EPA-660/2-75-004 APRIL 1975 Environmental Protection Technology Series Activated Carbon Treatment of Unbleached Kraft Effluent for Reuse I I S32Z LU CD National Environmental Research Center Office of Research and Development U.S. Environmental Protection Agency Corvaliis, Oregon 97330 ------- RESEARCH REPORTING SERIES Research reports of the Office of Research and Development, U.S. Environmental Protection Agency, have been grouped into five series. These five broad categories were established to facilitate further development and application of environmental technology. Elimination of traditional grouping was consciously planned to foster technology transfer and a maximum interface in related fields. The five series are: 1. Environmental Health Effects Research 2. Environmental Protection Technology 3. Ecological Research 4. Environmental Monitoring 5. Socioeconomic Environmental Studies This report has been assigned to the ENVIRONMENTAL PROTECTION TECHNOLOGY STUDIES series. This series describes research performed to develop and demonstrate instrumentation, equipment and methodology to repair or prevent environmental degradation from point and non-point sources of pollution. This work provides the new or improved technology required for the control and treatment of pollution sources to meet environmental quality standards. EPA REVIEW NOTICE This report has been reviewed by the Office of Research and Development, EPA, and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the Environmental Protection Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. ------- EPA-660/2-75-004 APRIL 1975 ACTIVATED CARBON TREATMENT OF UNBLEACHED KRAFT EFFLUENT FOR REUSE by E. W. Lang W. G. Timpe R. L. Miller Grant No. 12040 EJU Program Element 1BB037 ROAP/TASK 21 AZX/027 Project Officers John S. Ruppersberger George R. Webster Pacific Northwest Environmental Research Laboratory National Environmental Research Center Corvallis, Oregon 97330 NATIONAL ENVIRONMENTAL RESEARCH CENTER OFFICE OF RESEARCH AND DEVELOPMENT U.S. ENVIRONMENTAL PROTECTION AGENCY CORVALLIS. OREGON 97330 For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 ------- ABSTRACT A four-year pilot plant program was carried out to investigate the tech- nical and economic feasibility of treating unbleached kraft pulp and paper mill effluent for reuse. Preliminary laboratory studies and cost estimates indicated that the following treatment sequences should be investigated in the pilot plant: 1) primary clarification, carbon adsorption; 2) lime treatment, carbon adsorption; 3) primary clarification, bio-oxidation, carbon adsorption. Water of reusable quality can be provided from unbleached kraft effluent by several combinations of treatment utilizing activated carbon. Un- bleached pulping effluents typically contain about 1000 color units, 250 mg/1 TOC, and 250 mg/1 BOD. Reusable water quality as defined in this study is 100 color units and 100 mg/1 TOC. The most economical treatment is the microlime-carbon process that utilizes low dosages of lime and clarification followed by carbon adsorption in down-flow granular carbon beds. Capital cost for treatment by this process of 9.6 mgd of unbleached kraft effluent from an 800-ton-per-day mill was estimated to be approximately $6.7 millions. Operating costs, inclusive of capital depreciation, were estimated to be $0.30 per 1000 gal and $3.58 per pulp- ton, including credit for the reused water. Carbon adsorption in con- tinuous counter-current stirred contactors was found to have promise of lower operating cost and substantially lower capital costs as compared to adsorption in fixed beds. This report was submitted in fulfillment of Grant #12040 EJU by St. Regis Paper Company under the partial sponsorship of the Environmental Protection Agency. Work was completed in December, 1973. ii ------- CONTENTS Abstract it List of Figures viii List of Tables x Acknowledgments xiii SECTIONS I CONCLUSIONS ' 1 II RECOMMENDATIONS 3 III INTRODUCTION 4 IV LABORATORY STUDIES 6 Introduction 6 Selection of Carbons for Evaluation 9 Results of Carbon and Combined Treatments 9 Final Selection of Commercial Carbon for Pilot Plant 19 Additional Studies 21 V SELECTION OF EFFLUENT TREATMENT SYSTEMS TO BE EVALUATED IN THE PILOT PLANT 28 Introduction 28 Selection of Effluents to be Treated 29 Preliminary Selection of Treatment Sequences and Carbon Adsorption Systems 30 Order of Magnitude Cost Estimates 38 Final Selection of Treatment Sequences and Carbon Adsorption Systems 46 VI PILOT PLANT DESCRIPTION AND OPERATION 48 Objectives 48 Description of Pilot Plant 48 Operating Procedures 57 iii ------- CONTENTS (Continued) VI PILOT PLANT DESCRIPTION AND OPERATION (Continued) Page Sampling and Analytical Procedures 61 Manpower Requirements 63 VII OPERATING RESULTS: PRIMARY CLARIFICATION AND BIO-OXIDATION 64 Bio-Oxidation 64 Primary Clarification 66 VIII OPERATING RESULTS: LIME TREATMENT AND CARBONATION 67 Objectives and Description of Operations .67 Removal of Color and TOC in Lime Treater 68 Sludge from Lime Treater 74 Clarification in the Lime Treater 77 Operation of Carbonator 78 IX OPERATING RESULTS: BIO-CARBON COLUMNS 80 Objectives and Description of Operation 80 Removal of Color and TOC by Carbon Adsorption 81 Dosage of Carbon Required 86 Loadings on Carbon 88 Rates of Removal of Impurities by Carbon 89 Effect of Flow Velocity on Rates of Adsorption 94 Biological Activity in Carbon Columns 94 Backwashing of Carbon Columns 94 Correlation of Variables nc X OPERATING RESULTS: PRIMARY-CARBON COLUMNS gg Objectives and Description of Operation nr Primary Clarification and Filtration ofi Removals of Color and TOC by Carbon Adsorption 97 ------- CONTENTS (Continued) X OPERATING RESULTS: Page PRIMARY-CARBON COLUMNS (Continued) Dosage of Carbon Required IQQ Rates of Adsorption 102 Correlation of Variables 105 Pressure Drops and Backwashing of Columns 106 Displacement of TOC and Color from Columns During Backwashing 107 Non-Adsorptive Removal Mechanisms in Carbon Columns 108 XI OPERATING RESULTS: LIME-CARBON COLUMNS 111 Objectives and Description of Operation 111 Removals of Color and TOC; Effect of Soluble Calcium Concentration 111 Dosage of Carbon Required 118 Rates of Removal of Color and TOC 122 Effect of Operating Variables on Removal of Color and TOC 122 Cumulative Removals by Carbon 124 Removal of Low Molecular Weight Compounds 127 Removal of BOD and Turbidity 127 Removal of Metal Ions 128 Pressure Drop and Backwashing of Columns 129 XII OPERATING RESULTS: LIME-FACET ADSORPTION 130 Objectives and Description of Operation 130 Results from FACET Operation with Lime-Treated Water 130 Removals of Color and TOC 132 Adsorption Performance by Lime-FACET Compared to That by Lime-Carbon Columns 133 Attrition of Carbon in FACET Operation 135 v ------- CONTENTS (Continued) Page XIII SUPPORTING LABORATORY STUDIES 137 Color Increase in Presence of Activated Carbon 137 Adsorption Isotherms 141 Dynamic vs. Equilibrium Concentrations in Pilot Plant Columns 145 Foaming Tendency of Treated Effluent 145 Effect of Operating Conditions on Adsorption of Color and TOC on Carbon 146 XIV TREATMENT PLANT DESIGN STUDIES AND COST ESTIMATES 147 Overall Design Criteria 147 Design Methods and Assumptions 150 Cost Estimates 156 Results of Cost Estimates 157 Conclusions From Cost Estimates 163 XV REFERENCES 166 XVI PUBLICATIONS AND PATENTS 170 XVII GLOSSARY 172 XVIII APPENDICES 173 A. Water Quality Standards for Specific Uses 175 B. Procedures for Isotherms, Laboratory Lime Treatment, Laboratory Bio-Oxidation and Analyses 179 C. Detailed Information on Pilot Plant Equipment 183 D. Printout of Daily Summaries of Conditions and Results for Bio-Carbon Sequence 185 E. Printout of Daily Summaries of Conditions and Results for Primary-Carbon Sequence 186 F. Printout of Daily Summaries of Conditions and Results for Lime Treatment 187 G. Printout of Daily Summaries of Conditions and Results for Lime-Carbon Sequence 188 vi ------- CONTENTS (Continued) XVIII APPENDICES (Continued) Page H. Generalized Procedure for Designing Plant for Treating Kraft Pulp Mill Effluent by Adsorption in Granular Activated Carbon Columns 189 I. Relationships Useful in Design of Downflow Carbon Adsorption Columns 193 J. Estimate No. 1 - Microlime - Carbon Treatment of 9.6 MGD of Effluent from a New 800 Ton/Day Unbleached Kraft Pulp and Paper Mill 194 vii ------- FIGURES No. f" 1 Carbon Adsorption Isotherms for Color Removal from Total Mill Effluent 11 2 Carbon Adsorption Isotherms for Total Organic Carbon Removal from Total Mill Effluent ' 12 3 Comparison of Cumulative Loadings on Darco and WV-M Carbons in Simultaneous 2-Inch Column Tests 20 4 Carbon Adsorption Isotherms for Specific Compounds in Turpentine Underflow 26 5 Effluent Treatment Pilot Plant 50 6 FACET System 53 7 Exterior of Pilot Plant Showing Lime Treater, Carbonation Tank, Storage Tank, and Basin 55 8 Interior of Pilot Plant with Carbon Columns, FACET Tank (foreground) and Duomedia Filter (right foreground) 56 » 9 Influence of Calcium in Lime-Treated Water on Percent of Color Removed and on Color Remaining 72 10 Soluble Calcium in Lime-Treated Water Versus Increased Conductivity due to Lime Treatment 73 11 Concentrations of Color and TOC During Bio-Oxidation- Carbon Adsorption, Feb. 1 through Mar. 11, 1972 82 12 Color and TOC Remaining Versus Length of Carbon Bed During Bio-Carbon Sequence 87 13 Cumulative Removals of TOC on Carbon, Feb. 1 to Mar. 11, 1972 90 14 Rates of Removal of Color and TOC by Carbon Adsorption During Bio-Carbon Sequence, Feb. 4 to Mar. 11, 1972 92 15 Rates of Removal of Color and TOC by Carbon Columns as a Function of Average Concentration During Bio-Carbon Operation, Feb. 4 to Mar. 11, 1972 93 viii ------- FIGURES (Continued) No. Page 16 Color and TOG Remaining Versus Length of Bed During Primary- Carbon Sequence 101 17 Rate of Removal of Color and TOC by Carbon Adsorption During Primary-Carbon Operation at 10 gpm 103 18 Effect of Calcium Concentration on Removal of Color and TOC by Carbon Adsorption (Averages from Operating Periods) 112 19 Concentrations of Color and TOC Remaining Versus Length of Carbon Bed 117 20 Remaining Color in Water from Columns During Lime-Carbon Sequence 119 21 Remaining TOC in Water from Columns During Lime-Carbon Sequence 120 22 Color and TOC Versus Stage of Lime-Carbon Treatment for Medium Calcium Run 121 23 Cumulative Removal of Color Versus Time for Lime-Carbon Operation 125 24 Cumulative Removal of TOC Versus Time for Lime-Carbon Operation 126 25 Changes in Color of Dilute Black Liquor with Time of Stirring with Activated Carbon 139 26 Average Color Isotherms After Various Treatments During Pilot Plant Operations 142 27 Average TOC Isotherms After Various Treatments During Pilot Plant Operations 143 28 Flow Diagrams and Water Quality for Cost Estimates 1-11 149 29 Process Flow Diagram for FWPCA-Kellogg Study (17) 153 30 Flow Diagram for Microlime-Carbon Treatment of Unbleached Kraft Effluent for Reuse ix ------- TABLES No. ISSS 1 Laboratory Treatments of Effluents 8 2 Carbons Selected for Adsorption Studies 9 3 Comparisons of Carbons for Removal of Color and TOC from Two Mill Effluents 14 4 Comparison of Carbons for Removal of Color and TOC from Paper Machine White Water from Mill B, at pH 7.7 15 5 Effect of Treatment Sequences on Removal of Color from Mill Effluents 16 6 Effects of Treatment Sequences on Removal of TOC from Mill Effluents 17 7 Lime Treatment of Unbleached Kraft Effluent 22 8 Lime-Carbon Treatment of Unbleached Kraft Effluent 22 9 Color Change of Kraft Effluent 24 10 Ranges of Desired Properties of Process Water for Unbleached Kraft Manufacture 32 11 Effect of Treatment Sequences on Removal of Color from Total Mill Effluents 34 12 Effects of Treatment Sequences on Removal of TOC from Total Mill Effluents 35 13 Order-of-Magnitude Cost Estimates for Various Treatments of Total Mill Effluent from Mill A Effluent 40 14 Order-of-Magnitude Cost Estimates for Various Treatments of Total Mill Effluents from Mills A and B 41 15 Summary of Conditions of Treatment and Estimated Capital and Operating Costs for Various Treatments of Total Mill Effluent from Mills A and B 42 16 Properties of Water Routinely Measured 62 ------- TABLES (Continued) No. Page 17 Bio-Oxidation and Clarification for Period of 2/4/72 to 3/11/72 at Flow Rate of 20 gpm 65 18 Average Results from Lime Treatment 70 19 Relation of Specific Gravity and Solids Concentration of Lime Treater Sludge 74 20 Distribution of Calcium in Lime Treatment 76 21 Operating Summary of Biological Oxidation - Carbon Adsorption Sequence at 15 gpm, Low Carbon Dosage 84 22 Average Results for Bio-Carbon Sequence for Period of February 4, 1972 to March 11, 1972, Low Carbon Dosage 85 23 Results from Bio-Carbon Adsorption with Differing Carbon Dosage Rates 88 24 Summary of Primary Clarification-Carbon Adsorption Sequence 98 25 Average Results for Primary-Carbon Sequence for Period of April 4 through May 20, 1972 at Flow Rate of 10 gpm 99 26 Displacement of Color and TOG during Backwashing of Columns 108 27 Comparison of Removal Rates of Exhausted Carbon with Normal Operation 109 28 Lime-Carbon Sequence of Treatment: Removals of Color and TOC 114 29 Summary of Carbon Adsorption Following Lime Treatment: Removals of BOD and Turbidity 115 30 Carbon Adsorption Treatment of Lime-Treated Effluent Flow=10 gpm 116 31 Organic Compounds in Lime-Carbon Operations 127 32 Metal Ion Concentrations in Lime-Carbon Sequence 129 33 Results from FACET Operation with Lime Treated Water at 10 gpm 131 34 Comparison of Lime-FACET and Lime-Carbon Column Operation at a Water Flow Rate of 10 gpm 134 xi ------- TABLES (Continued) No. 35 Average Equilibrium Loadings on Carbon from Isotherms Prepared During Pilot Plant Operation 141 36 Bases Used for Design and Cost Studies 148 37 Average Composition of Water To and From Treatment 148 38 Comparison of Costs for Treating Unbleached Kraft Mill Effluent in New and Older Mills by Various Treatment Sequences with Carbon at $0.27/lb 158 39 Comparison of Costs for Microlime and Minilime Processes 162 40 Effect of Carbon Price on Treatment Costs 164 41 Potential Water Quality Standards for Specific Water Uses 176 xii ------- ACKNOWLEDGEMENTS The authors gratefully acknowledge the invaluable contributions to this program made by the many people in operations, engineering and research and development who made their time freely available during this program. The list of these contributors is too long to mention separately. It includes people from operating companies including St. Regis Paper Co., engineering firms, manufacturers of activated carbon, academic and feder- ally sponsored research, development and demonstration facilities. The National Council of the Paper Industry for Air and Stream Improvement (NCASI), particularly through H. F. Berger, Southern Regional Engineer, lent its assistance throughout this program. Dr. C. W. Dence and Dr. P. Luner of Syracuse University and Dr. H. Dugal of the Institute of Paper Chemistry assisted in developing an understanding of problems connected with effluent color. Mr. E. L. Spruill, Mr. C. L. Davis, and Mr. M. Gould lent valuable assis- tance in the area of lime treatment. The support of the program by the Environmental Protection Agency, and the help provided by Mr. W. J. Lacy, Mr. G- R. Webster and Mr. J. Gallup, is gratefully acknowledged. Major St. Regis contributors to this study include Mr. W. W. Spangler, Mr. D. Barnes, Mr. M. R. Garrett (laboratory studies and pilot plant operations), Mr. J. N. Rockwell (analytical), Mr. J. L. Gillespie (pilot plant construction), Mr. J. Moore (computer data analysis), and Mrs. B. Halfacre (secretarial). xiii ------- SECTION I CONCLUSIONS 1. Unbleached kraft mill effluent can be treated by sequences of lime treatment or primary clarification plus bio-oxidation or primary clarification followed by carbon adsorption in granular carbon columns to provide water suitable for general reuse in the mill. 2. Cost estimates indicate that a microlime-carbon sequence would be the least costly for treating effluent containing 1000 APHA-NCASI color units (CU) and 250 mg/1 total organic carbon (TOG) to provide reusable water containing 100 CU and 100 mg/1 TOC. For treating 9.6 mgd of effluent from a 800 ton/day mill, the capital cost was estimated to be $6-75 million and the total operating cost inclusive of capital depreciation was estimated to be $0.30 per 1000 gal, or $3.58 per pulp ton, with a credit of $0.06 per 1000 gal for the re- used water. 