estimates.  The  sequence  of analytical
steps is:

     1.  Review and analyze applicable performance data.

     2.  Estimate  best  performance  under  optimum  treatment
         conditions.

     3.  Estimate   achievable   performance   under   expected
         industrial operating conditions.
                              103

-------
8-1.        WKEER TREKIMENr OPTIONS AND            DMA.         -
      MTEQDNY AND ARSENIC REMOVAL

Treatment ^technology
Antdmony
Lame/Filter
Ferric cMoride/Filter
M.wiyS'il'ter
Arsenic
Lime Softening
Sulfide/Filter
Lime (260 mg/1) /Filter
Lime (600 rng/1) /Filter
Ferric sulfate
Ferric sulf ate
Ldbme/Ferric Chloride/
Filter
Activated alumina
(2 mg/1)
Activated carbon
(3 mg/1)
Ferric Chloride
Ferric Chloride
pH

11.5
6.2
6.4
—
6-7
10.0
11.5
5-7.5
6.0
10.3
6.8

3.1-3.6

_
—
Initial
Concen-
tration
(mg/D

0.6
0.5
0.6
0.2
-
5.0
5.0
0.05
5.0
3.0
0.4-10

0.4-10

0.3
0.6-0.9
Final
Concen-
tration
(mg/1)

0.4
0.2
0.2
0.03
0.05
1.0
1.4
0.005
0.5
0.05
<0.4

<4.0

1 0.05
<0.13
Removal
(%)

28
65
62
85
-
80
72
90
90
98
96-99+

63-97

98
-
References

40
40
40
9,
9,
41
41
42
41
9,
43

43

9,
9,




10
10




10




10
10
                              104

-------
8-2.              TREMMENT OPTIONS MID             DftTA         -
      BEEYLLIUM -KND CMMIOyi REMOVAL

Treatment Technology pH,
Beryllium
Lime/Filter - 11.5
Cadmium
Line (260 mg/1) /Filter 10.0
Lime (600 rag/1) /Filter 11.5
Lime Softening 5-6.5
LiirB/SxxLficte 8.5-11.3
Perrons Sulfide (Sulfex) 8.5-9.0
Ferrite cx>precipitation/ neutral
Filter
Initial
Concen-
tration
Cmg/1)

0.1
5.0
5.0
0.44-1.0
0.3-10
4.0
240
Final
Concen-
tration
(mg/D

0.006
0.25
0.10
0.008
0.006
<0.01
0.008
jRernoval References
(%)

99.4 40
95 41
98 41
92-98 8
98+ 44
99+ 7,8,11
99+ 5
                               105

-------
TASTE 8-3.   WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
             COPPER REMOWL

Treatment Technology
lajre/Filter
Lime (260 ng/1) /Filter
Line (600 mg/1) /Filter
Ferric sulfate/Pilter
Line
Lima
Mum
Idne/Sulfide
Ferrous sulfide (Sulfex)
Ferrous sulfide (Sulfex)
Ferrite Coprecipitation/
PH
8.5-9.0
10.0
11.5
6.0
>8.5
9.5
6.5-7.0
5.0-6.5
8.5-9.0
8.5-9.0
_
Initial
Concen-
tration
(mg/1)
3.2
5.0
5.0
5.0
10-20
3.0
3.0
50-130
3.2
4.0

Final
Concen-
tration

-------
TSBLE 8-4.  WftSTE WKEER TRES3WEOT OFTICKS SND EERFQi8fflHCE DKEA SWMBJSf -
            CHBCMIOM III MJD GHKMIUM VI SEMCWSL
Treatment Technology
Chromium
Lime (260 rag/1) /Piltar
Line {600 mg/L) /Filter
Ksduct ion/Litre
Reduction/Liire
lame Softening
Lime/Filter
lane
lams
Ferrite aaprecipitatlon/
Filter
Iterric sulfate
Ferric sulfate/Filter
Chrcmiura VI
Activated carixsi
(pulverized, Pitts-
burgh type EC)
Same as above
Activated carbon
(granular)
Ferrite cxjprecipitation
Sulfur dioxide reduction
Bisulfite reduction
pa
10,0
11.5
7-8
7-8
10.6-11.3
7-9
9.5
9.5
	
6.5-9.3
	
3.0
2.0
6.0
	
. —
	
Initial
Concen-
tration
Omg/l)
5.0
5,0
140 (as
Cr VI3
1300 (as
Cr VI)
—
—
15
3.2
25
	
5.0
10
10
3
0.5
	
	
Final
Concen-
tration
(mg/1)
0.1
0.1
1.0
0.06 CrIII
0.15
0.05
0.1
<0.1
0.01
	
0.05
1.5
0.4
0.05
not
detectable
0.01-0.1
0.05-1.0
Rancfval Eef erenees
(%)
98 41
98 41
	 9,10
	 3,9,10
98+ 46
	 47
	 45
	 45
n,,,,,^, e
984- 46
99 41
85 48
96 48
98 41
	 5
	 9,10
	 9,10
                         107

-------
TOBT.E 8-5.  WSSTE WATER IKEATMENT OPTIONS MID PERFOKffiNCE Dftim SOMMARY
            EEM) KEMOWAL

Treatment Technology
Lime (260 mg/1)
Lime/filter
Idme (260 mg/1) /Filter
Line (600 nng/1) /Filter
Ferrous sulfate/Filter
Sodium hydroxide (1 hour
settling)
Sodium hydroxide (24 hour
settling)
Sodium hydroxide/Filter
Sodium carbonate/Filter
Sodium carbonate/Filter
Sodium carfDonate/Filter
Ferrous sulfide (Sulfex)
Ferrite coprecipitation/
PH
10.0
8.5-9.0
10.0
11.5
6.0
5.5
7.0
10.5
10.1
6.4-8.7
9.0-9.5
8.5-9.0
___
Initial
Concen-
tration
(mg/1)
5.0
189
5.0
5.0
5.0
	
	
1700
1260
10.2-70.0
5.0
189
480
Final
Concen-
tration
(mg/D
0.25
0.1
0.075
0.10
0.075
1.6
0.04
0.60
0.60
0.2-3.6
0.01-0.03
0.1
0.01-0.05
Removal
(%)
95.0
99.9
98.5
98.0
98.5
— _
	
99+
99+
82-99+
99+
99.9
99.9
References
41
5
41
41
41
10
10
49
49
10
9,10
8
5
  Filter
                                      108

-------
TKBLE 8-6.                        OPTIONS AND                         -
                           MERCURY II BEMDVKL
Treatment Technology pH
Sulfide
Sulfide 10.0
Sulf ide/Filter 5 . 5
Sulf ide/Filter 4 . 0
Sulf ide/Filter 5.8-8.0
Ferrite coprecipitation/
Filter
Activated Carbon
Activated Carbon/Mum
Activated Carbon -
Initial
Concen-
tration
fog/1)
0.3-50.0
10.0
16.0
36.0
0.3-6.0
,6.0-7.4
0.01-0.05
0.02-0.03
0.06-0.09
Final Removal
Concen- (%)
tration
fog/l)
0.01-0.12
1.8 96.4
0.04 99
0.06 99.8
0.01-0.125 87-99.2
0.001-0.005 ,99.9
<0.0005
0.009
0.006
Reference!
9,10
50
50
50
50
5
9,10
46
50
                                  109

-------
TABLE 8-7.  WASTE WATER TREATMENT OPTIONS MID PERFORMANCE DATA SUMMARY -
            NICKEL REMOVAL

Treatment Technology pH Initial
Concen-
tration
(mg/1)
Line 8.5-9.0 75
Lime (260 mg/1) /Filter 10.0 5.0
Line (600 mg/1) /Filter 11.5 5.0
Caustic Soda/Filter 11.0
Iferrous sulfide (Sulfex) 8.5-9.0 75
Herrite coprecipitation - 1000
Final
Concen-
tration
(mg/1)
1.5
0.3
0.15
0.3
0.05
0.20
Rsnoval References
(%)
98 8
94 41
97 41
49
99.9 8,11
99.9 5
                                   110

-------
TABLE 8-8.  WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA SUMMARY -
            SILVER REMOVAL

Treatment Technology pH
Sodium hydroxide 9.0
Ferric sulfate (30 mg/1) 6-9
Lime Softening 9.0-11.5
Chloride precipitation -
Initial
Concen-
tration
(mg/1)
54
0.15
0.15
105-250
Filial
Concen-
tration
(mg/1)
15
0.03-0.04
0.01-0.03
1.0-3.5
Removal
(%)
72
72-83
80-93
97+
References
13
46
46
9,10
  (alkaline chlorination
  in the presence of
  cyanide)

Ferric chloride/Filter

Sulfide precipitation
6.2

5-11
0.5
0.04      98.2      40

 -      very high    9,10
                                     111

-------
T&BEE 8-9.  WASTE WATER TREATMENT OPTIONS AM) PERPOMSNCE DATA         -
            SEESJIDM MED THALLIUM REMOVAL

Treatment Technology
Selenium
Ferric chloride/Filter
Ferric chloride/Filter
Mum/Filter
Ferric sulfate
Ferric sulfate
IdirMFilter
Line/Filter
Thallium
Lime/Filter
Ferric chloride/Filter
MiWFilter
pH
6.2
6.2
6.4
5.5
7.0
11.5
11.5
11.5
6.2
6.4
Initial
Concen-
tration
(mg/D
0.1
0.05
0.5
0.10
0.10
0.5
0.06
0.5
0.6
0.6
Final
Concen-
tration
(mg/1)
0.03
0.01
0.26
0.02
0.03
0.3
0.04
0.2
0.4
0.4
Removal
75
80
48
82
75
35
38
60
30
31
Beferences
40
40
40
51
51
40
40
40
40
40
                                     112

-------
8-10.  WASTE WATER TREATMENT OPTIONS AND PERFORMANCE DATA. SUMMARY -
                         ZINC EEMOVSL
Treatment Technology
Line/Filter
Lime (260 mg/1)
Lime (260 mg/1) /Filter
Lime (600 rag/1)
Lime (600 mg/1) /Filter
Lime/Filter
Sodium hydroxide
Sulfide
Ferrous sulfide (Sulfex)
Ferrite ODprecipitation
pH
8.5-9.0
10.0
10.0
11.5
11.5
-
9.0
_
8.5-9.0
_
Initial
Concen-
tration
(mj/D
3.6
5.0
5.0
5.0
5.0
16
33
42
3.6
18
Final
Concen-
tration
(rag/1)
0.25
0.85
0.80
0.35
1.2
0.02-0.23
1.0
1.2
0.02
0.02
Removal
93
83
84
93
77
-
97
97
99+
99+
References
8
41
41
41
41
5
13
5
8,11
5
                             113

-------
     The  third step  involves  the  consideration .of treatment
system  variables  under  full-scale  operating  conditions  in
industrial situations  where the design objective would be the
simultaneous removal of  several  waste  load constituents.  Each
industry  designs  for  maximum removal  and/or recovery  of the
major   process-related  waste   substances  and  utilizes  an
appropriate  technology   which  is  both   reliable  and  cost
effective.  Optimum  treatment conditions  for  the removal  of a
particular pollutant, can rarely be achieved consistently and any
given set of conditions  will  be  somewhat  less than  optimum for
most, if  not all, of  the treatable constituents.  In any well-
operated production facility the  normal variations in production
rates,  raw  material quality,  the  desired product  mix  in some
cases,  and  contact  water  use requirements  may cause severe
hydraulic and pollutant  load  input excursions which at best can
be moderated by effective equalization in  the treatment  system.
This  is considerably less  of a problem in batch treatment than
with  a  continuously  operating  system.    The latter  requires
continuous  feedback monitoring  for  pH  control and  chemical
dosag*e  in  order  to  maintain  the  effluent  quality  within
acceptable  limits  for  a  number  of parameters.    Under these
conditions, the 30-day averages derived from the actual  treated
effluent monitoring data  (NPDES, etc.)  would equate to what has
been  identified  in Step  3   above  as  the  estimated  30-day
achievable  performance   using  the   same  general  treatment
technology.

     A  statistical  evaluation of long-term  monitoring  data is
described below  and the results are  presented  in  Appendix  A
where various  derivative quantities  such as long-term averages
and  standard  deviations  are  tabulated  and  the  bases  for
formulating  the   variability   factors    applicable   to  each
subcategory are explained in detail.

     For   each  nonexcluded   subcategory,  a  step   by  step
presentation of the logic used to  develop effluent limitations
is  given, based on performance  estimates for  30-day   average
concentrations for specific pollutants.   When available, these
concentrations  are  based  on  industry  monitoring  data.   When
long-term data are not available from  industry,  as  is  the case
with most toxic pollutants, achievable concentrations are based
on the treatability of these pollutants as discussed in Section
8 and summarized in Table 8-11.

     Variability factors applied to these  concentrations  for the
development of monthly average and daily   maximum  limitations
are based on statistical analysis of  long-term data  as presented
below and in Appendix A.   In many  cases, due  to  the  limited
amount of long-term data available, variability factors observed
in  one  subcategory  are  applied  in  other  subcategories where
similar treatment technologies are practiced.
                              114

-------
8-11.                               30-MY          FOR THE


Antimony, Sb
Arsenic V
Beryllium, Be
Cadmium, Gd
H Copper, Cu
Chrcmium III,
Cr+3
Lead, Pb
Mercury II,
Hg
Nickel, Ni
Silver, Ag
Selenium, Se
Thallium, Tl
Zinc, Zn
Lime
Settling
0.8-1.5
0.5-1.0
0.1-0.5
0.1-0.5
0.5-1.0
0.1-0.5"
0.3-1.6
Lime
Filter
0.4-0.8
0.5-1.0
0.01-0.1
0.05-0.1
0.4-0.7
0.05-0.5
0.05-0.6
Final
Sulfide
Filter

0.05-0.1

0.01-0.1
0.05-0.5

0.05-0.4
Concentrations (mg/1)
Ferrite
Coprecip- Soda Ash Soda Ash Alum
itation Settling Filter
Filter



<0.05
<0.05
0.01
0.20 0.4-0.8 0.1-0.6
0.01-0.05 <0.01

0.2-1.5
0.4-0.8
0.2-1.0
0.2-1.0
0.5-1.5

0.1-0.5
0.2-0.4
0.1-0.5
0.1-0.5
0.4-1.2

0.05-0.5
0.05-0.2


0.02-1.2




0.2-0.5
0.02-0.5
                               (continued)

-------
       TSBIE 8-11   continued
CFs


Arsenic V, As
Chromium VI,
Cr+6
Jfercury II,
Hg
Ferric Activated
Chloride Carbon
0.05-0.5 0.3
0.1
0.01
Final Concentrations (mg/1)
S02 Bisulfite Lirte/FeCl2
EMuction Ifeduction Filter
0.02-0.1
0.01-0.1 0.05-0.5

Mkaline
Chlori-
nation



Silver, Ag

Selenium, Se

Thallium, Tl

Cyanide  (Free),
 0%
0.05-0.1

0.05-0.1

  0.7
                                                                                             0.1-0.5

-------
8.2  THE USE OF HISTORICAL POLLUTANT DATA


8.2.1    Determination  of  Limitation  Guidelines  Based  Upon
Historical Performance

     In cases  where  there has been long-term monitoring of the
pollution levels  in  the effluent  stream discharged  by a plant,
it is possible to assess  in-plant treatment performance through
analysis of  historical data  that  has been collected for this
purpose.  The appropriateness of standards constructed from data
collected  from  a  single  plant  performance   is,  of  course,
dependent on the plant's  current performance in relation to the
performance of  other plants  in  the  manufacturing subcategory.
As   economically   feasible   alternative   waste   treatment
technologies become  available, pollutant  discharge guidelines
need to be reviewed  and revised to'reflect these  advances.

     Statistical  analysis  of  historical monitoring data  is
required to  assess  a  plant's ability to  discharge  within set
guidelines.  To perform this  analysis certain assumptions must
be  made  regarding  the   nature  of  applicable statistical  or
probabilistic  models,  the  constancy of  the  operation  of  the
treatment facility,  and the quality of the monitoring methods.

     The  statistical  analyses contained  in  this  development
document  belong  to  either of  two  principal  types:  those for
daily observations of  pollutant concentrations, and the others
for 30-day average pollutant  levels.

     Tables  in Appendix  A  provide  a  summary of  traditional
descriptive   measures,   i.e.,   number   of   observations(No),
mimima(Min),    arithmetic   average(Avg),   maxima(Max),   and
coefficient   of  variation(CV).    In  addition,  a descriptive
statistic, the variability factor, pertinent to the development
of performance standards  for pollution  monitoring,  is included.
These tables,  prepared for both daily measurements  as  well as
30-day  averages,  are  statistical  summaries derived  from data
offered by industry  in response to Section 308 Questionnaires,
Data .in  these  tables  are representative  of  currently achieved
pollutant  discharge  performance levels  in the  several  plants
presented.

     Formulation   of  variability   factors   to  be   used  in
determination  of  limitation  guidelines based  upon  historical
performance was accomplished by employing standard statistical
analysis  of  the  data resulting  from long-term  monitoring  of
effluent stream discharges of plants in the inorganic chemical
manufacturing  subcategory.  In  the  following paragraphs  are
presented  details   of  the   theory   and  derivation  of  these
statistical  procedures,   and  of  the  resulting  formulae which
                              117

-------
relate  variability  factors  to  estimated  long-term  parameter
averages,  standard  deviations,  coefficients  of  variation,  and
"Z-values"  computed  from the normal  probability distribution.
These details are given both for the analysis applying to daily
maxima criterion and for that applying to 30-day averages.

     The term "variability factor"  is  used  in referring  to the
multiple of  the  long-term  average which is used in formulating
performance  standards.    This  factor allows for  variation in
pollution level measurements due to sampling error, measurement
error,  fluctuations  in  the  amount  of  the  pollutant  in  raw
materials, and other process variations.

     In the recording of actual data, as reported by industrial
point sources in their responses to 308 Questionnaires, certain
data values  were entered as "less  than"  detectability limits.
In  these  cases,  the  sample  of  monitoring  data  has  been
"censored"  in the  process  of  data  recording  since  only  the
threshold  value  has  been  retained   (i.e.,  if  a  pollutant
concentration was  reported  as <0.050  mg/1,  the values of 0.050
mg/1 was used).   In the statistical  analysis of monitoring data,
censored  values  were included  with  measured  values  in  the
sample.  This practice provides a reasonable approach, both for
assessing  industry's capability  to perform  and environmental
concerns for valid pollutant limitations.

     First,  since  censoring  was  done  only  for  "less  than"
bounds,  any bias  from  their  inclusion would  cause  a  slight
increase in the long-term average, moderately affecting  (in the
direction on leniency toward industry)  the estimate of long-term
average pollution levels.

     On the other hand,  the use  of censored  values combined with
measured values tends to reduce the variability slightly  (or in
the direction of less leniency toward  industrial point sources).
For  illustration,  if the sample  consisted solely of censored
values,  the  estimated  long-term  average  might  be  shightly
overstated.   Nevertheless,  the  point  source  ought have  no
difficulty  with  the threshold  or  detectability limit as   a
performance  guideline,  since  none  of  the  historical  data
exceeded that limit.

8.2.2  Assumptions Concerning Daily Pollutant Level Measurements

     In  the  formulation  and  calculation  of  the  following
performance   standards,   individual   sample   measurements   of
pollutant   levels   were  assumed   to  follow   the   lognormal
distribution, a  well known and  generally accepted statistical
probability  model  used  in  pollution  analyses.   Under  this
assumption the logarithms of these measurements follow a normal
probability model.   It was also  assumed that monitoring  at  a
                              118

-------
given plant  was conducted  responsibly  and in  such  a way that
resulting   measurements   can   be   considered   statistically
independent and amenable to standard statistical procedures.  A
final assumption  was  that, treatment facilities  and monitoring
techniques had  remained substantially  constant  throughout the
monitoring period.

     As an indication of the appropriateness of  this assumption,
the  following  plot  of  the  cumulative  distribution of  daily
pollution concentration  logarithms  on  normal  probability paper
is illustrated in Figure 8-1.

     The linearity of  the  cumulative plot indicates the degree
to  which actual  monitoring  data  are  in  agreement with  the
theoretical lognormal model for their distribution.

     In  addition,  Figure  8-2,  reproduced  here from ,a report
prepared by industry for consideration by  EPA, also demonstrates
the validity of the lognormal assumption  for daily data.

     In the analysis of daily data, the inherent variability of
measured pollutant levels in the effluent stream from inorganic
chemical  manufacturing  processes   must   be   incorporated  in
calculating upper limits for daily  pollutant  discharge levels.
Even well treated and controlled plants  may experience some days
when an atypically high level of pollutant discharge is present
in their  waste  stream.  Such  high  variations may be due to a
variety  of  factors,  such  as  short-term  maladjustments  in
treatment facilities,  variation in flow  or pollutant load, or
changes in the influent stream.  To allow for this variability,
performance standards must necessarily be set above the plant's
long-term  average   performance  and   occasional,   infrequent
excessive discharges  permitted.  Since pollutant  discharge is
often expressed in terms of  average level,  it is convenient to
describe standards of performance and allow variability in term
of  multiples of  this  average.   Such  a  method of  computing
standards as functions of multiples  of average level performance
is explained below.  The ratio  of  the  pollutant standard level
to  the  estimated long-term  average  is   commonly  called  the
"variability factor".

     This   factor  is   especially   useful   with   lognormally
distributed pollutant levels because its value is independent of
the  long-term  average,  depending   only  upon   the .day-to-day
variability of the process and the expected number of excessive
discharge periods.  For a lognormal population, the variability
factor  (P/A),  the performance  standard P,    and the long-term
average A,  are related by:

     sLn(P/A) = S'(Z - S'/2)
                              119

-------
K
o
            30.0
            20.0
10.0

 8.0

 6.0

 5.0

 4.0

 3.0



 2.0
             1.0
              0.01     0.1      1   2   5   10   20 30 40 50 60 70 80   90  95   98 99
                                                                                99.9
                                                  PEKCENTAS!
                  Figure 8-1.    Cumulative distribution of daily concentrations of mercury in treated
                                effluent from plant #251.

-------
H
       0.30
       0.20
0.10

0.08


0.06



0.04


0.03



0.02





0.01
         0.01     0.1   .512    5   10   20 30 40 50 60 70 80    90  95   98 99
                                                                               99.9
                                             EEK3MTAGE
            Figure 8-2.   Cumulative distribution of daily concentrations of cyanide in treated
                          effluent from plant #765.

-------
     where

     1.  "In"  represents  the  natural  logarithm  (base  e)  of a
numerical quantity.

     2.  S1   is   the  estimated   standard   deviation  of  the
logarithms   of  pollutant  level  measurements.        In  the
calculations  which follow, S1  is computed  by  the statistical
procedure known as the "method of moments".

     3.  Z   is  a  factor  derived  from  the  standard  normal
distribution.  Z  is chosen to give  performance limitations which
provide a  balance  between appropriate consideration  of day to
day variation in  a  properly operating plant  and the necessity to
insure that  a plant is functioning properly.

     The value of  Z used  for  determining performance standards
for daily  measurements  of pollutant concentration is chosen as
Z=2.33.  This Z-value corresponds  to the 99th percentile of the
lognormal  distribution meaning   that  only  1  percent  of  the
pollutant observations taken from  a plant with proper operation
of  treatment facilities  would be  greater  than  the performance
standard, P.  This percentile  is equivalent  to allowing a plant
in normal operation 3 to  4 exceedances per year.

Calculation of Variability Factors

     As mentioned  above,  development of variability factors for
daily pollution level measurements was  based on the  assumption
that these data,  (XI,X2,...Xn), follow a  lognormal distribution.
When this distribution is  not a precise model, lognormally based
procedures tend to somewhat overestimate  variability and produce
liberal standards which act to the benefit of permittees.

     Following  this  assumption,  if  Yi=ln(Xi),  where  ln(Xi)
represents the natural logarithm or log base e of the pollution
measurement,  then  the  Yi;     i=l,2,...,n  are   each normally
distributed.   If A1  and S1 are the mean,and standard deviation
of Y=ln(X) respectively,  then  the  probability is k percent that
an  individual Y  will not exceed A'+ZS1, where  Z is the k-th
percentile of  the  standard normal  distribution, e.g.  Z=2.33 is
the 99-th  percentile  of  the  standard normal distribution.   It
follows  that  A'4-ZS1   is   the  natural  logarithm  of  the  k-th
percentile of X  and that  the  probability is k percent  that X
will not  exceed  a performance standard  P=  exp(A'+ZS') .   It is
also known that the  average value of X  is A= exp(A'+S'(S'/2)).
The variability  factor VP,  is  obtained by  dividing P  by A,
hence,

     VF = P/A = exp(S' (Z  - S'/2  )) , and

     ln(VF)  = ln(P/A) = S'(Z - S'/2)

                              122

-------
     To estimate the W for  a particular  set of monitoring data,
where  the  method of moments- is used, S1  is  calculated as the
square root of ln(l_.0 +  (CV)  ) , where the  sample1 coefficient of
variation, CV  -  S/X,  is the ratio of sample standard deviation
to sample average.

Example Calculation of Variability Factors From Long-Term Data

     Given the following descriptive statistics for a particular
parameter, as might be  found  for lead  fmg/1)in Appendix A.

     No     Min    Ayg     Max      CV

     128   0.002   0.068   0.100    0.609

     Calculate the estimated  standard deviation of logarithms

     (S1)2 = In  (1.0 +  0.6092) = 0.315

     S1 = 0.56

     Then

     ln(P/A) = 0.56(2.33 - 0.56/2) - 1.148

     The variability Factor VF is,

     VF = P/A =  exp{1.148) =  3.15

     The performance standard P;

     P = A(VP) = A (P/A) = (0.068) (3.15)  =  0.2*14

     The statistical distributions relevant for the analysis of
daily  data are shown in Figure 8-3.

     The  statistical   interpretation  of  P,  the  performance
standard,  is  that  one estimates  that  99   percent   (for  the
selected Z=2.33  value corresponding to the 99-th percentile) of
the daily pollution  level measurements  will  not  exceed P.   For
large  data sets, P is roughly equivalent to an upper 99 percent
confidence bound for an individual daily measurement.

8.2.3   Assumptions  Concerning 30-Day Average  Pollutant Level
Observation

     While  individual  pollution  level  measurements  should be
assumed   lognormally  distributed,  that   assumption   is  not
appropriate  when  analyzing   30-day  averages.    These   averages
generally are not distributed as lognormal  quantities.  However,
for  averages  of daily  (lognormal)  measurements,  a statistical
                              123

-------
               NOKMAL DISTRIBUTION
               (MODEL DENSITY OF LOGARITHMS OF POLLUTION VALUES)
                                        ln(P) = A1 +  2.33(S')
                                           Y  = ln(X) = Logarithm  (mg/1)
                 A1

         LOGNORMAL DISTRIBUTION
          (MODEL DENSITY OF
         POLLUTION VALUES)
                                                   X(mg/l)

                                           _ P(Performance Standard)
                       "—A (Long Term Arithmetic'Average)
         SAMPLE DISTRIBUTION OF N MEASUREMENTS
             (LONG TERM MONITORING DATA)
                                       X(mg/l)
                X" (Sample Average)
   Note: (a)   S1 is estimated as (S1)- = Cln(l + CV2)J
Figure 8-3.
  t?         2


 X^ ZX/N

Statistical distribution for daily pollution measurements.
                     124

-------
principle, the  "Central  Limit Theorem", provides the basis for
using the normal probability model.  Therefore, the methods used
in computing historical  performance characteristics for 30-day
averages  differ  from those used  for daily  samples.   In this
case,  the  sample  coefficient  of  variation is   the  primary
determinant of  the variability  factor,  and  there is no need to
resort   to   logarithmic  transformation.     Examples   of  the
appropriateness   of   this    assumption  is   the   cumulative
distribution of 30-day averages shown in Figures  8-4 and 8-5.  A
straight  line  plot  here on normal  probability paper indicates
validity of this model.

     Under these conditions,  the 30-day average values (XI, X2,
     Xm), for m months behave approximately as random data from
a  normal distribution with mean  A and  standard deviation S"_.
Therefore, the probability is  k  percent  that a monthly average X
will not exceed the performance standard P, where

     P = A + Z(S")

     The variability factor is

     VF = P/A o 1.0 + Z(S"/A)  and will be estimated by

     VF = 1.0 + Z{CV)

    Where

     1.  z  is  a  factor  derived  form  the  standard  normal
distribution.   If one wishes  a  performance  standard based upon
expecting  95   percent   of  monthly  averages   to  be  within
guidelines, then Z=1.64 should be used.

     2.  CV is the estimated coefficient of  variation of the 30-
day  averages  and is  computed by Sx/X,  the  ratio  of  standard
error  of sample means  to  overall or grand  average of monthly
.averages.

Calculation of Variability Factors

     A  sample  calculation of  30-day average variability factor
is shown  below.   The  descriptive  statistical  data  'is  for zinc
(mg/1) from Appendix A.

     No     Min    Avg     Max      CV

     30     0.010  0.151   0.815    1.03
     VP - 1 + Z(CV) = 1.0 -f 1.64(1.03) = 2.69

     P - A(VF) *  (0.151) (2.69)  - 0.406


                              125

-------
-»    3.00
!


      2.00
      1.00
       0.01       0.1   0.5  1  2     5  10   20  30 40 50 60 70  80   90  95  98  99      99.9
                                             EEFCENTAGE
           Figure 8-4.    Cumulative distribution of  30-day averages of total cyanide in treated
                         effluent from plant  #782.

-------
   t-I
   H
H
to
-J
   fl
          700 --FFOT
          600
          400
          200
          100
                                            . .1
                                                  ^
Ml
                                                       :•:,
                                                         1
                                                        an
                                                            :::
                     1
           0.01    0.1       125   10  20  30 40 50 60 70  80    90  95   98 99
                                         99.9
           Figure  8-5.    Cutmilative distribution of  30-day averages of ammonia  in treated
                          effluent from plant #782.

-------
     Given the previous descriptive statistics for a particular
sample, one obtains the performance standard  P, by multiplying
the mean  of the  30-day averages  in  the data  set by VF,   An
appropriate  statistical  interpretation'  is  that,    for  the
selected value of Z=1.64 corresponding to the  95th percentile of
a normal distribution, one estimates that 95 percent of the 30-
day average pollution level measurements will not exceed p.

     Figure  8-6  shows  the  relationship  between  the  normal
probability model  and frequency distribution of  set  of 30-day
averages.
                              128

-------
                    NORMAL DISTRIBUTION
           DENSITY OF  30-DAY AVERAGE POLLUTION MEASUREMENTS)
                                       XCmg/1)

                                 — P (Performance Standard)

                    _ A  (Long Term Average)
            SAMPLE DISTRIBUTION OF M MONTHLY AVERAGES
                   (LONG TEfiM MDNZDORING DATA)
         ^>* M-i
            Min
I
               X (rag/1)

X (Average of 30-Day Averages)
          Note:   (a)  P/A = 1+1.64(CV)

                       07 = S^/X

                      {S-)2=(Z  (

                     f=Z  X/M  ,
Figure 8-6.  Statistical distributions for 30-day average pollution measurements.
                                   129

-------

-------
                           SECTION 9


               TREATMENT TECHNOLOGY APPLICATIONS

                  FOR TOXIC POLLUTANT REMOVAL


9.1  SELECTION OF POLLUTANTS TO BE CONTROLLED


    In order  to  determine which  toxic  pollutants,  if any, may
require effluent  limitations,  the pollutants  observed  in each
subcategory were evaluated with regard to their fereatability and
potential environmental  significance on  the basis of  the raw
waste concentrations  and mass  loadings  found during screening
and verification.   In  an attempt  to prioritize the  need for
regulation the toxic metals were divided  into  two groups:

    Group 1 - Those metals which appear at concentration levels
              that  are   readily   treatable  using  available
              technology   and   which    have   environmentally
              significant mass emission rates.                 *

    Group 2 - Potentially  significant metals observed  in the
              subcategory.   These  include  toxic  metals which
              exist   at  concentrations   below   the  minimum
              treatability   limit   and   above  the   minimum
              detection level.

    Table 9-1  presents the  significant  toxic pollutant metals
found in each group.  In general, those metals occurring in the
first group  are  of prime  concern and  may require regulation,
while those occurring  in  the second group are of somewhat less
concern and are not expected to require regulation.


9.2    APPLICATION  OF  ADVANCE  LEVEL  TREATMENT  AND  CONTROL
ALTERNATIVES
            i

9.2.1  General Design Objectives

    Beginning with  Section  11 of this  document,  the selection
and  application  of  toxic  pollutant  treatment  and  control
                              131

-------
    TABLE 9-1.  PKKDRITIZATICN CP TCKTC METALS root® IB EACH SOBCSTEGOIS
SUBCHZEGQBY
Gilorine-diaphragm cell





Chlorine-mercury cell







Hydrof luoric Sold





Group 1 (1)
Chromium
Copper
Lead
Nickel
Zinc

Arsenic
Cadmium
Copper
Xead
Mercury
Nickel
Silver
Zinc
antimony
Chromium
copper
Lead
Nickel
Zinc
Group 2 (2)
Jteitiitcny
Srsenia
Cadmium
Mercury
Selenium
Kiallium
antimany
Chranium
Biallium





Arsenic
Cadmium
tfercury
Selenium
ffliallium

Titanium Dioxide -                      Chromium               Lead
  ChlorMe Process                                             Nickel
                                                               Zinc

Titanium Dioxide -                      flntimsny               Selenium
  Sulfate Process                       Arsenic                Thallium
  and                                   Cadmium
  Chloride llnenite Process             Chromium
                                        Copper
                                        Lead
                                        Nickel
                                        Zinc
 (1)  Group 1 - dominant raw waste pollutants selected as control parameters
               for the proposed effluent limitations.

 (2).  Group 2 - secondary raw waste pollutants found less frequently and at
               lower concentrations.  These pollutants have not been selected
               as control parameters but are expected to receive adequate
               treatment as a result of the proposed effluent limitations on
               the Group 1 pollutants.
                                                              (continued)
                         132

-------
      9-1 - continued.
SUBCATEGORY                            Group  1                 Group 2


Aluminum Fluoride                      Copper                 Arsenic
                                       Nickel                 Cadmium
                                                               Chromium
                                                               Jfercury
                                                               Zinc

Chrome Pigments                        Jtatuiwiy               Cyanide
                                       Cadmium                 Mercury
                                       Chronium
                                       Cyanide
                                       lead
                                       Nickel
                                       Zinc

%drogen Cyanide                       Cyanide

Sodium Dichromate                      Chrcmium               Copper
                                       Nichel                 Selenium
                                       Zinc                   Silver

Copper Sulfate                         antimony
                                       arsenic
                                       Cadmium
                                       Chromium
                                       Copper
                                       laad
                                       Nictel
                                       Selenium
                                       Zinc

Nickel Sulfate                         Antimony               Arsenic
                                       Chromium               Cadmium
                                       Copper                 Mercury
                                       Lead                   Selenium
                                       Nickel                 "Thallium
                                       Zinc

Sodium Bisulfite                       Chromium               Antimony
                                       Copper                 Cadmium
                                       Lead                   Mercury
                                       Nickel
                                       Zinc

Sodium Bydrosulfxfce                    Chromium               Copper
  formate Process                      lead                   Pentachlorophenol
                                       Hickel                 Phenol
                                       Zinc
                         133

-------
technology for model plant systems for each of  the subcategories
proposed  for  regulation  are described.    Several  levels of
treatment  are  proposed.    Level  1  represents  existing  BPT
treatment systems and the advanced levels  (Level 2,  3, etc.) are
the  selected  technologies for step-wise  improvements in toxic
pollutant removal  over that  achieved by  the BPT system.   Flow
diagrams show  BPT  components as  a  starting point for advanced
level treatment additions  and incremental  cost  estimates.

    For  both  existing  and   new  sources,  the  advanced level
technology options are selected as candidates for BAT with toxic
pollutant  removal  as  the primary   objective.   Although  the
advanced  level systems  chosen  also  give  improved  performance
over the Level 1  (BPT)  systems  for the removal of conventional
and nonconventional pollutants, this  is regarded as  a secondary
design objective.

9.2.2,    Pretreatment .Technology

Since untreated heavy metal  ions will usually pass  through the
treatment provided  in a typical  POTW,  or  will be precipitated
with the POTW  solid residue, pretreatment of wastes  containing
significant amounts of heavy metals  is necessary.  As a  general
rule, alkaline precipitation, followed by settling  and  removal
of the solids  will  suffice.   In  certain subcategories,  such as
the  chlorine  industry,  specific  treatment will be required for
highly   critical   constituents   (mercury,  lead,   chlorinated
organics and  asbestos).   Normally  the Level 2 model  treatment
processes shown in the following  subsections will be  appropriate
for pretreatment prior to  discharge  to a POTW.

9.2.3  New Source Performance Standards

    New Source Performance Standards  are at  least equal  to  BAT.
In a few cases where  new plants  have the opportunity  to design
systems  for  better  toxic removal performance without expensive
retrofitting the higher  technology systems have been used  as a
basis.
9.3    ESTIMATED  ACHIEVABLE  PERFORMANCE  CHARACTERISTICS  FOR
ADVANCED LEVEL APPLICATIONS


    Advanced  level  control  and  treatment  alternatives  for
reduction  of  pollutant  discharges  and  their  applicability to
each  subcategory are  presented in  the sections  dealing  with
individual products.   With few  exceptions,  these alternatives
were  selected  specifically for  removal of priority pollutants
and were des-igned for end-of-pipe treatment.
                              134

-------
    Treatment technologies  practiced outside  the  industry are
recommended when  appropriate  and, in most  cases,  apply to the
removal of toxic pollutant metals.  The estimated 30-day average
treatability  levels  (Sections  8,  Table 8-11),  long-term data
parameters, and the screening  and  verification results are all
utilized   in   the  development  of   estimated   performance
characteristics for  the recommended treatment applications in
each subcategory.

9.3.1  Advanced Ley_el Removal of BPT  Pollutants

    Performance estimates for these systems, when possible, were
based   on   effluent   quality   achieved   at  plants  currently
practicing  these  technologies.   However,  in most  cases,  the
advanced  levels are  not  currently being practiced  within the
specific  subcategory  of  concern,  and  performance information
from other appropriate sources  is necessarily  utilized.

    When established waste  water treatment practices,  such as
clarification or  filtration,  form a  part of advanced treatment
alternatives, the  specified  achievable effluent quality has been
based  on  concentrations accepted  as achievable  through proper
design  and  control.    The  prime  example  of  this  is suspended
solids  reduction by filtration.

9.3.2  Advanced Level Removal of Toxic Pollutants

    Performance estimates for  toxic pollutants were also based,
when possible, on effluent quality achieved at plants currently
practicing these technologies.  However, in most subcategories,
toxic  pollutant analyses are  not  conducted  unless  a specific
pollutant is regulated and requires monitoring.  Where transfer
of technology is applied as a treatment alternative, performance
estimates  for  toxic  pollutant  removals  were  based  on  the
demonstrated    performances    in   other    industries   while
incorporating  allowances for  specific  differences  in process
waste  characteristics  and operating  conditions.   Statistically
derived long-term  monitoring  data parameters  were described in
Section 8 and are  compiled  in  tabular  form  in Appendix A.  The
screening  and  verification  data  are  used  to  supplement  the
available  long-terra data  applied  to  each   subcategory.    A
judgment  is  made  whether  the  screening  and  verification data
represent  a  well-performing   system  or   one  which  is  not
performing   at   its   technological   potential.      For   a
well-performing system, the data are regarded  as representative
of  30-day   averages   and   are  compared  with  the  estimated
treatability  ranges  from Table 8-11,  as  well  as  the 30-day
averages developed from thte  long-term data.  In this manner, the
performance  estimates  for  each pollutant,  at  each  treatment
level  for  the  nonexcluded  subcategories,  are  developed  and
                              135

-------
presented in tabular summaries.  By starting with the estimated
achievable  30-day  averages,  the  specific  variability  factor
ratio derived  for  each  pollutant is used to estimate the daily
maximum values.

    The model  plant waste flow  per unit of production is then
taken to calculate the estimated  mass  emission values of the 30-
day average  and  daily maximum limits  for  each  pollutant to be
controlled.
9.4  POLLUTION CONTROL PARAMETERS TO BE REGULATED


9.4.1  Conventional Pollutants

    Waste  water  quality parameters  which  are  identified  as
conventional pollutants include the following;

    PH
    Total Suspended Solids  (TSS)
    Biochemical Oxygen Demand, 5-Day  (BOD-5)
    Fecal Coliform
    Oil and Grease

    Only  the  first two  parameters  (pH and TSS)  in this group
have been selected for  regulation  in  the  Inorganic Chemicals
Manufacturing  Point  Source  Category.   For  direct dischargers,
the  pH  range  of  6 to  9 has  been  established  as  the general
control limitation and the permissible frequency and  duration of
excursions beyond  this  range  is to  be  specified in individual
plant discharge permits.  The limitations on TSS are specified
for  both  BPCTCA and BATEA-based  regulations,  the former being
largely  a  function of  industry  performance  and   the  latter
stemming  from  treatability  estimates  with   the   appropriate
technologies.

9.4.2  Nonconventional Pollutants

    The   waste   water   quality   parameters    classified   as
nonconventional pollutants  include  the  nontoxic metals tsuch as
Al, B, Ba, and Fe  along with chemical oxygen demand  (COD) , total
residual  chlorine,  fluoride,  ammonia,  nitrate, and "phenols,"
etc.  Of these, only Fe,  COD,  total  residual chlorine, fluoride,
and  ammonia are  considered  for  regulation  in  the inorganic
chemicals  industry.    Due to  its  toxicity, chlorine  would  be
controlled  in direct  discharges,  but  would  be  excluded  from
control in pretreatment regulations.  A similar  argument  is made
for the control of ammonia.  However,  since  many POTW1s are only
capable  of  about  20  percent  ammonia  removal,  both  direct
discharge  and  pretreatment   regulations  would  specify  NH3
limitations.   Similarly, the type  of  COD found  in inorganic


                              136

-------
chemical  industry  discharges  is not  amenable  to biochemical
oxidation in a POTW.   In addition, compounds which  contribute to
the COD  are likely  to create odor  and corrosion problems in
sewer systems.  Therefore, its control  would also be retained in
pretreatment regulations.  Fluoride control  is also required for
both direct and indirect  discharges largely because  the most
practical  technology  for  fluoride  removal  (precipitation as
CaF2) must  be  applied  to relatively  concentrated waste water
sources.  This treatment method achieves  removal  levels which at
best are still unacceptable for direct  municipal  or  agricultural
water uses.  POTWs  are not  effective  for  fluoride removal and
unless  sufficient  dilution  occurs  prior to the  reuse  of the
water,  special  techniques   (e.g.,   adsorption  on  activated
alumina) would have to  be applied for further fluoride removal.

9.4.3  Toxic Pollutants

    The  toxic  pollutants  found  at  significant levels  during
screening and  verification are listed by subcategory  in Table
9-1.   Out  of these,  toxic  pollutant control  parameters were
selected largely on  the basis of treatability.   Since several
toxic  pollutants  may  be  controlled  by   a common  treatment
technology,  it  is  possible  to  select one or  more  control
parameters  which will  act as a surrogate for others exhibiting
the  same   treatability  characteristics.     Treatment  system
operating conditions would normally be optimized for the removal
of the specified control parameters which would  be  monitored on
a regular basis.  The other toxic pollutants would  be monitored
much less frequently  as a periodic check  of  the effectiveness of
surrogate control.

    The - following toxic  pollutants  have  been designated as
control parameters in this point source  category:

    Cadmium
    Chromium  (Total)
  - Copper
    Cyanide  (amenable to chlorination)
    Lead
    Mercury
    Nickel
    Selenium
    Zinc

    The   specific   control   parameters   selected  for   each
subcategory  are  presented  in  the  tables  entitled  "Control
Parameter Limitations"  in the sections  of  this report dealing
with the  individual  industries.    Some  general  comments about
them are given  here.
                              137

-------
    The  most  common technology  applied  in industry  for  the
removal of chromium from waste waters involves a  reduction  step,
whereby Cr  (VI)  in solution  is converted  to the less toxic Cr
(III) form which can then be  removed  by alkaline precipitation.
The  efficiency of this  treatment  depends  upon  the presence of
excess reducing agent and pH control to,drive the reduction step
to  completion.   When  treated effluent  samples are  taken to
monitor  residual  Cr   (VI)   and   total  chromium  levels,   the
analytical results  for  Cr  (VI) are subject to several factors
which adversely affect  the  accuracy and reproducibility of  the
diphenylcarbazide (DPC)  colororaetric method.  The problem  is  not
so much one  of analytical  interferences with the Cr (VI)   ~  DPC
color  development,  but  rather the  actual changes  in  Cr  (VI)
concentration  that  can  take  place  during sampling,  sample
preservation and storage, and analysis.  The major  cause of such
changes is the presence  of  excess  reducing  agent in the treated
effluent.   This tends  to  give false  low  readings  for  Cr  (VI)
although  in  some  cases  the opposite may occur as  a result of
sample   preservation  and   storage   under   acidic  oxidizing
conditions.

    Thus,  in  view  of  the  questionable   reliability  of  the
presently accepted Cr (VI) monitoring procedure,  total chromium,
Cr  (T) ,  is  recommended as  the  control parameter  to be used in
the  inorganic  chemicals  industry.   The adequacy of Cr (T)  as a
control  parameter  is  predicated   on its  effectiveness   as a
surrogate for  Cr  (VI)  control.  Since the concentration   of Cr
(T)  represents the  summation of all  forms of chromium normally
found  in  solution or suspension  including  Cr  (VI),  the  final
concentration  of  Cr (T) in a treated effluent  is dependent on
the  effectiveness  of  both  the   reduction  and  the  alkaline
precipitation  steps.   In  this way,  the use of Cr  (T)  as  the
control parameter assures  that adequate removal of  Cr  (VI) is
being  achieved  as  a   direct  consequence   of  the  treatment
technology required.

    Special  consideration  is  given  to  the  control  of  copper
which may enter a POTW.   At high  enough concentrations, copper
may impose toxicity effects on  the  microorganisms in a POTW  and
may accumulate in municipal sludges  rendering them unusable  for
certain land applications.  Thus,  copper may  be designated as a
control parameter for pretreatment  even though it may not  be so
designated for direct discharges.
                              138

-------
                          SECTION  10


             COST OF TREATMENT AND CONTROL SYSTEMS
10.1  INTRODUCTION
10.1.1  Purpose of Cost Data

     More  complex  treatment  methods  and  higher  levels  of
pollutant  removal  are  reflected   in   increased   costs    of
equipment, energy,  labor  and  chemicals.   At some  point,  the
increasing costs  of  treatment  will  outweigh the benefits  of
such  treatment.    Therefore,  it  is important   that  for  each
subcategory the Agency   know the base cost and the  incremental
costs of  each level   of    treatment which  it might prescribe.
These "options" of  internal  costs,  which are the  industry's
annual costs  of providing  the   necessary waste treatment,  will
result in related  increases in  product  costs, which are termed
external  costs.    Thus  annual  costs  of  waste  treatment  are
expressed in  terms of  dollars per  unit of annual production of
the principal product.

     Because plant  visits  revealed  very   few treatment plants
serving a single product manufacturing line,  it was not feasible
to seek actual waste   treatment   facilities which  could  serve
as  real models for estimating purposes.  Accordingly*, the cost
data  were taken  from   similar  construction  projects  by   tthe
contractor, and from unit process  equipment costs assembled f'rom
vendors and  other commercial sources.  Because the model costs
apply  to  a  wide range of   climate, material sources and labor
conditions, they should be  considered  as preliminary estimates
within plus or minus 15 to  25 percent.

     Actual costs  incurred  by  individual  plants  may be more or
less than the presented model plant costs.   The major causes of
variability are:


     1.  Waste water  treatment  combined  with  the  treatment of
         other product effluents.

     2.  Site  dependent conditions, as   reflected  in  piping
         lengths,   climate,  land   availability, water  and power

                              139

-------
         supply  and  the  location  of  the  points  of  final
         discharge and solids disposal,

     3.  Material   (reagent)   costs,   due  to   variation   in
         availability and distance from the source.


     The construction costs  are  based on  the  Engineering News
Record Construction Index for  July 1978   (ENRCI=2800), and other
costs are expressed in mid-1978 dollars.

10.1.2  general Approach

     Since  few  single  product  waste  treatment plants  were
available  for  detailed   study,   the  costs presented  in  this
section  are  based on model  plants which  closely resemble  the
types and  capacities  of waste  treatment facilities  needed  for
each separate product subcategory.  The  model plant selections
are  based  on  review  of  Section  308  Questionnaire  responses,
plant   visits,   development   documents,   contacts   with   the
industries to  verify  treatment practices and to   obtain data on
size, waste  water flow,  and solid   waste   disposal systems.
Thus, each model is synthesized  from  actual data  as  a typical
plant  in  its  subcategory with  a  level   of  waste  treatment
equivalent to  BPT.    Variations   in treatment  plant  capacity
are accounted  for by selecting sets of  models which  represent
the  range of  existing  production plant  capacities  in   the
subcategory;  large,  medium and small.   Thus the model plants are
not set  up  as  exemplary  plants,  but as  typical  plants of
adequate  design  which  represent  the  range  of  plants  and
treatment facilities found in the subcategory.

10.1.3  Cost References and Rationale

     Cost information contained in this   report   was obtained
directly from  industry,  engineering firms, equipment suppliers
and  current  experience of the  contractor.  Whenever possible,
costs   are   based  on   actual   industrial   installations   or
engineering estimates for projected facilities  as  supplied by
industries consulted  during  the study.   In the absence of such
information,   cost estimates  have been  developed from either
current  costs for  similar waste  treatment installations   at
plants  making  other  inorganic  chemicals or from general  cost
estimates for  specific  treatment technologies.

     Treatment  costs  are  based  on  model  production  plant
characteristics   which   determine   the   treatment   processes
selected for each  operation.  Under  set  effluent limitations,
treatment costs  are  primarily functions of the  pollutant  load
(i.e., kg/kkg  of product),  waste water  flow rate (i.e.,  cubic
meters/day).   Available data  indicate  that both pollutant loads
                              140

-------
and flow rates can vary significantly among plants manufacturing
the  same  product.

10.1.4  Definition of Levels of Treatment and Control
Cost Development

     For the purpose  of establishing  the base   level treatment
costs,   each   industry  is  assumed   to be  practicing  Best
Practicable Control Technology  Currently Available  (BPT), which
the  EPA Effluent Limitations  Guidelines required  by 1977  for
certain  pollutants  (conventional  and nonconventional,   as well
as some of the toxic pollutants) specified for each  subcategory.
The investment costs and annual costs  of  such BPT   systems are
shown in this report as the base level or Level 1.  This level of
treatment  may also  provide  incidental  removal   of additional
toxic pollutants  not previously specified in the regulations.

     The advanced treatment levels {Level 2,  Level 3, etc.) are
aimed primarily  at reduction of  toxic   pollutants to   levels
considered acceptable  for July lf  1984,   performance, utilizing
Best  Available  Technology  Economically Achievable  (BAT)  at
incremental  investment and  annual costs beyond those shown for
Level 1.  For  example, for  Level  3  treatment,  the   incremental
cost as  given in the table is directly  added to base  or 1st Level
cost  to  obtain  the total  cost of  the  treatment system.   The
addition of the Level  2 incremental  cost  is   not   required to
obtain   the Level 3  total.   The  wa'ste   water  treatment  flow
diagrams for  Levels   2,  3, etc., as   given in  this   report,
include  the flow  diagram  for base  or Level 1 of treatment.

10.1.5   Treatment and Disposal Rationale Applied to
Cost Development               ~*~

     The  following  assumptions   are   employed  in  the  cost
development:

     1.  Noncontact cooling water generally is  excluded from
          treatment   (and  treatment  costs)  provided  that  no
         pollutants  are introduced.
              •
     2.  Water  treatment,  cooling  tower and  boiler  blowdown
         discharges  are  not  considered .process  waste  water
         unless   such   flows  contain  significant   amounts  of
         pollutants.

     3.   Sanitary sewage  flow is  excluded.

     4.  The plants are assumed to operate 24-hours  per day, 350
          days a year,  except where otherwise  noted.

      5.  Manufacturing plants  are assumed to be single product
          plants.

                              141

-------
     6.  The  inorganic  chemical industry  extensively  uses in-
         plant control  techniques  such as in-process abatement
         measures,  housekeeping  practices,   and  recycling  of
         process waste  waters  to  recover, valuable materials or
         use  these materials  as  feed for  other by-products.
         Segregation of uncontaminated cooling and other waters
         prior to  treatment  and/or disposal, and other similar
         measures can contribute to waste  load  reduction.   All
         such  costs  have   not  been  included  in  the  cost
         estimates.

     7.  Excluded  from  the estimates  are  any costs associated
         with  permits,  reports   or  hearings   required  by
         regulatory agencies.

10.1.6  Expression of Costs

     Investment  costs   for  Level  1  treatment  systems  are
expressed in mid-1978 dollars to construct  base level facilities
for each  single  product manufacturing   subcategory at various
production rates.

     Similarly, operation, maintenance and amortization of the
investment are expressed as  base  level  annual  costs for Level
1  and  as incremental annual costs for Level 2 and above.  Where
a  single  product plant   produces more  than one waste stream
requiring treatment, the respective investment and annual costs
are the combined costs of all treatment.

     Total annual costs  per metric ton of  product  are shown in
the summaries for each product subcategory.

Direct Investment Costs for Land and Facilities

     Types  of  direct  investment  costs   for  waste  treatment
facilities and criteria for estimating major components  of the
model plants are contained in the following subsections:

     Construction costs   -   Construction  costs  include site
preparation,   grading,   enclosures,   buildings,  • foundations,
earthwork, roads, paving and concrete.

     The  costs  of   constructing   lagoons  can  vary  widely,
depending  on  local  topographic  and soil   conditions.    The
required  areas  of   lagoons   and   settling   ponds   and  their
consequent  costs   are   developed   as  a  function  of  volume
(capacity).   It  is  assumed  that   reasonably level  sites are
available, consisting   of  sandy   loam   with high clay content
and no  large rocks or rock  formations.    It  is  assumed that
two rectangular  lagoons  are  furnished in  parallel, with one
common dike to permit alternate dewatering  for sludge removal by
                              142

-------
the clamshell  method.    Using balanced  cuts  and  fills, earth
dikes with 2:1  slopes  provide  liquid  depths  from  three  to
five  meters.   Earth moving  costs are   significantly  affected
by    site  conditions    and quantities.    To  express    these
variations for a  range  of  sizes at three  depths, the cost of
clearing,  excavation,  dewatering, compaction,  finish  grading,
riprap  and  associated  indirect expenses for   earthen lagoons
were  plotted  against  liquid volume. Piping,  valving and dike
roads not included are added separately   in  the cost summaries.
Lagoons are unlined  unless  the contents are highly pollutional
or, acidic.  The liner  material  employed for  impervious lagoons
is Hypalon..  The  installed  cost   of  the liner  is $11.00 per
square   meter  ($9.20 per    square   yard), which includes the
trenching  and  backfilling  necessary  for anchoring the liner.
In some  subcategories, clay lining  has been used  in  place of
Hypalon at a   cost of  $5.40 per  square  meter  ($0.50 per square
foot) .

     Costs of  buildings  may vary  from $25.00  to $45.00  per
square  foot.  For the purpose of this  study, building cost is
estimated  at $377.00 per square meter  ($35.00  per square foot).

     Concrete construction for. miscellaneous  work  varies from
$260.00  to $785.00   per cubic  meter  ($200.00  to  $600.00 per
cubic  yard).   For   foundations   and  flat  slabs, concrete has
been  estimated   at  $395.00 per  cubic meter  ($300.00  per cubic
yard) in  place.  Asphalt  paving  which has been   used  on lagoon
dikes and  for  miscellaneous roads,  is  installed at a cost  of
$9.70 per  square meter  ($0.90 per  square   foot),  A  width  of
three meters  is generally  assumed.

     Equipment costs  -  Depending upon  the method of treatment,
equipment  for waste  water  treatment  consists  of  a combination
of  items  such  as  pumps,  aerators,  chemical  feed  systems,
agitators,   flocculant   feed    systems,   tanks,   clarifiers,
thickeners,  filters,  etc.   Cost   tables for  these  items  were
developed  from  vendors'   quotations  on  a  range  of  sizes,
capacities and motor horsepowers.   Except for large size  tanks
and  chemical  storage bins, the   cost  represents    packaged,
factory-assembled units.  Mechanical  components  are generally
skid mounted, prepiped   and prewired;  and    include associated
pumps,  meters- and  instrumentation.    Critical equipment  is
assumed to be installed in  a weatherproof  structure.  Chemical
storage, feeders and feedback  equipment include such items as
probes,   instruments,  controls,   transmitters,  valves,  dust
filters  and   accessories.     Bulk " chemical  storage  bins  are
designed to hold a   standard bulk truck . load,  plus five  days
needs,  between  ordering  and  delivery.    Critical pumps  are
furnished in duplicate and when clarifiers are used, the flow is
split between  two units,  permitting  one  to be bypassed   for
repairs.  Single  units    are used    for   small flows,  batch
treatment  and  intermittent service.


                              143

-------
     Installation cost   -  Installation  is  defined to include
all services, activities,  and  miscellaneous material necessary
to implement the described   waste  water   treatment and control
systems, including piping, fittings, and  electrical work.  Many
factors  can  impact  the cost  of  installing equipment modules.
These  include  wage   rates, manpower  availability,  whether the
job is  performed by outside contractors  or  regular employees,
new construction versus modification   of   existing  systems, and
site-dependent conditions (e.g., the availability of sufficient
electrical  service).    In these estimates,  installation costs
were   chosen  for  each  application,   based  upon  average site
conditions and considering the  complexity of  the* system being
installed. An appropriate cost  is  allowed   for interconnecting
piping, power circuits and controls.

     Monitoring equipment  - In  this report,  it is assumed that
monitoring  equipment will  be installed at the treated effluent
discharge  point. It  will consist  of an indicating, integrating
and  recording  type  flow  meter,   pH   meter  with  sensor  and
recorder, alarms and controls and  an automatic sampler.

     Land -  Land  availability and  cost  of  land can   vary
significantly, depending upon geographical  location, degree of
urbanization  and the nature of   adjacent  development.  Land for
waste  treatment,  and  in some cases for inert  solids disposal,
is assumed to  be  contiguous   with  the  production plant  site
and  reasonably convenient  to   a  waterway which    can  receive
permitted discharges of   waste  water.   Where inert solids  are
retained at the  plant site, enough land is included  in the base
level' model plant investment cost  to accept  residual solids for
a  normal operating period  of ten years at  the same production
rate for which  the  plant is sized.     For the purpose of this
report,  land  for lagoons,  treatment   facilities and   on-site
residual  waste  disposal  is  valued   at   $30,000  per  hectare
($12,000 per acre).

     Investment  costs  for  supporting  services -   Engineering
design and inspection  are typical services necessary to  bring a
project  from  a concept  to an operating system.   Such services
broadly  include  laboratory  and  pilot  plant  work  to establish
design parameters, site  surveys  to fix  elevations and  plant
layout,   foundation   and  groundwater    investigations,   and
operating   instructions;   in   addition   to   design   plans,
specifications and inspection during construction.  These costs,
which  vary  with   job  conditions,  are   often  estimated  as
percentages  of  construction   cost,  with   typical  ranges  as
follows:

      Preliminary survey and construction surveying  1 to 2%

      Soils and groundwater investigation            1 to 2%
                             144

-------
      Laboratory and pilot process work              2 to 4%

      Engineering design and specifications          7 to 12%

      Inspection during construction                 2 to 3%

      Operation and maintenance manual               1 to 2%


     Prom  these  totals of 14 percent   to 25 percent,  a  mid-
value  of   20   percent  of  in-place  construction  (installed
equipment and construction)  costs has been used  in  this study to
represent  the  engineering  and  design  costs applied  to model
plant cost estimates.

     The contractor's  fee and contingency, usually expressed as
a  percentage of  in-place construction  costs,    includes  such
general items as temporary utilities,   small   tools, dewatering,
field  office  overhead    and  administrative  expense.    The
contractor is  entitled to a reasonable profit on his activities
and  to  the cost    of  interest  on  capital  tied  up  during
construction.  Although not  all of  the   above  costs  will be
incurred on every job,  an  additional 20  percent of the in-place
construction costs has been used to cover related  costs broadly
described  as  contractor's  fees,  incidentals,  overhead  and
cont i ng enc i es.

Operation  and Maintenance Costs

     Annual operation and  maintenace  costs   are   described and
calculated as follows:

     Labor and supervis ion costs -  Plant operations are assumed
to be conducted  24-hours  per   day  350  days  per year,  with
attendance for  only part of each working day.   For batch waste
water treatments  systems adjustment are made for  the number of
working  days  in a year.  Personnel costs are based on an hourly
rate  of $20.00. This  includes  fringe  benefits and an allocated
portion of costs for management, administration and supervision.

     Personnel are assigned for  specific  activities  as required
by the complexity of the system, usually  4 to 12 hours per day.

     Energy costs - Energy  (electricity)  costs are based on the
cost of  $306.00 per horsepower operating 24 hours per  day and
350 days per year.   For batch processes,  appropriate adjustments
are  made  to  suit  the  production  schedule.   The  cost  per
horsepower year is computed as follows:

         Cy= 1.1 (0.7457HP x Hr x Ckw)/('E  x P)               1)
                               145

-------
     where
         Cy = Cost per year
         HP *  Total  horsepower rating of motor  (1  hp = 0,7457
         kw)
         E  = Efficiency factor (0.9)
         P  = Power factor (1.00)
         Hr = Annual operating hours  (350 x 24 = 8400)
         Ckw = Cost per kilowatt-hour of electricity  ($0.040)

     Note:  The 1.1 factor in  equation  (1) represents allowance
for incidental energy used such as lighting, etc.,
     It is assumed  that no other forms of energy are used in the
waste treatment system.
     Chemicals - Prices for  the  chemicals  were  obtained from
vendors and  the  Chemical Marketing   Reporter.   Unit costs  of
common  chemicals  delivered  to  the plant site are  based  on
commercial  grade  of   the  strengths or   active   ingredient
percentages as follows;
      Hydrated Lime (Calcium Hyroxide) Bulk      $ 80/metric ton
                                       Bag       $ 85/metric ton
      Quicklime                        Bulk      $ 70/metric ton
      Ground Limestone                           $ 13.20/metric ton
      Soda Ash (58% Bulk)                         $ 85/metric ton
      Caustic Soda (58% NaOH)                    $200/metric ton
      Sodium Sulfide (60-62%)                    $435/metric ton
      Sulfuric Acid                              $ 75/metric ton
      Hydrochloric Acid (32%)                    $ 70/metric ton
      Aluminum Sulfate (56% Alumina)              $250/metric ton
      Flocculant  (Polymer)                       $2.00/kg
      Sulfur Dioxide (Ton Containers)            $335/metric ton
                               146

-------
      Chlorine (ton Containers)                  $220/metric ton

      Sodium Bisulfide (72-74%)                  $385/metric ton

      Ferrous Sulfate                            $ 70/metric ton

      Diatomaceous Earth                         $  0.30/kg

      Activated Carbon                           $  2.00/kg


     Maintenance - The annual  cost  of  maintenance is estimated
as  10 percent of the investment cost, exluding land.

     Taxes and insurance  - An annual provision of  three percent
of  the total  investment cost has been  included for  taxes  and
insurance.

     Residual waste disposal - Sludge  disposal   costs can vary
widely.  Chief cost  determinants  include  the amount apd type of
waste, and the choice  of either on-site  disposal  or contract
hauling which depends on   the size of the  disposal operation and
transport distances.  Off-site  hauling and disposal  costs are
taken  as  $13.00  per cubic meter  ($10.00  per  cubic yard)   for
bulk hauling, with appropriate increases  for small quantities in
steel  containers.    For  on-site  disposal  from  lagoons,  a
clamshell  at  $600.00  and  front  end  loader  at  $300.00   per
disposal day are  used.   For very  large  sludge   quantities,
lower  unit  costs have been  assumed.     The  computed   sludge
quantities  are spread on  land valued at $12,000 per acre.

     Monitoring,   analysis   and " reporting   -   The   manpower
requirements covered by  the   annual labor and supervision costs
include  those  activities associated with  the  operation   and
maintenance of monitoring instruments, recorders, and automatic
samplers  as well  as  the taking  of  periodic    grab samples.
Additional costs for  analytical laboratory  services  have  been
estimated  for  each  subcategory   assuming that  sampling  takes
place  three times  a  week at  the point of discharge  and  that
an analytical  cost   of    $20.00 per constituent  is  incurred.
Approximately 10 percent   of  the total analytical  cost has been
added  for  quality control and water  supply  samples.   Unless
otherwise stated,  continuous discharge   is   assumed   and the
analytical  costs  associated  with compliance monitoring  at the
BPT level  are based on  the  determination of four constituents.
At  the   advanced   (BAT)   levels,  the  determination  of  six
constituents is assumed.  A  reporting  cost of  $1,500 per year
is added  for clerical support.   Monitoring costs for periodic
batch  treatments  are reduced  in  proportion to    the  number of
days per year  when discharges occur.
                               147

-------
Amor t i zation

     Annual depreciation  and  capital  costs  are  computed as
follows:

         CA = B%r(l+r)n /M3+r)n -1                          (2)

     where

         CA = Annual cost

         B  = Initial amount invested excluding cost of land

         r  = Annual interest rate  (assumed 10%)

         n  - Useful life in years


     The multiplier  for B  in equation (2)  is often referred to
as the capital recovery  factor,  and  is  0.1627 for the assumed
overall useful  life  of  10  years.   No residual or salvage value
is  assumed.

Items Hot Included in Cost Estimates

     In some  subcategories,  a portion of  the  waste water is
returned  to  process from  an intermediate treatment step.   In
these  cases,  the  costs  of  return piping  and   pumping   are
considered as  water  development  and not   as  waste treatment.
Costs for  subsequent  treatment are   based on   the remaining
flow  after diversion of the return-to-process flows.

     Although specific plants may encounter  extremes of climate,
flood  hazard  and  availability  of water,  the costs  of  model
plants   have    been   estimated  for   average  conditions   of
temperature, drainage and natural resources.  It is assumed that
any necessary site drainage, roads, water development, security,
environmental  studies  and permit costs  are already  included
in production facilities  costs.  Therefore, the model costs are
only for facilities,  supplies  and  services  directly related to
the treatment and disposal  of waterborne  wastes, including land
needed for treatment and on-site sludge disposal.  Air pollution
control equipment required by the Clean Air Act is not included,

     Dust   collectors   normally   associated   with   package
treatment, chemical transfer and feeding systems  are included.
Raw  wastes  from various  sources are assumed to be delivered to
the treatment facility at sufficient head  to fill the influent
equalization basin, and final effluent is discharged by gravity.
Costs of  pumps, pipes  lines  etc., necessary to deliver   raw
waste water  to  the  treatment  plant or to  deliver  the treated
effluent to the  point of  discharge are not  included in the cost
estimates.

                               148

-------
     Since the treatment  models  are  designed  to  serve single
product manufacturing plants,   no  emergency  holding basins  or
internal bypasses are provided.  Any such  necessary facilities
are more  appropriately  furnished  as part of a combined  waste
treatment system serving several product lines.


10.2  COST ESTIMATES FOR EACH SUBCATEGORY
     Estimated costs  for  the waste water  treatment plants for
the  different  annual  productions  and  at various    levels  of
treatment are  calculated  in terms of  total  annual  costs.   The
total annual cost is the summation  of   the  annual amortization
of the investment costs and the annual  operation and maintenance
costs.
     The types of costs shown for each model plant are:

     (a) Investment

     (b) Annual operation and maintenance

     (c) Annual  amortization  of  investment costs   (excluding
         land)

     The total annual  costs  per  metric  ton  of  product have
been calculated.

     For the purpose  of the cost estimate,  the first level  of
treatment  represents the  base  cost  of  the treatment  system
(BPT).   The other  levels  (second, third,  etc.)  represent the
incremental cost  above  the   base cost.   The actual  additional
costs  a  plant  would  incur  in  implementing  the  described
treatment processes  depend  on  current treatment practices, and
to some extent on the  availability of land.
                                                             \
     In some cases,  land for economical   on-site sludge disposal
for a ten year  period has been  provided  in  the BPT model plant
costs.   Since land  cost is  not amortized,  its  value  appears in
the initial investment cost   but  not in  the  total annual costs.
Where land  is  a  major factor in  the BPT estimated costs, its
significance will be  mentioned in  the  separate reviews of each
subcategory.

     For the  purpose of  cost  estimating,  a  set  of generally
representative  model  plant specifications are given for  each
nonexcluded subcategory  starting with the Chlor-Alkali industry
in Section 11.  These  specifications, together with  the  basic
assumptions on  co,st estimating  detailed  in this  section, form
                               149

-------
the  basis for  the  baseline  cost  estimates   for  alternative
treatment  systems.   These  cost estimates  are presented  in a
tabular   format   in   the  cost  development  portion  of  each
applicable subcategory section.  In order  to take into account
more  fully   the   wide   range  of  plant  specific  variables,
additional cost elements which  may   add to  the baseline costs
are then  considered  on a case-by-case  basis.  The results are
either expressed graphically  as a   cost envelope or are  given
as an estimated percentage factor to be  applied  to the baseline
costs.
                               150

-------
                           SECTION 11


                     CHLOR-M.K&LI INDUSTRY



11.1  MERCURY CELL PROCESS INDUSTRY PROFILE


1-1.1.1  General Description

     Chlorine and its co-product  caustic soda  (alkali) are used
in large quantities  in  the production of plastics, organic and
inorganic chemicals,  in the pulp and paper industry, in water
and waste water treatment  and in  a number of other industries.

     The production rate in the United States is  approximately 9
million metric tons  (10  million short tons)  of  chlorine per year
and over 95 percent of that production is by the  electrolysis of
a sodium  or potassium chloride  solution via  one  of  two major
processes.  The two processes, mercury cell  and  diaphragm cell,
differ in cell design and  in the quantity  and quality of waste
water generated, and because of  these difference they are being
addressed separately under the Chlor-Alkali Subcategory.

     Other  processes  for  chlorine  production such  as  the
recently  developed   membrane process  are  not  addressed  here
because  only  pilot-scale  production  exists "or  no data  is
available from fully operating facilities.

     Approximately 30 percent of the U.S. production of chlorine
is by mercury cell  plants.   Of 27 known plants,  308 data was
available for 15.  Table 11-1 presents a summary profile of the
subcategory.    Table  11-2  presents  the  current  status  of
discharge regulations for mercury cell chlorine  plants.

11.1.2  General Process Description and Raw Materials

Brine System

     The sodium  chloride solution  (brine or  salt  dissolved in
water) is treated with sodium carbonate1and sodium hydroxide to
precipitate  impurities  such  as  calcium,   magnesium  and iron.
The  precipitated  hydroxides  and  carbonates  are  then  settled


                               151

-------
         TABLE 11-1. SOBeaiEQORY ERCFUE DMCA
SUBCKDSQOOT
ODJQBINE MERCURY
Total subcategory capacity rate
Itotal subcategory production rate
Number of plants  in this subcategory
308 Data on file  for
    With total capacity of
    With total production  of
    Representing  capacity
    Representing  production
    Plant production range;
            Minimum
            Maximum
    Average production
    Median production
    Average capacity utilization
    Plant age range:
            Minimum
            Maximum
    Waste water flow range:
            Minimum
            Maximum
    Volume per unit product:
            Minimum
            Maximum
                         3,545,000 kkg/year
                         2,750,000 kkg/year
                               27
                               15
                         1,280,600 kkg/year
                         1,090,000 kkg/year
                               36 percent
                               40 percent

                           19,100 kkg/year
                          198,000 kkg/year
                           77,900 kkg/year
                           70,400 kkg/year
                               75 percent

                                2 years
                               26 years

                                4 cubic meters/day
                            2,100 cubic meters/day

                               < 1 cubic meters/kkg
                               11 cubic raeters/kkg
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Commerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chemical Industry,"June, 1978, and "Economic Analysis  of Proposed
Revised Effluent Guidelines and Standards for the Inorganic Chemicals Industry,"
March, 1980.
                                    152

-------
TABLE 31-2.  STATUS OF REGULATIONS - EFFLUENT LIMITATION GUIDELINES
SUBCATEGORY

SUBPART
CHLORINE MERCURY CELL

F  (40 CFR 415.60, 3/12/74)
                                           STANDARDS
Product   Para-
                           BPCTCA
                          (1)
                         BATEA*
              .(2)
  Max."   Avg; '      Max.   Avg.
    NSPS

Max.    Avg.
Process meters
Mercury
Cell TSS
Process
Hg,
(kg/kkg)
0.64
0.00028
0.00028
(kg/kkg)
0.32
0.00014
0.00014
(kg/kkg) (kg/kkg)
No discharge
of pwwp3
No discharge
of pwwp
(kg/kkg) (kg/kkg)
0.64 0.32
0.00014 0.00007
* Section 415.63 was remanded and is presently reserved (41 FR 51601,
  November 23, 1976).
(1) Max. = Maximum of any one day.
(2) Avg. = Average of daily values for thirty consecutive days shall not
    exceed.
(3) pwwp = Process waste water pollutant.
                                       153

-------
usually in a clarifier and the underflow,  known as  brine mud,  is
sent  to  a  lagoon  or filtered.   Brine muds  from  mercury cell
plants usually contain  small amounts  of mercury  because the
spent brine from the cells is  recycled.  Consequently brine mud
filtrate is recycled or treated before discharge and solids are
disposed of in secure landfills.

     Before it is sent to the cells,  treated  brine  is evaporated
if necessary to remove excess water and then  pH adjusted.  Spent
or depleted brine from the cells  is  acidified  and dechlorinated
using vacuum  and/or air stripping before  being saturated with
salt and recycled.

Mercury Cell Process

     The mercury cell,  in general,  consists  of  two sections:
the   electrolyzer   and  the   decomposer   or   denuder.     The
electrolyzer  is  an elongated  steel  trough  that  is  inclined
slightly from the horizontal.  Mercury flows in a thin layer  at
the bottom forming  the cathode of the  cell,  and the brine flows
cocurrently on  top  of  the  mercury.  Parallel graphite or metal
anode plates are suspended from the cover  of  the cell.  Electric
current  flowing   through  the   cell   decomposes   the  brine,
liberating  chlorine  at  the  anode  and  sodium  metal   at  the
cathode.  The metallic sodium  forms  an amalgam with mercury.

                2 NaCl(aq)  +  Hg =  C12 + 2  Na(Hg)

     The amalgam  from the electrolyzer flows  to  a denuder and
the spent  brine is  recycled  to the  brine purification process.
In  the  denuder,  the  amalgam  becomes an  anode  to a  short-
circuited iron or graphite cathode.  Deionized water  is added  to
the denuder which reacts  with the amalgam to form hydrogen and
caustic  soda.    In  modern   mercury  cells,   the  denuder   or
decomposer  is a horizontally or vertically laid graphite-packed
bed.  The water and the amalgam flow countercurrently.  Mercury
is then returned to the electrolyzer.

Product Purification

     Chlorine from  the cell  is cooled  to remove water and other
impurities.    The  condensate is  usually  steam   stripped" for
chlorine recovery and returned to the brine system  or discharged
After cooling,  chlorine gas  is dried further by scrubbing with
sulfuric acid.   The diluted  acid is then usually regenerated,
sold or  used  for pH control.  When  chlorine gas  is compressed
and liquified,  it leaves behind  noncondensible gases  known  as
tail  or  sniff  gas.   The  tail gas  is  usually  scrubbed with
caustic  or lime, generating a hypochlorite  solution  which  is
then  decomposed, used on-site,   sold  or  discharged  with   or
without -treatment.
                              154

-------
     The  sodium hydroxide  or  caustic  product  formed  at  the
denuder  has  a concentration  of  50  percent NaOH.   Some of the
impurities present in the caustic can  be removed or reduced by
the  addition of  certain chemicals,  and  the caustic  is  then
filtered.  In most cases it is sent to storage or is evaporated
if a more concentrated product is required.

     Hydrogen gas  is cooled  by  refrigeration to  remove water
vapor  and mercury,  and can  be  treated further  by molecular
sieves  or carbon.    Condensate from hydrogen. cooling  is  then
discharged or recycled to the denuder after mercury recovery.

     Figure  11-1  presents  a  general  process flow  diagram of
chlorine production by mercury cell.


11.2  WATER USE AND WASTE WATER SOURCE CHARACTERISTICS

11.2.1  Water Use

     Water  is  used   at  mercury cell  plants   for  noncontact
cooling,   tailgas    scrubbing,   cell    washing,    equipment
maintenance,  floor washings and in  the decompositon of  sodium-
mercury  almagam in  the denuder  to  produce  sodium hydroxide.
Because  most  brine systems at  mercury cell  plants  are closed
systems,  water  use in  the brine system  is minimal.   The total
water usage at  plants was found  to  range from 7.6 to 204 cubic
meters per metric  ton  (1800 to 49,000 gallons  per  short ton),
with  noncontact cooling  water  which  is  not covered  by  this
effluent  guideline comprising approximately 70  percent of the
total.

11.2.2  Waste Sources

     The following waste sources are or can be contaminated with
mercury and would therefore require treatment if discharged.

Brine Hud

     This is the waste produced during  the purification  of brine
before  it  is  introduced into  the cell for  electrolysis.   The
metals  commonly  removed during purification   are  magnesium,
calcium,  iron  and  other  trace  metals  such  as  titanium,
molybdenum chromium,  vanadium and tungsten.  Calcium  and  iron
are  removed  as  hydroxides.   Brine mud is  the major portion of
the waste solids produced from the process.  The solids  content
of the stream varies from 2 to 20 percent and the volume varies
from  0.04  to  1.5  cubic  meters per  metric  ton  of   chlorine
produced.  The waste  is  either sent  to  a  pond  for settling or is
filtered.    The overflow   from  the pond or  the  filtrate  is
                               155

-------
U1

BRINE •*•
1
C<
(

*
OtlOJUHe
PURIFICATION
SYSTEM


1
BRINE
HUBS
TO WASTE
AND RECYCLE
TONCONTACT
(CONTACT*)
XH.IHG WATER
i
C
o
L
E
R


PURIFIED ^ uimmrnv
-BRIHB •>• "£££** ~ r
* * t
HOBCOHTACT
COOLING HATER
^ BACKHAS
»ACID ^CELL" "*
T
D
R 	
Y ^
E ^ miHUiSyoR

' WEAK SULFURIC
:ONDENSATE ACID
TO HASTE OR REUSE
— 	 	 IIYDROUEH 	
l •
«ALGAH 	 ^ (DECOMPOSER) "*~ DBHINERALIZED
HROnlBV 1 1 "*~ "VNO- JO ATMOSPHERE
S0» iiUUlUM HOMCOIITACT K1LTBK) TOR (JsR
HYDROXIDE SOLUTIOH COOLING _^
X HATRR

COOLER CONDENSATE
t (SCRUBBING HASTE*) ^>TO O*fiSTE
en< cnr,Ttiu 1" 	 	 1 RECYCLE COOLING
HATER — +. FILTER -5°» SODIIW — (PACKAGING 1 	 *• TOWER
HYDROXIDE 1 1 -- -rilr~
•t 	 1 	 SOLUTION - ""'-'"-
BACKWASH SOLUTIOH TO
	 I 	 ATMOSPHERE
11 RECYCLE 1 BACKWASH A 1
^ | FILTER CAUSTIC (LIME) [ f
| 	 "" ANU HATER | BLOW DOWN
SOIfDS 	 TO INERTS
f~ LANDFILL
r i
\ 	 1 	 ^-J R
^ -MERCURY 	 ' ' 	 SOLIDS ~ -*-70 "ASTE B
- 	 	 TAIL GAS f» B SODIUM (OH CALCIUM)
5 	 HYPOCHLORITE..^.
SOLUTIOH
LIQUID _^TQ SAtES TO USE, SALES,

. j 	 PRODUCT
REFRIGERATION
\ SYSTEM
t t
NONCONTACT
COOLING
HATER
             USED AT SOME PLANTS ONLY
            Figure 11-1.  General process diagram for production of chlorine/caustic by mercury cells.

-------
recycled  to  the  process  as  makeup water for the brine.  In the
mercury  cell process, only 16  percent  of the NaCl solution  is
decomposed in the cell and  the  unconverted  brine is recycled  to
the purification unit after dechlorination.  This recycled brine
is contaminated with mercury so the resulting brine mud contains
small amounts of mercury.

Cell Room Wastes

     The  major components of  this  stream  include leaks, spills,
.area  washdown and  cell  wash waters.   The  amount  varies from
plant  to plant and  depends  largely on housekeeping practices.
Data  indicate a  range  of  from 0.01 to  1.5 cubic  meters  per
metric  ton  of chlorine  produced.   Cell room waste constitutes
the major stream requiring  treatment because of the high levels
of mercury present  in these wastes.  If graphite anodes are used
in  the  cells, the  wastes  may   also contain lead  (used  as  an
electrical  contact  at  the  anode)  and  chlorinated  organics.
However most mercury cell plants  have converted to metal anodes.

Chlorine  Condensate

     Condensation   from  the  cell  gas   is  contaminated  with
chlorine.  At  some  plants,  the condensates are recycled to the
process  after  chlorine  recovery.   Both contact and noncontact
water  is used for  chlorine  cooling  and  for  removal  of water
vapor.    Because of this,  the amount  and  type of waste water
varies  from  plant  to plant.   Data from  one plant indicates  a
waste  condensate flow  of  approximately  0.01  cubic  meter  per
metric ton of chlorine produced.

Spent Sulfuric Acid

     Concentrated sulfuric  acid is used  in  the dryer to remove
the  residual  water  from  the chlorine gas  after the first stage
of  cooling.   In most cases,  the  acid  is  used until a constant
concentration of 50-70 percent  is  reached.   The spent acids can
be  regenerated for  reuse,  used for pH control in a treatment
system,  or sold.

Tail Gas  Scrubber Liquid

     The  tail  gas  containing the  uncondensed chlorine gas from
the  liquefaction stage,  along with some air and other gases,  is
scrubbed  with sodium/calcium hydroxide  to  form sodium/calcium
hypochlorite  solution.    When  the  equipment  is  purged  for
maintenance,  the tail gas is  also  absorbed  in calcium or sodium
hydroxide, producing the corresponding  hypochlorite solution.
The  hypochlorite can be used in another process on site, sold,
discharged   to  treatment  or  decomposed  'before   discharge   or
                               157

-------
treatment.   The  amount of tail gas  scrubber  water  varies from
0.04 to 0.58 cubic meter per metric  ton of chlorine.

Caustic Filter Washdown

     The 50 percent caustic produced at the denuder is filtered
to remove salt and other  impurities.  The filters are backwashed
periodically  as  needed,  and the  backwash  can be discharged to
treatment or  filtered  with the filtrate recycled  to  the brine
system  and  the solids sent  for  disposal  or  mercury recovery.
Waste water volume from  caustic filter  backwashing is variable
and no flow data are available.

Hydrogen Condensate

     Hydrogen  produced  at the  denuder  is  cooled  to  remove
mercury and  water  carried over in the gas.   The condensate is
either sent to treatment facilities or  to mercury recovery after
which it can be returned to the denuder.  Data on the volume of
this waste stream are not  available.

Summary of Waste Water Flow

     Summing the flow ranges presented above  for specific waste
sources results  in a maximum  mercury-contamined waste  flow of
2.1  cubic meters  per  metric  ton  (m3/kkg)  for plants  where
specific  stream  data  were available.    This  does  not  include
brine mud flows  which are reused instead of  discharged,  and
therefore do not affect total  flow.

     Data  available on  total  discharges   at 13 mercury  cell
plants  are  presented  in  Table  11-3.    The  average  discharge
volume  indicated  is  also 2.1  m3/kkg, although flows as high as
6.3 m3/kkg do exist.


11.3  DESCRIPTION OF SPECIFIC PL&HTS


     The  following  descriptions  of  specific plants  includes
those that 'were  sampled during the  screening and  verification
program.  The discussion  primarily  covers plant  practices  in
waste water control and treatment.

11.3.1  Screening Progran

     Plant f299  was  visited in. the  screening and  verification
phase of  the program.   The mercury-contaminated waste streams
include outlet end-box wash  water,  spills and  cleanup  water,
brine mud saturator  sludge,  and  pump  seals waste  water.   The
combined waste water is sent to a  surge pond.  The effluent from


                               158

-------
11-3.         OF             PLOW      FOR CHLORINE MERCURY CELL HANTS

SUBCATBGORY CHLORINE ME8CURY CELL
Plant
Number
317
907
299
167
747
343
106
131
589
898
741
553
769
Average of 13 plants
feste Water Flow
^m /kkg Chlorine)
0.51 '
0.36
1.6
5.6
0.69
1.6
0.67
1.7
5.8
0.98
0.51
1.0
6.3
2.1
                              159

-------
the surge  pond is  mixed  with sodium  bisulfide  and sent  to a
settling pond.    The overflow  from the  pond is  pH adjusted,
filtered (in a filter press)  and  passed through activated carbon
towers before  discharge.   In  the  sampling program waste water
influent to  the  surge pond and  the  overflow  from the settling
pond  were   sampled.    Figure  11-2  gives  the general  process
diagram  and  shows all the waste streams  sampled.   Table 11-4
presents  major  pollutant  concentrations  and   loads  for  the
sampled streams.

11.3.2  Verification

     Four  more plants  (1*747,  $167,  £106 and f317)  producing
chlorine/caustic  by  mercury  cells were visited  and sampled in
the  verification  program.    Table  11-5  presents  pollutant
concentrations for  the sampled  streams—and loads  for  TSS and
mercury.

     At  Plant  f747r  the  brine  dechlorination system  has been
converted from barometric condensers to a  steam  ejector system.
The  conversion  resulted   in  increased  chlorine  recovery and
reduced  contact  waste  water.    By  providing  settling  and
secondary filter  facilities, the brine  filter backwash has been
eliminated.    The   tail   gas   scrubber  liquid   (hypochlorite
solution)  is  offered for  sale and if  not marketed, is treated
for removal of chlorine and  discharged.   Mercury bearing waste
waters  are treated  with  sodium  sulfide  (Na2S)   and filtered.
Solids are retorted for mercury recovery and the  filtrate is
mixed with the other process  waste  waters and  the pH adjusted
before discharge.  A flow diagram  of the manufacturing process,
including the waste water  treatment facility,  is  given in Figure
11-3.

     At  Plant f!67,  the  waste water streams,   consisting  of
filter backwash,  cell  room wash,  rain  water  runoff,  and leaks
and spills,  are  combined  and treated for mercury removal.  The
water is sent to a holding lagoon and the  overflow is reduced by
reaction with ferrous chloride, which precipitates mercury.  The
reacted solution  is  sent  to  a  clarifier and the  underflow from
the clarifier  is disposed of  in a  landfill.  The overflow is
filtered and the filtrate  is  passed through activated carbon and
an  ion  exchange  column prior  to  discharge  to   a  lagoon.   The
effluent from the lagoon is pH adjusted and discharged.  Figure
11-4 shows  the  simplified  process  flow diagram  for Plant f!67,
including the sampling locations.

     At Plant  f317, the brine purification mud  is  mixed with
spent sulfuric acid  and   sodium  hypochlorite  solution.    The
treatment  removes mercury  from the mud and transfers it to the
solution.  The solution is filtered and the solids landfilled.
The  filtrate is  mixed with other  mercury-contaminated  waste
                               160

-------
                                                                                               PMM,
                                                                                               DISCHARGE
                                                                                -JO BWEB
                      Kibte stxewa saqpled.
                      Huntiers to ( ) were sw^ded in
                      soreavtng, others during verif IcatJoi),
Figure 11-2.   General process flow diagram at plant  #299 showing the  sampling points.
                dilorine/caustic  (nercury cell)  manufacture.