3. Use of lower-cost carbon produced on-site at 10$/lb vs. a purchased cost of 27/lb reduces operating costs of the microlime-carbon se- quence by 137o, and capital costs by 87°. 4. Operating costs are reduced about 1270 per 1000 gal treated when the pulp production and effluent volume are doubled. 5. Effluent treatment by the microlime-carbon sequence is about 57= lower in capital and 147° lower in operating cost then by the minilime-carbon sequence. 6. Removal of color to 100 CU and of TOC to 100 mg/1 by the microlime- carbon sequence requires maintaining a dissolved calcium concentration of about 80 mg/1 Ca (minimum of about 40 mg/1) by the use of about 500 mg/1 of added CaO, no carbonation, and requires 2.5 Ib of activated carbon per 1000 gal effluent treated at a contact time of 1.8 hrs. At these levels, the microlime treatment alone removes 707° of the color, compared to 85-907o at a dosage of about 1000 mg/1 CaO in the minilime treatment. 7. The lime-color sludge obtained at 80 mg/1 dissolved Ca contains up to 3570 solids and filters to a solids content suitable for feeding to a lime kiln. 8. The primary-bio-carbon sequence provides reusable water quality at 10.5 Ib activated carbon per 1000 gal of effluent at a contact time ------- of 3.8 hrs. These requirements caused high capital costs and the resulting total operating costs were $0.58/kgal or $10.50/pulp ton, which were about double those for microlime-carbon. 9. The primary-carbon sequence provides reusable water quality at a dosage of 23.4 Ib activated carbon per 1000 gal effluent treated and a contact time of 5.3 hrs. These requirements caused high capital and operating costs and made this sequence much more costly than the lime-carbon sequence. 10. Granular carbon rather than powdered activated carbon is the economical choice because high dosage requirements (including the microlime-carbon sequence) necessitate regenerative use of activated carbon even if available at very low cost from mill site production. Regeneration systems for powdered carbon, though under development, have not as yet been applied on a large scale. 11. Fixed-bed carbon adsorption of raw clarified effluent or of water treated by bio-oxidation or lime gave no operating problems. Back- washing of the beds every one or two days prevents buildup of pressure drop. 12. A novel continuous countercurrent process (FACET) for carbon adsorption using stirred tanks of carbon slurries was shown to have considerable promise for reducing capital and operating costs, and its development is being continued. », 13. Non-adsorptive mechanisms are responsible for as much as 30% of color removal and 20% of TOG removal, particularly in the primary-carbon sequence. 14. Anaerobic digestion in the equalization basin resulted in substantial increases in non-adsorbable TOG in the form of low molecular weight organic compounds. 15. Interactions between carbon and effluent lead to not-fully-understood color and suspended solids phenomena, particularly in the FACET system. 16. The color of lime treated effluent is subject to increases of approx- imately 50% upon aging or bio-oxidation. 17. Carbon adsorption removed almost all of the foaming tendency of effluents. ------- SECTION II RECOMMENDATIONS It is recommended that the concept of treatment and reuse rather than discharge be given serious consideration by operating mills and by those planning to install new production and effluent treatment facilities. This concept appears to be a sound one from an economic standpoint partic- ularly when, as it appears, less severe quality criteria can be applied to treatment for reuse as compared to treatment for discharge, particular- ly with respect to BOD and TOG. Although a working definition of reusable water quality has been established in this study, i.e. 100 CU and 100 mg/1 TOG, and although many internal streams are used today in the kraft mill of even poorer quality, no generally accepted criteria exist and it is recommended that these be established by an industry committee. These criteria necessarily must include additional parameters which were not primary performance criteria in this study. It is also recommended that grants be established to demonstrate reusability of treated effluents on a mill scale, first in selected areas and subsequently on a maximum feasi- ble basis while appropriate operating and product quality parameters are monitored. It is recommended that Part II of this grant be continued to conclusion to assure availability of low cost carbon from mill sources for application in effluent treatment and reuse. It is recommended that the development of the FACET carbon adsorption system be continued as a means of reducing the cost associated with acti- vated carbon treatment. In view of the cost per 1000 gal of water treat- ed, it is recommended that application of advanced effluent treatment for color removal be preceded, in order to reduce overall water costs, by a study and/or implementation of suitable and economically advantageous methods of reducing; a) the total effluent load to the treatment system; b) the color that is allowed to enter it; and c) the fluctuations in concentration in the effluent. Specific questions recommended for further investigation include permissi- ble levels in reused water of dissolved solids such as sodium and calcium salts, heavy metal salts and others; effect of treatment and reuse on temperature levels in critical process areas, effect of dissolved salts on sheet formation, drainage, corrosion, pulp washing, lime mud washing, and scale formation. ------- SECTION III INTRODUCTION St. Regis Paper Company with partial support by EPA has been engaged since July, 1969, in a program for the development of an economical system for maximum water reuse in the kraft pulp and paper industry as a means of water pollution control and water conservation. This program is based on two key concepts: (1) effluent treatment using activated carbon and, (2) on-site production of activated carbon from readily available raw materials, particularly black liquor, with full integration into the kraft mill recovery and power systems to achieve the lowest net cost of activated carbon. This report covers the work accomplished under Part 1 of the program: effluent treatment using activated carbon. Preliminary results from the laboratory phase (15) and pilot plant (45,46,47) have been reported previously. Maximum water reuse, though not strictly definable, is assumed to be in the order of 90% reuse. Ten percent of the usual fresh water input to the kraft mill, it is assumed, will be discharged primarily to purge inorganic salts at nonpolluting levels. This effluent would discharge a total organic contaminant load equivalent to 10% or less of present discharges. Cost estimates made prior to starting this program indicated that treat- ment of unbleached kraft pulp and paper mill effluent by process combina- tions involving activated carbon, produced at the mill, and reuse of the treated water would be less costly than treatment by other processes to meet anticipated discharge standards, including standards for color. The first task of this program was a survey of the literature and a review of information from industry on treatment of effluents and reuse of treated water in pulp and paper mills. The findings of this survey were published in the first report of this program (40). All other activities of Part I are covered by the present report. The basic possibility of using activated carbon to obtain high-quality effluents, particularly in terms of color removal, was indicated in studies by McGlasson (10,11) and Berger (4). Treatment of kraft effluents with lime, as reported a number of years ago by Berger (3) and commercially applied by Davis (5), achieves substantial color removal and thus appeared to merit consideration as a treatment ------- step prior to activated carbon. Further reports on commercial installa- tions for lime decolorization of pulp effluents have appeared more recent- ly (7,8,14,18). The success of biological oxidation systems, particularly aerated lagoons in the southern states, in achieving as much as 90% reduction in BOD-5, and the trend towards more installations of this type, provided sufficient argument to include in this study an investigation of the effectiveness of activated carbon treatment of bio-oxidized kraft effluent. Thus it appeared that the role of activated carbon treatment might be as the sole treatment step for color and dissolved organics, or as a major second step after lime or biological treatment, or as a polishing step following a combination of lime treatment and biological oxidation. On the basis of laboratory studies and preliminary cost estimates, given in this report, a decision was made to build and operate a pilot plant at the Pensacola, Florida, St. Regis pulp and paper mill with a capacity of about 30 gpm of unbleached kraft effluent to investigate three sequen- ces of treatment: Primary clarification, activated carbon adsorption Lime treatment, carbon adsorption Primary clarification, biological oxidation, carbon adsorption. The results from operating this pilot plant from November, 1971, through December, 1972, and cost estimates for treating pulp mill effluents for reuse constitute a major part of this report. ------- SECTION IV LABORATORY STUDIES INTRODUCTION The purpose of the laboratory study was to help define where and how activated carbon might be used most economically in order to achieve treatment and up to 90% reuse of effluent from kraft pulp and paper mills. Based on laboratory results, and with the aid of preliminary designs and economic estimates as presented in Section V, the available choices were to be narrowed down to three systmes and a pilot plant designed and a pilot plant operation program developed. In this study, two main approaches were pursued in terms of effluents to be treated. These were the treatment of specific effluents for specific reuse, and the treatment of the total effluent for general reuse. It was recognized that the best approach might be a compromise of the two, such as separate treatment of evaporator condensate and turpentine underflow for reuse in lime mud and pulp washing, while combining all other effluents for treat- ment and general reuse. For the total effluent, the major treatments studied were primary clarifi- cation followed by activated carbon; primary clarification, lime treatment, activated carbon; primary clarification, lime treatment, biological oxida- tion, and activated carbon; primary clarification, biological oxidation, and activated carbon; and primary clarification, biological oxidation, lime treatment, and activated carbon. In these laboratory studies, "primary clarification" (simulated by filtration) was always carried out separately as a first treatment. Mill scale applications (14,18) and hence design and cost studies (Section XIV) use no clarification ahead of lime treatment. Laboratory treatment methods are described in Appen- dix B. These same treatment combinations were studied with the caustic and acid bleach effluents, although general reuse of bleach effluents is considered unlikely until chloride removal can be achieved economically. Only minor reductions can be expected in the chloride content as a result of any of the treatment sequences investigated. Hence, reuse of the bleach effluents in the pulp mill would eventually lead to an undesirable build-up of chloride in the black liquor. For other specific effluents, the use of activated carbon treatment alone was investigated in this study, although other treatments may also be required. For example, fiber removal from paper machine white water will be necessary before applying activated carbon treatment for the removal of dissolved organics. ------- By using effluents from the three southern mills of St. Regis, and one other mill with lime treatment, representing a spectrum of modern and older plants, the chances were increased that this program would be of value to the entire pulp and paper industry. The study included labora- tory treatment of effluent samples from various mills before any treatment and after full-scale primary, biological, and/or lime treatments. Table 1 summarizes the treatments used in the laboratory study with each of the effluents. In terms of activated carbon treatment per se, the primary function of the laboratory study was to determine the dosages of carbon required to reach desired concentrations of color and TOC in the various effluents to be tested. These dosage data usually are obtained in the form of adsorption isotherms as described in Appendix B. Carbon dosage data are part of the information required to prepare preliminary designs and make preliminary cost estimates for potential treatment systems. More specif- ically, for instance, dosage data together with other cost data for the carbon itself as well as for equipment, are major determining factors in deciding between single use of carbon (usually as powdered carbon) and regenerative use of carbon (usually as granular carbon). Commercially available carbons vary widely in their characteristics and more particularly in their effectiveness in removing dissolved TOC or color from different effluent streams. Although it is an objective of the two-part program to develop an activated carbon for effluent treatment from pulp mill sources (particularly black liquor), this carbon was not available in sufficient quantity for the laboratory program or the subse- quent pilot plant operation. By using a wide variety of commercially available activated carbons in the laboratory program, it was possible to select suitable commercial carbons for the pilot plant operation, as well as for potential full-scale applications, and it was also hoped that the information gained would help in the definition of characteristics to be obtained in the carbon made from black liquor. The major part of the laboratory study was devoted to establishing adsorption equilibrium data for carbon as discussed above. The final selection of the carbon to be used in the pilot plant was made on the basis of additional isotherm work on Pensacola No. 2 Mill effluent (unbleached kraft) and on the basis of comparative adsorption column work. This work was done after reaching the decision (see Section V) to treat unbleached kraft total effluent using granular carbon columns, and it also helped to determine the size of the pilot plant carbon columns. After selecting the lime-carbon treatment sequence as one of the three pilot plant treatment sequences, (see Section?), additional studies were performed in the area of lime and lime-carbon treatment. ------- Table 1. LABORATORY TREATMENTS OF EFFLUENTS Specific effluents Type of treatment Turpentine underflow Carbon Evaporator condenstate Carbon Paper machine white water Carbon Condensate cooling tower Carbon Woodyard Carbon fe Caustic bleach Carbon and combined. Acid bleach Carbon and combined Total mill effluents Total mill Carbon and combined High solids sewer Carbon Low solids sewer Carbon Total mill without biological oxidation Carbon Total mill with biological oxidation Carbon Total mill lime treated Carbon and bio-oxidation Simulated pulp mill effluents 0.05% black liquor solids Carbon 0.5% black liquor solids Carbon All treatments preceded by filtration Including lime and bio-oxidation in various combinations ------- Special studies were carried out in two other areas. One of these areas was studies leading to a continuous, countercurrent, fine carbon contact- ing system called FACET, for which a patent has been awarded (41), and which underwent initial development in the pilot plant. The other area is that of carbon adsorption in an aerated system with or without biolog- ical activity. Significant observations which have some bearing on the selection of a treatment system and its results are discussed in a separate part of this section, including observed color reversion after lime treatment, and adsorption or non-adsorption of compounds in effluents. SELECTION OF CARBONS FOR EVALUATION Each of the major manufacturers or suppliers of activated carbons for water treatment was contacted to determine which of their carbons would likely be best for removal of color and TOG from each of the effluents to be studied. Help in selecting carbons for the demonstration was also obtained from a study of the literature and discussions with other investi- gators working on water treatment. In general, the cost of the carbon and whether it is available in granular or powder form were disregarded. Samples for adsorption studies in powder form were supplied by Westvaco, Calgon, Atlas, Norit, and Barnebey-Cheney. A list of the carbons used in this study is given in Table 2. Table 2. CARBONS SELECTED FOR ADSORPTION STUDIES Barnebey- Westvaco Calgon Atlas Norit Cheney AN F-100 S-51 A SC C-190 F-300 KB F XZ WV-L F-400 DARCO FQA WV-G XPT NC WV-W XPS SG WV-M RESULTS OF CARBON AND COMBINED TREATMENTS Evaluation of Activated Carbons The major means of determining the effectiveness of each type of acti- vated carbon was carbon adsorption isotherms, the procedure for which ------- is described in Appendix B. Six samples of the effluent were treated with six different dosages of powdered carbon, and the results were plotted on log-log graphs of loading of impurity on the carbon versus concentration of impurity remaining in the water sample. Isotherms for four carbons for removal of color from the total mill effluent from mill A are shown in Figure 1. Note that the isotherms all had definite breaks at a color concentration of about 300 CU. Typical isotherms for removal of TOG are shown in Figure 2. These isotherms employ the same four carbons and the same effluent as the color isotherms above. The TOC isotherms are almost identical for all four carbons and can be represented by a single line. Such isotherms for color and TOC removal were prepared using two to four commercial carbons for each of the various effluent streams and after pre-treatment by lime or bio-oxidation or both. Over 150 isotherms, not including duplicates and repeats, were prepared. The reasons for using several carbons for each stream were (1) to indi- cate which of the commercially available carbons recommended by the manufacturers are best for treating each stream, and (2) to indicate which physical properties, such as pore size, of carbons are best for each stream. Some of the results of these carbon treatments are less definitive than had been hoped owing to several factors: The feed concentration and composition of a given stream often varied considerably over the period of time that samples were taken for the pre-treatment and carbon adsorption experiments. Some of the isotherms had very steep slopes owing to high concentra- tions of poorly adsorbed compounds, such as methanol. Some isotherms, particularly for color, were made up of lines of several slopes caused by the several types of components in the water. Comparisons of Carbons The most reliable method of comparing carbons for removal of impurities from water samples was to compare the loading of color or TOC per unit weight of carbon at selected concentrations of color and TOC. These loadings were read from the isotherms at color concentrations of 1000, 500, 250 and 100 CU and TOC concentrations of 150, 100 and 50 mg/1, representing intermediate and fairly complete treatment levels. 10 ------- 5000 I N 60 1= U O 53 H I I I I I I I I LEGEND OF CARBONS TESTED 50 100 500 1000 COLOR, APHA color units Figure 1. Carbon adsorption isotherms for color removal from total mill effluent 11 ------- SOOrTTTT 100 50 g M § 10 LEGEND OF CARBONS TESTED 1 A D a © AN F-400 KB NC 5 10 50 100 500 1000 TOTAL ORGANIC CARBON, mg/1 Figure 2. Carbon adsorption isotherms for total organic carbon removal from total mill effluent 12 ------- Another commonly used method is to compare the concentration of the impurity remaining in the water after equal dosages of the several carbons. Disadvantages of this method are that it requires that all carbons be tested with the same sample of water and that it provides no information on the effects of concentration on loading which isotherms would provide. In addition to loading capacity, other properties of carbons must be considered in choosing a carbon for a given application. These properties include whether powdered or granular carbon is best, the rates of adsorp- tion, the friability of the (granular) carbon, handling problems, and cost. If a carbon is to be used in a column contactor, it should be evaluated in continuous column tests with the particle size most likely to be used in the future plant scale of operation. In fully counter-current adsorption, the carbon dosage required to reduce the color or TOG from any feed concentration to some desired level can be calculated by using the following formula: Dosage, g/1 = Co - D (1) Xo where: Co= Initial concentration in CU or mgTOC/1, D = Desired concentration of color or TOG to be reached Xo= Loading at Co, CU/g or mgTOC/g. If a batchwise single - stage treatment is used, the dosage required is: Dosage, g/1 = Co - D (2) Xd Where Xd is the loading at the desired final concentration In practice, the maximum loading reached in a packed column system using granular carbon is about 65% of Xo from the isotherm, resulting in a. corresponding increase in the required dosage amounting to about 50% in excess of the dosage calculated using equation (1) above. This difference in loading is due to two major factors. First, a packed column is not a true counter-current contacting system, and only a fraction of the carbon actually is in contact with effluent of the feed concentration, Co, while most of the carbon is in contact with effluent of decreasing concentration in the direction of effluent flow. Second, the equilibrium that is estab- lished in a flow system is between the adsorptive surface and the effluent in immediate contact with it, while the bulk effluent remains at a higher concentration. This concentration difference provides the driving force for the transfer of the dissolved substances from the bulk phase through the stagnant water film surrounding each carbon particle and the water in 13 ------- the pores of the carbon. Results taken from a total of 29 isotherms are given in Table 3. These results were obtained using the total mill effluent from two southern kraft mills. The carbons are ranked according to the relative effectiveness of each for removal of color at 500 CU remaining in solution. Note that the ranking would have been different at a lower remaining color concentration, except that the top carbon performed best at either concentration. Note further that top performers in terms of color removal are not the top performers in terms of TOC removal. For Mill B effluent, S-51 and AN carbons were best, while for Mill A the KB and AN carbons were best. For pulp and paper mill effluents, color removal tends to be the more important characteristic of a good carbon. Mill A effluent generally resulted in higher loadings than Mill B effluent, probably because of the presence of sea water in Mill A effluent. Another comparison of carbons is shown in Table 4. This compares four carbons used to treat paper machine white water from Mill B. Results indicate that the AN and S-51 carbons were superior in color and TOC removal. The C-190 carbon, which showed good color loading, performed poorly on removal of TOCS Table 3. COMPARISONS OF CARBONS FOR REMOVAL OF COLOR AND TOC FROM TWO MILL EFFLUENTS Color removal TOC removal No. samples avg. loading,CU/g avg. loading, mg/1 Carbon averaged at 500 CU at 100 CU at 100 mg/1 at 50 tag/1 Mill A effluent KB 1 1200 200 200 30 AN 4 840 170 110 24 F-400 1 700 125 350 22 NC 1 500 180 • 250 23 Mill B effluent S-51 6 260 140 40 11 AN 8 250 90 30 7 NC 1 250 60 21 8 WV-G 1 145 120 30 WV-L 3 140 40 21 8 P-300 1 130 75 100 9 KB 1 110 100 110 30 WV-M 1 60 23 20 1 14 ------- Similar comparisons of carbons were prepared from isotherms of carbons made with other in-plant and total mill effluents. It was concluded that the most promising carbons for removal of color and TOG from the major mill effluents were AN, S-51, KB, F-300, and NC. Of these five carbons AN, KB, and NC are powdered carbons, and S-51 (as Granular Darco) and F-300 are granular carbons. Table 4. COMPARISON OF CARBONS FOR REMOVAL OF COLOR AND TOC FROM PAPER MACHINE WHITE WATER FROM MILL B, AT PH 7.7 Color removal No. samples avg. loading, CU/g TOC removal avg. loading, Carbon averaged at 100 CU mg/g at 50 mg/1 AN 3 577 137 S-51 2 279 285 C-190 3 247 65 F-400 2 115 77 Comparison of Treatment Sequences Involving Activated Carbon The purpose of this program was to determine the best means of treating in-plant and final effluents of pulp and paper mills for water reuse. The most economic method of treating a given stream might be with primary clarification and carbon treatment, or with carbon following lime treat- ment, primary clarification and bio-oxidation, or combinations of lime treatment and bio-oxidation. The effects of such combinations of pre- treatment and adsorption of color and TOC by carbon are illustrated by the results in Tables 5 and 6 for the treatment of a total mill effluent with AN carbon. The treatment methods are described in Appendix B. (All treatment sequences carried out in the lab start with paper filtra- tion, simulating primary clarification.) The carbon treatments are all with 1 g of AN powdered carbon per liter of effluent. 15 ------- Table 5. EFFECT OF TREATMENT SEQUENCES ON REMOVAL OF COLOR FROM MILL EFFLUENTS Treatment3 Initial colorb CU Final color CU Using total mill effluent from Mill A Carbon Bio Carbon Bio Lime Carbon Lime Carbon Lime Bio Carbon 948 833 570 833 570 45 1380 70 785 64 185 370 570 180 570 45 10 70 8 64 185 10 Using total mill effluents from Mills C and D Mill C No aeration Carbon Mill C With aeration Carbon Mill D Lime Carbon Mill D Lime Bio (in lab) Carbon 4590 4590 2025 525C 525 c 479 2500 2025 700 525 c 90 525 c 479 190 Initial PH 8.6 7.5 7.4 7.2 7.5 10.1 7.5 7.2 7.5 a Basis of treatments: filtration of the water samples before treatment to simulate primary clarification, single-stage powdered carbon (AN) with dosage of 1 g/1, biological oxidation for 40 hr with nutrients and seed added, lime treatment with dosage of 5 g/1 of CaO. b Four different samples of effluent used for five treatment sequences. c The high color levels after lime treatment are not representative of normal mill performance. 16 ------- Table 6. EFFECTS OF TREATMENT SEQUENCES ON REMOVAL OF TOG FROM MILL EFFLUENTS Treatments Using total mill Carbon Bio Carbon Bio Lime Carbon Lime Carbon Lime Bio Carbon Initial TOG, mg/1 effluent from Mill A 138 111 66 111 66 - 106 67 106 67 46 Using Mill C effluent: before and after No aeration Carbon With aeration Carbon 713 713 183 Final TOG mg/1 75 66 30 66 - - 67 38 67 46 23 aeration basin 570 183 90 Initial PH 8.6 7.5 7.4 7.2 7.2 10.1 7.5 a Basis of treatments: filtration of the water samples before treatment to simulate primary clarification, single-stage powdered carbon (AN) with dosage of 1 g/1, biological oxidation for 40 hr with nutrients and seed added, lime treatment with dosage of 5 g/1 CaO. 17 ------- When only carbon was used, the color was reduced from 948 to 370 CU with a dosage of 1 g/1 (see Table 5). When this effluent sample was first treated by bio-oxidation, the color was reduced only slightly (to 570 CU). The subsequent carbon treatment reduced the color to 180 CU with a dosage of 1 g/1. When lime treatment was used following bio-oxidation the color was reduced sharply to 45 CU, and a dosage of 1 g/1 of carbon reduced it further to 10 CU. The use of bio-oxidation following lime treatment was not as effective as the reverse treatment, since the color increased during the bio-oxidation. (This reversion of color after lime treatment is discussed later in this section.) This increased color was readily removed to a level of 100 CU with a carbon dosage of 0.1 g/1 and to 10 CU with a dosage of 1.0 g/1. The effects of the same treatment sequences on TOC removal from the same total mill effluent are given in Table 6. With carbon alone, the TOC was reduced from 138 mg/1 to 75 mg/1 with a single dosage of 1 g/1. Bio-oxidation reduced the TOC from 111 to 66 mg/1, and carbon treatment further reduced it to 30 mg/1 with a dosage of 1 g/1. Tables 5 and 6 also compare treatment methods on the basis of color and TOC removals using effluents from Mill C before and after bio-oxidation at the mill. Carbon treatment of the effluent before bio-oxidation with a dosage of 1 g/1 reduced the color from 4590 to 2500 CU and reduced the color after bio-oxidation from 2025 to 700. After lime treatment at Mill D, carbon treatment with a dosage of 1 g/1 reduced the color of the mill effluent to 90 CU. Bio-oxidation of this lime-treated effluent reduced the color only slightly, and carbon was less effective for color removal than before the effluent was bio-oxidized. The data in Table 6 for treatment of samples from Mill C show that the aeration basin reduced the TOC from 713 to 183 mg/1 and that treatment at 1 g/1 further reduced the TOC to 90 mg/1. The tentative conclusions drawn from this comparison of treatment se- quences on total mill effluent, when the goal is reduction to less than 100 CU and less than 50 mg/1 TOC, are that primary emphasis in the subsequent studies should be placed on carbon treatment, on bio-oxidation followed by carbon treatment, on lime treatment, and on lime treatment followed by carbon treatment. It appears that treatment sequences in- cluding both bio-oxidation and lime treatment before carbon treatment offer comparatively less promise for economical treatments of mill effluents for reuse. These conclusions are based on activated carbon being produced at low cost from black liquor at the mill site, which is the original basis of this program. Goals of different color or TOC levels for the treated water will affect the comparisons of treatment sequences, as will changes in the relative feed levels of color and TOC. It should be noted that in the comparison presented, the controlling impurity is color. 