-------
        TABLE 11-4.  POLLUTANT CONCENTRATIONS MUD LOADS £T PLANT I 299

SOBCATEGORY CHLORINE
Stream Stream
Number Description
Screening Phase: ^ '
1 Cell Waste
2 Mercury Treatment
Effluent
3 Tail Gas
Scrubber
Verification Phase: ^ '
1 Mercury Treatment
Influent
2 Mercury Treatment
Effluent
3 Cell Waste
4 Brijie Mod
5 Tail Gas Scrubber
(MERCURY CELL)
TSS
(mg/1) (kg/kkg)

12 0.016
5 0.007
NA NA

91 0.13
18 0.026
120 0.17
13,000 NA.
180 0.022


Mercury
(mg/1) (kg/kkg)

0.15
0.029
0.11

5,9
0.20
11.
0.54
0.17

0.0002
0.00004
NA

0.080
0.0003
0.015
NA
0.00002
NA = Not available.
 (1) = Data based on one 72-hour composite sample of each stream.
 (2) = Data based on three 24-hour composite samples of each stream.
                                    162

-------
           11-5.                           AND IOADS AT VMEPICATIDN

SUBC&TEGOEX CHLORINE {MERCURY CELL)
Stream Stream
Number Description
Plant
1
2
3
4
5
6
7
Plant
5
6
7
8
9

Plant
1
2

3
4

5

6

7
8
Plant
1
2

4
747
Cell Waste
Treated Waste
Acid Input
Acid Output
Dechlor System
G-2 Condensate
Tail Gas Scrubber
167
All C12 Wastes
Cell Wash
Brine Process
Treated Waste
Clarifier
Underflow 5
317
Cell Waste
Brine Mad
Filtrate
Tank Car Wash
Collection
Tank 21
Treated
Effluent
Deionizer
Effluent
N-C Cooling
Final Effluent
106
Cell Wash
Treated. Cell
Wash
Final Effluent
TSS
(mg/1)

700
60
NA
NA
9
"2
NA

560
57
4
2

,900

45

520
18

,000

110 ,

18
16
18

79 ,

20
2.0
Tkg/kkgJ

1.6 x 10~^
1.4 x 10
NA
NA
0.0037
2.7 x 10~5
NA

" —A
5.7 x 10 ^
7.1 x 10 „
1.3 x 10

4.0

NA

NA
NA

8.6

4.4 x 10~2
•— "\
5.2 x 10
2.2
2.4





Mercury
(mg/1)

18
0.10
0.023
0.003
0.035
0.27
0.039

3.8
0.72
0,005
0.32

10.4

14

34
0.033

123

0.10

0.001
0.001
0.002

3.9

0.015
<0.0005
(Kg/KKg)

4.3 x 10~f
2.3 x 10~g
3.5 x 10
7.2 x 10~i
1.5 x 10 jj
1.8 x 10~2
8.0 x 10

1.3 x iol:
6.7 x 10"t
9.0 x 10,
1.8 x 10
-5
8.7 x 10

NA

NA
NA

5.0 x 10

4.3 x 10~5
-7
2.9 x 10 J
1.4 x 10 j|
3.6 x 10




NA
NA = Not available.
 (1) = Data based on three 24-hour composites,
                                      163

-------
Had
                                                                                       DISCHARGE
        lECfrO

      SWP,.II!C POINTS,
                                          MKW1 TO PPOCES5
  Figure 11-3.   teneral process flow diagram at plant #747 showing the sanpling points.
                 Qilorine/caustic (mercury cell) manufacture.

-------
                                                        RECYCLE
H
CTl
in
                  SALT
NONCONTACT H,0      „ n

  TO WASTE *       H-°
                                 BBIHE PROCESS WA1£K
            TO (>H ADJUST-

            HEHT AND FIIIAL
            DISCHARGE
•«—
LAGOON
«to
* W
18

ION
EXCHANGE


LEGENO

SAWLIN6 POINTS




ACTIVATED
CAHSQH




SANtt FHTE.IS



CLARIFIER
*
j

** Fel
i UNDERFLOW TO
UKBFILL

            Figure 11-4.   General process flow diagram at plant 1167  showing the sampling points.
                            Oilorine/catistic  (mercury cell)  manufacture.

-------
waters, which includes the brine purge, cell room liquid wastes
and plant  area wash water.   This is  then  reacted  with sodium
hydrosulfide to precipitate  the mercury  as  mercury  sulfide and
then filtered.  The  solids  are sent  to a mercury recovery unit
and the filtrate  is  sent  to  a holding tank.  The effluent from
the holding  tank  is  mixed with de-ionizer waste and noncontact
cooling  water  before  discharge.    The  process flow .diagram
showing the waste streams sampled is given  in Figure 11-5.

     At Plant §106,  mercury-bearing  wastes  are segregated from
other waste  waters  and combined for  batch treatment.  Mercury-
bearing  leaks,  spills,  and    precipitation are  contained and
collected  by curbing  around  the  cell room  and  collecting the
wastes in  a  common  sump.   From the  sump  the combined waste is
pumped  for  treatment.   In  the  treatment  system,  the  pH  is
initially  adjusted  using waste sulfuric    acid  and  20  percent
caustic solution as required.  Sodium sulfide and filter aid are
added and the waste  agitated in fiberglass  reaction  tanks.  The
effluent  from the  tanks  is  filtered and  the  filter  cake  is
retorted  for  mercury  recovery.    The  residual waste,  after
mercury  recovery, is  placed  in  a  lined solid  waste disposal
area.  The filtrate  is sent  to the  first of two lined lagoons.
Primary pH adjustment  is  made using  waste sulfuric  acid and 20
percent  caustic  before entry  into  the first  lagoon;  final  pH
adjustment is made between the first and  the second lagoons.

11.3.3  Descriptions of Plants Not Sampled

     At  Plant  f589,  the waste  water  going  to the  mercury
treatment  system  consists of  cell room  washdown,  brine filter
backwash,  leaks,  spills, cleanup water, and  hydrogen  cooling
condensate.  The waste  waters are  reacted  with hydrochloric acid
and sodium bisulfide  and then sent to  a settling  basin where
mercury sulfide precipitates.  The overflow  is passed through a
series of effluent filters before discharge.

     At Plant £343,  the cell  room  wash  water, brine purification
sludge, and chlorine cooling condensate are combined  and sent to
a pond.  The suspended  solids settle  in the  pond  and  are dredged
out  once  a  year.   The  dredged  sludge  is "Chem  Fixed"  and
disposed of  in an appropriate landfill.   The overflow from the
pond is reacted with Na2S and  the reacted solution is sent to a
clarifier.    The  clarifier   underflow,   consisting  mainly  of
mercury  sulfide,   is  returned to   the  pond.   The clarifier
overflow is discharged.

     All  contact  waste  water  at Plant  1*907  is treated  for
mercury  removal  in  a  patented process involving reduction  of
mercury  to   the  metallic   state  using  sodium  borohydride.
Previously contaminated  wooden flooring  in the cell  room has
been removed and replaced with fiberglass gratings to reduce the
                               166

-------
H
a\
-j
         SALT
                                                                                                 CEF GAS
                                                                         DE-IONIZED  NQN3MSCT

                                                                         WATER VJVSTTi    COGtlHS
                        tEGBH



                      Waste streams canpled.
                                                                                                    DISCI IAICE
               Figure 11-5.   General process flow diagram at  plant  #317 showing the sampling points.

                               Chlorine/caustic (mercury call)  manufacture.

-------
amount of mercury in the effluent and for better waste control.
Molecular sieves have been installed on  cell  end boxes to reduce
the  mercury content  in the  air vented  from  the cells.   The
treatment not only cleans the air but is also believed to reduce
mercury in the plant area runoff.

     In  the treatment  system,   the  mercury-contaminated  waste
water  is  reacted with  sodium borohydride to  reduce dissolved
mercury to the metallic form.  The reacted solution  is filtered
prior  to  delivery  to one of  the  banks  of  three  columns packed
with anthracite  coal.   After passing  through  three absorption
columns in series,  the treated waste water  is delivered to large
holding tanks,  from which it may be discharged  or returned to
treatment,  depending on  its mercury   content.    Filter, cake,
resulting  from  the filtration of the  waste prior to  the coal
absorption step, is retorted  for mercury recovery.

     Waste solids at this facility,  including mercury treatment
sludges and  brine muds,  are  deposited  in an  on-site  disposal
area.  Chlorine  discharges  are  essentially eliminated by three
significant waste management practices:   the  chlorine condensate
is  collected  and  returned  to   the  brine  system,  tail  gas
scrubbing  effluents are  used in the  manufacture  of  another
product,  and  spent  sulfuric  acid  from chlorine  drying  is
dechlorinated in an air stripper and shipped  off-site  for the
manufacture of another product.   Gases  from the air stripper are
returned to the chlorine purification header.

     At  Plant  #324,  the  barometric  condenser   on  the  brine
dechlorination was  replaced with an indirect cooler, resulting
in  a  reduction of chlorinated  waste  water.    The tail  gas
scrubber  effluent  is  used   for the   manufacture  of  another
product, and the brine muds  are sent  to  a pond.  Small amounts of
mercury, when detected  in the brine mud, are leached with water
and  treated with  other  mercury-contaminated  waste waters which
include the cell room wash  water, caustic filter  backwash, and
brine  leaks.   The  combined waste water is mixed  with hydrogen
processing  waste  water,  reacted with  sulfuric  acid,   sodium
borohydride,  and  sodium  sulfide,   and  then  filtered.    The
filtrate is adjusted for pH and  recycled to process.

     At Plant #385, the brine mud sludge is  sent to  a retention
pond where it accumulates.  All  process contact waste,  water is
collected in an unlined pond where it is treated and  the treated
effluent  is used  as the scrubber liquid for tail gases.   The
spent  scrubber  solution is  sent  to  an adjacent paper plant for
use.

     At Plant  C416, the  cell room  wastes are used  for bleach
manufacture.  The waste water streams from the chlorine/caustic
plant  are sent to an adjacent paper company.
                              168

-------
     At Plant  f784,  the waste  water,  consisting  of  KC1 brine
filter  backwash  and  area  washdown  and  spills,  is sent  to a
basin.  The basin equalizes the  flow and  the overflow is  treated
with   sulfuric  acid   prior   to   reaction   with  NaHS    and
clarification.    The  clarifier  overflow  passes  through  an
activated carbon filter and  to  a  final  tank  where it undergoes
pH adjustment before discharge.

     The wastes are segregated at  Plant  £674.  The clarification
pond is used for  waste streams containing suspended  solids.  The
streams going  to the   pond include brine purification muds  and
spent chlorinated lime.  The mercury-contaminated waste waters
are  treated separately.   These  include the  brine saturation
waste,  brine filter  backwash,  cell room  sumps,  and  tank   car
washes.   The combined mercury-laden  waste water  is sent  to a
collection  pond  and  the  overflow  from the pond  is pH  adjusted
before  the  addition of Na2S.  The reacted solution is sent to a
another pond  and the, pond  overflow  is passed  through  a carbon
adsorption  column before final discharge.  A part  of the  treated
effluent is  re-injected into the brine well.

     At Plant  f012,  the brine  treatment area  is  paved to trap
all  spills,  leaks,   and  rain  runoff  from  that  area.    The
recovered  waste  is  recycled to the weak brine reservoir.   The
contaminated  waste  waters  from the plant  are  re-injected into
the  brine  wells  to  keep  the  hydraulic balance  and  maintain
pressure in the salt deposits.

11.3.4  Summary of the Toxic Pollutant Data

     Presented below  are the toxic  pollutants  found in the  raw
wastes  during screening and verification.

     Because  several  waste  streams  usually contribute  to   the
total raw waste at mercury cell plants,  a calculation was often
necessary  to determine the  pollutant  concentrations that would
exist  in the  streams before they were mixed prior to treatment.
An example  of  this calculation  is the "mixing" of the following
hypothetical streams:

     Stream A:   100 gallons per minute,  15 mg/1

     Stream B:   10 gallons per minute,  60 mg/1

      (Flow x concentration) + (Flow x concentration)
                        Total Flow

      =  concentration of mixed streams

      =  (100 gpm)(15 mg/1)  +  (10 gpm) (60  mg/1)     =  19 mg/1
                        110 gpm
                               169

-------
     The maximum  raw waste concentrations  observed during  any
single 24-hour sampling period were:


           Maximum Raw Waste Concentrations Observed
                             (yg/i)

Pollutant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Thallium
Zinc

Screening
Plant
(#299)
< 250
< 10
< 1
8
350
1
150
< 100
< 1
140
230
Verification
Plants
(f299, f747, f!67,
?206, $317)
770
400
790
180
2,300
1,900
180,000
2,400
870
440
34,000
     Section 5.1.2 of  this  report  describes the methodology of
the  screening  and  verification  sampling  program.     In  the
chlorine mercury cell  industry,  a  total of 18 days of sampling
were  conducted  at  Plants  f299,  t'747, f!67,  £317  and f!06.
Thirty-two  different  sampling  points  were  involved covering
various  raw waste streams  and  the treated  effluents at these
plants.  The evaluation of  toxic metal  content of  these  process
related  waste  streams  was based on 949 analytical data  points.
The screening  for  toxic organic pollutants  at  Plants i'299 and
$167 generated an additional 490  analytical data points.  The
daily raw waste loads were calculated  from  the waste stream flow
rates  measured or estimated  at the  time  of  sampling  and the
measured pollutant concentration.

     The daily loading  is determined  by:

         Daily loading  (as  kg of pollutant    (C)(Q)
         per day)                          =  1000

     Where:

         C  is  the concentration of the  pollutant  expressed in
         units of mg/1  (Note: kg/m3 = 1000 mg/1),  and

         Q  is  the waste stream flow rate expressed in units of
         m3/day  (m3, a cubic meter, is  equal  to 264.2 U.S.
         gallons).
                               170

-------
     Similarly,  the unit  loadings  were  calculated  from  the
reported chlorine production  rate,  the waste stream flow rate,
and the measured pollutant concentration:

     Unit loadino (as kg of pollutant per     (C)(Q)
     kkg of chlorine)                       = 1000P

     Where C and Q are the same as  described  above,  and P is the
chlorine production  rate expressed  in  units  of  kkg/day  (kkg is
1000 kg, a metric ton, which is equal to 2205 Ibs).

     The minimum, average, and maximum values are based on data
from those plants where the particular pollutant was found at a
concentration greater than the analytical  detection limits and
considered a "significant concentration".  The term  "significant
concentration" means an observed concentration in any 24- or 72-
hour composite  raw waste  sample  that is  above the analytical
detection  limit,  and  treatable   by   an  available  technology
regardless of economic considerations.

     In  Table 11-6,  the  toxic  pollutant  raw  waste data  are
presented  as  the  average  daily  concentrations and the  unit
loadings found  at  the individual plants.  These averages were
derived by averaging the concentrations and loads based on three
24rhour composite samples from each plant.

     In Table 11-7 daily loadings  (in kg/day) and unit loadings
(in kg/kkg)  are presented  as minimum, average  and maximum values
based on the data presented in Table 11—6.

     Based on  the  total annual production of this subcategory
and  the average  waste load  generated  per  unit  product,  the
estimated total pollutant  raw  waste loads generated  each year by
this subcategory are as follows:

                                       Raw Waste load
                   Pollutant               (kg/year)

                   Antimony                1,400
                   Arsenic                 1,000
                   Cadmium                  210
                   Chromium                 360
                   Copper                   960
                   Lead                     880
                   Mercury               44,000
                   Nickel                   820
                   Silver                   850
                   Thallium                 770
                   Zinc                    7,200
                               171

-------
TABLE 11-6.  TOXIC POLLUTANT RAW WASTE CONCENTRATIONS AND LOADS AT
             VERIFICATION PLANTS
^kg/kkgy

SUBCATEGORY
Pollttfcant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Thallium
Zinc
CHLORINE (MERCURY CELL)
299
0.48
0.00077
0.23
0.00037
0.010
0.000016
0.063
0.00010
0.30
0.00047
0.060
0.000096
5.9
0.0081
*
*
0.18
0.00029
0.27
0.00043
Plant
747
0.11
0.000078
0.030
0.000021
0.020
0.000014
0.10
0.000071
0.38
0.00027
0.16
0.00011
18
0.0043
0.093
0.000066
0.047
0.000033
0.022
0.000016
0.69
0.00049
#
167
*
0^33
0.0011
*
0.12
0.00040
0.075
0.00025
0.072
0.00024
3.8
0.013
0.060
0.00020
*
*
0.17
0.00057
317
*
0.10
0.00005
0.46
0.00023
0.080
0.000040
1.2
0.00060
1.4
0.00070
123
0.048
1.4
0.00070
0.11
0.000055
*
20
0.010
106
0.49
0.00070
*
0.031
0.000044
0.013
0.000019
0.12
0.00017
0.33
0.00047
3.9
0.006
0.17
0.00024
0.58
0.00083
0.38
0.00054
0.96
0.0014
* - Concentration below significant level as defined in 11.3.4.
                                      172

-------
                      11-7 „          OF RAW       LOADINGS AT
                             VERIFICATION PLSNTS

SUECATEGORY
Pollutant


min.
Antimony
Arsenic
Cadmium
Chratmm
Copper
Lead
Hercury
Nickel
gilver
0
0
0
0
0
0
1
0
0
.044
.0054
.0062
.0043
.045
.036
.6
.037
,0059
UsUiuE 0, 0090
Zinc
0
.14
CHLORINE
Daily
Loadings
(kg/day)
avg.
0.17
0.11
0.013
0.037
0.10
0.070
3.1
0.056
0.082
0.086
0.41
(MERCURY C

max.
0,30
0.27
0.025
0.098
0.18
0.12
5.1
0.075
0.22
0,14
1.1
at)



Unit
Loadings
(kg/kkg)
min.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
000078
000021
000014
000019
00025
000096
0043
000066-
000033
000016
00043
,0
0
0
0
0
0
0
0
0
,0
0
avg.
.00052
.00038
.000076
.00013
.00035
.00032
.016
.00030
,00031
.00028
,0026


Number of
Plants
Averaged*
max.
0.
0.
0.
0.
0.
0.
0,
0.
0.
0,
0.
00077
0011
00023
00040
00060
00070
048
00070
00083
00054,
010
3
4
4
5
5
5
5 .
4
3
3
5
* - Only those plants where the pollutant was observed at  "significant
    concentrations" are included in the averaging.   "Significant
    concentrations" is defined in 11.3.4.
                                      173

-------
11.4  POH.OTION ABATEMENT OPTIONS

11.4.1  Toxic Pollutants of Concern

     Mercury  is the  major toxic  pollutant of concern  in the
production of chlorine by the mercury cell process.  Other toxic
metals often  found  in significant concentrations in raw wastes
include  arsenic, antimony,   cadmium,  chromium,  copper,  lead,
nickel, silver,  thallium and  zinc.  Sources of  these metals are
assumed to be impurities in the raw salt  or brine and corrosion
products  from   the  reaction  between   chlorine  and  process
equipment materials of  constuction.    No  toxic  organics  were
found at significant levels.

11.4.2  Prey jailing Control and Treatment Practices

     Section  11.2.3 described  specific  control  and  treatment
practices at 14  plants.  All  known mercury  cell plants practice
treatment of mercury-bearing  wastes,  but  control practices such
as recycling  of brine mud filtrate or  pond overflow, chlorine
condensates,  hydrogen  condensates  and caustic filter backwash,
and  solids  handling vary  from plant  to plant.   Although all
known treatment  facilities precipitate mercury  and separate the
solids formed by clarification and/or filtration, sampling data
has  shown that  some treatment systems including those with more
advanced technologies  such as' adsorption or  ion  exchange, are
not  operating efficiently.

11.4.3  Process Modi£ications and Technology Transfer  Options

     The following process modifications  are being practiced at
one  or more  mercury cell  plants   and can significantly reduce
pollutant loads  discharged.

Anode Material

     Nearly all  mercury cell plants now use metal  anodes.  Their
use,  as  opposed   to   graphite   anodes,  improves  the  power
efficiency  of  the  cells  and  reduces the  potential  pollutant
load.

Liquefaction of  Chlorine

     Utilization of high pressure and  refrigeration for chlorine
recovery will reduce the chlorine content of tail gases.

Brine Recycling  •

     Although practiced  at many facilities  not all  plants are
using a closed-loop brine  system which eliminates a significant
waste volume requiring mercury treatment.
                               174

-------
Mercury Emissions

     Hydrogen gas  produced in the  denuder  can be refrigerated
and passed through treated carbon or molecular sieves to remove
the mercury escaping with the gas.'  This will  reduce  the mercury
emissions and reduce atmospheric fallout in the neighborhood of
the plant.  This  in turn will reduce mercury concentrations in
storm  runoff.     Two  plants  are  practicing   this  control
technology.

Tail Gas Emission Control

     When chlorine gas produced from the cell is compressed and
cooled,   chlorine    separates  '  as   liquid   chlorine,   and
noncondensable  gases  (tail  or sniff  gas  containing residual
chlorine  vapor)  are  produced  at  the  discharge  end  of  the
condenser.  The amount of chlorine present  in  the tail gas is
significant and has  to  be  removed and  treated  or recovered
before the  tail gas  is  vented  to  the  atmosphere.   The common
industrial practice  is to scrub the  gas with caustic  soda or
lime  solution  thus  producing the  corresponding  hypochlorite.
The hypochlorite solution  is either sold, used on-siter sent to
a waste water treatment plant, or discharged without  treatment.
Treatment  of  this  waste  is a    relatively  recent practice.
Decomposition is  a  common  method  of treatment using  catalytic,
thermal, and chemical methods as described below.

     Catalytic  decomposition  involves  the  addition of  small
quantities of  cobalt, nickel, and  iron chloride  to the waste
streams, followed by retention in reaction tanks for  periods up
to several days.  Of  the two plants employing this  technology,
one reports zero  discharge of chlorine,  and the other reports
respective average and maximum chlorine  discharge rates of 0.015
and 0.14 kg per metric ton of chlorine produced.

     Thermal  decomposition occurs  when the  temperature  of the
solution containing  hypochlorite  reaches 175 degrees F.   Lime
reacts with chlorine exothermically, producing heat  and calcium
hypochlorite.    If  the  hypochlorite solution  is not  cooled,
thermal decomposition  occurs .   One  chlorine/caustic plant is
using  this treatment  method and  another is planning  to use it.
The   plant   using   thermal  decomposition   reports  complete
conversion of hypochlorite to chloride.

     Chemical   decomposition  takes  place   by   reacting  the
hypochlorite solution with a chemical reactant which  is usually
sodium sulfite  or hydrogen peroxide.  Chemical decomposition is
expensive but complete and rapid.

     When chlorine is present in a  dissolved form  (hypochlorous
acid)  in water, a stripping technique may be applied  to recover
                               175

-------
the chlorine.  Chlorine  condensate  streams and spent chlorine-
drying  acid  are  most  commonly  treated  by  steam or  vacuum
stripping, with the chlorine frequently returned to process for
purification  and  recovery as  a  product.   The  tail  gas  is not
generally scrubbed with water because water does not effectively
remove  chlorine  and the chlorine concentration  in the exhaust
will  reach  0.1 to 4.5 percent by volume  after  scrubbing with
water.  One effective method of chlorine recovery from the tail
gas is  by the passage  of the  gas through an absorbing material
such  as carbon  tetrachloride  and  subsequent recovery  of the
chlorine.  The process is proprietary and little information is
available on  its design  or performance.

11.4.4  Best Management Practices

Area Runoff

     Provisions can be made  to divert and contain storm runoff
from  plant  areas.   Collected  runoff  can  then be sent  to the
waste water treatment system.

Leaks and Spills

     The brine treatment area and  the cell  room  areas  can be
paved with fiberglass gratings, and provision should be made to
collect the leaks  and spills from the operation.

Mercury Contaminated Solids

     The precipitated mercury waste should be stored in a lined
pond,  disposed of  in  a  secured landfill  or sent  to mercury
recovery operations.  Brine mud should be discharged to a lined
pond  or a  secure  landfill  after filtration.   The  brine mud
contains  small amounts  of mercury  which  can leach  into the
ground water  if proper safety precaution are not taken.

Transportation, Handling and Abnormal Operations

    'Provisions  should  be  made  to  remove  chlorine  from  air
emissions resulting from abnormal operating conditions such as
start up  and  shut down, or  from vents on returned  tank cars,
cylinders,  storage tanks,  and  process  transfer  tanks  during
handling and loading of  liquid chlorine.

11.4.5  Advanced Treatment Technologies

     Methods  available for the removal  of  elemental mercury or
mercuric  salts from plant  waste waters  include precipitation
with  sodium   sulfide  to  form   insoluble  mercuric  sulfide,
adsorption by activated  carbon,  adsorption by ion-exchange and
other  resins, reduction by  borohydrate,  hydrazine,  sulfite,
                               176

-------
hypophosphite or  iron,  and  biological reduction  (57).   All of
these  methods  are  patented?  many  of  these methods  have been
proven  on  a pilot  scale  only.   Sulfide  precipitation  and
adsorption techniques will also provide for the removal of other
toxic metals.


11.5  SELECTION OP APPROPRIATE TECHNOLOGY AND EQUIPMENT


11.5.1  Technologies for Different Treatment: Levels

     Following  the  evaluation of  significant  toxic pollutants
found in raw waste waters, current  industry treatment practices
and applicable treatment alternatives,  two levels of end-of-pipe
treatment were selected  as  alternatives  for  application in the
mercury cell chlorine subcategory.

Level 1

     This   treatment  consists   of  sulfide  precipitation  of
mercury-bearing  waste  water  followed  by  pressure filtration.
This  level  of  treatment,  which  will  also reduce  other heavy
metals,  includes recycle of  the  brine  waste  stream  back  to
process, and the settling and storage  of brine muds.  Mercury-
bearing solids can be sent to  mercury recovery or disposal.  The
flow diagram for  this treatment level is  shown in Figure 11-6.

Level 2

     The filtered Level 1 effluent  is passed through a granular
activated  carbon bed  where  residual  metal  sulfides and  any
metallic mercury will  be removed.   The  flow  diagram for this
treatment level  is shown in Figure  11-7.

11.5.2  Equipment for Different Treatment Levels

Equipment Functions

     In  Level  1, typical  of  existing  treatment  facilities,
mercury-bearing  wastes   are  equalized  in  a  surge  tank,  and
following chemical mixing,  sulfide precipitates are removed in a
conventional plate and frame  filter press followed by final pH
adjustment  of  the  filtrate before discharge.    In Level  2  a
conventional  granular  activated  carbon  filter  is  added  for
further removal  of residual metals  before pH adjustment.

Chemical Handling

     Sodium  bisulfide   is  used   with  filter  aid  after  pH
adjustment  to pH  5-7. Care  is needed to prevent escape of toxic


                               177

-------
                                                ]•	^
                              BRIM:
                            HUD STREW
1                        f
   «	*^V    MGOOH     X  »*
                                                                                        . RECYCLE TO
00
                                         sutftmicftdD
                      FEWER
                        AID
                 HERCUOT
                CXWBWJKMH)-
                VJASTESWEPH
                                                                          SODIUM
                                                                         BISULFIDE
                                                                         n
                                                                                                    FH/THt
                                      HOLDING WNK          MUttHS
                                                                              MIXING
                                     i—-g-
                                                                                                 SOLIBS TO
                                                                                                  MERCURY
                                                                                                  RBOOVHW
                                                                                                ORU11WIU.
                                                               BETCHCV BBUJHH UM!
                                                                                         	1	
                             Includes pH monitoring, flow monitoring and sampler
                   Figtire 11-6.   Level  1 waste water treatment for chlorine - mercury cell  subcategory.

-------
H
-4
vo
                                i' -'	^
                  BRIM: MUG
                   smsm
  mcuig
aOBBlMtNKTEB
 S&S1E S&T13R
                                         fflSCENCir W.TUPN LINE
                                                                     SDUBS TO
                                                                      Harare
                                                                      KBCWSW   i
                                                                    OR IMtnU.  I
                                                                                                             -a^-EFPUiwr
                        Inclndeii pH monitoring, flow monitoring and sampler
            Figure  11-7.  -Level 2  waste water treatment  for chlorine  - mercury cell subcategory.

-------
and obnoxious H2S fumes at neutral and acid  pH levels.  At Level
2 no  additional chemicals are used  since the activated carbon
bed is not regenerated but is periodically removed and replaced.
The.  handling  of  granular  carbon  may  cause  temporary  dust
problems but it causes no special hazards.

Separation and Removal of Solids

     Conventional settling and filtration methods are used, but
because of the toxicity of mercury, precipitated sludges should
be disposed of in a safe chemical waste area.

Monitoring Requirements

     Both  levels  of  treatment include provisions  for sampling
and monitoring  of  the waste water  discharge.   Monitoring  of
heavy metals is done by atomic absorption  methods at a qualified
commerical laboratory:  Simple field  tests for heavy metals as a
group are available for routine process control.


11.6  TREATMENT COST ESTIMATES


11.6.1  General Discussion


     To prepare treatment cost estimates,  a   model plant concept
was developed.  The proposed model plant characteristics are:

Waste Water Plow

     Data  presented  in  Table 11-3  indicate an average  waste
water flow of 2.1 m3/kkg for  13 plants, while the average of the
five plants surveyed during this  study averaged  ].7 m3/kkg.  The
latter value was used  for  developing  the detailed cost estimates
presented in the cost tables  because  the technology base for the
model plants was that observed in the  field.

     For effluent limitation calculations (see 11.7.2) the more
conservative unit flow from the larger  data  base and 2.1 m3/kkg
has been used.   Cost  estimates will  be adjusted to reflect the
larger unit flow before promulgation.

Chlorine Production

     Approximately 50 percent of  the  production  data for all the
chlorine/caustic  plants using mercury cells  is available  on
file.    Production   ranges  from  19,000  to  198,000  kkg  of
chlorine/year.   Three model  plants  with  productions  of 19,100
kkg/yrr  95,500  kkg/yr  and  191,000  kkg/yr  were  selected  to
                               180

-------
 represent  the  subcategory production range.   The flow per  unit
 of  production  is  assumed to be the same for each size of model
 plant.  Seventy-seven percent  of  the  plants  for  which flow  data
 was available  have flows per unit of production equal to  or  less
 than  the average  unit flow (Table 11-3).

 Solid Waste Produced

      Brine mud  constitutes the  major  source  of  solid waste
 generated  at chlorine plants.  Although flows and solids  content
 varies considerably from plant to plant, an average flow  of  0.42
 m3/kkg at  10  percent  suspended solids gave an estimated solids
 load  of 42 kg/kkg to be  used for  cost estimating purposes.

 11.6.2  Chlorine  Bearing Wastes

      In the selection of model plants, the following assumptions
 have  been made for the chlorine contaminated  waste streams.  The
 chlorine  condensate  waste  stream  has not been  included in the
 waste streams  going to the treatment  facility.   In  the majority
 of  the  chlorine/caustic  plants,  this  stream  is  stripped  of
 chlorine by steam or vacuum and the chlorine is  recycled to the
 purification operation.  The waste water is then returned to the
 process and introduced to the brine purification unit or  sent  to
 the  treatment unit.   The  quantity of waste water  generated  by
 this  operation is* small and does not significantly affect the
 flow  determination.   In some  cases  the  chlorine gas from the
 cells is  contact cooled with water  and  the  scrubbed  liquid,
 after steam  stripping,   is  reused.   The stripping  operation  in
 the recovery of chlorine is part  of  the process  and,  therefore,
 its cost  is not  included in the treatment  cost.   The spent  tail
 gas scrubber  solution,  which contains mainly calcium or sodium
•hypochlorite,  is  assumed to be used  or decomposed  before it  is
 discharged or  sent to  treatment.   Thermal decomposition can  be
 practiced  at  no  additional  cost at  some  facilities, while
 another  efficient treatment method is catalytic decomposition.
 The   cost  estimates  for decomposition  are  not included  here
 because at many plants the hypochlorite stream is sold, used on-
 site  or only  infrequently discharged  depending on market demand.

      However,  because  of  the environmental  effects of   high
 levels of  chlorine in waste water discharges, the  cost  for the
 dechlorination of total plant discharges  using  sulfur  dioxide
 has been  included because this is the treatment  method on which
 control of total  residual  chlorine is based.

 11.6.3  Model  Plant Treatment Costs

      On  the  basis of the model plant specifications  and design
 concepts presented earlier,  the estimated.costs of treatment for
 three models  having  different production levels are shown  in
                               181

-------
Table 11-8r 11-9 and 11-10.  The costs of Level 2 treatment are
incremental over  Level 1  costs.   Annual treatment  costs  as a
function  of  production are  shown  graphically in  Figure 11-8.
Similarly, treatment cost per metric ton of product is shown in
Figure 11-9.   Table 11-11 presents a summary  of  the  unit  cost
distribution between amortization and operation and maintenance
components.

     Variability  in  specific plant  water  use practices  and
treatment applications  may be responsible  for treatment costs
that are higher than those developed for  €he model plant.  These
variations have been considered.  Using  the model plant annual
cost curve as  a baseline, consideration of  the additional plant
specific  cost  factors  results  in the  cost  analysis as shown in
Figures 11-10  and 11-11 for Level 1 treatment  and Figures 11-12
and 11-13 for the Level 2  treatment.  The cost  envelopes reflect
the  impact  of  higher flows  (2.4 m3/kkg) which are required at
some plant locations and the consequent increase in costs due to
additional  chemical requirements  and the  variability  in  the
costs associated  with  solid waste  disposal.   A  combination of
these and other specific plant factors may  result in additional
costs ranging  from  30 to 125 percent of the baseline costs.

     Cost  estimates are  presented  in Table   11-12  for  plants
requiring dechlorination of waste waters  by  sulfur dioxide.  For
the range of model plant productions, the annual cost of sulfur
dioxide treatment varies from  $1.72  to $0.40  per  metric ton of
product.

11.7  BASIS FOR REGULATIONS
11,7.1  Basis for BPT Limitations

Technology Basis

     Existing  mercury  cell  chlorine  plants  are  controlling
mercury in  their  waste waters in  accordance with existing BPT
regulations  which  require  a  discharge of  less  than  0.00014
kg/kkg of product  as  a 30-day average.   These  BPT regulations,
40  CFR.415.62  (a)  presently  in effect will  not  be revised.
Pollutants  regulated  include  TSS and mercury.   The technology
basis of sulfide precipitation and filtration of mercury bearing
streams  (Level  1)  is currently  being applied  at  24  plants in
this  subcategory.    Other  plants  in  the industry  use  mercury
control methods that  are different  in detail but with the same
objective.

     The  existing  regulations,  presented   in  Table  11-2,  are
sustained by  the fact that  plants having properly operated BPT
technology  have  demonstrated  the achievability of the effluent
                               182

-------
                    TABLE  11-8 MODEL PLANT TREATMENT COSTS
   Subcategory  CHLORINE  Mercury cell

   Production        19,100 metric tons per year  (21,057 tcais pet year)
                         54 metric tons per day   (60 tons per day)
   Waste water flow      91 cubic meters" per day.
                                             LEVEL OF 7REMMENT*

                                           FIRST            SECOND
A.  INVESTMENT COST
    Construction 	               $49,100              $500
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls	                68,100            15,000
    Monitoring equipment
    in place...	                 9,000
    Engineering design
    and inspection...	                25,240             3,100
    Incidentals, overhead,
    fees, contingencies...                25,240             3,100
    Land.......	                21,000

    TOTAL INVESTMENT COST               $197,680           $21,700

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision.              $112,000  _         $14,000
    Energy	                 1,250
    Chemicals	                   500             1,400
    Maintenance	                17,668             2,170
    Taxes and insurance...                 5,930               651
    Residual -waste
    disposal...... „.„	                 4,400
    Monitoring, analysis
    and reporting...,,.....                15,000             7,500

    TOTAL OPERATION AND
    'MAINTENANCE COST                    $156,748           $25,721

C.  AMORTIZATION OF
    INVESTMENT COST                      $28,745            $3,530

    TOTAL ANNUAL COST                   $185,493           $29,251


    *First level represents the base cost of treatment system.
    Other levels represent the incremental cost above base cost.

                                      183

-------
                    TABLE .^11-9 MODEL PLANT TREATMENT COSTS
   Subcategory  CHLORINE  Mercury cell

   Production        95,500 metric tons per year  (105,288 tons per year)
                        272 metric tons per day   (300 tons per day)
   Waste water flow     455 cubic meters per day.
                                             LEVEL OF TREATMENT*

                                           FIRST            SECOND
A.  INVESTMENT COST

    Construction 	
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls	
    Monitoring equipment
    in place	
    Engineering design
    and inspection	
    Incidentals, overhead,
    fees, contingencies...
    Land	

    TOTAL INVESTMENT COST

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision.
    Energy	
    Chemicals	
    Maintenance	
    Taxes and insurance...
    Residual waste
    disposal	
    Monitoring, analysis
    and reporting	

    TOTAL OPERATION AND
    MAINTENANCE COST

C.  AMORTIZATION OF
    INVESTMENT COST

    TOTAL AtttJUAL COST
$134,500



 141,300

   9,000

  56,960

  56,960
  63,000
$461,720
$112,000
   3,700
   2,500
  39,872
  13,851

  21,400

  15,000


$208,323


 $64,871

$273,194
 $1,000



 61,000



 12,400

 12,400


$86,800




$14,000

  7,000
  8,680
  2,604



  7,500


$39,784


$14,122

$53,906
    *First level  represents the base cost of treatment system.
    Other levels  represent the incremental cost above base cost.
                                      184

-------
                    TABLE 11-10 MODEL PLANT TREATMENT COSTS
   Subcategory  CHLORINE  Mercury cell

   Production       191,000 metric tons per year  (210,577 tons per year)
                        545 metric tons per day   (601 tons per day)
   Waste water flow     910 cubic meters per day.


                                             LEVEL OF TREATMENT*

                                           FIRST            SECOND
A.  INVESTMENT COST

    Construction	              $257,700            $2,'000
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls	               213,200           115,000
    Monitoring equipment
    in place	                 9,000
    Engineering design
    and inspection	                95,980            23,400
    Incidentals, overhead,
    fees, contingencies...                95,980            23,400
    Land	               123,000

    TOTAL INVESTMENT COST               $794,860          $163,800

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision,              $112,000           $14,000
    Energy	                 6,400
    Chemicals.............                 5fOOO            14,000
    Maintenance	                67,186            16,380
    Taxes and insurance...                23,845             4,914
    Residual vaste
    disposal....	                42,600
    Monitoring, analysis
    and reporting	                15,000             7,500

    TOTAL OPERATION AND
    MAINTENANCE COST                    $272,031           $56,794

C.  AMORTIZATION OF
    INVESTMENT COST                     $109,311           $26,650

    TOTAL ANNUAL COST              -     $381,342           $83,444


    *First level represents the base cost of treatment system.
    Other levels represent the incremental cost above base cost.

                                      185

-------
500
o 400
o
•w-
1 30°
200
inn
100 500 " 1000 DAILY
1 1 1 (m3









!
1 !





1








! i








































.



























-------
u
u
     10

           LEVEL i'
                  \
                                              1  !
                                              I  I

                                                                  !  I
                50       100       150       200
                   PRODDCTICH (METRIC TONS/SEAR X 1000)

   Figure H-9.  Annual unit txeatment cxjst vs. production for the Chlorine
                   Sxibcatecpry (Mercury Cell Process)
                                 187

-------
                     TABLE 11-11.  MODEL ELMO? UNIT TR1MMENT COSTS


    Subcategory  CHLORINE  Mercury cell




                                               Annual Treatment Costs ($/kkg)


                                                     LEVEL OF TREATMENT

     COST HEM       PRODUCTION  FLOW      FIRST     SECOND*   THIRD    FOURTH
                      (kkg/yr)   (m3/<3ay)
    Annual Operation
    and Maintenance
    Annual
    Amortization
    Total Cost
 19,100
 95,500
191,000
 19,100
 95,500
191,000

 19,100
 95,500
191,000
 91
455
910
 91
455
910

 91
455
910
8.21
2.18
1.42
1.50
0.68
0.57

9.71
2.86
2.00
1.35
0.42
0.30
0.18
0.15
0.14
1.53
0.56
0.44
JN&t Applicable
* — These costs are incremental to first level costs.
                                     188

-------
500
•400
 300
 200
                                 f 1 COST
 100
               I
I
                    J_
I
I
50
                                     300
                       100       150      200      250
                  PKSOCTION (METRIC ims/mR x 1000)
     Figure 11-10.   Annual treatment cost vs. production for tihe Chlorine
                     Subcategory (Mercury Cell Process).

                                   189

-------
    14
    12
     10
to-
                        IEVEL fl COST
                  I
           I
I
50       100       150
    PRXDCTION (METRIC
                                              200
                                                  x 1000)
        Figure 11-11.  Annual unit treatment cost vs. production for :the
                       Chlorine Subcategory (Mercury Cell Process).

                                      190

-------
   600
   500
©400
   300
                             IEVEL #2 COST RSNGE
   200
   100
                           I
             J_
                50        100       150       200
                    PIO3UCTIQN (METRIC TCSNS/XE&R x 1000)
        Figure 11-12.
annual treatment cost vs. production for the Chlorine
 Subeategory (Msrcury Cell Process).

               191

-------
16
14
12
10
                                  RANGE
                    I
J_
I
          50       100       150       200
             PTODUCTION (METRIC TONS/YEAR x 1000)
 Figure 11-13.   Annual unit treatment cost vs.  production for the
                Chlorine Subcategory (Marcury Cell Process),

-------
TABLE 11-12.  ESTIMATED CHEMICftL DECHLORINATION COSES FOR THE CHDORrALKALI

SUBCATEGORY CHLORINE
Chlorine Production (kkg/yr)
A.













B.












C.


INVESTMENT COST
Construction. ..........
Equipment in place,
including piping,
fittings, electrical

Monitoring equipment

Engineering design

Incidentals, overhead,
fees, contingencies. . . .
Land 	 	 	 	
TOTAL INVESTMENT COST
OPERATING AND
MA3MTENANCE COST
Labor and supervision. .
Enerqv 	 	 	 	
Chemicals (S00).. 	

Maintenance. ...........
Taxes and insurance. . . .
Residual waste
disposal. ..............
Monitoring, analysis,
and reporting. .........
TOTAL OPERATING AND
MAINTENANCE COST
AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
(MERCURY CELL).
19,100

$3,000
•j'1— ' f W W V


20,100

^.^

4,600

4,600
— —
$32,200


14,000
500
1,500
3,220
966

,„„.„

7,500


$27,686

$5,239
$32,925

31, 850
.
$5,000



35,000

mmm—

8,000

8,000
—
$56,000.


26,000
659
2,000
5,600
1,680

ox... ..........

7,500


$43,439

$ 9,111
$52,550

191,000

$10,000
«!»**«. %* f \s w w


50,000

W.M.

12,000

12,000
w****.
$84,000

,
28,000
1,220
15,000
8,400 *
2,520

— .

7,500


$62,640

$13,660
$76,300
    COST PER KKG OF
    PHDDUCT (Dollars)
1.72
1.65
0.40
                                     193

-------
limitations based on available long-term monitoring data.  Table
11-13 presents  data from eleven mercury  cell plants, seven of
which are  meeting the  30-day average  limitations.   The other
four plants  have mercury control  technology  installed but are
not meeting BPT limits.

Flow Basis

     The existing  regulations contained only load limitations,
kg/kkg, and  no  flow  basis or  concentration limit was provided.
But  the regulations did  consider the  inclusion of noncontact
cooling water in determining  discharge  load limitations.

11.7.2  Basis for Proposed BAT Effluent Limitations

     The original BAT limitations  for this subcategory required
zero  discharge  of  process  waste  water  pollutants.    These
regulations were remanded and are  not in  effect.  The proposed
regulations  allow  for  the  discharge  of process  waste water
following treatment.

Technology Basis

     Utilizing  the cost estimates  presented in this report, the
Agency has analyzed  the cost  effectiveness of Level 1 and Level
2 treatment options  for pollutant  removal.  The  economic impact
on the mercury  cell  chlorine  subcategory  has  been evaluated in
considering the technology basis for proposed BAT limitations.

     For BAT, the Agency  is proposing limitations based on BPT
technology   (Level  1)   with  the  addition of   dechlorination.
Dechlorination  is being included in BAT because  the toxicity of
chlorine to  aquatic life is  well  documented  (59) and  it  is a
pollutant of concern to the Agency.   Dechlorination, currently
practiced at two plants, may be required only  at  fewer than half
of the plants in  the subcategory because  hypochlorite produced
in tail gas scrubbers is often sold or  used in other operations
while residual  chlorine in condensates is usually  stripped or
recovered.  Table 11-14 presents residual  chlorine discharges at
plants  that  have  reported   the  use,  sale  or treatment  of
chlorine-bearing  waste  waters.    This   data   indicates  that
dechlorination    technology    has    not    been   successfully
implemented.

     The Agency considered the addition of carbon adsorption for
additional mercury removal but rejected its use   because of high
cost and questionable performance  in  this  industry.
                               194

-------
    TABLE 11-13.   MERCURY DISCHARGES FROM SELECTED CHLOR-ALKKLI MERCURY

                  CELL PISHES*

SUBCATEGORY
Plant
#343
#907
#898
1195
#106
#589
#299
#747**
#317**
#195**
#324**

Average
0.000025
0.000020
0.000060
0.000040
0.000065
0.000055
0.000040
0.000055
0.000006
0.000022
0.00086
CHLORINE (?>IERCURY CELL)
Mercury
Daily Maximum
0.00094
0.00026
0.0025
0.00073
0.00022
0.00086
0.00019
0.000083
0.000048
0.00066
0.0022

laste Load (kg/kkg)
Maximum 30 -day Average
0.00029
0.000030
0.00043
0.00015
0.000096
0.00049
0.000056
0.000065
0.000010
0.00010
0.0018
*    See Reference 3
                                            t


**   Ercm Plant Long Term Monitoring Data presented in Appendix A.
                                     195

-------
TlfflEB  11-14.    RESHTOL CHLORINE DISCHARGES AT SELECTED
                CHDOR-JffiKKLI


Plant
1207
f 014
* 819
# 747
# 106
* 589
# 747**
* 324**

Average
0.33
0.04
ND
0.002
0.001
0.003
. 0.0025
3.72
Chlorine Waste Load (kg/kkg)
Range


1.4 maximum
0 to 1.29
0.016 to 0
0 to 0.006
0 to 0.14
0.001 to 0
ND
0.38 to 12

.14


.011

.2
*See Reference 3
**Frcro Plant long Teim Monitoring Data
                                  196

-------
Flow Basis

     The flow  basis  for  BAT limitations is 2.1 m3/kkg based on
the average of discharge data of 13 plants  presented in Table
11-3   The  order  of magnitude  of  this  unit flow  volume was
supported by data obtained during sampling  visits to five plants
at  which  flows  ranged  from 0.5  m3/kkg to 5.6 m3/kkg  with an
average of 1.7 m3/kk"g.

Selection of Toxic Pollutants to be Regulated

     The selection of pollutants  for   which  specific effluent
limitations  are proposed  was based on the  evaluation  of raw
waste  concentrations found during the  sampling program and on
t'he treatability of  toxic pollutants using BAT technology.
                                                *
     Table 11-15 presents the achievable concentrations of toxic
pollutants  using the BAT technology of  sulfide  precipitation
followed by filtration.   The concentrations, based on  literature
treatability  data presented  in  Section 8,1  and  summarized in
Table  8-11,   reflect  the  lowest  level  achievable  by  this
technology  for arsenic,   cadmium, copper,  lead,  nickel, silver
and  zinc.    For  antimony,  chromium  and  thallium,   literature
treatability data  are not available for this technology.  Also
presented in Table 11-15 are the maximum and average raw waste
concentrations  of toxic  pollutants  found during  the sampling
program with  an indication of the  number of  plants  where the
treatability concentration was exceeded.

     Based  on  the occurrence of treatable levels  of specific
toxic  metals  in  raw wastes  and  the  fact  that  the  sulfide
precipitation  technology  is  already utilized as  BPT  in the
chlorine mercury cell  subcategory,  arsenic,  cadmium,  copper,
lead,  nickel,  silver  and  zinc were selected as additional toxic
pollutants proposed  for BAT regulations. Antimony, chromium and
thallium were  included for  guidance  but no limits are proposed
because concentrations  found in the raw  waste load were below
treatable levels.

Basis  of Pollutant Limitations

     Limitations are presented as both concentrations  (mg/1) and
loads  (kg/kkg) for each pollutant.  The  relationship between the
two is based on the  unit flow rate.   Although actual unit flow
rates  at  plants  vary by  an order  of  magnitude  due  to such
factors as raw materials and plant control practices,  the Agency
has determined that  the  load limitations  can be met by well-
operated  treatment  facilities.   The concentration  or quality
limits are  included  below.

     BAT proposed limitations are presented in Table  11-16.
                              197

-------
   TABLE 11-15.   COMPARISON OF RAW WASTE CONCENTRATIONS OP TOXIC POLLUTANTS
                 WITH TREATAKLLITY

SUBCATEQORY
Pollutant
Antimony
Arsenic
Cadmium
Chroroium
Copper
Lead
Nickel
Silver
Thallium
Zinc
CHLORINE
Treatability * '
CtngA)
_ (2)
t
0.05
0.01
__ (2)
0.05
0.10
0.05
0,05
__ (2)
0.20
(MERCURY
Kfeximum
Plant
Average
tragA)
0.49
0.33
0.46
0.12
1.2
1.4
1.4
0.58
0.38
20
CELL)
Average of
5 Plants
(mg/1)
< 0.28
0.14
0.11
0.075
0.41
0.40
0.35
0.15
0.17
4.4

Number Plants out
of Five
Exceeding
Treatability
Level
__C2)
3
3
_J2)
5
3
2
2
__(2)
4
(1)  Literature-based txeatabJJLity estimates from Section  8.1.  Table 8-11,
    given as the lower limit of txeatability  expressed as a 30-day average.

(2)  No data available  on treatability with sulfide/f ilter.
                                     198

-------
                      TABLE 11-16.  PBDBOSED UMIXKE210NS

                            Chlorine - Mercury Cell
                           Best Available Technology

                         Wbste Eater flows  2.1 -n -
SOBCATEGORY
CHLORINE MRBTTTRY fTT.T.
Subcategory
Performance „,
Pollutant fog/1} VER* ;
Bfcaiconventional
Total Besidual
Chlorine(6)
Concentration Basis Effluent Limit
30-day 24-hour SSday 24-hour
avg. max. avg. max.
Pollutants:
0.2
1.7
0.20
0.34
0.00042
0.00071
Toxic Ksllutants
Antimony ^
Arsenic^53
Cadmium^53
Chrcmium '"
Copper^ 5)
Lead (5)
Mercury ^
Nickel153
Silver^55
1tellium(5)
Zinc {55
0.23 (33
0.10(3>
o.oso{3>
0.040(33
o.oso<2)
0.16(3)
0.020(2)
0.10 <23
0.070(33
0.17 (3)
0.15(3)
2.2
2.2
2.2
' 2.2
2.2
2.2
2.2'
2.2
2.2
2.2
2.2
0.23
0.10
0.050
0.040
0.050
0.16
0.048
0.10
0.070
0.17
0,15
0.51
0.22
0.11
0.088
0.11
0.35
0.10
0.22
0.15
0.37
0.33
_ W)
0.00021
0.00011
_ (4)
0.00011
0.00034
0.00010
0.00021
0.00015
— (4)
0.00032
-W)
0.00046
0.00024
-(4)
0.00024
0.00075
0.00022
0.00046
0.00032
__(4)
0.00070
(1)  V5^,  t±e variability factor ratio,  is the ratio of the -variability factor
    for daily measurements to the variability factor for 30-day average.

(2)  Lower limit of treatability fior sulfide/filter technology according to
    literature treatability data  (Table 8-11}.

(3)  Average effluent concentration  fron verification sampling,

(4)  tfo load limits proposed;  concentration limits are provided for guidance
    purposes.

(5)  Limits are also applicable to PSES and PSNS and HSPS.

(6)  Limits are also applicable to NSPS.
                                  199

-------
     C hlor ine - Total residual chlorine limits are based on data
transfer  from the  utility  industry   (58)  and  the  detectable
concentration  of  chlorine  (0.2 mg/1)  because treatment should
remove  essentially  all  chlorine.    Thus  the  maximum  30-day
average concentration limit was set at 0.20 mg/1.

     The daily maximum limit for  total  residual chlorine was set
at 0.34 mg/1 based on an evaluation of  long-term  monitoring data
for total residual chlorine as presented  in Appendix A  (Table A-
la and c).  The ratio of  24-hour maximum variability factors to
30-day average variability factors for two plants was 1.7, thus:

     VFR = 2.28 = 1.7
           1.38

    (       0.20 mg/1    \  /	1.7	j\ = 0.34 mg/1
    \30~day averagelimit/  \24-hour maximum  limit/


     The determination  of load  limitations  for  total residual
chlorine (kg/kkg) was calculated based on  the  unit flow rate of
2.1 m3/kkg, thus:

     (0.20 mg/1)  (2.1 m3/kkg) (kg/m) = -0.00042  kg/kkg
                                        3000 mg/1

for the  30-day average limit.   The 24-hour  maximum  limit was
calculated similarly, i.e.,

     (0.34 mg/1)(2.1 m3/kkg)/_kg/m3_V 0.00071  kg/kkg
     Mercury  -  The  proposed  BAT  limitations  for  mercury,
although based on  the  same  technology,  are more stringent than
BPT  limitations.    Dechlorination  does  not  affect  mercury
removal.     The   Agency  considered  the   following  data  in
establishing the BAT limits of 0.00010 kg/kkg for a maximum 30-
day average.
    «
     o   Half of  the 'plants with monitoring  data presented in
         Table 11-13'are meeting the limits.

     o   Three of  five plants  were meeting  the  limits during
         sampling of their wastes.

     o   Three of  four plants  with long-term  monitoring, data
         presented in Appendix A are meeting the limits.

     The daily maximum limit of 0.00022 kg/kkg for mercury was
based on an evaluation of long term monitoring  data from four
                              200

-------
plants presented in Appendix A.  The average variability factor
ratio for the four plants was 2.2.  Thus:

     (0.00010 kg/kkg) (2.2) = 0.00022 kg/kkg

     The  concentration   limitations   for   mercury  were  then
calculated based on the unit flow rate of 2.1 m3/kkg.  That is:


     (0.00010 kg/kkg) f WOO mg/lN  (2.1 m3/kkg) = 0.048 mg/1
                          —-   j
                       V

     and (2. 2) (0.048 mg/1) » 0.11 mg/1

respectively for the maximum 30-day average and 24-hour maximum.

     Additional  Toxic Pollutants  -  The effluent  limitations
proposed  for  the  selected  additional  toxic pollutants  were
derived  from two  sources of  information?   sampling  data  and
literature-based treatability  estimates.   Dechlorination does
not affect toxic metals removal.

     The  results of  analysis  of  treated  effluent  represents
plant performance observed during  three  days  of  sampling.   The
effluent  data  for  toxic  pollutants  found  above  treatable
concentations in raw wastes  are summarized in Table 11-17.  Data
are  presented  from  four  plants  practicing  BPT  technology
(sulf ide precipitation followed  by filtration) .   Sampling data
for  the  fifth  plant,  £299,  reflect effluent quality  prior  to
filtration.

     It  is  apparent  from  the   sampling   data   that  the  BAT
technology  systems  are  generally  achieving  higher  quality
effluents than treatability literature indicates.  This could be
a  reflection of  low  influent   concentrations  and  incidental
removal  of  metals,  which  indicate  that  applying  effluent
limitations  to  a  dominant  metal  pollutant  (mercury)  assures
effective control of other metals.

     The concentration bases  for  the  proposed  limitations are
derived  from average  effluent  sampling unless  the  observed
concentration was below  the  literature treatability level.  In
such cases the lowest  applicable treatability level  from Table
8-11 was used.   Because long-term monitoring data from mercury
cell  chlorine  plant  effluents  was  not available  for  these
metals,  the  variability  factor  ratio established  for mercury
limits (2.2) was also applied to these metals.  The VFR used in
the existing regulations  (2) agrees witli this.

     A.  Arsenic:   Because  the   sampling data  from  five plants
(Table   11-17)   indicated   an   achievable . average   arsenic
                              201

-------
              11-17.  EPELUMT CDNCENTRaTIONS OF TOXIC POIIOTMiTS
                      FEDM ITEKEFICmTION SAMPLING

SUBCA3DGORY
Pollutant
Jtofcunony
Arsenic
Caflmiuni
Geranium
Copper
Lead
Mercury
Nickel
Silver
Thallium
Zinc
CHLORINE (MERCURY
CHuL)

Plant Effluent Concentrations
(mg/1)
Plant
§299
0.15
' 0.063
0.073
<0.06
0.038
<0.050
0.029
<0.050
<0.015
0.20
0.100
#747
<0.2-5

-------
concentration of  less  than 0.096 mg/1,  this value, rounded to
0.10 mg/1, is supported by  the  estimated range of  treatability
from Table 8-11 and was selected in the concentration basis for
the proposed maximum 30-day average  limitation.   This effluent
limitation is:

     {0.10 mg/1) (2.1 m3/kkg)/_kg/m3  \ = 0.00021 kg/kkg
/  kg/m3  \
\1000 mg/1/
and the proposed daily maximum arsenic limitation is obtained by
applying the VFR value of 2.2 that is:

     (2.2)(0.00021 kg/kkg) = 0.00046 kg/kkg

     B,  Cadmium:     For   cadmium,   the  plant  sampling  data
indicated an achievable average concentration of  less than 0.050
mg/1 in  the sulfide/filter treated  effluent.   This falls well
within the  range of published  treatability values (Table 8-11)
and was used as the concentration basis for the  proposed 30-day
average effluent limitations.  Thus:

     (0.050 mg/1)(2.1 m3/kkg) /  kg/m3  \ = 0.00011 kg/kkg
                              VlQOO mg/1/

for the maximum  30-day average and  using the VFR value of 2.2,
the proposed daily maximum is:

     (2.2)(0.00011 kg/kkg) = 0.00024 kg/kkg

     C.  Copper:   In the case of  copper,  the average plant
performance derived from  sampling data (Table  11-17)  showed an
effluent concentration of less than 0.033 mg/1  which is slightly
below  the  accepted  lower  limit  of  treatability  based  on
literature data.  The latter  is approximately 0.050 mg/1  and was
selected as the concentration basis  for the proposed maximum 30-
day average limitation on copper.  Thus:

     (0.050 mg/1)(2.1 m3/kkg) /  kg/m3  N =  0.00011 kg/kkg
                              \iO00 mg/1/

and  the  proposed   daily maximum   limitation   is  obtained  by
applying the VFR value of 2.2,  that  is:

     (2.2)(0.00011 kg/kkg) = 0.00024 kg/kkg

     D,  Lead:   The  proposed maximum  30-day average limitation
for lead  is based  on sampling  data  shown  in  Table  11-17 which
indicate and achievalbe effluent concentration of less than 0.16
mg/1.  Thus:
                              203

-------
     (0.16 mg/1) (2.1 m3/kkg) /  kg/m3  *\  =  0.00034  kg/kkg
                             \1000 mg/1/

and, applying  the  VPR value of 2.2, the proposed daily maximum
limitation is:

     (2.2) (0.00034 kg/kkg)  =  0.00075 kg/kkg


     E.  Nickel;   The average plant  effluent  concentration of
less  than 0.074 mg/1  of  nickel  is  slightly  less  than the
accepted  lower  limit of  treatability  (0.10   mg/1)  based  on
literature data.  This lower limit of 0.10 mg/1 was  Selected as
the concentration basis for the proposed maximum 30-day average
limitation for nickel.  Thus:

     (0.10 mg/1) (2.1 m3/kko) /  kg/m3  \  « 0.0021 kg/kkg
                             VLOOO rag/1/
and  the  proposed  daily  maximum  limitation  is  obtained  by
applying the VFR value of 2.2, that  is:

     (2.2) (0.00021 kg/kkg) »  0.00046 kg/kkg

     F.  Silver:    For   silver,   the  average   effluent  data
indicated  an achievable  concentration of  less than 0.067 mg/1.
This is within the range of published treatability values  (Table
8-11) and  is used as  the concentration basis for .the proposed
30-day average effluent  limitation.  Thus:

     (0.067 mg/1) (2.1 m3/kkg) Skg/m3\ =   0.00014 kg/kkg
  /  kg/m3 N
  VjLOOO mg/1/
and the proposed daily maximum limitation derived from the VFR
value is 2.2 is:

     (2.2)(0.00014 kg/kkg)  =  0.00031 kg/kkg

     G.  Zinc:  The average plant effluent for zinc is less than
0.15 mg/1.   This  is greater than   the  accepted lower limit of
treatability which   is approximately 0.02  mg/1.  The observed
performance  level  of 0.15  mg/1 is  used as  the  concentration
basis  for  the  proposed  maximum  30-day  average  limitation of
zinc.  Thus;

     (0.15 mg/1)(2.1 m3/kkg)f   kg/m3  \  =  0.00032  kg/kkg
/   kg/m3  \
\1000 mg/1 /
and  the  proposed  daily  maximum  limitation  is   obtained  by
applying the VFR value of 2.2, that is:
                             204

-------
     (2.2)(0.00032)  =  0.00070 kg/kkg

     H.  Antimony and Thallium:   The sampling data indicate an
average effluent concentration of less  than  0,23 mg/1 antimony
and  less  than  0.17  mg/1  thallium.   These relatively  high
concentrations are  the  result of analytical  difficulties  with
some samples  which  gave high "less  than", results.*   Because of
this  and  the   fact  that  no   data  are  available  for   the
treatability of antimony or thallium with sulfide/fliter, these
concentrations  are being  offered  as  30-day average  maximum
limitations for guidance purposes only.

     I.  Chromium:  The  sampling data indicate that plants are
achieving  effluent  concentrations   of  less   than  0./044  mg/1
chromium.  Because no data is available for the treatability of
chromium with sulfide/filter, this concentration is used as the
basis for the proposed maximum 30-day average  limitation.  Since
there is no treatability data, the  limitation is being offered
as guidance and no load limitations  (kg/kkg)  are presented.

11.7,3  Basis for Proposed BCT Effluent  Limitations

     The BCT  limitation  (applicable only  to  TSS)  was set equal
to BPT because the  treatment  technology for BAT  is  the same as
for  BPT  except  for  dechlorination.   Dechlorination  does  not
affect conventional pollutants.

11•7-4  Basis for Mew Source Performance Standards

     For NSPSr the Agency is proposing limitations equal to BPT
for TSS and BAT for other.pollutants because of the prohibitive
cost of additional technology.  Pollutants to  be limited are pHr
TSS, mercury,  arsenic,  cadmium,  copper,  lead, nickel, silver,
zinc and total residual chlorine.

11 -7 *5  Basis f_or Proposed Pr e t reataaent  Standards

     For pretreatment standards  for new  and  existing sources,
the  Agency  is  proposing  limitations based  on  BAT  technology
excluding dechlorination.   Dechlorination is unnecessary  for
discharges  to   POTWs   because   chlorination  of  influent  to
treatment works  is  common.   Pollutants to be limited  are pH,
mercury,  arsenic,  cadmium,  copper,  lead,  nickel,  silver  and
zinc.
                              205

-------
11.8  DIAPHRAGM CSLL PROCESS INDUSTRY PROFILE


11.8.1  General Description

     Approximately 65 percent of the U.S.  production of chlorine
is by diaphragm  cell plants.   Of 40 known plants, 308 data are
available for 19.  Table 11-18 presents a  summary profile  of the
subcategory.    Table  11-19  presents   the  current  status  of
discharge regulations for diaphragm  cell  chlorine plants.

11.8.2  General Process Description

Brine System

     As  in  the  mercury  cell  process,   the   sodium  chloride
solution  (brine  or salt  dissolved in water)  is purified  before
it  is  sent to the electrolytic  cells.   Precipitation of major
impurities with  sodium  carbonate and sodium hydroxide followed
by clarification generates  a brine mud  waste which  is then sent
to a lagoon or filtered.   The  settled brine is  saturated further
by  the  addition of  salt from caustic  evaporators  and  then is
sent to the cells.

     The  fundamental  difference  between  diaphragm  and mercury
cell  brine  systems  is  that  unconverted  sodium  chloride  in
diaphragm  cell  processes is carried with the  sodium hydroxide
(caustic)  from  the  cell and  is then  removed  as  a  solid  in
caustic  evaporators.   In mercury cells the  unconverted  sodium
chloride  is discharged  as a spent brine from  the cell  and
recycled directly through the brine  system.

Diaphragm Cell

     The treated brine solution is electrolyzed in the diaphragm
cell to form chlorine, hydrogen,  and sodium hydroxide according
to the reaction:

     2NaCl + 2H20 = C12 + 2NaOH + H2

The  diaphragm   cell  contains   a  porous  asbestos  diaphragm
separating the anode from the  cathode.  Chlorine is  liberated at
the anode and hydrogen and  hydroxyl  ions  (caustic) are produced
at the cathode.  In the past, the predominant  material used for
anodes was  graphite with  lead used to  provide  an electrical
contact and support.  The lead was joined to the graphite anode
by  an  organic  binder.    In  recent years, many graphite  anodes
have been  replaced  by  stabilized metal anodes made of titanium
with  a  platinum or  ruthenium  oxide   coating.    (An  industry
association, estimate  is  that  approximately 49 percent  of U.S.
diaphragm  cell  capacity still  involves  graphite  anodes.)  The
                              206

-------
         2W3IE 11-18. SUBCATEGORY PROFILE DMCA SUMMftRY
                     CHLORENE  (DIAPHR&Q4 CELL)
Total subcategory capacity rate
Total subcategory production rate
Number of plants In this subcategory
308 Data on file for
    With total capacity of
    With total production of
    Representing capacity
    Representing production
    Plant production ranges
            Minimum
            Maximum
    Average production
                                    k
    Median production
    Average capacity utilization
    Plant age range:
            Minimum
            Maximum
    Waste •water flew range:
            Minimum
            Maximum
    Volume per unit products
            Minimum
            Maximum
8,272,600 kkg/year
6,427,000 kkg/year
       40
       19
6,397,000 kkg/year
4,200,000 kkg/year
       77 percent
       66 percent

   14,700 kkg/year
1,500,000 kkg/year
  221,000 kkg/year
  103,000 kkg/year
       67 percent

        4 years
       74 years

    1,100 cubic meters/day
    7,100 cubic meters/day

        1 cubic meters/kkg
       23 cubic inebers/kkg
Sources of data are Stanford Research Institute, Directory of CJhemical
Producers, U.S.A., 1977, U.S. Department of Commerce,  Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.;  Draft
Report, "Preliminary Economic Assessment of Effluent limitations  in the
Inorganic Chemical Industry*" June, 1978 and "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards for the Inorganic Chemicals Industry,"
March, 1980.
                                     207

-------
      11-19.
    STATUS OF FEGULffiTICNS  -  EFFLUENT LIMITATION GUIDELINES

GHODKENE
SUBP&KE P (40
(DIAPHRAGM
CFR 415.60
CKLTi)
, 3/12/74)


STANDARDS
Product
Process
BPCTC&
Max.1 Avg.2
Para- kg/kkg kg/kkg
meters (mg/1) (rag/1)
BATEA
Max. Avg.
kg/kkg kg/kkg
'(mg/1) (mg/1)
NSPS
Max. Avg.
kg/Meg kg/kkg
(mg/L) (mg/1)
Diaphragm   TSS    0.64
tell
Process
            Pb     0.005
                   0.32      No discharge
                             of pwwp

                   0.0025    No discharge
                             of
0.64
0.32
0.00008  0.00004
 Section 415.63 was remanded and is presently reserved (41 FR 51601,
 November 23, 1976).
"wax.  =  Maximum of any one day.
 Avg.
imum. of daily values for thirty consecutive days.
       =  process waste water pollutant.
                                     208

-------
advantages of using metal anodes compared to graphite  anodes are
increased power efficiency of the cells,  longer  anode  life and a
reduction  in  potential  pollutant loads of lead and chlorinated
organics.

Product Purification

     As  with  mercury  cell plants,  chlorine liberated  at  the
anode must be cooled  and dried  to  remove  moisture  and other
impurities.  The cooling generates a chlorine condensate stream
which can  be  stripped  to  recover chlorine then returned to the
brine  system  or  discharged.    Drying  the  chlorine   gas  is
accomplished  by  scrubbing with  sulfuric acid.   The  resulting
diluted acid  can  subsequently be regenerated, sold or used for
pH control.  When the chlorine gas is compressed and  liquified,
noncondensible gases known as tail or sniff gases  remain.  These
are  usually scrubbed  with caustic soda  or lime  generating  a
hypochlorite  solution  which  can  be  sold,  used on-site  or
discharged, with or without decomposition or treatment.

     The sodium hydroxide or caustic from,the diaphragm  cell has
a concentration of  about  14  percent  NaOH and a sodium chloride
content as high as  17 percent.  The caustic  is usually filtered
to  remove some  of  the  impurities and  then evaporated  to 50
percent NaOH  by  multiple  effect  evaporators.   Sodium chloride
remains  as a solid salt  which  is then  returned  to  the brine
system.   Further  purification of  the  caustic is necessary for
some applications  (such as rayon production) and extraction or
adsorption techniques have been used to  remove small  amounts of
impurities.   The  caustic can  be evaporated  further  if  more
concentrated products are  required.  The vapor evolved from the
last of multiple  effect evaporators  is condensed in  barometric
condensers  generating  contact  cooling  water,  or in   surface
condensers using noncontact cooling water.

     The hydrogen gas generated  in the process can be vented or
cooled by refrigeration to remove water vapor before sale or use
as a fuel.

     Figure 11-14 is a general flow diagram  for the manufacture
of chlorine by the  diaphragm cell process.


11.9  WATER OSS AND WASTE WATER SOURCES


11.9.1  Water Use

     Water  use  at diaphragm cell  plants  is similar   to that at
mercury  cell  plants with  one exception.   Common uses  include
noncontact  cooling, tail  gas scrubbers,  cell  wash,  equipment
                              209

-------
                  LI HE

i
SU1FATB gmi,
PURGE

HO
I
t t


COOMHS
TOHBR"
to
l_l *
O I
SLOWDOWN
TO WASTE
WATER
t
LEAKS,
SPILLS
WASIIDOWH
ETC.
t
TO WASTE
I DR1KB

HCOOTACT •
COOLING -W
WATER ^
11*
ECTC1B



f

PtMIFieWMOH
SKSTEH
t
DIAFHRAGH
CELL
1
ro 121 soDim
HYDROXIDB
SOtUTIOH
t
EWkPORMOR
ran-rur- ijjm.t 	 	 ^.*m wyTTR
DRIIW BUUS »• ^gnjuj HOHCOWACT COOMMG HATBB
t 4
COOIBB OR usg
WATER
— BAROHBIRIC
^ COHDEHSER

t
COOMNG
HAfEE
,_£*
COOLER I 	 ^ CHUJRlHATBn HATER COHDBNSATB
* f *
SALT
REMOVAL
!
SODMIH
MSDROXIDE
SOLUTION
1
FIWER
SUIHJMC-^
«nn


f •MMB) f 4

— — «""* iwwijK iX)!)lUH
r*"W4PMKKK
-------
maintenance,  floor  washings  and  filter  backwashing.    The
exception  at diaphragm  cell plants  is the  use of  wate'r for
barometric condensers in the evaporation of caustic.

11.9.2  Waste Sources

Brine Mud

     As with mercury cells, this  is  the waste produced during
purification of brine before it  is  introduced  into the cells for
electrolysis.    It  consists of  precipitated  hydroxides  and
carbonates of calcium, magnesium,  iron,  and  other  metals.   The
mud can be a major source of solid  waste depending on, the purity
of the raw salt  used.  At diaphragm cell plants brine muds are
filtered or  settled  in  lagoons.   The solids are landfilled and
the filtrate or overflow is discharged or recycled to the brine
system.

     Brine mud  is the major source  of  solid waste at chlorine
plants, and  discharges range from  0.04  to 1.5 cubic meters per
metric ton   (m3/kkg), with  a solids content of  from  two to 20
percent.

Cell Room Wastes

     These wastes include leaks, spills, area  washdown and cell
wash waters.   At  diaphragm cell  plants cell wash  waters are
heavily laden with  asbestos  and  are therefore  settled and/or
filtered  before  chemical  treatment or  discharge.    At plants
using graphite  anodes  in the cells,  the cell room wastes also
contain lead.  Data from diaphragm cell  plants indicate a waste
flow from 0.02 to 1.2 m3/kkg from  cell  room operations.

Chlorine Cooling Condensate

     Condensation  from  the indirect  cooling of  cell  gas  is
contaminated with chlorine.  The chlorine is removed  (stripped)
or  recovered  from  the  stream  before  discharge  or  recycle.
Condensate  flows  from  three  plants range  from  0.16  to 0.9
m3/kkg.

Spent Sulfuric Acici

     Concentrated  sulfuric acid  is used  to  dry  chlorine gas
after  the  first stage of  cooling.   Once  diluted  to  50  to 70
percent, the spent acid can be regenerated, sold or used for pH
control.
                               211

-------
Tail Gas Scrubber Liquid

     The uncondensed chlorine gas from the liquefaction stage is
scrubbed  with  sodium  or   calcium  hydroxide  producing  the
corresponding  hypochlorite.   The  hypochlorite can  be  used in
other processes, sold, decomposed,  or  discharged.  The amount of
tail  gas scrubber  water  generated  at  diaphragm  cell  plants
ranges from 0.1 to 0.29 m3/kkg.

Filter Backwashes

     Backwashing  of  filters used  to  treat brine before  it is
sent to  the  cells at one  graphite anode  diaphragm  cell  plant
generated a  waste water flow  of 0.45 m3/kkg.   Backwashing of
filters  used  to  clarify  caustic   product at  the   same  plant
resulted  in  an average flow of  5,4 m3/kkg.   At some diaphragm
cell  plants  these  waste   waters   are  partially  recycled  to
process.

     The relatively high flow of caustic filter backwash is due
to  the  need  to  remove sodium  sulfate,  an  impurity  in  the
caustic.  Sulfate ions, if  allowed to  accumulate  in the brine
system  at  graphite  anode   plants  will   interfere  with  cell
performance.

Hydrogen Condensate

     Cooling  of  hydrogen  gas  for  use  or   sale   produces  a
condensate stream which can  be  discharged.  Although  no data are
available on the volume of this flow, it is small.

Barometric Condenser Waste Water

     When vapors from caustic evaporators  are contact-cooled, a
significant amount of  waste water  can be  generated.   Plows of
from 90  to  300 m3/kkg  have been  reported at facilities  where
barometric   condenser  water   is   "once   through"  and   not
recirculated.    Recirculation   of   barometric condenser  water
requires a cooling step and  a  blowdown  discharge.   A number of
facilities are accomplishing this with a corresponding reduction
in  water  use.    The  necessary   blowdown  of  recirculating
barometric condenser  waste  water  at  two  plants ranges  from a
flow of 0.82 m3/kkg to Ol89 m3/kkg.

Summary of Waste Water Flows

     Table 11-20 summarizes unit waste water flow data available
by specific sources.  A separate list of  flows at  one graphite
anode  plant  is  presented  to  compare  waste  water  generation
between metal anode and graphite anode plants.
                              212

-------
  TABIE 11-20.  WaSTE WATER FLOWS AT DIAPHRAGM CELL CHDDRINE PLANTS
Stream Description
              Flow (nP/kkg)
    Plants with              Plant with
   jfetal Anodes            Graphite  Anode

 min.  avg.  max.
     room wastes
   and cell wash
Chlorine Condensate


Spent Sulfuric Acid

Tail Gas Scrubber

Caustic Filter Wash

Brine Filter Backwash

Caustic Cooling Slowdown

Brine Mud
0.02   0.38  0.67


0.16   0.49  0.90


       0.01

0.10   0,17  0.29

        NA.



0.82   0.86  0.89

0.04   0.42  1.5
1.2


0.78


 NA

0.11

5.4

0.45

 NA.

 NA
NA:  Not Available
                                    213

-------
11.10  DESCRIPTIONS OF SPECIFIC PLANTS


     The following descriptions of plants includes those plants
that were sampled during the screening and verification program.
The discussion primarily covers plant practices  in waste water
control and  treatment.   Plants were selected for screening and
verification sampling  because they were representative  of the
industry  in  that they  included  a  wide  range  of  sizes  and
variation in process detail.

11.10.1  Screening

     At Plant  f014r  visited during the  screening program, the
chlorine condensate is stripped with steam to remove and recover
chlorine.   Brine precipitates  (muds)   are land  disposed, while
the spent sulfuric  acid  and scrubber  solutions  are used at an
adjacent plant.  The condensate from the  hydrogen cooler  is used
as makeup water  for  a  cooling tower system,  and  the condensate
from the evaporative concentration of sodium hydronide is used
to dissolve salt reclaimed from the concentration process.  The
cell washings  are sent to  a collection pond where asbestos and
other suspended solids are  removed. In Figure 11-15 the general
process flow sheet is presented.  The waste streams sampled and
their waste loadings are presented in Table 11-21.

11.10.2  Verification

     Four plants  were visited and their waste  streams sampled
during the verification program.  The  results of  analysis of the
waste waters are presented in Table 11-21.

     At Plant  f261,  the  cathode wash  water  is  passed  through a
filter and  the asbestos  drummed and disposed of  in an off-site
landfill,  while  the  filtrate  goes  to  the  sewer.    Brine
purification  muds  at this  facility  are   utilized  for  their
alkalinity on-site and then they are settled prior to discharge
of  the  supernatant.     Spent  sulfuric   acid   is  used  for
neutralization of  waste  waters.   Dechlorination  of the drying
acid by  reaction with sodium  bisulfite  is  planned in the near
future.   Figure 11-16  shows  the process  flow diagram  and
sampling points.

     Plant  |738  has  two production lines,  738A  and 738B, that
are almost  identical.  At  the new  plant  (738B)  the NaOH is not
concentrated nor  is  the waste  from the chlorine disposal system
scrubbed.   In  addition,  the inert gases from  the liquefaction
step are put through the chlorine disposal system.  The process
flow sheets are shown in Figures 11-17 and 11-18.
                              214

-------
ro
H
tn






BW3

•»«
1 1
BRINE MUD










VENT GAS
1

1,00000001 	 *"«2 •Bffl.GBSIg.
98% H^SO, ** SCMUMHiH
4, A0 i3^
T 
-------
TABLE 11-21.  POEiOTSNT CmCENTRftTIONS AND LOADS AT
                        AND VppIPICRTIOP PLANTS

stBCASEoany CELCRINE DIAPHRAGM CEO
Plant &
Stream _ Stream
No. Description (mg/1)
#014
3
4
5
6
§261
1
2
3
4
5
#738A1
2
3
4
5

I738B6
7
8
9
10

11
12
13

14

0-2 condensate
Cell.wash
Brine mud
Bar. condenser

Brine mud
Cell wash
Asbestos filtrate
Filter cake
Bar. condenser
Cell room waste
Asbestos wash
Hypo scrubber
0-2 cooling water
Caustic cooling
tower
Cell room waste
Asbestos wash
Hypo scrubber
€3-2 cooling water
Caustic cooling
wtower
Chlorate sump
Plant effluent(B)
Final effluent
(Total)
Brine mud

.2
1600
NA
7

NA
4800
9
NA
6
27
57
290
35
48

95
72
160
20
4.7

32,
63
58

270
«r
TSS
(kg/kkg)

1.
2.




1.



1.
7.
2.
2.
4.

4.
8.
1.
1.
3.

7.
5.




4 x
4 x
NA
3.6

NA
8 x
NA
NA
NA
4 x
0 x
7 x
2 x
3 x

5 x
3 x
4 x
7 x
8 x

0 x
7 x
NA

NA

10
10




10



10
10
10
10
10

10
10
10

-3
—2




-1



-3
-3
-2
-1
-2

-3
-3
-2
10-2
10

10
10



-3

-3
—1



Lead
(mg/1) (kg/kkg)

0.
0.
0.
0.

0.
2.
0.
42
< 0.
0.
0.
0.
0.
0.

0.
0.
0.
0.

0055
26
72
005

36
0
075

010
077
031
18
28
51

067
13
20
20
< 0.010

< 0.
0.
0.

0.

010
12
078

10

5.
3.
1.

0x10
9x10
3x10
1.5x10

3.
7.



3.
3.
1.


-6
-6
-5
-3

OxlO"4
6x10
NA
NA
NA
9x10
8X10
7x10
1.3x10
4.

3.
1.
1.
1.
< 8.

< 2.
1.



5x10

2x10
5x10
7x10
7x10
2x10

3x10
1K10
NA

NA
-5



-6
-6
~5
-4
-4

-6
-5
-5
-5
—6

-6
-3



                                                 (Continued)
                             216

-------
                               TABLE 11-21 (continual)

Plant & Stream
Stream No. Description
#736
1
2
3
4
5
6
7
#9671
2
3
4
5
6
7

Cell wash
Cell room drain
Brine mud
50% Bar, condenser
70% Bar. condenser
95% Bar. condenser
* Chlorine condensate
Cell bldg wastes
Lead pond effluent
Caustic backwash
Brine backwash
Cell wash
Condensate and 112804
Scrubber waste
TSS
(mg/1) (kg/kkg)

934
283.5
20,000
32
21
90.33
2.4
1000
54
« 160
13,000
310
1100
270

6.0 x 10~2
4.6 x 10~3
33
NA
NA
NA
3.9 x 10""4
1.8 x 10"1
3.0 x 10"2
8.6 x 10-1
5.8
5.6 x 10~2
8.7 x 10-1
1.2 x 10~2
Le
(rag/1)

0.014
0.17
0.019
0.010
0.010
0.010
0.010
680
29
0.32
0.52
48
0.92
0.67
ad
(kg/kkg)

9,lxlO~6
2.8xlO~6
S.lxlO""5
NA
NA
NA
1.6xlO"6
1.2X10"1
1.6xlO"2
1.7xlO~3
2.3xlO~4
8.6xlO~3
7.3xlO~4
2.9xlO~5
NA: Not Available
                                     217

-------
SODIUM
CAR90KATE H»OH SALT
1 i '
HAW ... Hirrn<: ^ CLARI- ^ SAND KlS», grrriiUTfit -^^X—
IMKI 	 "*" "' " FLOCCULATORS ^ FILT£RS ~^j^ SATURATOR -J^—
I BACK
4*^f| ,rWASH
£ -{^ BRINE RECYCLED
f SETTLER TO PROCESS
MINE HUBS 1 	
ADJACENT PLANT w
HjSO^ SLUDGE TO LANDFILL
f COOLING WATER
i — I — j C\
-, 	 - NfX
CI2 ff
1 ' S
" 	 	 — -— • UQ f(jH
__ CELL
*1 ROOK

GD'
CELL
WASH ^~"
) 	 ^.HYDROGEN TO
^
^COOLING WATER
BEHSATE
n

•"""" 	 """** FIL'IER
STEAM
1 	 1»> 	 •»
MULTIPLE EFFECT
EVAPORATOR
V CI2 CONDENSATE CAUSTIC j
SPENT HjSOfc UATER Tfl AOjACEHT f (/-
tO PLANT FOR USE fc
^ UVPOCHLORIT
TOWER

"*• PURIFICATION 	 PURIFICATION 	 », LIOUrFlc.
^ BOTTOM TOWER *" TOWER ^ LIQUEFIER _ ^

_ 1 ' f Cl, STORAGE TANK
^ ' "" "' "~ 	 " 	 	 IHLUHINk AND RAILROAO CAR
, , TO STORAGE WASHOOWN
PURGE FOR DISPOSAL
BY .CONTRACT
LE6EMP
^•^ SAHPLIN6 POIHTS
E -*-T°tfMT COOL.C
UATER
HVPOCHLORITE STEAM
SOLUTION TO 1
ADJACENT PLANT V 1
m e«
EVAPOF
£
(
U5TIC i,,,,™ 	 	 	 £
ATOR
NaOH
BOILER
«
T l^\ ^ FILTRATE TO
'" '\±l ^"PROCESS SEWER
/•-^ ASBESTOS TO
	 Ij^-r-"**' LANDFILL
|li ^ TO
| ^"EJECTOR
BAROMETRIC COOLING
CONOENSEII wut
TO BOILER
<^N _ I FEED WAT£R
^K —tot T0 5L*K£ LIH£
* ^"(rOR LIHI PLANT}
	 ^-PROCESS SEWER

„ . c rf_ COOLING
' WATER
^^ PROCESS
^ SEWER
Figure 11-16.   General  process flow diagram at Plant #261 showing the sampling points,
               Chlorine/Caustic (Diaphragm Cell) manufacture

-------
tvJ
                 BRINK
          73% NaOH  •<
                        LEEESP

                  •« 8aif>Ung points.
(TYiint {12 - waste aim> (ccnfaination o£ all wastes) . )
        Figure 11-17.  General process flowsheet  at plant I738-A- stowing the sampling points.
                         Chlorine/caustic  (diaphragm cell)  manufacture.

-------
                                                                         VEW
             BRINE
to
w
o
              Figure 11-18.  General process flow diagram at plant #738-B showing the sampling points.

                             Oilorine/caustic  (diaphragm cell)  manufacture.

-------
     Plant §736  has  installed demisters  to  control the vapors
evolved  from  the  last  stage  of  the  evaporator  during  the
concentration  of  caustic soda.   In this  treatment,  the steam
evolved from  the  concentration of  cell  liquors  passes through
metal wool filters  to reduce entrained  solids.   The cell room
washings are  sent to  a settling chamber and the settled asbestos
is sent  to a landfill.  The  other  waste waters, consisting of
caustic  evaporator  washings  and  wastes  from  salt  separation,
brine purification operations,  and  caustic filtration backwash
waters,  are  combined  and  sent  to  one  of two  settling ponds.
Skimming  devices  on  the  settling  ponds  remove any  oil  that
separates, while the  settled solids  in  the ponds  are dredged and
disposed of in an abandoned brine well.  Figure  11-19 shows the
process flow diagram and sampling points.

     Plant 1967  uses  graphite  anodes  in  its  diaphragm cells.
The cell  washings  at  this  plant  are  sent to  an asbestos  pond
that has a continuous cover of water.   Periodically, the settled
solids  are  removed,   sealed  in  drums  and  disposed  of  in  a
landfill.  The overflow  from the  pond  is treated with soda ash
to precipitate lead,  and then filtered.   Sulfuric acid is used
to bring  the  pH down  to  the 6 to  9 range.  Figure 11-20  is a
general process flow diagram for Plant |967.

11.10.3  Descriptions of Plants Hot  Sampled

     At  Plant  1999   brine  mud  and other   streams  with  high
suspended solids are  collected  and  filtered with leaf filters.
The cake is disposed of in a landfill and the filtrate returned
to the brine system.

     At Plant |326, waste water from the diaphragm cell process
is combined with other process waste waters.  The combined waste
water is  sent  to two settling tanks in  series.   In one of the
settling tanks, skimmers have been  installed to  remove oil and
the overflow from the  second is filtered before  discharge.

     At Plant 1589, the brine mud from the clarifier underflow
is sent  to a  brine mud settling  pond.   The  overflow,  which is
mostly  brine, is  returned  to the process.    The cell  room
washings  are  sent  to  a  settling  pit  and  the  settled asbestos
fibers are removed by the use of a vacuum truck,  and disposed of
in a  landfill.   The  chlorine from  the cells is contact-cooled
with the tail gas scrubber water.   The resulting waste water is
steam stripped for chlorine recovery before discharge.