18 ------- FINAL SELECTION OF COMMERCIAL CARBON FOR PILOT PLANT As will be discussed in Section V, the selection from the many alternatives was narrowed to further work on three treatment sequences using total mill effluent (unbleached) and using a granular rather than powdered carbon adsorption system. A review of available information, such as is presented in Table 3, on treating Mill B effluent (Pensacola No. 2 Mill, chosen as the pilot plant site), indicated that Atlas S-51, Westvaco AN, Atlas KB, and Calgon F-300 gave the best overall results. Of the carbons available in granular form, the ranking of the carbons is (1) Atlas Granular Darco, (2) Calgon F-300, and (3) Westvaco WV-L. Similar comparisons were made of the candidate carbons using additional samples of Pensacola No. 2 Mill effluent, in-plant streams, and synthetic effluent prepared by diluting spent pulping liquor to the concentration equivalent to that of mill effluents (containing 0.05% black liquor solids). The conclusion from these isotherm evaluations was that the most promising carbons were Atlas Granular Darco and Westvaco granular carbons. The rates of adsorption of these carbons were compared in a series of dynamic column adsorption runs using Pensacola No. 2 Mill effluent at a superficial flow velocity of 1 to 2 gpm/ft^in 2-in. columns containing equal weights of carbon (about 4-ft beds). The carbons included Atlas Darco 20 x 40 and 8 x 30 and Westvaco WV-L 8 x 30, WV-W 20 x 50, WV-M 20 x 40. In these tests, the two Darco carbons were superior to the Westvaco carbons of corresponding size in the rate and degree of removal of color and TOG. These column adsorption runs showed that the removal of color from the mill effluent would be more difficult than the removal of TOG in reaching the selected goals of 100 CU and 50 mg/1 TOG. Therefore the properties of loadings and rates of removal of color were of prime importance in judging which carbon should be used in the pilot plant. Results of one of the side by side comparisons are shown in Figure 3, comparing cumulative loadings of color and TOG on granular Darco and WV-M carbons. The low rates of removal of color in both column tests indicated that the contact time of water in the pilot plant or commercial columns would be much longer than is commonly used in carbon treatment of municipal waste- water. To provide the longer time, it was decided to use a low flow rate of 1 to 2 gpm/ft of cross sectional area, of the columns. At this low flow rate, pressure drop would be low and pumping cost would not be an important factor in operating costs in a commercial unit. Therefore, it was decided that to gain as much external area of carbon as possible while still maintaining a reasonable pressure drop, 20 x 40 mesh carbon should be used. The 20 x 40 mesh carbon has an external surface area about four times that of the more generally used 8 x 30 mesh carbon and thus gives a 19 ------- Column A Column B Darco 20 x 40 lOOOg, 2.381 liters 1.01 gpm/ft2 2.1 bed vols./hr Westvaco WV-M 20 x 40 lOOOg, 1.981 liters 1.01 gpm/ft2 2.6 bed vols./hr 250 o o H 00 C3 U W O i-J O u PN o o a M 5 200 - 150 - A, DARCO, COLOR 100 - 50 100 150 CUMULATIVE BED VOLUMES 200 250 Figure 3. Comparison of cumulative loadings on Darco and WV-M carbons in simultaneous 2-inch column tests 20 ------- much higher rate of adsorption. A finer size of Granular Darco (Darco XPT) was available which had a particle size range (40 x 140 mesh) that was close to the 60 x 140 size considered desirable for the FACET system as described in Sections V and VI. Therefore, Darco XPT was also evaluated in isotherm and column tests and found to be very similar to Darco 20 x 40 in equilibrium adsorption properties and the rates of removal were about twice those with Darco 20 x 40. Both the Darco 20 x 40 and the 40 x 140 carbons were evaluated for resis- tance to attrition in stirring tests and by pumping the Darco XPT as a slurry. The resistance of the Darco carbons to attrition was not as high as some carbons but was considered satisfactory. Atlas Granular Darco was finally selected for use in the pilot plant on the basis of the following points: It had generally superior adsorption capacities and rates of adsorption for color and TOG. The Atlas carbon was available in the two size ranges needed for the pilot plant. The raw material for the Atlas carbon (lignite) is similar to the lignin-derived raw material to be used in producing activated carbon from kraft pulping liquor under Part II of the EPA - St. Regis program. - The Atlas carbons were lower in price - $0.25/lb. for the 20 x 40 mesh carbon and $0.17/lb for 40 x 140 mesh carbon than the other candidate granular carbons, which sold at about $0.29/lb. Attrition loss of the Atlas 40 x 140, which is an important considera- tion for the FACET carbon, was only slightly greater than that of other granular carbons. ADDITIONAL STUDIES Lime and Lime-Carbon Treatment After the selection of the lime-carbon treatment sequence of total mill effluent as one of the sequences to be piloted (see Section V), further experimental work was done to determine more closely the effect of different combinations of lime and carbon dosage on the resultant color and TOG levels. As described in Appendix B, a dosage of 5000 mg/1 CaO was used in earlier laboratory studies, as compared to the dosage of 1000-2000 mg/1 CaO that was later found to be satisfactory in plant tests at Interstate Paper Co. at Riceboro, Ga. and Continental Can Corp. at Hodge, La., (14,18). 21 ------- Table 7 shows results obtained at different lime dosages in one particu- lar effluent sample (Pensacola No. 2 Mill effluent). Table 8 shows the effect of different combinations of dosages of lime and carbon, where the sample was only filtered after lime treatment and before carbon treatment, i.e. no carbonation was used to remove soluble lime. Table 7. LIME TREATMENT OF UNBLEACHED KRAFT EFFLUENT CaO dosage, mg/1 Raw 200 400 800 1200 Color, CU 1585 1375 584 277 257 TOC, mg/1 245 204 181 146 136 Table 8. LIME-CARBON TREATMENT OF UNBLEACHED KRAFT EFFLUENT Dosage, mg/1 CaO 0 0 200 200 400 400 800 800 1200 1200 Carbon3 0 5000 1000 5000 1000 5000 1000 5000 1000 5000 Darco S-51 Color, CU 1675 413 438 177 61 6 56 8 21 1 CaO mg/lb 14 14 154 98 539 504 832 714 PH ,7 .1 10. 10. 11.8 11.8 12.3 12.2 12.4 12.4 12.7 12.7 Concentration found in filtered treated effluent 22 ------- Results in Table 7 indicate that 63% of the color was removed at a dosage of 400 mg/1 CaO, and that removal in this case leveled off at about 82%. Table 8 shows, at very similar initial concentrations, that 1 g/1 activated carbon is sufficient to reduce the color below 100 CU if 400 mg/1 CaO had been used. Higher lime dosages were un- necessary in conjunction with 1 g/1 carbon, while higher carbon dosages were unnecessary after 400 mg/1 CaO or higher, and insufficient at lower lime dosages. Table 7 indicates that higher lime dosages alone were also unable to reduce the color below 100 CU in this effluent. These results indicated generally that an optimum combination lime and carbon treatment might be developed in the pilot plant requiring substantially lower lime dosages than are employed if only lime is used for color removal. Color Changes of Raw and Treated Effluents In the course of laboratory work on the lime-bio treatment sequence, color increases were observed when lime treated effluent was bio-oxidized. The first two lines in Table 9 summarize these results. It was believed that these increases were rather significant, particularly since color increases were also observed at that time in the natural aeration bio- oxidation ponds of the Riceboro, Ga., mill of Interstate Paper Co. (18), and later at Continental Can Corp. at Hodge, La., (14) where lime treat- ments before bio-oxidation are being operated on a commercial scale. Subsequent laboratory work shown in lines 3-7 of Table 9 indicated that color of lime treated effluent usually increases upon aging by about 50%, with the exception of one out of 6 samples which showed a 28% decrease. At the same time it was observed that carbon treated effluent tends to decrease in color, while raw effluent exhibits some instability, showing a decrease first and then an increase. Table 9 (last line) also shows results observed during pilot plant bio- oxidation of unbleached effluent for comparison, indicating a 15% color reduction during bio-oxidation. These results indicate a lower degree of stability of color than expected, particularly since the color of kraft mill effluent is generally held to be non-degradable in the usual aerated lagoon treatment system. Partic- ularly in the case of lime treatment, treatment results should be assess- ed giving full consideration to potential color increases during subsequent bio-oxidation or storage. 23 ------- Table 9. COLOR CHANGE OF KRAFT EFFLUENT to Effluent No. of Samples Caustic bleach 2 Unbl. total Unbl Unbl . total . total Unbl. total Unbl Unbl Unbl Unbl Unbl a b . total . total . total . total . total mill 3 mill 2 mill 1 mill 1 mill 1 mill 1 mill 1 mill 1 mill0 Initial color, CU Final color, CU lime tr. carbon tr. raw bio-oxid. aged 67 130 46 104 166 - 242 68 100 184 ... 133 184 ... 313 184 - 268 119 - - 60 1180 - 1136 1305 950 810 Aging time, days - 0.21 6 5 4 4 20 4 11 - Sample size, Sample gal. initial 7.0 7.0 40 7.0 40 7.0 40 7.0 0.26 7.0 0.26a 7.0 0.26 400 10.1 400 b 10.2 PH final - - - - - 9.0 8.8 - 8.2 8.6 kept dark ft (\C\ rt. «1 1 fl *-IWW« 1 rt c pilot plant bio-oxidation ------- Adsorption of Specific Compounds Each of the effluents studied contains a mixture of organic chemcial compounds. Usually, a few or none of these compounds are identified in effluent treatment work beyond being reported as BOD, color, COD, or, in this study, primarily as color and TOC. The adsorption of organic sub- stances from mixed solution is a complex phenomenon, possibly owing to the effect of one substance on the adsorption of another. This can manifest itself in preferential adsorption of one substance over others, non-adsorption if a substance is only weakly adsorbed at best, or the displacement of a weakly adsorbed by a strongly adsorbed substance. This selective or preferential adsorption must be kept in mind in the evaluation of equilibrium adsorption data. Concentrations of some specific compounds have been determined in a number of tests in this study for two reasons. One is that, from a commercial applications standpoint, this information may indicate opportunities for recovery of by-products by adsorption - desorption cycles with activated carbon. The return from the scale of such by- products could be credited against effluent treatment costs. The turpentine decanter underflow, which was studied in the laboratory for removal of specific compounds, may be promising in this respect due to its content of turpentine fractions and other organics. Figure 4 pre- sents the adsorption isotherm of loading vs. concentration for carbon treatment of turpentine underflow. Also given is an adjusted isotherm with methanol subtracted from the TOC content, an isotherm for acetone, and an isotherm for dimethyl sulfide plus dimethyl disulfide. These compounds were identified by using gas-liquid chromatography. In this experiment, methanol was found not to be adsorbed and about 207o of TOC was not identified. The initial conclusion based on overall TOC ad- sorption in this example would have been that this carbon is ineffective. However, isotherms for specific compounds and the methanol-free TOC would suggest that this carbon is reasonably efficient for removing components other than methanol, thus potentially yielding a deodorized water containing some methanol that could be reused in the mill. This phenomenon has meanwhile been further explored in a separate study and a patent application has been filed. Adsorption Rates During the course of the laboratory studies, a limited amount of time was devoted to an examination of the factors affecting adsorption rates, i.e. the rate at which impurities defined as TOC or color are removed from effluent by activated carbon. 25 ------- 5000 fcO I o 55 H 1000 500 TOTAL, w/o MeOH 50 100 500 1000 5000 10000 TOTAL ORGANIC CARBON, mg/1 Figure 4. Carbon adsorption isotherms for specific compounds in turpentine underflow 26 ------- A major factor affecting adsorption rate is particle size. If one assumes that the equilibrium loading is not affected by particle size, then the smaller particles should adsorb at a higher rate because the distance a solute must travel within the porous structure of the carbon is decreased. Weber concluded (15) that rate is proportional to the inverse of the diameter squared, or l/d2. However, in another study, Weber reports for 50% ABS removal from solution by two different sizes (350 and 670 micron) that rates increased proportional to 1/d . Melter (11) reports rate proportional to 1/d2, while Berger (10) found on pulping effluents that rates were proportional to In this laboratory study, rates were evaluated in stirred batch tests where carbons of four different particle size ranges were added to unbleached kraft total mill effluent from Mill A at dosages of 1, 5 and 10 g/1. Results indicate a proportionality of TOG adsorption rates to the inverse of diameter to some power, with the power varying between 1.6 and 3.0 depending on the degree of loading at which rates were evaluated, i.e. power of 1.6 at 20%, 2.0 at 30% and 3.0 at 50% of load- ing. In other words, the more the carbon is loaded the more of a dis- advantage the larger particle size presents compared to the smaller size. However, this information was not later borne out fully by the comparison of rates in the adsorption columns and the FACET systems in the pilot plant. In these pilot plant runs the rate vs particle size relationship or TOG removal appeared to be closer to proportionality to 1/d2 at loadings normally above 50% of isotherm equilibrium loadings, and perhaps even closer to 1/d ' if rates had been compared at strictly equal loading on the carbon. The effect of dosage on rate would be expected to cancel out if rate is expressed as mgTOC/g carbon x hr. However, rates at higher dosages were actually found to be lower. 27 ------- SECTION V SELECTION OF EFFLUENT TREATMENT SYSTEMS TO BE EVALUATED IN THE PILOT PLANT INTRODUCTION A laboratory program (see Section IV) was carried out to determine the properties of various streams in three St. Regis kraft pulp and paper mills and to provide a basis for selecting which streams have the greatest promise for reuse when treated. Treatments of the various in-plant and final effluent streams were evaluated in the laboratory. These treatments included bio-oxidation, lime treatment, and activated carbon adsorption. Combinations of these treatments were also evaluated to permit selection of treatment sequences to be used in the pilot plant. As a further input to selection of treatment sequences for the pilot plant, order-of-magnitude cost estimates were made for treatment of effluents from two St. Regis pulp and paper mills with the candidate treatment se- quences. The selection of effluents to be treated and of effluent treat- ment sequences, as well as the result of the cost estimates, are presented in this section. These estimates were based on total mill effluents from kraft pulp mills not having bleaching operations because the chloride of the bleach water would not be removed in the treatments and would make it unsuitable for reuse as general mill feed water. The cost estimates included the follow- ing combinations of primary clarification (primary), lime treatment (lime), primary clarification combined with lime treatment (primary/lime), (as carried out on a mill scale (14,18)), biological oxidation (bio), and carbon adsorption (carbon): primary, bio primary, bio, carbon primary, carbon primary, lime primary, lime, carbon primary/lime, carbon The carbon treatments included use of powdered carbon without regeneration of the carbon, granular carbon with regeneration, and a new counter-current continuous process called FACET in which an intermediate size carbon is used. 28 ------- These estimates indicated that the use of powdered carbon was not practical nor economical and that both column adsorption using granular carbon and slurry contacting using the FACET process should be investigated in the pilot plant. The treatment sequences selected for study in the pilot plant to give reusable quality water were: primary/lime, carbon primary, bio, carbon primary, carbon (The actual execution of the primary/lime, carbon sequence in the pilot plant used a holding pond to smooth out concentration variations and inadvertently acted as a primary clarifier before lime treatment.) In each sequence both granular carbon column and FACET slurry carbon pro- cesses would be used. The cost estimates indicated that the total fixed and direct operating costs for these sequences would range from $0.16 to $0.29 per kgal in commercial installations. In view of the demonstrated success and reliability of carbon column adsorption process for treatment of municipal wastewater, compared with the unproven status of the FACET process, it was decided that the major emphasis in the pilot plant would be on the carbon column process. SELECTION OF EFFLUENTS TO BE TREATED Selection Procedure The selection procedure consisted of eliminating the less desirable effluent systems on the basis of quantitative comparison and judgement against different criteria which were derived from the project objectives. The effluent systems considered in this study are listed below, including a brief argument for the acceptance or rejection of each. Total Mill Effluent Without Bleach Effluent One of the effluent systems considered was the treatment of total mill effluent exclusive of bleach plant effluent, by carbon or by carbon following bio-oxidation and/or lime treatment. This was the approach selected for the pilot plant. Production of a general mill feed water from an unbleached kraft total mill effluent appeared to be technically feasible. This was indicated by laboratory results of this program and by results reported by others. Activated carbon plays an important role in various treatment sequences capable of producing reusable water. 29 ------- Economically, the production of reusable water appeared to be compet- itive when compared with treatment and disposal at anticipated more- stringent near-future disposal standards. Total Mill Effluent Including Bleach Effluent Treatment of total mill effluent, including bleach plant effluents, by carbon or by carbon sequentially combined with bio-oxidation and/or lime treatment had also been considered and had been rejected. The inclusion of bleach plant effluents (from chlorine-based bleach systems) in the total mill effluent was considered undesirable because the chlo- ride ion would build up in the recycled water. Laboratory investigations showed that the chlorine content of bleach effluents is essentially not reduced by activated carbon, or by any of the treatment sequences consid- ered in this program. If a chloride removal process such as ion exchange or reverse osmosis were to be considered at a later time, separate treat- ment of bleach plant effluent would probably be the approach to take. Chloride removal could then follow lime or lime plus activated carbon treatment of bleach plant effluent. In-Plant Effluents Treatment of in-plant effluents by activated carbon in a separate unit for each effluent had been considered as the major alternative to total mill effluent treatment. It had been found that a substantial part of the reuse of in-plant effluents that can ultimately be achieved does not, or not yet, require activated carbon treatment. Moreover, the favorable cost estimated for the total mill effluent treatment approach, and the lesser complexity of that approach in terms of actual operation and of its development, argued against pursuing treatment of individual effluents under this program. PRELIMINARY SELECTION OF TREATMENT SEQUENCES AND CARBON ADSORPTION SYSTEMS Selection Procedure The preliminary selection of treatment sequences to be studied in the pilot plant proceeded along the following steps: (1) Identification of potential sequences. (2) Determination of reusable water quality requirements. (3) Laboratory evaluation of the potential sequences with effluents from the three southern mills of St. Regis and from one other mill which uses lime treatment of the total mill effluent. 30 ------- (4) Preliminary selection of sequences based on laboratory results and judgement of comparative economics. (5) Consideration of variants within these sequences, particularly with regard to the carbon adsorption step. Identification of Potential Treatment Sequences The following treatment sequences were considered to have sufficient merit to warrant study in the laboratory: (1) Primary clarification followed by activated carbon adsorption; (2) Primary clarification, lime precipitation, activated carbon adsorption; (3) Primary clarification, biological oxidation, activated carbon adsorption; (4) Primary clarification, lime precipitation, biological oxidation, activated carbon adsorption; (5) Primary clarification, biological oxidation, lime precipitation, activated carbon adsorption. Determination of Reusable Water Quality Requirements Water quality requirements are usually expressed in terms of a numerical maximum or range of a number of properties such as color, dissolved solids, etc. Table 10 presents such numerical values based on several references. Although the TAPPI values are used as guidelines for new mills, their basis is a survey of industry practice dating back to 1948, rather than a specific determination of requirements. A TAPPI committee was set up at the 1970 TAPPI Air and Water Conference for the determination of actual water quality requirements at various points of use in the kraft process. The absence of BOD specifications in the TAPPI standards in Table 10 is noteworthy. 31 ------- Table 10. RANGES OF DESIRED PROPERTIES OF PROCESS WATER FOR UNBLEACHED KRAFT MANUFACTURE St. Regis TAPPI TAPPI Property proposal 5/1/69* E601 S-53D £603 S-49C Turbidity, mg/1 5-25 100 25 Color, CU 10-80 100 5 pH 6.5-8.0 Total alkalinity, mg/1 20-150 150 75 Hardness, mg/1 CaCOo 5-200 200 100 Dissolved solids, mg/1 50-500 500 250 Chloride, mg/1 10-150 . 200 75 Fe, mg/1 0.5 1 0.1 Mn, mg/1 0.3 0.5 0.05 COD, mg/1 0-12 BOD, mg/1 0-5 a Based on NCASI Bulletin 203 (48) b Specification for chemical composition of process water for kraft paper manufacture; tentative standard, corrected August, 1953. c Specification for chemical composition of process water for manufacture of soda and sulfate pulps; tentative standard, December, 1949. Early in this program an effort was made within St. Regis to collect water quality standards for the various points in the process of making pulp and paper. The results of this effort, compiled in table form with accompany- ing comments, are included in this report as Appendix A. As a result of this effort, it was realized that the tentative TAPPI stan- dards in Table 10 may well be over-specifying some of the water quality requirements of an unbleached kraft mill, except boiler feed water which requires further treatment. Water of 100 CU and 50 mg/1 TOG (total organic carbon), equivalent to about 25-100 mg/1 BOD, was tentatively considered to be adequate as a general mill feed water (later changed to 100 CU and 100 mg/1 TOC). It was considered possible that problems such as slime growth might arise at some of the water use points due to specific organic compounds or due to the general BOD. However, such problems as might arise will most probably have an economical remedy, such as the addition of slimicides or 32 ------- a downward adjustment of the TOG specification for part of the mill feed water. Thus, the tentative targets of 100 CU and 50 mg/1 TOG were considered a reasonable basis to proceed on. Other water properties were not considered in the selection of treatment sequences, although their ultimate importance is recognized. Consider- ation of color and TOG was believed to be sufficient, since removal of TOG and color is the main purpose of the treatment systems under devel- opment. Differences in efficiency and cost between the various systems are expected in this area and are of prime importance. Consideration of possible differences in removal of dissolved inorganics or in recycle water temperature was deferred. Laboratory Evaluation and Preliminary Selection of Treatment Sequences The laboratory evaluation of the above identified treatment sequences has been discussed in Section IV and has also been reported (15). Table 11 and 12 repeat the data on color and TOG removal presented in Tables 5 and 6 and additionally present data on dosages for two- and three-stage treatment with powdered carbon estimated from isotherm data by the procedure described in Appendix B. Carbon treatment alone is seen to be capable of achieving the tentative standards of 100 CU and 50 mg/1 TOG. The carbon dosages for both color and TOG removal are about equal when treating Mill A effluent. For Mill C effluent, substantially higher carbon dosages are required with TOG being the dosage determining factor. The use of lime or bio pretreatments results in substantial reductions in the carbon dosage in all cases. The bio treatment is, as expected, more effective for removing TOG than for removing color; therefore, color removal becomes the dosage determining factor after bio treatment (unless a partial bio treatment is used). Lime pretreatment, in contrast, is more effective for color removal. Thus, TOG removal after lime treat- ment becomes the dosage determining factor. (Caution is indicated, how- ever, by color increases observed in laboratory biological oxidations following lime treatment, as discussed in Section IV.) When comparing the carbon dosages required to reach 100 CU and 50 mg/1 TOG in a three-step treatment (lime-bio-carbon and bio-lime-carbon) with dosages required in the two-step lime-carbon treatment, it appears that their is little or no carbon dosage advantage in the three-step treat- ments. Since volumetric flow through a treatment system is a major cost determinant, a third treatment step cannot be justified on the basis of 33 ------- Table 11. EFFECT OF TREATMENT SEQUENCES ON REMOVAL OF COLOR FROM TOTAL MILL EFFLUENTS C Treatment with carbon with lg/1 carbon •a Treatment Mill A Carbon Bio Carbon Bio Lime Carbon Lime Carbon Lime Bio Carbon Mill C Carbon Bio (at mill) Carbon Mill C Lime (at mill) Carbon Lime (at mill) Bio Carbon Initial color, CU 948 833 570 833 570 45 1,380 70 785 64 185 4,590 4,590 2,025 525e - 525e 479 Final color.CU - 570 - 570 45 - 70 - 64 185 - - 2,025 - 525e e 525 479 - Initial pH 8.6 7.5 7.4 7.2 7.5 10.1 7.5 7.2 7.5 color, CU 370 180 10 8 10 2,500 700 90 190 7 10 Removal 61 68 78 87 95 46 66 83 60 Loading CU/g 580 390 35 62 175 2,100 1,300 430 290 dosage to 100 CU,g/l 123 stage stages stages 3.4 0.91 0.58 1.6 0.61 0.48 d - - , d 0.10 17.3 2.3 l.io 8.4 1.77 1.02 0.80 0.30 0.24 5.30f 1.30 0.74 All samples of water were filtered before treatment to remove suspended solids (simulating primary , clarification). Four different samples of effluent were used for five treatment sequences. Westvaco Aqua Nuchar. d No carbon needed to reach 100 CU. e High color not considered representative of normal mill performance. f Poor adsorption indicated by steep isotherm. ------- Ui Table 12. EFFECTS OF TREATMENT SEQUENCES ON REMOVAL OF TOG FROM TOTAL MILL EFFLUENTS Treatment with carbon Treatment3 Mill A Carbon Bio Carbon Bio Lime Carbon Lime Carbon Lime Bio Carbon Mill C Carbon Bio (at mill) Carbon Initial TOG, mg/1 138 111 66 111 66 - 106 67 106 67 46 713 713 183 Final TOG, mg/1 66 - 66 - - 67 - 67 46 - 183 - Initial PH 8.6 7.5 7.4 7.2 7.2 10.1 7.5 with TOG, mg/1 75 30 23 38 23 570 90 lg/1 carbon 7c Removal 44 55 - 43 50 20 51 Loading, mg/TOC/g 53 36 _ 30 23 140 90 dosage 1 stage 2.70 0.20 _ 0.19 - 29.0 4.40 to 50/mg/l TOC,g/l 2 3 stages stages 0.86 0.57 - _ - _ - - 5.20 3.20 1.20 0.73 All samples of water were filtered before treatment to remove suspended solids (simulating primary clarification). Westvaco Aqua Nuchar. ------- marginal contaminant load reductions for the remaining two steps. This comparison of treatment sequences for total mill effluent, when the goal is reduction to less than 100 CU and less than 50 mg/1 TOC, serves to eliminate the three-step sequences from further consideration. Goals of different color or TOC levels in the treated water will affect the comparison of treatment sequences. Changes in the concentration of color and TOC in the water being treated will also affect the eco- nomic comparison. Because of the variability of effluents from mill to mill, the effect of color and TOC of the feed water on capital and operating costs must be included in the economic comparison of treat- ment sequences. Variants in Carbon Treatment Systems The conventional choice to be made for carbon treatment is between (a) a granular carbon system including adsorption and carbon regenera- tion; or (b) a powdered carbon system including adsorption and carbon disposal after single use. A granular carbon adsorption system conventionally uses the carbon in packed columns, the liquid being purified while flowing through the bed of carbon. In a powdered carbon system, the carbon is contacted with the liquid in agitated tanks, followed by clarifiers to remove the carbon from the water. Clarification must usually be assisted by the use of flocculants. Thermal regeneration in a multiple-hearth furnace is the most common method for regenerating spent granular carbon. The development of thermal regeneration of spent powdered carbon has only recently been vigorously pursued (20,21,49,50,51). Disadvantages in regeneration of powdered carbon, compared to granular carbon, lie in the difficulty of dewatering the carbon before regeneration and in high carbon losses during thermal regeneration. No commercial system was in existence for regenerating powdered carbon, hence powdered carbon was considered as a single-use material in this evaluation. The purchased cost of powdered carbon is substantially lower than that of granular carbon $0.085/lb. vs. $0.25-0.30(at the time of this study). The conventional choice between the granular and powdered carbon systems for a specific application is based on an evaluation of several technical and economic factors. 