     At  Plant  1741,   chlorine,   caustic soda,  and  potassium
hydroxide are  produced using both mercury and  diaphragm cells.
Mercury-bearing effluent at this facility is treated by sulfide
precipitation.   Tail gas  absorption  wastes   are  treated  by
catalytic decomposition by a process which consists of scrubbing
                               221

-------
                                                                                                 MH08.
N)
to
                                          cait/ct ASBESTOS
                                          wrex  10 EOUD
                                         TORIVEH W3IE
                                               DISPOSAL
            Figure  11-19.  General process  floi'j diagram at Plant #736  showing the sanpling points.
                            CMorine/Caustic (Diaphragm  Cell)  manufacture

-------
                                                                                  TO PChtR HOUSE
NJ
          Figure 11-20.  Ger^ral process flow diagram at Plant 1967 showing the saropling points.
                         CMorine/Caustic  (Diaphragm Cell) manufacture

-------
with   caustic  soda   solution   and   treating   the  resulting
hydrochlorite  solution with nickel chloride and iron chloride.
Consumption of  iron  and nickel  chloride is approximately equal
and  consists  of  0.01  kilogram  per  metric  ton  of  chlorine
produced.    The  catalytic  decomposition  proceeds relatively
slowly,  and  wastes  are  retained  in  the  treatment  tanks  for
approximately three days,  after  which  time  no residual chlorine
is reported to  be present  (3).

11.10.4  Toxic Pollutant Concent r at ions

Analytical Data Base

     Section  5.1.2 of  this report describes the methodology of
the  screening  and  verification  sampling  program.    In  the
chlorine diaphragm cell industry, a total ,of 15  days of sampling
were  conducted at Plants  #014,  #261,  f738,  f967,  and  #736.
Thirty-seven  different sampling points  were  involved covering
various  raw waste streams  and  the treated effluents  at these
plants.   The  evaluation of  the toxic metal content  of these
process related waste  streams was  based on 975 analytical data
points.   The sampling  for toxic  organic  pollutants  at Plants
1014 and |967  generated 2300 analytical data points.   Analysis
of waste for  asbestos  generated -an additional 13 data points,

Asbestos

     Asbestos,  used as a diaphragm separating the cell anode and
cathode,  is  the  major toxic  pollutant consistently  found  in
process  waste  water  from  diaphragm   cell  plants.   It occurs
primarily in wastes resulting from activities such  as cell room
washdown and  cell repair and  cleaning.

     Table 11-22 presents the results of  asbestos determinations
of supply water and waste waters at three diaphragm, cell plants.
Results are expressed as total fibers per liter  (in millions) as
well as crisotile and  amphibole  fibers per  liter.

There   ar.e   no  standardized  analytical   techniques   and  no
definition  of  asbestos in water.   Because of  this, EPA  is
excluding   limitations  for   asbestos  from   these   proposed
regulations and deferring  regulation to  a later  date.

Toxic Metals

     Table  11-23 presents maximum  daily  concentrations of toxic
metals  found in  raw  waste samples during the  screening  and
verification  of  diaphragm  cell  chlorine plants.    Maximum
concentrations  observed   at one   graphite  anode  plant  are
presented separately.   It is clear that except for lead, toxic
metals   concentrations  at   the  graphite  anode  plant  are
                                224

-------
 -TABLE 11-22.  RESULTS OF ASBESTOS SAMPLING AT DIAPHRAGM CELL PLANTS
Plant
#261

#736



#967


Stream
Supply
Cell Wash
Filtered Discharge
Barometric
Condenser
Supply
Cell Wash
Cell Room Waste
Barometric
Condenser
Barometric
Condenser
Barometric
Condenser
Supply
Cell Ifeste
Pond Effluent
Caustic Wash
Brine Filter
Backwash
Cathode Wash Waste
Condensate & Spent
Acid
Neutralizer Waste
Total Asbestos
Fibers (MFL)*
8.0
2.1 X 108
1.6 X 103
0.4
• 0.7
2.0 X 107
2.9 X 102
1.8
5.3
1.4 X 102
9.7 X 102
2.4 X 104
2.4 X 103
7.8 X 103
8.0 X 102
3.2 X 105
2.7 X 102
2.1 X 103
Chrisotile
MFL
7.5
2.1 X 108
1.6 X 103
0.4
0.7
2.0 X 107
2.8 X 102
0
5.3
1.4 X 102
9.7 X 102
2.4 X 104
2.4 X 103
7.8 X 103
6.2 X 102
3.2 X 105
1.8 X 102
2.1 X 103
Amphibole
MFL
0.4
0
0
0
0
0
8
1.8
0
0
0
8 X 102
0
0
1.8 X 102
0
8.9 X 10
0
*Million fibers per liter
                                   225

-------
TABLE 11-23.          R?W                      OF TOXIC                 AT
              DIAPHR&Qd CELT. CHLORINE PLSNTS (mg/1)

SUBCMEGOBX'
Toxic
Metal
jtotimony
Arsenic
Beryllium
Cadmium
Qiromium
Copper
Leac1
Mercury
Nickel
Selenium
Silver
OSiallium
Zinc
CHLORINE DTKPHBMM CELL
Plants with
Metal Jtaodes
<0.25
0.17
<0.014
0.037
7.4
17
2.0
<0.003
22
<0.020
0.018
<0.25
3.0

Plant with
Graphite Anode
<0.065
0.59
<0.001
0.017
<0.048
0.27
44
0.004
0.070
O.Q30
<0.016
<0.050
0.25
                                    226

-------
essentially no higher than at  the metal  anode plants.   Because
several waste streams usually contribute to the total raw waste
at  chlorine  plants,  a  calculation  was  often  necessary  to
determine the pollutant  concentrations that would  exist when the
streams  were  mixed prior  to  treatment.    An example  of  this
calculation  is  the  "mixing"   of   the  following  hypothetical
streams:

     Stream A:  100 gallons per minute, 15 mg/1 of pollutant

     Stream B:  10 gallons per minute, 60 mg/1 of pollutant

     The weighted average for the mixed streams is given by:

     Concentration of mixed stream =

     (Flow A x Concentration A) +(Flow B x Concentration B)
                       (Plow A + Flow B)

     Substituting numerical values gives:

     (100 gpm) (15 mg/1)  + (10 gpm)  (60 mg/1) « 19 mg/1
                         110 gpm

     This method was used to calculate raw waste concentrations
of pollutants as  presented  in Table  11-23.  Barometric condenser
waste water when "once through" was not included because of the
high dilution effect of  these large  flows.  Brine mud flows were
also not included.

     The daily raw waste loads were  calculated  from the waste
stream flow rates measured or estimated at the time of sampling
and the measured pollutant concentration.

     That is,

     Daily loading  (as kg of pollutant per day) = (C) (Q)
                                                    1000

     Where:

     C  is  the  concentration  of the  pollutant expressed in unit
of mg/1  (Note:  kg/m3 = 1000 mg/1)  and

     Q  is  the waste  stream flow  rate  expressed in units of
m3/day  (m3, a cubic meter, is equal to 264.2 U.S. gallons).

     Similarly,  the  unit  loadings  were  calculated from  the
reported chlorine production rate,  the waste stream flow rate,
and the measured pollutant concentration.
                              221

-------
     Unit loading  (as kg of pollutant per kkg    (C)  fQ)
     of chlorine                              = 1000 P

     Where C and Q are as described above, and P  is the chlorine
production rate expressed in units of kkg/day  (kkg  is 1000 kg, a
metric ton, which  is equal to 2205 Ibs).

     The minimum,  average and  maximum values were  calculated
based on data  from those plants where the particular pollutant
was found at a detectable concentration.

     In Table  11-24,  the toxic  pollutant  raw waste  data  are
presented as the average daily concentrations  (based  on three
24-hour samples) and  the unit loadings found at the individual
plants.  Beryllium, selenium, and thallium  are  not included in
the table because  average concentrations were below detectable
limits.

     In'Table  11-25 plant average daily  and  unit loadings are
presented as minimum, average, and maximum values based on data
presented  in  Table 11-24  for metal  anode  plants only.    (The
graphite  anode  plant  is considered  separately  due  to  its
particular waste source characteristics.)

     Based  on  the  average  waste loads  generated  per  unit of
product at metal anode plants and one graphite anode plant, and
the estimated total subcategory production, the estimated total
pollutant   raw   waste  loads  generated  each   year   by  this
subcategory are as follows:


                         Raw  Waste Load

              Pollutant                    kg/year

              Antimony                         483
              Arsenic                        6,300
              Cadmium                           41
              Chromium                       3,100
              Copper                         4,400
              Lead                         470,000
              Mercury                           48
              Nickel                         3,600
              Silver                             5
              Zinc                           5,100

     Because cell  room wastes  including cell or  cathode wash
wastes,  leaks,  spills   and  washdown   are   usually   treated
separately  at  diaphragm  cell plants and  because other process
wastes  such  as filter   backwashes,  condensates  and  caustic
evaporation wastes are usually discharged  after  the settling,
                              228

-------
      11-24.
TOXIC                      AND LOADS AT           AND
VERIFICATION PLANTS
                     (mg/1)
(kg/kkg)

SlBCMEGOKf
Pollutant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Silver
Zinc
CHLORINE DIAPHRAGM nKT,T,
014
*
*
0.002
0.0000018
0.019
0.000017
0.015
0.000014
0.006
0.0000045
' 0.002
0.0000018
0.90
0.00081
*
*
Plant!
261 738A
*
0.17
0.0000064
0.037
0.0000014
1.9
0.000071
17
0.00064
2.0
0.000075
*
22
0.00081
0.018
0.0000007
'1.5
0.000054
*
*
*
*
0.52
0.0046
0.045
0.00039
0.082
0.00060
*
0.21
0.0018
*
0.29
0.0021
738B
*
0.011
0.000021
*
0.066
0.0012
0.12
0.00023
0.11
0.000021
*
0.067
0.00013
*
0.093
0.00018
736
0.010
0.0000033
0.057
0.000014
0.025
0.0000061
0.18
0.000044
0.43
0.00011
0.016
0.0000039
0.003
0.0000007
0.22
0.000054
*
3.0
0.00074
967**
0.011
0.00015
0.30
0.0021
*
0.004
0.000032
0.16
0.0011
21
0.015
0.002
0.000014
0.068
0.00049
*
0.19
0.0014
*  Below measurable concentrations
** Graphite Anode plant
                                     229

-------
11-25.
                OP
                          TOSK LQMXDKS KT SCREENING ABD VBRIFICKHCK MEEM. MJQEE EMNTS

SOBaflEQQfS
Pollutant
£* Antimony
o
Arsenic
Cadmium
Chromium
Copper
lead
Jfercury
Hickel
Silver
Zinc


min.
0.00077
O.OOlfl
0.00041
0.0042
0.0035
0.00090
0.00016
0.0066
0.00021
0.016

Loading
(kg/kkg)
avg.
0.00077
0.0084
0.00076
0.59
0.12
0.094
0.00030
0.31
0.00021
2.1
CHJJ08INE DI
AHnasicEii
Unit Loading
(kg/kkg)
HBX,
0.00077
0.020
0.0014
2.8
0.19
0.37
0.00044
1 1.1
0.00021
8.0
min.
0.0000033
0.0000064
0.0000014
0.000017
0.000014
0.0000039
0.0000007
0.000010
0.0000017
0,000054
aag.
0.0000033
0.000017
0.0000032
0.00096
0.00020
0.00016
0.0000012
0.00057
0.0000007
0.00078
mix.
0.0000033
0.000030
o.oooooei
0.0046
0.00064
0.00060
0.0000018
0.0018
0.0000007
0.0021
*Nuntoer of
Plants
Averaged
(out of 5)
1
3
3
5
5
5
2
5
1
4
Cnly those plants where the pollutant was observed at measurable concentrations.

-------
these  two  waste mixes were  evaluated  separately.   Table 31-26
presents average  raw waste  concentrations  and loads  of toxic
metals  found in  cell room  wastes  at  the  six  diaphragm cell
plants sampled.  Table 11-27 presents the similar data from the
sampling of other process wastes at these plants.

Toxic Organic Pollutants

     The use  of graphite anodes at  chlorine  plants results in
the generation  of a variety of  simple chlorinated hydrocarbon
compounds  as  a  result of  the attack  of chlorine on the anodes.
These  compounds are  carried  out of the cell  with  the chlorine
and  find  their  way  into  the various  waste  streams  which
originate  from  the  chlorine cooling,  drying,  compression,  and
liquefaction  steps.

     Table 11-28 presents the toxic organics that were observed
in measurable concentrations in the  raw wastes  at Plant |967.
The concentrations presented in the  table were calculated as a
mixture of  all  raw waste streams  weighted  on a flow  basis as
previously described.

     Table 11-29  presents  the  concentrations  of toxic organics
by individual raw waste stream at Plant $967.  It is clear from
the table  that  the  highest concentrations of organics occur in
wastes from chlorine treatment  (condensate, drying acid and tail
gas scrubber water)  and they account  for 83 percent  of the total
organic waste load.


11.11  POLLOTION ABATEMENT OPTIONS


11.11.1  Toxic Pollutants of Concern

     Lead  occurs  in  high  concentrations in the cell room waste
waters of  chlorine  plants  using graphite anodes.   Other toxic
metals  often found  in ,significant concentration  at diaphragm
cell plants include arsenic, cadmium, chromium, copper, nickel,
and zinc.  Antimony, mercury, and silver were also detected but
at concentrations that are not treatable.   These metals are not
considered  further.    The  sources  of  these  metals  may  be  raw
material  impurities  or  corrosion  products  from  the  reaction
between chlorine or acid and the process equipment materials of
construction.

     Toxic organic  compounds also occur  in waste  waters from
graphite anode plants  because of the  attack  of  chlorine  on the
anode  material.    They  appear  primarily  in  waste  streams
associated with the purification of chlorine.
                               231

-------
to
Ul
to
               TflBIE 11-26.    TOME MEEM, CCNCEOTRfimJOE MD LORDS IN CELL EDOM HRSEB HMERS KS SCREaflNG flND

                                          ELMHS/ ngA \

                                                \kgAkg;

Pollutant

Mtimony
Arsenic
Cadmium
Qircniuu
flapper
Lead
ffercury
Hictel
Silver
Zinc
014
*
0.010
0.0000001
*
0.94
0.000014
0.53
0.0000075
0.26
0.0000039
*
54
0.00081
*
*
261

*
0.17
0.0000064
0.037
0.0000014
1.9
0.000071
17
0.00064
2.0
0.000075
*
22
0.00081
0.018
0.0000007
1.5
0.000054
Plant f
73B&
0.050
0.0000081
*
*
*
0.24
0.000042
0.044
0.0000077
0.003
O.OOOOOOS
*
*
0.046
0.0000080
738B

*
*
*
0.075
0.000012
0.38
0.000061
0.11
0.000018
*
0.061
0.0000098
*
0.46
0.000074
736

0.038
0.0000031
0.17
0.000014
*
0.54
0.000044
1.1
0.000090
0.047
0.0000038
0.002
0.0000002
0.67
0.000055
*
0.58
0.000048
967**

0.41
0.00015
0.45
0.00017
0.016
0.0000059
0.086
0.000032
2.4
0.00089
370
0.14
0.001
0.0000004
0.36
0.00013
*
0.92
0.00034
                Below detection limits

                ** Graphite anode plant

-------
TSffiE 11-27.
RAH WASTE TOXIC METALS CCKCEKTEATIOH AND LOADS IN PROCESS STREWS OTHER THAN CELL BOOM WASTES
EBOM SCKffiKJHG SW> VERHTCKCICH PILOTS

Pollutant
Antimony
Arsenic
Cadniun
Chromium
een»r
lead
Mercury
Nickel
Silver
Zinc
|014
*
*
0.002
0.0000018
*
0,004
0.0000036
*
0.002
0.0000018
0.003
0.0000027
*
*
I738A
*
*
*
0.53
0.0046
0.041
0.00035
0.083
0.00060
*
0.21
0.0018
*
0.29
0.0021
Plant
I738B
*
0.011
0.000019
*
0.065
0.00011
0.094
0.00016
0.11
0.00019
*
0.067
0.00012
*
0,058
0.00010
§736
*
*
0.038
0.0000062
*
0.090
0.000015
*
0.003
0.0000005

*
4.3
0.00070
1967
*
0.29
0.0020
*
#
0.030
0.00020
0.40
0.0027
0.002
0.000014
0.052
0.00035
*
0.15
0.0010
Avg
*
0.15
0.0010
0.020
0.0000040
0.29
0.00014
0.043
0.00014

0.002
0,0000054
0.088
0.00072
*
1.5
0.0037
  Below detection limits

-------
TABLE H-28.
RAW mSTE TOXIC ORGANICS AT A GRAPHITE ANODE PLANT

CELL
Pollutant
benzene
carbon tetrachloride
1, 2-dichloroethane
1, 1, l-trichlorcjethane
hexachloroethane
1 , 1 , 2-trichloroethane
1,1,2, 2-tetrachloroethane
cdiloroform
1, 1-dichlorcethylene
2 , 6-dinitrotoluene
methylene cWLoride
brcmoform
dicMorcbrcnicinetnane
ctilorcxiibrcsnamethane
hexachlordbutadiene
bis (2-ethylheHyl) phthalate
di~n-butyl phthalate
tetxadiloroethlene
toluene
trichloroethylene
Concentration*
(mg/1)
0.00040
0.023
0.079
0.00014
0.010
0.00040
0.000044
0.085
0.000026
0.000026
0.00056
0.000063
0.035
0.002
0.004
0.00075
0.00078
0.036
0.0030
0.020
Load
(kg/day)
0.0011
0.066
0.23
0.00040
0.029
0.0011
0.00013
0.24
0.000074
0.000074
0,0016
0.00018
0.10
0.0057
0.011
0.0022
0.0022
0.10
0.0086
0.0057
 Flow-proportioned concentration
                                    234

-------
TKBIM 11-29.
RIW WaSTE TOXIC OBOTICS BY W5STE WKTER SOURCE M A
MODE PLffiNT
SUBCMEQORZ
          CfflJORINE DIM»HRfiGM ORT.T,

Stream
Cell building wastes
Caustic filter backwash
Brine filter backwash
Cell wash
Chlorine condensate and
Spent H2S04
Scrubber waste
Totals
Total Toxic
Qrganics
(fflg/1)
0.126
0.057
0.003
0.20
2.2

0.81
0.30*
Total Toxic
Organics
(kg/day)
0.0093
0.12
0.00050
0.014
0.70

. 0.015
0.86
Percent of
Total Toxic
Organics
1.1
14
0.06
1.6
81.5

1.7
100
  Flow-proportioned concentration
                                     235

-------
     Asbestos  occurs  in  all  waste waters  from  diaphragm cell
plants, and in large  quantities  in cell room waste waters when
cells are cleaned and repaired.

11.11.2  Prevailing Control and Treatment Practices

     Section  11.10  described  specific  control   and  treatment
practices  at  eleven   plants.    The  prevailing  practices  at
diaphragm cell plants  are to control asbestos wastes by settling
or filtering cell wash, waste waters and  to neutralize and settle
all  waste  waters before  discharge.  The recycle or  reuse of
waste streams  is  practiced to varying  degrees in the industry
depending on plant-specific factors such as  raw material quality
and type of anodes used.

     Plants  using  graphite anodes  are  treating lead-bearing
wastes by chemical precipitation and settling and/or filtration
before discharge.

     The control of toxic organic compounds  in the waste streams
at graphite anode plants  also varies in'the  industry.  At Plant
1967 where the end use of the chlorine is  captive,  involving its
direct application  to the manufacture of  a  chlorinated organic
product,  the  bulk  of chlorinated  organic   impurities  are not
removed.

     At Plant  f!95r where a more purified product is required,
the  organics  are accumulated  in the reboiler of the chlorine
scrubber.  The residues are treated batchwise for  separation and
recovery of the organic phase  materials which are then sold as
feedstock for  the manufacture of  related products.   Prior to
discharge  the  aqueous  phase is   vacuum  stripped  to  remove
additional  organics and  chlorine  for   recycle.   Normally, one
batch of organics  is  treated per week.   After  separating each
batch of organics and stripping  the residual  aqueous phase, the
quantity of waste water discharged  is approximately 5.7 m3/week
or 0.8 m3/day.  The organic loading in  this waste is not known,
however,  if  the assumption  is  made  that  the  discharge  is
saturated  with  carbon tetrachloride   (CC14)  (800 mg/1   8  20
degrees. C) , the waste load would be 0.5 kg/day.

     A-lthough the daily  mass emissions  from the  two plants are
likely to be similar and both would  require additional treatment
to achieve  acceptable discharge  levels, the wide difference in
concentrations of the  chlorinated organics as well as the manner
in which they  are handled would  necessitate the  application of
an advanced treatment technology  specifically suited  to  each
case.

     Where the flow is large and  the concentrations are low, the
application  of  activated  carbon  adsorption to   the  collected
                              236

-------
organic-bearing waste stream at Plant  f967  would be capable of
reducing a CC14 mass emission from 0.066  kg/day  to approximately
0.03 kg/day, assuming an  achievable  treatability level of 0.10
mg/1.

     In the case of Plant |195, where the volume of waste water
is small but the concentrations of  residual chlorinated organics
can be on the order of several  hundred  parts per million, a more
appropriate removal technology would be steam stripping with an
overhead return to the  process.   Assuming a treatability level
of 10  mg/1 for CC14  using  this  technology,  its  mass emission
could be reduced to approximately  0.001 kg/day.

11.11.3  Process Modifications and Technology Transfer Options

Anode Material

     The  use  of  metal  anodes   rather   than  graphite  anodes
increases   cell  power   efficiency  and   greatly  reduces  the
pollutant  loads of  lead  and toxic organics  in  plant  waste
waters.  Approximately half of the diaphragm cell production of
chlorine is now by metal  anodes.
                                    \
Caustic Evaporation Water

     The vapors  from the evaporative  concentration of caustic
soda are either contact-cooled or  cooled  in surface condensers.
Plants practicing  contact cooling  through barometric condensers
generate large amounts of waste water contaminated with caustic
soda and salt.  By changing from contact cooling of the vapors
to noncontactf cooling, or by recirculating barometric condenser
water,  the  amount of  waste  water  generated  can be reduced
considerably.   If  the change  is  considered  too expensive or is
not  feasible,  demisters_ or  similar  control  devices can  be
installed  to  reduce  the  salt  and  caustic  carryover  in  the
vapors.

Diaphragm Material

     Although not  in full scale use at  any U.S.  chlorine plants,
modified diaphragms have  been  developed  which can reduce power
consumption and minimize or eliminate asbestos  discharges.  The
modified diaphragms  include polymer membrane  and  ion exchange
membrane diaphragms.

     Polymer Modified Asbestos Membranes - These  consist  of a
polymer  treated asbestos  diaphragm baked  into  place on  the
cathode. ,   Its usage  results   in  power  savings  and has  an
environmental  benefit,  since, -at the  time of  rebuilding  the
                              237

-------
cathodes, the discarded material is produced in stablized pieces
instead of loose asbestos fibers.  Final disposal is thus safer
and easier.

     Polymer Membranes - These consist of a microporous Teflon®
type  polymer"andtheir   operation  has  been  demonstrated
successfully  in laboratory  and pilot plant  scale cells.   In
addition  to the benefits  of cost  savings through  energy  use
reduction and longer lifer their use eliminates the handling and
disposal problems associated with asbestos.

     Ion  Exchange   Membranes   -  These  membranes  allow  the
production of a concentrated  caustic similar to that produced by
mercury cells.  The production of salt-free  concentrated caustic
will  reduce  the  waste  water  associated  with  the  caustic
evaporation process.   Like  the polymer membranes, the problems
associated  with  the  handling   and  disposal  of asbestos  are
eliminated.

Itiquefaction of Chlorine

     Utilization of high pressure and refrigeration for chlorine
recovery will reduce  the chlorine content in tail gases.

Tail Gas Emission Control

     As with mercury  cell plants, chlorine in  tail gases has to
be  removed  and treated  or   recovered  before  venting to  the
atmosphere.    The  common practice  is  to  scrub the  gas  with
caustic   soda  producing  a  hypochlorite  solution.     This
hypochlorite  can  then be  sold, used  on-site  or,  discharged.
Decomposition  is  a common method  of removing the  chlorine in
this stream prior to discharge.  Catalytic,  thermal and chemical
methods  of  decomposition,   described  in  Section 11.4.3,  are
effective.

11.11.4  Best Management Practices

     The following Best Management Practices are  common industry
practices and  are provided for  guidance  purposes although they
may not meet  the  requirements of the Resource Conservation and
Recovery Act -(as amended, 42 USC 6901 et. seq.).

Area Runoff

     Provisions can  be made  to divert and contain storm runoff
from  areas  where lead  or  asbestos  contamination  could occur.
Collected runoff can  then be treated with other  wastes.
                               238

-------
Leaks and Spills

     Provisions can  be made in  cell  room areas to control  and
collect the leaks or spills contaminated  with lead or asbestos.

Contaminated Solids

     Asbestos  waste  and  precipitated metals wastes  should be
stored in a lined pond or disposed of in  a secure landfill.
                                         * 2
11.11.5  Advanced Treatment Technologies

     The methods  available  and  currently used  in the industry
for the removal of lead and other toxic metals  from plant waste
waters include hydroxide or carbonate precipitation followed by
settling  or  filtration.    Further  removal  of  metals  can be
effected  using   sulfide  precipitation,   adsorption  and   ion
exchange.

     Removal  of asbestos  from  cell wastes is improved with  the
addition  of  coagulating  agents  prior  to filtration  of these
wastes,
11.12  SELECTION OF APPROPRIATE TECHNOLOGY AND EQUIPMENT


11.12.1  Technologies for Different Treatment Levels

Level 1   (BPT)

     Level  1  treatment  addresses  the waste  characteristics
associated  with  diaphragm  cell  plants using  graphite anodes.
The  data  from  graphite  anode  plants were  used  because the
pollutant load is greater than for  metal anode plants.  Existing
plants that have changed from graphite anode to  metal  anode will
have residual effects that  increase their loads for an extended
time after  the  change - possibly  as  long  as  two years.   Waste
streams  from  the  cell  rooms  and  the cathode  wash  station
(asbestos pond overflow)  are collected in a holding tank  where
they  are   combined   with   any  other  process  waste  sources
containing  treatable levels  of  lead  and  other  toxic metals.
Alkaline  precipitation  of  the toxic  metals  is  accomplished by
the addition of soda  ash.  The solids  are removed by filtration
and  the  filtrate may  be  combined  with  other  process   waste
streams  such  as  chlorine condensate,  tail  gas  scrubber  water,
caustic filter backwash and  barometric condenser waters found to
be contaminated  with toxic metals  at levels usually below the
limits of treatability by alkaline precipitation.  Because the
other  process water  sources are  normally  alkaline  the  pH is
relatively  unchanged  and  clarification  for suspended  solids
                               239

-------
removal also achieves some additional  removal  of traces of toxic
metal  hydroxides.    Thus, the  combined  flow  is  clarified  and
discharged directly  or,  in some cases,  it may be combined with
noncontact  waste waters  and passed  through   a  polishing pond
system prior  to  final discharge.  At all  levels  of treatment,
the brine mud is collected in lagoons and the  effluent recycled
to process.  The flow diagram for Level 1 treatment is shown in
Figure 11-21.

     Level 1 treatment was ultimately selected as the basis for
BPT because it represents a typical  and viable  industry practice
for the control of asbestos  fiber, lead, and other toxic metals
in  waste waters  associated with  diaphragm cell  plants  using
graphite anodes.  Plants  utilizing metal anodes are expected to
have lower levels of toxic metal emissions and may not require
alkaline  precipitation to  meet the proposed  BPT limitations.
All 39 plants in the industry  presently  have  BPT or equivalent
treatment technology  installed.

Level 2  (BAT)

     The  objective  of  Level  2  treatment  technology   is  to
achieve,  at  a reasonable cost, a  greater degree  of  asbestos
fiber and toxic metals removal  than provided by Level 1.  Thus,
Level 2 adds dual-media filtration  to the combined effluent from
Level   1   treatment  excluding   noncontact   waste   streams.
Dechlorination of the  final  plant  effluent is also included in
Level  2  (BAT)  treatment.   This  assumes  treatment by  sulfur
dioxide or  bisulfite to remove  total residual chlorine  to the
detection limit of approximately 0.2 mg/1.  This  is a reasonable
value  for  a waste ^ater sample,  since  the  Iodine  Method  for
determining total residual chlorine is affected by the color of
the sample.

     Level  2  was   finally  selected  as   the  basis  for  BAT
regulations  on  the  strength  of  technology   transfer  options
within the inorganic chemicals  industry and because four out of
five plants sampled  were  meeting limits  derived from published
treatability  data.    In  addition,  two  plants  are known  to
practice dechlorination of the  final effluent.  The flow diagram
for Level 2 is shown  in Figure  11-22.

Level 3

     The  practice  of  sulfide  precipitation of mercury  in the
mercury cell segment of the  chlor-alkali industry suggested the
application of this  technology  for achieving greater removal of
toxic metals  in  diaphragm cell  plants.   Level  3  adds  sulfide
precipitation to Level 2  as  shown in Figure 11-23.  This option
was not  selected  due to  its relatively high  cost  per  pound of
additional metal removal  obtained.
                              240

-------
                     r-\
ro
*>.
H
                              Ds
                       incite WHK
                                 fcfcstff
                       OEUittW) .
                                   HDUJINGIM*
                                                 rnswMD
                                                  SMEGE

                                                  moans
                                                                                          1'  "»Q 	»• groan1
n»

 IWEFIU,
                                                                                          Includes pH monllorltif?, flaw tnonltoring

                                                                                          and atamptf t-
       Figure 11-21.  Level 1 wste -water treatmert:--fer chlorine - diaphragm cell subcategory.

-------
   p
   I
   •i
n
<  *<
1
i-.
                                       I
                                       J
                                      • I
                                       1
                                       II
                                       II
                                                    8
                                                   CM
                                                   CV

                                                   A
                            242

-------
10
*»
U)
       (tew,
                 r-\
/—n
        I   SCOllK

r
f












WaMSH










'

Jj WfDUKi 3SHK
i
1
i
i
1






1^

lUXniUTIMB
«>w
1

Icaj.
PO^H • — .^
. MSIES "

1
1
SMSMS
3MUWIB))









,



IBM
IKLC
W




























1
'C *
\ /
\ /
\




1
Ml
IMS
NK

**"
i
|r



r*— 
-------
11.12.2  Equipment for Different Treatment Levels

Equipment Functions

     Conventional sludge dewatering by filter press is used for
asbestos sludge before  disposal  and dual-media filter backwash
is  returned  to the influent surge  tank.   Level three requires
the  addition of  a  reagent mixing  tank  and  chemical solution
feeder.  Level 3 treatment requires  the  addition of a reagent
mixing  tank  and chemical solution  feeder  to introduce ferrous
sulfide ahead  of  the  Level 2  dual-media filter.  All equipment
is conventional and readily available.

Chemical Handling

     Nonhazardous  solutions  of  aluminum  sulfate and  sodium
carbonate are the only solutions used at Levels 1  and 2.  Inert
filter  aid  used in the  alum sludge filter process presents no
unusual hazard.   At Level  3  the potential  hazard of handling
sodium sulfide  is nullified by reacting it with ferrous sulfate
to  form ferrous  sulfide.   Any excess  ferrous  sulfide  will
oxidize to  a ferric  sulfide  precipitate.   At  the point where
sodium sulfide is reacted with ferrous sulfate, good ventilation
is  essential  to  avoid  the hazards  associated  with hydrogen
sulfide gas.

Solids Handling

     For all three levels  of  treatment,  brine mud  solids are
accumulated  in lined  lagoons on-site.    Asbestos  solids  and
precipitated metals wastes  are  to be sent to suitable chemical
landfills.
11.13  TKE&TMENT COST ESTIMATES


11.13.1  General Discussion

     To prepare treatment cost estimates, a model plant concept
was developed.  Because  higher pollutant loads and larger unit
flows  exist  at  graphite  anode plants as opposed to metal anode
plants  the characteristics  associated  with these  plants were
used  when  possible  for the  model  plant  characteristics  as
discussed  below.

     The  preliminary cost  estimates presented,  in  this  report
were   based  on   incomplete  industry  data  on  waste   source
characteristics and flow rates.   The cost  estimates assumed a
flow rate  of approximately  1.2 m3/kkg for the waste stream from
the cell  room^  asbestos  pond, and other sources (Table 11-20}.
                              244

-------
Later  revisions  of the  flow rates and  the inclusion of other
waste  sources in  the total  volume  of  process  related waste
streams to be handled brought the total model plant flow  rate up
to the  currently  estimated 8.8 m3/kkg  {Table  11-34).   This is
the  flow  used for the development of proposed regulations for
the  diaphragm  cell  segment  of  the  industry.    Since  cost
estimates  were  based on  the preliminary flow  estimate  of 1.2
m3/kkg, the Agency is assessing  the need for making appropriate
adjustments  in  the cost  estimates.   Such  adjustments will be
made prior to final promulgation of the regulations.  The model
plant  specifications  given below are those used for regulation
development purposes.

Chlorine Production

     Approximately  60 percent of  the production  data for all
chlorine  plants  using  diaphragm  cells  is  available  on file.
Production ranges  from  15,000 to 1,500,000  kkg of chlorine per
year.   Three model   plants  with  production  rates  of 19,100,
95,500, and 191,000 kkg per year were selected  to  represent the
subcategory production range.

Waste Water Flow

     Based on industry flow data (Table 11-20),  waste streams in
the model  plants  are segregated into brine mud,  cell  wash and
cell  room wastes,  and  other process  wastes  such as  filter
backwashes,  condensates   and  tail  gas  scrubber  wastes.   For
treatment  cost  estimates  at   all  levels  of  treatment  the
following flow basis  was used.

     A.  A brine mud  flow of  0.42  m3/k*kg is sent to lagoons for
solids  removal.    Solids  are disposed  of  on-site and  other
overflow is recirculated  to process.

     B.  Cathode  or  cell  wash   waters,  heavily  laden  with
asbestos  are  sent to asbestos removal  at  a flow  rate of 0.07
m3/kkg.

     C.  Cell room wastes consisting of leaks, spills, and area
washdown  contaminated with lead and  other  metals  are combined
with treated cell wash waters for  a total flow of 1.2. m3/kkg to
be treated for"metals removal.

     D.  Other  process  waste  water  sources  account for 'an
additional 7.6 m3/kkg which  is combined  with effluent  from the
treatment of  wastes  from the cell room and  cathode wash areas.
This brings the mod'el plant total  flow rate to an estimated 8.8
m3/kkg.   The final,  combined  process  waste  flow is  either
clarified and discharged  as  in Level  1 treatment or clarified,
                             245

-------
passed through dual-media filtration, and dechlorinated prior to
discharge as in Level 2 treatment.

Solid Waste Produced

     Brine  mud  constitutes  the  major  source  of  solid waste
generated at  chlorine plants.   Although solids content varies
from plant  to  plant,  an average of ten percent solids was used
for the model resulting in a solids load  of  42 kg/kkg.  Asbestos
from cell  wash  operations  and precipitated  solids from metal
treatment generate a solid waste of 0.83 kg/kkg.

Chlorine Bearing Wastes

     In the selection of model  plants, the following assumptions
have been made for the chlorine contaminated waste  streams.

     The chlorine condensate waste stream has not been included
in the waste  streams  going to the treatment  facility.   In the
majority of the chlorine/caustic plants,  this  stream is stripped
of chlorine by steam  or  vacuum and the chlorine is recycled to
the purification operation.  The waste water is  then returned to
the  process and introduced  to the brine purification  unit or
sent  to  the   treatment  unit.   The quantity  of  waste water
generated  by  this  operation  is  small.   In  some  cases  the
chlorine gas from the cells is  contact-cooled  with water  and the
scrubbed  liquid,  after  steam  stripping,  is  reused.    The
stripping operation in  the recovery of chlorine is part of the
process  and,  therefore,'  its   cost   is   not  included   in  the
treatment cost.

     The  spent  taii  gas  scrubber  solution,  which  is  mainly
calcium/sodium hypochlorite, is assumed to be used or decomposed
before it is discharged.  Thermal decomposition  can be practiced
at  no  additional  cost  at   some  facilities,  while   another
efficient method is catalytic decomposition. The cost estimates
for decomposition are not  included here  because at many plants
the  hypochlorite   stream  is   sold,  used  on-site   or  only
infrequently discharged, depending on market demand.

     However,  because  of  the environmental  effects of  high
levels of chlorine in waste water discharges, a separate set of
cost  estimates  have  been prepared  for  the  dechlorination of
total plant discharges using sulfur dioxide.

     Chlorinated  organic wastes  - The  chlorine-bearing waste
streams at graphite anode plants are  also those  streams carrying
the  highest concentrations  of toxic organics  as  indicated in
Table 11-29.  Section 11.11.2  discussed  the techniques  used to
recover and remove organics from waste streams at Plant f!95 and
the fact that  organic contaminated streams  can  exist  as either
                              246

-------
high volume-low concentration  or  low volume-high concentration
depending  on  plant  specific  factors.   Costs  for  removing
organics  are  not  included  in the  model plant  cost estimates
because organics  are not limited  in  the regulation.  Orqanics
occur at  low levels  at  most of  the  plants  and when  they are
present  the  appropriate treatment  method  is  site specific.
Althdugh the costs are not included  the  following information is
provided as guidance.  The additional costs for steam stripping
in  a plant  (such  as Plant  |195)  which  already has  a  vacuum
vaporizer,  would  be under  $10,000  for  modification of  the
existing  equipment.    Steam  costs  could vary  from $1,000  to
$5,000  per  year.   If  a vaporizer  is  not  in place, a  steam
stripper to process  5  to 30  m3/week would cost roughly $50,000
to  $100,000,  depending  on  the   input  concentrations  to  be
handled.  The corresponding steam costs would range  from $2,000
to $10,000 per year.

     The capital  costs of  an  activated carbon adsorption unit
for  handling the  relatively  high volume  wastes  with  a  low
influent  organic  loading  (as  found ~at Plant 1967)  cannot  be
reliably estimated in the absence of specific treatability data
on the waste streams in question.

     Alternatively,  incineration   of  the chlorinated  organic
residuals is an effective means of  destroying  and  disposing of
this  material provided  that  adequate measures  are  taken  to
control the release of HC1 to the atmosphere.       •

     A process evaluation should  be made to  determine the most
efficient means for isolating and collecting the organic bearing
waste streams prior to treatment.

     Incidental removal of chlorinated  organics will occur with
the  application of  model  plant   treatment  levels  previously
presented.  Such removal, however,  is  expected  to be  erratic and
therefore cannot be predicted.  Because  raw waste concentrations
of these organics vary considerably depending on plant practices
and  are marginally treatable at times, applicable  control and
treatment   technologies   will  need   to  be  assessed   on   a
case-by-case basis.

     For  these  reasons, the  Agency is  not  providing  specific
numerical discharge  limitations for  organic  pollutants,  but is
providing guidance for evaluating control options that could be
applied in the industry.

11.13.2  Model Plant Treatment Costs

     On the basis  of  the model plant  specifications and  design
concepts presented earlier,  the estimated costs of treatment for
three models  having different production levels are  shown  in
                              247

-------
Tables 11-30, 11-31, and 11-32.  The cost of Levels 2 and 3 are
incremental over Level  1 costs  and  provide  for higher effluent
quality with respect to toxic pollutants.

     Table  11-33   presents  a   summary   of  the   unit  cost
distribution between amortization and operation and maintenance
components.  Annual treatment costs as a function of production
rate is shown graphically in Figure 11-24.  Similarly presented
in  Figure 11-25  is the relationship  of unit cost  (cost  per
metric ton) to production rate.

     For  plants  requiring  dechlorination  of waste waters, cost
estimates  for  dechlorination of  plant effluents  using  sulfur
dioxide are discussed in Section 11.6.3.
11.14  BASIS FOR REGULATIONS


11.14.1  Basis for BPT Limitations

     BPT regulations  are  currently in effect for the diaphragm
cell  chlorine subcategory,  40  CFR 815.62(b) .   The  Agency is
proposing  to  revise  the  limitations,  however,  based  on an
increased unit flow rate.

Technology Basis

     For  BPT,  the Agency  is proposing  limitations based on
equalization,  alkaline precipitation  and settling  of lead and
asbestos-bearing wastes and  neutralization and settling of all
waste  waters before  discharge.    All diaphragm  cell chlorine
plants  are  known to be using this  technology (Level 1)  or its
equivalent.

Flow Basis

     As  described  in  Section  11.13.1, waste  water streams at
diaphragm cell plants1 are  separated into  two types, those  that
require treatment for asbestos-and metals removal and those  that
do n6t  require such treatment.   From data  presented in Table
11-20,  the  unit  flow rate of 1.2 m3/kkg  of  cell room and  cell
wash wastes  from one graphite anode plant was  selected  as the
flow basis for wastes  to be treated.  Graphite anode plant  data
were used  in this instance because the flows  were higher  than
those of other plants and  thus represent  a conservative estimate
of flow for other plants in the industry.  Using available  flow
data  the  remaining waste streams  total 7.6  m3/kkg as shown in
Table 11-34.   Thus the total unit flow discharge  'used  in the
development of effluent limitations is 8.8 m.3/kkg.
                              248

-------
                TABLE 11-30. MODEL PIANT TREATMENT COSTS
Subcategory CHLORINE Diaphragm cell
Production 19,100 metric tons per year
54 metric tons par day
Waste water flow 68 cubic meters per day
A.
B.
C.
INVESTMENT COST
Equipment in place,
including piping,
fittings, electrical
Monitoring equipment
in place ...«*.,.......
Engineering design
Incidentals, overhead,
fees, contingencies...
TOTAL INVESTMENT COST
OPERATION AND
MAINTENANCE COST
Labor and supervision.


Taxes and insurance...
Residual vaste
Monitoring, analysis
TOTAL OPERATION AND
MAINTENANCE COST
AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
FIRST
$57,100
106,850
9,000
34, 590
34,590
21,000
$263,130
«
$112,000
2,200
1,500
24,213 '
7,893
5,800
15,000
$168,606
$39, 394
$208,000
( ,21,057 tons per
( 60 tons per
*
year)
day )
LEVEL OF TREATMENT*
SECOND THIRD
$1,800 $2,250
17,900 20,400
3,940 4,530
3,940 4,530
$27,580
$14, 000
300
2,758
827
7,500
$25,385
$4,487
$29,872
$31,710
$14,000
300
100
3,171
951
7,500
$26,022
$5,159
$31,181
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.

                                 249

-------
                TABLE 11-31. MODEL PLANT TREATMENT COSTS
Subcategory CHLORINE Diaphragm cell
Production 95,500 metric tons per year
272 metric tons per day
Waste water flow 340 cubic meters per day.
( 105,288 tons per
( 300 tons per
year)
day )
LEVEL OF TREATMENT*

A. INVESTMENT COST

Equipment in place,
including piping,
fittings, electrical
work and controls.....
Monitoring equipment

Engineering design

Incidentals, overhead,
fees, contingencies. . .

TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.


Maintenance. ..........
Taxes and insurance ...
Residual waste

Monitoring, analysis

TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
FIRST

, $148,100



219,700

9,000

75,360

75,360
63,000
$590,520


$112,000
4,900
7,500
52,752
17,715

29,000

15,000


$238,867

$85,827
$324,694
SECOND

$2,900



27,000



5,980

5,980

$41,860


$14,000
600

4,186
1,255



7,500


$27,541

$6,810
$34,351
THIRD

$3,350



29,500

9

6,571

6,571

$46,002


$14,000
600
500
4,600
1,380



7,500


$28,580

$7,484
$36,064
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost

                                 250

-------
                TABLE 11-32 MODEL PLANT TREATMENT COSTS
Subcategory CHLORINE Diaphragm cell
Production 191,000 metric tons per year ( 210,577 tons per year)
545 metric tons per day ( 601 tons per day )
Waste water flow 680 cubic meters per day.
A. INVESTMENT COST
Equipment in place,
including piping,
fittings, electrical
Monitoring equipment
Engineering design
and inspection ........
Incidentals, overhead,
fees , contingencies ...
TOTAL INVES1MENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Energy. ...*..»........


Taxes and insurance...
Residual waste
disposal ..............
Monitoring, analysis
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
DIVESTMENT COST
TOTAL ANNUAL COST
FIRST
$271,900
295,500
9,000
115,280
115,280
123,000
$929,960
$112,000
8,000
15,000
80,696
27,898
58,000
15,000
$316, 594
$131,292
$447,886
LEVEL OF TREATMENT *
SECOND
$4,800
43,500
9,660
9,660
$67,620
$14,000
600
6,762
2,028
7,500
$30,890
$11,001
$41,891
THIRD
$5,250
46,000
10,250
10,250
$71,750
$14,000
^600
1,000
7,175
2,152
7,500
$32,427
$11,673
$44,100
*First level represents the base cost of treatment system.
Other levels represent the incremental cost above base cost.

                                251

-------
                 TABLE  11-33.  MODEL PLANT TREATMENT COSTS
Subcategory  CHLORINE  Diaphragm cell
                                           Annual Treatment Costs ($/kkg)
 COST ITEM
PRODUCTION  FLOW
(kkg/yr)  (m3/day)
         LEVEL OF TREATMENT

FIRST     SECOND    THIRD    FOURTH
Annual Operation
and Maintenance
Annual
Amortization
Total Cost
   19,100      68      8.83      1.33      1.36     Not
   95,500     340      2.50      0.29      0.30   ApplieabL
  191,000     680      1.66      0.16      0.17
   19,100      68      2.06      0.23      0.27
   95,500     340      0.90      0.07      0.08
  191,000     680      0.69      0.06      0.06

   19,100      68     10.89      1.56      1.63
   95,500     340      3.40      0.36      0.38
  191,000     680      2.34      0.22      0.23
                                 252

-------
  500
  400
  SOO
u

D

<
  200
  100
       TT
          i t  I/
           f
         W I/I "

             i
                ,  .  I  I
                    Z
                      I
                           z
                        TTT
                            JZ
                        /TT/T
z
 i r
                                   !  i  '
         Z
                                     /LEVEL, #1

            I  I !
                                 J	I
                                          1
                                              i  i
                           i  '  i
                             	L.L
                                                     I  I
                                                        1
                                                           J	i
                                                            I I
                                                               i  I
                                                            I I
              50       100       150      200
                 PRODUCTION (METJUC TONS/YEAR x 1000)

Figure 11-24, Annual treatment cost vs. production for the Chlorine
                 Subcategory (Diaphragm Cell Process)
                            253

-------
                                   1  I
                                               I  i
                                                         TT
               J	I
                                               I  i
 12
 10
         I	i
S  i
                                    IT
                                                      I  i
          LEV! LS i»2 EC #3
          I  I
                                                    i  I
                                                             I  !
         \
      TTTv
                                  i  i
            v\
            T\
               \r
       I !
        I TV
  i  i
         i :
          : LEA 'EL #11  ;
                                                      I  I
          !  i
                                                TT
                    i i
                                              i  !
                                                               i  i  i
            I
                                       i  !
                    I
                 !  I
I  i
I  I  I
             50       100       150       200

            PRODUCTION  (METRIC .TONS/XEAR X 1000)
Figure 11-2 5. Annual unit treatment cost vs. production for the
             Chlorine Subcategory (Diaphragm Cell Process)
                              254

-------
      11-34
OF UNIT FECWS Jffi DiaPHRSSM CFT,T,
SUBCATEGOKSf
         DlftPHRAQf
Stream Description
         Onit Plow
         Data
        Source
Cell room and cell
    wash wastes

Chlorine eondensate

Tail gas scrubber waste

Caustic filter wash

Brine filter wash

Caustic cooling blowdown


Spent sulfuric acid
           1.2


           0.78

           0.11

          "s.4

           0.45

           0.86


           0.01
Graphite anode plant


Graphite anode plant

Graphite anode plant

Graphite anode plant

Graphite anode plant

Ifetal anode plants
      average

Mstal anode plants
      average
      Total Unit Plow Discharge 8.8
                                     255

-------
Selection of Pollutants to be Regulated

     The  selection of  pollutants for which  specific effluent
limitations are being proposed was based  on an evaluation of raw
waste  data  from screening and verification sampling and on the
treatability of toxic pollutants.

     Table 11-35 presents the achievable concentrations of toxic
metal  pollutants  (found  at detectable  levels  in  raw  waste
streams)  using  the  available  treatment  technology  options.
Based  on literature  treatability data presented in Section 8.1
and  summarized in Table 8-11,  the  concentrations reflect the
lowest level achievable  by  these technologies.  Also  presented
in  the  table  are  the  maximum three-day  average  raw  waste
concentrations  observed during  the  sampling program  with  an
indication   of   the   .number   of   plants   where   treatable
concentrations were exceeded.

     Based  on  the occurrence  of treatable levels of specific
toxic  metals,  arsenic,  cadmium,  chromium, copper, lead, nickel
and  zinc  were  selected as  candidate toxic pollutants proposed
for- BPT  regulations.   Antimony,   mercury,   and   silver  were
detected but at less than treatable  levels.

Basis  of BPT Pollutant Limitations

     Limitations are presented as both concentrations (mg/1) and
loads  (kg/kkg), and the relationship between the two is based on
the  unit  flow rate  (8.8  m3/kkg).    The  concentration  basis
therefore  represents  the  concentration  of   the   total  plant
discharge including both treated and untreated waste waters.

     BPT proposed limitations are presented in Table 11-36.

     Convent ional Pollutants -

     A. pH: The treated effluent is  to be controlled within the
range  of  6.0  to  9.0.   This limitation  is based  on  the data
presented in Appendix B of this  report^and  the JRB Study  (52).

     B. TSS:-The proposed BPT limitations for  TSS are based on a
summary of  monitoring data  from Plant 1207  (3) .   The average
discharge  load of  0.30  kg/kkg  is  used  to  develop  discharge
limitations.    Because variability  factors  for TSS  were  not
available for this plant, factors obtained from the hydrofluoric
acid subcategory were used.  In that subcategory, where the same
technology of  alkaline precipitation and settling is used, the
average variability factor for daily measurements of TSS is 3.5
.and  the average  factor  for 30-day averages  is  1.7.   Thus,
utilizing  the  long-term average discharge  load of 0.30 kg/kkg
one obtains a maximum 30-day average load limit  ofs
                              256

-------
    TABLE  11-35.  COMPARISON OP TOXIC METALS TREATABILITY WITH SCREENING
                  AND VERIFICATION SAMPLING DATA


Arsenic
Antimony
Cadmium
Chrcmium
Copper
Lead
Mercury
Nickel
Silver
Zinc
Tree
(i
Level
1
0.5
0.8
0.1
0.1
0.5
0.3
_J3>
0.2
0.4
0.5
ng/D
Level
2
0.5
0.4
0.05
0.05
0.4
0.05
_»)
0.1
0.2
0.4
Plant Raw
Level Waste Average
3 (rag/1)
0.05
NA
0.01
NA
0.05
0.05
0.01
0.05
0.05
0.02
0.30
0.011
0.037
1.9
17
21
0.003
22
0.018
3.0
Number of Plants
Exceeding
Treatability
3
0
2
4
4
4
0
6
0
3
(1)   Literature-based treatability estimates from Table  8-11.
(2)   Of 6 plants, number exceeding treatability by sulfide/filter.
     (Level 3)
(3)   Treatability with this technology not available.
NA  Not Applicable
                                   257

-------
                             11-36.  PROPOSED LTMITOTIONS
                         Chlorine -  Diaphragm Cell

           Best Practicable Control Technology Currently Available
                         Waste Water Plow:  8.8m3Akg
Pollutant
              Subcategory
              Performance
                (mg/1)
                                  Concentration Basis
                   24-hr
                    max.
                  30-day
                   avg.
                      Effluent Limit
                         (kg/kkg)
                      Max
                      30-day      24-hr
                       avg.        max.
Conventional Pollutants
TSS
Toxic
57
2,1
57
120
0.51
1,1
Arsenic
Cadmium
Chromium
Copper
Lead
Nickel
Zinc
0.50^ ' 2.6
0,1Q(3) 2.6
G.10t3) 2.6
0.50(3) 2.6
1.1^) 2.6
OCT f\ V*""/ *"i f~
,50 2.6
0.50{3) 2.6
0.50
0.10
0.10
0.50
1.1
0.50
0.50
1.3
0.26
0.26
1.3
2.9
1.3
1.3
(5)
(5)
0.00088
0.0044
0.010
0.0044
0.0044
(5)
(5)
0.0023
0.011
0.026
0.011
0.011
(1) - VFR:  ratio of the 24 hour variability factor to the 30-day variability
            factor

(2) - Verification sampling

(3) - Lower limit of literature treatability (Table 8-11): used when observed
      sampling data
(4) - Based on long-term monitoring data

(5) - No effluent limitation proposed
                                     258

-------
     (0.30 kg/kkg)  (1.7) = 0.51 kg/kkg

and a maximum daily limit of:

     (0.30 kg/kkg)  (3.5) - 1.1 kg/kkg

     The concentration  bases then are  derived  by applying  the
model plant flow  rate of  8.8 m3/kkg  to obtain a 30-day average
concentration of  57 mg/1 derived as follows:
    (
0.51 kg/kkgN/lOOO ing/A = 57/mg/l
 8.8 m3/kkg/\kg/mi
and a daily  maximum concentration of 120 mg/1 derived from  the
variability factor ratio  (VFR:  3.5/1.7 = 2.1) as follows:

     (2.1) (57 mg/1) = 120 mg/1

     Toxic Pollutants -

     A.  Lead:  The proposed BPT  limitations for lead are based
on long-term  monitoring  data from one graphite  anode plant as
presented  in Appendix A.   The plant  is  achieving a long-term
average lead discharge of 0.0064  kg/kkg.

     Statistical  analysis of  monitoring data from  the plant
established a 30-day average variability  factor of  1.6 and a  24-
hour  variability  factor  of  4.1.    The  ratio  of  the   two
variability factors, VFR,  is  2.6.  The proposed 30-day  average
limitation  for  lead  was  then  obtained  by  multiplying   the
variability factor for 30-day averages by the  long-term  average
load; i.e.,  1.6  x  0.0064  kg/kkg = 0.010  kg/kkg.   Similarly  the
daily maximum limitation  was  obtained  by multiplying the daily
maximum variability factor by the long-term average load; i.e.,
4,1 x 0.0064 kg/kkg = 0.026 kg/kkg.

     The  concentration  basis  for  lead  is  derived from   the
relationship between concentration  (C), unit lead  (L), and unit
flow (Q) .

     C  (mg/1) *  L  (kg/kkg) /1000 mg/1
                 ,Q  (m3/kkg)\  kg/m3

     Thus the concentration basis for the maximum 30-day  average
for lead is:
    /
    V
0.010 kg/kkg\ /1000 mg/I\ -  1.1 mg/1
 8.8 m3/kkg ) \  kg/m3  ~)
     The concentration basis for the daily maximum limitation is
obtained -similarly or by  applying  the variability factor ratio
(VFR) of 2.6 to the maximum 30-day average concentration:

                              259

-------
     (1.1 mg/1)  (2.6)  =  2.9 mg/1

     Monitoring data from six diaphragm cell plants presented in
Table 11-37 indicates that plants  using metal anodes are meeting
the BPT lead  limitations.   One of two graphite anode plants is
meeting the limitations.

     The  limitations  proposed for  additional  toxic pollutants
are   derived   from   two   sources  -   sampling   data   and
literature-based  treatability  estimates.   The  concentration
bases  are derived from  effluent  sampling  unless  the  observed
concentrations are below treatability estimates.  In such cases
the lower  limit  of  the applicable  treatability  level  was used
(Table 8-1).

     B.   Chromium;   Raw waste concentrations  of chromium were
observed  as   high as  1.9  mg/1   (Table  11-26).    Table  11-38
presents  effluent data from the sampling of two diaphragm cell
plants   which    indicate   an    achievable   final   discharge
concentration of  0.05 mg/1  chromium.' Because this is below the
treatability  estimate  of 0.10 mg/1  with  BPT technology  (Table
11-35), the treatability concentration has  been used as the 30-
day average basis for deriving BPT  limitations for chromium.

     Because  no  long-term  monitoring  data is   available  for
chromium  in  this industry,  the  same variability  factor  ratio
(VPR)  obtained  from  monitoring  lead  in  the discharge  at  one
plant  is used to  obtain daily concentration  limits.

     (0.10 mg/1)  (2.6) = 0.26 mg/1

     To obtain  effluent  lead limitations for chromium, the 30-
day average concentration is multiplied By  the unit flow:

     (0.10 mg/1)  (8.8 mS/kkg)/  kg/m3  N =   0.00088 kg/kkg
                             VlOOO  mg/1/

and the daily maximum effluent limit is

     (0.26 mg/1)  (3.8 m3/kkg)/  kg/m3  N =   0.0023 kg/kkg
                             V1000  mg/1/

     C.  Copper,  Nickel, and Zinc:  Raw waste concentrations of
these  metals  were observed  as high as 17 mg/1 copper,  22 mg/1
nickel, and 3.0 mg/1 zinc.  Table 11-35 indicates an achievable
final  discharge  concentration  of  less  than  0.10  mg/1 for these
metals.  Because  this is below the literature-based treatability
estimate  of  0.50 mg/1 using BPT  technology (Table 11-35),  the
treatability  concentration  has been used as the 30-day average
basis  for deriving BPT limitations  for these metals.
                              260

-------
TABLE 11-37.
LEAD AND TSS DISCHARGES ERCM SELECTED DIAPHRAGM CELL
CHLORINE PLANTS ^
SUBCATEGORY
       CHLORINE - DIAPHRAQ1 CRT.T,
Plant
               Lead Discharge
                   kg/kkg
                        .Average
                                    Maximum
#589*
#738*
§261*
#014*
1967 (3)
#207

Plant
#014*
1207
0.0020
0.0010
0.0025
0.0060
0.0064
0.021
TSS Discharge
kg/kkg
Average
2.8^
0.30
0.030
0.015
0.019
NA
0.026
0.054

Maximum
NA
0.57
(1)  Fran Reference 3
(2)  Plant has "once-through" barometric condenser water
(3)  Long Term Data Appendix A
* Plants with metal anodes
NA:  Not Available
                                     261

-------
        TSELE 11-38.  TOXIC POH1HSNTS IN DIAPHRAGM OKTiT. PLSNT EFFLUENTS
                                         Effluent Concentration
                                                (mg/1)
Pollutant
Arsenic
Cadmium
Chroitiium
Copper
Nidcel
Zinc
Metal .anode Plant
#261 (1)
influent effluent
0.17
0.037
1.9
17
22
1.5
0.12
0.004
< 0.050
< 0.025
< 0.050
< 0.025
Graphite Jtaode Plant
#967 , }
Lead Treatment Plant ^
influent effluent discharge
0.28
< 0.023
0.10
1.6
0.070
0.93
0.36
< 0.015
< 0.050
0.030
< 0.050
< 0.10
0.30
<'0.015
< 0.050
0.031
< 0.050
0.15
(1)
     Cell wash waste filtered with coagulant to remove asbestos
(2)   Flow-proportioned average discharge,  consisting of lead treatment
     discharge and untreated filter backwashes,  condensates  and scrubber
     wastes
                                    262

-------
     Osing  the  same VPR of  2.6  the daily concentration limits
are thus,

     (0.50 mg/1)  (2.6) » 1.3 mg/1

     To obtain  effluent  load limitations for these metals, the
30-day average concentration is multiplied by the unit flow:

     (0.50 mg/1)  (8.8 m3/kkg) f  kg/m3  \ = 0.0044 kg/kkg
                              V.1000 mg/1/

and the daily maximum effluent limit is:

     (1.3 mg/1)  (8.8 m3/kkg) /  kg/m3  \ = 0.011 kg/kkg
                             \1000 mg/1/

     D.     Arsenic  and  Cadmium:     The  maximum  raw  waste
concentrations observed  for  arsenic  and cadmium were below the
literature-based  treatability concentration  for  these  metals
(Table 11-35).  For this  reason only  the concentration bases are
presented   in  Table  11-36.    The  concentrations  represent
treatability for  these metals and  are meant to serve  only as
guidance should these pollutants be of  concern.

11.14.2  Basis for BAT Effluent It imitations

     Previous BAT regulations called for no discharge of process
waste  water  pollutants.   The  regulations were  remanded.   The
proposed BAT regulations provide for the discharge of pollutants
following appropriate treatment of process wastes.

Technology Basis

     Utilizing the  cost  estimates  presented  in this report the
Agency has  analyzed the cost  effectiveness  of  the  base level
system (BPT) and the advanced level options (Levels  2 and 3) for
conventional and  toxic pollutant removal.  The economic impact
on the diaphragm cell chlorine subcategory has  been evaluated in
consideration  of  the  technology  basis   for   proposed  BAT
limitations.  The need for a reevaluation of cost-effectiveness
based  on  new cost data  will be  assessed by the Agency before
promulgation.

     For BAT  the Agency is uproposing  limitations  based on BPT
technology with the addition of dual-media filtration (Level 2)
and dechlorination of all process waste waters.  Filtration will
remove additional toxic metals  and  has  been used successfully in
the mercury-cell  chlorine subcategory.  Dechlorination is being
included in BAT because the  toxicity  of chlorine  to aquatic life
is  well  documented  and  it  is  a  pollutant  of  concern  to the
Agency (59).   Two chlorine  plants are known  to be practicing
                              263

-------
dechlorination.  The  Agency considered the addition of sulfide
precipitation (Level 3)  to the treatment  of cell room wastes but
rejected  it  because  further  reduction of toxic  pollutants in
this stream only would not substantially  improve total discharge
quality.

Flow Basis

     The flow basis for  BAT limitations is the model plant total
discharge of  8.8 m3/kkg.   This flow  reflects that expected at
chlorine plants using graphite anodes.

Selection of Pollutants to be Regulated

     For BAT regulations, the Agency has  selected  the same seven
toxic  metals  identified  in  the  proposed BPT regulations,  and
total  residual chlorine.

Basis of Pollutant Limitations

     For BAT regulations, the Agency is proposing  more stringent
controls on the discharge of  the  seven toxic metals of concern
on  the basis  of  physical  removal  by  filtration.   Alkaline
precipitation  converts  most  dissolved metals into  less  toxic
insoluble  forms and  excess alkalinity  exists in most  of  the
process wastes generated  in this  subcategory.    Proposed  BAT
limitations are presented  in Table  11-39.

     Nonconventional Pollutant -

     Chlorine:  Total residual chlorine limits are based on the
detectable  concentration  of  chlorine   (0.2  mg/1)   and   on
performance of  dechlorination in  the electric utility industry
(58) because treatment  should remove essentially all chlorine.
Thus the 30-day average limit was set  at  0.20 mg/1.

     The daily maximum limit for total residual chlorine was set
at 0.34 mg/1 based on an evaluation  of long-term monitoring data
and  determination  of  variability  factors  for  total  residual
chlorine  as  presented  in Appendix  A.   The ratio  of  24-hour
maximum  variability  factors  to   30-day  average  variability
factors for two plants was 1.7, thus  the maximum  30-day average
is given.by:

     (0.20 mg/1) (1.7)  =  0.34 mg/1

     The determination  of load limitations  for  total  residual
chlorine  (kg/kkg) was calculated based on. the unit flow rate of
8.8 m3/kkgr thus the maximum  30-day average  is given by:
                              264

-------
                    TABLE 11-39.  PROPOSED LZMHATIOKS
                         Chlorine Diaphragm Cell
                        Best Available Technology
                      Waste Water Flew:  8.8 m^/kkg


Pollutant


Treatability VF
(mg/D

Concentration Basis
R. Max
30-day 24-hr
avg. max.
Effluent Limit
f\gTf /L*1b"#fc? i
* * \"^"J/ J\ISJ*-1 /
30-day 24-hr
avg. max.
Noneonventional  Pollutant


                    0.2      1.7
Total Residual
 " Chlorine
0.20
0.34
0.0018  0.0030 :
Toxic Pollutants
Arsenic
Cadmium
Chromium^)
O)
Copper w
Lead<2)
19\
Nickel u'
Z-c(2)
0.50^
0.05<3>
0.05' '

0.40(3)
0 . 22
._.
0.101 '
0.40<3>
*
2.
2.
2.

2.
2.

2.
2.
2
2
2

2
2

2
2
0
0
0

0
0

0
0
.50
.05
.05

.40
.22

.10
.40
1.1
0
0

0
0

0
0
.11
.11

.88
.48

.22
.88

0

0
0

0
0
(5)
(5)
.00044

.0035
.0019

.00088
.0035
(5)
(5)
0

0
0

0
0
.00097

.0077
.0042

.0019
.0077
 (1) - VPR: ratio of the 24 hour variability factor to the 30-day  average
      variability factor

 (2) - Also applicable for PSES limitations

 (3) - Literature - based treatability estimate

 (4) - Based on filtration for BPT subcategory performance

 (5) - No effluent limitation proposed
                                     265

-------
     (0.20 mg/1)  (8.8 m3/kkg)/   kg/m3  \ «  0.0018 kg/kkg
                             VlOOO mg/1 /

The 24-hour maximum limit was calculated similarly,

     (0.34 mg/1)  (8.8 m3/kkg) /  kg/m3  \ =  0.0030 kg/kkg
                              VlOOO mg/1/

     Toxic Pollutants -

     Dual media filtration of BPT  effluent will significantly
reduce suspended  metal  precipitates,  BAT  limitations for the
toxic  metals  of  concern  are based  on literature-treatability
studies  as  presented in  Section 8.1  and  summarized  in Table
8-11.

     A.  Lead:  Filtration of the BPT effluent is estimated to
fesult in an  80 percent reduction of  lead  (41)  giving a final
concentration  of  0.22  mg/1.    This  value  is  used  as  the
concentration basis for  the maximum 30-day average limitation of
0.0019 kg/kkg.   Application of  the  model  plant discharge' rate
results in a loading of 0.0019 kg/kkg.  That is,

     (0.22 mg/1)  (8.8 m3/kkg) /  kg/m3  \ =  0.0019 kg/kkg
                              VlOOO mg/1/

     The variability  factor ratio  (VPR)  of 2.2, used for BAT
limitations, is from the analysis of mercury monitoring data in
the mercury cell chlorine subcategory (Section  11.7.2).  Mercury
cell chlorine  plants  typically practice  filtration  of waste
water  and the  value of 2.2 represents  the  average  VFR of four
plants.

     The daily maximum limitation is then,

     (2.2)  (0.0019 kg/kkg)  - 0.0042 kg/kkg

and the daily maximum concentration basis is:

     (2.2)  (0.22 mg/1)  = 0.48 mg/1

     B.  Chromium:  Filtration of the BPT effluent is estimated
to reduce the  chromium concentration by approximately 60 percent
(41) to give a final concentration of  0.050 mg/1.  This value is
used as the concentration  basis  for  the  maximum 30-day average
effluent limitation.  Application of the model plant discharge
rate results  in a corresponding loading limitation of 0.00044
kg/kkg.  That is,

     (0.050 mg/1)   (8.8 m3/kkg)/   kg/m3«  \ = , 0.00044 kg/kkq
                              VlOOO mg/1 /
                              266

-------
and, for  the daily maximum  limitation using  the  VFR value of
2.2, one obtains;

     (2.2) (0.00044 kg/kkg)  =  0.00097 kg/kkg

The corresponding concentration basis  is:

     (2.2) (0.050 mg/1) =0.11 mg/1

     C.   Copper and  Zinc:   Filtration of  the BPT effluent is
estimated  to reduce  the  copper  and zinc  concentrations by 20
percent  (41)  to give  a final concentration of 0.40 mg/1.  This
value is used as the  concentration basis for the maximum 30-day
average  effluent limitation.   Application of  the model plant
discharge rate results in  a loading  limitation  of 0.0035  kg/kkg.
That is,

     (0.40 mg/1) (8.8 m3/kkg)/   kg/m3  A =  0.0035 kg/kkg
                             V1000 mg/1 )

and for the daily maximum limitation using  the  VFR  value  of 2.2,
one obtains:

     (2.2) (0.0035 kg/kkg)  = 0.0077 kg/kkg

and the daily maximum concentration basis is:

     (2.2) (0.40 mg/1) =  0.88 mg/1

     D.  Nickel:  The addition of filtration to the BPT effluent
is  estimated to achieve  a 50  percent reduction of  the nickel
concentration.   The  basis of  the  proposed BAT  limitation is
therefore  0.10  mg/1  and  results in  a maximum  30-day  average
loading limitation of 0.00088 kg/kkg.  That is,

     (0.10 mg/1) (8.8 m3/kkg) /  kg/m3  N =  0.00088 kg/kkg
                              V1000 mg/1/

and the daily maximum is,

     (2.2) (0.00088 kg/kkg)  -  0.0019 kg/kkg

with a corresponding  concentration basis of:

     (2.2) (0.10 mg/1) =  0.22 mg/1

     E.   Arsenic and  Cadmium:   Filtration of BPT  effluent will
reduce  the  cadmium concentration  to  0.050  mg/1  but  will not
significantly  reduce  the  arsenic  concentration of  0.50 mg/1.
Because   maximum   plant   raw   wastes   were    below   these
concentrations,  no effluent limitations  are  being  proposed.
                              267

-------
Concentration values  presented  in Table 11-37 are intended for
guidance only.

11.14.3  Basis for BCT Limitations

Technology Basis

     Utilizing the cost estimates presented in this report, the
Agency  has  analyzed  the  costs of  BAT technology  in removing
conventional  pollutants.     This   technology   of  dual-media
filtration of all process waste water was  found by the Agency to
be cost effective in  removing TSS.  Proposed BCT limits for TSS
are given in Table 11-40.  This calculation is shown in Section
3.3.3.

Flow Basis

     The flow basis  for  BCT  limitations is the same 8.8 m3/kkg
used for both BPT and BAT limitations.

Selection of Pollutants to be Regulated

     BCT regulations only apply to total suspended solids  {TSS)
and pH.

Basis of Pollutant Limitations

     TSS  limitations  are  based  on  technology  performance of
filtration of waste water flow from the mercury cell segment of
the chlorine  industry.  Appendix Table A-l presents long-term
TSS monitoring data from a chlorine plant practicing filtration
of process wastes.   The  maximum 30-day average limitations are
based on  a  30-day average concentration  of 12  mg/1  using the
diaphragm cell model plant discharge  flow  rate  of 8.8 m3/kkgr
namely:

      (12 mg/1) (8.8 m3/kkg) f  kg/m3  \ =  0.10 kg/kkg
                           \1000 mg/1/

     The  variability  factor   ratio  for  this plant  was  1.9.
Applying this value  to the 30-day limit,  one  obtains  a 24-hour
maximum limit of:

      (0.10 kg/kkg)   (1.9)  -  0.19 kg/kkg
                               ^
with a corresponding concentration basis of:

      (12 mg/1) (1.9) = 23 mg/1
                              268

-------
                    TABLE 11-40.  PBQPOSED LIMITATIONS
                           Chlorine Diaphragm Cell
                        Best Conventional Technology
                         Waste Water Flow: 8.8 m^/kkg
Pollutant
Treatability
      Concentration Basis
VFR(1) Max  (mg/1)
      30-day      24-hr
       avg.        max.
                                                            Effluent Limit
                                                            30-day    24-hr
                                                             avg.      max.
Total Suspended
     Solids
               (2)
     12
 1.9   12
23
0.10
0.20
 (1) -  WR:  ratio of the 24 hour variability factor to the 30-day variability
             factor
 (2) -  Limitations based on technology transfer from mercury-cell chlorine
       subcategory; long-term monitoring data from .Appendix A-l
                                    269

-------
11.14.4  Basis for New Source Performance Standards

Technology Basis

     The Agency is basing NSPS limitations on the BAT technology
of alkaline precipitation  filtration  and dechlorination and on
the  performance achieved  at plants  using metal anodes.   The
conversion to metal anodes has largely eliminated the source of
lead  in waste  waters,  but  residual  lead contamination  at  a
converted plant may exist  for as long as  a  year or  more.   New
metal   anode   plants   should   have   relatively   low   lead
concentrations in their waste waters.  Proposed NSPS limits are
presented in Table 11-41.

Flow Basis

     The  flow  basis  of  8.8  m3/kkg  used   for  BPT   and  BAT
limitations is conservatively being used for new sources.

Selection of Pollutants to be Regulated

     For NSPS regulations, the  Agency initially considered the
same  BAT pollutants   (seven  toxic metals and  total   residual
chlorine), pH and TSS.  However, following an evaluation of raw
waste characteristics  at a new  metal anode  facility (shown in
Table 11-42) where residual  metals contamination from previous
graphite anode  use does not  exist, only two toxic  metals were
selected for regulation. A discussion of the selection  of these
metals is presented below.

Basis of Pollutant Limitations

     For NSPS regulations the Agency is proposing more stringent
controls on  the discharge of toxic  metals  of concern  on the
basis of lower  raw waste loads  generated at plants  using metal
anodes.  NSPS proposed regulations ar_e^shown in Table 11-39.

     Conventional and Nonconventional Parameters -

     A.  pH:   The treated effluent is  to  be controlled within
the range of 6.0  to 9.0.   This  limitation is based  on  the data
presented in Appendix B of this  report and the JRB Study (52).

     B.   TSS:   Limitations  for TSS  are the  same  as  in BAT
regulations.

     C.    Total  Residual Chlorine:     Limitations  for  total
residual chlorine are the same as  in BAT regulations.
                              270

-------
TSBU3 11-41.  PROPOSED LIMITAIIONS
    Chlorine Diaphragm Cell
New Source Performance Standards
  Waste Water Flow: 8.8 m3/tckg

,.^ C3oncentration Basis Effluent Limit
Pollutant Treatability VERUJ yigx (mg/1) Max k9>/KfcEr
(mg/1) 30-day 24-hr 30-day 24-hr
avg, max. avg. max.
Conventional and
Non-Conventional
TSS 12
Total Kesidual
Chlorine 0.2
Toxic Pollutants
jy:senic 0.50
Cadmium 0.050
Chromium^2^ 0.050
Copper 0.40
Lead(2) 0.050
Nickel 0.10
Zinc 0.40

1.9 12 23

1.7 0.2 0.34
2.2 0.50 1.1
2.2 0.050 0.11
2.2 0.050 0.11
2,2 0.40 0.88
2,2 0.050 0.11
2.2 0.10 0.22
2.2 0.40 0.88
(1) - VER: ratio of the 24 hour variability factor to
, ^variability factor
(2) - Also applicable to
PSNS limitations

0.10

0.0018
(3)
(3)
0.00044
(3)
0.00044
(3)
(3)
the 30-day


0.20

0.0030
(3)
(3)
0.00097
(3)
0.00097
(3)
(3)


(3) - No effluent limitation proposed
                271

-------
TABLE 11-42.  COMPARISON OF RSW WSJIE CHSRACTEBISTICS AT A NEW METAL
              PLAM? WITH TEEATABILITY OF TOXIC METALS
SUBCATEGOiy
CHEDIIKE DISPHRA^I CKTJ,

Pollutant
Arsenic
Cadmiuoa
Qaomittn
Copper
Lead
Nicskel
Zinc

Treatability (1)
0.50
0.050
0.050
0.40
0.050
0.10
0.40
Cancentxation(mg/D
Plant |738B(2)
Raw Waste
0.011
<0.025
0.066
0.12
0.11
0.067
0.093
(1)- Literature based treatability estimates using BIT technology
     of dual media filtration following alkaline precipitation of
     metals  (Table 8-11)

(2)- Verification sampling at new metal anode facility
                                    272

-------
     Toxic Pollutants -

     Table 11-42  presents  the results of verification sampling
of raw wastes at  a  new  chlorine plant using metal anodes.  The
total  raw  waste concentrations  of  toxic metals  are,  with the
exceptions  of  chromium  and  lead,  substantially  below  the
estimated treatability of these metals using BAT  technology, as
shown  in  the table.   For  this  reason  only chromium  and lead
effluent limitations  are proposed.   Other metals are presented
on a concentration basis for guidance purposes only.

     Lead  and Chromium:   The  treatability  of   both  lead and
chromium using  the  BAT technology  of  alkaline  precipitation
followed by  dual-media  filtration  is  estimated  at  0.05 mg/1
(Table 8-11) .   This value  was used as  the  concentration basis
for   the   proposed   maximum  30-day  average  NSPS  effluent
limitations.   Application of the  model plant discharge rate
results  in a  corresponding  loading  limit  of 0,00044 kg/kkg.
That is,

      (0.050 mg/1) (8.8 m3/kkg)/   kg/m3  \ =  0.00044 kg/kkg
                              \1000 mg/1 )

and  for  the  proposed daily  maximum  limitation   using  the VPR
value of 2.2, one obtains:

      (2.2)  (0.00044 kg/kkg)  =  0.00097 kg/kkg

The concentration basis for  the daily maximum is,
                             i

      (2.2)  (0.050 mg/1) = 0.11 mg/1

11.14.5  Basis for Pretreatment: Standards

Existing Sources

     For Pretreatment Standards for  Existing Sources  (PSES), the
Agency is proposing  the same  limitations  as  for BAT based on the
identical  treatment  technology  without dechlorination  being
used for indirect dischargers (see Table 11-39).   Dechlorination
is unnecessary because chlorination of publicly-owned treatment
works influent is fairly common.  The pollutants  to be limited
are chromium, copper, lead,  nickel, and zinc.

New Sources

     For Pretreatment Standards  for New Sources  (PSNS),  the
Agency is proposing  the same limitations as for  NSPS  based on
the identical treatment technology without dechlorination being
used for indirect dischargers (see Table 11-41).   Dechlorination
is unnecessary because chlorination of publicly-owned treatment
                              273

-------
works influent  is  fairly common.   The pollutants to be limited
are  chromium  and   lead.    The  pollutants  (arsenic,  cadmium,
copper, nickel,  and zinc)  are  not being  limited  based  on the
assumption that all new plants will use metal anodes.  As shown
in Table  11-42,  these  pollutants are below treatability levels
at such a plant.
                              274

-------
                          SECTION  12


                   HYDROFLUORIC ACID INDUSTRY
12.1  INDUSTRY PROFILE
12.1.1  General Description

     Hydrofluoric acid  (Hydrogen  fluoride-HF)  is produced both
as  anhydrous  and  aqueous  products.    It  is  used  in  the
manufacture of fluorocarbons  which  are used  as refrigerating
fluids,     and  plastics,  for  pressurized  packing   and  as
dispersants in  aerosol sprays.  It  is used  in  the production
of aluminum,   in  the  refining and enriching   of uranium fuel,
pickling of  stainless  steel,  in petroleum  alkylation, and  for
the  manufacture  of fluoride   salts.  The industry data profile
is given  in Table 12-1.  The status of regulations is given in,
Table 12-2.

12.1.2  General Process Description and Raw  Materials

     HP is  the  most  important manufactured  compound of  the
fluorine  family  in  volume of  production.   Fluorspar  (mainly
CaF2) 'and sulfuric  acid  are  the  raw  materials used for  its
manufacture.     Fluorspar  and     sulfuric     acid     react
endothermically  at 200-250 degrees C and the reaction  time is
20-60 minutes.   The reaction is given as:

               CaF2  + H2S04 .+   heat = CaSO4  +  2HF           (1)

     The  reaction kinetics  and the  yield of  product  depends
on the purity and fineness of  the  fluorspar.  The concentration
of sulfuric acid,  the temperature of  the reaction,- and the ratio
of sulfuric acid to fluorspar are among important variables.

     Crude fluorspar, as  mined, varies  in CaF2 content  from 35
to 90 percent.  The ore is upgraded by flotation which results
in 98 percent  CaF2  being  available for the production of  HE.
The analysis of a typical upgraded fluorspar is given as:

                        CaF2      Minimum 97.5-98%
                        SiO2      Maximum 1.0%
                        S           "     0.05%
                               275

-------
      12-1     -     SUBCATEGORY PROFILE DATA SUMMARY

SUBCATEGORY          HYDROFLUORIC ACID

Total subcategory capacity rate                   363,000 kkg/year

Total subcategory production rate                 261,800 kkg/year

Number of plants in this subcategory                    9

308 Data on file for                                    8
     With total capacity of                             *
     With total production of                     177,§00 kkg/year
     Representing capacity                              *•
     Representing production                           68 percent

     Plant production range:
          Minimum                                   7,300 kkg/year
          Maximum                                  62,000 kkg/year

     Average production                            22,100 kkg/year
     Median production                             15,800 kkg/year
     Average capacity utilization                      83 percent

     Plant age range:
          Minimum                                       7 years
          Maximum                                      58 years

     Waste water flow range;
          Minimum                                       0 cubic meters/day
          Maximum                                   4,700 cubic meters/day

     Volume per unit product:
          Minimum                _,                      0 cubic meters/kkg
          Maximum                                      86 cubic meters/kkg

Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Commerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.j and Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chemical Industry, "June, 1978 and "Economic Analysis of Proposed Revised
Effluent Guidelines and Standards for the Inorganic Chemicals Industry," March,  1980.
 * Data incomplete because certain plants did not respond to this question.
                                      276

-------
TflBLE 12-2  -
      OP BEGOLKTICNS
    GOIEEEilNES
SUBCAIEQOPY


SUBPAOT
Hvdrof luoric Acid

H  (40 CER  415.80, 3/12/74)
STANDARDS
Product
Process
Hydro-
fluoric
Acid
BPCTCA* ' BKE2A*
1 2 i •>
Max, Avg. Max.-1 Avg.''
Para- kg/kkg kg/kkg kg/kkg kg/kkg
rosters (rog/1) (mg/1) tog/l) Pig/i)
Fluoride (30) (15) » £^f^
NSPS*
Max. ! Avg.2
kg/kkd kg/kkg
ung/1) (rag/1)
, No discharge
of pwwp
TSS       (50)
                                 (25)
                         Ho discharge
                         of pwwp
No discharge,
of pwwp
  Sections 415.82, 415.83, and 415.85 were rananded and are presently
  reserved (41 FR 51601, November 23, 1976).
 "Max.  =» Maximum of any one day.
 2
  Avg.  — Average of daily values for thirty consecutive days shall not exceed.

  pwwp a* Process vastewater pollutants.
                                    277

-------
                        H20         "     0.1%
                        CaCOS     Principal remainder

     Silica is a  highly objectionable  contaminant, since each
pound  consumes  2.0  pounds  of  fluorspar  and  3.3  pounds  of
sulfuric acid by the following reaction:

         SiO2 + 2CaP2 + 2H2SO4 = SiF4 + 2CaSO4 + 2H2O       (2)

     Sulfuric acid having a concentration as  low  as  93 percent
or as high  as  99  percent is  generally  used.  Dilute sulfuric
acid enhances   better mixing  and  liberation of fluoride but
has  two   disadvantages;    viz,,   the    dilute  acid   is  very
corrosive  and  the water present in   the acid   evaporates and
distills   off   with  the  HP  gas,     thus   reducing   product
concentration.    Concentrated sulfuric   acid   (greater   than
98   percent)    offsets  these disadvantages  but  creates new
problems.  The  vapor pressure of concentrated   sulfuric acid is
sufficiently high   to  cause large amounts of  sulfuric acid to
be carried away by  the HF.  Excess sulfuric acid,  when  used,
will  leave with the gypsum as  part of the residue.

     HP generators are,  in the  majority of   cases, externally
fired  rotary   kilns   in which  acid  and  fluorspar are  fed
continuously through  a screw conveyor   at  the  forward end and
gypsum  is removed from the  other end through an air lock.   The
product HP  may  discharge   from  either  end.    The theoretical
amount  of  gypsum  produced  is  3.4 kg/kg of HP produced,   but
because of the  impurities  in the  fluorspar  the actual amount
°f gyPsum produced  is higher and  varies  from  3.6  to 4.8 kg/kg of
HP.

     One  manufacturer    uses   a patented  process  to  supply
internal  heat  to  the  reactor.    The  heat   is  supplied  by
introducing   sulfur trioxide  (SOS)  and  water  (as   steam).
The  exothermic  heat liberated by  the reaction of S03 and water
is used  for the heat required for HF generation.  Thus a part of
the  sulfuric acid  is supplied as 803.

     The HF gas  leaving the  reactor is cooled in a precooler  to
condense high   boiling  compounds.    The condensables are known
as drip acid and largely consist  of  fluorosulfonic acid  (HS03F)
and  unreacted sulfuric acid.   In 1978,  nine  plants out  of   a
total  of  eleven   returned  the    drip acid to   the  reactor,
while   the remaining  two   sent   the  drip  acid to the  waste
treatment  plant.   The  HP gas  from the precooler is  cooled
further  and  condensed  in  a cooler/refrigeration  unit.   The
uncondensed gas containing the HF  is scrubbed   with  sulfuric
acid and refrigerated  to  recover the  product.  The  scrubbed
acid liquor is  returned  to  the   kiln,  and residual vent gases
are  scrubbed  further  with  water  to  remove  HF   and    other
fluoride  compounds before   they are vented  to the  atmosphere.
The   scrubber   water is  sent  to the  waste water treatment

                              278

-------
plant.   Figure   12-1   is   a   block  flow   diagram  of  the
manufacturing process.

     The crude HF  is  then  distilled  to remove the  residual
impurities, and  "the  condensate,   which is   anhydrous HF, is
stored in tanks.  If  aqueous HF is desired, the crude product  is
then diluted with water to form  a 70 percent  HF solution as the
final product.


12.2  WATER USE AND WASTE SOURCE CHARACTERISTICS
12.2.1  Water Use

     Water is  used in HF production in noncontact cooling,  air
pollution   control,   product  dilution, seals   on  pumps and
kilns,   and   for  equipment  and  area   washdown.    Although
noncontact cooling  constitutes the  major  use  of  water, water
is also used, in a majority of cases, in the   transport of gypsum
as a  slurry   to  the waste   water treatment   facility.   The
water for  gypsum transport is provided by   either  reusing the
water from the   treatment   facility   or    by  using   once-
through cooling water.  Table  12-3   summarizes  the water  usage
found in this study.

12.2.2  Waste Sources

Gypsum Solids

     Gypsum   solids   are generated   as     a by-product.   The
amount  produced   is  in   the  range of 3.6  4.8 kg/kg    of  HF
produced.  The  gypsum  also  contains  small  amounts of sulfuric
acid, HF,  and     calcium fluoride.    Minor amounts  of other
impurities  present in  fluorspar are  also  removed  with   the
gypsum.  In  five  out  of eleven   plants producing HF,  gypsum
is slurried  with   treated waste    water, neutralized with lime
or soda ash, and pumped to a gypsum  storage   pond.  In one plant
the gypsum  slurry  is   pumped    to the   storage  pond without
treatment  and    in  another  plant   partial   neutralization  is
employed.  Three  plants transport the gypsum as  a dry solid and
dispose of  it  as  a  solid waste  after mixing with  lime  for
neutralization.   The  disposal  method   of  one plant    is not
known.  It  should  be  noted  that two of the   eleven plants have
recently  discontinued HF  production,  one of which is  in the
group of five.