36 ------- Usually, the use of powdered carbon is favored when the carbon consumption is low due to low liquid volume and/or low dosage require- ments (e.g. due to low contaminant concentration). Granular carbon would be favored where countercurrent adsorption is advantageous, and where the separation of a powdered carbon from the liquid being purified would present a problem. An economic evaluation must take into consid- eration the capital and operating costs associated with single-use powdered or granular carbon and with regenerative use of granular car- bon. The latter would require the higher capital investment, but operating cost for make-up carbon would be lower. As a third choice, St. Regis suggested a counter-current contacting system that uses a fine activated carbon of a size intermediate between powdered and standard granular carbon. A preliminary engineering assessment of this system (dubbed FACET) indicated that it combines some of the advantages of both conventional systems. The FACET system is shown schematically in Figure 6 (Section VI) as a three-stage counter- current contacting system. A patent (41) has since been issued on this process. Carbon size and concentration are chosen such that contact time require- ments are short, while the size is large enough to provide fact settling without the assistance of flocculating agents. Therefore, FACET permits use of simple stirred tanks as contactor-settlers and avoids the need for preclarification of the water commonly required in granular adsorp- tion systems to prevent column pressure build-up and avoids the need for frequent back-washing. Present indications are that a suitable particle-size range for FACET is about 100 to 250 microns (140 to 60 mesh). The FACET particle size is believed to be large enough to permit use of a standard multiple-hearth furnace for regeneration. The FACET process is discussed further in Section VI. The choice among the three carbon systems as applied in this project is influenced by the proposed production of activated carbon at the mill for the treatment of mill effluent. Such mill-site production of carbon makes all three types of carbon available at a substantially lower cost and also at a smaller cost differential between granular, FACET, and powdered carbon. Moreover, another dimension is added to the economic evaluation since carbon cost depends not only on type of carbon but on production volume. The latter is very materially dependent on whether the carbon is discarded after use or is regenerated. As a result of this preliminary evaluation of alternative carbon systems, it appeared that all three systems should be further evaluated in the economic estimate. 37 ------- ORDER OF MAGNITUDE COST ESTIMATES Bases of Cost Estimates Costs of equipment and overall plant costs for primary clarification, bio-oxidation, lime treatment, and carbon adsorption were obtained primarily from references 17 and 23-34. Additional sources of cost data were Infilco, Eimco, and St. Regis' internal sources. Most of the reference data on activated carbon systems are based on municipal effluent treatment. The estimates are of the order-of-magnitude type which might have inaccuracies of 20% to 40%. Allowances were made in the capital costs for installation, engineering, and contingencies. Most capital costs were based on overall plant costs from curves or from similar plants with the cost adjusted to the desired flow rate by suitable values of capacity-cost exponents. St. Regis Mill A total mill effluent, on which the laboratory tests were based, had a flow rate of 38 mgd, which included about 20 mgd of cooling water. Even though the flow without the cooling-water dilution would have been only 18 mgd and would have contained about twice the measured color and TOC concentrations, cost estimates were based on the 38 mgd flow because the lab data were based on this diluted flow. The average composition of this stream was 140 mg/1 TOC, 1000 CU, 130 mg/1 BOD, and pH 8.6. The average composition of the St. Regis Mill C total mill effluent during the lab program was 640 mg/1 TOC, 4450 CU, and pH 9.9. The color was abnormally high because of mill upsets during this period. The average flow rate was 15.3 mgd. A limited dosage lime process (1000 mg/1 CaO) similar to that used at Interstate Paper Company's Riceboro Mill (5,18) was used in these estimates. The amounts of TOC and color reduction by lime treatment that were used in these estimates are those obtained in the lab pro- gram with a dosage of 5000 mg/1 CaO which gives about the same color reduction as a dosage of 1000 ppm. The capital costs for lime treat- ment include lime slaking, reactor-clarifier, carbonation, CaCOo clarifier, and sludge filters, but do not include any portion of the capital and operating costs of lime calcination. The TOC reductions by bio-oxidation used in the estimates were 140 to 66 mg/1 TOC for Mill A and 640 to 140 for Mill C, which were the reductions obtained in the laboratory experiments. It was assumed that the bio-oxidation would be carried out in mechanically aerated basins with a retention time of 10 days. 38 ------- Three carbon treatment processes were considered: powdered carbon (about 10 micron average particle size), fine-carbon FACET system (100 to 250 micron particle size to 140 x 60 mesh), and granular carbon which is used in a fixed bed. Carbon dosage estimates were based on laboratory results with Westvaco Aqua Nuchar powdered carbon. The dosage used in these estimates was that required to re- duce the color to 100 color units and TOG to 50 mg/1 plus 50% excess carbon to allow for incomplete equilibration. Water with these values of color and TOG is considered satisfactory for general plant reuse. The influence of the price of carbon on treatment costs was determined by assuming prices of $0.02, $0.085, and $0.27 per pound of carbon. The price of commercially available powdered carbon was $0.085/lb and that for granular carbon was $0.27/lb at the time of these estimates. Charges related to fixed capital are included in operating costs on the basis of 16-year, straight line depreciation (6.257., of TCI/year), plus a total of 12% of TCI/year for repair, maintenance, taxes, and interest charges. General administrative overhead (GAO) was assumed to be 75% of labor costs plus 1.5% of TCI/year. Pertinent operating conditions are given in Tables 13, 14, and 15 for each estimate, along with results of the cost estimates. Results of Estimates Overview - Tables 13 and 14 present a summary of the conditions and cost data of the estimates prepared. The presentation in Table 15 is less detailed and shows a wider range of combinations including those of Tables 13 and 14. Estimates A-2, A-ll, and A-3 in Table 13 represent the three treat- ment sequences under consideration, using granular carbon adsorption. The effect of capital and operating costs on treatment system selec- tion will be discussed first. The influence of carbon cost on capital and operating costs and indirectly on treatment system selection will be discussed subsequently. Capital Investment - In terms of capital investment, the primary/lime-carbon system (A-ll) represents the lowest cost, mainly because of the significantly lower cost of the carbon adsorption and regeneration portion of the total system. The primary-bio-carbon system (A-3) requires the highest capital investment. 39 ------- Table 13. Estimated No. Treatments ORDER- OF-MAGNITUDE COST ESTIMATES FOR VARIOUS TREATMENTS OF TOTAL MILL EFFLUENT FROM MILL A; EFFLUENT: 140 mg/1 TOG, 1000 CU Effluent Flow, mgd. Carbon treatment, type Dosage, Ib/kgal , Carbon cone'n.,lb/ft Retention, hours, carbon water Carbon regeneration Final filter Treated water comp'n: TOC, mg/1 Color,CU Capital costs. $ million Price of carbon, $/lb 0.02 Primary 2.5 Bio-oxidation Primary/lime treatment Filters Carbon contactors Carbon dewatering Carbon drying & regeneration Carbon in system Total cost of investment(TCI) Operating cost. $/kgal Amortization @6.25% TCI/yr 0.040 R&M, tax, int.@127. TCI/yr 0.077 Labor + GAO 0.026 Elec.(0.5<:kwh)& gas(35<:MMBTU) 0.010 Coagulants 0.005 Make-up carbon (5%) 0.005 Lime (1000 mg/lCaO)@ $16/t - A-2 Primary- granular carbon Mill A 38 gran. 5.0 25 950 1 yes no 50 100 0.085 2.5 6.3 0.6 9.4 0.042 0.081 0.026 0.010 0.005 0.021 0.27 2.5 6.3 2.0 10.8 0.049 0.094 0.026 0.010 0.005 0.067 A-ll Primary/ lime- granular carbon Mill A 38 gran. 1.42 25 950 1 yes no 50 100 0.02 0.085 0.27 4.0 4.0 2.8 0.031 0.032 0.060 0.061 0.034 0.034 0.004 0.004 0.001 0.004 0.067 0.067 4.0 2.8 1.1 7.9 0.036 0.068 0.034 0.004 0.013 0.067 A-3 Primary-bio- granular carbon 0.02 2.5 3.6 h j 0.1 11.6 0.052 0.100 0.036 0.009 0.010 0.004 Mill A 38 gran. 4.3 25 950 1 yes no 50 100 0.085 2.5 3.6 i5"4 ) 0.5 12.0 0.054 0.104 0.036 0.009 0.010 0.019 0.27 2.5 3.6 h ) 1.7 13.2 0.059 0.114 0.036 0.009 0.010 0.059 Total operating costs 0.163 0.185 0.251 0.197 0.202 0.222 0.211 0.232 0.287 ------- Table 14. ORDER-OF-MAGNITUDE COST ESTIMATES FOR VARIOUS TREATMENTS OF TOTAL MILL EFFLUENTS FROM MILLS A AND B Mill A: 140 mg/1 TOC, 1000 CU,; Mill B: 640 mg/1 TOC, 4450 CU Estimate No. Treatments Effluent Flow, mgd. Carbon treatment, type Dosage, Ib/kgal _ Carbon conc'n.,Ib/ft Retention, hours, carbon water Carbon regeneration Final filter Treated water comp'n: TOC, mg/1 Color,CU Capital costs, $ million Price of carbon, $/lb Primary Bio-oxidation Primary/lime treatment Filters Carbon contactors Carbon dewatering Carbon drying & regeneration Carbon in system Total cost of investment (TCI) Operating cost, $/kgal Amortization (36.25% TCI/yr R&M, tax, int.(§12% TCI/yr Labor + GAO Elec.(0.5ckwh)& gas(35cMMBTU) Coagulants Make-up carbon* Lime (1000 mg/1 CaO)@ $16/t Total operating costs A-6 Primary- powd. carbon Mill A 38 powd. 13 0.1 1.5 1 no sand 50 100 0.02 0.085 2.5 A-10 Primary/lime- powd. carbon Mill A 38 powd. 6.9 0.05 2 2 no sand 50 100 C-2 Primary- gran. carbon Mill B 15.3 gran. 12.5 25 950 1.5 yes no 50 100 0.02 0.085 0.02 1.5 0.085 1.5 C-3 Primary- powd. carbon Mill B 15.3 powd. 36 0.3 1.5 1 no sand 50 100 0.02 0.085 1.5 - 1.7 0.8 0.5 1.2 negl. 6.7 0.030 0.058 0.035 0.047 - 0.013 _ - 1.7 0.8 0.5 1.2 0.2 6.9 0.031 0.059 0.035 0.047 - 0.055 _ 1 1 0 ne 6 0.029 0.056 0.036 0.020 0.010 0.260 _ - .7 .5 .8 gl. .5 0.029 0.056 0.036 0.020 0.010 1.100 - 4. 1. 0. 0. neg 6. 0.030 0.058 0.025 0.001 0.005 0.138 0.067 0 7 5 5 1. 7 0.030 0.058 0.025 0.001 0.005 0.590 0.067 - _ I" 0.1 5.0 0.056 0.104 0.031 0.010 0.010 0.012 - - _ I" 0.5 5.4 0.060 0.112 0.031 0.010 0.010 0.053 - - 0. 1. 0. neg 4. 0.046 0.089 0.046 0.020 0.030 '0.720 - 8 0 8 1. 2 0.046 0.089 0.046 0.020 0.030 3.05 _ 0.183 0.227 0.411 1.25 0.324 0.776 0.223 0.276 0.95 3.28 (5% for A-8, C-2; 100% for A-6, A-10, C-3) ------- Table 15. SUMMARY OF CONDITIONS OF TREATMENT AND ESTIMATED CAPITAL AND OPERATING COSTS FOR VARIOUS TREATMENTS OF TOTAL MILL EFFLUENT FROM MILLS A AND B NJ Cost Estimate Primary Clarif n Bio- oxid'n Lime Activated carbon 3wd.fine8 erar Final filtra- tion Mill A total mill effluent 1 2 3 4 5 6 7 7A 8 9 10 11 x X X X X X X X X X X X X R R R NR 75,R 75,NR 150,R x x X NR R Mill C total mill effluent before oxidation pond 1 2 3 x X X R NR x a b c d e f g Average particle size in microns Based on minimum price of carbon - $0.02/lb Based on intermediate price of carbon - $0.085/lb Same as c, but incl. $0.06/kgal credit for reusable water Product water contains 570 CU Product water contains 63 mg/1 TOC Product water contains 140 mg/1 TOC, 2000 CU Capital cost $mi11ion Operating Cost $/kgalb $/kgalc $/kgald _ - _ - - X X X X - X - 6.1 8.9 11.6 5.5 8.3 6.5 7.4 6.2 6.7 4.0 6.7 6.9 0.12e 0.16 0.21 0.17f 0.23 0.41 0.19 0.37 0.18, •p 0.14r 0.32 0.20 0.12e 0.19 0.23 0.17t 0.23 1.25 0.24 1.20 0.23, T 0.14 0.78 0.20 0.12e 0.13 0.17 0.17f 0.17 1.19 0.18 1.14 0.17, T 0.14 0.72 0.14 3.5 5.0 4.2 0.16» 0.22 0.95 0.166 0.28 3.28 0.16 0.22 3.22 R - regenerated NR - not regenerated ------- Compared to the primary-carbon system, the addition of a biological treatment system requires a substantial added investment while not substantially reducing the investment requirement in the carbon part of the treatment sequence. Thus, when aiming at a reuse treatment system for new mill design, the primary/lime-carbon and the primary- carbon sequences appeared to be the logical first and second choices for further development, based on capital cost estimates. However, when aiming at a reuse treatment sequence adaptable as a modification for an existing mill with a biological oxidation system, the choice would appear to become much more complex. It would in- clude the following options: continued operation of existing facili- ties, with addition of a carbon system; shut-down of biological oxida- tion, with addition of a carbon system; shut-down of biological oxida- tion, addition of a lime treatment system including modification of the existing primary treatment, and addition of a carbon system. Based on the limited cost detail on individual steps of the three treatment sequences presented in Table 13 the mill system modifica- tion most closely represented by the primary-bio-carbon sequence could represent the economical choice and thus appeared to merit further development. An elimination of any one of these three treatment sequences thus did not appear warranted on the basis of the capital investment comparison considering the objective of assuring broadest utility and applicability of successful development of new treatment systems to both new and existing mills. Operating Cost - Comparing operating costs, including fixed costs based on capital investment, at a carbon cost of $0.02 or $0.085/lb., the primary- carbon system represents the lowest cost system. Total operating costs for the primary/lime-carbon system are 15% above the primary- carbon system, and the primary-bio-carbon system is yet 10% higher in cost. These differences are within the accuracy of these estimates and at most indicate a possible cost trend. In the case of an exist- ing mill with a biological oxidation system, and where that system is already fully or partially depreciated, the primary-bio-carbon system would again tend to compare more favorably. Thus on the basis of operating cost, using mill-site produced carbon at up to about $0.085/lb.t no justification was seen to eliminate any of the sequences from further development under the objectives of this program. 43 ------- Carbon Cost - The fraction of total capital investment (TCI) represented by the carbon inventory merits closer inspection. It is seen to be an insignificant factor at a carbon cost of $0.02/lb., and a fairly low fraction (about 5%) of TCI at a carbon cost of $0.085/lb. However, at a carbon cost of $0.27/lb. (equivalent to purchased carbon), carbon inventory amounts to 13%-19% of TCI. Assuming successful development of mill-site carbon production at a cost of $0.02-.085/lb. carbon, system selection is essentially insensitive to the capital investment cost aspect of the carbon inventory. Carbon costs above the stated range tend to favor the primary/lime-carbon treatment sequence. Total operating costs, including fixed costs based on capital invest- ments, of all three sequences vary with carbon cost for two reasons: (a) make-up carbon represents a (variable) direct operating cost item; (b) the initial carbon charge (inventory) represents a (variable) capital investment cost item. However, the operating cost of the primary/lime-carbon system is relatively insensitive to carbon cost. Over the range of carbon prices examined ($0.02 to $0.27/lb.), the operating cost of the primary/lime-carbon system increases by only 15%, while the operating cost of the primary-carbon system increases by 54%. This makes the primary/lime carbon system the less costly system to operate with (purchased) carbon at $0.27/lb. The estimated operating costs of the primary/lime-carbon system, using purchased carbon at $0.27/lb., do not compare too unfavorably with the primary-carbon system using mill site-produced carbon at $0.085/lb. or less. The lower indicated investment cost of the primary/lime-carbon system using purchased carbon may offer suffi- cient reason in a specific case to give it serious consideration in comparison with systems offering lower operating costs. Carbon Adsorption Systems Table 14 brings out the effect of different carbon systems on capital and operating costs. It should be noted that the estimates in Table 14 were prepared only for two different carbon costs of $0.02 and $0.085/lb. The FACET system is represented by estimate A-8. It shows a moderate increase in direct operating costs over the granular carbon system A-2 (Table 13). This increase is due to higher estimated carbon dosage and correspondingly higher make-up requirements and to higher regen- erating utility costs. Comparing the capital cost estimates A-8 and A-2, the potentially significantly lower TCI of the FACET system is observed. This lower capital requirement would appear to make the 44 ------- FACET system an attractive alternative to the granular carbon system, if the lower costs can be substantiated on the basis of further development and subsequent cost estimates. The effect of using powdered carbon in the primary-carbon sequence with no regeneration of the carbon is indicated in estimate A-6, and may be compared with the corresponding granular carbon system estimate A-2. Operating costs for A-6 are higher, even at the lowest assumed carbon cost of $0.02/lb, due to higher carbon requirements in the absence of carbon regeneration. On the other hand, the TCI for A-6 is lower in the absence of investment costs for carbon regeneration. However, non-competitive operating costs of the powdered carbon system provided a basis for ruling it out from further development. A-10 is an estimate of the powdered-carbon version of the primary/lime- carbon sequence. A comparison with A-ll (primary/lime-granular carbon) shows no capital cost advantage of A-10, while operating costs are substantially higher than A-ll, even at the lowest carbon cost of $0.02/lb, and in spite of the relatively low carbon dosage required. Based on this and on the preceding comparison of A-6 with A-2, it appeared that further development of a powdered, non-regenerative carbon system was not warranted. Effluent Contaminant Concentration - Estimate C-2 in Table 14 is for primary-carbon treatment of Mill C total mill effluent which is more concentrated than that of Mill A but has a lower daily flow. In comparison with estimate A-2 in Table 13, C-2 shows a higher operating cost primarily because of higher fixed costs per unit volume of water and to a lesser extent because of increased carbon dosage. Estimates for treating Mill C effluent by the other two sequences were not prepared, since it was unlikely that the pilot plant system selection would have been influenced by these estimates. Estimate C-3 emphasizes the unattractiveness of a non-regenerative powdered carbon system as discussed in conjunction with A-6 and A-10. The higher carbon dosage required in C-3 makes it comparative- ly even less attractive. Comparison of Reuse Systems with Other Systems - The comparison of treatment systems in Table 15 indicates that the use of carbon treatment (either granular or FACET) might be only slightly more costly than biological oxidation, A-l, Table 15. Carbon treatment, and in some situations lime treatment, would provide a water suitable for reuse. In general, primary-bio treat- ment will not reduce the color sufficiently and primary-lime will not 45 ------- reduce the TOG, and perhaps color, sufficiently for reuse in the mill, and will not reduce BOD sufficiently for discharge to rivers in the future. The cost estimates indicate that lime treatment would be 15% higher than bio-oxidation. However, firm design and cost data were not available for the lime treatment process and there was some doubt as to the reliability of the values used in the esti- mation of the costs for lime treatment. FINAL SELECTION OF TREATMENT SEQUENCES AND CARBON ADSORPTION SYSTEMS The final selection of treatment sequences to be piloted was as follows: (1) Primary-carbon (2) Primary/lime-carbon (3) Primary-bio-carbon (The actual execution of the primary/lime carbon sequence in the pilot plant used a holding pond to smooth out concentration variations and inadvertently acted as a primary clarifier before lime treatment.) An examination of the economics and consideration of other factors did not provide a basis for eliminating any of the three treatment sequences considered. This program's orientation towards both new mills and exist- ing mills precluded the elimination from development, on economic grounds, of the primary-bio-carbon system, as discussed previously. Such elimina- tion, moreover, would require strong justification considering the fact that the primary-bio-carbon system has the advantage of minimum departure from established technology. In the final selection of carbon systems within the above treatment sequences, the single-use powdered carbon system was eliminated from further development because of its comparatively high cost. The remain- ing two systems; i.e., granular carbon and FACET, both indicated certain comparative economic and technical advantages. In terms of cost, the granular system indicated somewhat lower total operating cost, while the FACET system indicated lower capital investment. From a cost estimate reliability standpoint, the granular carbon system estimate was probably a more reliable one than that for the FACET system because of the avail- ability of substantial reference material for the granular carbon systems compared with the conceptual status of the FACET system. From the technical standpoint, the FACET system would appear to offer a better chance of avoiding certain problems that could develop in the operation of large-scale adsorption systems. Granular carbon columns tend to develop anaerobic biological activity. In a FACET system, it would be a relatively simple matter to control anaerobic activity or 46 ------- take advantage of aerobic activity by the injection of air. Another problem is often caused by suspended solids, leading to plugging and the need for frequent backwashing of granular carbon columns. Suspended solids are not expected to be a problem in the FACET system. Problems of this type can only be properly assessed in a pilot plant. Design data for a more reliable capital cost estimate for a commercial system also can only be obtained on the basis of pilot plant operating data. Hence, it was considered desirable to include both the granular carbon and FACET system in the plans for the pilot plant. 47 ------- SECTION VI PILOT PLANT DESCRIPTION AND OPERATION OBJECTIVES A pilot plant was designed and constructed that would meet the objectives of this program: to determine the technical feasibility of the selected treatment sequences involving carbon adsorption for removal of color and TOC from unbleached kraft mill effluent, to develop suitable conditions of operation, and to provide engineering design data for commerical systems, The pilot plant was designed so that effluent treatment by any of the following three sequences could be investigated: Primary-carbon Primary/lime-carbon (abbreviated lime-carbon) Primary-bio-carbon (abbreviated bio-carbon) Although the system selection detailed in Section V provided for inclusion of the primary/lime-carbon system, i.e., lime addition followed by simul- taneous removal of lime-color bodies and settleable solids in a primary clarifier, pilot plant plans were for practical reasons developed on the basis of lime treatment following primary clarification. (A holding pond, required to smooth out concentration variations , inadvertently acted as a primary clarifier before lime treatment.) The design incor- porated two carbon adsorption systems, i.e., downflow columns and FACET, and provided for simultaneous operation of both, resulting in a total of six treatment options. DESCRIPTION OF PILOT PLANT Location The pilot plant was located alongside the effluent ditch from the St. Regis Pensacola No. 2 Mill that normally produces 620 tons of unbleached pulp per day from a charge of 25% hardwood and 75% southern pine. This mill, built in 1948, uses batch digesters and produces mainly bag paper. The Pensacola mill effluent was selected for use in the pilot plant because this mill is typical of the more prevalent older pulp and paper mills. In addition, the Pensacola location affords the opportunity of testing a wide range of individual and combined effluents including bleach effluents from the No. 1 Mill. 48 ------- Basis for Design of Pilot Plant The pilot plant was designed to treat 30 gpm of mill effluent in the bio-oxidation and lime treatments, 15 gpm in the carbon columns, and 15 gpm in the FACET system. These flows accommodate the desire for large flows to provide reliable commercial design data and yet are small enough to keep equipment and operating costs within reasonable limits. The 30 gpm capacity of the bio and lime units gave sufficient flow for side-by-side evaluations of the two carbon systems with the same feed water. The pilot plant was designed for continuous operation seven days per week because intermittent operation would not have provided useful data for evaluation and designs. To provide operation round-the-clock, it was decided to use operators one shift per day for five days per week and to provide sufficient automatic controls and instrumentation to permit continuous operation without operator attendance during the remaining shifts, thus avoiding the cost and manpower scheduling prob- lems connected with round-the-clock operator attendance, particularly for pilot plants. An aeration basin (187,000 gal) was used for bio-oxidation because this method is the most commonly used method for bio-oxidation treat- ment of pulp mill effluents. A retention time of 4.3 days was used in designing the equilization-aeration basin. The depth of the earthen- dam rectangular aeration basin (8 ft) and the size of the floating aerators were selected to make the operation of the aeration basin typical of full-scale aeration basins. A large amount of flexibility was incorporated in the pilot plant to permit any treatment step to be by-passed or to permit simultaneous operation of as many as three of the six treatment options. Pilot Plant Equipment A flow diagram of the pilot plant is shown in Figure 5. Effluent is pumped from the ditch into a basin having a volume of 187,000 gallons and providing 4.3 days retention time when operating at 30 gpm. The basin is of compacted earth construction with the interior sloping sides coated with gunnite concrete to prevent water erosion of the sides. The basin serves to dampen the wide fluctuations of impurities, which may vary as much as 6-fold within a few days. One section of ^ the basin provides primary clarification at a rise rate of 100 gpd/ft at 30 gpm. Water from the basin passes through slots in a wooden wall placed across one corner of the basin. The wall consists of vertical boards with % to % inch spacing between boards. Only the lower one^or two feet of the slots are open, the remainder is covered with plastic to minimize water circulation from the aeration section. 49 ------- Ln O LIME 1 . r \\ \ '- + Y CO ,i ' • ,! " i .2 'l -' iiis^alll •- A i •! !! '•• ! ! Jl M " 1 llpiHr, . ik / M M ' 1 V U'1 i: !: Jr-^ j X \i~jy J-> SLUDGE <-} f clj Q^ ATER CARBONATOR pH ADJUST- MENT 5~\ r-Q-, (— 0~^ ^ 1 " 1 ' S^ \ UC^ ! // ^ \ \ / f- \ \_ 73k j ^T± FILTER -4 — 1 ^J ^ ^ ?, i i i i 1 1-. / k ^V ~s ^CT f / t I i i i i rpk ACTIVATED CARBON COLUMNS STORAGE TANK CLARIFICATION No.2 MILL EFFLUENT EQUILIBRATION OR BIO-OXIDATION BASIN SPENT CARBON ACTIVATED CARBON CONTACTORS FILTER SiORAGE TANK Figure 5. Effluent treatment pilot plant ------- The basin has a 5 hp surface aerator which is operated when biological oxidation is desired as part of the operating sequence. In that case, nutrients at normally used dosages are also fed to the basin. Further details on the basin and on each portion of the pilot plant are given in Appendix C. Effluent from the basin is pumped either directly to one or both of the two activated carbon adsorption systems, or it is pumped to the lime treatment system and then to the activated carbon systems. The lime treatment equipment consists of a 10 ft diameter reactor- clarifier with a 39 in. diameter center well, providing a retention time of 20 minutes in the reactor section and a clarifier rise rate of 550 gpd/ft at 30 gpm. Lime addition in the reactor raises the pH of the effluent to about 12 and causes precipitation of a mixture of CaCOo and a. Ca-organic floe. Lime addition is by pump from a mix tank into which hydrated lime is fed continuously from a hopper where it is slurried with about 5 gpm of recycled lime-treated effluent. A second reactor-clarifier permits carbonation of the lime treated, clarified effluent for removal of excess dissolved lime at pH 10.5. In the pilot plant,bottled COo is used for this purpose. At 30 gpm, this 7.5 ft diameter reactor-clarifier provides a retention time of 27 minutes in the reaction zone and a clarifier rise rate of 1000 gpd/ ft . A 200 gallon tank is provided for final pH adjustment with CC^. The lime treatment unit, the carbonation unit, the pH adjustment tank, and a 10,000 gal surge tank are located on an outdoor concrete pad. The remaining equipment items are located inside a 30 x 52 ft pre- fabricated steel building having a 20 ft wall height. The building also provides storage for truck-load quantities of lime and carbon. Two activated carbon adsorption systems are provided in the pilot plant, intended for simultaneous comparative operation if desired, both at 15 gpm. One system consists of four standard downflow carbon columns, 3 ft diameter by 15 ft high, each with a carbon bed of 10 ft (1600 Ib of granular carbon of 20 x 40 mesh size). At 15 gpm, each column provides 0.6 hour retention time in the carbon bed at a superficial velocity of 2 gpm/ft . All water connections to the columns are by means of 1% in. hoses with quick disconnects so that any desired arrangement of series or parallel flow could be used, either up-flow or down-flow through the columns. This arrangement greatly simplified the piping, especially for backwashing. A 4-ft diameter duomedia filter with a filter bed of 1 ft of anthracite and 1 ft of sand, supported on graded gravel, is provided for filtering the feed to the carbon systems when desired. 