     When gypsum  solids from the kiln  are  slurried with water
for  treatment,  the resulting stream    constitutes the  major
source  of waste water.   When kiln residue  is disposed of as a
solid waste,  scrubber   waste water is  the   major source  of
waste.  Table 12-4 gives  the data  for the  direct and indirect

                              279

-------
SLOBBY WATER
           TO WASTE
 1OXER'
                       SULFURIC
                         ACID
                      FLUORSPAR
                   SOLFATE  (GYPSUM)
                       SOLIDS
                                                 SULFONIC ACID
                   NONCONTACT
                   COOLING OR
                 REFRIGERATION
                     SYSTEM
                                     CRUDE HYDROGEN
                                                         SOLFURIC ACID
                          AQUEOUS HYDROGEN
                                             HYDROGEN FLUORIDE
                           FLUORIDE PRODUCT
                                                                        LEGEND
                                                                        COMMON PRACTICE
                                                                        INIERMICTSHT
                                                                        PROCESS (OR PRO
                                                                        CESS KS DULY
                                                                        SOME PLANTS)
 EJECTOR


 WSTSHMER

    to
TREATMENT
                            TO SALES
         Figure 12-1.   General process flow diagram  for production
                                of hydrofluoric acid.
                                           280

-------
  TABLE 12-3.  WATER USAGE IN THE HYDROFLUORIC ACID SUBCATEGORY
                               Water Usage at Plants
                                  {m3/kkg of HP)
                     (1)    (11                                        (1)
   Source         §987   #251   #753   #426   #120   #722   #167   #705
ton-contact       154    NA     63.5   110    NA     13.6   116     30.0
 Cooling

Gypsum Slurry     NA     64.0   NA     *      NA     22.5    41.6   30.0
 Transport

Maintenance,      NA      2.40   2.11  NA     0.1    12.2     5.00  16.9
 Equipment and
 Area Washdown

Air Pollution      7.90  14.4    4.23  NA     0.586  14.5     40.0  11,3
 Control
NA = Not Available

*  = Not Applicable

(1)  Discontinued HP production.-
                                  281

-------
      TABLE 12-4.        WATER FLOW AND            FOR THE
                  HYDROFLUORIC ACID SU8CATEGORY
I
Plant
1120
1426
(3)
1987
$837
1967
(3)
1251
(3)
1705
1167
1753
1928
1664
1722
Averages:
(1) D * DE
Ciln Residue
(D
Handling
D
D
D
5
S
S
S
S
S
S
S
S
(S only)
•y disposal
Reuse for Influent to
Kiln Residue Treatment
(2)
Slurry Facility
(Percent) (m3/kkg)
HF
(4)
— 9.10
(4)
— o
(4)
"**•*""' JL%MP* O
0 120
0 125
0 84.7
30.0-35.0 58.2
47.0 166
65.0 31.4
83.0 55.5
94.0 96.6
92.0-100 120
42.8 percent 95.4 m3Akg
S = Slurried to treatment
Treated
Effluent
Discharged
(m3/kkg)
HP
9.10
Not available
13.6
120
125
84.7
39.3
88.2
11,1
9.40
5.80
7.20
54.6 m3/kkg

(2>  Percent of waste water flow reused for residue slurry after
    treatment.
(3)  Dicontinued HF production.
(4)  Not Applicable.
                                   282

-------
process  contact waste  water  going  to   treatment facilities.
Noncontact cooling   water  has not been included in the figures
given  in  Table   12-4.    Figure  12-2     is   a    graphical
representation  of production   versus  waste water  flow  to in-
plant    treatment  facilities  for   plants  whose  waste  water
includes the  gypsum  slurry and  for those  "practicing disposal
of kiln residue as a solid waste.

Drip Acid

     This is  formed  in the  first stage of  the cooling  {i.e.,
in the precooler) of the gases   emitted from  the kiln.    Drip
acid   mostly contains   high   boiling  compounds consisting  of
complex fluorides,   especially  fluorosulfonic  acid,  and small
amounts    of   hydrofluoric   acid,   sulfuric acid,  and water.
Pluorosulfonic     acid     is  formed     by  reaction  between
hydrofluoric  acid   and  sulfuric  acid in the absence of water.
The quantity  of drip acid produced is relatively small.  In the
plants which  recycle the drip acid back  to the reactor,  it is
mixed with the  sulfuric acid feed stream before it enters the
kiln     where  it  is   hydrolyzed  to form   sulfuric acid and
hydrofluoric  acid.  The  critical factors  for hydrolysis  are
temperature   and  retention time  and   enough water is normally
present in the  kiln for the reaction.

Noncontact Cooling Water

     Noncontact  cooling  water   is  used   for  precooling  the
product gases emitted from  the kiln.  The  possibility  of product
or other  process compounds leaking  into  the cooling water is
very small?  however,  in the event that  the cooling water  does
become contaminated, the proposed limitations for  fluoride  may
be  exceeded.   Depending on   the merits of  the situation, the
upset   and bypass provisions may apply.   In some plants, the
cooling  water  is used  to  transport the waste gypsum.

Scrubber Haste Water

     Scrubber water  is    another waste water source,  and  in
plants whieh  practice   dry  disposal  of gypsum, scrubber water
constitutes the predominant and major  source of  waste  water.  It
contains  fluoride,  sulfate,   and  acidity.    The  fluoride   is
present as HP, silicon tetrafluoride (SiF4), and hexafluosilicic
acid (H2SiF6).  Silica  present in the  ore  as  an impurity reacts
with HP forming silicon tetrafluoride  as  shown  in Equation 3.

                   Si02  + 4HP  , »   SiF4 + 2H2O               (3)

     In  the  scrubber, the   tetrafluoride  is  converted  to
hexafluosilicic acid according to the  following equations:
                              283

-------
   15,000 +
   12,500 4
   10,000 -f

i
 g  7,500
 1
 CO
    5,000  -j-
   2,500  --
    2,000  -•

   1,000  --
                                               0
                                                   0

                          /
                                          Slurrying Kiln Waste	~-'	
                                                       +
                         75     100            150
                          HP Broduction, kkg/day
Figure 12-2.  Production versus waste flow data for HF plants,
                            284
                                                                      200

-------
                     SiP4 + 2HF  =  H2S1F6                  (4A)


                3S1F4 + 2H2O  »  2H2S1F6 + S1O2             (4B)


Distillation Wastes

     The   distillation waste generally contains  HF and water.
In some cases the vent  gases  from the distillation column are
scrubbed before  they  are  emitted   to  the  atmosphere,  and the
resulting scrubber water requires treatment.

     The  range  of   waste  water    quality of  the different
streams generated from  the production of HF  is  summarized in
Table 12-5.  The  data  are  taken from  the prior  development
documents, 308 Questionnaire responses, and industry visits.

Other Solid Wastes

     The  total  solids  generated from  the  process   and   the
treatment    system   consist of   gypsum    and  the  fluoride
precipitated as calcium fluoride.   Table 12-6 gives the amount
of  suspended solids   generated  from   the  process  and   the
quantity   of total  suspended   solids  generated at  the   waste
water treatment  plant for the  HF plants  visited in  screening
and verification.  The  data-  indicate  that the gypsum  waste
constitutes  more  than    95   percent   of  the  total  solids
produced.   Table 12-7   gives    the   amount   of  gypsum solids
produced   at different HF manufacturing facilities.   The data
shows that  3.8  to  4.7 kg gypsum solids  are  produced per kg of
product.


12.3  DESCRIPTION OF PLANTS VISITED AND SAMPLED


12.3.1  Screening

     Plant  $705  was   visited  and process waste  water samples
were collected  and   analyzed   for conventional, nonconventional
and  tonic  pollutants.     The   process  used  at this   site  is
similar to    the   conventional    HF manufacturing    process
described earlier.    The drip  acid is sent   to the  waste  water
treatment facility   and the gypsum   produced from  the reactor
is  slurried with  water   and    also  sent  to    the treatment
facility.   The  waste waters from the  HF   production facility
are  combined with  the aluminum  fluoride plant waste  waters.
The  combined raw waste water   is  treated with lime and   sent
to settling ponds before  discharge.   Figure   12-3 shows   the
general process   and   the locations  of the sampling   points.
Table 12-8  gives  the   flow data  and  the  total    suspended
solids  (TSS)  and  fluoride emissions.

                               285

-------
                          12-5.        ffiCW     HXDK3EKXBIC ACID MMfOEmCTORING ELSNTS
to
00
01

3
Flow in m /Kkg of %drofluoric Acid
Plants
Source of
Waste Water
Gypsum Slurry
Drip Acid
Scrubber
Waste Water
Other
#251 (1) #987 (1>
64.0 Dry
disposal
0.0490 0
14.4 8.30
0.530 0.530
#753 #426 §120
NA Dry Dry
disposal disposal
00 0
2.30 N& 0.624
8.40 NA 5.55
#722
(Total
Recycle)
0
(total
Recycle)
M
1167 #705 (1)
122 (Total
Recycle)
NA 0.0180
40.0 11.3
5.20 22.5
#837
6.50
0
1.12
NA
    (1)  Discontinued HP production


   m = Not Available


   *    Other does not include wasteflows from storm water runoff.

-------
   TABLE 12-6.  SOLID WASTE GENERATED AT THE HH3RQFLUORIC ACID FLAM'S SAMPLED
 Plant                   Gypsum Solids Going To      Total Solids Produced
                           Treatment Facility            (kg/kg of HF)
                          " -  (kg/kg of HP)


1705 ^                          4.73                        4.78

#251(1)                          3.81                         NA

#167                             3.94                         NA
(1)  Discontinued HF production.

NA = Not Available
                                     287

-------
 TKEIE '12-7.  GXPSUM SOLIDS PRODUCTION IN TEE HYDROFHJORIC ACID SDBCMESORY

Plant
1837
f705 <1)
*167
#722
#120
§426
%87 (D
1251 ^
*753
1967
f928
Kiln Residue Produced
(kgAg of HP)
3.86
4.73
3.94
NA
NA
4.00
4.13
3.81
m
NA
NA
Kiln Residue
Disposal/^Ireatroent Method
S
s
s
s
D
D
D
S
S
s
• s
S" * Slurried with •water and sent to wastewater treatment facility.
D » Dry disposal.
N& = Not available.
 (1)  Discontinued HP production.     288

-------
                                              vats
ro
oo
                       Waste streams sarpled.
               Figure 12-3.  General process flow diagram at plant #705 showing the sampling points.

                                         Hydrofluoric acid manufacture. •

-------
 TABLE 12-8.   FLOW AND POLLUTANT CONCENTRATION DATA OF THE SAT/PLED WASTE
               STREAMS OF PLANT #705 PRODUCING HYDROFLUORIC


Stream
No.
1
2

3

4

Sanpled
Stream
Description
Kiln Slurry
Scrubber Waste
Water
Surface Drains
Cooling Tower
Slowdown
Treated' Effluent
Screening
Flow
(m3/kkg of HP)
26.6
10.0

20.0

23.3(3>
DataC2>
Fluoride
(kg/kkg of HP)
15
9.6

6.9

1.6
Total
Suspended
Solids
(kg/kkg of HP)
4700
0.070

3.9
•
1.9
(1)   This plant has discontinued the production of HF since the time of
     sartpling.

(2)   One 72-hour coirposite sample of each waste water stream.-

(3)   The discharged effluent consists of the treated waste waters from
     hydrofluoric acid and aluminum fluoride plants.
                                    290

-------
12.3.2  Verification

     The same  streams at Plant  |705  were  sampled again in  the
verification phase.  The  variation in  the flow of  the  streams
in  the   two sampling phases was negligible.  Table 12-9 gives
the TSS and fluoride load summary of the sampled streams.

      Two more HF  plants (Plant  |251  and f!67)  were sampled in
the verification phase.   The  drip  acid at  both facilities is
also sent to  the waste  treatment   plant  and the  hydrofluoric
acid waste  waters   are combined    with   aluminum   fluoride
plant waste  for  treatment.    In  addition to  drip  acid, Plant
1251 waste  water consists   of  scrubber water, gypsum slurry,
and plant area  hose down.  The  treatment consists  of  gypsum
ponds where   the suspended  solids  are  removed.   The overflow
from the last  gypsum pond   is  neutralized and   the pH adjusted
with wastes  from other product lines.  Figure 12-4  is  a block
diagram of  the process showing the  sampling locations at Plant
1251.

     At Plant  #167,  the major  raw  waste sources are the kiln
waste  slurry,    the  absorber  tails from the condensate  {drip
acid)  recycle  system,  and  the ejector water which  is  used to
quench the  off-gases  from  the  absorber.   All  three of these
waste streams are collected  in a common neutralization pit where
lime slurry  is added.    The waste then  flows into  a series of
three lagoons for  solids removal and  final pH adjustment prior
to discharge.  Most  of   the gypsum  settles out   in the first
lagoon  and   the  overflow   enters   the  second  lagoon  where
commingling  with  wastes from  other  processes  takes  place.
Verification   sampling   data   from   this  plant were obtained
from  four  sampling  points.   These are:    1)  the kiln   waste
slurry, 2) the absorber tails,  3)  the ejector water,  and 4)  the
effluent from the  first  lagoon.  The  fourth sampling point  is
the  last  point  at which  all waste water originating  in the HF
plant can be intercepted.


12.3.3  Summary of the Toxic Pollutant Data


     Eleven  toxic  pollutants  were  found   in  the    raw  waste
samples from HF  Plant §705. They  were also verified at three
other typical HF  plants practicing BPT treatment.  No organic
toxic pollutants  were  found at detectable  levels.  The results
were:
                              291

-------
TAKEE 12-9.  FlOW AND PCMG3TANT (XftKEIOTR&TION DATA OF THE SAMPLED WASTE
             FOR PIANTS #705, 1251, AND 1167 PRODUCING HYDBDFLUORIC ACID

Plant
*705(2)





1251 (2)





#167






Stream
Kb.
1
2

4

5
5

6

2
3
1
2

3

4

Verification Data
Sampled 3 Flow
Stream Cm /kkg of HP)
Description
Kiln Slurry
Scrubber Waste
Water
Surface Drains
Cooling Tower
Slowdown
Treated Effluent
AHF Plant
Hbsedown
SO2 Scrubber
Waste
Gypsum Pond Inlet
Gypsum Pond
Outlet
Kiln Slurry
Ejector & Absorber
Unit Wastes from
Kilns #1,#2, and
14
Ejector & Absorber
'Unit Wastes from
Kilns 15 and §6
Effluent from
First lagoon
26.6
10.0

20.0

23.3
1.20

14.4

84.7
84.7
122
25.0

14.6

162

(1)
Fluoride
(kg/kkg of HF)
3.8
1.5

3.4

0.54
1.9

0.31

58
27
4.9
14

20

11

Total
Suspended
Solids
(kg/kkg of HF)
4700
0.019

4.0

0.040
0.26

0.10

3800
0.80
170
0.36

0.41

22

(1)  Three 24-hour composite samples of each waste water stream.

(2)  These plants have now discontinued their HF production.

(3)  Consists of the combined flow from hydrofluoric acid and aluminum
     fluoride plants.
                                     292

-------
K3
                       VENT
                       DUST
                    COLLECTION
         WET
         SPAR"
    SPAR
   DRYING
          HOSE DOWN
            WATER  '
                                 AIR
                             i/iiu
                                N
                            RESIDUE
                          HANDLING
                           LOSSES
  DRIP
  ACID
  AUO

WATER
                                                         A1F3 PRODUCT
                                                                                LIQUEFACTION
                                                                             •*-
    AHF
PURIFICATION
                                                     DILUTION WATER
              O
                       LEGEND
SAMPLING POINTS.
                                                                            15
                                                             I
                                                            44
                                                            ^^
                       HOSE DOWN WATER
                          AHF PLANT
                                                        #2
                            NEUTRALIZATION
                                SYSTEM
                                                                    ALKALINE STREAMS AND ««-
                                                                  ACID  FROM OTHER PLANTS
                                                                                                               WATER
                                                                                      EFFLUENT
              Figure 12-4.  General process flow diagram at Plant #251 showing the sampling points.
                            Hydrofluoric acid manufacture.

-------
                   Maximum Raw Waste Concentrations Observed
                                       (pg/1)
           Pollutant          Screening       Verification
                              Plant f705    Plants 1705,
                                                 *167
Copper
Lead
Selenium
line
Antimony
Arsenic
Cadmium
Chromium
Mercury
Nickel
Thallium
770
5200
25
8100
70
10
2
73
2
150
5






.0

.0

.5
600
200
230
13000
2800
160
60
1200
43
2000
63
     Section 5.1.2 of this  report  describes  the methodology of
the   screening  and  verification  sampling   program.  In  the
Hydrofluoric Acid  industry,   a  total  of  12   days  of sampling
were conducted   at  Plants   £705,  £251,   and  f!67.  Sixteen
different sampling points were  involved covering  the raw waste
source,   the  various  raw    waste   streams,  and  the  treated
effluents at these  plants.   The  evaluation of  toxic   metal
content of  these  process related waste  streams  was  based on
572 analytical  data points.   The  screening  for  toxic  organic
pollutants   at    Plant   f'705  generated an    additional  635
analytical   data  points.    The daily  raw  waste   loads  were
calculated  from  the  waste   stream    flow  rates  measured  or
estimated  at the  time of  sampling  and  the  measured pollutant
concentration.

     That is,

     Daily loading  (as kg of pollutant    (C)(Q)
     per day)                          = 1000

     Where:

     C is the concentration of the pollutant expressed in units
     of mg/1 (Note: kg/m3 = 1000 mg/1), and

     F  is  the  waste  stream  flow  rate expressed  in units  of
     m3/day.  (m3,  a  cubic   meter,   is   equal  to   264.2  U.S.
     gallons)

     Similarly, the  unit   loadings   were calculated from  the
reported hydrofluoric  acid production rate, the waste  stream
flow rate,  and the measured pollutant  concentration.

     Unit loading  (as  kg of pollutant     (C)  (Q)
     per kkg of hydrofluoric acid)    =  1000 fP)

                              294

-------
     Where C and F are the same as  described  above, and P is the
     hydrofluoric  acid  production rate  expressed in units  of
     kkg/day.   (kkg is 1000 kg, a metric ton, which is equal to
     2205 Ibs.)

     The minimum,  average,  and   maximum  values  are  based on
data from  those  plants  where the  particular  pollutant was
found  at concentrations greater  than the analytical detection
limits  and  significant   in   that it   could   conceivably  be
treated  by  an available    treatment    technology regardless
of economic considerations.

     In Table  12-10,  the  toxic  pollutant   raw waste data  are
presented  as  the  average  daily   concentrations and  the  unit
loading found  at  the individual plants.  The overall averages
are also shown  and  were  subsequently  used in the  calculations
of the  average daily  loadings and the  average   unit loadings
shown in Table  12-11  along with the  corresponding  minimum and
maximum values.

     Based on  the total annual production of this subcategory
and  the average  waste  load  generated   per  unit  product,  the
estimated total  toxic pollutant raw waste  loads generated each
year  for this subcategory are as follows:


                     Pollutant       Waste Load  (kg/year)

                     Copper                      6600
                     Lead                       10000
                     Selenium                     260
                     Zinc                      110000
                     Antimony                    8900
                     Arsenic                     1400
                     Cadmium     .                  79
                     Chromium                    4700
                     Mercury    ,                  130
                     Nickel                     10000
                     Thallium                     840
12.4  POLLUTION ABATEMENT OPTIONS


12.4.1  Toxic Pollutants of Concern
                                          i

     Toxic pollutants  in raw waste waters and slurries typical
of the  HF  industry  include  the   heavy  metals often  found
as  impurities  in  fluorspar.    These  metals are   zinc,  lead,
                              295

-------
          TABLE 12-10.  TOXIC POLLUTANT RAW WASTE DATA
SUBCATEGORY:  HYDROFLUORIC ACID
                                                                   (1)
Average Daily Pollutant Concentrations and Loadings at Plants Sampled
                              (mg/1)
(kg/kkg of Anhydrous HP)
#705 (S) #705 (V) 1251 (V)
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Thallium
Zinc
0.018
0.0010
0.051
0. 0029
0.0014
0.000080
0.062
0.0035
0.41
0.023
2.47
0.14
0.00090
0.000050
0.062
0.0035
0.0070
0. 00040
*
*
4.0
0.23
0.010
0.00057
*
*
0.0060
0.00034
0.26
0.015
0.26
0.015
0.044
0.0025
0.0053
0.00030
0.48
0.027
*
*
*
*
0.21
0.012
0.12
0.010
0.11
0.0091
*
*
0.47
0.040
0.12
0.010
0.059
0.0050
0.018
0.0015
1.18
0.10
. 0.017
0.0014
0.039
0.0033
0.28
0.024
I167CV)
0.74
0.12
0.028
0.0046
0.0030
0.00047
0.074
0,012
0.32
0.051
0.062
0.010
0.0010
0.00016
0.15
0.025
0. 0074
0.0012
0.019
0.0030
8.2
1.3
Overall
Average
0.22
0.033
0.062
0.0055
0.0035
0.00030
0.22
0.018
0.28
0.025
0.66
0.039
0.0060
0.00050
0.47
0.039
0.011
0.0010
0.029
0.0032
3.2
0.41
S - Screening data from one 72-hour composite  sample  of
    individual or combined raw waste streams.
V - Verification data  from three  24-hour composite  samples, averaged,
    from each raw waste sampling  point.
* - Concentration below significant level.
(1) The methodology of the sampling program  is described  in Section
    5.1.2, and Section 12.3.3  presents the scope of sampling  in  the
    Hydrofluoric Acid  industry.
                                  296

-------
      TABLE 12-11.  SUMMARY OF RAW WASTE LOADINGS FOUND  IN
               SCREENING AND VERIFICATION SAMPLING
SU8CATEGORY:  HYDROFLUORIC ACID
                     Daily      "                 Unit
                    Loadings                   Loadings         No. Of
                    (kg/day)                    (kg/kkg)         Plants
 Pollutant   Minimum Average Maximum   Minimum Average Maximum  Averaged*
Toxic
Antimony
Arsenic
Cadmium
Chromium
Copper
lead
Mercury
Nickel
Selenium
Thallium
Zinc
Conventional

0.023
0.01-2
0.0031
0.15
0.60
0.10
0.0021
0.14
0.016
0.16
0.49

2.0
0.50
0.014
1.7
1.4
1.8
0.057
4.1
0.093
0.31
21

6
1
0
5
2
5
0
14
0.
0.
72

.4
.2
.025
.4
.80
.4
.21

20
45



0.00057
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0"0030
000077
0035
0096
0025
000050
00035
00040
0030
012

0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.

034
0055
00030
018
025
039
00050
039
0010
0032
41

0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
1.

12
0090
00047
040
051
14
0015
10
0014
0033
3

4
3
. 3
' 4
4
4
4
4
3
2
4
& Nonconventional
TSS , 190000
Fluoride
13
310000
2900
520000
7900
3800
8.
8
4200
34

4800
58

3
4
* Only those plants where  the pollutant was observed at significant
  levels were  included.
                                   297

-------
nickel, antimony, chromium,  arsenic,  copper,  and selenium.  Raw
waste  waters    from plants   practicing  dry disposal  of kiln
wastes   may include   some   of   the same  heavy   metals   in
scrubber and  area washdown  wastes,  but  in considerably smaller
amounts, since   the spent ore  is  hauled  as  a  solid  waste and
bypasses the waste  water  treatment facilities.    Although  the
fluorosulfonate  anion  is found  in   HP  wastes  containing drip
acid, organic   compounds  are not   anticipated in waste  waters
from  this  industry.  No  toxic  organic pollutants were found at
significant levels.

12.4.2  Process Mod if i cat i ons and  Technology Transfer  Options

     1.  Gypsum  produced  in the  kiln can  be disposed of  as a
solid waste instead  of being slurried with  water  and  sent to
the waste  water  treatment facility.  The -solids  in  this  case
are mixed with  lime and stored in  piles  on  the land surface
until alternative    disposal methods   are     found   or   the
site abandoned.     Although  the  dry   disposal    method  is
labor intensive (involving transporation   and  landfill operating
cost) , it has  been found  to  be   less   expensive   due to  the
reduced  initial  capital   requirement  and    operating costs
relative  to the wet  slurry  method    which   requires  a more
extensive system of pipes, pumps,  and on-site impoundments.

     2.  The  use   of  soda  ash   in  place  of     lime  for
neutralization has some   advantages.    It   eliminates or reduces
the  problem of  scale formation in  the pipelines  and  scrubbers
when the  treated waste water  is recycled.  It offers a faster
reaction time  and better  control of pH than lime.  Even though
the cost of  soda ash is higher than lime,  soda ash has been found
overall to be a less expensive alternative  at some plants.  One
plant reported  that a combination  of  brine and  soda ash has
been found to  present the best  alternative  for   operation of
the recycle system  at minimum cost.   After  the   use of  soda
ash,  the treated  effluent water   can  be   totally  recycled,
either to the  scrubber  or to the  kiln for transportation water
for the gypsum.

     As  the pH  approaches  6,  sodium  in soda   ash replaces
calcium  present  in   the   gypsum   waste.    This  frees  enough
calcium ion to precipitate fluoride as calcium fluoride.  Where
the   scrubber water is  the predominant source of waste  water',
the  water  has  to be  treated  first    with  enough lime  to
precipitate fluoride  as  calcium    fluoride.    Soda ash  can
then   be   added   to    the supernatant to  precipitate calcium
followed    by   neutralization with HC1 to  reduce  scaling
problems.

     3.  Two   out  of  a    total  of 11  plants manufacturing
hydrofluoric  acid  send  the drip  acid   to  the  waste  water
treatment facility.  The  rest of   the plants  recycle  it to the

                              298

-------
reactor.   When discharged  to  the waste  treatment  system,  the
fluorosulfonic  acid does  not  hydrolyze  and  leaves with  the
treated effluent as a complex  fluoride in   soluble  form.  The ,
total   fluoride concentration of  the  effluent  will be higher
for  the plants  discharging drig acid  compared to  those which
do notr  after the  same  neutralization   treatment.     The  two
plants  discharging  drip   acid  to  waste  looked  into    the
feasibility of  returning  it to  the  kiln,   but because of  the
unique design of the   kilns,  they found  it to  be economically
unattractive.  Bench   scale studies   have  shown  that  the drip
acid can be hydrolyzed to free the HP.


               HSO3F + H2O + heat  =  H2S04 + HF            (5)

     The two  plants  not  returning  the  drip  acid to  the kiln
should be  able to   hydrolyze  the material in  a  separate unit
before  commingling  it with  other  wastes,  thus  avoiding the
treatability problem associated with complex fluorides.

12.4.3 Best Management"Practices

     1.  Runoff can be collected  from  raw  material  and product
storage, process, and impoundment  areas.   It should  be treated
with    other   process  waste    at  the waste  water treatment
facility.  Leachate and permeate control needs  to be practiced
on the  solid waste  stored in  many   plant premises as gypsum
piles.   There  is a risk   that  uncontrolled stockpiling  _may
contaminate the local ground, water.

     2.  Ponds designed for solids removal must be deep enough
to have a  minimum  of  disturbance from  wind and rain.   In those
areas where   the  rainfall  rate  exceeds yearly   evaporation,
the collection of  runoff  from   raw  material,  product  storage,
process, and impoundment areas may lead to serious water balance
problems.   Recycle ponds would  have to be  designed  to handle
this excess loading.

12.4.4  Prevailing Control and Treatment Practices

     Plant  §705   combines   the    hydrofluoric  acid  wastes,
including  the  gypsum  slurry,  with aluminum   fluoride waste.
The combined  waste  water,  after neutralization, is   sent to
settling lagoons before   discharge.  This plant was visited in
both the  screening and verification phases of the  project and
a fuller description of waste treatment practice is given below.

     Plant f837  combines  the  gypsum  slurry   and plant  area
hosedown waste water with  the  equipment  washings,  leaks,  and
spills  etc.  from  the  aluminum  fluoride plant and  neutralizes
them with  lime.   The solids are  removed  in   settling  ponds
before discharge.   The  waste    water from  scrubbers  of  both

                              299

-------
hydrofluoric acid  and  aluminum fluoride  plants ' is   sent  to
an  adjoining facility for use.

     Plant |251 also combines the hydrofluoric acid and aluminum
fluoride  waste  water.   The suspended  solids in  the combined
waste water   are removed  in  the  gypsum ponds.   The  overflow
from the  gypsum  ponds   is  neutralized and the pH adjusted with
the  waste water   from  other products which are  manufactured on
the site.  The plant  is   in the  process of installing a  new
proprietary treatment process to  further reduce  the fluoride in
its  waste waters.

     Two plants,  $120 and f987f dispose of the kiln residue as a
solid waste  after lime addition.  The  waste water in both cases
is treated with lime  and the solids are separated?  in one case
in a   clarifier  followed by  a filtration,  and   in  the other
by lagooning.

     At Plant  §167,  the combined  waste  water   (including the
gypsum) is  neutralized with  lime  and  then settled in  lagoons
before discharge.

     Plant f722  practices complete recycle.  The gypsum slurry,
scrubber   water,  and other   waste waters  are  combined  and
treated with  soda  ash  for  neutralization.    The neutralized
solution  is  settled in  lagoons  and then   is recycled  to the
scrubbers and  to the kiln to slurry the gypsum.

     Plant f426 disposes of the gypsum solids from the kiln as a
solid waste after lime addition.  The scrubber water is  used to
make   another   product.    The  noncontact cooling    water is
neutralized   when  required    with caustic  soda and   settled
before discharge.

12.4,5  Advanced Treatment Technologies

     Although alkaline   precipitation,  sulfide  precipitation,
the xanthate   process,  and ion  exchange might  be  applied to
clarified solutions  for  control of  metal ions, only alkaline
precipitation can be  readily used for slurri.ed kiln wastes from
HF production.   Sulfide precipitation  from  cleared solutions
could be   used   to provide additional  removal  of zinc,  lead,
nickel, and copper and to a lesser extent, antimony.
                              300

-------
12.5  SELECTION OF APPROPRIATE TECHNOLOGY AND EQUIPMENT


12.5.1  Technologies for Different Treatment Levels

Level 1 (BPT)

     Neutralization with lime, used widely in the  HP industry,
is   shown   as the BPT   treatment, principally  to control  pH
and   the   nonconventional   pollutant   fluoride   -   which  is
precipitated as calcium  fluoride.   Sufficient settled effluent
may  be reused    to  transport  kiln  waste  to   'the  treatment
facility  as  a slurry,  and  the remainder  is adjusted to  a pH
between 6  and   9  before  discharge.  The flow  diagram is shown
in Figure 12-5,

Level 2 (BAT)

     Treatment   is  alkaline precipitation, using   additional
lime     and     close control  of     pH       in second-stage
neutralization,  followed    by lagoon   settling.   Sufficient
lagoon effluent  is  reused  to   transport  kiln  waste to    the
treatment facility as a slurry  and  the   remainder JLS filtered
to remove finely divided metal hydroxides.   The  flow diagram is
shown in Figure 12-6.

Level 3

     It is assumed  that 65 percent of the Level 2 effluent   is
reused for   transporting   spent  kiln waste to   the  treatment
facility.   For   the  remaining  35 percent,  pH adjustment and
sulfide precipitation  are  used  ahead   of  the  Level   2 dual
media  filter,  to   react  with  residual  lead,  copper,  nickel,
zinc, and antimony which may not have   reached their optimum pH
levels for  alkaline precipitation.    The flow diagram for this
treatment is shown  in Figure 12-7.

Level 4

     As an   alternative to Level 2, Level  4  employs soda ash
instead   of lime for  neutralization,  depending  on   the spent
ore     to contain   enough  ' calcium  to   precipitate  calcium
fluoride.   Use   of     soda   ash permits  increased-effluent
recycling   without scaling problems    associated with calcium
sulfate.   To  control salinity and  sodium  alkalinity, a final
effluent blowdown of at least 10  percent  of  the  influent rate
is maintained.   The common heavy metals   will  be precipitated
as carbonates  and   hydroxides    with   varying degrees    of
effectiveness at  pH   levels attainable with  soda ash.     The
effluent is filtered and adjusted  to a  pH between 6 and 9 before
discharge or process recycling.   (Figure  12-8.)
                              301

-------
        UUS

       I—-fin
     .RAW	Ti
 WASTE WATER
                EQUALIZATION
                                              RECYCLE FOR SLURRY TRANSPORT   •*-
                                              MIXING
                                                                 LAGOON
                                                                  LAGOON
pH ADJUSTMENT
                                                                                         -»•»
        EFFLUENT
       	 	»•
Includes flow monitoring, pH monitoring and sampler
   Figure 12-5.   Level 1 waste water treatment for hydrofluoric acid sx±)category.

-------
                                                                BACKWASH
o
U)
r — nuw;
l^J— fi»-





HAW
WASTE WATER


r--@
i
i
i
f3
f
k i /,
KOUAUZATION I
l.-.torf-n 	 a.


" : 	 -~1
i
I
1 '
J u
,— L. -1* »X
MIXING ,


RECYCLE fOB
SUJRRY TBAHSPOBr*""
—Ji, LAOOON r-~\



•
-\ UODON /I*.










                                                                                .   l'»    I
                                                                                ADJUSTMENT,
                                                                       4_h

                                                                        SUMP Lrt_
                                                                                       FILTBO    *
                                                                                 1JSOENU




                                                                                 i Uuw niDRllitiriiig«
               Figxire 12-6.   Level 2 waste water •treatment  for hydrofluoric acid subcategory.

-------
*»
                WSIBtKtEK
                                       r
                                         —fun

0
                                    L
                                                                               SUKW TRANSPORT
                                                                                                     "1
                                                                                                M
                           Include* flow monltoriisg, pH manllaring aad
                    Figure 12-7. Level  3 v?aste water treatment for hydrofluoric acid subcategory.

-------
                                                 r
en
               SMI
_r**-






•
E
a

f dfi)




/
— , /
SQKHSKfSCH J











f

HKKK




I
I
li
**






,-A UOOCH /»,
n ™
r '
1

1

V ,
L— £^ LKCX30N £m^i*


atmur n«i3H
,t,

SUMP





3RT






•rfU—





r**



1

i



X

t

racsfi
&^U£
i



,
.1. i.Wt,.l

                             Includes flow mojtitoHiig, pH mQUitarmg a&d *
               Figure  12-8.  Level 4 waste water treatment for bycfrofluoric acid subcategory.

-------
Level 5 (NSPS)

     The proposed NSPS treatment  is  dry handling and off-site
chemical   landfill  for    the    kiln   waste  and   two-stage
alkaline  precipitation with  clarification  and filtration  for
the  liquid process wastes.    Heavy  metal precipitation  with
soda ash  permits partial  recycling  for  uses other than slurry
transport.   (Figure 12-9.)

12.5.2  Equipment for Different Treatment Levels

Equipment Functions

     Level  lr   typical  of   existing  treatment    facilities,
utilizes  very   little   equipment,   but  depends    on    lime
neutralization in settling lagoons, with final pH adjustment.

     In Level 2,   conventional dual  media filtration is added to
the  Level 1  system.   In  Level  3,  standard reagent mixing and
solution  feeding units  are  added  to the Level 2 system.   In
Level  4,  which is an alternate to Level 2,  the same  type of
chemical   feed  equipment is used for soda  ash  as was used for
lime in Level  2.   Conventional lagoons and dual media filters
are used  in Level  4,   but  special  attention  to selection of
materials   is   required   because  of   the  high   salinity of
recycled effluent.  In the  NSPS  model,  dry  kiln waste disposal
is  recommended  with    conventional    dry  solids  handling
equipment.   Lagoons,  clarifiers, and filters   are    used for
scrubber,   noncontact   cooling,    and    other  miscellaneous
liquid  wastes.    In  this  case,  equipment  for  storing    and
handling  the  dry  kiln  waste  is not considered  to   be waste
water  treatment,  and  the cost is   not  included  in the cost
estimates.

Chemical Handling

     Lime  (as CaO) is the major chemical used   in Levels  1 and
2, along with minor  amounts of hydrochloric acid  for   final  pH
adjustment.    With   normal  precautions, these   chemicals pose
no special hazards.   In  Level 3,  ferrous sulfide  is  prepared
on-site   by  mixing  sodium bisulfide   and   ferrous   sulfate.
Although  sodium  bisulfide  can release toxic  H2S at pH  levels
below    7,  the  hazard  can    be  mitigated  by  avoiding  acid
conditions and by providing adequate ventilation.  After  mixing
its components,  the  ferrous  sulfide solution, is  stable at the
pH  levels employed in  the process.   In  Level 4,  only  sodium
carbonate  and hydrochloric   acid    are  used,  without  unusual
safety hazards or special handling problems.   In the NSPS system
only  lime,   soda  ash   and   hydrochloric  acid    are    used,
introducing no special problems of safety or handling.
                              306

-------
                                                                     ll ADJUSTMENT
                                                                                  HECYCI.B *O
                                                                                   SCRtfUBER
   {3-iUO OfASTEJ
        Include* flow munltorlng, pH motiltartog aiul B
Figure 12-9.   Waste water treatment new source performance standard  for
                hydrofluoric  acid subcategory.

-------
Separation and Removal of Solids

     Solids are accumulated in  unlined  settling   lagoons.   In
Level  4,  calcium  fluoride will    still precipitate    in  the
lagoons but the total sludge quantities  will be  less  than in
fievels  If 2, and 3  where lime is  used.   Solids from Level 4
treatment will   be  alkaline,  very,  saline,  and   difficult to
consolidate.   Dry  solids from the Level 5  (NSPS)   model  are
not  subjected  to treatment,  except   for  nominal  application,
of  lime before  hauling   in dry  form to an approved  chemical
landfill.
12.6  TBEATMEHT COST ESTIMATES


12.6.1  General Discussion

     To  prepare   treatment  cost  estimates,  a   model  plant
concept    was  developed.   The  proposed  BPT model  treatment
consists of:

     A.  Slurry transportation of kiln solids to an equalization
         basin.

     B.  Application of  lime  to  precipitate fluoride and toxic
         metals, followed by lagoon settling,

     C.  pH adjustment before final discharge.

     D.  Scrubber,  cooling,  and  distillation  wastes  enter  the
         equalization basin.

     It  is assumed  that   drip  acid is  recycled to the process
reactor  and does not appear directly in the waste stream.

     For new  or remodeled production facilities,  the NSPS model
treatment  system is based  on  hauling  dry kiln residue directly
to  a  landfill. Miscellaneous liquid wastes  in the  NSPS model
are     subjected     to     two      stage      lime-soda     ash
neutralization/precipitation,   followed   by  filtration   and
partial  return of effluent for use in scrubbers.

Haste Water Flow

     The   data   in  Table 12-4   for  plants sending the gypsum
solids to  the  treatment  facility indicate that   the  unit flow
varies from  approximately 31.0 m3/kkg of HP  to  166   m3/kkg of
HF.  For the model plants, a constant unit flow of 95.4  m3/kkg
of HF was  assumed.
                              308

-------
HP Production

     In   the HP subcategory,  production  ranges from a minimum
of 7,300 kkg/year to a maximum of 62,000 kkg/year  with a mean of
22,100   kkg/year and a  median of 15,800  kkg/year.   For waste
water  treatment cost estimates,  three production  levels were
selected  as  model plants.   These are  19,100  kkg/year,   38,200
kkg/year, and 57,300 kkg/year.

Waste Water Pollutant Load

     The amount of kiln residue varies from 3.8  to 4.1 kg/kg of
HP produced.  The waste  water  going  to treatment model  plants
is assumed to contain 3.8  kg of solid  kiln  residue per kg of
HF.  Fluoride  emissions in  waste water have been shown to vary
as indicated below:
      Source of Data                 Fluoride,  (kg/kkg)

      Reference 3                            20

      Reference 3                            37

      Screening and Verification          3.8 to 58
      Phase Sampling
       (Tables 12-8 and 12-9)

     For  the  model plants,  the average  fluoride  loading from
kiln   wastes  of   31  kg/kkg HF produced  was  used  to establish
treatment  requirements and related costs.

     The costs  shown  at  each level of  treatment correspond to
the    model  plant  BPT   system  (Level  1}   and  one  or  more
alternative BAT  systems  (Level  2, 3',  and  4)  which  may add
to  or  modify the  existing  BPT  system  to meet more stringent
priority pollutant  removal  requirements.  The BAT system also
provides   a   higher   effluent  water  quality with  respect to
the  conventional  and  nonconventional parameters.

     At  each    level    of    treatment,   the   cost  elements
associated with  the  typical rates of effluent  reuse  have also
been  included.   However,   the   hydraulic   loading   on  the
treatment  system  is  unaffected by reuse, and,  therefore  the
total costs (including  reuse) are independent of the particular
rate of reuse  that may  be practiced.

     The estimated   costs ,for  three  models   having different
production levels  are given   in Tables 12-12, 12-13, and  12-
14.    For these  models,  both the hydraulic  and "the pollution
loads  per   unit   of   production   are held  constant  over the
entire  range   of   production.   Annual  treatment cost  as  a

                              309

-------
                    TABLE 12-12.   MODEL PLANT TREATMENT COSTS
Subcategory HYDROFLUORIC
Production 19,100 n
54 n
Waste water flow 5220 <
A. INVESTMENT COST
Construction .........
Equipment in place,
including piping,
fittings, electrical

Monitoring equipment..
Engineering design
Incidentals, overhead,
fees, contingencies...

TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.

Maintenance. ..........

Taxes and insurance...
Residual waste disposal
Monitoring, analysis

TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
ACID
letric tons per
netric tons per
ruble meters pe
FIRST
$877,500
356,000
30,000
9,000
254,500
254,500
1,020,000

$2,801,500
$56,000
14,000
534,800
172,650
6,500
84,045
350,000
15,000

1,232,995
$289,850
$1,522,845**
(1)
year (21
day (60
r day.
LEVEL OF
SECOND
$24,500
89,500

22,800
22,800

$159,600
$14,000
1,500

15,960

4,788
7,500

$43,748
$25,966
$69,714
,057 tons pe
tons per da
TREATMENT*
THIRD
»
$25,000
92,000

23,400
23,400

$163,800
$14,000
1,800
3,400
16,380

4,914
7,500

$47,994
$26,650
$74,644
(1)
r year)
y)
FOURTH
$24,500
89,500

22,800
22,800

$159,600
$14,000 '
1,500
367,700
15,960

4,788
7,500

$411,448
$25,966
$437,414
    *First level represents the base cost of treatment system.
    Other levels represent the incremental cost above base cost.

  **!neluding $11,100 for the reuse of treated effluent to slurry
    kiln residues, etc.
(1)  Production year  is 350 days.
                    310

-------
                    TABLE 12-13.   MODEL PLANT TREATMENT COSTS
Subcategory HYDROFLUORIC ACID
(1) (1)
Production 38,200 metric tons per year (42,115 tons per year)
109 metric tons per day (120 tons per day)
Waste water flow 10450 cubic meters per day.
A. INVESTMENT COST
Equipment in place,
including piping,
fittings, electrical
Reuse facilities......
Monitoring equipment..
Engineering design
Incidentals, overhead,
fees, contingencies...
TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.



Taxes and insurance...
Residual waste disposal.
Monitoring, analysis
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST
LEVEL OF TREATMENT*
FIRST SECOND THIRD FOURTH
$1,354,500 $35,000 $35,500 $35,000
493,500 131,000 137,500 131,000
40,000
9,000
379,400 33,200 34,600 33,200
379,400 33,200 34,600 33,200
1,944,000
$4,599,800 $232,400
$56,000 $14,000
19,500 3,100
1,069,600
257,580 23,240
10,000
137,994 6,972
700,000
15,000 7,500
2,265,674 $54,812
$432,098 $37,811
$2,697,772** $92,623
$242,200 $232,400
$14,000 $14,000
3,400 3,100
6,700 735,350
24,220 23,240
7,266 6,972
7,500 7,500
$63,086 $790,162
$39,405 $37,811
$102,491 $827', 973
    *First  level  represents the base cost of treatment system.
    Other levels  represent the incremental cost above base cost.

  **Includes $16,500 for the reuse of treated effluent to slurry
    kiln residues, etc.
(1)  Production year is 350 days.
                     311

-------
                     TABLE 12-14.   MODEL PLANT TREATMENT COSTS
   Subcategory  HXDRQffLUORIC AGED

                                              (1)                    (1)
   Production        57,300 metric tons per year   (63,173 tons per year)
                        163 metric tons per day    (180 tons per day)
   Waste water flow   15700 cubic meters per day.
A.  INVESTMENT COST

    Construction 	
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls.....
    Reuse facilities	
    Monitoring equipment..
    Engineering design
    and inspection.	
    Incidentals, overhead,
    fees, contingencies...
    Land	
                                             LEVEL OF TREATMENT*

                                   FIRST      SECOND       THIRD


                              $1,755,500     $49,000     $50,000
                                 848,000
                                  50,000
                                   9,000
203,500     215,500
 FOURTH


$49,000



203,500
532,500
532,500
2,880,000
50,500
50,500
53,100
53,100
50,500
50,500
B.
    TOTAL INVESTMENT COST

    OPERATION AM)
    MAINTENANCE COST
                              $6,607,500    $353,500    $371,700    $353,500
Labor and supervision.




Taxes and insurance...
Residual waste disposal.
Monitoring, analysis


TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION Of
INVESTMENT COST
TOTAL ANNUAL COST
$56,000
28,000
1,604,400
362,350
13,000
198,225
1,050,000

15,000


3,326,975

$606,464
$3,933,439**
$14,000
4,600

35,350

10,605


7,500


$72,055

$37,514
$129,569
$14,000
4,900
10,070
37,170

11,151


7,500


$84,791

$60,475
$145,266
$14,000
4,600
1,103,025
35,350

10', 605


7,50i


$1,175,080

$57,514
$1,232,594
    *FIrst level represents the base cost of treatment system.
    Other levels represent the incremental cost above base cost.

  **Includes $21,200 for the reuse of treated effluent to slurry
    kiln residues, etc.
 (1) Production year is 350 days.
                      312

-------
function  of  production  is shown graphically in Figure 12-10.
Similarly, treatment  cost   per  metric ton of product is given
in Figure 12-11.

     To indicate the effect on costs  of  an increased pollution
load per unit of production  for the 38,200  kkg/year model plant,
the  pollution  load  was  increased  by 100   percent and  the
hydraulic load was held  constant. The  cost estimate indicated
that the annual unit cost per metric ton of product at first"  and
fourth    (incremental)    levels    of    treatment    increased
approximately 40 percent and  90   percent  respectively  over the
original  model unit cost.  The increased  cost is  due mainly to
the additional cost of chemicals. Increase of pollutant loading
had  no effect on the unit cost   of treatment at other levels of
treatment.

     Similarly, for  the same model  plant, the  hydraulic load
was increased  by  100   percent   and  the  pollutant load  was
held constant.   The  cost  estimate  indicated  that  the annual
unit cost per metric ton of  product   at  the  second  and fourth
levels  of treatment  increased approximatley   70 percent and 10
percent respectively over the original model unit  cost.  There
was no  significant impact   on the unit cost  at other levels of
treatment.

     Table   12-15   presents  a   summary  of  the   unit  cost
distribution between amortization and operation and maintenance
cost   components  at various  production  rates  and  levels  of
treatment.

     At the second,  third  and  fourth  levels of treatment, the
cost estimates are  based on part  of the waste water  flow  being
recirculated  and  the  remaining flow being treated,  thus  the
subsequent  treatment units are   sized and estimated  for  lower
flows than if recycling  were not practiced,

12.6.2  Model Plant Control Costs for  Existing Sources

     For the  model plant  control costs  for existing sources at
the  first -level of treatment,    the  disposal  of the sludge is
on-site  and  hence  the land  requirements are  fairly  large.
Chemicals,   sludge  hauling,   and .disposal   costs   have     a
significant  impact on   the total annual  costs.  At   the  second
and  third levels of treatment however, amortization, labor and
supervision costs constitute a  major portion of the additional
annual costs.

     The fourth level of treatment is  designed for recirculation
of the major portion of  the  treated effluent and therefore, soda
ash is used  for   neutralization  in place  of lime.  Due to this
change, chemic.al cost has a  significant impact on the additional
annu al cos ts.

                              313

-------
  a
  o
  o
   ^
  o
  o
  o
                         •9
                                 <4
                                 Cd2
                                                     14
                                                      x
                                                        X
                                             •BV:SL
Figure 12-10.
 10       20       30        40       50       60


    PRODUCTION (METRIC TONS/YESR X 1000)



Jtonual treatinent cost vs.  production for the Hydrofluoric

            Acid Subcatecpry
                                314

-------
   110
   100
    90
   80
    70
    60
              10       20       30       40        50

                  EEODDCTICN (METRIC 1OTS/YEAR X 1000)


Figure 12-11.  jtonual unit treatment cost vs. production for the
                 Hydrofluoric Acid Subcategory
                              315

-------
              TABLE 12-15.         PLANT           COSTS
Subcategory  HYDROFLUORIC ACID
                                   Annual Treatment Costs  ($/kkg) of HP
COST ITEMS
PRODUCTION  FtOW
 (kkg/yr)   (m3/day)
  LEVEL OF TREA1MENT

FIRST      SECOND    THIRD   FOURTH
Annual Operation
and Maintenance
Annual
Amortization
Total Cost
   19,100   5,220     64.55      2.29      2.51
   38,200  10,450     59.31      1.43      1.65
   57,300  15,700     58.06      1.26      1.48
   19,100   5,220     15.18      1.36      1.40
   38,200  10,450     11.31      0.99      1.03
   57,300  15,700     10.58      1.00      1.06

   19,100   5,220     79.73      3.65      3.91
   38,200  10,450     70.62      2.42      2.68
   57,300  15,700     68.65      2.26      2.54
                              21.54
                              20.68
                              20.51
                               1.36
                               0.99
                               1.00
                              22.90
                              21.67
                              21.51
                                316

-------
12.6.3  Model Plant Control Costs for New Sources

     The basis of the selection of the model plant representing
a  new source   is  described  earlier  in   this  section.   The
estimated  costs  for   three   different  models,  having three
different  production levels are  given   in Tables 12-16,  12-17,
and 12-18.   Both the hydraulic and pollutant loads are directly
proportional to  the  production,  i.e., -the   waste flow per unit
of production and the pollutant  loading  per  unit of production
are held constant.

     Annual treatment cost as a function of  production is shown
graphically  in Figure 12-12.  Treatment cost per metric ton of
product is given in Figure 12-13.

     Table   12-19   presents  a   summary  of  the   unit  cost
distribution between amortization and operation and maintenance
components.

     For the model  plant, the dry solids generated in the kiln
are  hauled  to approved chemical dump   sites, eliminating kiln
waste slurry.  The waste water sources  are air pollution control
(scrubbers), leak, spills, and washdowns.

     The cost  of  transporting  dry kiln  waste 'sludge to  the
approved  chemical  dump  site has been  included in  the  cost
estimates.  The cost of conveying the   dry solids  from the kiln
operation to the trucks  (for transporting to the  dump site)  is
not included  in  the cost  estimate.  Such costs,  which can vary
widely   with site  conditions,  are considered  to be   process
costs  and  not part  of  treatment.  However, if  such costs are
to  be  considered  as  part of the treatment costs,  then  the
estimated  total  annual costs per metric ton of product for the
three model  plants would ,be as follows:

      Production          Flow          Total Annual Cost
      (kkg/year)         (m3/day)              ($/kkg)

       19,100             680                 14.81
       38,200           1,370                  9.68
       57,300           2,030                  8.03


     Since the  sludge disposal   is not  on site,  the land cost
has  negligible impact on total annual  cost.   However, the cost
of  transporting the dry  solids   to the dump  site  constitutes
about 75  percent of the annual costs.
                              317

-------
                     TABLE 12-16.
               MODEL PLANT TREATMENT COSTS
   Subcategory  HYDROFLUORIC ACID
   Production
   Waste water flow
                        (1)
19,100 metric tons per year
    54 metric tons per day
   680 cubic meters per day.
                  (1)
(21,057 tons per year)
(60  tons per day)
A.  INVESTMENT COST

    Construction	
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls	
    Monitoring equipment
    in place	
    Engineering design
    and inspection	
    Incidentals, overhead,
    fees, contingencies...
    Land	

    TOTAL INVESTMENT COST

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision.
    Energy	
    Chemicals.	
    Maintenance	
    Taxes and insurance...
    Residual waste
    disposal	
    Monitoring, analysis
    and reporting	

    TOTAL OPERATION AND
    MAINTENANCE COST

C.  AMORTIZATION OF
    INVESTMENT COST

    TOTAL ANNUAL COST
                          LEVEL OF TREATMENT*

                                 FIRST


                               $64,000



                               327,000

                                 9,000

                                80,000

                                80,000
                                30,000

                              $590,000
                               $56,000
                                 6,100
                                44,000
                                56,000
                                17,700

                               742,000

                                15,000
                              $936,800


                               $91,112
                            $1,027,912
    *First level represents the base cost of treatment system.
    Other levels represent the incremental cost above base cost.
 (1)  Production year is 350 days.
                                    318

-------
                     TABLE 12-17.    MODEL PLANT TREATMENT COSTS
   Subcategory  HYDROFLUORIC ACID

                                             (1)                        CD
   Production        38,200 metric tons per year    (42,115 tons per year)
                        109 metric tons per day    (120 tons per day)
   Waste water flow    1370 cubic meters per day.
                                               LEVEL OF TREATMENT*

                                                      FIRST
    INVESTMENT COST
    Construction 	                          $94,500
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls	                          468,500
    Monitoring equipment
    in place....	                            9,000
    Engineering design
    and inspection...	                          114,400
    Incidentals, overhead,
    fees, contingencies...                          114,400
    Land	                           60,000
    TOTAL INVESTMENT COST                          $860,800

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision.                          $56,000
    Energy	                            8,300
    Chemicals	                           88,000
    Maintenance	                         •  80,080
    Taxes and insurance...                           25,824
    Residual waste
    disposal..	                        1,480,000
    Monitoring, analysis
    and reporting	                           15,000
    TOTAL OPERATION AND
    MAINTENANCE COST          "                   $1,753,204

C.  AMORTIZATION OF
    INVESTMENT COST                                $130,290
    TOTAL ANNUAL COST                            $1,883,494
    *First level represents the base cost of treatment system.
    Other levels represent the incremental cost above base cost.
 (1) Production year is 350 days.
                                   319

-------
                     TABLE 12-18.   MODEL PLANT TREATMENT COSTS
   Subcategory  HYDROFLUORIC ACID

                                             (1)                       (1)
   Production        57,300 metric tons per year   (63,173 tons per  year)
                        163 metric tons per day    (180 tons per day)
   Waste water flow    2030 cubic meters per day.
                                               LEVEL OF TREATMENT*

                                                      FIRST
    INVESTMENT COST
    Construction 	                         $120,700
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls	                          601,000
    Monitoring equipment
    in place	                            9,000
    Engineering design
    and inspection	                          146,140
    Incidentals, overhead,
    fees, contingencies...                          146,140
    Land	                           84,000

    TOTAL INVESTMENT COST                        $1,106,980

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision.                          $56,000
    Energy	                           12,250
    Chemicals	                          132,000
    Maintenance	                          102,298
    Taxes and insurance...                           33,209
    Residual waste
    disposal	                        2,226,000
    Monitoring, analysis
    and reporting	                           15,000
    TOTAL OPERATION AND
    MAINTENANCE COST                             $2,576,757

C.  AMORTIZATION OF
    INVESTMENT COST                                $166,438
    TOTAL ANNUAL COST                            $2,743,195
    *First level represents the base cost of treatment system.
    Other levels represent the incremental cost above base cost.
 (1) Production year is 350 days.
                                   320

-------
o

«   2
O
O
o
w-
             10        20       30        40        50       60

                 PIODUCTICN CMETOIC TCNS/YEAR x 1000)


    Figure 12-12.  Annual treatment cost vs. production for the

                Hydrofluoric Acid Subcategory
                            321

-------
60
50
                                                LIE''
                                          'EL
40
30
                                J_
10        20        30        40       50

     PKCOJCTION CMETKCC TOTS/YEAR X 1000)
                                                       60
   Figure 12-13.  Annual unit treatment cost vs. production for
            the Hydrofluoric Acid Subcategory  (NSPS)
                         322

-------
                 TABLE 12-19.   MODEL PLANT TREATMENT COSTS
   Subcategory  HYDROFLUORIC ACID
                                      Annual Treatment Costs ($/kkg) of HF
   COST ITEM
                       LEVEL OF TREATMENT*

PRODUCTION   FLOW    FIRST   SECOND   THIRD   FOURTH
(kkg/yr)    (m3/day)
Annual Operation
and Maintenance


Annual
Amortization


Total Cost


19,100
38,200
57,300

19,100
38,200
57,300
19,100
38,200
57,300
680
1,370
2,030

680
1,370
2,030
680
1,370
2,030
                                          49.05
                                          45.90
                                          44.97
                                           4.77
                                           3.41
                                           2.90

                                          53.82
                                          49.31
                                          47.87
                                Not Applicable
* Only applies to first level,
                                  323

-------
12.7  BASIS FOR REGULATIONS


12.7.1  Evaluation of BPT Treatment Practices

     Control  and   treatment     practices  for  eleven  plants
producing  HF  are   presented in Table  12-20.  Also  indicated
are  other  product-related waste water  sources   and pollutant
loads  discharged.

     It  is  clear  from  the  table  that  a  wide  variation  in
effluent   quality  exists  within this subcategory.   The factors
believed   to  cause  these variations are  the  following:

Dry Residue Handling

      The  disposal  of  kiln waste   by   dry handling  rather than
slurrying  is practiced  currently at  three plants.  This process
eliminates   the major source of waste water generated at   most
plants, greatly  reducing  the raw  waste loads  to   be treated.
The only sources of waste water remaining  are from  air pollution
control and washdown,

Effluent Reuse

     Reuse of treated waste water for  slurry  transport of kiln
wastes is  commonly practiced  to varying degrees and  clearly has
a major  effect on  pollutant  loads  discharged.   Although four
plants  do not  practice  reuse,   it   has  been  demonstrated
sufficiently  that  this  practice   is   both  technologically and
economically  feasible.

Recycle of Condensables

     Recycling  of  drip  acid  or   condensable cooler  bottoms
reduces  the  loading   of  fluoride  in the  treated  effluent
since   the  fluoride   species   (fluorosulfonic acid)   in this
material is not   removed  by conventional  lime treatment.   Only
two  plants do  not  recycle  drip  acid.

Other Related Products

     Most  hydrofluoric  acid plants  also  discharge  wastes from
related  products  such   as   aluminum   fluoride, fluorocarbons,
hexafluorosilicic  and  tetrafluoroboric  acids  to  treatment.
These  other product wastes  can  account  for higher  raw waste
loadings   and  increase  the  potential  for  complex   fluorides
formation  and  can also impact treatment efficiency  by diluting
the  raw   waste.    In  addition,  commingling of other  product
wastes will  limit  the  percentage of reuse  of the total plant
treated effluent.
                              324

-------
         , TKBCB 12-20.  SUMMARY  OS WASfE SATES QUaStUi HO} TREATMENT radWOKXBf EMPLOYED AT HYDROFLUORIC ACID PUNTS(1)
Plant
* 426
i 664
* 167
1 120
Product-Related
Haste Water Sources
Hydrofluoric acid
fluosilicic acids
production
Hydrofluoric acid
production
Hydrofluoric acid,
fluorocarbon,
Color tn*/sodiu»
hydroxide, and
hydrochloric acid
production
Hydrofluoric acid
production
Control and Treatment
Technology Employed
Dry residua hauling
and dumping; neutra-
lization with caustic
of noncontact cooling
water and floor
drainage
Residue slurry, neutra-
lization with sodium
carbonate, settling,
recycle
Residue slurry, lime
treatment , settling ,
recycle
planned dry residua
handling, lime
Amount of Cooler Bottoms Effluent volume
Treated (Condenaables) in m3 /metric ton
Waste Water Recycled? (gal/short ton) of
Reused Actual Production
0 Tea 465 (111,397)
includes noncon-
tact cooling
water
94% Yes 5.78 (1,360)
47% Ye» 103 (24,200)
0 Yes m
Long Term
Average Pollutant
wasteload Discharged
(kg/metric ton)
(lb/1000 Ibj
Fluoride »SS
1.2 HD
0.10 0.27
18 0.4S (Net)
ND NO
                                                                                                                                <3
t 967    Hydrofluoric acid,
         fluorocarbon, and
         sulfuric acid
         production
treatment, clarification

Residue slurry, settling Present:  0
(Recycle and pH          'planned:  70%
polishing facilities            to 75%
under construction.)
* 928    Hydrofluoric acid
         and aluminum
         fluoride production
I 837    All hydrofluoric
         acid genorated as
         used captivoly for
         aluminum fluoride
         production

f 753    Hydrofluoric acid
         production
Residue slurry, settling,   83%
recycle (Flocculation,
lime treatment, and
clarification facilities
under construction.)

Residue slurry, line         0
treatment, settling
Residue slurry, lime        65%
treatment, settling,
recycle, pH polishing
                                                                         Yes
                                                                         Yes
                                                                         Ye»
                                            Yes
125 (30,000)   Present:  24      16
               Expected
               with       1.8     2.1
               additional
               facilities

9.44 (2,260)   Present!   1       1.7
               Sxpected
               with       0.65    0.75
               additional
               facilities
                                                                                    134  (32,200)
11.0 (2,650)
                                                                                                              1.8      3.1
                                                                                 0.64     0.38
t 251 (2!
* 70S<2»
* 722
* 987 (2>
HF, AlFt, chlorine/ Residue slurry, settling, 0
sodium hydroxide, neutralization
aluminum oxide, and
fluorocarbon
production
Hydrofluoric acid Residue slurry, lime 30% to 35%
and aluminum treatment, settling,
fluoride production recycle, pH polishing
Hydrofluoric and, Residue slurry, lime 92% to 100%
in recent past, treatment, settling,
fluoboric, recycle, pH polishing
acid production
Hydrofluoric! Acid Dry residue hauling 0
1 Kiln: Yes 22,2 x 10* 46 530
3 Ktlnss Ho (553 x 10*5
So 25.9 (8,204) 3.2 0.64
Yes 0-10.3 (0-2,460) 0-0.81 0 to 0.54
Yes 8.8 ND ND
(1)  Adapted from Calapan (Reference 3).
(2)  Hydrofluoric Acid production has been discontinued at these plants since the  time of sampling.
(3!  affluent loading less the influent loading.
HD » Hot determined,

-------
     In addition to the above  factors,  the design and operation
of  the  treatment  facilities affect  effluent quality.   Solids
removal   depends   on   retention   time  and   surge  capacity.
Precipitation of fluoride  requires careful pH  control  and in
areas of heavy rainfall or winds,  adequate freeboard or multiple
ponds  are  necessary to limit the discharge   of  high pollutant
loads due to unfavorable climatic conditions.

Pollutant Removal with BPT Treatment

     Treatment level 1 is equivalent to the  proposed BPT in the
Hydrofluoric Acid industry.  Table 12-21  presents a summary of
long  term  effluent monitoring data  on total suspended solids
(TSS)  and   fluoride  from  four  plants.     Means,   standard
deviations, and variability  factors  are given where sufficient
data are available.  These performance characteristics are later
utilized for the development of the proposed  regulations on TSS
and fluoride.

     The  ability  of BPT treatment   to remove toxic pollutants
can  be estimated  by comparing the raw waste data presented in
Table  12-10  with  the  corresponding  treated  effluent  data
presented  in  Table 12-22.   The latter expresses   the  removal
efficiency  as  the  calculated average percent removal observed
at these  plants.   The  BPT  removal efficiency for  some  of  the
toxic  metals  is undoubtedly  augmented to some   degree  by the
fact that the raw waste may carry insoluble  forms of the metals
that were  never  completely leached  out of  the ore.  Removal of
these forms would  take place  simply  by settling out? however,
the  effluent  concentrations  of some metals  such  as  chromium,
nickel, and zinc remain at concentrations higher than  should be
achievable by alkaline  precipitation.  This suggests that these
metals are largely  in solution  coming  into the treatment system
and that the  optimum conditions  for metal  hydroxide formation
were not  being  attained at the  time of sampling.
                                            \
     The original BPCTCA limitations  for  this subcategory shown
in  Table 12-2 required zero  pollutant discharge except during
periods  of excess rainfall.   Objections  to   the-zero-discharge
limitations concerned the feasibility of using gypsum-saturated
water  for reuse in  the  air pollution control  scrubbers.

     The  proposed  BPT  waste  water  control  and  treatment
technology allows for the discharge of process waste water after
appropriate treatment.   This  technology is practiced widely in
the   industry   and   should   pose   no   technical   problems.
Implementation of BPT at all sites in the industry will  achieve
the indicated pollutant discharge levels.

     The nine plants  presently producing hydrofluoric acid  all
have installed BPT treatment or the  equivalent.  At the time  *bf
                              326

-------
  TABLE 12-21.  SUMMARY .OF LONG TERM MONITORING DATA FROM FOUR
                                           (1)
                    HYDROFLUORIC ACID PLANTS
                      Treated Waste Load (kg/kkg) or (Ib/lOOOlb)
                          Daily Data          30-Day Average Data
                    Long Term           (2) Long Term            (2)
Plant               Avejrage    St.Dav.  W   Ave£age  St.Dev.  VF
 No.   Parameter      (X)    (S)   (Sg)          (X)       (S)
#664

#753

#722

Fluoride
TSS
Fluoride
TSS
Fluoride
TSS
0.10
0.29
0.72
0.38
0.81
0.54
0

0

0
0
                             0.090 0.77 4.5    0.10     0.040   1.7


                             0.27  0.36 2.2    0.64     0.15    1.4
                             0.52 ' 0.59 3.3    --       —      —
                             0.37  0.62 3.5    —       —      —

    (3)
#705   Fluoride      —      —    —   —     0.49     0.22    1.7
       TSS           —      _•____     o.84     0.37    1.7
(1)
   Based on Reference'3 data.

(2)
   In the case of daily measurements, the variability factor, W,
   for a lognormal distribution is found by the expression ln(W) -
   S1(Z - 0.5S1), where S" is the estimated standard deviation of
   the logarithm derived from the arithmetic mean, X, and the
   arithmetic standard dev:
   (S')2 = in
[1.0 +f_Sj\2'
L     ^x/ _
ation, S, according to the relationship,
.When the value of Z is 2.33, the
   variability factor for the 99 percentile is obtained.
   For 30-day average measurements, a normal distribution is
   obtained and the variability factor is found by the expression,
   W = 1.0 + Z  f S\.  When the value of Z is 1.64, the         >
                 VxV
   variability factor is for the 95 percentile.  Please refer to
   Section 8.2 for a more detailed discussion of the statistical analysis
   of long term data.

(3)
   Although Plant #705 does not recycle the drip acid, the TSS
   data is not adversely affected and is used as the basis for
   the 30-day average W.

— Not Available.

-------
       TABLE 12-22.  TOXIC POLLUTANT TREATED EFFLUENT OATA

SUBCWEGQRY:  HXDROFLUORIC &CID

                                                                    (U
Average Dally fellutant Concentrations and loadings at Plants Sampled
                                 (1B3/1)
                         (kgAkg of anhydrous HP)
                                                             (2>
                                                     Overall   Average
            f705(8)    1705 (V)    1251 (V)    H67(V) Average  S Removal
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
ffercury
Nickel
Selenium
Thallium
Zinc
<0.010
<0. 00021
<0.0030
<0. 000063
0.00030
0.0000060
0.014
0.00029
0.10
0.0021
0.0060
0.00012
<0. 00040
<0. 0000080
0.050
0.0010
0.033
0.00069
0.0070
0.00015
0.071
0.0015
<0.0020
<0. 000042
<0.010
<0. 00021
<0.0017
<0. 000035
<0.046
<0. 00096
<0.020
<0. 00042
<0.022
<0. 00046
<0. 00050
<0. 000010
<0.010
<0. 00021
<0.0050
<0. 00010
<0.0012
<0. 000025
0.053
O.OOU
<0.17
<0.017
<0.020
<0.0020
<0.0020
<0. 00020
0.22
0.022
0.070
0.0069
<0.031
<0,0031
<0.0010
<0. 00010
0.52
0.052
<0.071
<0.0070
<0.0070
<0. 00069
0.16
0.01S
0.047
0.012
0.016
0.0040
0.0087
0.0022
0.050
0.013
0.060
0.015
0.010
0.0026
0.0065
0.0017
0.090
0.023
0.010
0.0026
0.0030
0.00069
1.9
0.49
<0.057
<0.0073
<0.012
<0.0016
<0.0032
<0. 00060
<0.083
<0.0091
<0.063
<0,0061
<0.017
<0.0015
<0.0020
<0. 00044
<0.17
<0.019
<0.030
<0.0025
<0.0045
<0. 00039
0.55
0.13
74
81
9
9
62
77
97
67
<54
Effluent
>Influent
85
83
 (S) Screening data from one 72-hour composite sample of treated
    effluent.
 (V) Verification data from three 24-hour composite samples.
 (1) The effluent data presented here corresponds to the raw waste
    data shown  in Table 12-10.  The methodology of the sampling
    program  is  described  in Section 5.1.2, and the scope of
    sampling in the Hydrofluoric Scid  industry is described in
    Section  12.3.3,
 (2) Wien averaging values indicated as "less than" «}, the
    absolute value MBS used and the resulting average was  indicated
    as a "less  than" value.
                    328

-------
sampling, seven  of  the  12 plants  operating were  meeting  the
proposed  fluoride  limitations  and  eight  were  meeting  the
proposed   TSS  limitations  according  to  the   data  available.
Although  there   is  practically  no long term  monitoring data
available to  support  the  additional proposed   limitations  on
toxic   metals,  the  screening and verification data  indicate
that  all   three  plants   sampled  are  meeting   the  proposed
limitations on  antimony,  copper,  and lead,  while  two of the
plants are meeting the proposed zinc  limitation  and one plant
is   meeting  the   proposed  chromium and   nickel  limitations.
With the limited amount of  toxic metal data,  it  is not possible
to  estimate  compliance  or  noncompliance on   a   statistical
basis.   The  Agency is   conducting   additional  treatability
studies.

12.7.2  Basis for Proposed BPT Effluent Limitations

Technology Basis

     For BPT, the Agency  is proposing  limitations  for which the
technology  basis  is,  or is equivalent  to,  equalization, lime
neutralization/alkaline   precipitation,    solids   removal   by
settling  or  thickening,   final pH adjustment,   and  discharge
of clarified  effluent.   The   in-house  process  recycling of
the   reactor  condensables  (drip acid)   is required for meeting
the  proposed  fluoride limitations.

Flow Basis

     The  reuse  of  treated  waste water  to  slurry kiln residues
to  the  treatment  system  is  not required  for meeting the BPT
limitations.  BPT  or  its equivalent is practiced by all plants
in  this  industry  including  six  which  reuse, for  slurrying
residues, proportions of their treated waste water ranging from
30 to 100  percent of  the plant flow as shown in Tables 12-4 and
12-20.

     The practice of reusing waste water  in this manner has two
opposing effects on the plant effluent:

     A.  A  decrease  in  the  net  discharge  unit  flow  rate
          (m3/kkg), and
     B.  An increase in the fluoride concentrations (mg/1).

     As a  result,   the  fluoride   unit loading  (kg/kkg) in  the
effluent  does  not decrease  as   a  direct proportion  to  the
decrease  in  the flow rate, but  is partially offset due to the
increase  in  fluoride  concentration  as  a function of  percent
reuse.  The   relationship of  percent  water reuse  to fluoride
concentrations  and unit loadings  is shown in Figure 12-14.  The
apparent  reason  for  the  increase in   fluoride concentration
with  reuse  is a  calcium  deficiency which  may  result  from the


                              329

-------
                    2'5             *
                              Percent Reuse
                                                             I'OO
                               SO
                          Percent Reuse
                                                  75
100
                     I12GEND
•  Long-term data
O  Expected with treatment system upgrading
D  Screening  and verification  sampling results
Figure 12-4.4
           FluorMe loads and concentrations discharged at
           selected hydrofluoric acid plants.
                        330

-------
buildup  of  sulfate    or  bicarbonate concentrations  in  the
treatment systems at  plants  where  reuse   is practiced.   Other
pollutants  such as   TSS   and metals would not be expected  to
exhibit  similar  concentration offset effects in these systems.

     It  should  be noted   that  while the practice  of reusing
waste water  for  kiln  residue  slurrying may be  advantageous in
some locations with  respect to alternative water  supply costs,
there is no associated  reduction in the  hydraulic  load,  size,
or cost  of the BPT treatment system  itself.

     The net result of water reuse is a moderate decrease in the
effluent  fluoride  loadings  which   is  achieved  at  a  small
additional annual cost of  less than  one percent of the estimated
BPT treatment systems cost  (Tables 12-12r 12-13, and 12-14).

     The model plant BPT treatment system is based on an inflow
rate  of  95.4 m3/kkg  derived from  the average of  nine plants
which  handle the kiln  residues in  a  slurry system as shown in
Table   12-4.   The treated  effluent flow rate is  54.6 m3/kkg
which is the  average  effluent  flow  rate  for  the same nine
plants and  corresponds to the  reuse of about 43 percent of the
flow for  residue slurrying and other uses.

Selection of Pollutants to be Regulated

     The selection  of pollutants for which  specific numerical
effluent limitations are proposed was based on  an evaluation of
raw  waste  data  from  the   screening  and verification sampling
program.  The following two major factors were considered:

     Raw waste  pollutant concentrations - A tabular summary of
maximum   raw  waste   concentrations   is  presented  in Section
12.3.3.    Data  from the one   plant sampled  for screening were
used  to determine  the need for verification  sampling.    The
maximum concentrations found during   verification are also shown
for comparison.   For  each pollutant, the maximum concentration
observed  gave   a  preliminary   indication  of   its  potential
significance in the subcategory.  On this  basis, the preliminary
selection  of candidates  for  regulation  included  zinc,  lead,
antimony,  nickel,  chromium, and copper in decreasing  order of
their  apparent  pollution  potential.   These  pollutants   were
observed   at   least   once  during   the   sampling   program  at
concentrations considered treatable  in this  industry using one
of  the  available  treatment  technology  options.     The  other
metals,  cadmium,  thallium,   and   mercury  exhibited  maximum
concentrations that were considerably  lower.

     Total  subcategory  _raw  waste pollutant loadings -Pollutant
raw   waste  loading  data   were used to  evaluate the overall
magnitude of the pollution potential for the subcategory.   Data
from  the plants sampled  are presented in Table  12-10 and the

                              331

-------
daily  and unit loadings  are  summarized in Table .12-11.   This
information,  coupled   with  the estimated  total   hydrofluoric
acid   production  rate  of    261,800   kkg/year,   yielded   the
approximate    total  annual  pollutant  loading  rates  for  the
subcategory shown  in   Section 12.3.3.    This method of ranking
the pollution potential  of  the observed toxic metals confirmed
the  maximum concentration  based   ranking  and  indicated  that
zinc,  nickel,  lead,  antimony,  copper, and chromium were the
six  dominant   toxic metals  in terms of    both    total   mass
loadings   and   treatable   raw waste  concentrations.

     In view of  the treatment technology already  implemented in
this  industry, the added  BPT regulation  of  any one of  these
pollutants   may provide assurance  that  all of  the   observed
toxic  metals  would  receive  adequate   treatment and   control.
This   includes taking credit  for  incidental removal of metals
which are  either   below   practical treatability  limits  or
are    not   particularly  amenable  to   removal   by   alkaline
precipitation  methods.  The latter  includes cadmium, selenium,
thallium,  and   mercury.   Thus,  because  zinc,  nickel,  lead,
antimony, copper,  and  chromium were  observed most  frequently at
treatable concentrations and may serve as reliable indicators of
overall treatment system performance, these metals were  selected
as the  additional parameters proposed  for  BPT regulations.

Basis of Pollutant Limitations

     Conventional  and  nonconventional parameters  -

     A.  pH:   The  treated effluent  is  to be controlled within
the  range of  6.0   to   9.0.   This  limitation  is   based on  the
data  presented  in Appendix  B of this report and  the JRB Study
(52).

     B.  TSS and  Fluoride:   The data  presented  in Tables 12-
20  and  12-21  were used for   the   development   of   TSS  and
fluoride limitations.  However, because  of  the   wide range  of
product  mixes,  significant  differences  in  residue handling,
waste water treatment, reuse practices, and dilution with other
product waste -streams,  it was necessary  to select  only those
plants where   the   effect   of  BPT  technology could be clearly
observed.  The  plants excluded are:

     |426  and  f!20 because  kiln residues  are handled as a dry
     solid,

     1167, 1967, and  £251  because the combined treatment of HF
     wastes  along  with  the  waste   waters  from other  major
     products  generated high  fluoride  loadings   in  the large
     volume discharges with fluoride  at  its minimum treatability
     concentration,
                              33-2

-------
     #705 because  cooler bottom condensables  (drip acids) are
     not recycled back to the process  but are  added to the raw
     waste contributing  complex  fluorides which tend to remain
     in  solution  after  lime  treatment,    TSS data  are not
     affected.


     Data from the remaining five plants  are presented in  Table
12-23  which   summarizes  the  development  of  the  proposed
regulations for total suspended solids  and  fluoride.  Since the
BPT level of treatment  does not  require  the  reuse of treated
waste  water  for  slurrying  kiln  residues,   the  performance  of
Plant  f837 was used as the  long  term average unit loading  basis
for the  TSS  and fluoride limitations.  The  variability factors
used for fluoride  are  based on  the  long term  data from Plants
1664 and #753 and those used for  TSS are derived from Plant #722
for daily   measurements   and   Plant  §705  for  30-day average
measurements  as  indicated in Table 12-23.

     The proposed  maximum  30-day  average   TSS limitation was
obtained   by  multiplying  the  variability factor for  30-day
averages from Table 12-23 by the  long   term average waste  load;
i.e.,  1.7  x 3.1  kg/kkg =   5.3   kg/kkg.  Similarly,  the  daily
maximum  TSS   limitation   was obtained  by multiplying   the
variability  factor  for daily measurements  by  the long  term
average?  i.e., 3.5 x  3.1 kg/kkg = 11 kg/kkg. The same approach
was taken  to obtain the proposed  fluoride  limitations;    i.e.,
1.6 x  1.8  kg/kkg = 2.9 kg/kkg for  the maximum 30-day average,
and  3,4  x 1.8  kg/kkg =  6.1  kg/kkg   for the    daily   maximum
limitation.   These  computations  are shown   on  Table 12-23 and
the proposed BPT  limitations are presented  in Table 12-24.

     The  concentration   basis   (C)    for   each  effluent  is
derived   from   the  relationship  between   concentration (C) ,
flow (Q), and  unit loading,

         C (as mg/1)  =  1000  (L)
                               (Q)

     Where L is  the effluent limitation expressed as  a unit
     loading in kg of pollutant per  kkg of product  (kg/kkg) , and
     Q  is  the flow  rate expressed  as  cubic meters  per  kkg of
     product  (m3/kkg).   (Note: kg/m3 =  1000  mg/1.)

     Thus,  the  concentration  basis   for  the   maximum  30-day
average TSS limitation is:

           (5.3  kg/kkg)      (WOO mg/l\  =  97  mg/1
           (54.6  m3/kkg)      \ kg/m3   )

and the concentration basis for the daily maximum limitation is
obtained  by  a similar  calculation or simply  by  applying the

                               333

-------
    TABLE 12-23.  DEVELOPMENT OF TSS AND FLUORIDE LIMITATIONS
Long Term Average
Waste Load Discharged
Plant
#837
#753
#928
#722
#664
Average of
practicing
(excluding
Reuse
(percent)
0
65
83
92
94
four plants
effluent reuse
#837)
Fluoride
(kg/kkg of HF)
1.8
0.72
1.0
0.81
0.10
0.'66
TSS
(kgAkg of HF)
3.1
0.38
1.7
0.54
0.29
0.73
Variability Factor for                    3.4^                  3.5(5)
 Daily Measurements

Variability Factor for                    1.6(1)                  1.7(6)
 30-Day Averages
                                              (?}                     (7}
Variability Factor Ratio  (VFR) 3.4/1.6 =  2.r ;        3.5/1.7 = 2.r ;

Effluent Limitations for BPT
  (from Plant #837)                         ,_.                         ,,.
 a. Daily Max       3.4 X 1.8 kg/kkg = 6.1JJ   3.5 X 3.1 kg/kkg = 11 *~>
 b. Max 30-Day Avg  1.6 X 1.8 kgAkg = 2.9^ '   1.7 X 3.1 kg/kkg =  5.3W

Effluent Limitations for BAT
  (from average of four plants)              /.,%
 a. Daily Max       3.4 X 0.64 kgAkg = 2.2JJ                     NA
 b. Max 30-Day Avg  1'.6 X 0.64 kg/kkg = 1.01  ;                     NA
NA - Not Applicable
 (1)  Variability factor average of Plants #664,1722 and #753 from Table 12-21.
 (2)  Ratio of the daily  (24-hr) variability factor to the 30-day
     average variability factor.  This value appears on the Proposed
     Limitations tables.
 (3)  The long term average loading in kgAkg multiplied by the
     variability factor for daily measurements as shown.
 (4)  The long term average loading in kgAkg multiplied by the
     variability factor for 30-day measurements as shown.
 (5)  Variability factor from Plant #722, Table 12-21.
 (6)  Variability factor from Plant #705, Table 12-21.
                                   334

-------
             TABLE  12-24.  PROPOSED LIMITATIONS
                      Hydrofluoric Acid
   Best Practicable Control Technology Currently Available
          Waste Water Flow: 54.6 m3/kkg  of HP (43% Reuse) *
Subcategory
Pollutant Performance
(rag/1)
Conventional and
Nonconventional
Pollutants:

Total Suspended
Solids

Fluoride
Toxic
Pollutants:

Antimony

Arsenic

Chromium

Copper

Lead

Nickel

Selenium

Zinc



(2)
57

(2)
33


(3)
0.80
(3)
0.50
(3)
0.10
(3)
0.50
(3)
0.30
(4)
0.17
(3)
0.20
(4)
0.55
Concentration Basis
(1) (mq/1)
VFR




2.1


2.1



2.0

2.0

2.0

2.0

2.0

2.0

2.0

2.0
Max
30 -day
Avg




97


53



0.80

0.50

0.10

0.50

0.30

0.17

0.20

0.55
24-hr
Max




200


110



1.6

1.0

0.20

1.0

0.60

0.34

0.40

1.1
Effluent Limit
(ko/kkct) of HP
Max
30-day
Avg




5.3


2.9



0.044

_ » '

0.0055

0.027

0.016

0.0093

— (5)

0.030
24-hr
Max




11


6.1



0.088

— **'

0.011

0.054

0.033

0.019

_'5)

0.060
 (1) - VFR: ratio of the 24 hour variability factor to the
      30 day variability factor.

 (2) - Long term average based on loading data and
      variability factors selected from Table 12-21.

.(3) - The lower limit of the literature treatability estimate
       (Table 8-11) is used as the basis for the  30-day average
       limitation when the observed average of the sampling data
       is below this level.

 (4) - Average effluent concentration from screening and verification
      sampling data.

 (5) - Tito effluent limitation proposed.

  *   From Table 12-4.
                      335

-------
variability factor ratio, VFR,  from Table 12-24   to the maximum
30-day average concentration;  that  is,

          (VFR) (max. 30-day average  concentration or loading)

              = daily  maximum concentration or loading In this
case,  the daily  maximum TSS   concentration is 2.1 X 97  mg/1 =
201 mg/1.

     In  the  same manner,  the  concentration basis  for   the
maximum  30-day average fluoride limitation  is,

            (2.9 kg/kkg)    /  1000 mg/l'N  = 53  mg/1
            (54.6 m3/kkg)   V
and  the daily maximum fluoride concentration is  2.1 X 53 mg/1 =
111 mg/1.  (Note:  due  to rounding  off,  this value differs just
slightly  from the  value  that appears in Table  12-24 which was
obtained  by  calculating  the concentration  directly  from the
daily  maximum limitation;   i.e.,

           (6.1 kg/kkg)     /1000 mg/A  =  112 mg/1.
           (54.6 m3/kkg)    \  kg/m3/

In either case, only two  significant figures should be taken.)

     Performance evaluation and  review  of discharge  quality has
been complicated by problems associated with  chemical analysis."
Prior to July 1976, the  methods  generally used for the analysis
of  fluoride   in  industry  were  specific  ion  electrode  or
colorimetry.  These  methods did   not detect the soluble complex
fluoride  species present  in the  waste  water.   The best  method
of  total fluoride detection (free   as  well  as   complex)  is
distillation  followed  by   analysis  using  the  specific  ion
electrode.  Using the distillation method, the complex  fluorides
are  hydrolyzed and the  resulting  HF  is carried over with the
distillate along with  any   free HF  in  the  sample.  Thus, the
method of total  fluoride analysis used  for effluent monitoring
is capable  of measuring free fluoride  and the fluoride  present
in   the  form  of  complex  ions  which  are  not removed  by  lime
treatment.   Monitoring  data on   effluent fluoride levels using
the  revised  method are  likely   to be  higher  than  the  levels
previously  reported under  the same treatment conditions.

     Toxic pollutants  - The  effluent  limitations proposed for
the  selected  toxic pollutant   control  parameters are derived
from  three  sources  of  information.    These are 1) screening
and   verification   sampling    data,   2)  literature  based
treatability   estimates   (Section   8.1), , and    3)   a limited
amount of long term  monitoring data from Plant &251.
                               336

-------
     The sampling  results  represent plant performance observed
during  three days  of  sampling  at  each  of  the  plants.   The
treated    effluent  data  on   the  toxic  pollutants  found  at
significant levels  are summarized in Table 12-22.  The average
values shown  for  each  pollutant   are  interpreted  as   being
approximately   equal  to  a   maximum  30-day   average unless
there  is   some reason  to  believe  that   abnormal conditions
existed either  in the process operation  or  in the  treatment
system at  the time of  sampling.   Abnormal  conditions   would
dictate   that  high  values  should   either be   excluded   or
regarded  as  daily   maxima    rather   than  monthly  averages.
For this  subcategory, the screening  and verification  data are
believed  to  represent  normal influent and   effluent values at
the plants sampled.

     For a number  of the  metal  pollutants, the  sampling data
demonstrate that  the effluent quality   and percent removal with
full   scale   BPT   systems   are   considerably better  than the
literature treatability data in Section 8.1  would indicate for
that  particular technology.   For  example,  even though arsenic,
cadmium,  mercury, and thallium average  influent concentrations
are   well   below   the   accepted   treatability   limits   for
lime/settling  shown  in  Table 8-11,  greater than  60 percent
removals  were  observed for   all but cadmium  as  is shown in
Table  12-22.   This high degree of  incidental removal supports
the contention that by  applying   effluent  limitations just  to
the dominant  metal   pollutant(s),  an  effective control  of the
other metals may  also  be assured.

     In Table 12-24, the  concentration bases for  the proposed
BPT    limitations are  derived  from   the   averaged  effluent
sampling data  unless the observed  pollutant concentration is
actually below the   literature treatability level.   In  such
cases,  the  lowest  applicable treatability level from  Table
8-11 is  used.  This approach  results in   the   setting   of
achievable   effluent  limitations for all of the pollutants of
concern and provides for the  possibility  of  wider  variations
in  the influent  quality.  Such  variations  may be associated
with   different   fluorspar   impurity  levels or  other  process
variables  not  fully taken  into   account by the  limited data
obtained.

     The basis  for the  proposed BPT limitations  on each of the
selected metals is given below.

     A.  Zinc:  The  raw waste concentrations of zinc ranged as
high   as   11.3   mg/1  (Section   12.3.3,   Table   of  Maximum
Concentrations  Observed)  and  averaged about 3.2  mg/1  (Table
12-10)  for  the  plants  sampled.    BPT  treatment  achieved  an
average   removal  of  better  than  80  percent  with an average
performance   concentration  of about 0.55  mg/1  in the treated
effluent  shown in   Table 12-22.    This  level  of performance

                              337

-------
approximately  equals  that    obtained   from    the  literature
treatability data    in Table   8-11.    The average performance
value  is  used  as  the  concentration basis   for   the proposed
maximum  30-day average  effluent limitation  of   0.030  kg/kkg
using the model plant flow of  54.6  m3/kkg   (Table  12-4).  This
limitation was  achieved  by all  but one of the plants sampled.
Using the model plant flow of 54.6  m3/kkg from Table 12-14, the
limitation was calculated as follows:

      (0.55 mg/1)(54.6 m3/kkg)  /   kg/m3  \  =  0.030 kg/kkg
                              \1000 mg/1/


     Since  long   term  monitoring  data   on    zinc are  not
available  from  the HF industry,   the variability factor ratio
(VFR) of  2.0 was selected  on  the basis of lead monitoring data
from Plant #251  presented  in Tables A-lOa  and A-lOc. This  is
justified by the   similarity  in  the  chemistry of lead, zinc,
and the other metals  of  concern under  BPT  treatment conditions.
Thus,

      VFR  »  VF of  da ily measu rements
              VF of  30-day averages

           =  2.0

and the daily maximum limitation  for zinc is,

      ,(2.0) (0.030 kg/kkg)  =  0.060 kg/kkg.

The proposed effluent  limitations on zinc and the other metals
of  concern are given in Table 12-24.

     B.  Nickel:   The sampling  data  indicate  better than  60
percent  BPT    removal of    nickel  resulting    in an  average
effluent   quality   of   about   0.17  mg/l.A     The  literature
treatability data    in   Table  8-11   show   an  effluent level
approximately equal  to this  value.  Thus,  0.17  mg/1  is used as
the  concentration    basis   for   the proposed  maximum  30-day
average effluent limitation of 0.0093  kg/kkg.  A VFR of   2.0 was
used  following the  same rationale  described  for  zinc.  Thus,
the proposed maximum 30-day  average limitation is,

       (0.17 mg/1)(54.6 m3/kkg) /  kg/m3  %  »  0.0093 kg/kkg,
                               V1000
and the daily maximum limitation is,

      '(2.0)  (0.0093 kg/kkg)  =  0.019 kg/kkg.


     C.  Lead:   Because   the   observed   average   raw  waste
concentration of  lead (0.66- mg/1)  was  very close  to the 0.30

                              338

-------
mg/1  lower  limit  of  its  estimated treatability  according to
literature  data,  the latter was selected as the concentration
basis  for  the   proposed    maximum 30-day  average   effluent
limitation rather  than   using the observed performance average
of   less than  0.02   mg/1.     This   results  in  setting the
limitation at 0.016 kg/kkg, a  level  which would be achievable
with BPT  treatment  even  when higher  influent  levels  occur.
It  also   avoids  taking  credit for  incidental removal   and a
higher removal  efficiency than can be  justified by the use of
this technology.   A WR of  2.0 was used  for lead   on  the  basis
of  long term data from  Plant £251.  The  proposed maximum 30-
day average limitation is,

       (0.30 mg/1) (54.6 m3/kkg) /  kg/m3  \  =  0.016 kg/kkg,
                               V1000 mg/V

and the daily maximum limitation is,

       (2.0) (0.016 kg/kkg)  =  0.032 kg/kkg.


     D.  Antimony: In a   manner   similar to that described for
establishing the  lead regulation, the  concentration  basis for
the   proposed  maximum  30-day average   effluent  limitation on
antimony  was set   at 0.80 mg/1  in  accordance  with literature
treatability  data.   The  resulting  limitation  of  0.044 kg/kkg
was met in two of  the four sampling  data  sets.  A VFR of  2.0
was also used for  antimony although a wider range of variation
may be observed when  more operating  data are collected.  The
proposed maximum 30-day  average limitation isr
      (0.80 mg/1)(54.6 m3/kkg) (  kg/m3  ^  =  0.044 kg/kkg,
                               \1000 mg/1/

and the daily maximum is,

      (2.0) (0.044 kg/kkg)  =  0.088 kg/kkg.

     E.  Copper :   The  concentration  basis  for   the  proposed
maximum 30-day average effluent   limitation on copper was set at
0.50  mg/1 in accordance with the literature  treatability data.
All  of   the   plants sampled had   average loadings below  the
proposed  0.027  kg/kkg limitation.    A  VPR  of   2.0  was  used
following the same  rationale  described  for  zinc.  Thus,  for
copper, the proposed maximum  30-day  average  limitation is,

      (0.50 mg/1)(54.6 m3/kkg)  /  kg/m3  \ =  0.027 kg/kkg,
                                \1000 mg/1/

and the daily maximum is,

      (2.0) (0.027 kg/kkg)  =  0.054 kg/kkg.

                              339

-------
     F.  Chromium:   The  concentration  basis   for  the proposed
maximum 30-day  average limitation on  chromium was  set  at 0.10
mg/1 in  accordance  with the literature treatability data.  Two
of   the  plants sampled   exceeded the proposed   limitation of
0.0055   kg/kkg.    A VFR   of  2.0  was used  following  the same
rationale  described  for  zinc.   The  proposed   maximum 30-day
average  BPT  effluent limitation is,

       (0.10 mg/1)(54.6 m3/kkg)  /   kg/m3  \ =  0.0055 kg/kkg
/  kg/m3  \
V1000 mg/1/
andf the daily maximum is,

       (2.0)(0.0055 kg/kkg)  =  0.011 kg/kkg.

     G.  Other metals:   The concentration   bases  for  arsenic
and   selenium are   also  presented in Table 12-24.   These are
intended   to  serve  as guidance in cases where these pollutants
are found to  be of serious concern.

12.7.3  Basis for Proposed BCT Effluent Limitations


     For  the  Hydrofluoric Acid  Subcategory,  the  Agency  is
proposing BCT  limitations applicable to total suspended solids
(TSS)  based   on  the  estimated  performance  of  Level  2   (BAT)
treatment.  Assuming  that the  addition of dual media filtration
to  "the  BPT  system  removes  approximately  30  percent  more
suspended solids, the maximum  30-day average TSS loading of 5,3
kg/kkg  (Table 12-24) would   be  decreased  to  3.7  kg/kkg  as
follows:

      (1.00 - 0.30)(5.3 kg/kkg) = 3.7 kg/kkg

     By adjusting  the loading to  account for the decrease in
effluent flow rate from BPT (54.6 m3/kkg) to BAT (33.4 m3/kkg),
the  proposed  BCT  maximum  30-day  average  effluent  limitation
becomes,

      (3.7 kg/kkg) /33.4 m3/kkg\
                  V54.6 m3/kkg/

     *  2.3 • kg/kkg

     The corresponding daily maximum limitation is  then obtained
by applying the VFR value of 2.1 (Table 12-24).  That is,

      (2.1)(2.3 kg/kkg) =  4.8 kg/kkg
                              340

-------
12.7.4  Basis for Proposed BAT Effluent Limitations

The Application of Advanced Level Treatment

     Utilizing the cost  estimates presented in this  report, the
Agency  has  analyzed  the  cost/ effectiveness  of  the base level
systems  (BPT)  and  the  various  advanced  level options  for
conventional, nonconventional and toxic pollutant removal.  The
economic impacts  on  the Hydrofluoric  Acid industry  have been
evaluated  in  detail  (53)  and   taken into  consideration in the
selection  of  the   technology  basis  for  the  proposed  BAT
regulations.

     For BAT,  the  Agency  is  proposing  limitations based on
treatment  consisting  of  Level  2  technology.    It  has   been
estimated  that this will remove  11,100 pounds  per year of toxic
metals and 104,000 pounds per   year   of fluorides  in  addition to
the pollutant removals  already being achieved  by BPT treatment.

     The  Agency  considered  the  use  of  treatment  Level  3
(addition  of  sulfide  precipitation)   but  rejected  it  due  to
lack  of  performance  data.   EPA  also considered  Level    4,  a
variation of Level 2,  that would substitute soda ash  in the lime
precipitation step and  allow  90  percent  recycle of effluent.
This  option  was  rejected due  to being prohibitively expensive.
Pollutants limited by the proposed  BAT are fluoride, antimony,
chromium,  copper, lead, nickel, and  zinc.

Technology Basis

     For BAT, the  Agency is  proposing more  stringent effluent
limitations  on  fluoride  and  the  toxic  metals  based   on  the
addition of dual media filtration or its equivalent to  the  BPT
treatment system,  coupled with the  requirement of  at least 65
percent  effluent  reuse  for   kiln  residue   slurrying.   This
technology  aims at  both  the  reduction  of   suspended  solids
containing fluorides  and   metal precipitates and the reduction
of the dissolved  component  loadings  of these substances  in the
final effluent.   The minimum reuse  rate  of 65  percent  was
selected because  it is  typical of the  five plants  (Plants f!67,
#753, 1928,  ^664, and f-722) which  presently practice reuse as
is shown   in Table 12-4.

Flow Basis

     With  the model plant  inflow rate  of  95.4 m3/kkg and  the
reuse  of  65  percent  of  the  treated  effluent,  the   quantity
discharged  is 33.4 m3/kkg;  i.e.,  (1.00  -0.65) (95.4 m3/kkg)  =
33.4 m3/kkg.
                              341

-------
Selection of Pollutants to be Regulated


     For the BAT  regulations,   the Agency  has selected fluoride
and  the  same six  toxic  metals identified in the proposed  BPT
regulations.  The rationale for their selection   is discussed in
Section 12.7.2.

Basis of Pollutant Limitations

     Nonconventional pollutants  -    The   only  nonconventional
pollutant is fluoride.   The limitation proposed  for BAT is based
on the performance of the four  plants shown in Table 12-23  that
presently reuse  at least 65 percent of their treated effluent.
The long   term average effluent loading taken from Table 12-23
is  0.66  kg/kkg  for  the four  plants and this   is equal to  the
performance of Plant  f753  which  reuses 65 percent, the lowest
reuse rate  of  the four  plants.   Although these plants do not
employ  the  filtration  technology which  is  the basis   for the
BAT regulation, the use of  this performance in  conjunction with
the   30-day average  variability  factor  of 1.6  and  the model
plant   net   discharge rate  of   '33.4  m3/kkg  results  in  a
calculated  maximum   30-day average  concentration of  30  mg/1
total   fluoride.   Thus,   the maximum 30-day average limitation
is,

      (1.6) (0.66 kg/kkg) =  1.1 kg/kkg

and its concentration basis is,

      (1.1 kg/kkg?   /1000 mg/lN  =  33 mg/1
      (33.4 inS/kkg)  \kg/m3    )

     This  represents  a  58 percent  reduction  in    fluoride
concentration in  going from BPT  (43 percent reuse) to  BAT  (65
percent reuse  plus  filtration).   The  use  of a fixed loading
limitation allows  the  permissible  concentration to  increase
as a  function of  percent reuse.   The proposed daily maximum
limitation on fluoride  is obtained by  utilizing the long term
average and  variability  factor for daily measurements,

      (3.4) (0.66 kg/kkg) =  2.2 kg/kkg

and the concentration basis is,
      (2.2 kg/kkg)  /lOOOmg/l\  =  66 mg/1
      (33.4 m3/kkg) ykg/mS    )

     The variability  factors used  for the  BAT limitations on
fluoride are the  same  as  for BPT shown  in. Table 12-23.   The
proposed    BAT   limitations    for    the  Hydrofluoric  Acid
Subcategory  are presented in Table 12-25.


                              342

-------
              TABLE 12-25.   PROPOSED LBHTATIONS
                     Hydrofluoric Acid
                 Best  Available Technology
         Waste Water Plow:  33.4 m3/kkg of HF (65% Beuse)*

                          Concentration Basis   Effluent Limit
                              (1)    (mg/1)       (kg/kkg of HF)
rua.j.uucuiu J
Nonconventional
(2)
Fluoride, F
Toxic
Pollutants:
(2)
Antimony1 '
Arsenic
Chromium^
Copper'2'
Lead ~(2)
Nickel (2*
Selenium
Zinc (2)
', L cauauii jLi_y
(mg/1)
Pollutants:
33
0.70
0.50
0.040
0.29
0.060
0.15
0.18
0.52
vrn — — — — •—
30-day
2.1 33
2.0 0.70
2.0 0.50
2.0 0.040
2.0 0.29
2.0 0.060
2.0 0.15
2.0 0.18
2.0 0.52
24-hr
Max
66
1.4
1.0
0.080
0.58
0.12
0.30
0.36
1.0
30-day
1.1
0.023
_J4)
0.0013
0.0097
0.0020
0.0050
_J4>
0.017
24-hr
Max
2.2
0.047
_J4)
. 0.0027
0.019
0.0040
0.010
_J4>
0.035
(1)  - VFR:  ratio of the 24 hour variability factor to the
     30 day variability factor.

(2)  - Also  applicable for PSES limitations.

(3)  - 30-Oay average calculated for the model plant based on
     data  In Table 12-2J..

(4)  - No effluent limitation proposed..
 *    !he effluent flow rate is 35 percent of the average influent
      shown in Table 12-4  (i.e., 0.35 X 95.4 is3/kkg = 33.4 m3/kkg).
                                   343

-------
     The  estimated   performance  of  Level   3   and   Level  4
alternative technologies   are    presented    in  Tables  12-26
and 12-27,   respectively.    For  these options,   the   expected
performance is   expressed   in   the  achievable  maximum 30-day
average  and daily   (24-hour)   maximum  concentrations.  These
are  presented   for  comparison   purposes   and  are  not   the
bases  for any proposed  regulations.

     The Agency is  currently conducting treatability studies on
dual  media  filter performance  with HF   industry wastes.  The
results will be available prior to promulgation.

     Toxic  pollutants  -  For  BAT  regulations,  the  EPA  is
proposing  more stringent controls on  the discharge of the six
toxic metals  of  concern  on the basis  of  a   reduced volume of
dicharge  and   physical    removal  by  filtration.    Alkaline
precipitation converts  most of the dissolved  metals into less
toxic, insoluble  forms  such as  hydroxides  and hydrated oxides.
Other  mechanisms  of  removal    including coprecipitation   and
flocculation   are   undoubtedly  involved  during  the treatment
process and probably  account  for a substantial portion of the
removal of certain toxic metals.  Because there  is no  directly
applicable  data   on  filter performance  in  the HF   industry,
literature treatability studies (40,  41) have been evaluated in
order to  estimate  the  probable efficiency of  filtration for
the  removal  of residual  suspended metal   precipitates.   The
following information   was derived from   pilot  scale tests on
raw municipal, waste  water samples spiked   with toxic metals,
treated  with  lime,   and   settled,   followed  by  dual  media
filtration of the clarified effluent:

      Removal of Suspended Metal Precipitates by Filtration

                                       (Percent)
      Antimony                             7
      Arsenic                              0
      Chromium (III)                      60
      Copper                              42
      Lead                                80
      Nickel                              14
      Selenium                            12
      Zinc                                 6

     The filter efficiency values have   been   used  in setting
the  proposed BAT  limitations.   The  basis for  the limitation on
each  metal is given below.

     A.  Zinc:   Filtration of  the BPT effluent  is estimated to
reduce the  zinc  concentration by 6  percent   to approximately
0.52  mg/1.   This value  is  used  as the  concentration basis
for the ma'ximum 30-day average effluent limitation.  Application


                              344

-------
               12-26.              OF
                          Hydrofluoric Acid
                        Level of Treatment:  3
              Waste Water Plow:  33.4 m3/kkg of HP  (65% Reuse)
Pollutant Treatability WR
(rag/1)
Nonconventional
Fluoride, F
Toxic
Pollutants:
Antimony
Arsenic
Chromium
Copper
Lead
Nickel
Selenium
Zinc
Pollutants:
33 2.1
0.70 2.0
0.50 2.0
0.010 2.0
0. §50 2.0
0.060 2.0
0.10 2.0
0.18 2.0
0.20 2.0
Achievable Concentration
Max
30-day 24-hr
Avg Max
33 66
0.70 1.4
0.50 0.10
0.010 0.020
0.050 0.10
0.060 0.12
0.10 0.20
0.18 0.36
0.20 0.20
(1)  - WR:   ratio of the 24-hour variability factor to the 30-day
     variability factor.
                                  345

-------
          ISBLE 12-27.   PERFORMANCE OF ALTERNATIVE TECHNOLOGY
                           Hydrofluoric Acid
                         Level of Treatment: 4
               Waste Water Plow:   9.5 m3/kkg of HF (90%  Reuse)
Pollutant Treatability WR(15
(ng/i)
Nbnconventional
Fluoride, F
TOXIC
Pollutants;
Antimony
Arsenic
Chrcniium.
Copper
I^ead
Nickel
Selenium
Zinc
Pollutants:
33 2.1

0.70 2.0
0.50 2.0
0.040 2.0
0.29 2.0
0.060 2.0
0.15 2.0
0.18 2.0
0.52 2.0
Achievable Concentration
(irer/1)
Max
30-day
Avg
33

0.70
0.50
0.040
0.29
„ 0.060
0.15
0.18
0.52
24-hr
Max
66

1.4
1.0
0.080
0.58
0.12
0.30
0.36
1.0
(1) - WE: ratio of the 24-hour variability factor to the 30-day
     variability  factor.
                                  346

-------
of the BAT model plant discharge  rate  results  in a corresponding
loading  limitation of 0.017 kg/kkg.  That is,

       (0.52 mg/1) (33.4 m3/kkg) /  kg/m3  \ = 0.017 kg/kkg
                               VLOOO mg/y

and,  for  the  daily maximum  limitation  using the VFR  value of
2.0, one  obtains,

       (2.0) (0.017 kg/kkg)  =  0.034 kg/kkg.

     This represents an  overall reduction of  43 percent from the
BPT  loading limitation.   The VFR value  of 2.0 used  for BPT was
also  used  for BAT  because the  variability   of  the   filtrate
quality  is   expected  to   be no   greater  than  the  observed
variability of the  unfiltered  effluent  at Plant  1251  (Tables
A-lOa  and  A-lOc).   Treatability  studies .are being  conducted
by the EPA  and the  results  on filter  performance   will  be
available prior to  promulgation.  The proposed BAT limitations
on  the toxic metals  are  included in Table 12-25.

     B.  Nickel:    The   addition   of  filtration  to  the  BPT
effluent   is  estimated   to achieve a  14   percent  reduction in.
the nickel  concentration.   The  concentration  basis for  the
proposed   BAT   limitation is therefore  set  at  0.15  mg/1  and
results  in a  maximum   30-day  average  loading  limitation of
0.0050 kg/kkg.  That is,

       (0.15 mg/1)(33.4 m3/kkg) /  kg/m3  \ =  0.0050 kg/kkg
                               \1000 mg/1/

and the daily maximum is, .

       (2.0) (0.0050  kg/kkg)  =  0.010 kg/kkg.

     This  represents  an overall 46 percent  decrease  from  the
corresponding BPT level.   A VFR value of  2.0  was used following
the same rationale  as applied to zinc.

     C.  Lead:  With  the addition  of  filtration,  providing an
estimated  80   percent   removal   of  the   residual lead,  the
concentration  basis  for the proposed BAT  limitation is   set at
0.060  mg/1.   This  is   in  close  agreement with the  literature
treatability  data  summarized  in  Table  8-11.  O'n this  basis,
the  maximum  30-day  average effluent limitation .for  lead  is
0.0020  kg/kkg.  That is,

       (0.060 mg/1) (33.4  m3/kkg)  /  kg/m3  N =  0.0020 kg/kkg
                                VLOOO  mg/1/
                              347

-------
and, the proposed daily maximum  is,

       (2.0) (0.0020 kg/kkg)  =  0.0040 kg/kkg.

     This represents nearly  an 88  percent  reduction  from  the
BPT level.   A VPR of 2.0 is used  following the same   rationale
as applied  to zinc.

     D.  Antimony: The addition   of  filtration  is expected to
decrease the antimony  concentration by  approximately  7 percent.
Thus,   the   maximum  ' 30-day  average  concentration   basis  is
estimated  as 0.70  mg/1.    This  establishes  the corresponding
loading  limitation  at 0.023  kg/kkg which  is  about 48 percent
lower  than  the  BPT  limitation.  The proposed maximum 30-day
average  limitation is,
                    i
       (0.70 mg/1) (33*.4 m3/kkg) /'  kg/m3  "\  =   0.023 kg/kkg
                               VLOOO mg/y
and the daily maximum  is,

       (2.0)(0.023 kg/kkg)  =  0.046 kg/kkg.

     The VFE  is estimated to be 2.0  for the  reason that this
value was  used for the BPT limitation on antimony.

     E.  Copper:  Filtration of the BPT effluent is estimated to
achieve  approximately  a 42 percent  reduction  in the average
copper  concentration.  Thus,  a  value of 0.29 mg/1 is used  as the
concentration  basis  for  the proposed 30-day  average  effluent
limitation of 0.0097 kg/kkg.  That  is,

       (0.29 mg/1) (33.4 m3/kkg) f kg/m3  \ =   0.0097 kg/kkg
                               \1000 mg/1/

and, the daily maximum limitation is,

       (2.0)(0.0097 kg/kkg)  =   0.019 kg/kkg.

     This  represents  an overall reduction of  64 percent below
the BPT    level.   A   VFR value of 2.0  was  used  for  the BAT
limitations for   the same reason described for zinc.

     F.  Chromium:  For chromium, an average additional removal
of  approximately   60  percent  is  expected  with  the use  of
filtration.   For  this reason,  the  concentration basis for the
proposed 30-day average  BAT   limitation is  set at 0.040 mg/1.
This results  in  a  corresponding  loading limitation of  0.0013
kg/kkg.  That is,

       (0.04!0 mg/1) (33.4 m3/kkg) /   kg/m3  \ =  0.0013 kg/kkg
                                \1000 mg/1/

                              348

-------
and, the daily maximum limitation is,

       (2.0){0.0013 kg/kkg)  =  0.0026 kg/kkg.

     This  represents  an  overall  reduction of  approximately 76
percent from the BPT  level.    A   VFR value  of   2.0  was also
used   for  chromium for the same reasons presented for zinc.

     G.  Other metals:   The   concentration   basis for arsenic
and selenium are also given assuming 0  and  12 percent removals,
respectively,  by the  addition of filtration to  the BPT system.
The values presented  in Table  12-25 for these toxic pollutants
are  intended  to be  used as guidance  in cases  where  they are
found  to be of serious concern.

12.7.5  Basis for Proposed Hew Source Performance Standards

Application of Advanced Level Treatment

     Examination of  raw  waste loads  indicates  that  the prime
source  of  pollutants  at  HF    plants  is  the  kiln  waste.
Currently,   two plants  handle their kiln   waste   as  a solid
greatly  reducing    the total  raw   waste load  and  subsequent
effluent.  Based on this  and  an   examination   of control and
treatment   alternatives   available to  this industry, it  has
been determined   that new  HF   facilities should achieve  the
effluent   quality   attainable by   NSPS, Level  5,  technology.
The   control parameters  for NSPS  are   pH,  TSS,    fluoride,
nickel,  and  chromium.   The  recommended treatment technology
for new sources as described is dry  handling  of  kiln wastes and
chemical treatment,  filtration and  reuse  of   other   treated
wastes.   The use of  soda ash for  precipitation of   fluorides
will  allow  approximately 60  percent  reuse  for air  pollution
control  scrubbers, the second  major source  of waste  water.

     Raw  waste  toxic  pollutant    metal loadings  from sources
other   than kiln wastes  were  minimal   and   only occasionally
observed at  potentially   significant  levels.    It is  assumed
that   following   chemical precipitation  for fluoride removal,
the effluent loads  discharged  will be  insignificant with regard
to these metals.

Technology Basis

     For new  plants  in  the  hydrofluoric   acid  industry, the
specified waste treatment technology  is  the reduction of waste
flow and pollutant loadings by  the  dry handling of  kiln wastes
and the  treatment of  other  wastes  by alkaline precipitation
followed  by  settling of   solids  and   filtration  of   the
effluent.    The technology also  incorporates   the reuse of 60
percent or more   of the  treated  effluent for the  air pollution
control  scrubbers.   Two plants  now practice dry handling of

                               349

-------
kiln  residues.  This  technology greatly reduces the waterborne
raw waste loads  and   is   available to  new plants because they
have   the  opportunity  to  design  the  most efficient systems
without retrofitting.  The dry  solids generated  would have  to
be handled and disposed of in a  manner  consistent  with   any
applicable  requirements  of   the  Resource   Conservation  and
Recovery Act  (RCRA), 42  USC 6901 et.   seq.

     Pollutants  limited  by  the proposed  NSPS  regulations are
pEr   TSSr fluoride,  chromium, nickel, and  zinc.    Metals from
scrubber  water  and   other  plant  waste  streams facilitates
the reuse of   60  percent  or  more  of  the treated  effluent
for scrubber  operation.   Effluent reuse for this purpose  is
presently practiced  in the hydrofluoric  acid  industry.   Plant
f722  reuses  92  to 100 percent of its  soda  ash  treated  waste
water  for  both scrubber  operation  and  kiln  residue transport.
Plant   1664  reuses  approximately  94    percent of  similarly
treated waste   water   for the same  purposes.   Information on
these  plants  is  summarized in Tables 12-5 and 12-20,

Plow Basis

     The basis   for  the   model  plant   total  treatment system
influent is  the  flow  data on  scrubber  and other waste water
sources  (excluding  gypsum slurry  water)  for five of  the nine
plants  presented  in   Table   12-5.    Plants f426,  1722,  and
1837  were  excluded   because of incomplete data for  scrubber
effluent  and  Plant   f!67 because  of an unusually  high  flow
rate  for  the   scrubbers.   The average   raw waste  flow  rate
for  the  five   remaining plants   is   14.9   m3/kkg  and  with a
reuse rate of  60  percent the net effluent is 6.0 m3/kkg.

Selection of Pollutants to be Regulated

     For NSPS,  the  two major  waste  water4 sources  of   concern
are   the  air  pollution control   scrubbers and "other"   process
wastes.   The latter  includes  the  indirect  contact wastes from
surface  drains  but excludes storm water runoff.  The pollutant
parameters   of  concern   are    pH,   TSS,   fluoride,  and  the
toxic  metals.    Screening and  verification sampling  data on
the raw  scrubber and   other  sources   confirm  the need   for
limitations on  pH,   TSS,  and   fluoride.   The  four sets   of
sampling data  from  these  sources  indicate  that the  relative
importance  of   the toxic metals is considerably less  than was
found  for  the BPT and  BAT  sources   which included the  kiln
residue slurries.  The observed maximum  and  average  raw  waste
concentrations of the  toxic metals  are   shown  in Table 12-28.
In  the NSPS  raw waste sources,  nickel,  zinc, chromium,  and
selenium   were   the  only toxic  metals  which   showed  maximum
concentrations that would be treatable by alkaline precipitation
and  filtration.  Of   these,   selenium is marginal  even at  its
maximum  concentration   and    has    an average concentration

                             350

-------
      TABLE 12-28.   TOXIC POLLUTANT RAW WASTE DATA USED TO
                               NEW SOURCES*
SUBCATEGORY: HYDROFLUORIC
Pollutant
Antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Thallium
Zinc
ACID
Concentration
(I)
MaKimum
(tng/1)
0.030
0.014
0.021
0.41
0.12
0.029
0. 0020
0.81
0,24
0.0040
0.45

(2)
Average
(rag/1)
0.014
0.0090
0.0080
0.11
0.049
0.011
0.0010
0.18
0.068
0.0020
0.15
* Based on four sets of screening and verification sampling
  data from Plants |705, #251, and #167 taking only the
  scrubber and "other" waste sources.

(1) Maximum value observed from screening and verification
    sampling data.

(2) Average value derived from screening and verification
    sampling data.
                                  351

-------
that  is far  below  the minimum  level treatable.   For   this
reason, only  nickel,  chromium,  and zinc  have  been selected as
toxic pollutant control parameters for NSPS regulations.

Basis of Pollutant Limitations

      Conventional Parameters -

     A.  pH:    For  NSPS,   the  BPT    limitation  is  retained.
Control of  the final effluent  within  the range  of pH  6.0 to
9.0 is required on the basis of data  presented in Appendix  B of
this report and the JIB Study (52).

     B.  TSS:   Although there   is no  applicable  performance
data  available on the filtration of   treated and clarified NSPS
waste   water,    taking   the  proposed    BPT    maximum 30-day
average  concentration  of  97  mg/1  of  TSS  and assuming  30
percent    additional  removal    by   filtration,  an  estimated
performance level  of   68   mg/1   is    obtained.  Pilot  scale
studies    (41)    have    demonstrated  an  average removal  by
filtration  of  approximately 30   percent  from waste  water
containing   suspended   metal hydroxides  after  lime treatment.
A WR of 2.1  is used on the basis of  long  term data  summarized
in  Table 12-21  and  described in the  BPT section.  Thus, the
proposed maximum 30-day limitation on TSS  is,

       (68 mg/1)(6.0 m3/kkg) /  kg/m3  \ »  0.41  kg/kkg
                            V1000 mg/1/

and, using the VFR value of 2.1,

       (2.1)(0.41  kg/kkg)   =  0.86 kg/kkg  is the proposed daily
maximum.

The proposed NSPS limitations are presented in Table 12-29.

     Nonconventional pollutants   -   The  only  nonconventional
pollutant  of concern is  fluoride.  The concentration basis for
the proposed maximum NSPS  30-day average  limitation  is   set
equal  to  the 30   mg/1  BAT model plant  performance level (Table
12-25),  because'the  treatment technology  is the same.   A VFR of
2.1  is  used on the same basis given for  the use of this ratio
in  the BPT   and  BAT limitations.   Thus, the  proposed 30-day
average is,

       (30 mg/1)(6.0 m3/kkg) /  kg/m3  \ =  0.18  kg/kkg
                            V1000 mg/1/

and, using the VFR value of 2.1, the  daily maximum is,

       (2.1)(0.18 kg/kkg)  =  0.38 kg/kkg.
                              352

-------
            TABLE  12-29.  PROPOSED LIMITATIONS
                     Hydrofluoric Acid
             New Source Performance Standards
         Waste Water Flow:  6.0 m3Akg  (60% Reuse)
Pollutant Treatability
(mg/1)
Conventional and
.Nonconventional Pollutants:
Total Suspended 68
.Solids, TSS
(2)
Fluoride, F 30
Toxic
Pollutants :
Antimony 0.70
Arsenic 0 . 5
(2)
Chromium 0.040
Copper 0.29
Lead 0.060
(2)
Nickel 0.15
Selenium 0.18
(2)
Zinc 0.52
Concentration Basis Effluent Limit
(1) (rog/1) (kg/kkg of HP)
*je»r* —-.——,—,— ——«__.. —«.««,„— ,.__™™__
30-day 24-hr 30-day 24-hr
Avg Max Avg Max
2.1 68 143 0.41 0.86
2.1 3H 63 0.18 0.38

(3) (3)
2.0 0.70 1.4 — —
(3) (3)
2.0 0.5 1.0 — —
2.0 0.040 0.080 0.00024 0.00048
(3) (3)
2.0 0.29 0.58 — -—
(3) (3)
2.0 0.060 0.23 -- —
2.0 0.15 0.30 0.00090 0.0018
(3) (3)
2.0 0.18 0.36 — —
2.0 0.52 1.0 0.0031 0.0062
(1)  - WR:  ratio of ±he 24 hour variability factor to the
      30 day variability factor.
(2)  - Also  applicable for PSNS limitations.
(3)  - No effluent limitations proposed.

                                353

-------
      Toxic pollu tants -

     A.  Nickel:   For the proposed  NSPS  limitation on nickel,
the  BAT  concentration basis  and  VFR value  are  used because
the   treatment technology  remains essentially the same.   The
proposed  maximum  30-day average limitation  is,

       (0.15 mg/1)(6.0 m3/kkg) /  kg/m3  \  =   0.00090 kg/kkg
                              V1000 mg/lj

and, with a VFR of 2,0, the daily  maximum  is,

      (2.0)(0.00090 kg/kkg)  =  0.0018 kg/kkg

     The toxic pollutant limitations  for NSPS  are  presented  in
Table  12-29.

     B.  Chromium:    Similarly    for    chromium,    the    BAT
concentration basis  and  VFR  value  are   again   used  for  the
proposed NSPS  limitations.   Thus,  for chromium,  the proposed
maximum 30-day  average limitation is,

       (0.040 mg/1)(6.0 m3/kkg) /   kg/m3  \ =   0.00024 kg/kkg
                               \1000 mg/1/

and the daily maximum is,

       (2.0)(0.00024 kg/kkg)  =  0.00048 kg/kkg


     C.  Zinc:   In the case   of  zinc,  the concentration basis
for the  proposed  maximum  30-day  average   is  the  same  as BAT
(0.52 mg/1).  Thus, the maximum 30-day  average limitation is,

       (0.52 mg/1)(6.0 m3/kkg) /  kg/m3  \  =  0.0031 kg/kkg
                              \1000 mg/1/

and, the daily maximum is,

       (2.0)(0.0031 kg/kkg)  =  0.0062 kg/kkg

     D.  Other  metals:  The  concentration bases   for antimony,
arsenic, copper, lead, and selenium are also provided  in  Table
12-29 to be used as guidance  in cases where one or more of these
toxic metals may be of more serious concern.

12.7.6  Basis for Proposed Pretreatment Standards

Existing Sources

     For Pretreatment Standards for Existing Sources (PSES), the
Agency is proposing limitations based  on BAT.    The pollutants
to be  limited  are fluoride,  antimony,  chromium,  copper,  lead,

                              354

-------
nickel, and  zinc  as  indicated in Table 12-22.  However, at this
time, there are no indirect dischargers in the HF industry.

New Sources

     For Pretreatment  Standards  for  New Sources   (PSNS), the
Agency is proposing limitations  based  on  NSPS.   The pollutants
to  be  regulated  are  fluoride,   nickel,   chromium, and  zinc
as  indicated in Table 12-29.
                              355

-------

-------
                          SECTION  13


                  HYDROGEN PEROXIDE INDUSTRY
13.1  SUMMARY OP DETERMINATIONS


     It has  been  determined  that  no further effort be given to
developing or  revising  BAT,  NSPS,  or  pretreatment regulations
for  the   Hydrogen  Peroxide  Subcategory   using  either  the
electrolytic process or the organic process.

     The bases  for this recommendation  are:  1)  only one plant
exists   that   manufactures    hydrogen  peroxide  using   the
electrolytic process  and  2)  no toxic  pollutants were found in
the  wastes  using  the  organic   process.     Therefore  this
subcategory  is excluded under Paragraph 8 of the Consent Decree.

13.2  ASSESSMENT OF THE WATER POLLUTION POTENTIAL

13.2.1  Production Processes  and Effluents

     In the electrolytic process,  ammonium  (or other) bisulfate
solution is  electrolyzed,  yielding ammonium persulfate  at the
anode and  hydrogen gas  at  the  cathode.  The presulfate is then
reacted  with  water  to  yield  hydrogen  peroxide  and original
bisulfate.   Hydrogen peroxide is  separated from bisulfate by
fractionation, after which it is concentrated and filtered.  The
only waste is a  stream of  condensate from  the fractionation
condenser.

     The   organic   process    involves    the   reduction   of
alkylanthraquinone by hydrogen over a  supported metal catalyst
to  produce  the  corresponding alkylhydroanthraquinone.    The
reacted  mixture  is  oxidized  to  form  hydrogen peroxide  and
original  alkylanthraquinone.   The  peroxide is  extracted with
water and the organic material  in  the solvent is  recycled to the
process.   Since  hydrogen peroxide  manufactured  by the organic
process consists  of a series of  exothermic chemical  reactions,
the bulk of  the water usage is for  process cooling (contact and
noncontact).  Noncontact cooling accounts for over 90  percent of
the  total  water  usage  in this subcategory.  The  waste water
sources  include  contact cooling  (barometric-condenser)  water,

                              357

-------
purification   washing   of   the  organic   working   solutions,
regeneration waste from the deionizers, and leaks and spills.

13.2.2  Plants

     Only   one  plant  exists   in  the  United  States   that
manufactures hydrogen peroxide  using  the electrolytic process.
The  hydrogen  peroxide  subcategory  profile  data received  in
response to 308 letters is given in Table 13-1.

     Three  plants  produce  hydrogen  peroxide  by the  organic
process.

13.2.3 . Toxic Pollutants

     Data has been received on 100 percent of the industry as a
result  of  section 308  letters.   A"" sampling survey  for  toxic
pollutants  was   made   for  three  plants.     At  one  plant,
pentachlorophenol  was  found  in  significant  concentrations.
However, it was determined that is  presence was due  to its use
as  a   weed  killer  at   the plant   site  and  this  use  was
discontinued.  Two more plants were sampled in the verification
phase,  and  the survey  indicated that  no  toxic pollutants were
being discharged  in significant quantities.

     Toxic pollutants found during sampling were as follows:

                                       Maximum Concentration
         Pollutant                       Observed (pg/1)

         Zinc                                 256
         Pentachlorophenol                   4850
         Bis(2-ethylhexyl)phthalate            20
         Chloroform                            11
         Naphthalene                           11

13.3  STATUS OF REGULATIONS

     Since  no  toxic  pollutants  were  found  in  significant
concentrations, the subcategory is excluded under Paragraph 8.
                              358

-------
      13-1
SIBCKCEQORY PROFILE DATA
SUBCATEQOKf
ffiTDBQGEN PEROXIDE
Total subcategory capacity rate
Total subcategory production rate
Number of plants in this subcategory
308 Data on file for
    With total capacity of
    With total production of
    Representijig capacity
    Representing production
    Plant production range:
            Minimum
            Maximum
    Average production
    Median production
    Average capacity utilization
    Plant age range;
            Minimum
            Maximum
    Waste water flow'range:
            Minimum
            Maximum
    Volume per unit product:
            Minimum
            Maximum
                            85,700 kkg/year
                               4
                               4
                           102,200 kkg/year
                            57,000 kkg/year

                               66 percent

                            5,560 kkg/year
                            28,730 kkg/year
                               NA
                               NA
                               NA

                               15 years
                               27 years

                               NA
                               NA

                               NA
                               NA
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Commerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.? Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chemical Industry, " June,  1978 and  "Economic Analysis of Proposed
Revised Effluent Guidelines and Standards  for the Inorganic Chemicals Industry,"
March,  1980.
                                   359
NA = Not Available.

-------

-------
                          SECTION 14



                  TITANIUM DIOXIDE  INDUSTRY

        (RUTILE/UPGRADED ILMENITE — CHLORIDE PROCESS)
14.1  INDUSTRY PROFILE
14.1.1  General Description

     Titanium dioxide  is  manufactured by a chloride process, a
sulfate  process,  and  a  chloride-ilmenite  process.     This
subcategory  is  subdivided  into three  segments,  one for  each
processs because of the difference in raw materials used, waste
water  flows,  and  raw  waste characteristics.   Ti02  is  a  high
volume  chemical,  ranking within  the first  fifty of all  U.S.
chemicals  production.     Over  fifty  percent  of  the  titanium
dioxide produced  is  used in   paints,  varnishes  and lacquers.
About  one  third  is used  in the paper and plastics industries.
Other uses are found in ceramics, ink and rubber manufacturing.

     The industrial  profile data for the  chloride  segment are
presented  in  Table  14-1, while the  status  of regulations  is
given  in Table 14-2.

14.1.2  General Process Description and Raw Materials

     In the chloride process, the raw materials used are rutile
or upgraded  ilmenite ore, which are  relatively pure materials
with a high titanium  and a low iron content.   For  upgrading
ilmenite (FeTiOS), a  beneficiation process removes a part or all
of  the  iron.      Several  patented  processes exist  for  the
beneficiation step and two or three are in current operation on
a commercial  scale.     The   wastes  from the  chloride  process
using  beneficiation  of  ilmenite in titanium dioxide production
are different from those produced using high grade titanium ore
(rutile or upgraded ilmenite).   The Titanium Dioxide Subcategory
has  been  classified  further  into  three  separate  categories:
sulfate  process   using ilmenite  ore,  chloride  process  using
rutile  or  upgraded  titanium  ore,  and  chloride  process  using
ilmenite  ore.    This  section   is  restricted  to  the  chloride
process using rutile ore.

                              361

-------
IftBEE 14-1
SDBCATBQOEY ERCFIIE DKCA SUMMARY

                              DIOXIDE                  )
Total subcategory capacity rate
Stotal subcategory production rate
Number of plants in this subcategory
308 Data on file for
    With total capacity of
    With total production of
    Representing capacity
    Representing production
    Plant production ranges
            Minimum
            Maximum
    Average production
    Median production
    Average capacity utilization
    Plant age range:
            Minimum
            Maximum
    Waste  water flowirange:
            Minimum
            Maximum
    Volume per unit product:
            Minimum
            Maximum
                          610,000 kkg/year
                          389,000 kkg/year
                                5
                                5
                          184,600 kkg/year
                          142,000 kkg/year
                               30 percent
                               37 percent

                           16,900 kkg/year
                           45,200 kkg/year
                           28,400 kkg/year
                           25,600 kkg/year
                               77 percent

                                6 years
                               15 years

                            1,140 cubic meters/day
                            4,770 cubic meters/day

                             29.3 cubic meters/kkg
                              110 cubic meters/kkg
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Commerce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary iksonomic Assessment of Effluent Limitations in the
Inorganic Chemical Industry," June, 1978, and "Economic Analysis of
Proposed Revised Effluent Guidelines and Standards for the Inorganic Chemicals
Industry," March,  1980.
                                   362

-------
TKBXE  14-2 -     STftTOS OP HEGOIATICNS  -  EEET£ENT IJMEEKEECN GUIDELINES
SCffiCXCEGQRX Titanium Dioxide
SUBPAKT -V (40 CFR 415. 220, 3/12/74)
STANDARDS
Product Para-
Process meters
Chloride ^g
Process
Iron
SUlfate _gg
Process
YWM.
BPCTCA* BKCEA*
1 2
Max. Avgr. Max. Avg.
kg/kkg hg/kkg ka/Mcf kg/kkg
(nig/1) ftngA) fcig/1) ftng/1)
4.6

0.72
21.0 ^
(100.0)
1.7
2.3 2.6 1.3

0.36 0.36 0.18
10.5 10.6 5.3
(50.0)
0.84 0.84 0.42
NSPS*
Max. Avg.
kg/kkg kg/kkg
(mg/1) (rag/1)
2.6 1.3

0.36 0.18
10.6 5.3

0.84 0.42
                       (8.1)      (4.0)
 Sections 415.220, 415.222, 415.223, and 415.225 were remanded and are
 presently reserved (41 FR 516CEL, Ifeventoer 23, 19761.
 Ttox, at Maximum of any one day.

  Avg. - Maximum average of daily values for thictv mr>s<=«utive dairs.
**£low basis  210,000 1/kkg.
                                    363

-------
    In the chloride process, the ore and  coke  are dried and then
reacted  with  chlorine  to  form  titanium  tetrachloride.   The
chemical reaction taking place  in the reactor is given as:

     3C + 2TiO2 + 4C12  =   2TiC14 + C02 -I- 2CO                (1)

     The  reaction  takes  place  at  a  temperature  of  800-1000
degrees C  and a fluidized  bed  reactor is  generally used.  The
product  gases  leaving  the   reactor  consist   of  titanium
tetrachloride,  unreacted   chlorine,   carbon  dioxide,   carbon
monoxide and minor  amounts  of heavy metal chlorides.  The gases
are initially  cooled to 250 degrees C to remove the  impurities,
although in  some  cases purification is accomplished by washing
the gases with liquefied titanium tetrachloride.  Iron chloride
and small amounts of vanadium,  zirconium, and other  trace metal
chlorides are  removed by centrifugation and the liquid recycled
to the absorber.  Titanium  tetrachloride is  liquefied from the
gases after  the first stage  of cooling by  further cooling to
ambient  temperature.    Copper, hydrogen sulfide  and,  in some
cases,  proprietary  organic complexing  agents  are  added  for
purification  to  the  condensed  solution.    Copper  acts as  a
catalyst to  decompose  the  phosgene formed in the TiC14 stream.
Organic complexing  agents  aid  in separation  of  the T1C14 from
other   chlorides   such   as    cupric   chloride   and   silicon
tetrachloride.

     The  residual  uncondensed  gases  generally  consist  of
hydrochloric  acid,  chlorine,  carbon monoxide,  carbon dioxide,
nitrogen, and  some  titanium tetrachloride.  They are treated to
remove acidic materials before  being vented to the atmosphere.

     The liquefied  titanium tetrachloride  contains impurities
such as  aluminum  chloride,  silicon  tetrachloride,  etc.,  which
are removed  by distillation.   The distillate  is  the purified
titanium tetrachloride and  the  impurities  remain as a residual
which  becomes waste.   The tail  gases  from  the distillation
column are scrubbed to  remove  acidic materials.   The titanium
tetrachloride  product is   then  reacted  with oxygen,  as air,
forming titanium dioxide and chlorine:

     TiC14 + 02  =  Ti02 +  2C12                              (2)

The rate of  reaction is negligible below  600  degrees  C  but
increases  rapidly  above  this  temperature,  and  is generally
maintained between  1200-1400  degrees C  for  efficient reaction
and conversion.   The  needed heat  is  supplied  by  passing  the
reactants through heat exchangers,  by  electric dischargers, or
by use of fluidized  beds.    After the oxidation reaction,  the
titanium dioxide  forms a solid  and  is  separated  from the gases
either in cyclones,  baghouse filters, or Cottrell precipitators.
The residual  chlorine is refrigerated  and  liquefied.  The tail
                              364

-------
gases are scrubbed with caustic  soda  to remove chlorine before
being vented to the atmosphere.   When  air  is used for oxidation,
chlorine recovery is  achieved by  absorption in trichlorethylene,
followed  by distillation  to  remove  chlorine.    The  titanium
dioxide  is  then sent  to  the finishing operation where  it is
vacuum degassed  and  then treated with  alkali,  using a minimum
amount  of  water  to  remove traces  of  absorbed  chlorine  and
hydrochloric acid.  The pigment  is then milled, surface treated
for  end-use application,  dried, and packaged  for  sale.     A
generalized process  flow diagram,  including  the waste streams,
is shown in Figure 14-1.


14.2  WATER USE AND WASTE SOURCE CHARACTERISTICS


14.2.1  Water Use

Water Use

     Water is used in noncontact cooling,  for  scrubbing the  tail
gases  from  the  purification  and  oxidation  reactor  to  remove
contaminants,  and  in  the  finishing operation  of  the product.
The  total amount of  water  usage  varies  from  45.3 to 555 m3/kkg
of Ti02 produced, as  shown in Table 14-3.  The  table also shows
that cooling water  constitutes  the major use of  water and varies
from 10.7 to 426 m3/kkg of Ti02  produced*

14,2.2  Waste Sources

Wastes from Cooling Chlorinator  Gas

     The  waste consists of  solid  particles  of unreacted  ore,
coke, iron, and small  amounts of vanadium, zirconium, chromium,
and  other heavy metal chlorides.  They are either dissolved in
water and sent to the waste water treatment facility or disposed
of in landfills as a  solid waste.

Chlorinator Process Tail Gas Scrubber Waste

     The uncondensed  gases,  after  the liquefaction of titanium
tetrachloride,  are initially wet  scrubbed to  remove hydrogen
chloride,  chlorine,   phosgene,  and titanium  tetrachloride  and
chlorine.   In a second stage,  they are  scrubbed  with caustic
soda to  remove chlorine as hypochlorite.

Distillation Bottom Wastes

     These  contain  copper, sulfide,   and  organic  complexing
agents  added  during  purification in  addition   to  aluminum,
silicon,  and  zirconium  chlorides.    These   are  removed  as
waterborne wastes  and reaction with water converts silicon and
anhydrous aluminum chlorides to  their respective oxides.

                              365

-------
                                OR

                              awns one
                                                           ATHOSPHEHS
                                                        HSSBS
                                                        HATER
Figure 14-;!. •  General process diagram  for production of titanium dioxide
               (chloride process) from  high grade ores.
                                   366

-------
     14-3.        USAGE IN          DIOXIDEHSHLQRIDE PROCESS/HIGH


            ORES SUBC&TEGORY
                         Water usage at plants


   Water Use               (m3/kkg of TiO_)
                                         «£

                               Water Use
Plant #102
Noneoataat cooling
Direct process contact
Indirect process contact
Maintenance, equipment
cleaning and wark area
wasbdown
Mr pollution control
Istoncontaofc ancillary uses
Sanitary & potable water
fbtal
182
10.5
NA
6.65
0.25
11.60
0.23
211.23
Plant 1172
10.66
15.53
0.72
0.52
7.14
10.4
0.31
45.28
Plant #199
426
73.2
26.5
2.80
11.3
9,5
5.6
554.9
N& = Not available
                                 367

-------
Oxidation Tail Gas Scrubber Wastes

     The   gases   from   the   oxidation  Unit   are   cooled  by
refrigeration to liquefy and  recover  chlorine.   The uncondensed
off-gases  are scrubbed  with water  or  caustic soda  to remove
residual chlorine.  When caustic  soda is used as  the scrubbing
solution,  the  resulting  solution  of  sodium  hypochlorite  is
either  sold,  decomposed,  sent ,to  the  waste  water  treatment
facility,  or  discharged without treatment.  The scrubber waste
stream also contains titanium dioxide particulates.

Finishing Operations Waste

     The  liquid  wastes  from the finishing  operation contains
titanium  dioxide as  a  suspended solid  and  dissolved  sodium
chloride  formed  by  the  neutralization of  residual HC1  with
caustic soda.

     The  range  of  waste  water  flows  requiring  treatment  is
summarized in Table 14-4.  The wide range of  flow occurs because
some plants use  additional water  to wash solid process residues
to the waste treatment system.


14.3  DESCRIPTION OF PLANTS VISITED AND SAMPLED


14.3.1  Screening

     Plant =P559  was visited  and the  waste effluents sampled in
the screening phase of  the program.   Plant #559 makes titanium
dioxide using both the sulfate and the chloride processes.  The
waste waters from both processes  are mixed and undergo combined
treatment.

     The solids  from  the chloride process,  called  pit solids,
(mainly unreacted ore,  coke,  iron,  and trace  metal  chlorides
including  TiC14)  are  separated from  the first stage cooling of
the chlorinated gases  and are slurried with water.  The slurried
pit solids and the distillation column bottom residue effluents
from the  chloride process are sent  to  a large  settling  pond
(called the weak acid pond)  where they are mixed  with the weak
acid from  the sulfate process.  The  overflow from the settling
pond is neutralized with ground calcium carbonate in a reactor.
The reactor  effluent  is  filtered,  aerated  to  remove  iron and
combined with neutralized  strong-acid waste effluent (from the
sulfate process).   The  combined  scrubber and  contact  cooling
waste water  from both  sulfate  and  chloride processes  is  also
combined at this  point.   The  combined  waste water is neutralized
and solids settled out  in a pond prior  to final  discharge,   A
flow diagram  of  the  treatment facility  including  the sampling
locations  is shown in Figure 14-2.

                              368

-------
ISfflEE 14-4.  mSEE       FLOW K)R          DIC3XID1-CHIJORIDE
             SUBC&TEGORY
SUBCATEQORY           TITANIUM DIOXIDE  CChlQrjx3e Process)
Plant                    Unit Ifeste Water Plow Going to Treatment Plant
  .                                        (m3/Mcq of TiD2)


102                                   29.3(1)

172                                   34.7(1)

559                                   91.0(2)

199                                  110.0(2)
(1)  Offsite disposal of process solid residues.
(2)  Process solid residues are slurried to «aste treatment.  The average
     flow of Plants #559 and #199 »s used as the model plant flow for
     cost estimating and regulation development.
                                    369

-------
KlVtll
SWPU MATE*'

MUNICIPAL
SUPPLY WATER

uai
SUPPLY WATER
e.
   vo
U)
<1
o
              OTHER PRODUCT
               (WEAK ACID)  '
               WASTE VATEft
                                 SAI1ITAKY
                             AMD TIO- FIHISHIHG
                             AMA HASTE VATEP.
                                                                  OTHEK PRODUCT
                                                                  (STRONG ACID)
                                                                   VAST! WATER
SETTLING
URRIEO —p.
T SOLIDS
PCN
|9

i
«
^.ri-i Dl
	 	 	 .„„„., ft:
\> £-*.

I
STIILATIOM BOTTOM
WASTE WATER
?°
                                                                                                         FINAL EFHUiMT
                                   e
                          LEGEND

                        SAMPLING POINTS
                                                                                CHLORIDE PROCESS SCRUBSEH
                                                                                      WASTE WATER
                                                                     OTHER PRODUCT
                                                                      WASTE HATiR
    Fimire 14-2
                  *
                                      General &cw diagram at  Plant #559 showing the sanpling points.
                                       (Titanium dioxide — chloride process manufacture.)

-------
     Problems were  encountered  during the sampling  of  the pit
solids  and  the  distillation bottoms.   The pipes  carrying the
wastes from the  process  discharged  at  the bottom of the settling
pond and  it  was not possible to take  the  samples  right at the
outlet of the pipe.  The combined sample of the two streams was
taken at the surface of  the discharge.  It  is probable that some
solids settled before the stream reached the surface.  Table 14-
5 gives  the  waste  flows and pollutant loadings for the streams
sampled at Plant 1559.  Because of the intermixing of the waste
effluents  from both   chloride  and   sulfate  processes,  the
pollutant   loadings  in  Table   14-5  were   calculated   by
proportioning according  to the  relative hydraulic loadings.

14.3.2  Verification

     Plant f!72  was sampled in the  verif-ication phase.  Titanium
dioxide is made at  this facility by the  chloride  process only.
The strong acid wastes  and the  spent  coke  and ore residues are
hauled  to a  secure chemical landfill  for  disposal.   The waste
water from the  process, mainly  the scrubber water, is collected
in  trenches  and   sent   to  a  central reactor basin.    Other
discharges, including a part of  the total  rain runoff, are also
collected in  ditches and  sent  to  the reactor basin.   In the
reactor basin,  sodium hydroxide  is used for neutralization, and
the resulting effluent  is mixed with  the  remaining  rain water
runoff and sent  to the first of  two retention basins arranged in
series.  The overflow from the  second retention  is pH adjusted
with sulfuric acid before discharge.   A  simplified  diagram of
the treatment system, including  the sampling points, is shown in
Figure  14-3.    Table 14-6 gives the  waste flow  and pollutant
loadings for the streams sampled.

14.3,3  Toxic Pollutant Concentrations

     Five toxic pollutants  were found above  the  treatability
levels in the raw waste  of plant $559.   It  is possible that some
of the pollutants might  be from  the sulfate process waste water
as the two raw  waste effluents  are intermixed  before treatment.
One pollutant was found  above the treatability level in the raw
waste of  Plant   fl72.   No organic toxic  pollutants  were found
above treatment levels  in  the  raw  wastes of  either plant.  The
maximum concentration of the toxic pollutants  found  in the raw
waste   in  significant   concentration in   the  screening  and
verification program weres
                              371

-------
     OBOE 14-5.  FICM SND KEIW»W CSXCSOimnm D?3& OP THE SfiMELED
   STEAMS OF PWOTfl72 BKTOCING IHMmM XttCBOBE W CHMSCDDe-KOTIIE HCCESS
STREAM |2 STREAM f 5
(li
CAICULOTED ESSJKKEB SERESM f6
Bit a>lidsand Scrufcter and • IWal Raw WSste Treated Effluent
Distillation Bottoms Contact Cooling Witet *«»«« ^.ij.ucin.
A B" C D 1
RAlutant (ftxBsKf3)
Ohit Flow OXKJ. , Unit Load (Unit Flow) Gone.
(nfl/kkg) (tng/I) (kg/Hcg) {m3/Kkg) (mg/1)
10.9 80.1
TSS 6903 75.2 314
Iron 1348 14.7 143
w chromium 112 1.2 0.11
to
Lead 3.53 0.04 0.009
Nickel 3.46 0.04 0.016
Zinc 2.12 0.02 0.13
F G H I J K
(DxExlO"3) (AH)) (CW) (KdO~3/G)
ttoit toad ttiit Flow Unit Load Cone. Unit Flow Oonc.
(kg/Beg) (ra3/lckg) (kg/Meg) tog/1) (nQ/kkg) 6ngA)
91 91
25.2 100.4 1103 23
11.5 26'2 mB 4.4
0.01 1.21 13.3 0.03
0,001 0.041 0.5 0.002
0.001 0.041 0.5 0.005
0.01 0.03 0.3 0.06
I.
Unit toad
(kg/kkg)

2.1
0.4
0.003
0.0002
0.0004
0.005
(1) See Figure 14-2 for location of sampling points

-------
                                    PROCESS
                                  WASTE WATER
           HOLDING POND
               FOR
           RETREATMENT
                                          #1
                                  .RETENTION
                                     BASIN
                                   RETENTION
                                     BASIN  -
                                          ^,
                                          #3
                                   DISCHARGE
                                                  NaOH
                                    MIXING
                                    BASIN
                                  NEUTRALIZE
                                        RAIN RUNOFF
                                  RAIN RUNOFF
                                                 pH ADJUSTMENT
                                                        LEGEND
                                    ft SAMPLING  POINTS
                                    ^
                                     *   THE TOTAL  RETENTION TIME
                                        OF WATER  IN  THE  TWO- PONDS
                                        IS 5  DAYS.
Figure
General flow diagram at Plant #172 showing the sampling points.
Titanium dioxide (chloride process) manufacture.
                                      373

-------
   TABLE 14-6.  FLOW AND KHXUTAMF CONGEKmATE)N mm OP THE  SftMELED
             FOR       §172                    DIOXIDE  (CHLORIDE
             SAMPLED STREAM #1
                           SAMPLED STREAM #3
JEtollutant    Saw Wfe.sfce Influent
                           Treated Effluent
              A          B           C-D          E            F   ,
                                 (A+BxlO  )                          (D+ExlO  )
          Unit Plow  Avg'. Cone.  Unit Load  Unit Plow  Avg, Cone.   Unit Load
                                  (kg/kkg)    (roVkkg)     (rag/1)        (kg/kkg)
             34.7
TSS
171
                       34.7
                                   5.93
6.7
0.23
Iron


Chromium


Lead


Nickel
                        2.9
                        0.72
                        0.005
                        0.08
           0.10
           0.03
           0.0002
           0.003
0.33


0.02


0.002


0.01
0.01


0.0007


0.00007


0.0003
Zinc
0.3
                                   0.01
0.09
0,003
                                     374

-------
            Maximum Raw Waste Concentrations Observed  (jig/1)
Pollutant
Plant 1559
Chromium
Lead
Nickel
Zinc
Screening
Plant f!72
152,000
5,150
6,320
3,300
Verification
1800
NS*
NS
NS
*NS - Concentration was found below the lower limit of
      treatability;  i.e., not significant.

     The  screening and verification  sampling program  and the
methodology used have  been  described in Section  5.1.2  of this
report.  A total of six days  of sampling was conducted at Plants
1559  and  f!72.   Five  effluent waste  streams were  sampled  at
Plant  |559  and  three  streams  were sampled at Plant  #172.  At
each  sampling  point,   three  24-hour  composite  samples  were
collected for analysis.   The evaluation of toxic metal content
of  these  process   related   waste   streams  was   based  on  550
analytical  data   points.     The   average  unit  loadings  and
concentrations  for  conventional,   nonconventional,  and  toxic
pollutants found in the raw waste effluents for Plants f559 and
1172 are given in Table 14-7.

     The  total  quantities of  toxic pollutants  generated each
year  for  this  subcategory  (calculated  as   total  subcategory
production times average  unit toxic pollutant  load  from Table
14-7) are as follows:

          Pollutant            Waste Load  (kg/year)

          Chromium                    241,000
          Lead                          8,200
          Nickel                        8,500
          Zinc                          7,800


14.4  POLLUTION ABATEMENT OPTIONS


14.4.1  Toxic Pollutants of Concern

     The dominant  toxic pollutant in untreated effluents in the
Titanium  Dioxide   (chloride  process)  Subcategory  is  chromium.
Chromium  was  found in treatabl.e concentrations  at  both plants
sampled in the screening and verification phase.  Lead, nickel,
and zinc were'found in the raw waste of Plant 1559 at treatable
levels, but- were not present in the Plant $172  raw  waste.  At
Plant 1559, the chloride process waste effluents  are mixed with
                              375

-------
      14-7.   RAW msiE
        mtm        OF THE SAMPLED
SUBCMBGORY:
DIOXIDE (CHDORIDE PROCESS)
Average Daily
Pollutant
Toxic •
Iron
Cteomium
Lead
Nictel
Zinc
Conventional!
TSS
Pollutant Concentration and
(kg/kkg of Ti02)
(mg/1)
Plant
1559
26.2
(288)
1.21
(13.3)
0.041
(0.5)
0.041
(0.-5)
0.03
(0,3)
100.4
(1103)
Loadings at Plants
Plant
1172
0.10
(2.9)
0.03
(0.72)
0.0002
(0.005)
0.003
(0.08)
0.01
(0.3)
5.93
(171)
Sampled
Overall
Average
13.15
0.62
0.021
0.022
0.02
53.17
          376

-------
the sulfate  process waste effluents  before treatment.   It is
likely  that  the  three  major  toxic  pollutants  found  were
contributed  by  the sulfate  process  wastes, as  it uses  a low
purity ore (ilmenite),  At Plant 1172,  the  solids generated from
the  chloride process  (which  consist  of   solid  particles  of
unreacted ore, coke, iron, and other  heavy  metals)  are hauled to
a  landfill  for  disposal.   It seems  probable that  the  three
pollutants  are  present  in  this solid  waste and  hence  do not
appear in the waste waters.

14.4.2  Process Modi£i cation and Technology Transfer Options

     1.  Research  to  develop economical techniques  to recover
the  vanadium and  other  metal values  from  the   solid  wastes
generated from  the process  waste treatment system would appear
to be a fruitful area of  investment.

     2.  New plants can utilize refrigeration and high pressures
for chlorine liquefaction.   This would  reduce  or  eliminate the
chlorine residual  problem in the tail gases.  The capital cost
to modernize old plants is high, but  these plants should have a
caustic  soda or lime  scrubber  instead of  a water scrubber to
remove residual  chlorine  from  the  tail  gases. . Caustic or lime
scrubbing removes a significant portion of  the  chlorine from the
tail gases  as seen from the  analagous  data for  the chlorine
subcategory given  in Section 11.

14.4.3  Best Management Practices

     Provision  should be made  at  all plants  to  collect storm
water runoff  from  the plant site and send  it  to  the treatment
facility.   Three  out of a  total  of five  existing  plants are
presently treating storm  water runoff.

14.4.4  Prevailing Control and Treatment Practices

     At Plant §172, the solid wastes consisting of spent ore and
coke are hauled  to an off-site landfill.   Process waste waters
consisting of scrubber and contact-cooling  effluents and a part
of the surface runoff are sent to a mixing  basin where they are
neutralized with caustic  soda.   The  effluent from the basin is
then  sent  to  two  retention  ponds in  series.    Additional or
residual  rain   water  runoff  is  added   to   the  ponds  for
clarification.  The overflow from the last  pond is monitored and
discharged to a surface stream.  At Plant ^559, the waste waters
from both chloride and sulfate processes  are mixed and treated
together.  The  distillation bottoms and the unreacted  ore and
coke from the chloride process are combined with the weak acid
effluent from the  sulfate process in a pond.  The overflow from
the pond is neutralized with limestone and  oxidized with air for
the removal  of  iron.    The waste water is then mixed with the
                              377

-------
neutralized  strong  acid  waste  (from the  sulfate  process)  and
scrubber waters  (from both  the  chloride and sulfate processes)
and  neutralized  with  lime  in  a reactor  and  sent  to  a final
settling  pond.   The  overflow  from  the  pond  is  the  final
discharge.

     At Plant  ^199,  all  the process waste waters are combined,
including storm  water  and sanitary waste  water.   The combined
was.te water  is sent  to a four-stage neutralization  system, and
the effluent from each of the four  stages of neutralization is
sent to a thickener.   The thickener overflow is transferred to
the  first  of  three  settling  ponds,  also  in  series.    The
underflow from the thickener is heated to improve  its filtration
characteristics  and filtered  in  four  rotary drum  filters.  The
thickened solids from the filters are disposed of in a landfill
and the filtrate is  combined with wash  water,  and  vacuum pump
seal water  prior to  being recycled to  the fourth stage of the
neutralization train.  The overflow from the last settling pond
is discharged.

     The process waste water streams at Plant #102 are received
in two tanks, neutralized with lime,  and then sent to a settling
basin.  The settled  solids are retained  in the settling lagoons.
The plant  has  future plans  for treating  boiler  blowdown,  and
cooling tower blowdown, leaks and spills with the process waste
water.

     At Plant 1605,  the unreacted ore and coke  is disposed of as
a solid waste  in the pit.   The  waste water from the process is
passed to two tanks  for  flow equalization,  and  the water  is then
reacted with ground  limestone slurried  in  water.   The treated
solution is centrifugally treated to remove  coarse solids which
are separated and landfilled.  A flocculating agent  is added to
the  centrate and the  solution  is  sent  to a clarifier.   The
clarifier overflow is degassed and the pH  adjusted with caustic
soda (if required) before discharge.

14.4.5  Advanced Treatment Technologies

     Neutralization and settling are practiced  for the treatment
of chloride process raw  waste  effluents at all the  five plants
for  which  308   data  are available.    Air  oxidation,- sulfide
precipitation, xanthate precipitation, and ion exchange might be
applied  to  the  clarified  solutions  for  control   of  metals.
Sulfide precipitation or  the  xanthate process  could  be used to
provide additional removal of zinc, lead,  and nickel*
                               378

-------
14.5 SELECTION OF APPROPRIATE TECHNOLOGY AND EQUIPMENT


14.5.1  Technologies for Different Treatment Levels

Level 1  (BPT)

     The chloride process wastes are equalized, neutralized with
lime  to  a pH range of  6 to  9,  and settled  in  lagoons before
discharge.   Level  1 treatment  is  typical of industry practice
and for this reason was  selected as  the  technology basis for BPT
regulations.

Level 2

     Alkaline precipitation as a second-stage lime treatment to
an  optimum pH  (9  to 10)  is  added  to  Level 1  to  precipitate
metallic  hydroxides, which  are  then filtered before discharge.
Filtration  removes  traces of metallic  hydroxides which do not
separate   in  a gravity  system.     Level  2  technology  was
utilimately  selected   as   the   basis   for  the  proposed  BAT
limitations  because it  provides  an economical  method  for the
removal of additional toxic metals.

Level 3

     Ferrous  sulfide  treatment   is  added  ahead of  the  Level 2
filter to precipitate the heavy metals more effectively.

     Alkaline  precipitation was chosen as Level 2  because it
readily  supplements existing  lime  neutralization by the simple
addition  of  filtration  and  increasing the Level 1 lime dosage.
Sulfide precipitation was chosen at Level  3 because it provides
a  polishing treatment  for  most  residual heavy  metals beyond
Level 2  treatment.

     Figures  14-4, 14-5,   and  14-6  show the model  treatment
systems  adopted for the  chloride process wastes.

14.5.2  Equipment for Different Treatment Levels

Equipment Functions

     BPT   treatment   is   essentially   lagooning  with   lime
neutralization, using no special equipment except a lime feeder
and mixer.

     In  Level  2,  second stage  lime treatment  is  followed by
gravity   clarification   and   multi-media   filtration,   with
necessary pH controls.
                               379

-------
                                             LIME
          RAW
     WASTE WATER
                  — \ "   JLAGOON JL — .
                        LAGOON  / —
*9
W
                                         MDttNG
           LAGOON
                                                                                        *o
                                                          •*»
                                                                                              EFFLUENT
           LAGOON
                   >J
w
00
o
                       Includes flow monitoring, pH monitoring and sampler.
                 Figure 14-4.   Level, 1 waste water treatment  for titanium dioxide  — chloride

                               process.

-------
                                                           BACKWASH
          RAW
        WASTE WATBi
U)
00
                                                                                      pH ADJUSTMENT
                                                          SUMP
                                             TO LANDFILL



                      'ncludes flow monitoring, pll monitoring and sampler.
           Figure 14-5.  Level 2 waste water treatment for titanium dioxide  — chloride
                           process.

-------
-A   LAGOON  ym»-
         RAW
      WASTE WATER
                       LAGOQN
oo
                                    r
LIME       ±

  !V
                                                   LAGOON
                                       MIXING
                                 JLAGOON
                                                                        FEilROUS      SODIUM
                                                                        SULFATE     BISULFIDE
                                                               BACKWASH
                                                                                         4
                                                                                 t-
0
                                                   FILTER PRESS
                                                                 SUMP
                                                   TO LANDFILL

                        Includes flow monitoring, pH monitoring and sampler
                                                                                  r
                                                               'h^l
                                    SUMP    DUAL
                                            MEDIA
                                            FILTER
                                 CLARIFIER  -,
                                                                                            * EFF.
            Figure 14-6.   Level 3 waste water treatment for titanium dioxide
                          process.
                                                             chloride

-------
     In Level 3,  ferrous  sulfide  is added ahead of the Level 2
filter, to react with residual  heavy metals more completely than
in the alkaline precipitation  step  at Level 2.

Chemicals and Handling

     Lime  and  hydrochloric  acid  are  fed  with  conventional
equipment at all levels, and ferrous sulfide is prepared on-site
by mixing  ferrous sulfate  with sodium bisulfide.   When normal
dust control and  good ventilation are used,  there should be no
adverse  effects  from handling these  chemicals,  although care
should be taken that hydrogen  sulfide gas is not generated.

Separation and Removal of Solids

     Inert ore fractions and precipitated solids are accumulated
in clay-lined  lagoons,  which  are  alternately  drained.   Solids
are mechanically  removed  to self-draining 18  ft.  high storage
piles  on land  provided  at  the site  for a  10-year  operating
period.   At  Levels  2  and 3,  small  amounts  of  heavy  metal
precipitates in the  clarifier  underflow are  filter pressed and
hauled to a secure landfill.
14.6  TREATMENT COST ESTIMATES


14.6.1  General Discussion

     To determine the treatment cost, a model plant concept was
developed.  A raw waste unit flow  was  selected and pollutants to
be  treated  were selected,  based  on  the  treatment  system data
available  for  the   five  TiO2  plants  and  the  screening  and
verification  sampling program.   Three production  levels were
then selected to cover the  entire subcategory range.  Treatment
costs  for Levels 1, 2, and 3 were calculated for  each of the
model  plant production  ranges using  the unit  flow  and unit
pollutant loads.   The preliminary  cost data given  in the cost
tables and  figures  were generated using a low unit  flow of 31
m3/kkg of Ti02 based on incomplete  industry  data.  The new unit
flow  of  100  m3/kkg  used   for  the model  plant  in  regulation
development has  been selected to  be more representative of the
subcategory and  it  is assumed that the unreacted ore and coke
are slurried and sent to the  treatment system, instead of being
disposed  of in  a landfill as a  solid  waste.   The  need  for
revising the  preliminary  cost estimates  is  being  evaluated by
the Agency  and  any  appropriate changes  will be made prior to
promulgation.   The  model  plant  specifications  presented here
were used in regulation development.
                              383

-------
Waste Water Flow

     The  unit waste  effluent flow  varies  from  29.3  to 110.0
m3/kkg of Ti02 for the four plants as shown in Table 14-4.  The
primary reason for the variation in the flow is  that some plants
slurry  the  spent ore  and  coke  (solid waste  from  chloride
process)  and  send it to  the  treatment  system,  and others haul
the  dry solids  to a  landfill.    The  flow variation  is  also
dependent  on the  difference  in the  chlorine  recovery  process
from  the tail  gas and  the amount  of  scrubbing  liquid used.
Small  variations  in  flow  also  result   from   the  finishing
operation which is dependent on the type of  titanium dioxide end
product desired.  Plants #559 (unit flow of  91 m3/kkg) and #199
(unit  flow  of  110  m3/kkg)  sent  the  solid   waste  from  the
manufacturing process to  the treatment  facility.  It is.assumed
for  treatment  system  cost  estimation that   the  solids  are
included  in the  raw waste  flow to  the treatment system.   A
constant unit flow of 100 m3/kkg of  Ti02 has been used  for the
model plants, which  is  an average of  the unit  flows  of Plants
f559 and #199.

Pollutant Load

     The  primary  pollutants occurring  in  the  waste  water are
suspended  solids,  acidity, and  the  chlorides of  ferric iron,
chromium  and  other trace  metals.   The  suspended solids  (TSS)
loading  values  for  Plants #559  and #172  are  100.4  and  5.93
kg/kkg of TiO2  (Table 14-7).   The low value represents  a plant
that  hauls ore  and coke off-site,  while  the  high  value  is
believed to be due to nonrepresentative sampling.  The amount of
solids produced  are higher  than the values  indicated  for, the
sampled plants.   Consequently, a higher  suspended  solids  loading
of 500 kg/kkg of TiO2 ("reported in the 308 data  from Plant #199)
is  assumed  for  the model  plants.    To   establish  treatment
chemical  requirements  and related  costs,   the  toxic  pollutant
loadings for the model plant are taken  as the average values of
the  unit  pollutant  loadings  of  the  plants   sampled   in  the
screening and verification program (Table 14-6)  and the selected
pollutant values are:

         Pollutant           Unit Loading (kg/kkg of Ti02)

         Chromium                         0.62
         Lead                             0.021
         Zinc                             0.020
         Iron                            13.15
         Nickel                           0.022

Production Rates

     Five  plants  produce  titanium  dioxide  from  rutile  ore or
ilmenite ore, using  the chloride process at a total production

                              384

-------
rate of 142,000 metric tons per year.  Production ranges from a
minimum of 16,900 kkg/year to a maximum of 45,200 kkg/year with
a mean of 28,400 kkg/year and a median of 25,600 kkg/year.  For
waste water  treatment  cost estimates, three  production levels
were selected as model plants.  These are 16,900 kkg/year, 25,500
kkg/year,  and  45,200   kkg/year.    This  range  of  production
includes all United States plants.

     The estimated costs  for  the  three models having different
production levels  are  given in Tables 14-8, 14-9,  and 14-10.
Annual  treatment  costs as a  function of production  are shown
graphically  in Figure  14-7.   Similarly,  treatment  costs  per
metric  ton  of product  are given  in Figure 14-8.   Table 14-11
presents  a  summary  of  the  unit  cost  distribution  between
amortization, and the operation and maintenance cost components
at various production  rates and levels of treatment.  The costs
shown at each  level  of treatment  correspond to the model plant
BPT (Level 1} system  and higher level  (2 or  3) systems which may
add to or modify the existing BPT system to meet more stringent
toxic pollutant removal requirements.  The higher levels (2 and
3) also  furnish a better effluent  quality  with  respect to the
conventional and nonconventional  parameters.  For  model plants
at the base level of treatment, amortization, chemicals and the
residual waste disposal costs  have  a  significant  impact on the
total annual costs.  At treatment levels 2 and 3, amortization,
chemicals and labor constitute  a major  portion of the additional
annual costs.


14.7  BASIS FOR REGULATIONS


14.7.1  Evaluations of BPT Treatment Practices

     All the plants  producing  titanium dioxide by the chloride
process  using  rutile  ore or  upgraded  ilmenite ore  practice
neutralization  and  settling for  control  and treatment  of  the
waste  effluents.    A  variation  in  the  effluent  quality  is
expected because of the method  of  handling the unreacted ore and
coke  (generated  as solid  residue from the  chloride  process).
Two of the five plants  haul the residue to a secure landfill for
disposal while the remainder slurry the residue  with  water and
send  it  to the treatment system.   No information  is available
about recycling the treated waste water at any of the plants.

Pollutant Removal with BPT Treatment

     Treatment Level 1  is equivalent to the proposed BPT in the
TiO2 subcategory (chloride process).

     Plants |559 and f!72 practice neutralization  and settling
of the raw  waste.  At Plant 1559,  the chloride process raw waste

                              385

-------
                    TABLE 14-8.    MODEL PLANT TREATMENT COSTS
   Subcategory  TITANIUM DIOXIDE-Chloride Process

   Production        16,900 metric tons per year   (18,632 tons per year)
                         48 metric tons per day    (53 tons per day)
   Waste water flow    1485 cubic meters per day.


                                             LEVEL OF TREATMENT*

                                           FIRST            SECOND
A.  INVESTMENT COST

    Construction-	              $368,500           $49,000
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls	               209,000           389,000
    Monitoring equipment
    in place	                 9,000
    Engineering design
    and inspection	               117,300            87,600
    Incidentals, overhead,
    fees, contingencies...               117,300            87,600
    Land	               192,000             6,000

    TOTAL INVESTMENT COST             $1,013,100          $619,200

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision.               $56,000           $84,000
    Energy	                 3,700             4,300
    Chemicals	               140,000            34,100
    Maintenance	                82,110            61,320
    Taxes and insurance...                30,393            18,576
    Residual waste
    disposal	               108,000             9,000
    Monitoring, analysis
    and reporting	                15,000             7,500
    TOTAL OPERATION AND
    MAINTENANCE COST

C.  AMORTIZATION OF
    INVESTMENT COST

    TOTAL ANNUAL COST
$435,203


$133,592

$568,795
$218,796
 $99,767
$318,563
    *First level represents the base cost of treatment system.
    Other levels represent the incremental cost above base cost.
                                    386

-------
                    TABLE 14^9.    MODEL PLANT TREATMENT COSTS
   Subcategory  TITANIUM DIOXIDE-Chloride Process

   Production        25,500 metric tons per year   (28,113 tons per year)
                         72 metric tons per day    (80 tons per day)
   Waste water flow    2240 cubic meters per day.


                                             LEVEL OF TREATMENT*

                                           FIRST            SECOND
A.  INVESTMENT COST

    Construction	              $525,000           $50,800
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls	               228,000           450,000
    Monitoring equipment
    in place	                 9,000
    Engineering design
    and inspection	               152,400           100,160
    Incidentals, overhead,
    fees, contingencies...               152,400           100,160
    Land	               276,000             6,000
  i?

    TOTAL INVESTMENT COST             $1,342,800          $707,120

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision.               $56,000           $84,000
    Energy	                 4,000             5,500
    Chemicals	„	               211,000            51,000
    Maintenance		               106,680            70,112
    Taxes and insurance...                40,284            21,213
    Residual waste
    disposal	               164,000            11,000
    Monitoring, analysis
    and reporting...	                15,000             7,500

    TOTAL OPERATION AND
    MAINTENANCE COST                    $596,964          $250,325

C.  AMORTIZATION OF
    INVESTMENT COST                     $173,568          $114,072

    TOTAL ANNUAL COST                   $770,532          $364,397


    *First level represents the base cost of treatment system.
    Other levels represent the incremental cost above base cost.

                                     387

-------
                 TABLE  14-10. MODEL PLANT TREATMENT COSTS
Subcategory  TITANIUM DIOXIDE-Chloride Process

Production        45,200 metric tons per year   (49,833 tons per year)
                     129 metric tons per day    (142 tons per day)
Waste water flow    3980 cubic meters per day.
                                          LEVEL OF TREATMENT*

                                        FIRST            SECOND
A.  INVESTMENT COST

    Construction 	
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls	
    Monitoring equipment
    in place	
    Engineering design
    and inspection	
    Incidentals, overhead,
    fees, contingencies...
    Land	

    TOTAL INVESTMENT COST

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision.
    Energy	
    Chemicals	
    Maintenance	
    Taxes and insurance...
    Residual waste
    disposal	
    Monitoring, analysis
    and reporting	

    TOTAL OPERATION AND
    MAINTENANCE COST

C.  AMORTIZATION OF
    INVESTMENT COST
                  t

    TOTAL ANNUAL COST
                                     $815,500



                                      283,000

                                        9,000

                                      221,500

                                      221,500
                                      504,000
                                   $2,054,500
                                      $56,000
                                        4,600
                                      374,000
                                      155,050
                                       61,635

                                      294,000

                                       15,000


                                     $960,285


                                     $252,266

                                   $1,212,551
 $76,800
 590,000
 133,360

 133,360
   6,000

$939,520
 $84,000
   7,650
  95,000
  93,352
  28,185

  20,000

   7,500


$335,687


$151,883

$487,570
  *First level  represents the base cost of treatment system.
  Other  levels  represent the incremental cost above base cost.
                                  388

-------









2n





,_,
o
o
o
o
0 1 5
rH
X


EH
O
o
EH
8
§ 1.0

3

H
g
S



0.5


























































































































































































































































































































































































































































•4





<3










































/





/









































y
^





/

.







































/
s





/









































Ci
'





(•'









































/






/
f








































/






{
f








































/







/








































/







/








































/







/







































t
/







/







































/
/







/







































/
/







/







































/
'








/







































<£









-------
60
50
40
                         I
30
                         ffl
                                               I El
rat.
il
20
           10       20       30        40        50

             ERODOCTICN .(METRIC TONS/YEAR X 1000 )
 Figure 14-8.  Annual unit treatment cost vs. production for the
      Titanium Dioxide Subcategory, Chloride Process

                           390

-------
                 TABLE 14-11.  MODEL PLANT TREATMENT COSTS
Subcategory  TITANIUM DIOXIDE-Chloride Process
                                           Annual Treatment Costs {$/kkg)
COST HEM
PRODUCTION   FLOW
   (kkg/yr)  (m3/day)
                    LEVEL OF TREATMENT

           FIRST     SECOND    THIRD    FOURTH
Annual Operation
and Maintenance
Annual
Amorti zation
Total' Cost
    16,900
    25,500
    45,200
1,485
2,240
3,980
16,900
25,500
45,200
16,900
25, 500
45,200
1,485
2,240
3,980
1,485
2,240
3,980
25.75
23.41
21.25
                        7.90
                        6.81
                        5.58

                       33.66
                       30.22
                       26.83
12.95
 9.82
 7.43
                     5.90
                     4.47
                     3.36

                    18.85
                    14.29
                    10.79
13.27
10.09
 7.65
                     6.07
                     4.60
                     3.47

                    19.33
                    14.68
                    11.12
   Ntot
Applicable
                                 391

-------
water  is  mixed  with  the   sulfate  process  waste  water  for
treatment.   Also at Plant 1559, the spent  ore  and coke  (solid
residues from the chloride process) are slurried with water and
sent to the treatment facility whereas at Plant ^172, the solid
residues are hauled to  a  chemical  landfill.   Long-term treated
effluent data have been submitted by both Plants 1559 and 1172.
The derivation of the variability  factors  for  daily and 30-day
averages for both plants  are given in Tables 14-12 and 14-13.

     The concentration  of the  raw waste and  treated  effluent
along  with  the  percent  removal   of   the  pollutants  by  the
treatment  system for  Plants  1559  and 1172   sampled  in  the
screening and verification program are given in Table 14-14.

14.7.2  Basis for Proposed BPT Effluent Limitations

Technology Basis

     For  BPT,   the  Agency  is  proposing  limitations  based on
equalization,  neutralization,   and  settling or  clarification.
All plants  in  this  segment  of the  industry have BPT technology
installed.

Plow Basis

     The flow going to  the treatment system at different plants
varies and  is  dependent on  the  method  of disposal of the spent
ore and coke  (pit solids) and on the finishing operation.  The
spent  ore  and  coke are either hauled  to  a landfill  as  solid
residue or sent to the treatment system.  For the purpose of the
model  plant  treatment system, the solid   residues  from  the
manufacturing   process  are assumed to  be  slurried  with water
and sent to  the treatment system.  Plants f559 and f!99 do, in
fact,  send  the  solid residues to  the treatment  system.   The
model  plant  treatment system  is  based  on an inflow rate of 100
m3/kkg of Ti02 which is  an average  value of  the effluent flow of
Plants §559  and  |199.   The  treated effluent flow is assumed to
be  the same as  the  influent   flow.    The   water  added   or
removed  in  the   treatment  system through chemical addition,
precipitation,  and    evaporation have   been   neglected,  as it
varies from plant to  plant   and is dependent  on  the selection
of  treatment chemicals   as  well  as climatic conditions and is
insignificant in comparison to the total flow.

Selection of Pollutants to be Regulated

     The  selection  of    pollutants  for which  regulations  are
being  proposed was based  on  an   evaluation  of  the waste data
from the screening  and  verification   sampling  program.  The
two  major factors  considered   were  the  individual plant raw
waste  concentrations  and   the  total  subcategory  pollutant
loadings.

                              392

-------
                      14-12.   HISTORICAL
                    DATA SUMMARY WHS mRIABILirY FACTOR
                              Daily Measurements
                       Subcategory:  Titanium Dioxide
                        Chloride Process  (Rutile Ore)
                                 Plant  #559
                        April 76 through September 78

Pollutant

Daily Data
TSS
(1)
No. of Points 889
Average x,
ppn 21
Cadmium Chromium Iron Lead

109 128 854 128
0.058 0.072 0.620 0.068
Nickel

128
0.08
Zinc

128
0.151
  Standard
 Deviation,  S
  Standard
 Deviation,  S1
65.93    0.044     0.054     3.46   0.041  0.07    0.20
 1.54    0.68      0.67      1.86   0.56    0.76     1.02
                                           13.5    3.2     4.4     6.4
30-day (1)
Averages
No. of Points 30 26 30
Sfa^d 21.84 0.042 0.038
Deviation
Variability •* 04 24 2 04
Factor
Variability
Factor Ratio
WR 3.6 1.6 1.9
28 30 30 30
0.94 0.04 0.05 0.155
4.0 2,1 4.4 3.1
3.4 1.5 1 2.1
(1)  Section 8.2 presents a discussion of the approach and methodology employed
    in the statistical evaluation of data.
(2)  VER is the ratio of the variability factor for daily measurements to the
    variability factor for 30-day averages.
                                    393

-------
      14-13.   HISTORICAL EEFHJHST MONZDORJNG niaca. SUMMARY WITH VARIABILITY
                                  FACTORS
                             EMLY MEASUREMENTS

SOBCATEGORY: TITANHM DIOXIDE-ChlorMe Process
(RutxLe/Opgraded Ilmenite Ore)
Plant 1172


Daily Data{1)
No. of Points
Average ic, ppm
Standard deviation, S
Standard deviation, S1
Variability factor
30-Day Averages
No. of Points
Standard deviation, S
Variability factor
/ 1*
Variability Factor Ratio ^
VFR

TSS
454
5.39
9.13
1.16
7.6
15
6.31
2.92
0
2.6
Pollutant
Chromium
454
0.008
0.016
1.27
8.6
15
0.012
3.46
2.5

Copper
454
0.02
0.03
1.08
6.9
15
0.028
3.29
2.1

Einc
454
0.02
0.027
1.02
6.4
15
0.026
3.13
2.1
(1)  Section 8.2. presents a discussion of the approach and methodology employed
    in the statistical evaluation of data.

(2)  VFR is the ratio of the variability factor for daily measurements to the
    variability factor for 30-day averages.
                                   394

-------
TABLE 14-14.  TREATMENT PERFORMANCE DATA OF SAMPLED PLANTS #559 MID #172
SUBC&TEGOKY:  THANIUM DIOXIDE-ChlorMe Process
Pollutant
Plant #559 Plant #172
Pollutant
Concentration
(rag/1)
A
Raw
Waste
B
Treated
Effluent
Percent
Removal
cK^oo
Pollutant
Concentration
(rag/1)
D
Raw
Waste
E
Treated
Effluent
Percent
Removal
-f^oo
TSS
Iron
Chromium
Lead
Nickel
Zinc
1103
288
13.3
0.5
0.5
0.3
23
4.4
0.03
0.002
0.005
0.06
97.9
98.5
99.8
99.6
99.0
80.0
171
2.9
0.72
0.005
0.08
0.3
6.7 .
0.33
0.02
0.002
0.01
,0.09
96.1
88.6
97.2
60 -
87.5
70.0
                                    395

-------
     Raw waste  concentration -  Plant  #559 was visited in the
screening  phase   for sampling  of the  raw and  treated waste
water.  For each  pollutant, the maximum  concentration observed
gave  a preliminary   indication  of   its   pollution potential.
Five pollutants were found  above treatabilitv levels  in  the raw
waste  of  Plant f559  and they  were:    chromium,  iron,   nickel,
lead,  and  zinc  (Section 14.3.3).  A second  plant,  f.172, was
sampled  in the verification phase and  chromium was  the  only
pollutant  found above treatability levels in the raw waste.  At
Plant f559, the waste water  from the  chloride process is mixed
with   the  sulfate process waste water  and the chloride  process
effluents  were  sampled  at the  point  of mixing.   It  is  highly
probable that the sampled waste  included   the sulfate   process
effluent impurities.  The sulfate process  for the manufacture of
TiO2 uses  an ore of lesser purity.  For this reason the  nickel,
lead,  and  zinc  found  are attributed to the sulfate process and
are not further considered in this segment.  The nonconventional
and  toxic  pollutants  of  concern  include chromium  and iron.
Iron,  a nonconventional  pollutant is  significant because it is
present as a major  impurity in the rutile or  upgraded ilmenite
ore  and   was found at  treatable levels  in the  Plant |559 raw
waste.

     Total subcategory  raw  waste  pollutant loading - Chromium
was the only  toxic pollutant found in  significant concentrations
in  the raw waste of  both plants sampled  in  the screening and
verification phase. The  average unit  raw  waste chromium  loading
(Table 14-7)  obtained from the plants sampled was multiplied by
the total TiO2 subcategory production  by the chloride process to
evaluate the  overall  magnitude of the  pollutant potential for
the  subcategory.    The  value  of  241,000  kg/year  of chromium
discharged by  the  subcategory  in  the effluent  indicated the
necessity of control of this pollutant.

     The treatment technology practiced by the industry  removes
the  chromium  and  iron to low levels  as  seen  from the effluent
quality of the plants sampled and shown in Table 14-14.

Basis of Pollutant Limitations

     Conventional and nonconconventional parameters -

     A.  pH:    The  treated effluent is  to be  controlled within
the range  of 6.0 to 9.0.  This limitation is  based on the data
presented  in Appendix B of this report and the JRB Study (52).

     B.  TSS:  Long-term effluent data is available for TSS for
Plants f559  and #172.   At Plant  §172,  the  solid residues  from
the manufacturing process are sent to a landfill.  Although the
amount of  solids sent to the treatment system at Plant ^559 is
high compared with that selected for the model  plant (because of
                              396

-------
intermixing   of   sulfate   waste),    the   long-term   average
concentration  of  21  mg/1  (Table 14-12)  calculated  from the
monitored  data  submitted  by  Plant   $559  is  selected as the
treatment performance basis for the subcategory.  The daily and
the 30-day  average variability factors  (11 and  3.04)  derived
from the long-term data  of  Plant  #559 and given in Table  14-12
are used  to calculate the  concentration basis.   The proposed
unit effluent  limitations  are calculated using the model  plant
unit flow of 100 m3/kkg.  The calculations are given below:

     Proposed 30-day average concentration

       =  (21 mg/1)(3.04)  =  64 mg/1

     Proposed 24-hour maximum concentration

       =  (21 mg/1)(11)  =  230 mg/1

     Proposed 30-day average effluent  limit

       = (64 mg/1) (100 mS/kkg)/'  kg/m3  \
                              VlOOO mg/1/

       =  6.4 kg of TSS
           kkg of Ti02

     Proposed 24-hour maximum effluent limit

       =  (230 mg/1) (100 m3/kkg) f  kg/m3  >\
                                 V1000 mg/V

       =  23  kg of TSS
           kkg of TiO2

     C.  Iron:   The subcategory  performance  standard  of 0.62
mg/1 selected for iron is based on the long-term average of the
effluent data submitted by Plant f559  (Table 14-12).

     For the model plant, it is assumed that iron is present in
the ferric state in the raw waste  from the chlorination process.
Using the daily variability factor  of  4.0 and the 30-day average
variability  factor  of   13.5  estimated  from  the  long-term
monitored effluent data  of Plant $559 for  iron (Table 14-12),
and  the model plant  unit flow  of   100  m3/kkg,  the proposed
concentration basis and  effluent  limitations  are determined as
shown below.

     Proposed 30-day average concentration basis is:.

        (0.62 mg/1)(4.0)  =  2.5 mg/1
                              397

-------
     Proposed 24-hour maximum concentration basis is:

        {0.62 mg/l}{13.5)  =  8.4 mg/1

     Proposed 30-day average effluent limit is:

        (2.5 mg/1) (100 m3/kkg)/ m3/kkg  N
                              \1000 mg/lj

        =  0.25 kg of iron
             kkg of Ti02

The proposed daily maximum effluent limit is:

      (8.4 mg/1)(100 m3/kkg) / m3/kkg  \ =  0.84 kg of iron
                            V1000 mg/lj      kkg of Ti02

     The proposed  maximum 30-day  average  concentration of 2.5
mg/1 is higher than the  achievable  effluent concentration of 1.6
mg/1 reported in the literature  (10).  The latter concentration
is based on  the performance  of  lime neutralization followed by
settling of acid mine drainage waste,  and may  not be appropriate
for this subcategory.

     Toxic pollutants   -  Chromium is the only regulated toxic
pollutant because of its presence in the raw waste of the plants
surveyed at treatability levels.

     A.  Chromium:  The proposed chromium limitations are based
on the  long-term  data of  the  treated effluent of  Plant 1559.
The  influent to  the treatment  system  at Plant  fS59  contains
significant  amounts  of  chromium because of  the intermixing of
sulfate process waste.    The  long-term  average  of  0.07  mg/1
derived from   the  monitored data  of  the treated  effluent of
Plant 1*559  (Table  14-12)  is taken  as  the proposed subcategory
performance concentration.  The daily variability factor of 2,0
and 30-day  variability  factor of  3.8  estimated from the long-
term  data  of  Plant |559  (Table  14-12)  and  the model  plant
effluent flow  of  100  m3/kkg  of Ti02 are  used to  derive the
proposed concentration  basis  and  effluent  limitations.   The
calculations are shown below.

     The proposed 30-day average concentration basis is:

        (0.07 mg/1)(2.0)  =  0.14 mg/1

     The proposed daily maximum concentration basis is:

        (0.07 mg/1)(3.8)  =  0.27 mg/1
                              398

-------
     The proposed 30-day average effluent limit is:

        (0.14 mg/1)(100 m3/kkg) / kg/m3   N
                                V1000 mg/lj

        =  0.014kg of chromium
               kkg of Ti02

     The proposed daily maximum effluent limit is;

     (0.27 mg/1}(100 m3/kkg)/kg/m3  \ =  0.027 kg of chromium
                            \1000 mg/1/       kkg of Ti02

     B.  Other metals:  Lead, nickel, and zinc were found  in the
raw waste of  Plant |559 in the treatability  range.   They were
not found in the raw waste of Plant  |172.  Plant 1559 intermixes
the chloride  and  sulfate  process waste  before  treatment.  The
presence  of  these pollutants  in the  raw waste  at Plant £559
might result  from  the sulfate process waste.   The limitations
for the three pollutants are given and are intended to serve as
guidelines in cases where  the pollutants are  found to  be  of
serious concern.

     The  selected  30-day  concentration basis for lead, nickel,
and zinc are based on the lower treatability limits achieved by
the lime precipitation and settling of metal contaminated waste
(Table 8-11)  and the values are:

          Lead   = 0.3 mg/1
          Nickel = 0.2 mg/1
          Zinc   = 0.5 mg/1  ,

     The variability factor ratio of 1.9 for chromium estimated
from the  long-term data of  Plant  |559  (Table 14-12) is used to
obtain the daily  maximum  proposed limits for lead, nickel, and
zinc.     This  variability  factor  ratio   was   used  because
precipitation   of   chromium   is  similar   in  performance  to
precipitation of other metals.  Calculations are as follows:

     The  proposed  24-hour  maximum concentration basis for lead
is:

        (0.30 mg/1) (1.9)  =  0.60 mg/1

     The proposed 24-hour  maximum concentration basis for  nickel
is:

        (0.20 mg/1)(1.9)  =  0.40 mg/1

     The  proposed  24-hour maximum concentration  basis for zinc
is:

        (0.50 mg/1)(1.9)  =  1.0 mg/1

                              399

-------
The proposed limitations for BPT are given in Table 14-15.
14.7.3  Basis for Proposed BCT Effluent Limitations
     The BCT limitation  (applicable only to TSS)  was set equal
to BPT because BAT is equal to BPT.
14.7.4  Basis for Proposed BAT Effluent Limitations
The Application of Advanced Level Treatment
     The advanced  level  technologies,  viz,,  the use of sulfide
and xanthate as  a polishing  step  to the base  level treatment
system (BPT) f were considered for BAT and NSPS but were rejected
on the basis of cost  (Level  3 Table 14-11).   Level 1, used for
BPTr is selected for BAT treatment technology.
Technology Basis
     Alkaline precipitation  followed by settling  used  for BPT
(Level 1) is proposed for BAT.
Flew Basis
     A unit waste water flow rate of 100  m3/kkg of TiO2 used for
the BPT model plants has been selected for BAT.
Selection of Pollutants to be Regulated
     Chromium and  iron are  the  two pollutants  identified for
regulation.
     Nonconventional pollutants -
     The  proposed  iron  limitations  are   the  same  as  those
selected for BPT.
     Toxic pollu t an t s -
     A.  Chromium:    The  limitations  proposed  for  BAT  are
selected for BPT.
     B. ' Other metals:   Concentration limits for lead, nickel,
and zinc are  not proposed as limitations.   However,  they are
contained in this document for use  if these pollutants are found
to be of concern.   The values are the same as those selected for
BPT.
     Table 14-16 gives the proposed limitations  for BAT.
                              400

-------
                      TABLE 14-15-  PROPOSED LMECATIONS
     TITANIUM DIOXIDE - CHLORIDE PROCESS (HJTILE OR UPGRADED nMENTTE ORE)

          Best Practicable Control Technology*Currently Available
               Waste Hfeter Flow:  100 m3/kkg of Ti02

                                          Concentration       Effluent Limit

                    Subcategory           Basis (mg/1)        (kg/kkg of Ti°2)
Pollutant           Performance  VFR(1)   ffex                ^^
                      (mg/1)             30-day     24-hr    30-day     24-hr
                                          Avg      Max      Avg        Max
Conventional and
Non Conventional
Pollutants:
(2)
Total Suspended Solids 21 v ' 3.6
(2) 3 4
Iron 0.62^' J'
Toxic Pollutants:
Chromium 0.070^ 1.9
Lead 0.30(3) 1.9
Nickel 0.20(3) 1.9
Zinc 0.50(3) 1.9
64 230 6.4 23
2.5 8.4 0.25 0.84
0.14 0.27 0.014 0.027
0.30 0.60 _ (4) _ (4)
0.20 0.40 _ (4) - (4)
0.50 1.0 _<4> _(4)
 (1) VFR:  Ratio of the 24-hour variability factor to the 30-day variability
          factor.

 (2) Long term average based on loading data and variability factors of plant
    #559 selected from Table 14-11.

 (3) The lower limit of the literature treatability estimate (Table 8-11)  is
    used as the basis of or the 30-day average limitation.

 (4) No- effluent limitation proposed.
                                     401

-------
                      TABLE 14-16.  PROPOSED LMTATIONS
      THENIOM DIOXIDE - CHLORIDE PROCESS  (RUTILE/UPGRADED ILMENITE ORE)

                         Best Available Technology
                 Waste Water Flow;  100 m3/kfcg of Ti02
Pollutant
 Concentration    Effluent Limit
 Basis (mg/1)     (kg/kkg of TiO )

30-day  24-hour   30-day  24-hour
 Avg      Max      Avg      Ma,x
Nonconventional
Pollutants ;
W4>
Toxic
PoXtatants:
Chromium^ '
Nickel
Zinc
0.62 3.4 2.5 . 8.4 0.25 0.84
0.070 1.9 0.14 0.27 0.014 0.027
0.30 1.9 0.30 0.60 (3) {3)
0.20 1.9 0.20 0.40 (3) (3)
0.50 1.9 0.50 1.0 (3) (3)

 (1)  See Table 14-14 for details.

 (2)  WE:  Ratio of the 24-hour variability factor to the 30-day variability
     factor.
 (3)  No effluent limitation proposed.

 (4)  Limitations are applicable for PSES.
                                    402

-------
14.7.5  Basis for Proposed New Source Performance Standards

Application of Advanced Level Treatment

     The  concentration of  conventional,  nonconventional, and
toxic  pollutants  can  be  reduced  by filtering  the clarified
effluent from BPT  in a dual media filter.

Technology Basis

     For  new  plants,  the  recommended  waste  water treatment
technology  is  lime neutralization and precipitation, settling,
and  dual  media  filtration  (equivalent  to  Level 2) .   All the
existing chloride  process plants using rutile/upgraded  ilmenite
ore  currently practice lime  neutralization  and  settling, but
only published treatability data is  available  on the  performance
of dual media filters.

Flow Basis

     The  raw  effluent flow rate  is the same as  that used for
BAT,  namely  100  m3/kkg  of  Ti02.    It  is  assumed that the
unreacted ore  and  coke are slurried with water and  sent to the
treatment system.  The selected flow value  is  an  average of the
unit effluent flow rate of two plants (|559  and f!99) practicing
this method of solids disposal.

Selection of Pollutants to be Regulated

     It is  proposed  that  the pollutants regulated  for  BPT are
also regulated  for NSPS.   The  pollutant parameters of concern
are  pH,  TSS,  iron,  and  chromium.    Concentration  limits are
provided  for  lead,  nickel,   and   zinc  in  cases  where   these
pollutants  become  of concern.

     C onv e n tiona1  parameters ~

     A.  pH:  For NSPS, the BPT limitation  is  retained.  Control
of  the  final  effluent within  the  range of  pH 6.0  to 9.0 is
required  on the basis of data  presented in Appendix B of  this
report and  the JRB Study  (52).

     B.  TSS:  There are no in-plant performance  data available
on the filtration  of treated and clarified  NSPS waste water, so
a 30 percent additional removal is  assumed.   This assumption is
based  on treatability studies  (41)  using  filtration.   This
reduction is applied to the selected BAT (or BPT)  maximum 30-day
average  of  64  mg/1.   The  proposed   maximum  30-day  average
concentration basis is then given by:  64 mg/1  (1.00 - 0.30) = 45
mg/1.   Likewise,  the  proposed 24-hour maximum  concentrations
and  unit   effluent  limitations  are  obtained   from   the BAT
limitations (Table 14-15)  as shown  below.


                              403

-------
     The proposed 24-hour maximum concentration is:

        (230 mg/1)(0.70)  -  160 mg/1

     The proposed 30-day average effluent limit is;

        (6.4 kg/kkg)(0.70)  =  4.5 kg of TSS
                                 kkg of Ti02

     The proposed 24-hour maximum effluent limit is:

        (23 kg/kkg)(0.70)  =  16 kg of TSS
                              kkg of TiO2

     Konconventional pollutants -

     A.  Iron:  No in-plant performance data is  available on the
effect of dual media filtration on the removal of iron from the
lime treated and clarified waste water.  The removal efficiency
of 30  percent obtained  for  TSS  from  the  treatability studies
(41)  is assumed also to apply to  iron,  since  the iron is present
as  a  floe.   The  proposed  concentration  basis  and  effluent
limitation for NSPS  are obtained  by multiplying  the  selected BAT
(or BPT) limitations (Table 14-15) by 0.70 as follows:

     The proposed 30-day average concentration basis is:

        (2.5 mg/1)(0.70)  »  1.8 mg/1

     The proposed 24-hour maximum concentration basis is:

        (8.4 mg/1)(0.70)  -  5.9 mg/1

     The proposed 30-day average effluent limit is:

        (0.25 kg/kkg)(0.70)  =  0.18 kg of iron
                                 kkg of T1O2

     The proposed 24-hour maximum effluent limit is:

        (0.84 kg/kkg)(0.70)  =  0.59 kg of iron
                                 kkg of TiO2

     Toxic pollutants -

     A.  Chromium:     For  NSPS,   the  Agency is proposing  more
stringent controls  on  the  discharge  of chromium.    There  is no
directly  applicable data  on  filter   performance  in  the  TiO2
(chloride   process)   industry.     Therefore,   the   proposed
limitations  are  based  on  literature  treatability  studies
(40,41).  In pilot scale treatability tests,  raw municipal waste
                              404

-------
water samples were  spiked with toxic metals, treated with lime
and settled.  This was followed by dual media filtration of the
clarified effluent.   For  chromium,  an additional removal of 60
percent was obtained  by  filtration.   This  reduction factor is
assumed to be applicable.   The proposed  limitations  for NSPS are
obtained by multiplying the respective BAT  (or BPT) limitations
(Table 14-16)  by  0.40 as follows:

     The proposed maximum 30-day concentration basis is:

        (0.14 mg/l)(0.40)   =  0.060 mg/1

     The proposed daily maximum concentration basis is;

        (0.30 mg/1) (0.40)   =  0.10 mg/1

     The proposed 30-day average effluent limit is:

        (0.014 kg/kkg) (0.40) =  0.0060 kg of chromium
                                   kkg of TiO2

     The proposed maximum daily effluent limit is:

        (0.030 kg/kkg) (0.40)  =  0.010 kg of chromium
                                   kkg of TiO2

     B.  Other metals:    Treatability  studies  have indicated
that the following increased removals of lead, nickel, and zinc
can be achieved by filtration  (40,41).


                                   Additional Removal by
                            Filtration Using Settled Effluent
               Lead                      60
               Nickel                    14
               Zinc                       6
The   additional   levels   of  removal   are  applied   to  the
corresponding BAT (or BPT)  concentration  for the above metals to
get the NSPS concentrations.

     The proposed 30-day average lead concentration basis is:

        (0.30 mg/1) (0.40)  =  0.12 mg/1

     The proposed 24-hour lead concentration basis is:

        (0'.60 mg/1) (0.40)  =  0.24 mg/1                   ^


                              405

-------
     The proposed 30-day average nickel concentration basis is:

         (0.20 mg/l)(0.86)  =  0.17 mg/1

     The proposed 24-hour maximum concentration basis for nickel
is:

         (0.40 mg/1) (0.86)  =  0.34 mg/1

     The proposed  30-day average concentration  basis  for  zinc
is:

         (0.50 mg/1)(0.94)  =  0.47 mg/1

     The proposed 24-hour maximum  concentration basis  for zinc
is:

         (1.0 mg/1) (0.94)  =  0.94 mg/1

The proposed conventional, nonconventional, and  toxic pollutant
limitations for NSPS are given in Table 14-17.

14.7.6  Basis for Pretreatment Standards

Existing Sources

     For Pretreatment Standards for Existing Sources  (PSES), the
Agency   is   proposing   limitations   based    on   BAT.   The
pollutants  to  be limited are iron  and  chromium (see Table 14-
15).

New Sources

     For  Pretreatment  Standards  for New Sources  (PSNS), the
Agency is proposing  limitations  based on  NSPS.  The pollutants
to be regulated are  iron and chromium (see Table 14-17).
                              406

-------
 Pollutant
                      TABUS 14-17.  PROPOSED LIIOTIATIONS
                     TITANIUM DIOXIDE - CHLORIDE PROCESS

                      New Source Performance Standards
                    Waste Water Flow:  100 m3/kkg of TiO2
Treatability   VFR
                                         (1)
Concentration    Effluent Limit
Basis,  (ragA)    (kg/kkg of TioJ
 Max             Max
30--day  24-hour  30-day  24-hour
 Avg    Max     Avg    Max
 Conventional and
 Nonconventional-
    Pollutants  :
Total Suspended
    Solids


lron<2>


Toxic Pollutants;

         (2)
Chromium


Lead


Nickel


Zinc
                          15
                         0.40
                         0.030


                         0.060


                         0.17


                         0.47
               3.6
               3.4
1.9


1.9


1.9


1.9
 45       160     4.5     16
1.8       5.9     0.18    0.59
                         0.060     0.12    0.0060  0.012
                         0.12
                         0.17
                         o.47
          0.24
          0.34
                                                         0.94
                                              (3)     (3)
                                              (3)     (3)
(1)   VFR:   Ratio'of 24-hour variability factor to> the 30-^day variability
     factor.

(2)   Also  applicable for PSNS limitations.

(3)   No effluent limitations proposed.
                                     407

-------
14.8  TITANIUM DIOXIDE - SDLFATE PROCESS
INDUSTRY PROFILE

14.8.1  General Description

     The  industrial profile for the Sulfate Process Segment of
the Titanium Dioxide Subcategory is presented in Table  14-18 and
the status  of regulations  is  shown in Table  14-2.

14.8.2  General Process Description and Raw Materials

Sulfate Process - General  Description

     Among  the various  titanium ores,  ilmenite  is available in
abundance.   Ilmenite is  a low-grade titanium  ore with a Ti02
content varying  from 45 to 60  percent.   Ilmenite ore and slag
from  iron  production generally comprise the raw materials used
for the preparation of titanium dioxide by the  sulfate process.
Large  amounts  of   water  and  sulfuric  acid  are used  in this
process,  and the majority of  the plants are  co-located with
sulfuric acid plants.  Table 14-19 gives the analysis of various
ilmenite ores.  The preparation of T102 by the  sulfate process
utilizes three important steps:

     1.  Digestion:  FeO.Ti02  + 2H2S04 = PeS04  + TiO.S04
                                           + 2H20

     2.  Precipitation:  TiO.S04 + 2H20 • Ti02.H20 4- H2S04

     3.  Calcination:  Ti02.H20 =  Ti02 + H20

     The ore is  dried,  ground, and  then  reacted with sulfuric
acid*  The  reaction takes place at 160  degrees C and  the reacted
mixture consists of titanyl,  ferrous, and ferric sulfates.  The
product  is  dissolved in water.  The total  iron in  the reacted
product  is  kept  in the ferrous state by  the addition of scrap
iron.   After the reduction,  the  product  is dissolved in water
and  clarified with  the  aid  of flocculating agents to remove
insoluble  impurities such  as silicon, zirconium,  and  unreacted
ore.  The iron is removed from the clear  solution by  cooling the
solution  to 10  degrees -C when FeS04.7H20  crystallizes.   The
ferrous sulfate crystals,  commercial copperas,  are mechanically
separated  from the solution by  filtration  or  centrifugation.
The concentrated titanyl sulfate solution is diluted with water
and heated  to form titanium dioxide hydrate, which  is known as
strong  acid, is separated and either discharged  or  recycled.
The  Ti02.H20   filter  residue  is  slurried   with  water  and
conditioning agents are  added to control particle size, color,
dispersibility,  and photochemical stability.   The conditioning
agents include potassium, zinc, antimony,  and calcium compounds,
"and phosphate salts.  The  solution is filtered  and the filtrate
                              408

-------
             TABLE 14-18. - SUBGffiEGQRY PROFILE DHSA SUMMARY
SIBCMEGORY
TITANIUM DIOXIDE
SULEATE PROCESS
Total subcategory capacity rate
Total subcateogry production rate
Number of plants in this subcategory
308 Data on file for
     With total capacity of
     With total production of
     Representing  -.capacity
     Representing production
     Plant production range:
             Minimum
             Maximum
     Average production
     Median production
     Average capacity utilization
     Plant age range:
             Minimum
             Maximum
     Waste water flow range:
             Minimum
             Maximum
     Volume per unit product
             Minimum
             Maximum
                         401,000 kkg/year
                         259,000 kkg/year
                               4
                               5
                         320,000 kkg/year
                         246,000 kkg/year
                              80 percent
                              95 percent

                          31,000 kkg/year
                          74,500 kkg/year
                          49,000 kkg/year
                          43,000 kkg/year
                              76 percent

                              23 years
                              54 years

                          35,000 cubic meters/day
                         125,000 cubic meters/day

                             300 cubic meters/kkg
                             780 cubic meters/kkg
Sources of data are Stanford Research Institute, Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Cotanetce, Current Industrial
Reports, December 1977; Energy and Environmental Analysis, Inc.; Draft
Report, "Preliminary Economic Assessment of Effluent Limitations in the
Inorganic Chemical Industry," June,  1978 and "Economic Analysis of Proposed Revised
Effluent Guidelines and Standards for the Inorganic Chemicals Industry "
March, 1980.
                                    409

-------
                                                  5MBIJ3 14-19.  SHMXSIS OF HMEKCHE OSES1
H
O

Chemical
Constituent
TiO,
PeO
Fe203
si°2
A1203
P2°5
ZrO2
MgO
MnO
CaO
^2%
Cr203
Virginia
Piney
River Eoseland
44.3
35.9
13.8
2.0
1.21
1.01
0.55
0.07
0.52
0.15
0.16
0.27
51.4
37.9
1.6
4.6
0.55
0.17
NA
2.35
0.70
0.59
0.07
NA
UNITED SMCES
Hew York Florida
44.4
36.7
4.4
3.2
0.19
0.07
0.006
0.80
0.35
1.0
0.24
0.001
64.1
4.7
25.6
0.3
1.5
0.21
NA
0^35
1.35
0.13
0.13
0.1
California
48.2
39.1
10.4
1.4
0.2
NA
0.05
0.6
0.1
0.1
0.05
0.03
Ivry
42.5
39.1
20.7
0.88
1.05
NA
NA
2.0
0.04
0.1
0.36
0.15
CANADA
Bourget
22.4
36.9
31.2
1.0
6.01
0.93
NA
1,50
NA
0.55
NA
NA
Allard
37.3
26,3
30.0
NA
NA
0.004
NA
NA
0.10
NA
0.39
NA
                 Constituents expressed as weight percent.

                NA!  Not Available

-------
is  known  as  weak  acid.   Residual  acids  and  iron originally
present  in  the  precipitate  are  removed  with  the  water  of
hydration by calcination.  The resulting Ti02  pigment  is sent to
finishing operations,  which  vary according to  the end product
requirement  and  application.   The wet finishing operations may
include  some,  or   all,  of   the  following   steps:  repulping,
milling, surface treatment with proprietary agents  in solution,
washing, and drying.   The  alternative dry finishing operations
may include one or more milling steps followed by packaging.  A
simplified  block  diagram  of  the sulfate  process  is  shown in
Figure 14-9.


14.9  WATER USE AND WASTE SOURCE CHARACTERISTICS


14.9.1  Water Use

     Water is used in the preparation of titanium dioxide by the
sulfate process  for  noncontact cooling, air  emission control,
and for  process  reactions.   In the process, water is  used to
leach the  soluble  sulfate  salts from the  reaction  mass  and to
convert the titanyl sulfate to titanium dioxide hydrate.  Water
is  also used  to  wash  the  titanium  dioxide hydrate precipitate
free  from  residual  acid  and  iron.    Water  is  used for  air
emission control during  the  drying of  ore, on  digester  units,
and for the cleaning of the kiln gases before  they are vented to
the atmosphere.  In  the  digester  unit,  water  seals are used to
maintain a vacuum on the digester units. Large  amounts of water
are also  used in the  finishing  operations.   Table 14-20  is a
summary of water usage in the titanium dioxide subcategory using
the sulfate process.

14.9.2  Waste Sources

Digester Sludge

     After  the  digestion  of  the  ore  in  sulfuric  acid,  the
result-ing  sulfates are  dissolved  in  water  and  the  insoluble
impurities are removed in  a clarifier or filter.  These include
silica, alumina, sulfuric acid,  and  unreacted  iron.  The quality
of  this waste varies and depends on the type  and quality of ore
used.   Data on  the  quantity  of  this  waste   indicates  that
approximately 210 kg/kkg is produced.

Copperas
           \
     The recovered ferrous  sulfate is marketed or disposed of as
a solid waste.   The amount of copperas  generated is  about 950
kg/kkg of Ti02.  The copperas generally contain  small  amounts of
adsorbed "sulfuric acid.
                              411

-------


HH5JUTB
STROHG-ACID {
WAfSR — •
DIGESTER

HATER ... , fc,
WATER"""*

STEAM — »,
WEAK-ACID
	 RECfCLE 	 ^
WATER— »
CLARIFIER
*
EVAPORATOR
*
PRECIPITATION
1
FLASH
COOLER
— EHI8SIOHS — »"
FPRAY COK01HSEBS uvurjieu'r fci
WATER — to ADO VEOTORI SCRUBBERS «rrl««ia. ^
	 Mnn qt.flTIBV itmn.it sfTnV «-

_• EMISSIONS — », ' ' -
„.___ CONDENSERS — EFFLUENT —*• 	
WATER 	 1>

CONDEHSERS — EFFLUENT ~* 	
tjumoo ^

J
1 FIRST MOORE
PTr.TBn

HATER — *»
SECOND MOORE
FILTER
STEftli *^ 	 •
1 	 RECXCLE 	 »
HATER — •>
STEAM 	 »
— *
DZ

„„ 	 , 	 „, HEAR M"TP 	 m

! '
CALCINER

HEX MILL
— KMIB5IOMS 	 » COOLIMG SPRAYS AHD ELECJ-._..fc ..,„,., J,CID_
HATER 	 UROSTATIC PRECIPITATORS | ~~ 	
1
EMISSIONS
*
HATER 	 * MIST ELIMINATORS 1 	 EFFLDEHT — »• 	
	 |»BI.nBHT ^
" ""'•"• 	 ff , i ....
1

JET MILLS
i- - •» i H5» i un^ fc
IfATEH •— -9
I
TITANIUM
OXIDE PIGMENT
4 '•
PACKA6IHG '
HATER — - * Jgl MILL SCRUBBERS 	 ^.EFFLUENT——
t
TO SALES
r.m

Figure 14-9.  General process flow diagram for production of titanium dioxide by
              sulfate process.
                                                                                         £TE DISPOSAL •

-------
      14-20.              IN          DIOXIDE -         PROCESS SOBCKEBGOEZ
Uses
Water Usage per Unit of Production
        s( m3/kkg of TiO,}
Noncontact  cooling

Direct process contact

Indirect process contact
(pumps, seals, leaks,
spills, etc.)

Maintenance, equipment
cleaning and work area
washdown

Mr pollution control
Nbncontact ancillary
uses  (boilers, utilities,
etc.)
Plant #555

      47.8

     390
       6
     258

      36
Plant #694

     408

     588

       1.6
       1.8



      78

      33
Plant #696

     149

     297
       4
      81

      NA
NA:  Not Available
                                     413

-------
Strong Acid Waste

     When water  is  added to titanyl sulfate solution after the
removal of copperas,  sulfuric  acid and the hydrate of titanium
dixoide are  formed.   The acid contained in solution is removed
by filtration and the filtrate is known as  strong acid solution.
The concentration of  sulfuric  acid varies  from 15  to 30 percent
as  H2S04.    In  addition to  sulfuric  acid,  the  waste  stream
contains  ferrous sulfate,  titania,  antimony, and other  heavy
metal salts.  A part  of  the acid  is returned to the process and
the rest sent to the  treatment facility.

Weak Acid Waste Stream

     The  waste  generated  from   washing  the  titanium  dioxide
hydrate precipitate is known as weak  acid.  The concentration of
sulfuric acid  in this waste varies from two to four percent as
H2S04 and contains various  impurities, including  iron sulfate,
titania,  antimony,  and  other  heavy  metal  salts.   It  also
includes, in some  cases, the  conditioning agents  added  to the
precipitate prior to washing, to control and  improve the quality
of the final product.  The  weak acid may also include the kiln
exhaust scrubber waste.

Scrubber Wastes

     Scrubber waste  water results from the scrubbing of vapors
emitted during  the  drying  of the  ore, during  digestion,  and
during kiln drying.  The amount of waste water generated depends
on  the amount  of water used  and type  of   emission  controls
practiced.    The  scrubber  water  contains  titanium  dioxide
particulate, acid mist,  sulfur trioxide and sulfur dioxide.  Of
all  the  waste produced from  titanium dioxide-sulfate  process
manufacture  subcategory, the  scrubber  waste  water constitutes
the major portion.

Wet Milling Waste

     These  wastes  are  generated during  wet finishing of  the
titanium  dioxide  pigment.    Wet  milling  is  used to  produce
pigment particles of  the desired  size  and  surface  character and
requires steam  and water for  repulping the  pigment.   Caustic
soda  is  also  used  to  remove any  residual  acidity from  the
titanium dioxide  pigment during  the  finishing operation.   The
waste  water  from wet  finishing  operation, therefore,  contains
titania, sodium  sulfate, and other agents added  to improve or
achieve desired properties  in  the final product.
                              414

-------
14.10  DESCRIPTION OP PLANTS
14.10.1  Screening

     Plant |555 was visited and its  waste  streams  sampled in the
screening  phase  by  an  EPA  Region  II   team.    The  pigment
manufacturing  operation   utilizes  a  titania  slag  for  the
production of Ti02  by  the sulfate process.  After digestion of
the  slag  in  sulfuric  acid  the  residual  gangue material  is
filtered   out   and  the   clear   liquor   is   concentrated  by
evaporation.   The crude  pigment  is formed  by hydrolysis  with
water and  steam  and processed to form  both  anatase  and rutile
type pigment products.  Table 14-21  presents  raw waste flows and
pollutant characteristics for Plant f555.

     Waste  water  samples were  collected at  five  points and
analyzed  for  the  conventional,  nonconventional,  and  toxic
pollutants.   These sampling  points were  designated as  1) the
digestion suppression  flume  containing  waste water from direct
contact  air scrubbers  on  the digesters, 2) the black end flume
containing  wastes from  major cuttings,  filter  sludges,  acid
filtrates, and evaporator and condenser  waters, 3)  the white end
flume carrying finishing  process  filtrates,  noncontact cooling
water, and sanitary wastes,  4) northside jet air scrubbers, and
5) southside jet air scrubbers.

     At present , all of the process waste streams  are collected
in a  settling  basin which  is open  to  tidal fluctuations  that
provide  diurnal  flushing  of the effluent  into  the receiving
waters.

14.10.2  Verification

     Plant  f559  was surveyed  in  both  the screening phase and
verification phase of  the study.  At this  plant the strong acid
is sent to a lined holding pond for  equalization.   Effluent from
the pond is neutralized with ground calcium carbonate limestone
in a  reactor.   A minimum  amount is added to raise the pH to a
level such  that calcium sulfate, but  not  ferrous  hydroxide, is
precipitated.  The CO2 formed during  the reaction is vented to
the  atmosphere  and  the  , calcium   sulfate  slurry  goes  to  a
clarifier.   The  underflow from the  clarifier is  filtered to
produce  pure  gypsum crystals  at  a concentration  of 70  to 80
percent.

     The  weak  acid  is sent  to a  settling, pond, where  it is
combined with  a  small quantity of other  wastes.   The effluent
from  the  weak  acid  pond  is  mixed  with  the  calcium  sulfate
clarifier overflow and neutralized with  ground  calcium carbonate
in a three-stage reactor.   Pebble and  slaked  lime  are also added
                              415

-------
14-21.
BSW       C3BWM3ERISTICS (INDUSTRY DATA)
(PRODUCTION OP Ti02 BY
                                                              PMNT #555

Waste Source
Digestion
Clarification
Evaporation
Cooling
Unit
(nvVkkg pH*
of TiC>2)
115
3.58
113
20
Strong Acid from 8.49
first Moore Filtration
Weak Acid from
12.2
3.0
2.5
4.0
6.1
< 0.5
2.0
Pollutant Waste Loads, (kg/kkg of TiC^)
Acidity NEU Fe TSS TDS
(asH2S04> (asN)
20.8
26.7
18.7
2.49
2.360
88.3
NA
NA
NA
NA
NA
NA.
0.042
8.42
1.14
0.099
139
3.8
9.3 35.7
175 40 .'8
3.2 20.2
0.46 3.09
0.959 2.815
0.23 98.8
first Moore Filtration
Weak Acid from
10.4
1.7
148
NA
0.29
0.13 151
second Moore Filtration
Weak Acid from
12.0
2.0
20.8
NA
0.22
2.0 7.50
first stage
Calcination
Weak Acid from
second stage
Calcination
                     40.0
                 2.2
                                    19.2
NA    0.64
NA:  Not A-yailable

* Value in pH units
(1) - Response to 308 Questionnaire, 1976
4.92 33.1
Calcination Mist
Bliminators
Wet Milling Washing
and Drying
Jet-Mill Condenser
Jet-Mill Scrubbers
Boiler and Water
Plants
38.

11.

27.
18.
16.

7

1

0
0
6

3.

8.

6.
7.
9.

0

0

5
4
0

7.50

NA

NA
NA.
NA

NA

8.6

NA
NA
NA

0

0

0
0
0

.02

.01

.01
.13
.66

0.

2.

1.
1.
5.

21

13

1
7
25

27.9

11.0

2.7
3.58
8.92

                                    416

-------
to raise  the  pH and precipitate more  calcium  sulfate.   Air is
also  introduced  to convert  the ferrous  iron  to ferric.   The
effluent  from  the reactor goes to another  clarifier,  and the
clarifier underflow is filtered to concentrate the solids to 70
percent.  The  overflow  from  the second clarifier is mixed with
the other  process waste  waters.   These  include the scrubber,
finishing,  and cooling  waste  waters.   The combined  water  is
neutralized with slaked  lime   before  it  is sent  to  a  final
settling pond,  the  effluent  from  which  is discharged.   Figure
14-10 gives the flow diagram of the treatment process and shows
the  sampling   locations  for  both  screening and verification.
Table  14-22 gives  the  flow data  for the  waste  streams  and
conventional and nonconventional pollutant emissions,

14.10.3  Other giant Descriptions

     At Plant  #694, the clarification  sludge which contains the
unreacted  ore is  sent  to  waste  disposal.      The  weak  acid
effluent from  the plant is neutralized with slaked lime and the
grit is settled out for  landfill disposal.  After the separation
of  grit,  the  aqueous   stream  is  discharged   to  a  municipal
treatment system.  The  other wastes,  together  with runoff from
the plant site,  are  collected,  and sent to a lagoon for solids
removal, and the overflow discharges to a river.

     At Plant  1696, the  raw wastes are  sent  to thickeners to
remove  the  suspended  solids and  the overflow  is  discharged.
Depending  on   the  titanium  content,  the  underflow from  the
thickeners  is  either recycled  or  disposed  of  in  a landfill.
This plant has discontinued operations.

     At Plant  |605f  the  process raw waste streams are combined
and sent to a  reactor for neutralization with a water slurry of
finely ground  calcium carbonate.  The  effluent from the reactor
is  hydrocycloned into  three fractions.   The  first fraction,
which is the coarse gypsum slurry,  is separated from the reactor
effluent at a concentration of  85 to 90 percent,  and placed in a
self-draining  dewatering  system.   The  "dry"  solids are finally
trucked" to  a   landfill.   The second  fraction  separated in, the
hydrocyclone is  a fine  gypsum  slurry  which  is  recycled to the
neutralization reactor.  The  residual gel slurry forms the third
fraction, and  this  is sent to  a thickener after C02 degassing.
A flocculating agent is added   to  the  flow to  the thickener to
promote solids separation and   thickening.   The underflow from
the  thickener  is centrifuged  and  the solids  landfilled.   The
filtrate from  the centrifuge is recycled  to the thickener, and
the thickener  overflow is discharged.

     The volume and characteristics of waste water streams from
different sulfate process titanium dioxide plants do not differ
greatly.   Some variations,  however,  are  noted  as  a  result of
                              417

-------



WtAK ACID .__,,»,
WASTE STREAK


Itl
SI
H
OTHiR MOPUCT
WASTE VATEA
k
WEAK ACIO rONO


VER
PPL* WATER " 	
mteiMi
SUPPLY WATER
\

H
00

STRONG AC IB .^^
WASTE STMAH



tt\\ r
iUPPUf WATER




STROKG ACID
MHO












n
/o


i\




^
ft



















REACTORS






































FILTER


SOLIDS
Tft




,

















REACTORS










FILTER
1
»




	 •" ' HHAL f~+*
i

SOLIDS TO
STORAGE/ ,,
LAHOF1LL k






























pnmj ^_7
IS


Vs

^ 	 OTHER PRODUCT
WASTE WATER




TIO. {SULfATI PROCESS)
Z SCRUBBER
WASTE WATER












STORAGE/LANDFILL













LEGEND
mm SAMPLING POINTS
Figure 14-10.  Gaieral flow diagram at plant 1559 shewing the sanp-ling points.
               (Titanium dioxide - sulfate process.)

-------
      14-22.         AND                          FOR THE
                      FOR
   Stream
     No. W>
Sampled          Unit       TSS             Iron
Stream           Plow       Load            Load
Description    (irtVkkg) (kg/kkg of Ti02) (kg/kkg of Ti02)
                  Weak Acid
                  Pond Overflow
                  68
.I1"2'   1.23
1.23
                  Strong Acid
                  Pond Overflow

                  Scrubber and
                  Contact Cooling
                    Water

                  Final Treatment
                    Effluent
                   6.1     205.85
                    (1) (2)
                 361.4     113.5
                   (1) (2) (3)
                 436        10.0
                      106.34

                       51.68



                        1.92
(1)  - The flow is contributed by the sulfate process stream.

(2)  - The pollutant load was calculated by multiplying the flow contributed
      by the sulfate process stream tines the concentration of pollutant.
      Pollutant Load = (total stream flow)  x (fraction contributed by sulfate
      process waste)  x stream pollutant concentrated.
(3)  - "While calculating the "unit flow the contributions to the treatment
      process from precipitation, the water in the treatment chemicals,.
      losses from evaporation and from solids leaving the process have
      not been considered.

(4)  - See Figure 14-10 for sampling point location
                                    419

-------
differences  in  ore  qualities,   in  location  and  in  process
details.   The majority  of the  dissolved pollutants  in  waste
water from this segment of the Ti02  industry  consist  of acidity
and  iron.    Segregation  of  the  waste  water is  important  for
control  and  treatment   practices   and   aids   in  developing
economically feasible  treatment  systems.   Generally,  weak  and
strong  acid  streams are  segregated from  each other  as well as
from the less  contaminated waste waters  which  include contact
cooling, scrubbing,  and  some finishing operation  wastes.   The
unit flows  for the  segregated  raw waste  streams  at  different
facilities are shown in Table 14-23.

     The average total effluent  flow  rate is 475 m3/kkg (Table
14-23)   for Plants #555,  £694,  and  #559.  Complete flow data is
not available for Plants #696 and $605.

14.10.4  Toxic Pollutant Concentrations

     Section  5.1.2   of  this  report  describes   the  scope  and
methodology  of  the  sampling  program.   In the  Sulfate Process
segment  of  the Titanium Dioxide  Subcategory,  18  different
sampling points were  selected for  studying the  toxic pollutant
characteristics of  the water  supplies,  the   raw  process  waste
waters,  and  the  plant  effluent  at  two major  manufacturing
facilities.  For the inorganic constituents 575 analytical data
points  were  generated  and an additional 1,824 data points were
obtained for the organic  toxic pollutants excluding  blanks  and
duplicates for quality control.

     The only organic toxic pollutant found during the screening
program was  phenol  which was observed  at only  one of  the  two
plants  sampled.  The maximum raw waste concentration of phenol
was  0.020  mg/1, however the  raw  water  source for the  plant
contained  as much  as  0.007 mg/1.    This  is  well below  the
treatability  level  for  phenol,   therefore,  phenol   is  not
considered a significant or process related pollutant.

     Daily raw waste loads were  calculated from the  flow rates
measured or  estimated  at the time  of sampling and the measured
pollutant concentrations.,  That  is,

     Daily loading  (as kg of pollutant per day)  = (C)(Q)
                                                   1000
                                              *
     Where the concentration  (C) of  the pollutant is  expressed
     in units of mg/1  (Note: 1 kg/m3 = 1000 mg/1), and the flow
     rate  (Q)  is  expressed  in units of  m3/day   (m3,  a  cubic
     meter,  is equal to 264 U.S.  gallons).

     Similarly,  the unit  loadings  were  calculated  from  the
reported Ti02 productions rate (P),  the waste stream flow rate
(Q)/ and the measured pollutant concentration (C).


                              420

-------
TABLE 14-23.  PROCESS WASTE WATER FLOW AT PLANTS # 555, #694 and #559
                 TITANIUM DIOXIDE (SULFATE PROCESS)

Plant
#555
#694
#559
Average

A
Strong
8.49
16
6.10
10
Flow in (m3/kkg
B
acid Weak acid
78.2 '
67
69
72
of Ti02)
C
Scrubber and
contact cooling
water
362
457
361
393

D = A
Total
449
540
436
475

+ B 4- C
Effluent




                                421

-------
     Onit loading  (as kg of pollutant
     per day kkg of Ti02)
                                (C) (Q)
                                1000 (P)
     Where C  and Q  are  expressed in  the  same units described
     above, and  the production  (P)  is  expressed in  units of
     kkg/day  (kkg  is 1000 kg, a metric  ton,  which is equal to
     2205 Ib).

     The maximum concentration of toxic pollutants found in the
raw waste at concentrations above the  treatability level in the
screening and verification program were:•
             Maximum Concentration Observed  (ug/1)
Pollutant
     Screening
(Plants  f555  & 1559)
Verification
(Plant H559)
Cadmium
Chromium
Copper
Lead
Nickel
Zinc
Antimony
Arsenic
Thallium
Selenium

124
1
3
6
3




340
,000
,500
,700
,400
,800
20
11
19
360

31,
1,
5,
1,
17,
1,


12
000
000
200
300
000
400
340
41 ,
Below detection









limit
     A summary  of  daily and unit  (per unit  of production) raw
waste loads for all plants sampled can be found in Table 14-24.
Individual plant  raw waste  loads  and concentrations  found in
sampling are given in Table 14-25.

     Based on  the  total annual production of this industry and
the average waste load  generated per unit product, the estimated
total toxic  pollutant  raw waste loads  generated  each year for
this subcategory are as follows:
                              422

-------
         TABLE 14-24.  SlMffiRY OF BAH WASTE LOADINGS FOUND IN SCREENING AND VERIFICATION SAMPLING

SIBCKIEQOKX THSWUM DIOXIDE - SUIESE PROCESS
Pollutant: Loading Range,
(kg/day)
Minimum Maximum
Toxic
it1- antimony 5.0 28
to
U)
flrsenic 1.9 4.0
Cadmium .068 7.2
Chromium 140 530
Copper 8.2 19
Lead 3.0 65
Nickel 3.7 23
Selenium 7 9.5
Thallium .47 1.2
Zinc 1.8 85
Conventional and Nonconventional
IBS
Icon, Pe
Minimum
.032
.012
.00044
1.1
.065
.024
.029
.0020
.0030
.014


Unit Loading,
(kg/Meg)
Average Maximum
0.11 0.22
0.19 .032
0.19 .057
2.0 3.4
.085 .12
.18 .42
.080 0.15
.031 .060
.0055 .0080
.34 .55
320
600
Ito. Of
Plants'1*
3
3
3
3
3
3
3
2
2
3
1
1
(1)  - Data are  taken only from those plants where pollutants were found above detection limits, or, in the
      case of  TSS and Iron, where data are available.

-------
                                     14-25.  "rescte wuumnsi AVERSCE raw      K»BS AND
K)

SOBCffilEGOHX
t
TEEBflltM DIOXIDE - SULEKTE PROCESS
Screening

.antimony
.arsenic
Cactnium
Chroniun
Copper
Lead
Nickel
Selenium
Thalliun
Zinc
Cmj/1)
0.77
0.11
0.29
3.8
0.20
0.075
0.091
MA
m
0.088
Plant S555
(kg/kkg)
0.22
0.032
0.057
1.1
0.065
0.024
0.029
< 0.06
MA
0.014
Plant
0.16
0.029
0.0020
7.0
0.25
0.20
0.31
NA
0.020
1.1
1559
(kg/kkg)
0.080
0.014
0.0009
3.4
0.12
0.10
0.15
HA
0.0080
0.55
Verification
Plant
(rnq/1)
0.074
0.028
0.0010
3.1
11
0.96
0.14
0.0050
0.0070
1.04
1559
(kg/kkg)
0.032
0.012
0.00044
1.4
0.070
0.42
0.061
0.0020
0.0030
0.45
                   KB. = Not Available

-------
                   Pollutant           Total Annual Raw
                                       Waste Load (kg/year)


                   Cadmium                 5,000
                   Chromium              510r000
                   Copper                 22,000
                   Lead                   47,000
                   Nickel                 21,000
                   Zinc                   88,000
                   Antimony               29,000
                   Arsenic                49,000
                   Selenium                8,000
                   Thallium                1,400


14.11  POLLUTION ABATEMENT OPTIONS


14.11.1  Toxic Pollutants of Concern

     The  toxic  pollutants  found  above treatability  levels  in
this  industry  were  evaluated  on  the  basis  of  the  maximum
concentration observed in the process  raw waste waters.   These
values  are shown  in  Section  14.10.3.   Using  cadmium  as  an
example of a borderline case, its maximum observed concentration
of   0.34   mg/1  is   considered   significant   because  removal
efficiencies  ranging  from 70 to  97 percent  could possibly  be
achieved on the basis of the lower limits of treatability shown
in Table 8-11  for lime/settling, lime/filter, and sulfide/filter
technologies.    The  BAT  utlimately selected  as  a   basis  for
regulations may not   be  as  effective  as  the  most  advanced
technology  considered  at  this   stage  of  the  evaluation  of
alternatives.

     The  sampling  data  from this  industry indicate  that the
toxic pollutants of  concern are chromium,  zinc,  nickel,  lead,
copper, antimony,  arsenic,  and  cadmium  in  decreasing order  of
the  amounts found.   Selenium  and  thallium  were detected  at
levels  too  low  to  be   treated  effectively.    The  relative
pollutant concentrations and loadings  in the  raw waste largely
reflect the amounts of impurities  in the  ilmenite ore or titania
slag being  processed.  The  major  impurity found in  the various
grades of raw material is ferrous  iron  as indicated in Table 14-
19.  The toxic metal  impurities would also be expected to occur
in a wide range of concentrations in the raw materials.

     The  advanced  treatment  technology  options  evaluated for
sulfate process segment of  the industry were selected for their
ability  to  remove   toxic   metals  of  concern  with  greater
efficiency  than the prevailing (BPT) practice  which also removes
TSS, iron,  and sulfate from the waste waters.

                              425

-------
14.11.2  Process Modifications and Technology Transfer Options

     Specific  process  modification  recommendations  are  not
made.  However, several areas for further investigation suggest
themselves.  They are:

     1.  One of the water borne wastes,  the  strong sulfuric acid
produced  from  the Ti02  sulfate process, has a  sulfuric  acid
concentration  that varies  from  15  to 30  percent as  H2S04.
Currently, only a small portion of it is recycled.  Research is
needed to find cost-effective ways to concentrate  the acid to 90
percent  and  to eliminate the  impurities (especially  iron)  so
that it can be reused  in the digester.   This will eliminate much
of  the  alkali  requirements  for  neutralization  and  relieve
disposal problems associated with solid waste gypsum.

     2.  Economical  methods  need  to  be  developed  for  the
recovery of iron oxide,  aluminum,  and vanadium from the waste to
the extent that markets are available for these materials.

     3.  If markets could be developed  for  the sale of ferrous
sulfate  (copperas),   solid  waste  disposal   problems  would  be
reduced.  Currently, a portion is sold  and the rest disposed of
as a solid waste.

14.11.3  Best Management Practices

     Storm water runoff from the plant  site should be collected
and sent to the treatment facility for  the removal of suspended
solids.

14.11.4  Prevailing Control and Treatment Practices

     The treatment practices of the plants producing Ti02 by the
sulfate process is given in Sections 14.10.1  to 14.10.3.

14.11.5  Advanced Treatment Technologies

     Although  sulfide precipitation,  the xanthate process,  and
ion-exchange might be  applied to the clarified solution obtained
by  alkaline precipitation,  oxidation  and  settling  the  cost
incurred are high because of  the  large  quantity  of  water (more
than  400 m3/kkg of Ti02)  that must  be treated.   The  sulfate
process  is one of  two subcategories  (the other being  Soda  Ash
Solvay Process)  in  the  Inorganic  Chemicals  Industry studied in
this  report  that  generates,  the  largest quantities  of  waste
effluent.
                              426

-------
14.12  SELECTION OF APPROPRIATE TECHNOLOGY AND EQUIPMENT


14.12.1  Technologies for Different Treatment Levels

Level 1 (BPT)

     In the  Level  1 treatment the blended strong and weak acid
streams  are neutralized  with  calcium  carbonate.   The  toxic
pollutants  are  precipitated  and separated along with gypsum in
first stage thickeners.  Aeration then oxidizes  any ferrous iron
present and  removes C02  before mixing with miscellaneous plant
waste  containing   minor   amounts  of  heavy  metal  priority
pollutants.  The combined stream is then given lime treatment of
pH 9 and  settled  in polishing lagoons  before discharge.   This
three-step  system  is  patterned  after  existing  systems  which
separate the acid streams from miscellaneous  wastes in order to
make possible  the recovery of  pure  and impure gypsum from the
relatively consistent acid streams.   Alkaline precipitation of
heavy metals,  and significant removal  of  arsenic occur during
the  last  two-stages  of  lime  neutralization,  and  settling  of
precipitated  toxic pollutants  occurs  in  the  final polishing
lagoons.   Because waste flow  rates  are unusually  high  in the
sulfate  process,   long-term   lagoon  settling  is  more  cost
effective than  dual media  filtration.   The mechanical aeration
step used for oxidizing ferrous iron may contribute an important
mechanism  for  the  simultaneous  removal of other  heavy metals
present  very  similar to  the  ferrite coprecipitation  method
described in the, Treatment Technology Assessment section.   The
flow diagram of the treatment system is shown in Figure 14-11.

     Although  the Model Plant  does  not include  equipment for
gypsum  recovery,  it  is based on  separation  of waste streams,
making pure  or  impure gypsum  recovery possible by intercepting
thickener  underflow(s).   Recovery of gypsum  as a saleable by-
product is not  a viable option since no market  appears to exist
at this time.

Level 2

     Level  2 for  the sulfate process employs the described BPT
treatment  for  strong acid,  weak  acid, and   55% of  the "other
wastes".  The remaining other wastes receive  soda ash treatment
and  settling,   to  permit  recycling  a nonscaling  effluent for
scrubbers and miscellanous uses.   Heavy metal pollutants in the
separated   recycle   stream  are   settled  as  carbonates   and
periodically removed  to a secure  l;andfill.  The flow diagram of
this treatment  is shown in Figure  14-12.
                               427

-------

               I
               I
                              I
                             I
a
CO
                             rH
                             3
                              CD
428

-------
V£>





GROUND
LIMESTONE
CiCOj




C.C










-•"•"





"^

X.




S


,1



















fin 	 ri son*
Y LJ ASII
1
' u
1


»BC1tCl« %i I, 1|
(Misc. west V | jyr | y






/






1





L





«












45%




OTHER
WASTES


HAP1O MIX AND SETTUNC1




1



^p"V- —
X
nr
^Q-1
Oj SLUHHT
TANK




















>»\ WEAK ACin/*~1pP1f
BilSTgOMfi AraliA— f\J^
t*
k


t f

JQ
IUI.T1PL

RECYCLED EFFLUENT
-fr. C08*0
ATMOSPIIEHE
__ 	 --"
*
E
REACTORS


























if



^—



-V













tIME
"^
*^-^

X_X" ^"*^i
I' I «l — - | FOMTMEU
1 -J—WI ^ | *
I***" §
MIXING I
| f _» A -»


1


n
i
Jy








1 AERATION * MIXING
X^ ^X^ OOZ REMOVAL
lit STAGE
TIBCKENER
f










k,/


.



'C")
1
i¥
i
roy




7







MIXING
TANK
I 2ND STAGE
I THICKENER .
i \



«*



















ffi*J
I r










/ " EFFIAJEHT
FOUSIBHa
IAGOON
RECYCLE , SOUU.SW5TOSAI-
                                   flow iRoaitorlag, pH tnonitdirlnf
              Figure  14-12.  Level 2 waste water treatment for titanium dioxide - sulfate process.

-------
14.12.2  Equipment for Different Treatment levels

Equipment Functions

     Treatment of waste water from the sulfate process involves
the mechanized  handling of  large  quantities of  chemicals and
reaction products, primarily gypsum.

     The  BPT model  includes  rail  car  deliveries of  ground
limestone  and lime,  bucket  elevators, storage  bins,  multiple
feeders, mechanical feeders, mechanical aerators, and two-stage
thickening for  removal  of pure and iron-bearing gypsum for the
treated  acid  waste  streams.    Calcium  saturated  thickener
overflow  and  miscellaneous  other  waters  are  subjected  to
alkaline precipitation and settled in a one-day polishing pond.
In Level 2,  to  reduce the mass discharge of heavy metals, only
55% of the BPT  "other waste"  flow joins the treated acid waste
stream,  for  BPT  treatment as  described  above.   However, the
remaining 45% of "other  wastes"  is  given separate  treatment with
soda  ash settled  in  a lagoon,  for  recycle  to miscellaneous
scrubber  and noncontact  cooling purposes.   Treatment  of the
strong and weak acid streams,  including oxidation and settling
of ferrous iron, remains the same as in the BPT model.

Chemicals and Handling

     First stage neutralization  employs ground limestone, while
lime is used for second stage and final alkaline precipitation.
Oxygen is supplied from atmospheric air,  and polymer is added to
assist in the second stage settling of  iron  hydroxide.   Aside
from  the  bulk  handling  of  large  amounts of  these  common
chemicals, there are  no special  hazards involved in their use.

Separation and Removal of Solids

     Large  quantities  of thickener  underflow  are pumped  to
spreading areas for consolidation of the  solids, which are later
pushed into 18 foot high piles on land provided for 10 years of
operation.   Solids from  occasional  draining of  the polishing
lagoon  and  the Level 2 recycling  lagoon are  returned  to the
aeration step of the  waste acid streams,  after  which they will
be settled out  in  the second stage thickener, being handled as
part  of  the thickener  underflow.    Although  no  dewatering
equipment is'provided, the first and second stage  thickeners can
be  sources  of  pure  and'  impure gypsum  for future  byproduct
recovery.
                               430

-------
14.13  TREATMENT COST ESTIMATES


14.13.1  General Discussion

     To prepare treatment cost estimates, a model plant concept
was  developed.    For conceptual  design a  representative  unit
waste  flow was  selected,  together  with three  different  Ti02
production rates.  The latter  were  chosen  to cover most of the
rates  typical for  the Ti02 subcategory (Sulfate Process).   The
selected  daily  Ti02  production   for   the  model  plant  was
multiplied by  the selected unit  flow to obtain the volume of
influent to the  treatment  system.   The selected unit raw waste
pollutant  loads  were  also  multiplied  by  the  model  plant
production rate to determine the pollutant load  on the treatment
system.  Capital and equipment costs were then  calculated based
on developed conceptual design parameters  for each model plant
production  rate.    The  rationale  used  for  the model  plant
selection is given below.

Waste Water Flew

     Waste effluent  data  is available  for  three plants and is
given  in Table  14-23.   For the model  the  average value of the
three  plant data  has been  used.   The unit  flow data for strong
acid ranges from 6.10 to 16 m3/kkg of Ti02.   (Table 14-23).  For
the  model  plant the average value  of  10 m3/kkg has been used.
Unit flows for the weak acid stream range from  67 to 78 m3/kkg.
For  the model plants, a unit flow of 72 m3/kkg  of Ti02 is used,
The  third  segregated  stream   includes  contact  cooling  water,
scrubber water,  and  finishing  operation waste water.  The unit
flow for  this stream  varies  from  plant  to  plant  and depends
largely on the type  and quality of  the Ti02 pigment end product
desired.  For  model plants, a unit flow of 393 m3/kkg of Ti02 was
used.  For model plants a total effluent flow which consists of
the  strong acid,  weak  acid, and scrubber effluent, etcetra, of
475 m3/kkg of Ti02 was used.

Production

     Five plants produce titanium dioxide by the sulfate process
at  a total production  rate  of 259,000  metric  tons  per  year.
Production ranges  from a minimum  of 31,000 kkg/yr to a maximum
of 74,500  kkg/yr  with  an average  of 49,000 kkg/yr and a median
of   43,000  kkg/yr.     For  treatment  cost  estimates,  three
production  levels were  selected.   These  were 31,800 kkg/yr;
47,700 kkg/yr, and 74,500 kkg/yr.

Waste Water Pollutant Load

     As  stated  before, the  principal pollutants  occur  in the
strong and weak acid streams and  include high acidity  (sulfuric

                               431

-------
acid)r suspended  solids,  iron and  other  heavy metal sulfates.
The  other  waste  waters  contain  titanium  dioxide  and  small
amounts  of  other  heavy  metals  as  suspended  solids.    Iron
concentrations vary depending on the grade of  ilmenite ore used,

14.13.2  Model Plant Control and Treatment Costs

     The average raw waste pollutant loadings given  in Table 14-
23 were used for the model plant.  For  the model plants, a total
iron loading  of 600 kg/kkg  was  used with  the assumption that
two-thirds  was suspended ferric  hydroxide  and one-third (200
kg/kkg of Ti02) was soluble  ferrous iron.  The  unit sulfate and
suspended solid loadings  for the different waste water streams
for the model plant were:


                   Sulfate Loading     TSS Loading
Stream              (kg/kkg of Ti02)      (kg/kkg  of Ti02)


Weak Acid               2,300                300
Strong Acid             1,800                200
Other Waste Water      Negligible            113


Chemical Useage

     In  the model BPT  system,  powdered  limestone  is used for
first stage  neutralization  of mixed strong  and weak  acids,  at
the unit rate of 3,000  kg/kkg of Ti02.  Pebble lime (CaO) is used
for second  stage  neutralization of  the mixed  acid  streams and
for  the  final  neutralization  of   the  total combined  flow,,
including the other miscellaneous wastes.  The  unit application
of CaO  for  all purposes  is  0.235  kg/kkg of Ti02.   In Level 2
(which is not used as a regulation basis), soda ash is added to
45% of  the  "other waste" flow at an  approximate dosage of 130
yg/1, to permit partial recycle for miscellaneous purposes.

Solids Produced

     Although some existing plants have attempted  to produce two
grades of saleable gypsum from the strong  and weak acid streams,
at  present   there is  not a sufficient  market for  gypsum  to
justify  byproduct gypsum recovery  in  the  model  plants.   The
solids produced from  the  treatment  facility consist  of gypsum,
iron oxide, and the original suspended solids  introduced in the
influent.   The total  solids produced  in the  model  plant are
assumed to be 5,500 kg/kkg of Ti02.

     Additional solids  generated  in the  soda  ash treatment of
Bother wastes" at Level 2 are only a few hundred pounds per day,
                               432

-------
and  are  considered  a  negligible  increase  in  total  solids
production.     These   additional   solids   are   periodically
transferred  from  the  recycle  polishing  ponds  to  the  main
treatment system just ahead of the aeration step.  In this way,
the additional quantity of toxic metals will be subjected to the
ferric iron flocculation, lime treatment, and settling sequence
in the BPT system.

     The  estimated  costs for  three  models  having  different
production levels  are  given  in Table  14-26,  14-27,  and 14-28.
Annual  treatment  costs as a  function of production  are  shown
graphically  in Figure  14-13.    Similarly,  treatment cost  per
metric ton of product is given in Figure 14-14.

     Table 14-29 presents a summary of  the unit cost distibution
between   amortization   and   operation   and  maintenance   cost
components at different productions and at the BPT and the Level
2 treatment.

     For  existing  sources at  the first level of treatment,  the
disposal  of  sludge  is  on-site,  hence  land  requirements  are
fairly  large.    Amortization,   chemicals,  labor, and  residual
waste  disposal costs  have  significant  impact  on  the  annual
costs.  The treatment Level 2 amortization,  chemicals, and labor
constitute a major portion of the additional costs.

14.14  BASIS FOR REGULATIONS


14.14.1  Evaluation of BPT Practices

     Out  of  a  total  of four Ti02 plants (sulfate process)  that
are  currently  in operation,  only one plant  (|559)  has a  BPT
treatment  system.     The other  3   plants  practice  partial
neutralization and settling.   The proposed  BPT limitations  are
based on  available long-term data from plant |559.

Pollutant Removal with BPT Treatment

     Treatment Level 1 is equivalent to the proposed BPT in the
Titanium  Dioxide  (sulfate process) industry.   Means,  standard
deviations,  and  variability  factors were  calculated  from  data
submitted  by Plant  §559  for  final  effluent quality,  and  the
results   are  given  in  Table  14-30.      The   performance
characteristics are utlized for the development of the proposed
BPT regulations.

     The  ability of the"treatment system  to  remove conventional,
nonconventional, and toxic pollutants was estimated by comparing
the  treated  effluent qualities  with  the raw waste qualities of
the sampled waste streams. The  data  for Plant  #559 are given in
Table 14-31.


                              433

-------
                    TABLE  13-26.  MODEL PLANT TREATMENT COSTS
   Subcategory  TITANIUM DIOXIDE  Sulfate

   Production        31,800 metric tons per year   (35,059 tons per year)
                         90 metric tons per day    (100  tons per day)
   Waste water flow  42750  cubic meters per day.
                                             LEVEL OF TREATMENT*

                                           FIRST            SECOND
    INVESTMENT COST
    Construction 	              $701,200          $117,500
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls	             2,328,400           233,000
    Monitoring equipment
    in place	                 9,000
    Engineering design
    and inspection	               607,720            70,100
    Incidentals, overhead,
    fees, contingencies...               607,720            70,100
    Land	             1,272,000            12,000

    TOTAL INVESTMENT COST             $5,526,040          $502,700

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision.              $504,000           $56,000
    Energy	                96,000             9,000
    Chemicals	             1,589,000           176,000
    Maintenance	               425,404            49,070
    Taxes and insurance...               165,781            15,081
    Residual waste
    disposal	               210,000
    Monitoring, analysis
    and reporting	                15,000             7,500

    TOTAL OPERATION AND
    MAINTENANCE COST                  $3,005,185          $312,651

C.  AMORTIZATION OF
    INVESTMENT COST                     $692,132           $79,836

    TOTAL ANNUAL COST                 $3,697,317          $392,487


    *First level represents the base cost of treatment system.
    Other levels represent the incremental cost above base cost.

                                     434

-------
                 TABLE 14-27.  MODEL PLANT TREATMENT COSTS
Subcategory  TITANIUM DIOXIDE  Sulfate

Production        47,700 metric tons per year   (52,589 tons per year)
                     136 metric tons per day    (150 tons per day)
Waste water flow   64600 cubic meters per day.
                                          LEVEL OF TREATMENT*

                                        FIRST            SECOND
A.  INVESTMENT COST

    Construction 	
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls	
    Monitoring equipment
    in place	
    Engineering design
    and inspection	
    Incidentals, overhead,
    fees, contingencies...
    Land	

    TOTAL INVESTMENT COST

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision.
    Energy	
    Chemicals	
    Maintenance	
    Taxes and insurance...
    Residual waste
    disposal	
    Monitoring, analysis
    and reporting	

    TOTAL OPERATION AND
    MAINTENANCE COST

C.  AMORTIZATION OF
    INVESTMENT COST

    TOTAL ANNUAL COST
                                     $958,700



                                    2,980,200

                                        9,000

                                      789,580

                                      789,580
                                    1,920,000
                                   $7,447,060
                                     $672,000
                                      138,000
                                    2,384,000
                                      552,706
                                      223,411

                                      315,000

                                       15,000
                                   $4,300,117


                                     $899,252

                                   $5,199,369
$161,000
 278,000
  87,800

  87,800
  18,000

$632,600
 $56,000
  12,000
 265,000
  61,460
  18,978
   7,500
$420,938


 $99,995

$520,933
 *First level represents the base cost of treatment system.
 Other levels represent the incremental cost above base cost.
                                  435

-------
                    TABLE 14-28.  MODEL PLANT TREATMENT COSTS
   Subcategory  TITANIUM DIOXIDE  Sulfate

   Production        74,500 metric tons per year
                        212 metric tons per day
   Waste water flow  100700. cubic meters per day.
              (82,136 tons per year)
              (234 tons per day)
                                             LEVEL OF TREATMENT*

                                           FIRST            SECOND
A.  INVESTMENT COST

    Construction	
    Equipment in place,
    including piping,
    fittings, electrical
    wark and controls	
    Monitoring equipment
    in place	
    Engineering design
    and inspection.	
    Incidentals, overhead,
    fees, conting enc ies...
    Land	

    TOTAL INVESTMENT COST

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision.
    Energy.	
    Chemicals	
    Maintenance	
    Taxes and insurance...
    Residual waste
    disposal	
    Monitoring, analysis
    and reporting	

    TOTAL OPERATION AND
    MAD1TENANCE COST

C.  AMORTIZATION OF
    INVESTMENT COST

    TOTAL ANNUAL COST
 $1,293,500



  3,914,500

      9,000

  1,043,400

  1,043,400
  2,940,000
$10,243,800
   $672,000
    199,000
  3,719,000
    730,380
    307,314

    420,000

     15,000


 $6,062,694


 $1,188,328

 $7,251,022
$208,000
 322,000
 106,000

 106,000
  24,000

$766,000
 $56,000
  18<000
 412,000
  74,200
  22,980
   7,500


$590,680


$120,723

$711,403
    *First level represents the base cost of treatment system.
    Other levels represent the incremental cost above base cost.
                                      436

-------
                                j	L
                                1  i
                                                          ; L.EVEL, f2  j

                                              XT i  i  x
                                                 i  i
            i	!	I
             I  I

\    \
                             ' yf \,
                                      II
                                       I  i  I
       20
30       40       50
         EJOTJCTICH (MESOK:
                                            60       70
                                                x 1000)
              80
Figure 14-13.  Jtonual txeatanent cost vs. production for the titanium dioxide
               subcategory, sulfate process.
                                      437

-------
130
1ZO
no
100
90
z


( 1
1 1 '
' '
'
i ! •
• l I f
t
t
. * '
: • ; :
t ' ,
I . j
i ;
' ! ' !
i ' '
• 1








j


l





;







i





































































t
1





















































i
i
i
*
i , , >

;
® i ' •
•\
\ •
\




\


:


1
• ' i



fi

1 i

^ •












X '














































































x
N




















































.





1 •
• i . :

, ;
• ; l i
X' ' '
X • •

X
; ; \_
? '^X
, i . i N
i
;
i

!

X







X




X^i



















































"X












t
















!






























; i
!

i ' ,


i
1 !

, ;



l
• i
i i
\i !
j\
iXi
X






\







X















V




IX





1





















X
^
r























t






















i
L> JU 4U SU b

,
1 ,
' , ' ,
! '
1 1
I !
; | ,
! i i
!
! i !





!
! ,
i 1 ,
;
I
! i !
! ! !
;


i


*s


i
X

X



X

I






i >






X!



x
^





-
































.





S^














i












1 i
1
, i t

!
• 1

: i •
' I
i i
i ', .
'
i
1


1

J

1
I i
| , '



|


1
1
!
1
1
1





!

i


. l
j i
SJ 1LEVEJ
N -
i _®i i







Sy


































,
IwEVEJ
\fflkl
















!
i























4
u 70 a

•

1
' '
!
1 j 1 <
i i :
1 I
1 ' ' '.
1 • I
.
i !
1 1
'
j
1
111.
i : i






i
I
i
j
i
!











1
i i
!

,
;

!
• i
i
1
1

•
^ rfi







1



























!
!





1





1
0
                                      (METRIC TCNS/YERR x 1000)
Figure 14-14.  Amual unit treatment cost vs.  production for the titanium dioxide
               subcategory, sulfate prooass.
                                     438

-------
                 TABLE 14-29.  MODEL PLANT TREATMENT COSTS
Subcategory  TITANIUM DIOXIDE  Sulfate
                                           Annual Treatment Costs ($/kkg)
                  PRODUCTION  FLOW
                   (kkg/yr)  (m3/day)
                             LEVEL OF TREATMENT

                    FIRST     SECOND    THIRD    FOURTH
Annual Operation
and Maintenance
Annual
Amortization
Total Cost
31,800  61,600
47,700  92,600
74,500 144,000


31,800  61,600
47,700  92,600
74,500 144,000

31,800  61,600
47,700  92,600
74,500 144,000
 94.50
 90.15
 81.38
 21.77
 18.85
 15.95

116.27
109.00
 97.33
 9.83
 8.82
 7.93
 2.51
 2.10
 1.62

12.34
10.92
 9.55
Not Applicable
                                439

-------
TABLE 14-30.  KESTORICMi EEFKM3T I-OCCTORING
              SUBCMBGOKr - T33SNIUM DIOXIDE
               snr.TWFE BRQCESS W3NT  4559

Pollutant
TSS Cadmium Chromium
Daily Data
No. of Boints 899 109 128
Average, x 21.0 0.060 0.070
nof-f^ cd) 65.93 0.044 0.054
§ Standard (2) 1.54 0.68 0.67
Deviation, S1
XSfity<3> *•» . 3-85 , 3.a
30-Day Average
No. of Boints 30 26. . ,30
Standard 21 84 0 042 0 038
Deviation , 1*B4 O'°*z °'038
Variability , „ . 2 ._ _ fl.
Factor
Variability
Bactxar Ratio
VFR(5) 3.62 1.58 1.87
Iron .Lead . Nickel Zinc
854 128 128 128
0.62 0.068 0.08 0.151
3.46 0.041 0.071 0.204
1.86 0.56 0.76 1.02
13.65 3.16 4.39 • 6.41

28 30 30 30
0.94 0.04 0.048 0.16
4.00 2.14 4.39 3.05

3.38 1.48 1.00 2.10
(Continued)

-------
TSBLE 14-30.  Continued
 (1)      S is the arithmetic standard deviation and is given by
             »-,/-?
                           n-1
         vtere xi is the data value for point i

               x  is the mean value
               n  is the number of data points

 (2)      S1  is the estimated standard deviation

            S1  =\/ ln(l  4
         iwhere  S  is .the arithmetic standard deviation

               x~  is the mean value

 (3)      Ite variability factor (W) of daily measurements for lognormal  distribution
         is found by the expression,

                             - In (W)  = S1  (2- 0.5 S')
         •where S1  is the estimated standard deviation
                               Z- 2.33 for 99th percentile

                                                                      (Continued)

-------
TfiBIE 14-30.  Continued
 (4)        The variability factor (W) for 30-day average raeasuretnents is found by the
           expression
                    W = 1.0 + Z
                                    8
           TOiere x  is the mean value

                 S is the arithmetic standard deviation

                 Z = 1.64.for 95th pereentile

 (5)       WR:   Ratio of the 24-hour variability factor to the 30-day variability factor

-------
14-31.  VEEOMEATION -RESULTS      - SULFATE
                 DIOXIDE       #559
•"
Bollutant
Total Suspended
Solids
Iron
Cadmium
Chromium
Copper
Lead
Nickel
Zinc
Arsenic
Antimony
Selenium
Thallium
Raw
Waste Treated Effluent
A BCD
Unit Load Concentration Unit Load Concentration
(kg/kkg) (mg/1) (kg/kkg^ fag/1)
116
364
0.00045
1.3
0.070
0.040
0.060
0.45
0.012
0.030
0.0020
0.0030
266 10.0 - 23
835 1.92 4.4
0.0010 0.000040 0.00010
3.1 0.011 0.025
0.16 0.002 0.0050
0.96 0.00090 0.0020
0.14 0.0020 0.0050
1.0 0.030 0.062
0.028 0.0040 0.010
0.074 0.0060 0.015
0.0050 0.0020 0.005-0
0.0070 0.0010 0.0030

E
Rstioval
Efficiency
(%)
91
99
90
99
97
99
96
94
64
80 "
' 0
60
                    443

-------
14.14.2  Basis for Proposed BPT Effluent Limitations

Technology Basis

     For  BPT,  the Agency  is  proposing limitations  based  on
equalization,    limestone    neutralization,    clarification,
aeration,  alkaline precipitation  and  settling followed  by  pH
adjustment  before  final  discharge of   the   effluent.    This
technology is chosen because  it has been  installed  and operated
successfully by a plant in the  industry.

Flow Basis

     Waste flow data  is  available for  3 plants and the average
value of 475 m3/kkg of Ti02 (Table 14-23)  is taken as the inflow
for  the model  plant  treatment system.   The  treatment  plant
effluent is taken to be the same as the  influent and the loss or
addition    of    water    through   chemicals,     evaporation,
precipitation, and through solid removal  have been  neglected.

Selection of Pollutants to be Regulated

     The  selection of pollutants  for  which limitations  are
proposed is based on the evalutation of  raw waste data from the
screening  and verification program.   The following  two  major
factors were considered:

     Raw waste pollutant concentration - Initially one plant was
visted  and the  waste  effluent  sampled  in the screening phase.
The discovery of any toxic pollutants  in the raw waste above the
detection level and in the treatability  range was the basis for
proceeding with verification sampling.   The presence of the same
pollutants in the verification phase confirmed the  significance
of the  screening  program observation.   The pollutants found in
signifcant levels  in  the  raw waste of  the   plants sampled  in
screening and verification are  given in Section 14.10.3.

     On the basis of maximum concentration of pollutants in the
raw waste the preliminary selection of candidates for regulation
included chromium, cadmium, copper, lead,  nickel, zinc, arsenic,
and antimony.

     Total subcategory raw waste pollutant loading - The average
unit toxic pollutant loadings of the raw  waste of plants sampled
in  the  screening  and  verification program   (Table  14-23)  was
multiplied by the  total Ti02  production rate of 259,000 kkg/year
to yield  an  estimate  of  the total annual pollutant loading for
the  subcategory  (see  Section  14.10.3).   The  data  give  an
indication of the  overall magnitude of the pollution potential
for the subcategory.
                              444

-------
Basis of Pollutant Limitations

Conventional and nonconventional parameters

     A.  The  treated effluent  is  to be  controlled within  the
range of pH  6.0  to 9.0.  This  limitation is based on the data
presented in Appendix B of this report  and the JRB  Study  (52).

     B.  TSS and Iron:  The long-term average values of 21 mg/1
for TSS and 0.62 mg/1 for iron derived  from  the monitoring data
of Plant £559 (Table 14-30)  was used as  the proposed subcategory
performance values.  The  variability factors for daily and  30-
day average estimated from Plant 4'559 long-term data  (Table  14-
30)  were  used  in  calculating the concentration basis   and
effluent limitations as shown below.

Total Suspended Solids

     The proposed  TSS maximum  30-day  average concentration  is
given by:

      (21 mg/1) (3.04) = 64 ing/1

     The proposed TSS 24-hour maximum by

      (21 mg/1)(11.0) « 230 mg/1

     The proposed TSS maximum 30-day average effluent limit  was
obtained by using  the  model  plant unit  flow of  475  m3/kkg,
namely

      (64 mg/1)(475 m3/kkg)  /  kg/m3  \   =  30 kg  of TSS
                            UOOO mg/1/       kkg of Ti02

     from:

     The proposed  iron 24-hour maximum  effluent limit:

        (230 mg/1)(475 m3/kkg)  /  kg/m3  \  »   110 kg of TSS
                                \1000 mg/1/       kkg of Ti02

     The proposed  iron maximum 30-day average concentration:

     =  (0.62 mg/1) (4.0) = 2.5 mg/1

     The proposed  iron 24-hour maximum  concentration:

     =  (0.62 mg/1) (13.65)  = 8.5 mg/1

     The proposed  iron maximum 30-day average effluent limit?

        (2.5 mg/1)(475 m3/kkg) /  kg/m3  \  -  1.2  kg of iron
                               \1000 mg/1/      kkg of Ti02

                              445

-------
     The proposed iron 24-hour maximum effluent limit

     =  (8.5 mg/1)(475 m3/kkg) /  kg/m3  N  -  4.1 kg of iron
                               \1000 mg/1/      kkg of Ti02

     Toxic Pollutants  - The  effluent  limitations proposed for
the selected  toxic  pollutant  parameters  are derived  from two
sources of information.  These are 1} long-term monitoring data
for Plant |559, 2) literature-based treatability  estimates.

     If  the   long-term  data  of  a  certain  pollutant  was not
available or  the 30-day  average obtained from the  long-term data
was less  than the lower  level of treatability values, then the
lower limit of treatability was used as the concentration basis
for the maximum  30-day  average limitation.  The  long-term data
of most  of the  toxic  pollutants for Plant  #559 are  given in
Table 14-30.

     A.  Antimony:    The  maximum  concentration of  antimony
observed in the raw waste during the screening and verification
program was  1.4  mg/1 (shown  as  1400  yg/1 in Section 14.10.4).
At Plant  1-559, 80  percent of  the  antimony is  removed during
treatment   (Table  14-31).     The  proposed   30-day  average
concentration  of 0.8  mg/1  is  based  on  the  lower  limit  of
treatability  as determined  by literature  studies (Table 8-11).
A variability factor ratio of 1.9 (ratio of 24-hour variability
factor to the 30-day variability factor) determined for chromium
(Table 14-30) from the long-term data for  Plant $559 was used to
obtain the 24-hour maximum concentration.  Thus:

     The  proposed  antimony  24-hour  maximum concentration  is
     given by;

     (0.80 mg/1)(1.9)  =  1.5 mg/1

     The  proposed  antimony  30-day  average  effluent  limit is
     given by:

     (0.80 mg/1) (475 m3/kkg) /   k'g/m3  \   =   0.38 kg of antimony
                            UOOO mg/1 /        kkg of Ti02

     The  proposed antimony 24-hour  maximum effluent  limit is
     given by:

     (1.5 mg/1)(475 m3/kkg) /  kg/m3  \  =   0.71 kg of antimony
                            VLOQO mg/1/         kkg of Ti02

     B.  Cadmium:  The maximum concentration of cadmium  found in
the raw wastes during the  screening and verification program was
0.340 mg/1 (shown as 340 ng/1  in Section 14.10.4). The data for
Plant f559 indicated a removal efficiency  of 90.0  percent (Table
                              446

-------
14-31).  Thus, the long-term average value of  0.060 mg/1 and the
variability factor of 3.85  for  daily maximum and 2.43 for 30-day
average estimated  from the long-term monitoring  data  of Plant
1559  (Table  14-30)   were   used   in calculating  the  proposed
concentrations and effluent limitations as shown below:

     The proposed cadmium 30-day average concentration is given
     by:

     (0.060 mg/1) (2.43)  =  0.15 mg/1

     The proposed cadmium 24-hour  maximum concentration is given
     by:

     (0.060 mg/1) (3.85)  =  0,24 mg/1

     The proposed cadmium 30-day effluent limit is given by:

     =  (0.15 mg/1)(475 m3/kkg)  /   kg/m3  \= 0.070 kg of cadmium
                                \1000 mg/1/       kkg of Ti02

     The  proposed cadmium  24-hour  maximum  effluent  limit is
     given by:

     (0.24 mg/1)(475 m3/kkg)/  kg/m3  \ = 0.11 kg of cadmium
                            \1000 mg/1/      kkg of Ti02~

     C.  Chromium:  The proposed subcategory limitation of 0.070
mg/1 is based on the average of  the long-term monitoring data
for Plant f559 given  in Table 14-30.  The variability factor of
3.81 for  the  daily data  and the variability  factor of  2.04 for
30-day  averages were  estimated from  the  same data  for Plant
If559r and an established model plant unit  flow of 475 m3/kkg was
used in setting up the proposed limitations.

     The proposed chromium maximum  30-day average concentration
     is given by:

     (0.070 mg/1)(2.04)  =  0.14 mg/1

     The  proposed chromium 24-hour  maximum  concentration is
     given by:

     (0.070 mg/1) (3.81)  =  0.27 mg/1

     The  proposed chromium 30-day  average  effluent  limit is
     given by:

     (0.14 mg/1)(475 m3/kkg) /  kg/m3  \ =  0.070 kg of chromium
                             VlOOO  mg/1/       kkg of Ti02
                              447

-------
     The  proposed  chromium  24-hour  maximum effluent  limit is
     given by:

     (0.27 mg/1)(475 m3/kkg) /  kg/m3  \  = 0.13 kg of chromium
                             V1000 mg/1/       kkg of Ti02

     D.  Copper:   The value of 0.5  mg/lf  which  is  the lower
limit  achieved  from  the  lime-settling of  copper  contaminated
waste  water  from  the  treatability  studies  (Table  8-11)  was
selected  as•the proposed maximum  30-day  average concentration
because  no  long-term  data  for   copper   is  available.    The
variability  factor  ratio of 1.87  developed  from the long-term
data for  Plant  f559  for  chromium  (Table  14-30)  was  used to
estimate  the 24-hour  maximum concentration because performance
of the treatment system is expected to be  the same  for copper as
for chromium .  The  calculations for  the proposed concentrations
and effluent limits are given below:

     The proposed copper 24-hour maximum concentration is given
     by

     (0.50 mg/1)(1,87)   =   0.95 mg/1

     The proposed copper 30-day average effluent limit is given
     by:

     (0.50 mg/1)(475 m3/kkg) f  kg/m3  ^  =  0.24 kg of copper
                             \1000 mg/1/        kkg of Ti02

     The proposed copper 24-hour maximum effluent limit  is given
     by:

     (0.95 mg/1)(475 m3/kkg) /  kg/m3  N =  0.46 kg of copper
                             VLOOO mg/1/    .kkg of Ti02

     E.  Lead:   The lowest concentration of lead achievable by
treatment as determined by treatability  studies  (value of 0.30
mg/1 from Table 8-11 for  lime-settling)  was selected  as  the
proposed maximum 30-day average concentration.  The higher value
was selected because  if  lead is  present in large quantity this
represents  the  achievable level.  The  selected value is higher
than the value obtained by multiplying the long-term average of
.070 mg/1 by the 30-day  variability factor of  2.14 estimated
from the  monitoring  data  for Plant  f559  (Table  14-30) .   The
variability  factor  ratio of  1.48  obtained from  the long-term
monitoring  of lead for  Plant f-559  (Table 14-30)  was  used in
calculating the 24-hour maximum concentration.  The calculations
used  to  establish  the  proposed  concentrations  and  effluent
limitations  are shown below:
                               448

-------
     The proposed  lead 24-hour maximum  concentration is given
     by:

     (0.30 mg/1) (1.5)  =  0.45 mg/1

     The proposed  lead 30-day average effluent  limit is given
     by:

     (0.30 mg/1) (475 m3/kkg) /  kg/m3  \ =  0.14 kg of lead
                             \1000 mg/1/       kkg of T5.02

     The proposed 24-hour maximum effluent limit is given by:

     (0.44 mg/1) (475 m3/kkg) /  kg/m3  \ =  0.21 kg of lead
                             \1000 mg/1/       kkg of Ti02

     F.  Nickel:  The  proposed 30-day average concentration of
0.20  mg/1   is  based   on  the  lower   limit  established  by
treatability  studies  and  achieved  using  lime treatment  and
settling (Table 8-11).  The' proposed daily maximum concentration
was estimated by multiplying the 30-day average concentration by
the variability factor  ratio of 1.87 developed for chromium from
the long-term data for  Plant 1559 (Table 14-30) .  The variability
factor  for  chromium was  used  because the  treatment  system is
expected  to perform similarly for nickel  and chromium.   The
calculations for the proposed concentrations and effluent limits
are given below:

     The proposed nickel  24-hour maximum concentration is given
     by:
     (0.20 mg/H (1.87)  =  0.37 mg/1

     The proposed nickel maximum  30-day average effluent limit
     is given by:

     (0.20 mg/1) (475 m3/kkg) /  kg/m3  \  =   0.10 kg of nickel
                             \1000 mg/1/         kkg of Ti02

     The proposed nickel 24-hour maximum effluent limit  is given
     by:

     (0.37 mg/1) (475 m3/kkg) /  kg/m3  \  =   0.18 kg of nickel
                             UOOO mg/1/         kkg of Ti02

     G.  Zinc:   The  lower  limit established  by  treatability
studies, namely 0.5 mg/1,  (Table 8-11)  was used  as  the basis' for
the  proposed 30-day  average  concentration  limit  because the
observed  average  effluent  concentration   (Table  14-31)  was
considerably less. The variability factor ratio  of  2.1 developed
from  the  long-term  data  for  Plant  1559  for  lead   (Table
                              449

-------
14-30)  was  used   to  estimate  the  proposed  daily  maximum
concentration  since  similar  performance  with  this  treatment
technology is expected.  The calculations used to establish the
concentration basis and effluent limitations are shown below:

     The proposed zinc daily maximum concentration is given by:

      (0.50 mg/1)(2.1)  =  1.1 mg/1

     The proposed  zinc  30-day average effluent  limit is given
     by:

      (0.50 mg/1)(475 m3/kkg) /  kg/m3  \ =  0.24 kg of zinc
                             \1000 mg/1/       kkg of Ti02

     The proposed  zinc  24-hour  maximum effluent limit is given
     by:

      (1.1 mg/1)(475 m3/kkg) /  kg/m3  N =  0.52 kg of zinc
                            \1000 mg/1/      kkg of Ti02

     H.  Arsenic:  The proposed 30-day average concentration of
0.5 mg/1 is based on the lower limit established by treatability
studies for lime precipitation and settling  (Table 8-11) because
no  long-term data  for  arsenic  treatment is  available.   The
proposed   daily  maximum   concentration   was  estimated   by
multiplying the 30-day average concentration with a variability
factor  ratio of 1.9 developed  for  chromium  from the long-term
monitoring  data for  Plant  f559.   The  calculations for  the
proposed concentrations and effluent limits are given below:

     The proposed arsenic  24-hour  maximum concentration is given
     by:

      (0.50 mg/1) (1.9)  = 0.95 mg/1

     The proposed arsenic 30-day average effluent limit is given
     by:

      (0.50 mg/)(475 m3/kkg) /  kg/m3  N  =   0.24 kg of arsenic
                            \1000 mg/1/         kkg of Ti02

     The  proposed   arsenic  24-hour maximum  effluent  limit is
     given by:

      (0.95 mg/1) (475 m3/kkg) /  kg/m3  N  =  0.46 kg of arsenic
                             UOOO mg/1/        kkg of Ti02

     Summary  - A  summary  of   the   proposed  conventional,
nonconventional, and  toxic pollutant  limitations  for BPT  are
given in Table 14-32.
                              450

-------
                     TABLE 14-32.  BROWSED LIMITATIONS
                      THANHM DIOXIDE SULFATE HffiCESS

           Best Practical Control Technology Currently Available
                      Waste ite-iter Flow.  475 m3/kkg of TiO2
Pollutant
Conventional and
Nonconventional
Pollutants
Total Suspended
Solids
Iron
Toxic Pollutants
Antimony
Cadmium
Chrcroium
Copper
Lead
Nickel
Zinc
Arsenic
Sufccategory «,
Performance WR
(rag/1)
21 (2) 3'6
0.62(2) 3.4
0.80(3) 1.9(4)
o.06(2) 1.6
0.07(2) 1.9
0.50<3) 1.9(4)
0.30(3> 1.5<5>
0.20(3) 1.9(4)
0.50(3> 2.1(5)
0.50(3) 1.9
Consentration
Basis
I lax
30-day 24-hr.
Avg Max
64
2.5
0.80
0.15
0.14
0.50
0.30
0.20
0.50
0.50
230
8.5
1.5
0.24
0.27
0.95
0.45
0.37
1.1
0.95
Effluent
Limit
(kg/kka of TiO,)
Max
30-day 24-hr,
Avg Max
30
1.2
0.38
0.070
0.070
0.24
0.14
0.10
0.24
0.24
no
4.1
0.71
0.11
0.13
0.46
0.21
0.18
0.50
0.46
(1)  WR:  Ratio of the 24-hour •variability factor to the 30-day variability
    factor.

(2)  Long-term average based on loading data and variability factors of
    plant #559 selected from Table 14-30.

(3)  The lower limit of the literature treatability estimate  (Table 8-11)
    is used as the basis  for the 30-day average limitation.

(4)  Variability factor ratio of chromium developed from the long-term data
    of plant  #559 has been used   (Table 14-30).

(5)  Variability factor ratio estimated for this pollutant from long-term
    data of plant 1559 has been used.
                                   451

-------
14.14.3  Basis for Proposed BCT Effluent Limitations

     The BCT limitation  (applicable  only to  TSS)  was set equal
to BPT because BAT is equal to BPT.

14.14.4  Basis for Proposed BAT Effluent Limitations

     For  BAT,   the  Agency  is  proposing limitations  based  on
treatment consisting of Level 1 technology, and are the same as
BPT.   A  treatment system  requiring  55  percent recycle through
use  of  soda ash  precipitation  was considered  but  rejected
because  its  performance  has  not  been  demonstrated.    The
limitations proposed for BAT are given in Table 14-33.

14.14.5  Basis for Proposed Hew Source Performance Standards

     Level  1 treatment technology  (also proposed for  BPT  and
BAT) is selected as the basis  for NSPS limitations.  A treatment
system requiring  55 percent  recycle through  use of  soda  ash
precipitation   was  considered  but   rejected  because   its
performance  has not  been  demonstrated.    Compared to BAT,  NSPS
additionally  limits  pH,   TSS   and   iron.    The  proposed  NSPS
limitations are given in Table 14-34.

14.14.6  Basis for Proposed Pretreatment Standards

Existing Sources

     For pretreatment standards for Existing  Sources  (PSES), the
Agency is proposing limitations based on  BAT.  The pollutants to
be limited are iron, antimony, cadmium,   chromium, copper, lead,
nickel and zinc as indicated in Table 14-34.

New Sources

     For  pretreatment  standards  for New Sources  (PSNS),  the
Agency is proposing  limitations based on NSPS.   The pollutants
to be  regulated  are  iron,  antimony,  cadmium, chromium, copper,
lead, nickel and zinc as indicated in Table 14-34.
                              452

-------
                            14-33.  PEDPOSHD LMITATtCNS
                             '         DICKIEE
                        Best Available Technology
                  Waste Water Plow:  475 m3/kkg of TiO0

(2)
Subcategory ,,» WET Concentration
Pollutant Perfcomnce^ Basis (mg/1)


Nonconventional
Pollutants
iron (3)
Toxic Pollutants
Antimony
Cadmium
Chromium
Copper (3)
Lead*3)
Nickel (3)
ZJno(3)
Arsenic


0.62
0.80
0.060
0.070
0.50'
0.30
0.20
0.50
0.50


3.4
1.9
1.6
1.9
1.9
1.5
1.9
2.1
1.9
3u-3ay
Avg
2.5
0.80
0.15
0.14
0.5Q
0.30
0.20
0.50
0.50
24- hour
Max
8.5
1.5'
0.24
0.27
0.95
0.45
0.37
1.1
0.95
Effluent Limit
(kg/kkg of TiC>2)
30-ctay
Avg
1.2
0.38
0.070
0.070
0.24
0.14
0.10
0.24
0.24
24 -hour
Max
4.1
0.71
0.11
0.13
0.46
0.21
0.18
0.52
0.46
(1)   Proposed Lainitations  for BPT Table 14-32

(2)   VtR:  Ratio of the 24-tour -variability factor to the 30-day variability
     factor.

(3)   Also applicable  for PSES and PSJS limitations.
                                    453

-------
      Dollutant
Conventional and
                        TABLE 14-34.  PROPOSED LIMITATIONS
                         TITANIUM DIOXIDE SULFATE PROCESS

                         New Source Performance Standards
                      ftkste Tflater Flow:  475 m3/kkg of TiO.
Subcategory
Performance
                                  (1)   VFR
(2)   Concentration  Effluent Limit
     Basis,  (mg/1)   (kg/kkg of TiO,)
      Max           -  Max       ^
     30-day  24-hour  30-day 24-hour
     Avg     Max     Avg     Max
Nonconventional
Pollutants
Total Suspended
Solids
Iron
Toxic Pollutants
Antimony
CadmivHn
Chromium
Copper
Lead
Nickel
Zinc
Arsenic
21
0.62
0.80
0.060
0.070
0.50
0.30
0.20
0.50
0.50
3.6
3.4
1.9
1.6
1.9
1.9
1.5
1.9
2.1
1.9
64
2.5
0.80
0.15
0.14
0.50
0.30
0.20
0.50
0.50
230
8.5
1.5
0.24
0.27
0.95
0.45
0.37
1.1
0.95
30
1.2
0.38
0.070
0.070
0.24
0.14
0.10
0.24
0.24
110
4.1
0.71
0.11
0.13
0.45
0.21
0.18
0.52
0.46
 (1)  For basis, see proposed limitation for BPT T-able-32,

 (2)  VFR:  Ratio of the 24-hour variability factor to the 30-day variability
     factor.
                                     454

-------
14.15   TITANIUM DIOXIDE -  CHLORIDE  ILMENITE PROCESS INDUSTRY
PROFUSE
14.15.1  General Description

     Total subcategory  production  capacity is  given  in  Table
14-35 Profile   Data Summary.   The 308 data  available for the
Ti02  Subcategory does  not   adequately   cover   the one-step
chloride  ilmenite process;  however, supplementary information
has  been   submitted   by   industry   (55) .    The status   of
regulations    is   presented   in  Table  14-2.      Additional
information  on  the  chloride    process  industry  is   given  in
Section 14.1.

14.15.2  General. Process Description and Raw Materials

     For the manufacture of  titanium dioxide  by the combined ore
benefication-chloride  process,    a  generalized  process flow
diagram including the waste streams is shown in Figure 14-15.

     The direct  use  of   ilmenite  ore  for  the  manufacture of
titanium dioxide  pigments  requires the   application  of either
the       sulfate      process    or   the    one-step       ore
beneficiation/chlorination  process which  is   referred to   in
this  report   as  the chloride-ilmenite   process.    Processes
which  involve  a  separate ore  beneficiation step  (either  at
the  plant or at  the  ore source)  resulting  in an upgraded or a
synthetic  rutile  product to be  used   as  feed material  for  a
chloride process  would  not   be classified   as   a  chloride-
ilmenite  process.  A  separate  ore  beneficiation  process would
fall  within  the   Ore  Mining  and   Dressing  Category   for
regulatory purposes,  and the   manufacture of   TiO?   from   an
upgraded   ilmenite or  synthetic rutile would be   in  the same
classification as a chloride process using natural rutile ore.

     The central  feature of  the  chloride-ilmenite  process is
a  fluidized bed  reactor,   referred  to  as   the  chlorinator,
which   receives  the  ore,  coke,  and  chlorine.  For  any given
ilmenite ore  composition,  the    differential   rates  of the
various    metal    chlori'nation      reactions    taking   place
simultaneously  in  the  chlorinator make  impossible any clear
distinction  between     ore     beneficiation     and   titanium
tetrachloride    formation  steps.     The     reaction  mixture
composition is further  complicated by recycling   recovered ore
from the   quench tower back to   the   chlorinator.   Thus, the
wastes  generated  by  the  process  are    not  separable  into
beneficiation  wastes  and chlorination   wastes.  The chlorinator
acts  as    the  primary   source    of  concentrated  acidic wastes
which  are collected  for  treatment  and disposal from  the ore
recovery   and gas scrubber units.


                              455

-------
EfflLE 14-35
 .SOBCKCEGQRY PROFILE DATA SUMMARY
SOBCATEGORY
TZEKNIDM DIOXIDE Chloride Process CLlmenite Ore)
Total subcategory capacity rate

Total subcategory production rate

Stariber of plants in this subcategory

308 Data on file for
     With total capacity of
     With total production of
     Representing capacity
     Representing production


     Plant production range;
           Minimum
           maximum

     Average production
     Median production
     Average capacity utilization

     Plant age range:

           Minimum
           Maximum

     Waste water flow range:
           Minimum
           Maximum

     Volume per unit product:
           Minimum
           Maximum
                                       m.

                             522,775 kkg/year

                                   4
(1)
                             495,500 kkg/year
                                 NA
                                  95 percent
                             Unknown; wide variation
                             in production
                                    75 kkg/year
                                   228 kkg/year

                                151.50 kkg/year
                                       NA.
                                       NA
                                       NA
                                 8400 cubic meters/day
                               42,000 cubic meters/day
                                   29 cubic meters/kkg
                                  140 cubic meters/kkg
 (1)   Capacity included in Table 14-1.

      Sources of data are Stanford Research Institute,  Directory of Chemical
      Producers, U.S.A., 1977, U.S. Department of Commerce,  Current Industrial
      Reports, December 1977? Energy and Environmental  Analysis,  Inc.,  Draft
      Report, "Preliminary Economic Assessment of Effluent Limitations  in the
      Inorganic Chemical Industry,"      June, 1978,  and "Economic Analysis of
      Proposed Revised Effluent Guidelines and Standards for the Inorganic
      Chemicals Industry ," March, 1980.
         - not available
                                    456

-------
Figure 14-15.  General process flow diagram of the titanium tetrachloride portion
               of a titanium dioxide plant using the chloride-ilrnenitje'process.
                                  45?

-------
     Once  the  T1C14  intermediate  has    been  isolated  and
purified,  the  production of  Ti02  is  basically  the same as
described in  Section 14.1.2 for the ordinary chloride process.


14.16  WATER USE AND WASTE SOURCE CHARACTERISTICS


14.16.1  Water Use

     Water is  used in the manufacture   of titanium dioxide by
the  chloride-ilmenite    process   for   noncontact   cooling,
process   reactions,   air emission  control, product  treatment,
washing  and    transport  operations.   Table  14-36   presents  a
summary  of  water  use data for three plants.

14.16.2  Waste Sources

     The concentrated  process waste  stream  generated by  the
chloride-ilmenite  process contains  the  HC1  generated  in the
chlorination process  along with  iron and other metal chlorides
in  solution.  The waste   stream  also  carries  the spent  coke
and  unreacted ore solids in suspension  (TSS) .

     The other major  sources  of  process contact  waste   water
are  combined  in the  dilute  process waste stream.  These wastes
are    generated  in   the product  finishing  operations  which
include  the  application of  surface coatings  (usually alumina
or silica)  to  the titanium dioxide pigment particles, and the
final dewatering,  washing, drying,   and  sizing  of   the product.
The application of  surface coatings  requires  the  use of acid
and  alkali  to  maintain   the proper  pH  range  for  chemical
treatment  of  the Ti02  slurry.     The    resulting   salts   of
neutralization  are  washed  from  the  product.  These  dilute
acid wastes  are  high in  total   dissolved  solids  and contain
suspended Ti02 from the finishing operations.

     Table 14-37  summarizes  the   average    raw  waste  loads
carried  by the  concentrated  and dilute  process waste  streams
at  three  plants.

     In Table   14-36, considerable differences  in  water  usage
are  revealed  among the  three plants.   These  differences  are
largely  a   reflection of  plant, age in  the   sense  that  the
feasibility   and   economics  of effective  contact/noncontact
waste water segregation  and  recycling  are highly dependent on
the original   plant design  and facilities  layout.   Obviously,
Plant  $713 is   a new  plant which incorporates  modern concepts
of  water  use  and  waste  handling practices and is therefore
used as  the  basis for  the  chloride-ilmenite NSPS.  The  high
flow plant, Plant #237,  •  is  an   older,   existing  facility in
                              458

-------
          TABLE 14-36.   AVERAGE "HATER USAGE FOR TiO~  PRODDCTION

                        BY THE CHLORIDE -HMENTTE PfiOCESS

*
Use Plant #237
*
Nbncontact 73-140
Cooling
Process Contact 100-140
and Clearup
Noncontact 9- 11
Jtacillary Uses
(Boilers,
Sanitary, etc.)
Plant #550
om /kkg of TiOj
330-390

47- 59

6- 7



Plant #713

15-16

29-33 (2)

5- 6



Source of data, (55).

(1)  The average total flow of 120 m /kkg is used as the basis for BPT.


                            3
(2)  The average flow of 31m /kkg is used as the basis for NSPS.
                                    459

-------
                           TKELE 14-37.  AVERAGE MW MSCE LORDS FCR Ti02 ERODUCTION

                                         BY THE          -
ot
o


Plant
#237
Cone. Process Dil. Process
Stream Stream
(kg/kkg of Ti02)
TSS
HC1
Fed3
Other metal
chlorides
100-150
200-230
900-1150
140-155
20-35
8-10
1-3
Negl.
Plant #550
Gone. Process
Stream
(kg/kkg
150-200
250-300
1000-1200
190-210
Dil. Process
Stream
of Ti02)
15-20
0,5-0.8
2-3
Negl.
Plant
#173*
Cone. Process Dil. Process
Stream Stream
.(kg/kkg of TiO2)
200-240
120-240
1000-1200
120-150
5-20
Negl.
Negl.
Negl.
     *  "Ehese mLues are estimated for a new plant prior to start-up.



        Negl. -Negligible (< 0.5)

-------
which process contact water usage is  by far the highest of the
three plants.  Although some  reductions   in the volume and the
relative proportion of contact water usage  may  be feasible, the
economic incentives  are   lacking  and   it   is unlikely   that
older plants  will  be  extensively  modified  to improve water
use  patterns  alone.   This segment was  not further subdivided
because the basic process  is the same.


 14.17  DESCRIPTION OF PLANTS  VISITED AND SAMPLED


 14.17.1  Screening

     Plant f'55Q was  visited during  the  screening phase of  the
sampling program.    This   plant  is ._  capable   of  producing
titanium  dioxide  from  ilmenite  ore by means  of   a  one-step,
integrated   beneficiation/ chlorination process.   However,  at
the  time  of  the   sampling visit,  the  plant   was  not  using
ilmenite ore, but  rather   an  upgraded ore which was similar in
quality to  rutile.   For  this   reason,  the sampling   results
cannot  be  considered  representative   of  a chloride-ilmenite
process and  are  not presented in this  report.

     Plant $550   disposes of  its   concentrated acid  waste  by
deep   well   injection.   These  ferric chloride   laden acidic
wastes are  collected  first in a  system  of four settling ponds
where the bulk  of  the solids are removed.  Dredging  of the ponds
is a continuous  operation  and the  sludges  are landfilled in an
adjacent, on-site    area.   Unlike   the  wastes from  the Ti02-
Sulfate Process, the iron  content of  the  concentrated wastes
from  the  Chloride-Ilmenite  Process is  largely in the ferric
state after   chlorination   and  probably  would not   require
aeration if these wastes were treated  in  a  conventional  BPT
system   utilizing    neutralization   and   settling.    After
settling and  clarification, the acidic wastes  at   Plant  |550
are  deep   well  injected  without prior  neutralization.    The
other process  waste waters from  this plant include the dilute
acid  wastes  from  the  scrubbers   and  white  water  from  the
finishing  operations.   The  dilute  acid wastes are equalized,
neutralized  with  caustic  and  sent to a primary settling pond,
a   polishing   pond,  and finally    a   clear  pool   prior  to
discharge.  The   white  water-  from  product finishing,  first
goes  to  a  slip pond   for  pigment recovery before mixing with
the   neutralized dilute acid   wastes in  the primary    settling
pond.   Noncontact  cooling   water   and  sanitary wastes  are
handled separately.
                              461

-------
14.17.2  Verification Program

     No  verification  program   was   carried    out  in    the
subcategory  since    the    only    other   nominally   chloride-
ilmenite  plant   in operation during the sampling  program was
Plant |237.  However,   this plant was   not  using ilmenite ore
during the  period  of study   and therefore  was  not  visited or
sampled.

14.17.3  Toxic Pollutant Concentration

     Because  neither   of  the  two  operating  Ti02  chloride-
ilmenite  process   plants were  actually using  ilmenite ore at
the time when  the  sampling program  was being conducted,  the
toxic  pollutant    characteristics of  this   process  have been
estimated on  the basis   of sampling  results  from the   Ti02-
sulfate   process at Plant  #559   where  a typical ilmenite was
being used.   The  process  waste water   characteristics of the two
processes are  expected  to  be   similar  because  the  sources of
iron and toxic metal pollutants  are  related to the use of the
same type of ore material.   This segment was not combined with
the sulfate process segment because the manufacturing process is
different.  The  basic  difference between the two processes is
the chemical agent   used in the reaction   with the ore  and this
has     a  significant     impact     on  the   conventional  and
nonconventional  pollutant   parameters,      such   as   acidity,
suspended   and   dissolved solids,  sulfate,  chloride, and iron
(ferrous vs.  ferric).

     Thus, the toxic pollutants found at  potentially significant
levels  in the raw waste  during  sampling  of Titanium Dioxide
Sulfate  process  plants   (Section  14.10.3) are also   presented
here   to  be    used   as   the    basis  for evaluating   the
pollutant  characteristics  of    the   Titanium Dioxide-Chloride
Ilmenite  process.

                Maximum Concentrations Observed
                             (ug/1)
                           Screening              Verification
     Pollutant        (Plants £555 & f559)        (Plant f559)
Chromium
Nickel
line
Lead
Copper
Cadmium
Selenium
Antimony
Thallium
Arsenic
124,000
6,400
3,800
3,700
1,500
340
340
20
19
11
31,000
1,300
17,000
5,200
1,000
12
< 20
1,400
41
340
                              462

-------
     The only organic toxic pollutant found during the screening
program was  phenol which  was  observed  at  only  one of the two
plants  sampled.    The    maximum   raw waste concentration   of
phenol  was  0.020 mg/1,  however,  the raw  water  source for the
plant contained  as  much  as  0.007  mg/1.   Therefore, phenol is
not   considered  a  significant  or   process-related  pollutant
because it is well below the concentration which is treatable.

     Daily raw waste loads   were calculated from the flow rates
measured or estimated at the  time of sampling  and the measured
pollutant concentrations.  That is,

     Daily loading  (as kg of pollutant per  day) = (C)(Q)
                                                   1000

     Where the concentration  (C)  of  the pollutant  is expressed
in  units of mg/1.  (Note:   1  kg/m3  = 1000 mg/1), and  the flow
rate  (Q)  is  expressed in units of m3/day  (m3,  a cubic meter, is
equal  to 264 U.S.  gallons).

     Similarly, the  unit  loadings  were  calculated from  the
reported Ti02  production rate   (P) , the waste  stream  flow rate
(Q), and the measured pollutant concentration  (C).

     Unit loading  (as kg of pollutant per  kkg of Ti02)  =  (C)(Q)
                                                        • 1000(P)

where C and  Q  are expressed in the same  units described above,
and the production (P)   is  expressed  in units  of  kkg/day (kkg is
1000 kg, a metric ton, which is equal to 2205 Ibs).

     A  summary of  daily  and  unit per  unit  of  production raw
waste  loads for the plants sampled  is presented in Table 14-38
and  the  individual plant averages  are given in Table 14-39.

     The estimated  total annual raw  waste water load of  toxic
pollutants generated by  the Chloride-Ilmenite Process  is given
below.

                             Total Annual Raw Waste Water Load
      Pollutant                            (kg/year)

      Chromium                             1,050,000
      Nickel                                 42,000
      Zinc                ,                 178,000
      Lead                                   94,000
      Copper                                 44,000
      Cadmium                                 9,900
      Antimony                               58,000
      Thallium                                2,900
      Arsenic                                99,000
      Selenium                               16,000

                              463

-------
      TABLE 14-38.  SU-MflBY OF RAW HASTE LOADINGS FOUND IN SCREENING AND VERIFICATION SAMPLING

SUBCATEGORY TITANItH DIOXIDE - SULFATE PROCESS (Applied to Chloride Ilroenite Process)
Pollutant Loading Range
(kg/day)
Minimum Maximum
Priority
Antimony 5.0 28
Arsenic 1.9 4.0
Cadmium 0.068 7.2
Chromium 140 530
Copper 8.2 19
Lead 3.0 65
Nickel 3.7 23
Selenium 7.6 9.5
fliallium 0.47 1.3
Zinc 1.8 85
Conventional
OSS
Iron
Minimum

0.032
0.012
0.00044
1.1
0.065
0.024
0.029
0.0020
0.0030
,0.014


Unit Loading
(kg/kkg)
Average

0.11
0.19
0.019
2.0
0.085
0.18
0.080
0.031
0.0055
0.34
320
600
Maximum

0.22
0.032
0.057
3.4
0.12
0.42
0.15
0.06
0 .0080
0.55


No. of
Plants (1)

3
3
3
3
3
3
3
2
2
3
1
1
(1)  - Data are taken only from those plants where pollutants were found above detection limits, or
      in the case of TSS  and Iron, where data are available.

-------
                               TABLE 14-39.  TOXIC POLUUTANT AVERAGE RAW WftSffi LOADS WTO
cn
SUBCATEGORY TITANIUM DIOXIDE — Sulfate Process
(Applied to Chloride Ilmenite
Screening
Plant 1555
(mg/1) (kg/kkg)
antimony
Arsenic
Cadmium
Chromium
Copper
Lead
Nickel
Selenium
Thallium
Zinc
0.77
0.11
0.29
3.8
0.20
0.075
0.091
MA
NA
0.088
0.22
0.032
0.057
1.1
0.065
0.024
0.029
< 0.06
NA
0.014
Plant 1559
(ma/1) (kg/kkg)
0.16
0.029
0.002
7.0
0.25
0.20
0.31
NA
0.02
1.1
0.080
0.014
0.0009
3.4
0.12
0.10
0.15
NA
0.008
0.55
Process)
Verification
Plant 1559
(mg/1) (kg/kkg)
0.074
0.028
0.0010
3.1
U.
0.96
0.14
0.005
0.007
1.04
0.032
0.012
0.00044
1.4
0.070
0.42
0.061
0.002
0.003
0.45
                 NR = Not Available

-------
     Section 5.1.2  of  this  report  described  the  scope   and
methodology  of  the   sampling   program  and  Section  14.10.3
indicates   the  size  of  the  analytical  data  base   on  toxic
pollutants for the  sulfate   process   segment.    This  is   the
basis  for  selecting  pollutants of  concern in the  chloride-
ilmenite segment  of  the  Ti02 subcategory.


14.18  POLLUTION ABATEMENT OPTIONS


14.18.1  Toxic Pollutants of Concern

     Rationale  for   selection  of   the  toxic pollutants  of
concern  is presented in Section  14.11.1 for  the sulfate process
industry.  The  sampling  data   evaluations   resulted  in   the
selection of chromium,  zinc,  nickel,   lead,  copper, antimony,
arsenic,   and   cadmium on  the   basis  of   raw  waste  maximum
concentrations and  total annual industry loads.

     The major  impurity found  in  the  various  grades  of  raw
material  is  ferrous  iron as  shown in Table  14-19.   In  the
sulfate   process  the   unwanted  iron  remains largely  in   the
ferrous state   and may be  crystallized out of the  acid  waste
streams  and  sold  as    coppers   (ferrous  sulfate).   In  the
chloride-ilmenite  process,  the   same  ore impurity  is  largely
oxidized  to  the   ferric  state during  the  chlorination  step.
This  appears  in  the acid  waste streams  as   ferric chloride
(FeC13) in the amounts indicated in Table 14-38.

     Iron, in either the ferrous   or ferric state, is classified
as a nonconventional pollutant.   However,  when  present  in large
amounts,  such as  it  is  in the  Ti02 industry,  it  can  be a
considerable  aid  to toxic  metal removal in treatment systems
designed  to take  advantage  of coprecipitation processes.

14.18.2  Process Modifications and Technology Transfer Options

     The  comments  made  in  regard  to the  Titanium Dioxide-
Chloride  Process for  rutile  and  upgraded  ores  in  Section
14.4.2  are  generally applicable  to  the  Chloride-Ilmenite
Process.

14.18.3  Best Management Practices

     Storm  water    runoff   from  the    plant  site   should  be
collected and sent  to  the treatment  facility  for  the  removal
of suspended  solids.
                              466

-------
14.18.4  Prevailing Control and TreatJnent Practices

     The three   chloride-ilmenite plants from  which the water
use  and waste  source information was obtained  all handle the
disposal  of  the concentrated process  waste stream separately
by   either    ocean  dumping  or  deep  well  injection.    The
availability  of  either   of  these methods   of  disposal  for  a
particular plant is a matter  handled  on a case-by- case basis
by the   appropriate regulatory  agencies from   which  various
approvals     and  permits   are   required    under  the  Marine
Protection,  Research  and  Sanctuaries  Act of  1972   for  ocean
disposal  or  by   state  and  local authorities  for   deep  well
injection.    For the  purpose  of  developing  the model plant
concept  and specifying a generally applicable- waste treatment
technology for the  chloride-ilmenite  industry,  for the purpose
of  this study the  assumption  has  been made that neither the
ocean  dumping  nor   the  deep well  injection disposal  options
are  generally available,  and  that  the concentrated  process
waste  stream is, therefore, included  in  the  raw waste  influent
to the  model plant waste water treatment system.

     In  practice,  one  plant  disposes   of  the  entire  metal
chloride, HC1, and TSS waste by ocean dumping.  The  remainder of
the  plants  dispose of  the  concentrated  waste  by deep  well
injection after use of surface lagoons for removal of settleable
solids.

     The dilute  process  waste streams  are segregated to  the
extent   possible   from  noncontact  sources   and  treated   in
conventional  in-plant systems  utilizing   equalization and spill
diversion        facilities       followed        by       lime
neutralization/coagulation,   solid  separation  in  a settling
pond, and final  discharge  of  the   treated effluent.   Chemical
coagulating  agents  such as   ferric   chloride  and alum  may be
used either  before  or  after pH control as an aid in the removal
of metal hydroxides  and other suspended  solids.

14.18.5  Advanced Treatment Technology

     Advanced   treatment   technology   options    for   in-plant
treatment   of process wastes have been  evaluated  as   possible
polishing step  additions   to  a   conventional    system   for
equalization,   neutralization,   and   clarification  in   ponds
prior to discharge.   Such options include:

     1.  Aeration for  a) deearbonization if limestone  is  -used
for  neutralization,  and  b)  ferrite  coprecipitation, assuming
that sufficient ferrous  iron  is  aleady  present or  is added to
the  system  as needed  (the latter  may   also be accomplished by
adding  scrap iron  to the acid wastes).
                              467

-------
     2.  An   alkaline   precipitation   step   under   optimum
conditions  for metal hydroxide precipitation, i.e.  pH 9-10.

     3.  Dual-media  filtration   for   additional   removal  of
suspended solids including toxic metal hydroxides.

     4. Sulfide precipitation for additional toxic metal removal
followed by filtration.

     5. Other  metal  removal  technologies  including xanthate
precipitation,  ion exchanger  and membrane  applications, all  of
which  were   regarded  as categorically inappropriate  from  a
practical and economic point of view.


14.19  SELECTION OF APPROPRIATE TEC1NOLOGY  AND EQUIPMENT


14.19.1  Technologies for Different Treatment Levels

Level 1 (BPT)

     Figure  14-16 shows the  model treatment  system chosen for
this  • subcategory.   Calcium   carbonate   (limestone)  is  used
to   neutralize  the  concentrated acid   waste   stream.   The
priority   pollutants are precipitated    in  the   first  stage
thickeners.   Aeration  then  oxidizes  any   ferrous  iron present
and     removes    C02   before     mixing    with  dilute   and
miscellaneous plant  wastes.   The combined  stream  is then given
lime treatment to  pH 9-10 for additional  toxic  metals  removal
and   settled  in   polishing   lagoons before discharge.   This
treatment  system   is  patterned after  the  model  plant BPT
waste   water  treatment  technology for  the sulfate process as
presented in this  report.

     This technology was used as  the   treatment  model for  BPT
regulations  because of  the similarity of wastes to  those in the
Ti02-Sulfate Process  industry.   This  technology   is available
and,  to some degree, already  employed  in   the  Ti02-Chloride-
Ilmenite industry.   The  proposed BPT treatment   would remove
greater than 95  percent   of  the major pollutants  of  concern
including toxic   metals according to   preliminary  treatability
estimates.

Level 2  (NSPS)

     Level 2  treatment   adds   dual-media  filtration  to the
Level 1 technology  for additional removal of  suspended solids
and      toxic   metal    hydroxides    following  the  alkaline
precipitation and settling steps.  The flow diagram  for  Level
2   is shown  in  Figure    14-17.   This level of   treatment was
                              468

-------
                                     •a
469

-------
c!
     WlSTE
     siJjrtw
                                                                                                pl MHUSTHBJT
                                                                                                        EFFUJENT
                                                                                       DUO,
                                                                                       (EOIA
                                                                                      FZUTGR
                                        THICXQER
              Figure 14-17.
Level 2 waste water treatment  for titanium dioxide - chloride
(ilmenite ore)  process.

-------
selected  as    the  basis  for  NSPS  because    it provides  a
relatively   economical  method for  removing  additional  toxic
metals.

14.19.2  Equipment for Different Treatment Levels

Equipment Functions

     Unlike  treatment  of    the   waste  waters  from  the  Ti02
Sulfate  Process,   limestone  neutralization   of   the Chloride-
Ilmenite     Process   waste  waters  does  not  generate  large
quantities of solids   (e.g., 'gypsum)  which  require mechanized
separation  and   transfer   to  sizable  on-site or  off-site
disposal areas.   The solids that   are  generated from TSS and
metal  precipitate   separation  can   be   collected in moderate
sized  lagoons  and  periodically  transferred  to  appropriate
chemical   landfill   disposal   sites in  accordance   with   the
Resource Conservation  and Recovery  Act  (RCRA)   (as  amended,
42   USC   6901,  et    seq.) .    The  Level" 1 treatment  model
includes  rail   car  deliveries  of  ground  limestone  amd  line,
bucket elevators, storage  bins, multiple reactors and chemical
feeders,   mechanical    aerators and   thickeners   for   solids
removal.   The  clarified overflow   is   treated with  lime  for
additional   toxic  metals removal   and  settled  in  a  one-day
polishing  pond  prior  to   final pH  adjustment,  monitoring and
discharge.

Chemicals and Chemical Handling

     First  stage  neutralization    utilizes  ground   limestone
while  lime  is used  for second stage neutralization and  final
alkaline    precipitation.    Oxygen  is  supplied as  air  and
treatment  chemicals   may be  added as   required  for  removal of
precipitated metals   and  other suspended  solids.  Aside from
the  large  scale  bulk chemical   handling   requirements   for
limestone    and  lime,    there   are  no   particular  hazards
involved.

Disposal, of Solids

     Periodic  removal of   solids  from  settling   impoundments
will  require compliance with RCRA regulations as  applicable to
on-site  or off-site chemical disposal site operation.
                              471

-------
14.20  TREATMENT COST ESTIMATES


14.20.1  General Discussion

     Preliminary cost  estimates hae  been  prepared  for Level 1
 (BPT)  treatment  only.   Final cost  estimates,   including those
for    Level  2   (NSPS)  will  be   prepared   prior  to   final
promulgation   of     the   regulations.      The  model   plant
specifications given  below were utilized for preliminary  cost
estimating and for development  of,  the proposed regulations.

Production

     There  are  three  plants  at different  locations producing
 (or   capable  of  producing)   titanium dioxide by   the combined
ilmenite   ore  beneficiation  -  chlorination process.  Annual
capacity  of  these plants  varies from 136,000  metric tons  to
207,000  metric   tons.    For  treatment  cost  estimates,  four
production levels were  selected.  These  were 35,000  kkg/year,
70,000 kkg/year,  113,750   kkg/year, and  157,500 kkg/year.

Waste Water Flows

     Waste water  is  typically segregated  into  two  streams;
strong acidic  waste water flow  from beneficiation -chlorination
of  ilmenite   ore and air  emission   scrubbing facilities,   and
the   other   waste  water   from  process  reactions,   washings,
product   transport,  cooling tower blowdown, water   treatment
blowdown, and other  operations.  For the model plants, a unit
flow  of  6  m3/kkg  of  product for  the concentrated acidic waste
water and 114 m3/kkg  of product for  the dilute wastes  is used.
The treatment system  is  designed to handle a total  flow of 120
m3/kkg of product  (Table  14-36).

     For  the  NSPS  model  plant,  a  unit flow  of 6  m3/kkg  of
product   for  the concentrated  acidic    waste water  is used.
Because   of   improved   design  which   allows  for   recycle
systems   and   more  efficient   process   water   utilization,
dilute   waste   water   is   considerably reduced.    The total
combined  waste  water  flow  of  31   m3/kkg of product  is  used
 (Table 14-36) .   The treatment system is  Level  2  which   is BPT
plus dual, media  filtration.

Pollutant Load

     The principal pollutants occurring  in the waste waters are
TSS,  iron,  chromium,  zinc,  and hydrochloric   acid.   For  the
model   plants,  the following  unit  pollutant loads have  been
considered:
                              472

-------
                    TSS                  175 kg/kkg of Ti02
                    HC1                  230 kg/kkg of Ti02
                    Iron                 375 kg/kkg of Ti02
                    Chromium             1.4 kg/kkg of Ti02
                    Zinc                 0.5 kg/kkg of Ti02

     The loading   values for TSS,  HC1,  and iron  are  based on
data    submitted    by   industry   on    the  chloride-ilmenite
process.    The   chromium   loading  is  an estimated   average
derived from  a  wide   range  of ilmenite  ore  qualities and  the
zinc loading  is  taken   from  the  screening   and   verification
data  on  the Ti02  sulfate process.

Chemical Usage

     In the model  BPT  system,  powdered  limestone is used  for
first  stage .neutralization  of  strong  acidic waste flow at the
unit rate of 302 kg/kkg of Ti02.   Pebble lime  (CaO) is  used for
second stage neutralization of  the mixed  acidic  and other waste
waters  and final  neutralization  of the   total combined  flows.
Lime is used at the unit rate of 42 kg/kkg of Ti02.

Solid Waste

     The solids produced in  the   treatment facility  consist of
iron hydroxides,  the   original suspended  solids  introduced in
the  influent and solids  derived from the  treatment chemicals
added   for  neutralization.  The  total  solids produced  in the
model plant  are assumed to be 990  kg/kkg  of Ti02.

14.20.2  Model Plant Control and Treatment Costs

     The estimated  costs  for   four  models  having different
production   levels  are  given  in  Tables  14-40,  14-41,  14-42,
and  .14-43.

     Table   .14-44   presents   a   summary  of  the  unit  cost
distribution  between     amortization   and    operation    and
maintenance cost  components  at   different productions  at the
BPT level  of  treatment.

     For existing  sources  at  the  first   level   of treatment,
the  disposal of  sludges is  on-site,  hence  land requirements
are  fairly large.  Amortization,  chemicals, labor, and residual
waste   disposal  costs  have significant  impact  on  the  annual
costs.

     The unit  waste  flow  of  6  m3/kkg   of product for  the
concentrated acidic waste water stream is  the same for  BPT and
NSPS systems.   The   NSPS  treatment  technology is the  same as
BPT,  but   the total   combined acidic  and dilute waste  water
                              473

-------
                TABLE 14-40.  MODEL PLANT TREATMENT COSTS
   Subcategory   TITANIUM DIOXIDE  Chloride-Il. Ore

   Production            35,000 metric tons per year    (38,587 tons per year)
                            100 metric tons per day     (110 tons per day)
   Waste water flow      12,000 cubic meters per day.


                                            LEVEL OF TREATMENT*

A.   INVESTMENT COST                              FIRST

     Construction 	                    $300,500
     Equipment in place,
     including piping,
     fittings, electrical
     work and controls	                     696,500
     Monitoring equipment
     in place	                 '      9,000
     Engineering design
     and inspection	                     201,200
     Incidentals, overhead,
     fees, contingencies...                     201,200
     Land	                  	252,_000


     TOTAL INVESTMENT COST                   $1,660,400

B.   OPERATION AND
     MAINTENANCE COST

     Labor and supervision.                    $336,000
     Energy	   ,                   31,000
     Chemicals	                     260,000
     Maintenance	                     140,840
     Taxes and insurance...                      49,812
     Residual waste
     disposal	                     105,000
     Monitoring, analysis
     and reporting	                  	i5i222


     TOTAL OPERATION AND
     MAINTENANCE COST                          $937,652

C.   AMORTIZATION OF
     INVESTMENT COST                        	$229^146


     TOTAL ANNUAL COST     .                  $1,166,798


     *First level represents the base cost of treatment system.
     Other levels represent the incremental cost above base cost.

                                     474

-------
                    TABLE 14-41.        PLANT           COSTS
Subcategory   TITANIUM DIOXH3E  Chloride-Il.  Ore

Production          70,000 metric tons per year     (77,175 tons per year)
                       200 metric tons per day      (220 tons per day)
Waste water flow    24,000 cubic meters per day.

                                        LEVEL OF TREATMENT *

A.  INVESTMENT COST                           FIRST

    Construction,..........                $387,500
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls......                 865,000
    Monitoring equipment
    in place.	                   9,000
    Engineering design
    and inspection	                 252,300
    Incidentals, overhead,
    fees, contingencies....                 252,300
    Land......		                 492,000
    TOTAL INVESTMENT COST                $2,258,100

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision.                 $504,000
    Energy		                   43,000
    Chemicals	                  510,000
    Maintenance.	                  176,610
    Taxes and insurance...                   67,743
    Residual waste
    disposal	                  105,000
    Monitoring, analysis
    and reporting.........                   15,000
    TOTAL           AND
                COST                     $1,421,353

C.  AMORTIZATION OF
    INVESTMENT COST                        $287,344


    TOTAL ANNUAL COST                    $1,708,697
    *First level represents the base cost of treatment system.
    Other levels represent the incremental cost above base cost.
                                     475

-------
                    TABLE 14-42.        PLANT           COSTS
Subcategory    T33ANIOM DIOXIDE   Chloride-Il.  Ore

Production       113,750 metric tons per year   (125,409 tons pec year)
                     325 metric tons per day    (358 tons per day)
Waste water flow  39,000 cubic maters per day.


                                        LEVEL OP TREATMENT*

A.  INVESTMENT COST                            FIRST

    Construction	                  $508,000
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls	                 1,179,500
    Monitoring equipment
    in place.	                     9,000
    Engineering design
    and inspection........                   339,300
    Incidentals, overhead,
    fees, contingencies...                   339,300
    Land.	                	2§0£QOO

    TOTAL HSraBSTMENT COST                 $3,155,100

B.  OPERATION AND
    MAINTENANCE COST

    Labor and supervision.                 ,$588,000
    Energy	                    62,000
    Chemicals......	                   823,000
    Maintenance	                   237,510
    Taxes and insurance...                    94,653
    Residual waste
    disposal	                   210,000
    Monitoring, analysis
    and reporting	                	:L§iP_P_P_

    TOTAL OPERATION AND
    MAINTENANCE COST                      $2,030,163

C.  AbDRTIZATION OF
    INVESTMENT COST                       	!2§§ii3§

    TOTAL ANNUAL COST                     $2,416,591
   * First level represents the base cost of treatment system.
   Other levels represent the incremental cost above base cost.
                                     476

-------
                    TABLE 14-43.  MODEL PLANT           COSTS
Subcategory   TITANIUM DIOXIDE  Chloride-Il.  Ore

Production,          157,500 metric tons per year      (173,643 tons per year)
                        450 metric tons per day       (496 tons per day)
Waste water flow     54,000 cubic meters per day.


                                        LEVEL OP TREATMENT*

A.  INVESTMENT COST                             FIRST

    Construction	                 $638,000
    Equipment in place,
    including piping,
    fittings, electrical
    work and controls	                1,356,000
    Monitoring equipment
    in place................                    9,000
    Engineering design
    and inspection.	                  400,600
    Incidentals, overhead
    fees, contingencies.....                  400,600
    Land	                1,080,000
    TOTAL INVESTMENT COST                  $3,884,200

B.  OPERATION MID
    MAIMENANCE COST

    Labor and supervision.                   $588,000
    Energy.	                     71,000
    Chemicals	                  1,141,000
    Maintenance.	                    280,420
    Taxes and.insurance...                    116,526
    Residual waste
    disposal	                    210,000
    Monitoring, analysis
    and reporting	                	i§x.2§2

    TOTAL           AND
                COST                       $2,421,946

C.  AMORTIZATION OP
    INVESTMENT COST                       	li§§£.243

    TOTAL ANNUAL COST                      $2,878,189
   *F±rst level represents the base cost of treatment system.
   Other levels represent the incremental cost above base cost.
                                     477

-------
                    TABLE 14-44.  MODEL PLANT TREATMENT COSTS
Subcategory   TITANIUM DIOXIDE  Chloride-Il.  Ore
                                                    LEVEL OF TREfflMENT
                 PRODUCTION FLOW
                  (kkg/yr)  (m3/day)
                     FIRST
                      SECOND
                        THIRD
Annual Operation
and Maintenance
Annual
Amortization
Total Cost
 35,000
 70,000
113,750
157,500
 35,000
 70,000
113,750
157,500

 35,000
 70,000
113,750
157,500
12,000
24,000
39,000
54,000
12,000
24,000
39,000
54,000

12,000
24,000
39,000
54,000
26.79
20.31
17.85
15.38
 6.55
 4.10
Not Applicable
                                          .40
                                          .90
33.34
24.41
21.24
18.27
                                     478

-------
flow for  NSPS system is much less  than BPT  model;   however,
the reduced   flow has  negligible impact on  costs  because the
unit   waste  loads   are  the same.    There  is  insignificant
difference  in the estimated   total  annual  costs  per  kkg  of
product between BPT and NSPS levels  of treatment for the model
plant designs.
14.21  BASIS FOR REGULATIONS


14.21.1  Evaluation of BPT Treatment Practices

     The prevailing control and treatment practices in the Ti02
Chloride-Ilmenite  industry  have   been   reviewed  in  Section
14.18.4.    For  the  purpose of  regulations development,  it  has
been  assumed   that  neither  ocean   dumping  nor  deep  well
injection  methods  are  generally  available  as disposal options
for all  or  any  portion of  the process-related wastes.   Thus,
treatment technology   used in  the  Ti02-Sulfate Process Segment
of  the  industry has  been used as the  basis  for the proposed
regulations.

14.21.2  Basis for Proposed BPT Effluent  Limitation

Technology Basis

     The  Agency   is     proposing   BPT  limitations  based  on
technology  used  in  the Ti02-Sulfate Process  industry (Section
14.14.2)    involving   equalization, limestone  neutralization,
clarification,  aeration,  alkaline precipitation,  and  settling
followed by final   pH adjustment  and discharge.   The rationale
for   the selection  of  Level  1 technology  is  given in Section
14.19.1.

Flow Basis

     The BPT  model plant  flow rate is  based on the reported
average process  contact  and  clean up  waste  water  flow at Plant
£237 of  120 m3/kkg as  indicated in Table 14-36.

Selection of Pollutants to be Regulated

     The selection of  pollutants to be regulated   in  the Ti02
Chloride-Ilmenite industry is based on  the analysis of raw waste
maximum  concentrations and total  industry loadings as presented
in Section  14.3.3.   The  significant toxic pollutants include:
                              479

-------
                 Chromium
                 Zinc
                 Nickel
                 Iiead
                 Copper
                 Antimony
                 Arsenic
                 Cadmium

     Each of these pollutants  was  present at least once in  raw
waste  water  at  a  maximum  concentration  level  ' regarded   as
treatable   in  accordance  with   the  appropriate treatability
estimates  presented  in  Table  8-11.   Although  arsenic is  a
borderline  case  with an  observed maximum of 0.34  mg/1, it  is
regarded  as  a  candidate  for  regulation  based   on    alkaline
precipitation   technology.     All   of  the  other  pollutants
identified  here as  significant are definately treatable by this
technology.

     This  selection   follows  the    same  logic  presented  in
Section  14.14.2   for  the Ti02-Sulfite Process.   At   the BPT
level,  the  Agency  is also proposing  limitations  on Total
Suspended Solids  (TSS)  and   iron    which  are  classified as
conventional  and   nonconventional pollutants,  respectively.

     Conventional and nonconventional  parameters -

     A.  pH:  The  treated  effluent  is to be controlled within
the  range  of pH 6.0 to 9.0.   This limitation is based on the
data  presented  in  Appendix  B of  this report and the JRB study
(52).

     B,  TSS  and  iron;   The analysis of long  term monitoring
data from   Plant §559  (Table 14-30)   indicates  an achievable
long-term  average  of 21 mg/1 for TSS and  0.62  mg/1  for iron
(total).

     For TSS, the proposed maximum  30-day average  limitation
is  derived from the  long  term average, the variability factor
of  3.0   for  30-day averages  (rounded off  from  3.04  in Table
14-30),  and  the BPT model  plant  flow rate of   120   m3/kkg.
The proposed TSS  maximum 30-day average concentration   basis is
given by:

     (3.0)(21 mg/1) = 63 mg/1

and the proposed TSS maximum 30-day average by:
     (63 mg/1)  (120 m3/kkg)/*  kg/m3  \
                           \1QQO mg/1/
              =7.6 kg/kkg
                              480

-------
     With a  variability   factor  ratio   (VFR)   of   3.6, the
corresponding proposed TSS daily maximum limitation is given by;

     (3.6)(7.6 kg/kkg) = 27 kg/kkg

and the TSS daily maximum concentration basis is:

     (63 mg/1) (3.6)  =  230 mg/1

     Similarly  for  iron,  the  concentration  basis  for  the
proposed  maximum 30-day average  limitation is  derived from the
long term   average  and   a variability   factor  of   4.0.   The
proposed iron maximum 30-day average  (Table 14-31) concentration
are given as:

     (4.0) (0.62 mg/1) = 2.5 mg/1

and the proposed iron limitation is:
      (2.5 mg/1)(120 m3/kkg)/  kg/m3   \
                        •   V 1000 mg/1/
              =0.30 kg/kkg
     The  corresponding proposed  daily maximum  limitation for
iron   is  determined  by  applying  the  variability factor ratio
(VFR)  of  3.4 as follows:

     (3.4) (0.30 kg/kkg) = 1.0 kg/kkg

and the iron daily maximum concentration basis is:

     (3.4) (2.5 mg/1)  .=  8.5 mg/1

The proposed BPT limitations are presented in Table 14-45.

     Toxic Pollutants - For the Ti02 Chloride-Ilmenite process,
the   proposedlimitations  on the   toxic   metals  found at
significant    concentrations  are   based  on    estimates  of
achievable   30-day   „ average concentrations   as  presented in
Table   8-11  because no directly  applicable  industry treatment
performance   data   are   available.     The  lower•  limits  of
treatability  shown  for  lime/setting    are    taken   as    the
concentration  bases  for  the  proposed   maximum  30-day average
limitations  on the various  toxic  metals.     The variability
factor  ratio  (VFR)  used   for  each  pollutant is  identical to
the value  derived  from long-term  monitoring data on Plant f559
presented in Table 14-31.

     A.  Antimony:   The  concentration   basis  for the proposed
maximum  30-day  average   limitation  is  the lower  limit  of
                              481

-------
                    TABLE 14-45.  PROPOSED LIMITATIONS
           Titanium Dioxide - Chloride Process Using Ilmenite
        Best Practicable Control Technology Currently Available
                      Waste Water Flew:  120 m3/kkg

Pollutant
Estimated
Treatability
(mg/1)
Concentration Basis
WRW (rag/1)
Max
30-day
Avg
24-hr
max.
Effluent Limit
(kg/kkg)
Max 24-hr
30-day max.
Avg
Conventional and
Nbnconventional
Pollutants
Total Suspended
Solids
Iron
Toxic Pollutants
Antimony (6)
Arsenic (6)
Cadmium (6)
Chromium (6)
Copper (6)
Lead (6)
Nickel^)
Zinc (6)
2l(2) 3.6 63
0 fi? ' ) 34 75
VS * \J> <-« •w/*T± £,* * +J
0.80(3) 1.9(4) 0<80
0.50(3) l.gte) o,50
o.io (3) i.e(4) o.io
0.10(3) 1.9 (4) 0.10
0.50(3) 1.9(4) 0.50
0.30 (3) 1.5 (4) 0.30
0.20(3) 1.9(4) Qm2Q
0.50(3) 2.1(4) 0.50
230
8.5
1.5
0.95
0.16
0.19
0.95
0.45
0.38
1.1
7.6
0.30
0.096
0.060
0.012
0.012
0.060
0.036
0.024
0.060
27
1.0
0.18
0.11
0.019
0.023
0.11
0.054
0.046
0.013
(1) - WRs  ratio of the 24-hr (daily) variability factor to the 30-day
            average variability factor
(2) - Long term average from Plant t559 monitoring data (Table 14-31)
(3) - Estimated lower limit of treatability as* a 30-day average (Table 8-11)
(4) - Based on long term data from Plant #559(Table 14-30)
(5) - Set equal to the YFR for antimony
(6) - Applicable to proposed BAT and PSES limitations.
                                     482

-------
treatability,  0.8   mg/1 ,  from Table  8-11.   Applying  the BPT
model  plant flow rate of 120 m3/kkg.

     The proposed antimony limitation  is given by:

                                kg/m3
                               1000 mg/1.
     (0.80 mg/1) (120 mS/kkg)/*  kg/m3   \
                                        .)
                = 0.096 kg/kkg

and,  by  applying  the VFR value of 1.9, the proposed antimony
daily maximum limitation is,

     (1.9) (0.096 kg/kkg) = 0.18 kg/kkg.

for which the corresponding concentration is:

     (1.9) (0.80 mg/1) =1.5 mg/1

     B.  Arsenic:   For arsenic,  the lower limit of treatability
is 0.50 mg/1.  Although long-term monitoring  data on arsenic are
not  available,   arsenic is   expected  to behave  in  a  manner
similar   to   antimony  during   lime  treatment and    for  this
reason, the same  VFR value of 1.9  is utilized.

     Thus,   the  proposed   arsenic  maximum   30-day  average
limitation is given by:

     (0.50 mg/1)  (120 m3/kkg)/   kg/m3  \
                             \1000  mg/1/
             = 0.060 kg/kkg

and the proposed antimony daily  limitation by:

     (1.9) (0.060 kg/kkg) = 0.11 kg/kkg

     C.   Cadmiums   The lower limit of treatability for cadmium
is   estimated  at 0.10 mg/1 as a 30-day  average (Table  8-11).
Using  this  value  as  the concentration basis,  the  proposed
cadmium  maximum 30-day limitation  is given  by:
      (0.10 mg/1)  (120 m3/kkg)/"  kg/m3  \
                             VLOOO mg/V
            = 0.012 kg/kkg
and    the    proposed   cadmium  daily  maximum limitation    is
obtained by  applying the  VFR value  of 1.6  from Table 14-31.
That is:

     (1.6)  (0.012 kg/kkg) = 0.019 kg/kkg
                              483

-------
     D,   Chromium:    The   proposed  chromium   limitations are
based   on an estimated 30-day  average   treatability limit  of
0.10 mg/1  using  lime/settling  treatment.   The  achievability
of   this   concentration level  is  predicated on the assumption
that chromium  is  in the trivalent  state  and  no   significant
amount  of the  hexavalent  form is present.

     Thus, the  proposed   chromium   maximum   30-day   average
limitation is given by;
     (0.10 mg/1)  (120 m3/kkg)/  kg/m3  \
                             V1000 mg/1/
             = 0.012 kg/kkg
and  application of  the  WR  value of  1.9  gives  the proposed
chromium daily maximum limitation as,

     (1.9)  (0.012 kg/kkg) = 0.023 kg/kkg.

     E.    Copper:    Using     an  estimated  lower  limit  of
treatability   for   copper  of  0.50  mg/1  and  a  VFR value  of
1.9, the  proposed   limitations are  identical to  those given
above  for arsenic.  The  proposed  maximum  30-day  limitation
is  0.060 kg/kkg  and the proposed  daily maximum limitation is
0.11 kg/kkg.

     P.  Lead:   The  lower  limit  of  treatability for  lead  is
estimated at 0.30 mg/1 as  a 30-day average (Table  8-11) .   Osing
this value  as   the concentration  basis,  the  proposed  lead
maximum  30-day average limitation is given by:

     (0.30 mg/1)  (120 m3/kkg)/" kg/m3  \
                             UOOO mg/1/
              = 0.036 kg/kkg

and, applying the VFR of 1.5  from Table 14-31,  the proposed lead
daily maximum limitation  is:

     (1.5)  (0.036 kg/kkg) = 0.054 kg/kkg


     G.  Nickel:   In a similar manner for  nickel, the  proposed
limitations -are   based on   an  estimated  treatabiliy limit of
0.20 mg/1  and a VFR value of  1.9.  The proposed nickel maximum
30-day  average is given  by:

     (0.20 mg/1)  (120 m3/kkg)/  kg/m3  \
                             VLOOO mg/1/
              = 0.024 kg/kkg

and the proposed nickel daily maximum is given by:
                              484

-------
     (1.9) (0.024 kg/kkg) - 0.046 kg/kkg

     H.  Zinc:  The estimated  treatability limit for   zinc is
the  same as  arsenic and copper,  i.e.,  0.50 mg/1, however, a VFR
value  of 2.1 is applied instead of 1.9.

     Thus, the  proposed zinc  maximum 30-day average limitation
is  given by:
      (0.50 mg/1)  (120 m3/kkg)/  kg/m3  \
                             VOOO mg/1/
             = 0.060 kg/kkg
and the proposed zinc daily maximum is given by:

     (2.1) (0.060 kg/kkg) = 0.13 kg/kfcg

The proposed BPT limitations are presented in Table 14-45.

14.21.3  Basis for Proposed BCT Effluent Limitations

     For  BCT,   the   Agency  is proposing limitations  for TSS
equal  to the BPT limitations because BAT is equal to BPT.

14.21.4  Basis for Proposed BAT Effluent Limitations

     For BAT,  the  Agency is proposing limitations on  iron and
the   toxic pollutants  based   on  the   application of  Level 1
technology  which  is  equivalent  to BPT.  The  model plant flow
basis of 120  m3/kkg   used   for BPT is  also  used  for BAT.  The
proposed  BAT  limitations  are presented in Table 14-45.  A more
advanced technology using soda ash precipitation and recycle of
waste water was considered  for  the similar  sulfate process but
was rejected because its performance has not been demonstrated.

14.21.5  Basis for the Proposed Mew Source Performance Standards

Technology Basis

     For NSPS the Agency is proposing   limitations  based on the
application  of Level  2 treatment technology  which adds  dual
media  filtration  to the  BPT  system  for greater  efficiency
in  the   removal of  suspended   solids   including  iron   and
toxic  metal  precipitates.

Flow Basis

     The reported  data  on  process contact  and clean-up waste
water flow at Plant $713 is selected  as the  basis  of  a  model
plant for  new  sources.  Process modifications  resulting  in a
                              485

-------
greatly  increased  efficiency of  water  use  reduce  the average
flow rate to 31 m3/kkg as shown in Table 14-36.


Basis for Pollutant Limitations

     Conventional parameters -

     A.  pH:  The  treated  effluent is  to  be controlled within
the range of  pH   6.0 to  9.0.  This limitation   is  based on
data presented  in  Appendix B of  this report and  the JRB Study
(52).

     B.  TSS:  The  concentration basis  for  the  proposed NSPS
maximum  30-day average  limitation is   obtained  by  applying an
average filtration efficiency of 38 percent removal  (41)  to  the
corresponding BPT  concentration of 64 mg/1 (Table 14-45).  That
is:

     (1.00-0.38) (64 mg/1) = 40 mg/1:

Then,  the  proposed  maximum   30-day  average  limitation  is
obtained   by  applying  the  NSPS  model plant flow  rate  of 31
m3/kkg:
      (40 mg/1)  (31 m3/kkg)/  kg/m3  \
                          VLOOO mg/1/
             =1.2 kg/kkg
     The proposed   TSS daily maximum limitation is  determined
by   multiplying this  value by the  VPR of  3.6  (Table 14-31) ,
namely:

      (3.6)  (1.2 kg/kkg) =4.3 kg/kkg

and the daily maximum  concentration  basis  is:

      (3.6)  (40 mg/1)   =  140 mg/1

     The same VFR that was  used  in developing  the proposed BPT
limitations  is  also  used    for   NSPS  because  the  actual
variability  • of  the   BPT system with   added  filtration   is
expected   to  be   somewhat less than the statistically derived
VPR for BPT.

     The proposed NSPS limitations are presented in Table  14-46.

     Nonconventional pollutants - The    only  nonconventional
pollutantolconcern  is   iron.    For  NSPS,    the Agency  is
proposing a  maximum   30-day   average  limitation  based   on  an
average  filtration  efficiency of   38  percent   removal  (41).
                              486

-------
                      TABLE 14-46.   PROPOSED LIMITATIONS
              Titanium Dioxide - Chloride Process Using llmenite
                      New Source Performance Standards*
                       Waste Water Flow:  32 m3/kkg

Pollutant
Treatability (2)
(mg/1)
Concentration Basis
(1) (mg/1)
Max
30-day
Ava
24-hr
max.
Effluent Limit
(kg/kkg)
30*12 24~hr
ou— oay max.
Avq
(jonventional and

Nonconventional

  Total suspended   40
  Solids
  Iron              1.6
3.6   40


3.4   1.6
                                                   140
                                                     5.4
1.2
0.050
4.3
0.17
Toxic Pollutants
Antimony ")
Arsenic ^ '
Cadmium' 3)
Chromium (3)
Copper '^)
Lead (35
Nickel (3)
Zd*c<3>

0.80
0.50
0.075
0.040
0.29
0.060
0.17
0.47

1.9
1.9
1.6
1.9
1.9
1.5
1.9
2.1

0.80
0.50
0.075
0.040
0.29
0.060
0.17
0.47

1.5
0.95
0.12
0.076
0.55
0.090
0.32
0.99

0.025
0.016
0.0023
0.0012
0.0090
0.0019
0.0053
0.015

0.048
0.030
0.0037
0.0023
0.017
0.0029
0.010
0".032
(1)  -  VFR:   ratio of the 24-hour daily variability factor to the 30-day
             average variability factor.

(2)  -  Based on the application of pollutant - specific removal efficiencies
       for dual-media filtration  (41)    to adjust the BPT performance on
       treatability estimates shown in  30-day average concentrations in
       Table 14-45.
 *     Including pretreatment standards for new sources (PSNS)  covering iron
       and toxic metals which are expressed as concentrations.

(3)  -  Applicable to proposed PSNS limitations.
                                    487

-------
Thus, the  appropriate  concentration  basis  is derived from the
corresponding  concentration basis of 2.5 mg/1 of iron used for
the  BPT maximum  30-day average  (Table 14-45) .  That is:

     (1.00-0.38)  (2.5 mg/1) = 1.6 mg/1

and the limitation proposed for NSPS  is:
      (1.6 mg/1)  (31 m3/kkg)/  Kg/m3   \
                           V 1000 mg/1/
                = 0.050 kg/kkg

Again, applying the same VFR value of 3.4 that was  used for the
BPT limitations,  the  proposed NSPS daily maximum limitation for
iron  is:

      (3.4)  (0.050 kg/kkg) = 0.17 kg/kkg

and the daily maximum concentration basis is:

      (3.4)  (1.6 mg/1)  =  5.4 mg/1

     Toxic pollutants - The  Agency   is  proposing  new source
performance standards for the  eight  toxic  metals  identified at
significant    concentrations    during    the   screening   and
verification  sampling  program.    To  the  extent  possible,  a
specific   filtration  removal  efficiency derived from published
literature data  (41)  is   applied  for   each  toxic  pollutant
parameter.      The  filtration    removal  efficiency   (percent
removal)  is  applied  to   the  estimated    lower    limit   of
treatability   (Table  8-11)   for   lime/settling treatment   (BPT
basis) to arrive  at the  concentration basis for   each proposed
NSPS maximum 30-day average limitations.

     A«   Antimony  and  arsenic:   No credit  for  additional
removal   by   filtration is  taken   for  either   antimony or
arsenic because removal data is not available.   The  Agency is
proposing NSPS  limitations  for which  the  concentration  bases
are  identical   to  those   used  for    the  development   of BPT
limitations.   Thus, for  antimony,  the proposed   NSPS  maximum
30-day average is given by:

      (0.80 mg/1)  (31 m3/kkg)/  kg/m3  \
                            \1000 mg/1/
             = 0.025 kg/kkg.

and  the corresponding  daily maximum limitation is obtained by
applying the VFR of 1.9, that is:

      (1.9)  (0.025 kg/kkg) = 0.048 kg/kkg.
                              488

-------
     Similarly, for   arsenic   the  proposed  NSPS maximum   30-
day  average limitation is:

     (0.50 mg/1)  (31 m3/kkg) /  kg/m3  \
                             VJLOOO mg/1/
             = 0.016 kg/kkg

and the proposed daily maximum is:

     (1.9) (0.016 kg/kkg) = 0.030 kg/kkg

     B.  Cadmium:  Employing a filtration  removal efficiency of
25  percent  for    cadmium    (41)  results  in    the  following
concentration basis for the proposed NSPS maximum 30-day average
limitation:

     (1.00-0.25)  (0.10 mg/1) = 0.075 mg/1

Therefore, the proposed limitation is:

     (0.075 mg/1)  (31 m3/kkg)/  kg/m3  \
                             VlOOO mg/l}
             » 0.0023 kg/kkg

The corresponding proposed daily maximum limitation:

     (1.6) (0.0023 kg/kkg) = 0.0037 kg/kkg.

and the daily maximum concentration basis  is:

     (1.6) (.075 mg/1) =0.12 mg/1

     C.  Chromium:    For   chromium    the   filtration   removal
efficiency is reported to be approximately 60 percent  (4.1) .

     Thus, for   the   proposed NSPS   maximum   30-day   average
limitations, the concentration basis is given by:

     (1.4)0-0.60)  (0.10 mg/1) = 0.040 mg/1

and the proposed NSPS limitation  is:

     (0.040 mg/1)  (31 m3/kkg)/  kg/m3  \
                             \1000 mg/1/
             = 0.0012 kg/kkg.

     The  proposed  NSPS  daily  maximum limitation  for chromium
is  then  obtained by  applying the VFR  value of  1.9,  that is:

     (1.9) (0.0012 kg/kkg) = 0.0023 kg/kkg.
                               489

-------
and the daily maximum concentration basis  is:

     (1.9) (0.040 mg/1) = 0.076 mg/1

     D,  Copper:  The estimated filtration efficiency for copper
removal   is    approximately    42    percent  (41).  Therefore,
the concentration basis  for  the proposed  NSPS maximum 30-day
average effluent limitation is given by:

     (1.00-0.42)  (0.50 mg/1) = 0.29 mg/lr

and the proposed limitation is:

     (0.29 mg/1)  (31 m3/kkg)/  kg/m3  \ =  0.0090  kg/kkg
                            V1000 mg/1/


     The proposed  NSPS  daily   maximum   is  then  obtained by
multiplying   the maximum 30-day  average  by the VPR  value  of
1.9.   That is:

     (1.9) (0.0090 kg/kkg) = 0.017 kg/kkg

and the daily maximum concentration basis  is:

     (1.9) (.24 mg/1) = 0.55 mg/1

     E.  Lead:   Starting with  the  estimated  BPT  treatability
level   of  0.30   mg/1  for   lead  and applying  a    filtration
removal efficiency of 80 percent  (41), one  obtains:

     (1.00-0.80)  (0.30 mg/1) = 0.060 mg/1

This is the  concentration basis for the  proposed  NSPS maximum
30-day effluent limitation which  is:
      (0.060 mg/1)  (31 mS/kkg)/  kg/m3  >>
                             \1000 mg/1/
             = 0.0019 kg/kkg
     The  proposed  NSPS daily  maximum limitation for   lead  is
then  calculated by multiplying the   30-day average  limitation
by  the  VFR value of  1.5 as follows:

      (1.5)  (0.0019 kg/kkg) = 0.0029 kg/kkg

and the daily maximum  concentration basis is:

      (1.5)  (0.060 mg/1) = 0.090 mg/1
                              490

-------
     One  can    determine  the  concentration  basis  for    the
proposed  daily  maximum limitation by applying  the VFR to  the
concentration  basis for  the maximum 30-day average:

     (1.5) (0.060 mg/1) = 0.090 mg/1

     F.   Nickel:    For   nickel  the  estimated  efficiency of
removal by dual  media filtration   is  approximately 14  percent
(43).  Thus, the  maximum 30-day average concentration  is:

     (1.00-0.14) (0.20 mg/1) - 0.17 mg/1

and  the  proposed   NSPS   maximum  30-day    average   effluent
limitation  for nickel  is:
      (0.17 mg/1)  (31 m3/kkg)/  kg/m3  \
                            \1000 mg/1-/
            = 0.0053 kg/kkg
     The   corresponding   proposed   daily   maximum   effluent
limitation  is obtained by applying the VFR value of  1.9.  That
is:

      (1.9)  (0.0053 kg/kkg) = 0.010 kg/kkg

and the daily maximum concentration basis  is:

      (1.9)  (0.17 mg/1) = 0.32 mg/1

     G.   Zinc:   For zinc,  the removal  efficiency using dual
media filtration is  estimated  at approximately  6 percent  (41).
This  value   is applied  to   the  BPT  concentration basis   to
obtain the NSPS  concentration  basis  as follows:

      (1.00-0.06)  (0.50 mg/1) =  0.47 mg/1

Thus,  the  proposed   NSPS  maximum   30-day average  effluent
limitations for zinc is:

      (0.47 mg/1)  (31 m3/kkg)/   kg/m3  \
                            V.1000 mg/1/
             = 0.015 kg/kkg

and  the  corresponding  proposed  daily  maximum limitation is
obtained  by multiplying  this  limitation by  the VFR value  of
2.1.   That is:

      (2.1)  (0.015 kg/kkg) = 0.032 kg/kkg

and the daily maximum concentration basis  is:
                              491

-------
      (2.1)  (0.47 mg/1) = 0.99 mg/1

The proposed NSPS limitations are presented in Table 14-46.

14.21.6  Basis for Proposed Pretreatment Standards

Existing Sources

     The  Agency  is  proposing  Pretreatment  Standards    for
Existing   Sources   (PSES)   based   on  Level   1   (BPT/BAT)
treatment.   The  pollutants  to be  limited  are iron  and  the
toxic  metals as  indicated in Table 34-45.

New Sources

     Pretreatment Standards  for New Sources   (PSNS)  are being
proposed   by   the  Agency  on   the basis  of   NSPS treatment
technology   for   the   Tl02-Chloride~Ilmenite   industry.   The
pollutants  to be   limited are  iron  and the  toxic   metals as
indicated  in Table 14-46.
                              492

-------
                           SECTION 15


                  ALUMINUM FLUORIDE  INDUSTRY

 *t

15.1  INDUSTRY PROFIIJi


15.1.1  General Description

     Aluminum  fluoride  is  used  as  a  raw   material   in  the
production of cryolite  (sodium  fluoroaluminate) ,  which in turn
is used  in the production of aluminum.   Aluminum fluoride is
used also as a metallurgical  flux (for  welding  rod  coatings), as
a ceramic  flux  (for  glazes and  enamels),  and  as a brazing flux
(for aluminum fabrication).

     The industry profile data for this subcategory are given in
Table 15-1, while the status of regulations is  given in Table
15-2.

15.1.2  General Process Description and Raw Materials

     In  the  dry  process  for  the  manufacture  of  aluminum
fluoride,  partially  dehydrated  alumina hydrate is reacted with
hydrofluoric acid gas.  The reactions  is given as:

       A1203  +  6HF  =  2A1F3    +     3H20                (1)

     The product, aluminum fluoride, is formed as a solid, and
is cooled with noncontact  cooling water before being milled and
shipped.  The gases from the reactor are scrubbed with water to
remove unreacted  hydrofluoric acid before  being  vented  to the
atmosphere.  A simplified  flow  diagram of  the  process is shown
in Figure 15-1.


15.2  HATER USE AND WASTE SOURCE CHARACTERISTICS


15.2.1  Water Use

     Water  is  used  in  noncontact  cooling  of  the  product, for
seals on vacuum pumps and  for scrubbing the reacted gases before

                              493

-------
TKEOZ 15-1.
SOBCKIBQDSf PBCFXGE
SUBCATEGOS^
                     ALOMINtM FtOORIDE
Total subeatespry capacity rate
Total subcategory production rate
Number of plants in this sufacategory
308 Data on file for
    With total capacity of
    With total production of
    Representing capacity
    Representing production
    Plant production range:
            Minirraam
            Mastitaaa
    Average production
    Median production
    Average capacity utilization
    Plant age range:
            Minimum
            Maximum
    Waste water flow range:
            Minimum
            Maximum
    Volune per unit product:
            Minimum
            Maximum
                          134,700 kkg/year
                                5*
                                6
                          204,800 kkg/year
                          120,000 kkg/year
                                UR
                                m

                               38 kkg/year
                           45,600 kkg/year
                           24,300 kkg/year
                           35/500 kkg/year
                               59 percent

                                5 years
                               21 years

                              539 cubic meters/day
                            2,200 cubic meters/day

                                5  cubic maters/kkg
                               12  cubic meters/fckg
Sources of data are Stanford Research Institute., Directory of Chemical
Producers, U.S.A., 1977, U.S. Department of Ccnroerce, Current Industrial
Reports, Decsnber 1977? Energy and Bnviromental analysis, Inc.; Draft
Report, "Preliminary Economic Assessnent of Effluent Limitations in the
Inorganic Chanical industry." June,  1978- and "Eronomic Analysis of Proposed
Bevxsed Effluent Guidelines and Standards for the Inorganic Chemicals Industry,"
March, 1980.
NA =* Not Available
*  Seven plants were operating at the beginning of this study, but two closed down
   production after 1978.
                      494

-------
      15-2 .
SUBPART
                     OF             -          LIMZmTION GUIDELINES
              Aluminum Fluoride



              W  (40 CFR 415.230, 5/22/75)
STANDARDS
BPCTCA* BA3EA*
Product
Process
A1F3




Para-
meters
Fluoride
TSS

Aluminum

Max.1
kg/kkg
tog/l)
0.68
(40) 3
0.86
(51)
0.34
(20)
2
Avg. Max. Avg.
kg/kkg kg/kkcf kg/kkg
0.34
(20)
0.43
(25)
0.17
(10)
NSPS *
Max. Avg.
kg/kkg kg/kkg
(mg/1) (mg/1)





 Sections 415.230, 415.231, and 415.232 were revoked by the Agency

 (41 FR 51601, November 23, 1976)-.



Tflax. = Maximum of any one day.


2.
 'Avg. =Max±raann average-of daily values for thirty consecutive days.



      basis  17,000 1/Kkg.          ,„_
                                   495

-------
IO
         HYDRATED
         ALUMINA
         HYDROGEN
         FLUORIDE
                               WATER
                                I
                                           VENT
                                   SCRUBBER
                           CYCLONE
                          REACTOR
                                            1
                                         WSIG WATER
                                                NONCONTACT
                                                COOLING WATER
                                                  COOLER
  PRODUCT
 COLLECTION
AND STORAGE
 ALUMINUM
•FLUORIDE
 PRODUCT
                       Figure 15-1.  General process  flow diagram for production
                                        of aluminum fluoride.

-------
being vented to the atmosphere.   Water  is  also used for leak and
spill cleanup  and equipment washdown.   Table  15-3  summarizes
water usage in the aluminum fluoride industry.

15.2.2  Waste Sources

Noncontact Cooling Water

     Noncontact cooling water is used to cool the product coming
out of  the  reactor.   In some cases  it  is recirculated and the
blowdown  treated separately from  other process  contact waste
water or  it is discharged  without  treatment.  The water can be
monitored for  fluoride  and if  process contamination occurs, it
can  be  diverted to  the  waste water  treatment facility  for
fluoride  removal.

Floor and Equipment Washings               N

     The  quantity and  quality   of  waste  water  generated  from
these operations varies  and  depend  largely on the housekeeping
practices at the individual plants.

Scrubber Waste Water

     This is the major source of waste water  requiring treatment
before discharge or recycle to the scrubber.  It is contaminated
with  hydrofluoric  acidr aluminum fluoride  and  aluminum oxide,
and, in some cases, sulfuric acid and silicontetrafluoride have
been  detected.     These   originate  as   impurities   in  the
hydrofluoric acid used in  the process.  Table 15-4 presents the
waste water  flows at different  facilities  in  the subcategory.
Noncontact cooling water is excluded  from consideration since it
normally  does not contain  pollutants.

Solid Wastes

     In  aluminum  fluoride  production,   hydrofluoric  gas  and
solids,   such  as aluminum trihydrate  and  aluminum  fluoride,
escape  with  the  vent gases.  During scrubbing,  the  solids are
suspended in the scrubber  water, while hydrofluoric acid gas is
dissolved.    In, the  treatment   facility,  the  waste water  is
neutralized with lime and  calcium fluoride precipitates out and
settles with other suspended solids.  In the majority of cases,
the  solids  are  retained in the lagoon for  periods  up  to ten
years.   Table 15-5  gives  a summary of the amounts  of solids
generated at two aluminum  fluoride plants.

     Different  wastes from  the aluminum  fluoride process are
intermixed  before treatment.   As  mentioned  earlier,  scrubber
water  constitutes  the  major   source  of  waste  water   in  the
aluminum  fluoride  subcategory.   If  the  production  of aluminum
                              497

-------
        TABLE 15-3.  WATER USAGE IN THE ALUMINUM FLUORIDE SUBOYIEGORY
      Source                   Water use per unit of production
                                        3
                                       (in /kkg of  A1F3)

Non-contact cooling
Indirect process
Plant
# 837
14.5
12.2
Plant /9.
t 705 (Z}
NA(1)
1.15
Plant
# 188
6.95
NA
Plant
# 251 (2)
NA
NA
contact  (puiqps, seals,
leakst spills)

Maintenance, e.g.           1.13       2.39         NA        1.02
cleaning and work area
washdown

ScruKber                    3.45       8.92        3.46       18.7
 (1)  NA. = Not Available

 (2)  Currently not manufacturing aluminum fluoride.
                                       498

-------
        TABLE 15-4.  WASTE WATER FLOW AT PLANTS #837, #705 AND #251
                     FOR ALUMINUM FLUORIDE SUBCATEGORY



        Source                         Flow rate per unit of production

                                               ( m /kkg of AlFo)


                                     Plant #837      Plant #705(4)  Plant #251(4)

Scrubber water                          3.45            8.92(2)       18.7(3)

Maintenance equipment                   1.13            2.39           1.02-
cleaning and work area
washdown

Total raw waste flow                    4.58           11.3           19.7
Average of above                                       11.9
three flows


 (1)  All flow information is from 308 Questionnaires and plant visits.  Unit
     flow is calculated by dividing waste water flow in m-Vday by production
     in kkg/day.

 (2)  From Table 15-6  (see footnotes which describe basis of information).

 (3)  From Table 15-7  (see footnotes which describe basis of information).

 (4)  Currently not manufacturing aluminum fluoride.
       TABLE 15-5.  SOLIDS GENERATED AT PLANT #705 AND #251 PRODUCING
                             ALUMINUM FLUORIDE
        Plant                       Total Solids Generated(kg/kkg of A1F-.)
                                                                       < o


        #705(1)                 ,                    54

        #251(1)                                •     69


 (1)  Currently not manufacturing aluminum fluoride.
                                     499

-------
fluoride  is  integrated with hydrofluoric  acid,  then the waste
waters from both plants are combined and treated.
15.3  DESCRIPTION OF PLANTS VISITED AND SAMPLED


15.3.1  Screening

     Plant  f705  was  visited  in  the   screening  phase  of  the
program.   Both  hydrofluoric  acid and  aluminum  fluoride  are
produced  at  this  facility  by the  general  processes described
earlier.   The waste  water  from  the  hydrofluoric acid  and
aluminum  fluoride plants  is mixed and sent  to  the treatment
facility.  At the treatment facility the combined waste water is
neutralized  with  lime  and sent  to a  series of settling ponds.
The effluent from the last  pond  is given a final pH adjustment
before  a portion  is  discharged and  the rest  recycled  to the
process.   Figure  15-2  shows a simplified block  diagram  of the
process  including  the  waste  water   treatment  facility  and
sampling locations.  Table 15-6 presents a summary of flow data
of  the  sampled streams,  and the  data  for  important classical
pollutant parameters.

15.3.2  Verification

     Plant f705 was  visited again and the same streams sampled
in  the  screening  phase also  were  sampled and  analyzed  in the
verification phase.  The  variations in individual stream flows
were  small  during  the two  phases of sampling.    Table  15-6
summarizes   the   flow   data  and  important  conventional  and
nonconventional  pollutant emissions.   A  second  plant  (Plant
$251) was visited and sampled in the verification phase.  Figure
15-3  is a simplified  flow  diagram of the  aluminum- fluoride
manufacturing  plant and   the waste  water  treatment  facility
showing  the  sampling locations.   Table  15-7  presents  the flow
and pollution  concentration data  for  the  plant.   The aluminum
fluoride and hydrofluoric acid waste  streams  are  combined and
sent  to  a  gypsum pond  for  suspended solids  removal.    The
overflow from  the pond  is  mixed  with  alkaline and  acid streams
from  other plants for  neutralization  and  pH adjustment  before
final discharge.

15.3.3  Summary of the Toxic Pollutant Data

     Following  is a  list  of  toxic pollutants which identifies
their maximum  concentration  levels as  found in the raw process
waste streams sampled during screening  and verification.
                              500

-------
                                                        varan
                                                                   VEHf
o
H
                                                                                     'AQUEOUS
                                                                                      IIP
10

TIERE
\
'»
SCHU8BEK
VTER

1
REM2TOR
                                                                                               EFFLUENT

                                                                                         IS     DISCHARGE
                                                                                     IBGBP


                                                                                   Haste atreans sanpled.
          Figure 15-2.  General prowess  flew diagram at Plant  1705 showing the sampling points

                                    (aluminum fluoride manufacture).

-------
                           15-6.  BDOW AND                             OF TOE
                               STRERMS FOR PMBT §705 PBODUCENG MXMEMM FLUORIDE
Saitpling
Phase
Screening



Sanpled
Stream
No.
3
4l)
3&4
5
Verifica- 3
tion Sampling ...


3&4
5
Sanpled
Stream
Description
MF-s scrubber
Surface drains,
cooling tower,
blowdown, etc.
Total raw waste
load
(2)
Treated waste x '
AlFg scrubber
Surface drains,
cooling tower,
blowdown, etc.
Total load
Treated waste ' '
Unit
Flow
(m3/kkg)
8.92
2.39
11.3
24
8.92
2.39
11.3
24
Total
Suspended
Solids
(mg/1) (3) (kg/kkg) W
13,000 120
200 0.48
11,000(5) 120
80 2.0
1,400 13
200 0.48
1,200(5) 13
2.0 0.048
Fluoride
(3) (4)
(mg/1) (kg/kkg)
530 4.7
350 0.82
490 5.5
70 1.6
1400 12
170 0.40
1100 13
20 0.55
Muminum
(3) (4)
(mg/1) (kg/kkg)
780
40
620
10
460
27
370
1.0
7.0
0.10
7.1
0.17
4.1
0.060
4.1
0.012
en
o
to
     (1)   Consists of waste water from HF and A1F-, process.  Flow indicated is estimated portion of total
          flow contributed by MF3 nmintenance and washdown waste water from 308 Questionnaire,.  Total flow
          is 17.8 nr/kkg of product  for both process wastes combined,

     (2)   Consists of waste water from HF and A1F3 process.  Plant currently not manufacturing A1F3.

     (3)   Average of three daily composite  samples during verification and single value obtained during
          screening.

     (4)   kg/kkg of AlF^.                        (5)  Weighted average based on unit flows.

-------
                                                                                          VENT
U1
o
             VENT
             DUST
           COLLECTOR
                       H2S04
WET
SPAR"
     SPAR DRYING
       HANDLING
        LOSSES
HOSE DOWN
   WATER
                    r
HF KILN
                                      WASTE
                  AIR
                      OR|J>
                      ACID

                   WATER
       LEGEND

^ SAMPLING POINTS.
                                                                 AHF
                                                             PURIFICATION
                                                       DILUTION
                                                        WATER  ;
                                                  A!F3  PLANT

                                                   HOSE DOWN
                                                                           t
                                                                      HOSE  DOWN  WATER
                                                                        AHF PLANT
                                                                           NEUTRALIZATION
                                                                               SYSTEM
                                                                                                  WATER
                                                 _J
                                                                                            EFFLUENT
                                                                                            TO RIVER
                                                            ALKALINE STREAMS
                                                            AND ACID FROM OTHER PLANTS

     Figure 15-3,   General process flow diagram at Plant #251 showing  the sampling  points,
                   (aluminum fluoride manufacture).

-------
               OSffiEE 15-7.  HOT MD POELOEftNT OCMENTRATION DAEA. OP THE SAMPLED STOEM6
                              FOR PLANT 1251 PEODOCTOG fflUMMM EHX3RIDE
Stream
No.
Sanpled
Stream
Description
Obit
Plow
Cm3/kkg
of A1P3)
Verification
Sampling
4
6
45j6

2
3
AlP- fecrufcfcer
water
S02 scnabber
water '^
•Dotal raw waste
load
GypstM pond
influent^2'
Q/psum pond
effluent(25
12.6
6.10
18.7

25.1
25.1
Total
Suspended
Solids
(ng/1) (kg/kkg)


1200 16
0.0 0.0
1200 16

19,000 470
9.0 0.23
Fluoride
(mg/1) (kg/kkg)


470 5.90
20 0.14
320 6.0

660 17
320 8.0
Aluminum
(rag/1) (kg/kkg)


50 0.60
0.20 0.0010
50 0.60

26 0.65
22 0.55
8
     (1)  One half flow of SC>2 scrubber water is assumed to contribute to the AlP^ process since the
         total flow is oomnon to the AU^ and HP process.
     (2)  Consists of hydrofluoric acid and aluminum fluoride waste water.  Plant currently not
         manufacturing A1F-,.

-------
           Maximum Haw Waste Concentrations Observed
 Pollutant           Screening                 Verification
                     Plant |705             Plant #705 and f251
Arsenic
Selenium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Cadmium
Antimony
Beryllium
200
68
70
120
25
1.6
150
450
0.70
0
0.80
480
97
1100
250
91
11
290
450
33
3
0







.0
.80
     Section 5.1.2 of this report  describes  the methodology of
the  screening  and  verification  sampling   program.    In  the
aluminum  fluoride  industry,  seven   days   of  sampling  were
conducted at Plants ^705  and f251.   Seven sampling points were
identified and studied for the subcategory.   The evaluation of
toxic pollutant content of  these process-related waste streams
was based  on 637  analytical data points.   The  screening  for
toxic organic pollutants  at  Plant #705 generated an additional
645 analytical  data points.   The daily  raw waste  loads were
calculated ' from  the  waste  stream  flow  rates  measured  or
estimated at  the  time of  sampling and the  measured pollutant
concentration.

     That is,

         Daily loading (as kg of pollutant per day) =
                                                         1000

     Where:

         C is the  concentration of the  pollutant  expressed as
         mg/1 (Note:  kg/m3 = 1000 mg/1) , and

         Q is the Aluminum Fluoride process - waste stream flow
         rate expressed as m3/day.   (m3,  a cubic meter, is equal
         to 264.2 U.S.  gallons)
                              505

-------
     Similarly,  the unit  loadings  were  calculated from  the
reported  aluminum fluoride  production  rate, the  waste stream
flow rate, and the measured pollutant concentration.

     Unit loading  (as kg of pollutant      //-,%
     per kkg of aluminum fluoride)     =  -i — '
                                            1000P

     Where C and Q are the same as  described above, and P is the
     aluminum  fluoride production  rate expressed  as  kkg/day.
     (kkg is 1000 kg, a metric ton, which  is equal to 2205 Ibs.)

     The P  and  Q factors  are for the Aluminum Fluoride Process
     and thereby  the Agency  has segregated  that portion of the
     effluent   attributable   only   to  the  Aluminum  Fluoride
     Process.

     Table  15-8  and  15-9  are a tabulation of  the raw waste and
treated  toxic  pollutant  concentrations  and   loads  determined
during the three plant visits.  The loads  and concentrations are
based  on   the   average   of   three  composite  samples  during
verification  and  one composite sample during  screening.  These
unit loads  were  used  to determine  the minimum,  average,  and
maximum unit loading valves presented in Table 15-10.

     Based on the total annual production of 134,700 kkg/year in
this subcategory  and the  average waste  load generated per unit
product in Table  15-10, the estimated total toxic pollutant raw
waste  loads generated  each  year  for  this subcategory  are as
follows:

         _ Pollutant _ Waste Load  (kg/year)

              Arsenic                       180
              Selenium                      140
              Chromium                      400
              Copper                         94
              Lead                           20
              Mercury                         3.0
              Nickel                        180
              Zinc                          140
              Cadmium                        11
              Antimony                        0.70
              Beryllium                       0.30
                              506

-------
                      15-8.  TOXIC POLLOTRMT        wm       imns AND
    StBCATEGORy
AHMHM HTOKEDE
Pollutant
Arsenic
Selenium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Cadmium
Jtotiitony
Beryllium
Screening
Plant
fag/i) (1)
0.18
0.050
0.030
0.10
0.0050
0.00040
0.11
0.16
0.00020
_J3)
0.00020
#705
(kg/kkg) (2)
0.0020
0.0010
0.00030
0.0010
0.00010
0.0000040
0.0010
0.0020
0.0000020
__<3)
0.0000020
Verification
Plant
(mg/D
0.18
_J3)
0.44
0.070
0.020
0.00040
0.22
0.080
0.010
0.00040
"""*
#705
(kg/kkg)
0.0020
— <3)
0.0050
0.0010
0.00020
0.0000050
0.0030
0.0010
0.00020
0.0000050
— ™
Plant
(mg/1)
0.020
0.050
__(3)
0.010
0.010
0.0030
0.010
0.020
""""•
#251
(kg/kkg)
0.00030
0.0010
__(3)
0.00010
0.00010
0.000050
0.00020
0.00030
..(3)
__^
Average
Concentration
fag/1)
0.13
0.050
0.24
0.060
0.012
0.0013
0.11
0.090
0.0050
0.00040
0.00020
U1
o
     (1)  Concentrations based on average raw waste loads shown and total process production and waste

         flows.


     (2)  kg/kkg  of product.


     (3)  —  below analytical detection limit.                    *

-------
    TABLE 15-9.  TOXIC POLLUTANT EFFLUENT CONCENTRATIONS DURING SAMPLING
SUBCATEGORY
ALUMINUM FLUORIDE
Pollutant

Arsenic
Selenium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Cadmium
Antimony
Beryllium
Plant and Sampling Phase
#705
Screening
fag/D
HD(1)
ND
0.0070
0.10
0.0020
ND
0.050
0.0020
0.0020
ND
0.0020
#705
Verification
fag/D
ND
ND
0.040
0.0010
0.020
ND
ND
0.0010
0.0010
ND
ND
#251
Verification
fag/D
0.0050
0.070
0.22
0.070
0.030
ND
0.45
ND
ND
ND
ND

Average
fag/D
< 0.0050
< 0.070
0.090
0.060
0.020
ND
< 0.25
0.0020
< 0.0020
ND
< 0.0020
(1)  ND — Not Detected.
                                     508

-------
TABLE 15-10.
                            OF RAW WRSEB LOADINGS POUND IN SCHEMING AND VERIFICATION
    SDBCATEGOBy

Pollutant
Ttoxic
Arsenic
Selenium
Chromium
Copper
Lead
Mercury
Nickel
Zinc
Cadmium
Antimony
Beryllium
Conventional and
NDnconventional
TSS
Fluorine
Aluminum
Loading Bange,
kg/day
Minimum Maximum

0.050
0.030
.0.020
0.020
0.0030
0.026
0.026
0.040
0.00010
NA(2)
NA
600
250
100

0.080
0.16
0.22
0.050
0.020
0.0080
0.12
0.080
0.0070
0.00020
0.00010
5400
980
320
Minimum

0.00030
0.0010
0.00030
0.00010
0.00010
0.0000040
0.00020
- 0.00030
0.0000020
NA
NA
13
5.5
0.60
Unit Loading,
Average

0.0013
- O.OOJ.O
0.0030
0.00070
0.00015
0.000020
0.0013
0.0010
0.000080
0.0000050
0.0000020
50
8.1
3.9
kg/kkg
Maximum

0.0020
0.0010
0.0050
0.0010
0.00020
0.000050
0.0030
0.0020
0.00020
NA.
NA
119.0
13.0
7.0
No. of
Plants
Averaged

3
2
2
3
3
3
3
3
2
1
1
3
3
3
Ul
o
     (1)  Average unit loadings from Table 15-8.

     (2)  Not Applicable

-------
15.4  POLLUTION ABATEMENT OPTIONS


15.4.1  Toxic Pollutants of Concern

     The toxic pollutants found in actual plant waste waters are
lead, mercury,  cadmium, antimony,  beryllium,  copper,  arsenic,
chromium, nickel, zinc, and selenium,  in the case of selenium,
it is apparent  that  the source was the raw water supply and is
therefore  not  regarded as  a  process-related  pollutant,  but
control  of  selenium in  the  treated effluent may  be required.
The ones of most concern are chromium  and nickel.

     Copper and chromium may be present  as  trace impurities in
the  hydrofluoric  acid  used   to  react  with  bauxite  to  form
aluminum fluoride.   Arsenic, zinc, and nickel  may originate as
impurities in the bauxite ore.   Waste treatment  processes should
be designed to control TSS, fluoride,  and the significant toxic
metals.   Lead, mercury,  cadmium, antimony, and beryllium are
eliminated  as  toxic  pollutants  of  concern   because  levels
observed are too low to be considered  treatable.

15.4.2  Process Modifications and Technology Transfer options

     1.  Total  recycle  of waste water  to the scrubbers  is
feasible if final neutralization  is with soda ash.  The calcium
in the  waste  is precipitated  as calcium carbonate  and scaling
problems in pipes and scrubbers  are reduced.

     2.  passage of  the  vent  gases from  the reactor  through a
cyclone prior to scrubbing with water  will  remove the aluminum
oxide  and  aluminum  fluoride   particulates.    The  collected
material in the cyclone can be recycled to the  reactor.   The
installation of a  cyclone  will result  in material recovery and
will also reduce  the suspended solids  load  going to  the waste
water treatment facility.

15.4.3  Best Management Practices

     1.  Rainfall  runoff in plant areas,  treatment facilities
and other places  susceptible to  fluoride  contamination  can be
collected and sent to the waste  water  treatment  facility.

     2. If solid wastes containing fluoride  are  stored on land,
studies  should  be  conducted  to   ascertain   the   risk  of
contaminating ground water.  Where necessary, provisions can be
made for collection and  treatment of leachate, permeate, and
runoff.

     3.  Settling  ponds  in  the  waste  water  treatment  facility
should be deep enough (or provided with baffles)  to eliminate or
reduce turbulence caused by wind and rainfall.   This will reduce
the  incidence  of weather-related plant upsets,  and  suspended
solids limitations will be met more consistently.
                             510

-------
15.4.4  Prevailing Control and Treatment Practices

     Plant |705  practices lime neutralization  and  settling of
the  waste  waters.    Since  aluminum  fluoride  production  is
integrated with  hydrofluoric  acid  production,  the waste waters
from the two processes are combined before treatment.  The plant
does not treat noncontact cooling water.

     At Plant  §837 the  tail  gases are  scrubbed  with soda ash
solution,  and  the resulting  solution  is  sent to  an adjacent
facility  for  use.   The  water from  the wet  scrubbers  on the
hydrated alumina dryers  are also sent  to an adjacent facility
for use.  The waste waters  from area washdown  are combined with
other product waste water,  treated  with hydrated  lime and sent
to a settling lagoon before discharge.

     Plant |188  produces  aluminum  fluoride in small quantities
and  in  batches.    The  waste water  from  the  batch operation is
first sent to a  collection pond.  It then goes  to a second pond
where lime and alum are added  and it finally  enters  a  third pond
where the pH is  adjusted by recarbonation.

     Plant  f251  mixes  the   aluminum  fluoride  waste  with
hydrofluoric acid plant waste.  The combined  waste water  is sent
t° gypsum  ponds  for  suspended solids removal.  The supernatant
is treated with  an effluent from another  plant for pH  control
and  neutralization.    Because  of  the  presence  of   complex
fluorides  (from  the HF process)  in the waste waters,  the plant
is planning to use a new proprietary process in the near future
to further reduce  fluoride levels in the  final  effluent-.

15.4.5  Advanced Treatment Technologies

     Metal ions can be precipitated as  hydroxides  at alkaline pH
levels,  and  in  clarified  solutions they may  be  exchanged for
hydrogen  or  sodium ions  by ion exchange.   Metal  ions  at low
levels  may  also  be  controlled   by   xanthate  precipitation,
although the process is not widely used.' Sulfide precipitation
will reduce copper, nickel, and zinc to  low  levels but will not
control  chromium or  arsenic.   Although  the mechanism  is not
clear,  arsenic   levels   appear  to  be   reduced  -in  the  lime
neutralization  process  followed  at  most plants,  perhaps  by
entrapment or  adsorption of the oxide during  the precipitation
of calcium fluoride.   A  combination of lime  and ferric  sulfate
coagulation is probably  the most effective and  practical method
for reducing arsenic concentrations.
                              511

-------
15.5  SELECTION OF APPROPRIATE TECHNOLOGY AND EQUIPMENT


15.5.1  Technologies for Different Treatment Levels

Level 1  (BPT)

     Neutralization with lime  is widely used  in the industry to
remove   the   primary  nonconventional  pollutant   as  calcium
fluoride.   Because  lime  neutralization  to  pH  10  results  in
significant  incidental removal  of toxic pollutants,  alkaline
precipitation was chosen as BPT  (Level 1) technology.  The flow
diagram  is shown in Figure 15-4.

Level 2  (BAT and NSPS)

     A higher  removal of suspended metal hydroxides,  TSS, and
CaP2 can be achieved  by adding  dual  media  filtration  to the
Level I  system.  The flow diagram  is shown in Figure 15-5.

Level 3

     Sulfide precipitation is added to the proposed BAT level of
treatment  to attain  a higher  level  of  heavy metal  removal.
Chromium and  selenium  levels  are  not  appreciably  reduced
although other toxic pollutant levels are.  The flow diagram is
shown in Figure 15-6.

Level 4

     The technology is similar to Level 2, except that soda ash
is  substituted  for  part  of  the  lime   treatment,  permitting
partial  recycling of effluent.  Eighty percent recycle has been
demonstrated and is used  in the development of plant performance
estimates.  The flow diagram is shown in Figure 15-7.

15.5.2  Equipment for Different Treatment Levels

Equipment Functions

     Level 1 consists of flow equalization with first stage lime
application followed by second  stage lime application and lagoon
settling.  The final pH is  adjusted  with hydrochloric acid to
the 6-9  range  before discharge  through  an  effluent monitoring
system.

     In  Level  2,  dual media  filtration is added  to provide
better  control  of  suspended  solids,  including  heavy  metal
hydroxides,  which  are  returned  to  the  .lagoons  as  filter
backwash.
                              512

-------
in
H
                       UMS
O

                      -••Oh
                     BW
                   HASTE WATER
                                                                                                           i WXJUSTOENT
                                                              MIXING
                                                                                  XA90CH
                                               I	fc^H-•»!




                           'Includes flow Monitoring,  pH monitoring and sampler.
                  Figure 15- 4.  Level 1  waste water treatment for  the aluminum fluoride subcategory.

-------
                                                r
                                                                      BSOKKRSH
in
*»
^






RBH
WASTE

1






HATER
r--@-
i
i
i
"



/
^_ /^
— »* 	 /*-"*•»
ECWALIZATICM 1
L»^>T_fc
• ««_








«~ — •
|

MDCIN;





1
1
•jl
i!
-L§*
; f
r







•  i ,j —


             ^Includes flow ncnitoring, jtt nonltxirlng an! aanpler.
                Figure 15-5.  Level 2 waste water treatment  for the  aluminum fluoride subcategory.

-------
                                                                  FEtVUIS SUtHOE
Ul
H
tn
 MM
msrz-
*A   i   Z
                                              r-

                                           -ii
                                             i'A
                                             ll1^-
                                daw manttorlng, pff monitoring and lumpier
                                                                                          X
           Figiure 15-6.   Level 3 waste water  treatment for the aluminum fluoride subcategory.

-------
!
1

r
i
1
i

i
X


             1
                              W
                             I
                             I
                             H
                              ref
                             I
                             &
                             r-
                             in
                             H
516

-------
     In  Level  3,  ferrous  sulfide  is  prepared  on  site  from
ferrous  sulfate  and  sodium bisulfide and is added ahead of the
dual media filter shown  in  Level 2,  to reduce  heavy metals
(except chromium) to lower levels  by sulfide precipitation.'

     Level 4  is  a modification  of  Level 2 which allows partial
recycling of final effluent by substituting soda ash  for part of
the lime treatment, and settling the resulting calcium carbonate
in a clarifier before filtration.  This  step  reduces  the calcium
saturation  and  permits  recycling  of effluent  without serious
scaling  problems.    Although a  small  blowdown of  effluent is
maintained for control  of  salinity the  total mass discharge of
toxic pollutants is  less than  that achieved  in Level 2 due to
the lower effluent flow  rate.

Chemicals and Handling

     In BPT  (Level 1)  and in Level 2, two-stage neutralization
is  accomplished with  lime alone,  using  conventional  handling
equipment to deliver milk  of  lime  to two points of application.
In Level 3r a mixture of ferrous sulfate and  sodium bisulfide is
prepared   in  a  well-ventilated  space  and  applied  with  a
conventional  solution  feeder  to the inlet of the  Level 2 dual
media filter.   With  adequate ventilation and proper pH control
in this  chemical preparation, there  are no unusual problems in
chemical handling.  In Level 4,  soda ash is used to furnish part
of the alkalinity, employing  conventional dry chemical feeding
equipment for this nonhazardous  chemical.

Separation and Removal of Solids

     At  all levels  of  treatment  the precipitated  solids  are
removed mechanically from  the lagoons  at regular intervals and
are  piled  in  self-draining  areas near  the lagoons,  on  land
provided for  a  ten-year operating period.   Fluoride and toxic
pollutants  are   in the  insoluble  or  adsorbed form  and do not
constitute a  hazard  to the local  environment when  left at the
plant site under controlled conditions, i.e., with leachate and
permeate control.

Monitoring Requirements

     Control of  fluoride and toxic pollutants in the treatment
process can be  reasonably  assured  by  pH and  fluoride ion field
testing equipment.  At  advanced"levels very low values  of toxic
metals are detected best by atomic absorption methods,  normally
performed in commercial  laboratories on carefully collected and
composited samples.
                              517

-------
15.6  TREATMENT COST ESTIMATES

15.6.1  General Discussion

     A model plant concept was developed for the subcategory for
treatment cost estimation purposes.  The proposed BPT treatment
system specifications are outlined subsequently.

Waste Mater Flow

     The range of waste water data on file shows flow variations
from  4.58  m3/kkg of  A1F3 to  19.7 m3/kkg  of A1F3  (see Table
15-4).  Based on these values,  a unit flow of 11.9 m3/kkg  of A1P3
was taken  as the average for  the waste  water  treatment model
plant for cost estimating purposes.

Production

     Six  plants  manufacture  aluminum   fluoride  at  a total
production rate of 120,000 kkg/yr.  Individual plant production
rates range  from  a minimum  of  38  kkg/yr to a maximum of 45,600
kkg/yr with  an average of 24,300 and a median of 35,500  kkg/yr.
For  waste  water  treatment  cost  estimates,  three  production
levels  were  selected as  model  plants.    These  three  models
reflect the production levels of the plants for  which data is on
file  (excluding a small  batch operation  plant)  and  are 17,500
kkg/yr, 39,200 kkg/yr and 50,400 kkg/yr.

Pollutant Loadings

     Observed pollutant loadings varied from 14  to 27 kg/kkg of
A1F3 for suspended  solids and from 5.4 to 39.5 kg/kkg  of A1F3
for fluoride.  The data sources are as follows:

     Source  of Data	TSS (kg/kkg-AlF3)    F (kg/kkg-AlF3)

     EPA Document 1974 Ref-      16-20             15-20
     Screening and
     Ver ification
     Phase - Plant Data          14-27             5.4-40


     For model  plants,  pollutant  loadings of  20  kg of total
suspended solids and 18  kg of fluoride per kkg of A1F3 were used
to establish treatment requirements.

Treatment Chemicals

     Lime (CaO powder form)  is added to precipitate fluoride and
to raise the pH to a  six  to  nine  range.   For each of the model
plants, lime is  added at 25 percent  above the stoichiometric
                              518

-------
requirements   for   fluoride   precipitation.      For  advanced
treatment, ferrous sulfide  is  added  to give a concentration of
10  ppm.   This  acts  as  a  polishing  step  to  remove additional
trace metals  from  the  effluent.   For  a more  advanced  level of
treatment, soda  ash  is  added  in addition  to  lime  (CaO) .   The
soda ash dosage was assumed to be 770  kg/kkg.

Variation in Flow and Pollution Loading

     To  indicate  the  effect  on  costs  of  higher  and  lower
pollutant loadings, cost estimates were developed for one model
plant  (35r600  kkg-A!F3/yr)  at  27 kg  of TSS/kkg-AlF3 and  30 kg
fluoride/kkg-AlF3 and 14 kg fluoride/kkg-AlF3.  The waste water
flow for these additional estimates was held constant as in the
original  mode (i.e.,  15 m3/kkg-AlF3).   Unit flows were  also
varied to monitor  the sensitivity of  cost to plant size.   In
this case, the pollutant loadings were  assumed to  be  the same as
in the original model.   The range of  waste water flows used were
10.1 m3/kkg to 22.8 m3/kkg.

Generation of Solids

     From the pollutant  loadings and treatment chemi.-.als above,
the waste treatment residue consists of  20 kg/kkg of suspended
solids plus 46.2 kg/kkg  from  added chemicals.  Thus,' the total
solids generated are 66.2  kg/kkg of  product.   After mechanical
removal  to self-draining piles,  the  combined fluoride As(CaF2)
is  reasonably stable  at the   reaction pH  reached  during  lime
treatment.

Cost Estimates

     The  estimated  costs   for models  having three different
production and four  levels of  treatment  are given in  Tables
15-11, 15-12  and 15-13.   For  these  models, both  the hydraulic
and pollution loadings   per unit  of production are held constant
over the entire range of production.  Annual treatment cost as a
function, of  production  is shown  graphically in Figure  15-8.
Similarly, treatment cost per metric ton of product  is given in
Figure 15-9.

     To  indicate the  effects   on  cost  of  varying  the pollutant
load per unit of product, cost estimates were developed for one
medium-size  production  model  plant   at  higher  solids  and
pollutant (fluoride)  loadings.   For  these models  the hydraulic
load  per unit  of  production  was  held  constant.   The  cost
estimates for these models  are given in Tables 15-14 and 15-15.
The  effects  on  costs  of varying  the  unit pollutant  load are
shown graphically in Figures 15-10 and 15-11 at Levels 1 and 4.
Variation of  pollutant loads  has a significant impact  on Level
1,  but had no effect on the incremental  costs of treatment at
Levels 2 and 3.

                              519

-------
                     TADLE 15-11.    MODEL PLANT TREATMENT COSTS
Subcategory ALUMINUM FLUORIDE
(1)
Production 15,900 metric tons par year (17
45 metric tons per day (50
Waste water flow 540 cubic meters per day.



A. INVESTMENT COST

Equipment in place,
including piping,
fittings, electrical
work and controls.....
Monitoring equipment

Engineering design
and inspection. 	
Incidentals, overhead,
fees, contingencies...
Land 	
TOTAL INVESTMENT COST
8. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Enerdv. ...............

Maintenance. ..........
Taxes and insurance...
Residual waste
disposal. 	 	 	
Monitoring, analysis

TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OP
INVESTMENT COST
v TOTAL ANNUAL COST


FIRST

$39,800



192,000

9,000

48,160

48,160
24,000

$361,120


$56,000
3,400
35,000
33,712
10,833

5,400

15,000


$159,345

$54,849
$214,194

LEVEL OF
SECOND

$10,000



63,000



15,600

15,600


$109,200


$14,000
600

10,920
3,276



7,500


$36,296

$17,766
$54,062
(1)
,529 tons per year)
tons per day)
(2)
TREATMENT
THIRD

$14,000



74,000



17,600

17,600


$123,200


$14,000
900
800
12,320
3,696



7,500


$39,216

$20,044
$59,260


FOURTH

$20,500



172,000



38,500

38,500


$269,500


$14,000
2,500
9,800
26,950
8,085

"

7,500


$63,835

$43,847
$112,682
(1)   Production year is  350
(2)   First level represents
     Other levels represent
days.
the base cost of treatment system.
the incremental cost above base cost.
                          520

-------
                    TABLE 15-12.   sKJDSL PLANT miAWENT COSTS
Subcategory ALUMMlM FLUORIDE
1
(I) (I)
Production 35,600 metric tons per year (39,249 tons per year)
101 metric tons per day (112 tons per day)
Waste water Clow 1200 cubic meters per day.



A. INVESTMENT COST

Equipment in place,
Including piping.
fittings, electrical

Monitoring equipment

Engineering design

Incidentals, overhead,
Sees, contingencies...
tand. 	 	
TOTAL INVES1MSST COST
a. OPERATION AND
MAINTENANCE COST
Labor and supervision.



Taxes and insurance...
Residual waste

Monitoring, analysis
and reporting.........
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTISATION OF
INVESTMENT COST
TOTAL ANNUAL COST


FIRST

$63,600



238,000

9,000

62, 120

62,120
42,000

1476,840


$56,000
5,500
80,000
43,484
14,305

12,500

15,000


$226,789

$70,748
$297,537

LEVEL OP
SECOND

$15,000



84,000



19,800

19,800


$138,600


$14,000
900

13,860
4,158



7,50i


$40,418

$22,550
$62,968
(2)
TREATMENT
THIRD

$19,000



90,508



21,900

21,900


$153,300


$14,000
1,300
1,800
15,330
4,599



7,500


$44,529

$24,941
$69,470


P0URTH

$34,000



259,000



58,600

58,600


$410,200


$14,000
3,100
18,800
41,020
12,306



7,500


$96,72
-------
                     TABLE 15-13.    MODEL PLANT TREATMENT COSTS
Subcategory ALUMINUM FLUORIDE
(1) (U
Production 45,800 metric tons per year (50,494 tons per year)
130 metric tons per day (144 tons per day)
Waste water flow 1550 cubic meters per day.



A. INVESTMENT COST

Equipment in place,
including piping,
fittings, electrical

Monitoring equipment

Engineering design

Incidentals, overhead.
fees, contingencies...
Land 	
TOTAL INVESTMENT COST
8. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Enerav. ...............


Taxes and insurance. . .
Residual waste

Monitoring, analysis

TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
' INVESTMENT COST
TOTAL ANNUAL COST


FIRST

$76 , 500



281,000

9,000

73,300

73,300
60,000

$573,100


$56,000
7,400
100 , 000
51,310
17,193

16,000

15,000


$262,903

$83,481
$346,384

LEVEL OF
SECOND

$20,500



110,000

, *.«.»„« — -

26,100

26,100


$182,700


$14,000
1,500

18,270
5,481



7,500


$46,751

$29,725
$76,476
(2)
TREATMENT
THIRD

$24,500



116,500



28,200

28,200


$197,400


$14,000
1,900
2,400
19,740
5,922



7,500


$51,462

$32,116
$83,578


FOURTH

$43,000



317,000



72,000

72,000


$504,000


$14,000
4,300
26,400
50,400
15,120



7,500


$117,720

$82,000
$199,720
(1)   Production year is 350
(2)   First level represents
     Other levels represent
days.
the base cost of'treatment system.
the incremental cost above base cost.
                       522

-------
   500
o
o
H

X
   400
   300
   200
   100
                                                  OIVEI.
              10        20        30        40        so

                   HO3UCTIGN (METRIC TONS/HEAR X 1000)



   Figure' 15-8.  Annual treatment cost vs. production for the

                          Fluoride  Subcategory
                              523

-------
  20
   15
                       \
•to-
                    \\
                            |V
                        \
                        \i\
                         \\
   10
                                                      13
             10        20       30       40        §0
                 HODOCTICN (METRIC TON/YESR X 1000)

 Figure 15- 9.   Jtanual unit treatment cost vs> pioduction for the
                  Aluminum Fluoride Subcategory
                            524

-------
                    TABLE  15-14.   MODEL KANT TREATMENT COSTS
                                                               (3)
Subeategory AUMMtM FLUORIDE
(I) (1)
Production 35,600 metric tons per year (39,249 tons par year)
101 metric tons per day (112 tons per day)
Waste water flow 1200 cubic meters per day.



A. INVESTMENT COST

Equipment >in place,
including piping,
fittings, electrical

Monitoring equipment
in place. .............
Engineering design
and inspection* .......
Incidentals , overhead,
fees, contingencies...
Land......... 	 	
TOTAL INVESTMENT COST
8. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Snarctv. ...............
Chemicals. ............

Taxes and insurance...
Residual waste
disposal. .............
Monitoring, analysis
and reporting. ........
TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OF
DIVESTMENT COST
TOTAL ANNUAL COST


FIRST

$82,000



241,000

9,000

66,400

66,400
66,000

$530,800


$56,000
5,500
130,000
46,480
15,924

19,000

15,000


$287,904

$75,622
$363,526

LEVEL OF
SECOND

$15,000



84,000



19,800

19,800


$138, S0i


$14,000
900

13,860
4,158



7,500


$40,418

$22,550
$62,968
(2)
TREATMENT
THIRD

$19,000



90,500



21,900

21,900


$153,300


$14,000
1,300
1,800
15,330
4,599



7,500


$44,529

$24,941
$69,470


FOURTH

$34,500



270,000



60,900

60,900


$426,300


$14,000
3,100
31,500
42,630
12,789



7,500


$111,519

$69,359
$180,878
(1)   Production year is 350
(2)   First level represents
     Other levels represent
(3)   Sensitivity Analysis -
days.
the base cost of treatment system.
the incremental cost above base cost.
increased pollutant load.
                       525

-------
                     TABLE 15-15.    MODEL PLANT TREATMENT COSTS
                                                               (3)
Subcategocy ALlMBilM FLUORIDE
(1) (I)
Production 35,600 metric tons per year (39,249 tons per year)
101 metric tons per day (112 tons per day)
Waste water flow 1200 cubic maters per day.



A. INVESTMENT COST

Equipment in place.
including piping.
fittings, electrical

Monitoring equipment

Engineering design

Incidentals, overhead,
Eees, contingencies...
Land 	
TOTAL INVESTMENT COST
8. OPERATION AND
MAINTENANCE COST
Labor and supervision.



Taxes and insurance...
Residual waste
disposal. .............
Monitoring, analysis

TOTAL OPERATION AND
MAINTENANCE COST
C. AMORTIZATION OP
INVESTMENT COST
TOTAL ANNUAL COST


FIRST

$56,900



221,000

9,000

57, 380

57,380
30,000

$431,660


$56,003
5,508
60,000
40,166
12,949

9,000

15,000


$198,615

$65,350
$263,965

LEVEL 0?
SECOND

$15,000



34,300



19,800

19,800


$138,600


$14,000
900

13,860
4,158



7,500


$40,413

$22,550
$62,968
(2)
TREATMaiT
THIRD

$19,000



90,500



21,900

21,900


$153,300


$14,000
1,300
1,800
15,330
4,599



7,500


$44,529

$24,941
$69,470


FOURTH

$34,000



259,000



58,600

58,600


$410,200


$14,000
3,100
14,610
41,020
12,306



7,500


$92,536

$66,739
$159,275
(I)   Production year is 350  days,
(2)   fist level represents the base cost of treatment system.
     Other levels represent  the  incremental cost above basa cost,
(3)   Sensitivity Analysis  -  decreased pollutant load.
                        526

-------
  20
  15
•to-
  10


















































•





























































































































.





















































•
























































,
L















































/
/
DT






























^






s




/
/



w

































\



\
\




/





»llc


































\




s

s
/






RT1



































»



*^
\

s






> 1




































s



s,








'O





































\



\

X





r.T

















(




















X



s

V




m










































'


^






w


































tf
X,



•vj

—



IT





































/


*&


f>

•t





























/-



v^



-»v

"^*

i
rJrw













































TT




v.


^.

^


D









i


























ITT




*s»



"^
^




































?R





•^*.


""^

"•••I







































^.S





«.


(7y




































Fir

































































































T>

















































r.ii

















































m
















































w

















































IT

















































T<





















,



























^A

















































h


























































































































              A1F3 PRODUCTION (METRIC TONS/YEAR X 100Q)

 Figure 15-10.  Effect of variation of pollutant load on treatment
                    cost at level 1 technology
                             527

-------
20
                   k
                  ji
                                        S. Ell
                                           I
                                                        D?D
                   fc
15
                         \
                          \
10
** D'JlC^EMiElb
                          NT
IQ?D
             10       20        30        40       SO        60
              A1P3 PW1DDCTICN (MEaKCC TCNSAEaR X  1000)

Figure 15-11.-  Effect of variation of pollutant load on treatment
                   cost at level 4 technology
                            528

-------
     To judge the effects on cost of varying the hydraulic load
per unit  of  production, cost estimates  were developed for one
medium-size  production model plant at  a higher  and  a  lower
hydraulic loadings.   The pollutant load per unit of production
was held constant for these models.  Tables 15-16 and 15-17 show
the cost estimates.  At treatment Levels 2, 3, and 4 the effects
on  costs  of  varying  the  per  unit hydraulic  load are  shown
graphically in Figures 15-12, 15-13, and 15-14.  Hydraulic load
variation had no significant effect on  the costs of treatment at
Level  1.    Table  15-18  presents  a summary  of  the unit cost
distribution  between  amortization  and  the   operation  and
maintenance cost components at various production and levels of
treatment.   The  effects  on  cost  due to  variations   in unit
pollutant and hydraulic loads are also shown in Table 15-18.

     At the  first  level of  treatment,  chemicals,  labor,  and
amortization have  significant impact on the  annual  costs.   At
the second, third and fourth levels of treatment, the operation
and maintenance  cost comprises  approximately two-thirds of the
additional annual costs, and  the remaining  one-third is due to
amortization.

     Effects  on  annual  costs  arising  from  higher and  lower
pollutant  loads per  unit  of  product  for  a  medium  level  of
production model plant were studied. At high pollutant loading,
the annual  cost at  the first  and  fourth levels  of treatment
increased approximately by 25 and 35 percent, respectively, over
the  base   case   cost.    At  the  second  and third  levels  of
treatment, annual costs per unit of product are the same  as for
the original model.

     At lower pollutant loadings, annual cost at  the first level
of treatment decreased by  15  percent below  the  base case cost.
At other  levels,  annual  costs per unit of product  are the same
as for the original model.

     The  annual  costs arising from higher  and  lower hydraulic
load per unit of product for a medium level of production model
indicated  that  at the  first  level of  treatment,  variation of
hydraulic  loads had  an insignificant  impact  on  annual cost
compared to the original model annual cost.

     In the second, third, and fourth levels of treatment, at a
higher  hydraulic  load,  additional  annual  costs  per   unit  of
production increased by 24,  21,  and  18 percent respectively over
the original model costs.

     At a lower hydraulic load,  additional annual costs per unit
of  production  decreased  by  10  percent  at  second and  third
levels, and  by  16  percent  at  the fourth level,  compared  to the
original model cost.
                              529

-------
                     TABLE 15-16.    TOSL PLANT TREATMENT COSTS(3)
Subcategocy ALIMINIM FLUORIDE
(I) (I)
Production 35,603 metric tons per year (39,249 tons per year)
101 metric tons per day (112 tons per'day)
Waste water flow 2300 cubic meters per day.



A. INVESTMENT COST

Equipment in place,
including piping.
fittings, electrical

Monitoring equipment

Engineering design

Incidentals, overhead,
fees, contingencies...
Land............. 	
TOTAL INVESTMENT COST
B. OPERATION AND
MAINTENANCE COST
Labor and supervision.
Enerov. ...............


Taxes and insurance...
Residual waste

Monitoring, analysis

TOTAL OPERATION AMD
MAINTENANCE: COST
C. AMORTIZATION OF
INVESTMENT COST
TOTAL ANNUAL COST


TOST

$66,100



256,000

9,000

66,220

66,220
42,000

$505,540


$56,000
7,400
80,000
46,334
15,156

12,500

15,000


$232,420

$75,417
$307,837

LEVEL OP
SECOND

$21,000



117,600



27,720

27,720


$194,040


$14,300
1,500

19,404
5,821



7,500


$48,225

$31,570
$79,795
(2)
TREATMENT
THIRD

$25,000



124,000



29,800

29,800


$208,600


$14,000
1,900
1,800
20,860
6,258



7,500


$52,318

$33,939
$86,257


FOURTH

$43,500



321,000



72,903

72,900


$510,300


$14,000
4,700
18,800
51,030
15,309



7,500


$111,339

$83,025
$194,364
(!)   Production year is 350
(2)   First level represents
     Other levels represent
(3)   Sensitivity Analysis -
days.
the base cost of treatment system.
the incremental cost above base cost.
increased hydraulic load.
                        530

-------
                     TABLE 15-17.    MODEL PLANT TREATMENT COSTS ^
Subcategory ALIMDKM fLUORIDE
(I) (1)
Production 35,600 metric tons per year (39,249 tons per year)
101 metric tons per day (112 tons per day)
Waste water flow 1020 cubic meters per day.



(2)

LEVEL OF TREATMENT

A. INVESTMENT COST

Squiptent In place,
including piping,
fittings, electrical

"tonitorinq equipment

Engineering design

Incidentals, overhead,
fees, contingencies...
Land. 	 	 	
TOTAL INVESTMENT COST
3. OPERATION AND
MAINTENANCE COST
tabor and supervision.



Taxes and insurance...
Residual waste

Monitoring, analysis

TOTAL OPSRAf ION AND
MAMTENAICE COST
C. AMORTIZATION 0?
INVESTMENT COST
TOTAL ANNUAL COST
flRST

$63,600



237,000

9,000

61,920

61,920
42,000

$475,440


$56,000
5,500
80,000
43,344
14,263

12,500

15,000


$226,607

$70,520
$297,127
SECOND

$14,500



73,300



16,960

16,960


$118,720


$14,000
600

11,872
3,561



7,500


$37,533

$19,315
$56,848
THIRD

$18,500



76,000



18,900

18,900

$132,300


$14,000
900
1,800
13,230
3,969



7,500


$41,399

$21,525
$62,924
FOURTH

$30, 000



206,000



47,2i0

47,200


$330,400


$14,000
2,500
18,800
33,040
9,912



7,500


$85,752

$53,756
$139,508
(1)   Production year is 350
(2)   First level represents
     Other levels represent
(3)   Sensitivity analysis -
days.
the base cost of treatment system.
the incremental cost above base cost.
decreased hydraulic load.
                        531

-------
   zo
   IS
I?
   10






















































!






























































































































f
4.















































/
/
a














































/


x:












































/



SE




































k







f




H

































\
\
,
>
\
\
\ N
\


/





El




































\
^


s
^
/






) 1







































\

^
\
s





IY









































s

(\
>




DE









•
































\

s\




M











































\.

s^



L:











































+
\
>s
N


:c













































N
^x
N.

L









































s



V

^

Ob









































s




Is
im
TS
D








































S-







1 ^
s^









































-I






*«.
•*>
^«








































X.






•»
*^
**•
















1























SE







^-**
""""•J
"%i







































i&







•«.
~~(i
~~*







































!E








h








































p

















































En

















































D^

















































yy

















































E.

















































EC

















































I

















s































d

















































^D





























































































































              10       20       30        40        SO         60
                MP3 PBGDUCTTCN (B5ETRIC TClK/IiaR X 1000)

    Figure 15-12.  Effect of variation of hyojcaulic load on treatment
                        cost at level 2 technology
                           532

-------
ANNOMi TREATMENT COST ($/METRIC TON)
t- t- fJ
01 O W> O




































































































































Zl
















































/*
m













-

































/

T"














































/


?F













































/



A?
I


































*
n
1

^



(
*




FT
n?















.



















.
\
l
\
i
\
j
^





) l
D






































\

\
s






rv








































\
^
\
N k





DR










































\
S
V
^



PiT











































\
\




rrr












































\
v,
s


rr














































\
V

































































































>
Vr


y

vj x^-
^



















X
XI


















«S
Vkg










































/
'



k.
•x.
"•»









































/







^







































^
^







»
">•
"*






































•^*








•*^
• —







































"M










j!






































•rp















































7B

















































P|
T
















































T>
D7
6-A















































fl
,rv
UJ























,























iTi

















































pp

















































TTf



























1





















iTf
















































T





























































































































10 20 30 40 50 60
MP3 PRODUCTION (METRIC TONS/YEaR X 1000)
Fipare 15-13.  Effect of variation, of bydtaulie load on treatment
                  cost at level 3 technology
                        533

-------
N
O
C
•to
W





I
»-<

Ul
•-»

O
              c-rniY
                      ±
                    X
                         \
   \
jK3Eaii;
                                    y
                                       £=££
                          KDIMfll
                                                      i
IQIJ3
               10       20        30        40       SO        60

                 A1F3 HRODDCTICN (METRIC TCNS/XEAR X 1000)



      Figure 15-14.  Effect of variation of hydraulic load on treatment

                        cost at level 4 technology
                               534

-------
              TABLE 15-18.   MODEL PLANT TREATMENT COSTS
Subcategory  ALUMINUM FLUORIDE
                             Annual Treatment Costs/Metric Ton of Product
 <~OST ITEMS
Annual Operation
and Maintenance
Annual
Amortization
Total Cost
PRODUCTION   FLOW      FIRST
(kkg/yr)    (m3/day)
LEVEL OF TREATMENT

SECOND    THIRD     FOURTH
tion
nee 15,900
35,600
45,800
a 35,600
b 35,600
c 35,600
d 35,600
15,900
35,600
45,800
a 35,600
b 35,600
c 35,600
d 35,600
15,900
35,600
45,800
a 35,600
b 35,600
c 35,600
d 35,600
690
1,550
1,990
1,550
1,550
2,203
1,064
690
1,550
1,990
1,550
1,550
2,203
1,064
690
1,550
1,990
1,550
1,550
2,203
1,064
10i02
6.37
5.74
8.09
5.58
6.53
6.37
3.45
1.99
1.82
2.12
1.84
2.12
1.98
13.47
8.36
7.56
10.21
7.41
8.65
8.35
2.28
1.14
1.02
1.14
1.14
1.35
1.05
1.12
0.63
0.65
0.63
0.63
0.89
0.54
3.40
1.77
1.67
1.77
1.77
2.24
1.60
2.47
1.25
1.12
1.25
1.25
1.47
1.16
1.26
0.70
0.70
0.70
0.70
0.95
0.60
3.73
1.95
1.82
1.95
1.95
2.42
1.77
4.33
2.72
2.57
3.13
2.60
3.13
2.41
2.76
1.87
1.79
1.95
1.87
2.33
1.51
7.09
4.59
4.36
5.08
4.47
5.46
3.92
          a Increased pollutant load
          b Decreased pollutant load
          c Increased hydraulic load
          d Decreased hydraulic load
                                   535

-------
15.7  BASIS FOR REGULATIONS


15.7.1  Evaluation of BPT Treatment Practices

     EPA   is   proposing   BPT  limitations  based  on  Level  1
treatment.  All  plants  in this subcategory  have installed BPT
technology.  Pollutants limited by the proposed BPT regulations
are  TSS,  fluoride,  chromium,  nickel,   and pH.    The  major
pollutants previously regulated are TSS,  fluoride, and aluminum.
Aluminum is no longer considered  a pollutant of  concern due to
its relatively nontoxic  nature.   The treatment proposed as the
basis  for  BPT  regulations  will   actually  benefit  from  the
presence  of   aluminum  which  will  precipitate  under  mildly
alkaline conditions and act as  a coagulant to aid the removal of
toxic metals and suspended solids.

15.7.2  BPT Effluent Limitations

Technology Basis

     The   Agency  proposes  BPT   limitations   for   which  the
technology  basis is, or  is  equivalent  to,  equalization,  lime
neutralization/alkaline   precipitation,    solids   removal   by
settling or  thickening,  final pH  adjustment,  and  discharge of
the  clarified effluent.   This  technology  represents  current
practice  in the Aluminum  Fluoride industry and  was  therefore
selected as the  basis for  the proposed BPT effluent limitations.

Flow Basis

     The  basis of  flow  for  BPT  limitations  is  estimated from
data provided  in the 308 questionnaires  for three  of  the four
complete plant responses  received, including Plant 1837, f251,
and |705.  Plant f!88 was omitted in view of the batch process
utilized  for  the manufacture of  aluminum fluoride.   The other
three plants are continuous manufacturing processes.

     The two major  raw process waste water sources contributing
to the total plant flow  estimates  include scrubber and work area
washdown.   These waste  water  sources are  summarized  in Table
15-4 for the three  plants considered.  The model plant flow for
the A1F3  industry  is estimated as  the  average  total  raw waste
water  flow  for  the  three  plants,  and  is  used   to  estimate
pollutant  discharge loadings  for  the  purpose  of  regulation.
Exact measures of treated effluent  from the aluminum fluoride
industry  are  not  available,  since  aluminum fluoride  plants
normally integrate process waste streams  with those generated by
the hydrofluoric acid process prior  to treatment and discharge.
The unit  flow  rates varied widely for the  three plants  in the
range between  4.58  to 19,7 m3/kkg  of  product  which is largely


                              536

-------
dependant on  the scrubber  design  and water utilization.   The
A1F3  process  in  Plant §251  shares  an  S02  scrubber  with the
anhydrous hydrofluoric acid process.   Waste  water generation
from  this combined  use scrubber was  estimated on the basis of
hydrofluoric acid utilization in the two processes.

  The cleaning  and  work area washdown  flow  is similar for the
three plants considered, ranging  between 1.02 and  2.39 m3/kkg of
product.

     The average total  flow for the three plants  is 11.9 m3/kkg
of  product.   This  flow  is used  for the model  plant  in the
aluminum fluoride subcategory.

Selection of pollutants to be Regulated

     The  selection  of  pollutants  for  which  specific numerical
effluent limitations are proposed was based on an evaluation of
raw  waste data  from  the  screening  and  verification sampling
program.  Pollutant  data from the plant  sampled during screening
was  used to determine the  need  for  verification  sampling.
Verification sampling of Plant f705 and  |251 provided additional
pollutant raw waste  concentration data  needed  to  assess the
magnitude of the pollution potential.

     For conventional pollutants, the Agency  has selected pH and
total  suspended  solids for  specific  treatment   and  control.
Fluoride  was selected as  the  only  nonconventional  pollutant
parameter because it  is a  major  constituent  in the process raw
waste and is a pollutant of concern to the Agency.  A limitation
on  aluminum  is  not   proposed  because this  constituent  of the
process  wastes  will  be  effectively  controlled   by  treatment
required for removal of toxic metals.

     Results  of  the  screening  and  verification  sampling are
tabulated in  Section 15.3.3 for the  raw process  waste stream.
The  pollutant concentration  listed  under verification  is the
highest" value  observed  during  sampling at   the  two  plants
visited.  Toxic  pollutants are listed based on their presence,
during sampling, at  detectable concentration  levels.  Pollutants
from  this list  were  considered  candidates  for  regulation if
their concentrations appeared at  least once at  approximately the
lowest   level   estimated   as  treatable  using   any  available
technology appropriate for  their removal.  The only two metals
which passed  this test were chromium and nickel.   The only two
metals  which  passed this test were  chromium  and  nickel.   The
metals  arsenic,  copper, selenium,  and zinc were  never observed
in  the  raw waste at concentrations equal  to or above the lowest
level  estimated as  treatable as  presented  in Table  8-11 and
therefore are not proposed for regulation.
                               537

-------
     Specific   numerical  effluent   loading   limitations  are
proposed  for   chromium   and   nickel   for  which  the  average
concentration levels  (Table  15-8)  are considered treatable for
at least one plant visited during sampling.

     No limitation is being proposed for aluminum because  of its
relatively low  toxicity  and  its  beneficial effects in removing
toxic metals  by coprecipitation.   In addition,  control of the
major toxic metal  ions should provide  adequate  control of the
aluminum concentration, since the treatment pH for BPT  is  in the
region  considered  optimal for  alkaline precipitation  of most
metal hydroxides.

Basis of Pollutant Limitations

     Conventional and nonconventional parameters -

     A. pH:   The treated  effluent is to  be  controlled within the
range  of  6.0 to  9.0.    This  limitation  is  based on  the data
presented in Appendix B of this report and the JRB Study  (52).

     B. TSS  and Fluoride:   Pollutant limitations for  TSS and
fluoride  were  based  on the   evaluation  of  data  for  the
hydrofluoric acid subcategory.  This  evaluation  is described in
Section 12.7.2  under  "Basis  of Pollutant  Limitations."   There
are  no  plants   where  the  BPT  treatment  performance can  be
evaluated for  the treatment of  raw  aluminum  fluoride process
waste  water   alone.    Aluminum  fluoride  plants  integrate raw
process  waste  water   with  waste   waters  generated  from  the
hydrofluoric acid process.

     In view  of the  similar  waste water  characteristics, the
effluent concentration from a common  treatment system would be
the  same  for  TSS  and  fluoride whether it originates  from the
A1F3 industry or the HF  industry.   Therefore,  a maximum 30-day
average concentration  for the A1F3 industry  of  97 mg/1 and 53
mg/1 from the HF subcategory (Table 12-24) are proposed for TSS
and  fluoride, respectively.   These are  relatively high values
that are unique to this industry.  The variability factor ratio
of  2.1  was  selected  based  on   the  evaluation  in  the  HF
subcategory (Table 12-23). The unit effluent load limitation is
determined as follows:

     L  (as kg/kkg)  =  (Q) (C)
                        1000

     Where C  is  the  maximum  30-day  average  concentration  in
mg/1, Q is the  unit flow in m3/kkg, and 1000 is the conversion
factor for kg to grams.   (Note:  kg/m3 = 1000 mg/1.)
                              538

-------
     The  24-hour  maximum   is   determined  by  the  following
relationship:

         Maximum 30-day average X VFR  =  24-hour maximum
          (concentration or unit           (concentration or
           loading)                           unit loading)

     In this case, the daily maximum TSS concentration is 2.1 X
97 mg/1 = 200 mg/1.   The unit loading is then

     97 mg/1 (11.9 m3/kkg)/   kg/m3   \  =  1.2 kg/kkg
                          V 1000 mg/1 )

     In the same manner  the concentration basis for fluorides is
2.1 X 53 mg/1 = 110  mg/1.  The unit loading is then

     53 mg/1 (11.9 m3/kkg)/   kg/m3   \  =  0.63 kg/kkg
                          V 1000 mg/1 )

     The  24-hour  maximum   unit  loading  is   determined  by
multiplying 2.1 times the 30-day average unit loading determined
above.

     Toxic pollutants  - The effluent limitations  proposed for
the selected toxic pollutant  control parameters are derived from
three  sources   of information  including  1)  literature  based
treatability   estimates  (Section   8.1),   2)   screening  and
verification sampling  data,  3)  a limited amount  of long-term
monitoring data from Plant #251.

     The sampling results represent raw process waste pollutants
observed during three days of composite sampling at each of the
plants verified.  An assessment of treatment system performance
was not possible in view of the lack of representative effluent
data  available in  the  subcategory.    Effluent data  obtained
during verification  sampling  is  for  treated waste water from the
HF  and A1F3 processes  combined,  since  no plant  is available
which  treats A1F3 wastes alone.  Therefore,  the screening and
verification  data may  be  used  to  determine candidate  toxic
pollutants   for  regulation   without  specifying   achievable
concentration limits which represent the A1F3 plant performance
alone.   However,  review of  the  combined HF  and A1F3  waste
effluent data in Table 15-9  reveals that  all toxic pollutants of-
concern are treatable within the levels of treatability defined
in  Section  8.1 for  lime settling  (BPT) .   Removal of  toxic
pollutants from one  waste water  or the other would not differ in
light  of  the similar  nature  of  HF and A1F3 wastes.  Therefore,
the literature  estimates of treatability  discussed  in  Section
8.1  have  been used as  'the  basis  for   determining  specific
numerical limitations for toxic pollutants.
                              539

-------
     A.   Chromium;  The  literature treatability  value  of 0.1
mg/1  from  Table  8-11  for  lime  settling  is  considered  to
represent  a maximum  30-day  average  concentration value  for
chromium  in view  of  plant  performance  data in  the  HF  and
combined  HP/A1F3  industries.   The unit  load limitation  was
calculated as follows:

     (0.10 mg/1)  (11.9 m3/kkg)/   kg/m3   N =  0.0012 kg/kkg
                              \1000 mg/1  )

Since long-term monitoring data on chromium  is  not available,
the variability factor  ratio (VFR)  of 2.0  was selected  on the
basis  of lead  monitoring  data  from  Plant f251  presented  in
Tables A-lOa and A-lOc.  This is justified by  the similarity in
the chemistry  of  lead, nickel,  chromium, and other  metals  of
concern under BPT  treatment conditions.  Therefore,

     VFR = VF of daily measurements -  3L12
            VF  of  30-day averages      1.55

     VFR =2.0

The daily maximum limitation for  chromium was  determined  as
follows;

     (2.0)(0.0012  kg/kkg)  = 0.0024  kg/kkg

The proposed effluent limitations on  chromium are presented in
Table 15-19 for BPT treatment.

     B.   Nickel:   The raw  waste concentration of nickel was
observed as high as 0.29 mg/1 (Section 15.3.3, Table of Maximum
Concentrations  Observed)  to an average  value of  0.22  mg/1  at
Plant 1705  (Table  15-8).   The literature treatability value of
0.20 mg/1 from Table  8-11 for lime  settling is  used  for the
purpose of  regulation  in  view of  the absence of  actual plant
performance data.  The limitation is determined as  follows:

     (0.20 mg/1)  (11.9 m3/kkg)/   kg/m3   \ «  0.0024 kg/kkg
                              V 1000 mg/1 /

Therefore, the  24-hour maximum load limitation is:

     (2.0)(0.0024  kg/kkg)   =  0.0048 kg/kkg

where 2.0 is the VFR as discussed for chromium.

     C.   Other  metals:   The concentration bases  for arsenic,
copper, selenium,  and  zinc are  also presented in  Table 15-19.
These pollutants are listed to serve as guidance in cases where
these pollutants are found to be of water quality concern.  The


                              540

-------
               TABLE 15-19.  PROPOSED LIMITATIONS
                        Aluminum Fluoride
     Best Practicable Control Technology Currently Available
                  Waste Water Plow: 11.9
 Pollutant
                                                   Effluent Limit
                                                        (kg/kkg)
             Concentration Basis
Subcategoty     (1)    (mg/ll
Performance   WR	
 (mg/1)              30-day  24-hr   30-day  24-hr
                      Avg     Max     Avg     Max
Conventional and
Nonconventional Pollutants:
Total Suspended   97
Solids, TSS

Fluoride          53

Toxic Pollutants:
                     (2)
             2.1
     (2)
             2.1
           (4)
         97


         53
                         200
                         110
                       1.2   2.4


                       0.63  1.3
Arsenic

Chromium

Copper

Nickel

Selenium

Zinc
  0.50
  0.10
  0.50
       i

  0.20
  0.20
 (3)

 (3)
i
 (3)
l
 (3)
I
 (3)
I
 (3)
  0.50
2.0
2.0
2.0
2.0
2.0
       2.!
 (5)

 (5)
i
 (5)
i
 (5)
I
 (5)
i
 (5)
0.50

0.10

0.50

0.20

0.20

0.50
1.0

0.20

1.0

0.40

0.40

0.50
                                                           (65     (6)
0.0012  0.0024
    (6)    (6)
0.0024   0.0048
    (6)    (6)
                                           (6)
                                    (6)
   (1) - VFR:  ratio of the  24 hour variability factor to  the
        30 day variability factor.
   (2) - 30 Day maximum  average concentration based on  the proposed
        HP subcategory  regulation  (Section  12.7,2).
   (3) - The lower limit of the literature treatability estimate
         (Table 8-11) is used as the basis for the maximum 30-day
        average limitation and pubcategory  performance,  since no plant
        is available where BPT treatment can be evaluated for the A1P3
        waste water alone.
   (4) - WR based on HP subcategory evaluation.
   (5) - WR based on limited long term data.
   (6) - No effluent limitation proposed.
                                541

-------
concentration   limitations  are   also  based   on  literature
treatability levels presented  in  Table 8-11.   However in every
case   these   treatability  levels    were   above   raw   waste
concentrations observed for each of these metals.

15.7.3  Basis for Proposed BCT  Effluent Limitations

     The BCT limitation  {applicable only  to TSS)  was set equal
to  BPT because  the addition  of  more  treatment  technology to
remove  conventional pollutants  failed to  pass  the BCT  cost
comparison  test  (44 FR  44501  July  30, 1979)  as described in
Section 3.3.3.

15.7.4  Basis for Proposed BAT  Effluent Limitations

The Application of Advanced Level Treatment:

     The Agency has analyzed the cost  effectiveness of the base
level systems  (BPT) and  the various  advanced level options for
conventional, nonconventional  and toxic pollutant  removal based
on utilizing the  cost  estimates presented  in this report.   The
economic  impacts  on the  aluminum fluoride  industry have  been
evaluated  in  detail  and  taken  into  consideration  in  the
selection  of   the   technology  basis  for  the  proposed  BAT
regulations.

     For  BAT,  EPA  is  proposing  limitations  based  on Level  2
treatment.  This  treatment  option adds dual media  filtration to
remove  additional toxic metals and  fluorides.   This  level of
treatment removes 300 pounds per year  of toxic metals and 62,000
pounds  per  year of fluorides.  Pollutants  limited in proposed
BAT regulations are fluoride,  chromium, and nickel.

     EPA considered limitations  based on  Level 3 and 4 sulfide
precipitation  and  use  of soda  ash  to  increase  recycle,
respectively.  These options  were rejected because they remove
only  small  incremental  amounts  of  toxic  pollutants in  this
subcategory.

Technology Basis

     For  BAT,  the  Agency proposes   more  stringent  effluent
limitations  on  fluoride  and  the  toxic  pollutants based  on
addition of dual  media filtration or  its  equivalent to the BPT
treatment system  (Section 15.7.2).

Flow Basis

     The same flow established for BPT in Section  15.7.2 is used
in  the  development  of  the  BAT effluent limitation.   The  flow
used is 11.9 m3/kkg of product (Table  15-4).
                              542

-------
Selection of Pollutants to be Regulated

     The Agency  has selected  fluoride and the  same  two toxic
pollutants  identified  in the proposed BPT  regulations for the
BAT regulations.   The rationale for  their selection is discussed
in Section 15.7.2.

Basis of Pollutant Limitations

     Nonconventional  pollutants  -  The  only  nonconventional
pollutantselected is fluoride. The limitation proposed for BAT
is based on the  evaluation of plant performance data discussed
in Section  12.7.4  for  the HF  subcategory.   A  maximum 30-day
average  concentration  of  30  mg/1   for  total  fluoride  was
identified in the evaluation and is  used here  in Table 15-20 for
establishing   a   numerical  limitation.    Selection  of  the
concentration  is  based  on  the similarity  between  the  waste
stream from the HF  and A1F3 subcategories.

     The 24-hour maximum concentration is determined as follows
from the WR and maximum 30-day average concentration;

     (2.1)(30 mg/1)  =  63 mg/1

     The  effluent  limitation  for   fluoride  is  determined  as
follows:

     (30 mg/1) (11.9 m3/kkg)/   kg/m3   \ =   0.36 kg/kkg
                            V 1000 mg/1 )

     The 24-hour  maximum limitation is determined in a similar
manner as follows:

     (63 mg/1) (11.9 m3/kkg)/   kg/m3   N »   0.75 kg"/kkg
                            V  1000 mg/1 )


     The variability factor  of 2.1  used for  the development of
BAT limitations is  the same used for BPT in Section 15.7.2.

     The  estimated  performance  of  Level  3   and   Level  4
alternative  technologies  are presented  in  Tables   15-21  and
15-22, respectively.   The tables present the estimated maximum
30-day  average  and  24-hour  maximum  concentrations  for  the
purpose of comparison with the proposed regulations.

     Toxic pollutants - Addition of  dual media filtration to the
BPT  level   of treatment provides  additional  removal of  the
suspended  metal  hydroxides.   • Therefore,  BAT  provides  more
stringent control of the toxic pollutants.   Since  there is no
directly applicable data  on  filter performance for  the A1F3
                              543

-------
                     TABLE 15-20.  PROPOSED LIMITATIONS
                              Aluminum Fluoride
                          Best Available Techrology
                       Waste T/feter Flow:  11.9 m3/kj
           0.18
           0.47
(5)

(5)
          2.1   30
2.0

2.0

2.0

2.0

2.0

2.0
0.50

0.04

0.29

0.17

0.18

0.47
63




 1.0

 0.08

 0.58

 0.34

 0.36

 0.94
                          0.36
0.00048

  _J4)


0.0020

  __(4)

    (4)
                               0.75
0.00096

  _J4)

0.0040

  _J4>

    (4)
 (1) - WR:  ratio of the 24-hour variability factor to the 30-day variability
      factor.
 (2) - Also applicable for pretreatment standards for existing sources  (PSES)
      which are expressed as concentration.
 (3) - 30-day average calculated from the HP subcategory Table 12-21 and
      12-25.
 (4) - No effluent limitation proposed.
 (5) - literature treatability estimates.
                                      544

-------
            TABLE 15-21.  PEEM5KMANCE OF ALTERNATIVE TECHNOLOGY
                             Alxminum Fluoride
                           Level of Treatment: 3
                       feste Water Flow: 11/9 m3/kkg


Pollutant

Nbnoonventional
Pollutants:
Fluoride
Toxic
Pollutants:
Arsenic
Chromium
Copper
Nickel
Selenium
Zinc

Treatability
(mg/1)

25
0.050
0.040
0.050
0.10
0.18
0.20
Concentration Basi§
(1) (rag/1)
™* Max 24-hr
30-day Max
\
3.0 25 75
2.0 0.050 0.10
2.0 0.040 0.080
2.0 0.050 0.10
2.0 0.10 0.20
2.0 0.18 0.36
2.0 0.20 0.40
(1)  - VFR: ratio  of the 24-hour variability factor to the 30-day variability
     factor.
                                     545

-------
             T&BuE 15-22.              OP SLTEHmTIVE TECHNOLOGY
                             Aluminum Fluoride
                           Level of Treatment: 4          /,.,,
                K&ste Water Plow: 2.4 mVktag  (80% Recycle) u'

Ooncmtration Basis
Bollutant
Nonconventional
Pollutants:
Fluoride
Toxic
Pollutants:
arsenic
Chrcmiim
Copper
Nickel
Selenium
Zinc
Treatability
(rag/1)
30
0.50
0.04
0.29
0.17
0.18
0.47
(1 ) (mg/1)
WR^ '
Max
30-day
Avg
2.1 30
2.0 0.50
2.0 0.04
2.0 0.29
2.0 0.17
2.0 0.18
2.0 0.47

24 -hr
Max
63
1.0
0.08
0.58
0.34
0.36
0.94
(1)  - WR:  ratio of the 24-hour variability factor to the 30-day variability
     factor.
(2)  - The effluent flow rate is 20 percent of the average influent or basis
     of flow (i.e.,  0.20 x  11.9 m3/kkg = 2.4 m3/kkg).
                                     546

-------
industry, the literature  treatability  studies  cited in Section
12.7.4 under "Toxic Pollutants" for the HP industry are used in
the following determinations.  The estimated percentage removal
efficiency, presented in the aforementioned section, was applied
to  the 30-day  average concentrations  developed  for the  BPT
limitations to establish the proposed BAT  regulation.  The basis
for the BAT limitation on each toxic metal is given below.

     A.  Chromium:  Filtration of the BPT effluent is estimated
to reduce the chromium concentration by 60 percent.  Therefore,
the maximum 30-day  average  concentration  would be 0.04 mg/1 by
the following relationship:

     BPT 30-day average /100% - 60%\ =  BAT 30-day average
         concentration  \   100%   /        concentration

     The limitation is determined numerically as follows:

    /100 - 60\  0.10 mg/1  =  0.040 mg/1
    \  100   /

     Application of the BAT model plant discharge rate results
in the proposed chromium limitation as follows:

      (0.040 mg/1) (11.9 m3/kkg)/   kg/m3   \ =  0.00048 kg/kkg
                               V 1000 mg/1 /

and,  the  daily  maximum limitation using  the VPR  value  of  2.0
becomes

      (2.0) (0.00048 kg/kkg)  =  0.00096 kg/kkg

The VFR  value of 2.0  used  for BPT  is similarly  used  for  BAT
because the variability of  the filtrate quality is anticipated
to be no greater than  the  observed variability of the unfiltered
effluent.  The  variability  factor was  observed from long-term
data  at  Plant  f251   (Tables  A-lOa and  A-lOc).   Treatability
studies are being conducted by the EPA that assess the proposed
BAT level of treatment.

     B.  Nickel:  Filtration of the BPT  effluent is estimated to
reduce the nickel concentration by 14  percent.   Therefore,  the
maximum 30-day  average  concentration would be  0.17 mg/1 by the
following calculation:
    /100 - 14\ (0.20 mg/1)  =  0.17 mg/1
    V  100   )
Application  of the BAT  model plant  discharge rate  gives the
following load limitation for nickel:
                              547

-------
      (0.17 mg/1)  (11.9 m3/kkg)/_kg/in3j\ =  0.0020 kg/kkg
                              V. 1000 mg/1 )

consequently the  24-hour maximum value is

      (2.0)(0.0020 kg/kkg)  =  0.0040 kg/kkg

as presented in Table 15-20.

     C.   Other metals:   The concentration basis  for arsenic,
copper,  selenium,  and   zinc  are  also  given  in  Table  15-20
assuming   0,   42,  14,   and  6   percent  removal  efficiency,
respectively, by  the  addition of  filtration to the BPT system.
The values  presented  in Table 15-20 for these toxic pollutants
are intended for  use  in  cases where they are  of concern from a
water quality standpoint.  However in all cases the treatability
level was above the raw waste concentration levels observed.

15.7.5  Basis for Proposed New Source Performance Standards

Technology Basis

     For NSPS, the Agency proposes the  same treatment technology
that is proposed  for BAT.

Flow Basis

     The  same  flow established  for BPT  and BAT  is  used in the
development of the NSPS effluent limitations.

Selection of Pollutants to be Regulated

     The  Agency  has  selected TSS, fluoride  and the  same  two
toxic  pollutants  identified for  the  BAT regulations.    The
rationale for their selection is discussed in Section 15.7.2.


Basis of Pollutant Limitations

     Conventional pollutants -

     A.  pH:  For NSPS,  the BPT limitation is  retained.  Control
of the final effluent within the  range  of pH between 6.0 and 9.0
is required based on data presented in  Appendix B of this report
and the JKB Study  (52) .

     B. TSS:   In view of the absence of applicable performance
data concerning TSS, a value of 68 mg/1 was assumed from the HF
subcategory for  the  maximum  30-day average concentration.   The
value was  developed  by  assuming  a 30  percent  reduction  in TSS
over  the  30-day  average concentration  estimated  for BPT  (97
                              548

-------
mg/1).   The  assumption  is based  on pilot  scale  studies  (41)
which  have demonstrated  an average  removal by  filtration of
approximately 30 percent  from  waste water containing suspended
metal hydroxides after lime treatment.

     A  VFR of  2.1  is used  on  the basis  of  long-term data
presented  in  Table  12-21  of the HF  subcategory.   The proposed
30-day average limitation on TSS is determined as follows:

      (11.9 m3/kkg)/^  kg/m3  \ -  0.81 kg/kkg
                  V 1000 mg/1 /

     The  24-hour  maximum  consequently becomes  kg/kkg)  =  1.7
kg/kkg.

     The proposed NSPS limitations are presented in Table 15-23,

     Nonconventional   pollutants   -   Fluoride  is   the  only
nonconventional pollutant and is set equal to the BAT limitation
of 30 mg/1 for NSPS.

     Toxic pollutants  - Waste  water sources are expected to be
the  same  as currently identified  for new  source  A1F3 plants.
Therefore,  the  proposed  toxic  pollutant limitations  for NSPS
have  been set  equal  to  the  proposed BAT   limitations  by  the
Agency.   BAT  limitations  for  the toxic pollutants is discussed
previously in Section 15.7.4.

15.7.6  Basis for Proposed Pretreatment Standards

Existing Sources

     Pretreatment  Standards for  Existing  Sources  (PSES)  are
proposed by the  Agency to  equal  BAT  limitations.  The pollutants
to be limited are fluoride, chromium, and nickel.

New Sources

     Pretreatment Standards for New Sources  (PSNS) are proposed
by the Agency  to equal proposed  BAT  limitations.  The pollutants
to be regulated include fluoride, chromium,  and nickel.
                              549

-------
                           15-23.  PKOPSOSED
                              Aluminum Fluoride
                      New Source Performance Standards
                        Pfeste W&ter Flow: 11.9 m3/kkg
Pollutant
Treatability
     Cmg/1}
Concentration Basis
    (1)     (mg/1)
WR
                                                            Effluent Limit
                                                                (kg/kkg)
       Max
      30-day
       Avg
                                                24-hr
                                                 Max
                    Max
                  30-day
                   Avg
                                                       24-hr
                                                        Max
ODnventional and
Total Suspended
SolMs, TSS

Fluoride, F  (5)

Itoxic
lolltatants;

Arsenic

Chromium  (5)

Copper

Nickel  (5)

Selenium

Zinc
                    68
                    30
        (2)

        (2)
                     0.50(3>

                          (3)
                     0.04
                     0.29
                     0.17
                     0.18
                     0.47
           (3)

           (3)

           (3)

           (3)
2.1   68

2.1   30
2.0

2.0

2.0

2.0

2.0

2.0
0.50

0.04

0.29

0.17

0.18

0.47
       140

        63
1.0

0.08

0.58

0.34

0.36

0.94
          0.81

          0.36
0.00050

  _J4>

'070020

  — (4)

     (4)
            1.7

            0.75
0.0010

  _J4)

0.0040

  _J4>

    (4)
 (1) - WSs ratio of the 24-hour variability factor to the 30-day variability
      factor.
 (2) - 30-day average calculated from the HF subcategory Table 12-29.
 (3) - Literature treatability estimates from BM? level  treatment.
 (4) - No effluent limitation proposed.
 (5) - Also applicable for treatment standards for new sources (PSNS) which
      are expressed in concentration.
                                     550

-------
                          SECTION 16


                   CHROME PIGMENTS INDUSTRY
16.1  INDUSTRY PROFILE
16.1.1  General Description

     Chrome  pigments  are  a  family  of  inorganic  compounds
primarily used  as  colorants  in a number of  industries.   These
pigments are used  in paints,  ceramics, floorcovering products,
ink, paper,  and cements.  However,  certain  chromium compounds
(i.e., oxides)  may be used as  raw materials in the manufacture
of   certain  metals   and   alloys.     Chrome   pigments   vary
substantially  in  their  chemical  makeup.    The  various  types
include chrome  yellow, chrome  orange,  molybdate chrome orange,
anhydrous  and   hydrous  chromium  oxide and  zinc yellow.   The
industry data profile  is given in Table 16-1 and the status of
the regulations are shown  in Table 16-2.

16.1.2  General Process Description and Raw Materials

     The general  manufacturing process  for  each of the above
compounds is given below.

Chromium Oxide

     This  pigment  consists  of   two  compounds;  anhydrous  and
hydrated  chrome oxide   (Guigets  Green) ,    The*  amount of  the
anhydrous  salt  oxide produced is approximately  ten times  the
amount of hydrated  chromic oxide  produced.   It is offered in a
narrow  range of  shades from  light  yellowish to  dark  bluish
green.

     Anhydrous  oxide  is   almost  pure chromium oxide and  the
commercial grade consists  of  a minimum of 98.5 percent  Cr203.
It  is  prepared  by  calcination  of sodium dichromate with sulfur
or carbon according to the reactions given below:

     Na2Cr207 + S = Cr203 + Na2SO4                          (1)
                              551

-------
             .TSBEE 16-1.  SOBCKEBOOIg HSgHE QKOR.
     SUBCAXBGOIOT
                         CHBOME PIQ4EH1S
     Itotal subcategory capacity rate
     Total subcategory production rate
     Number of plants in this subcategory
     308 Data on file for
        With total capacity of
        With total production of
        Representing capacity
        Representing production
        Plant production ranges (2)
                 Miniimm
                 MaxJsuni
        Average production
      .  Median production
        Average capacity utilization
        Plant age range:
                 Minimum
                 Mscdmsa
        Wastewater flow range:
                Maxiinira
        Volurre per unit product:
                Minimum
                Maxiirum
63,000 kkg/year
64,500 kfcg/year
    12
     5

39,800 kkg/year

    62 percent

   100 kkg/year
18,000 kkg/year
 6,300 kkg/year
 6,400 kkg/year
    78 percent

    38 years
    60 years

   800 cubic meters/day
11,363 cubic meters/day

    32 cubic meters/kkg
   170 cubic meters/kkg
(1)   Sources of data are Stanford Research Institute, Directory of Chemical
     Brodacers, U.S.A.,  1S77/ U.S. Department of Oonnerce, Current Industrial
     Reports,  Decanfaer 1977; Energy and Environmental Analysis, Inc.? Draft
     Report,  "Preliminary Economic Assessment; of Effluent Limitations in the
     Inorganic Chanical  IMustry," June, 1978 and "Economic Analysis of Proposed
    Revised Effluent Guidelines and Standards for the Inorganic Chemicals Industry,"
    March, 1980.
(2)   Based on  production at 11 plants, all other figures are based on 308
     Questionnaires.
                          552

-------
    TKBJM 16-2.  SEXTOS OF HEGCJLAIICNS  -  EEFEOEOT LJMTEaTICN GUIDELINES

SOBCSTEQQH21 Chrome Pigments
SUBPAOT
AH
(40CPR
415.340, 5/22/75)
STANDARDS
Product
Process
Chrome
Pigment







Para-
meters
TSS
Cr(T)
Cr*6
Pb
Zn
CN
CM (A)
Fe
BPC3CA * BKEEA HSPS
Max. Avg, Max. Avg, Msec. Avg.
kg/W^ kg/Kkg kgAkq kg/kkg kg/kkg ko/kka
(mg/1) (itq/1) (mg/D W/L) ftnfA) Tmg/a)
5.1
(76.1)*
0.10
(1.5)
0.010
(0.2)
0.42
(6.3)
0.72
(10.8)
0.010
(1.5)
0.10
(0.2)
Oo72
(10.8)
1.7 Reserved Reserved
(25.4)
0.034
(0.5)
0.0034
(0.1)
0.14
(2.1)
0.27
(4.0)
0.0034
(0.5)
0.034
(0.1)
0.27
(4.0)
Sections 415.340, 415.341, and 415.342 were revoked by the Agency
£41 FR 51601, November 23, 19761.
 Max, = Maximum of any one day.
2
 Avg. * Average of daily values for thirty consecutive days.
* flow basis  67,000 1/kkg.
                                  553

-------
     Na2Cr207 + 2C = Cr203 + Na2C03 + CO                     (2)

The use of  sulfur as the reducing agent  eliminates  C02 and CO
emissions but increases the sulfates  in the raw waste as well as
producing SO2  and SOS  in the off-gases.   In the manufacturing
process using  sulfur,   the  raw materials  consisting of sodium
dichromate  and  sulfur   are mixed  with water  and  the resultant
solution  is fed  to  a   kiln.   The material  is heated  and the
reacted  materials  from  the  kiln  are   slurried  with water,
filtered, washed,  dried,  ground,  screened,  and packaged.   The
effluent  gases from  the  kiln containing  sulfur dioxide and
sulfur  trioxide   are   wet  scrubbed   before  venting  to " the
atmosphere.

     A  general  process  flow  diagram  of  the preparation  of
anhydrous chrome oxide  is given in Figure 16-1.

     Hydrated  chromium oxide,  Cr203.2H20  or  Cr2O(OH)4,  also
known  as  chromium hydrate  and Guigets Green, is a brilliant
bluish green.   It is  made by reacting sodium dichromate with
boric acid  as follows:

     2Na2Cr207 + 8H3B03 = 2Cr203.2H20 + 2Na2B407
                          + 8H20 + 302                       (3)

The raw materials  are  blended  in  a mixer  and then heated in an
oven at about"  550  degrees C.   The reacted material  is  slurried
with water  and  filtered.   The filtered solids  are washed with
water, dried, ground, screened, and packaged.  The filtrate and
the wash water  are treated with sulfuric  acid to recover boric
acid according to the reaction given below:

     Na2B407 + H2SO4 +  5H2O = 4H3BO3 + Na2SO4                (4)

     A  waste  stream  containing  some  boric  acid  and sodium
sulfate is  discharged from the boric acid unit.  Figure 16-2 is
a generalized flow diagram of the process.

Chrome Yellow and Chrome Orange

     Chrome  yellow  is one  of  the   more  important synthetic
pigments.  The  chrome yellows cover the  range of hues from light
greenish yellow  to reddish medium yellow  and consist mainly of
lead chromate.   They  are made by reacting  sodium  dichromate,
caustic soda, and lead  nitrate.  The reactions are given ass

     2HNO3 + PbO = Pb(NO3)2 + H20                            (5)

     Na2Cr207 + 2NaOH + 2Pb(N03)2 = 2PbCr04 + 4NaN03 +  H2O   (6)
                              554

-------
                                     van?
                                     1
U1
01
Sf
SCRUBBER 	 **
UQOTJ 3CH.
i
WRIER
SODIUM DICHBOMATE T
J2
IJHBER """"" 	 """""I
I
«p T®
MTVFTT ^ KT^" *^ ^^!Tr ^ FTTiTRR ^ HRVRR ^ ^^ TMT_r ^^
SULFUR ^


CHROME CKIDE
            Figure 16-1.   General process diagram for production of anhydrous chrome oxide.

-------
         SODIUM

       DICHROMftTE
BORIC ftCID
(Ji
tn
en
                                                                                                      HTOBftllC CHK»E

                                                                                                      C5XIDE TO GRINDINS,
                                                                                                              MID
                                               VRSTE TRUER
                        Figure 16-2.  General process diagram for production of hydrated diromic oxide.

-------
     Lead  chromate  is  formed  as   a  precipitate  during  the
reaction.    It  is  filtered   and  treated  with  chemicals  for
development  of  desired  specific   pigment  properties,  dried,
milled,  and  packaged.    The   filtrate  from  the  filtration
operation  is  sent to the waste  water   treatment facility,   A
flow diagram of the chrome yellow manufacturing process  is shown
in Figure 16-3.

Molybdenum Orange

     Molybdenum  orange  is made  by  the  coprecipitation of lead
chromate  (PbCr04)  and lead molybdate  (PbMo04).   The resulting
pigments are more  brilliant than chrome oranges.

     The  process  consists  of  dissolving  molybdic  oxide  in
aqueous  sodium  hydroxide  and  adding  sodium  chromate.    The
solution is mixed  and reacted  with  a solution-of lead  nitrate.
The precipitate  from the reaction  is  filtered, washed, dried,
milled  and packaged.   The filtrate is  sent  to  the treatment
facility.

     The reaction  is  given  as  follows:

         Mo03 +  2NaOH = Na2Mo04  + H20                        (7)

         PbO + 2HN03  = Pb(NO3)2  + H20                        (8)

         Na2MoQ4 + Pb  (NO3)2 = PbMo04 + 2NaNO3               (9)

         Na2Cr04 + Pb  (N03)2 = PbCr04 + 2NaN03              (10)

         PbMo04  +  PbCr04  = PbCr04.PbMo04                    fll)

     A simplified flow diagram for the manufacture of molybdenum
orange  is  given  in Figure 16-4.

Chrome Green

     Chrome greens are a coprecipitate of chrome yellow and iron
blues.  They  include a wide variety of hues from very  light to
very  dark  green.   Iron blues are  manufactured by  reaction of
aqueous  solution  of iron  sulfate   and  ammonium sulfate  with
sodium  hexacyanoferrate.   The precipitate  formed is separated
and oxidized  with sodium  chlorate  'or  sodium  chromate to form
iron  blues  (Fe  NH4  .Fe CN  6) .   Chrome  green is  produced  by
mechanically  mixing  chrome  yellow   and  iron  blue  pigments  in
water.  The coprecipitate formation  of chrome  green is  given bys

     PbCr04 + Fe(NH4) .Fe(CN)6  =  PbCr04Fe(Nfi4) .Fe(CN) 6       (12)
                              557

-------
(Jl
Ul
00
                      IBU) GXHE
                      HKBBH
                      tcraic ACID
                   SCDIIH HYDHOXIDE
                   SODIUM
                                                                                                     CHRCHE YELiCW
                                                                                                     1DDRONS, MHMNG
                                                                                                     AH) PACKAGING
                                                                                     HRSffi WIER
                          Figure  16-3.   (feneral process diagram for prod.tcrtd.on of dirone yellow.

-------
        VENT
        I
 SODIUM
CHROMATE  WATER
Mjbxuuu; uxjJJis
WATER
CAUSTIC SODA fc


(J\ '
VO


I£AD OXIDE

NITRIC flCID


DISSOLVER





VENT


UifctlULiVJSK

f f
















^^ FILTRATION
WASHING
*
WASTE WATER




                                                                                       DRYING
                                                                                       MULDG
                                                                                       AND
                                                                                       ESCKAGIN3
                                                                                       OF
                                                                                       tOLYBEENtM ORRNGE
                                                                                       (PbCr04.PbMo04)
                                                                                       PRODUCT
Figure 16-4.   General  process  diagram for production of molybdenum orange.

-------
     Figure  16-5   gives   a  process   flow   diagram  for  the
manufacture of chrome green.

Zinc Yellow

     Zinc  yellow,   also  called  zinc chromate,  is  a  greenish
yellow pigment.  It is a complex compound  of zinc,  potassium,
and   chromium   which    has    the    approximate   composition
4ZnO.K20.4Cr03.3H20.  It is made by the reaction of zinc oxide,
hydrochloric acid,  sodium dichromate,  and  potassium chloride.
Zinc yellow is formed as a precipitate and is filtered, washed,
dried, milled, and  packaged  for  sale.   The reactions are given
as:

     2KC1 + 2HC1 + 2Na2Cr207.H20 = K2Cr4013 + 4NaCl
                                   + 3H20      ,            (13)

     4ZnO + K2Cr4013 + 3H20 = 4ZnO.K20.4Cr03.3H20          (14)

     A general  flow  diagram  of  the manufacturing  process  is
given in Figure 16-6.


16.2  WATER USE AND WASTE SOURCE CHARACTERISTICS


16.2.1 Water Use

     In the  chrome  pigments industry, water  is used primarily
for noncontact  cooling, washing the precipitated product, and as
boiler feed  for steam  generation.    In  some cases,  water  is
introduced into the reactor along with the raw materials.

     In addition, substantial quantities of water may be used in
cleaning  equipment.    This occurs  during  product  changes  at
plants manufacturing a  number  of  pigments.   This partially
accounts  for  the increased  unit water  use  at larger plants,
since these plants have the most complex product mix.

     In anhydrous and hydrated chrome  oxide manufacture, water
is used for  slurrying of  the  reaction product and  in scrubbing
the reactor vent gases.  Table 16-3 is a summary of water usage
at different pigment plants in the chrome pigments subcategory.

16.2.2 Waste Sources

     Some   plants    produce    different    pigment    products
sequentially in  the  same equipment.    At  a  few  plants,  the
different pigment products are manufactured concurrently and the
waste waters  combined and treated  at  a  single  facility.   A
generalized flow diagram applicable  to all chrome pigment plants
                              560

-------
                                                               VRTER
                                       IRON BUS
                                                              RESLtMW
Ul
01
LEAD NITRATE ^^. • 	 |
SODIUM CHBDMRTE ^
SODUM SOHB.TE ^
BfflOTW




V 1
PIHER
AM)
1-


SHADE


TRNK



FILTER
I



DRYER

	 •» ORIKDING
BIfflDINQ
MO
PACKING
CHECMB GREEN
PKODUCT
                                          WASTE WATER
                                                                                  WASTE WATER
                           Figure 16-5.  General process diagram for production of chrome green.

-------
              'ZnO
                                                  warn

Ul
o\
to
	 ^
mi fc
RCI te
i 	 -
HBSCfKW 18NK



FH.TRATICN
ASHING


DR-mc
pHUUHG, PACKW3DK

 OFfflE ZQC MliOW

 (K,0'4ZnD-4ClO,'3H.O)

  2        J  4
                                                HASTE VftlER
                         Figure 16-6.- General process  diagram for production of  zinc yellow.

-------
      16-3.               IN THE CHROME          SUBCAGEGOKy
                                                           (1)
         USE
                                     #464
UNIT FLCW (m3/kkg)
Plant Designation
     #436          #214
Noncontact cooling
Direct process contact
Indirect process contact
Maintenance
Scrubbers
Boiler Feed
Total
. 9.50
18.6
7.18
12.0
3.30
2.52
53.1
6.45
147
m(2)
1.78
9.56^
11.1
176
NA
32.6
NA
0.152
NA
0.152
32.9
(1)   Includes all chrome pigment product mixes.   Values indicated only
     for those plants that reported complete information.
(2)   NA - Not applicable.
(3)   Iron blue pigment process.
                                   563

-------
is given in Figure 16-7.   The waste water  sources  are similar
for all pigment products except that at chrome oxide plants, an
additional scrubber waste  is  generated.  Table  16-4  gives the
waste water flow data Nummary for several plants.  The quantity
of waste water and  the pollutants vary for the different pigment
products since the pollutants  are  dependent  on the raw materials
used.  The  figures  in  Table  16-4   represent   actual • plant
discharges.

     The data  sources for  the plants  used in the determination
of unit flow values presented in Table 16-4 are outlined below:

     Plant f464.   Data based on  308  questionnaire submission.
Only chrome pigment production and flows were included.

     Plant §214,   Data based on  308  questionnaire submission.
Chrome pigment and iron blue production and flows were included,

     Plant f436.   Data based on  308  questionnaire submission.
Chrome pigment production and flows were included.

     Plant f002.  Data based on three days of sampling.  Chrome
pigment  and  organic pigment  (20%)  productions  and  flows  were
included.

     Plant |894.  Data based on three days of sampling.  Chrome
pigment, iron  blue,  and organic pigment  (15%)  productions and
flow were included.

     As previously discussed,  various plants make several chrome
pigments sequentially or concurrently.  Thus the unit hydraulic
load going to  the  treatment facility will be an average of all
the waste loads  from the different processes.   The  raw waste
from a  complex plant may  contain nearly all of  the following
substances:  sodium acetate,  sodium  chloride, sodium nitrate,
sodium sulfate, potassium chloride; lead, iron,  and zinc salts;
soluble chromium and pigment particulates.


16.3  DESCRIPTION OF PIA8TS


16.3,1  Screening

     Plant f894  was  visited during the  screening  phase of the
program.  The samples for this plant were analyzed for.all toxic
and conventional pollutants.

     This plant  produces over  100 products  including organic
pigments such  as copper phthalocyanine,  and all  the wastes are
combined and,treated together.  Treatment consists of chromium
                              564

-------
o\
Ul
WATER WASH WATER ^^ ^j, ^^
1 * T . -
RAW MATERIALS ^
	 • 	 ^

REACTOR




FILTER




DRYER




MILLING MB
SCREENING

	 *» PIGMENT
PRODUCTS
TO '
I 41 1 PACKAGING
WASTE WATER NON-CONTACT PAKnOTATE
(BY-PRODUCT SALTS, STEAM WRCTO
                                            UNREflCTED MATERIALS,

                                            ETC.)
                  Figure 16-7.   General process  diagram for production of chrome pigment complexes.

-------
           TSBIE 16-4.  SUMYRKY OF W&SIE WMER FI£X?
                       SUBCMEGOKT:         PIGMENTS
Plant Designation
#464
#214
#436
#002
#894
Waste Water Flow^
(m3/kkg)
41.1
32.8
149
78.4(2)
170 <2>
         Weighted Average Flow          ,            105^ '
(1)   mcludes waste water from all pigment product mixes.

(2)   Includes organic pigments.

(3)   Weighted on the basis of production since unit waste flow is
     directly related to   plant production:

                 Weighted average =  21 [(unit flow)(production)]
                                       2 (production)

                 i.e.  =
                 Where Q = Unit flow and P = production (which is considered
                 confidential information) .
                                   566

-------
VI  reduction,  equalization  and  neutralization,  followed  by
clarification and filtration.  Sulfur dioxide is added to reduce
the hexavalent chromium to the trivalent state at a low pH prior
to hydroxide precipitation.  The backwash from  the sand filters
is recycled to the equalization tank, while the  sludge from the
clarifiers is passed through filter presses  and then  hauled to a
landfill.   The landfill  has a  bottom  consisting  of  two  clay
layers  sandwiching  a gravel  layer to  allow  for collection of
leachate drainage.  Any water from the  sludge is trapped in the
gravel  layer,  and is pumped out and  returned to the plant for
retreatment.

16.3.2  Verification

     Two plants  were visited during  the  verification  phase of
the program.  The first plant, f002, has a rather large product
mix.   However, one  of  the larger continuous units  can  have a
major impact on the  raw waste characteristics.   This unit either
produces lead  chromate or zinc  chromate.   During  the sampling
period,  zinc  chromate was  being  produced.    All process  waste
waters  are treated continuously.  First, the wastes are treated
in  an  S02  reactor   to  convert  .hexavalent  chromium  to  the
trivalent  state.  The pH is then  adjusted  to 8.5  and  then the
waste is passed through precoated filters, followed by discharge
to the  sewer.  Figure 16-8 shows the treatment  flow diagram and
sampling points.  Table 16-5 shows  the waste flows and pollutant
loadings.   At  sample point 
-------
                        RAW WASTE  S02     ACID
                                #1
                              CHROME  TREATMENT
                                   TANK
                                  pH  3.0
           CAUSTIC
               CAUSTIC ADDITION
                    THROUGH
                    pH 8.5
                LAB FILTERED
                	1
     OUTFALL
     TO SEWER
                                FILTER  FEED
                                   TANK
                                     I
                                      FILTER AID
                                                BACKWASH
                            	J
                          (FILTERS NOT WORKING  SO
                          WERE BEING BYPASSED.
                          THIS WOULD BE THE  FLOW
                          PATTERN IF FILTERS UERF
                          OPERATING.)
                                                           LEGEND

                                                    ^| SAMPLING POINTS.
Figure 16-8.
 General waste water treatment process flow diagram at plant #002
"showing the sampling points.  (Chrome pigment manufacture.)
                                   568

-------
           16-5.  FICW, PODJOTSNT
                OF THE SAMPLED WASTE STREAMS FOR PLBNT # 002
                      SOBCATEGORY:  CHROME PIQfflTS
               Conventional and Nonconventional Pollutants

                                 (mg/1)
(kg/kkg of chrome pigments)
Stream
*
1
2-0
2-F
Stream
Description
Raw Waste
Unfiltered
Treated
Waste
Filtered
Treated
Waste
Flow
(ia3/kkg) TSS
78.4 700
55
78.4 970
76
78.4 NA/ *
Fe
1.6
0.13
2.3
0.18
0.06
0.0047
Cr(¥I)
300
24
120
9.4
m
(1)  NA - Not available
                                   569

-------
solids,  and  toxic metals.   No  organics were  analyzed during
verification.

     Figure  16-9  shows  the treatment  system  flow diagram with
the sampling points indicated.  Table 16-6 gives waste flows and
pollutant loadings.

16.3.3  Toxic Polliitant Concentrations

     The toxic pollutants  found  above  treatable concentrations
in the raw wastes during sampling are given in the table below.
Screening data  was obtained at  Plant  f894.   Verification was
completed at Plants  f894  and  |002.   The only organic pollutant
found in the raw  waste  above  the protocol  detectable limit (10
lig/1) was naphthalene at 14 pg/l.  It should be noted however
that some nitrobenzene  (56 pg/1)  and phthalates at levels up to
220 yg/1 were  found in the treated effluent  and  one raw water
intake.  Since  they were not present  in the  raw  wastes,  it is
presumed they are present as a results of sample contamination?
i.e.,  plasticizer  in  Tygon  Tubing.    No  organic  pollutant
sampling was made during verification.

     Section 5.1.2 of this  report describes  the methodology of
the screening and verification sampling program.  In the chrome
pigments industry, 9  days of sampling  were  conducted at Plants
#894 and 1002.   This involved 5 different  sampling points for
raw and treated waste streams.   The  evaluation of toxic metals
content of these process related waste streams was based on 195
analytical data points.  The screening  at Plant £894 for organic
pollutants generated  another  228 data points.   The  daily raw
waste loads  were calculated from  the  waste  stream  flow  rates
measured or  estimated at  the  time of  sampling and the measured
concentration.

     That is,

         Daily loading  askg of pollutant    =   (C)(Q)
                               day                 1000

     Where:

         C is  the concentration  of the  pollutant expressed in
         units of mg/1  (Note:  kg/m3 = 1000 mg/1), and

         Q is the waste-stream flow  rate expressed  in units of
         mS/day.   (m3,  a cubic  meter,  is  equal  to  264.2  U.S.
         gallons.)

     Similarly,  the  unit   loadings  were  calculated from  the
reported chrome pigments production rate, the waste stream flow
rate, and the measured pollutant concentration.
                              570

-------
               WASTE
WASTE (_
WATER
     (EPA SAMPLE
      POINT ALSO)
en
«j
H
                                                                SLAKED
                                                                 LIME
        EQUALIZATION
            TANK
                                                        1
NEUTRALIZATION
     TANK
  pH 6.2-6.5
NEUTRALIZATION
     TANK
  pH 8.0-8.3
                                  BACKWASH
                                   HOLDING
                                    TANK
                     BACKWASH
         (2)
        SAND
       FILTERS
  -9-