51 ------- A second carbon system, a novel system called FACET (for Fine Activated Carbon Effluent Treatment) and subject of a U.S. patent (41), was in- cluded in the pilot plant for initial development (Figure 6). The FACET system is a multi-stage counter-current agitated system with continuous counter-current transfer of carbon and liquid from stage to stage. The tanks used in the pilot plant are sized 3 ft diameter by 5 ft high, with a mixing zone of 170 gallons and a carbon charge of 212 Ib per tank at 15 g carbon per 100 ml of slurry, or 15 wt/vol %. Water to be treated flows by gravity through the three stirred tanks in series and then to a product storage tank or is pumped through a 3-ft diameter duomedia filter and then to the product water tank. Activated carbon, which has a mesh size of 40 by 140 mesh, is fed from a feed hopper into the third FACET tank. Carbon slurry, which typically has a concentration of 15 wt/vol %, is removed from the third tank at the same rate as the carbon is fed to that tank. The carbon is thus moved counter-current to the flow of water through the second tank, through the first tank, and to a spent-carbon container. The system endeavors to avoid some of the problems associated with the conventional granular and powdered carbon contacting systems, while combining advantages of both of these systems by using a carbon size intermediate to the standard powdered and granular classifications. Adsorption rate being inversely proportional to the square of the carbon particle diameter (12,13), the intermediate size carbon loads up to its capacity much faster than the coarser granular carbon. Although the ultimate capacities of both granular and intermediate size carbons are essentially equal, the higher adsorption rate of the intermediate size carbon results in a corresponding increase in carbon turn-over rate and a corresponding decrease in carbon inventory. Equipment size for the adsorber decreases correspondingly. The FACET system allows close control of the quality of the treated effluent by adjustments in carbon transfer rate or carbon slurry density, or both. Continuous discharge of spent carbon permits a closer tie between regenerator capacity and average carbon throughput, eliminating costly offstream inventory of carbon. Suspended solids present in the liquid to be treated do not present a plugging problem as in the granular carbon column system, hence primary clarifiers may be omitted. Comparing FACET to the powdered carbon system, the intermediate size carbon has a higher settling rate than the powdered carbon. This obviates the need for flocculants and substantially reduces the clar- ification area compared to that of a powdered carbon system (clarifi- cation area being a function of liquid flow rate and particle sedimenta- tion rate). Use of the intermediate size carbon allows a properly 52 ------- EFFLUENT CO r-ACTIVATED CARBON SPENT CARBON CONTACTORS FILTER STORAGE TANK Figure 6. FACET system ------- baffled agitated tank to serve as both adsorber and clarifier. As a result, equipment size as well as cost is significantly lower for the FACET system than for the powdered carbon adsorption system. Figure 7 shows the pilot plant building, the lime-treatment equipment, and equalization basin. Figure 8 shows the carbon columns, the duo- media filter in the right foreground, and the FACET tanks in the front foreground. Instrumentation Instrumentation of the pilot plant is sufficient to permit continuous 24-hour operation with only one operating shift for adjustments and sample taking. Conductivity, pH, turbidity, temperature, and flow rate are recorded continuously. Three water flow rates are automatically controlled and recorded: feed to lime treatment, feed to carbon columns, and feed to FACET carbon system. Automatic controllers are provided on the CC>2 feed lines to the carbonator reactor-clarifier and pH adjustment tank to control the pH in these tanks at a selected level. Timer controllers are provided for the sludge valves from the lime treater and the carbonator. The monitoring of the qualities of 12 streams in the process is pro- vided by a step programmer at the instrumentation-control center of the pilot plant. Lines from 11 sample points and a tap-water line are brought to the continuously operated sampling instruments through programmer controlled solenoid valves. The programmer opens one sample line at a time in sequence and the water from that sample point flows in series through a conductivity cell, pH measuring cell, a Hach light- scattering turbidimeter, and a Bausch and Lomb Spectronic-70 spectrophoto- meter for measurement of color. The signals from these four instruments and a thermocouple are conditioned and recorded on a twelve-point record- er during a five-minute measuring and printing cycle for the sample stream that is on. The stream number that is being measured is identified on the recorder. After the five-minute period on one stream, the pro- grammer steps to the next sample stream that is to be analyzed. The programmer can be modified to permit analysis of only the sample streams of real interest, which reduces the overall cycle time. This programmed stream analysis system permits the above five important parameters to be measured and recorded during the unattended evening shifts and on weekends. When an operator is present, the stream analysis system is valuable for showing trends and rate of change which permits better control of the operation. The pilot plant was built and shake-down runs made during 1971. The total cost of the pilot plant, excluding the building, was $194,000. 54 ------- Ln i ,i Figure 7. Exterior of pilot plant showing lime treater, carbonation tank, storage tank, and basin ------- Figure 8. Interior of pilot plant with carbon columns, FACET tank (foreground) and duomedia filter (right foreground) 56 ------- OPERATING PROCEDURES The normal procedures used in operating each part of the pilot plant, and the sampling and analytical procedures are discussed in this portion of the report. Procedures that were unique to a given treatment sequence are discussed in the sections on each treatment sequence. A simplified flow diagram is given in Figure 5. Bio-oxidation and Clarification The operation of bio-oxidation in the aeration basin was quite simple and required little attention. Water was pumped continuously from the mill effluent ditch through a 4-mesh strainer that was placed beneath the surface of the water in the ditch. At times when there was quite a bit of fiber in the effluent, the screen would accumulate solids and would require cleaning with a brush. The water was pumped into a basin at a flow rate of about 4 gpm greater than was being used in the pilot plant. The excess water flowed through a constant- head overflow attached to the drain line from the bottom of the basin. The water being pumped from the mill effluent ditch was discharged into the basin at the end opposite the withdrawal point. The retention time of water in the basin during bio-oxidation was 6.5 days, based on a flow rate of 20 gpm. Nutrients (ammonium nitrate and phosphoric acid) were added once per day in the ratio of 2.5 Ib N and 0.5 Ib P for each 100 Ib of BOD in the incoming water, which is the usual rate of addition in southern kraft mills. The 5 hp surface aerator was run continuously which kept the dissolved oxygen at about 6 mg/1. (A spare aerator was also placed in the basin for use if needed.) Primary Clarification When the operating sequence called for primary clarification without bio-oxidation, the aerator was turned off, nutrient feed was discontinued, and the aeration basin then served as an equalization basin to help level out the hourly and daily fluctuations of impurities in the mill effluent. The water from the equalization basin passed through the clarifier section of the basin and was pumped to the next treatment step. Lime Treatment When the operating sequence called for lime treatment, the aeration basin was again used as an equalization basin, with aeration and nutrient feed turned off, to level out impurities concentrations in the effluent. The water also passed through the clarification zone, although clarification was not called for at that point of the intended 57 ------- treatment sequence. Water from the clarification zone of the basin was pumped through a flow controller to the center mixing zone of the lime treater where a dilute slurry of lime was also fed continuously. Bagged hydrated lime was charged to a hopper on the feeder every one to three days. An oscillating screw feeder dropped the lime at the selected dose rate into a mixing tank to provide the slurry that was pumped to the lime treater. Water for slurrying the lime was recycled from the lime treater through a flow meter and a float valve which maintained a constant level of water in the slurry tank. The rate of feed or dosage rate of lime was manually set at a constant value for periods of at least several days and was varied only when attempting to relate color reduction to dosage. Later it became evident that for microlime treatment, the lime dosage should be varied to main- tain a constant concentration of soluble calcium. Without a suitable continuous monitor for soluble calcium, the control of dosage to main- tain a constant soluble calcium content was barely satisfactory. The mixture of lime and water in the axial cylindrical mixing chamber of the lime treater was agitated by a slow-moving multiple-arm stirrer. The water moved from the bottom of the mixing chamber upward through the clarification zone and overflowed to a pipe leading to the carbonator. Sludge that settled to the bottom of the lime treater could be discharged to the drain through a timer-controlled valve but was normally withdrawn once per day to a 50-gallon tank for measurement and sampling. Later in the program, it was found that better lime treatment was obtained by circulating sludge at about 5 gpm to the top of the mixing chamber to increase the solids content of the water in the mixing chamber. Carbonation The carbonator was of the same design as the lime treater except that the clarification zone had half the volume arid cross-sectional area of the lime treater, which provided twice the rise rate of the lime treater. Water, which flowed by gravity from the lime treater, was treated with cylinder C02 in the mixing zone to remove the excess lime in the water as CaCO-j. The C02 was fed from two banks of cylinders through pressure regulators, a flow control valve, and a flow indicator to a turbine-type gas disperser placed about 4 ft below the surface of the carbonator mixing chamber. An inverted-trough C0? bubbler located at the bottom of the mixing chamber was used initially for admitting the C02 to the water but the efficiency of absorption was very poor. The turbine type of CC^ disperser gave very good absorption efficiency. Sludge recycle to the top of the carbonator mixing chamber was later added by use of an air lift. Sludge from the carbonator was discharged once per day to the sewer. 58 ------- pH Adjustment Water from the clarification zone of the carbonator flowed by gravity to a 200-gallon tank in which C02 was admitted through a turbine disperser to adjust the pH to the desired level. The rate of admission of C02 to the carbonator and to the pH adjust tank was controlled automatically to the desired pH as measured by pH sensors in these tanks. Carbon Column Adsorption Water from the lime treatment unit, from bio-oxidation, or from primary clarification alone was pumped either directly to the carbon columns or through a duomedia filter and then to the columns. The duomedia filter operated quite satisfactorily if it was backwashed every one or two days for about 10 minutes at about 6 gpm/ft . Less frequent backwashing of the filter resulted in the formation of mud balls in the filter which were not broken up during backwashing. Backwash water was provided by well water stored in the 10,000 gal surge tank. Each column was charged with 1600 Ib of Atlas Darco 20 x 40 mesh activated carbon for all runs, which gave a bed depth of 10 ft and a bed volume of 70.7 ft . The columns were operated in the down-flow mode during all of the pilot plant runs. The water from the bottom of the first column flowed through the hose inter-connections to the top of the second column, and so forth, for as many columns as were to be used. The water from the final column flowed through a vertical riser pipe to an overflow equal to the height of the columns. This overflow device insured that the columns always remained full of water. The water pressure after each column was measured and the pressure drop across each column was checked several times each day and was recorded daily. At the relatively low flow rates used (0.7 to 2.1 gpm/ ft ), the pressure drop across all four columns remained at low values of 2 to 8 psi. Sample taps were available on each column at 2-ft intervals, but samples for analysis were normally obtained only after each column. The columns were backwashed every other day with well water stored in the 10,000 gal surge tank at 35 gpm (5 gpm/ft ) for 20 min. Before the backwash was started, compressed air was admitted to the bottom of each column to thoroughly agitate the carbon bed and break up any mud layer that might be forming on the top of the carbon bed. The backwash water was discharged to the sewer but, in a commercial plant, it would be clarified and reused for backwashing or added to the feed to the columns. 59 ------- When the quality of the product water from the columns became poorer than the selected limits, the first column was removed from the system and a column with a fresh charge of carbon was added as the last column. The second column then became the lead column. The carbon from the column that was removed from operation was discharged as a slurry by opening a 6-in hand-hole at the base of the carbon bed (top of the sand-gravel support bed). The spent carbon was discarded. Fresh carbon from bags was recharged to the column through a manhole at the top after water was added to the column. Bagged carbon was hoisted by an electric hoist to a charging platform at the top of the four columns. FACET Adsorption Water from the basin or from lime treatment was pumped through a flow control valve to the first of the three FACET contacting tanks. - The water flowed by gravity through the three tanks to a product storage tank or through a pump and duomedia filter. The speed of each of the slow-speed agitators in the FACET tanks was adjusted to barely keep all of the carbon suspended. Carbon to FACET (Atlas Darco 40 x 140 mesh XPT activated carbon) was charged to the feed hopper from bags and fed by a feeder and vibrating conveyor to the third FACET tank. The usual rate of carbon addition for treating 10 gpm of lime-treated water was only about 2-4 Ib/hr. The carbon was moved at this same rate through the three tanks counter- current to the water flow, while the concentration of carbon in each tank was maintained at the desired level (10-15 wt/vol %) by adjusting the flow of carbon from each tank. The carbon was transferred as a slurry from tank to tank by means of air lifts which performed reliably. Originally, we tried rotary-vane pumps having neoprene impellers for transferring the carbon slurry but these were not satisfactory. The solids concentration in each tank was measured several times during the day shift, and the rate of transfer from each tank was adjusted to give the desired solids concentration by adjusting air flow to the air lift. Spent carbon from the first FACET tank was collected, drained of water, weighed, and sampled for determination of moisture content. As indicated in Figure 6, baffles were provided in each tank to provide a clarification zone at the top of each tank so that no carbon would be carried by the water to the succeeding water treating tank. The baffles were inclined at an angle of 45° to allow the settled carbon to slide back into the agitated zone. These baffles had sufficient openings around the edges that some turbulence from the mixing zone persisted into the clarification zone which allowed some of the carbon to be carried by the water in the wrong direction. This lack of good 60 ------- clarification was particularly troublesome in the third FACET tank because 'some of the fresh carbon was lost from the system in the product water. The rise rate in this zone was 1.4 gpm/ft2 at 10 gpm, which allowed the 140-mesh particles to remain in the tank since they settled at a rate of about 18 ft/hr which is equivalent to 2.2 gpm/ft . SAMPLING AND ANALYTICAL PROCEDURES Samples were taken once per day during most of the pilot plant runs. Even though most of the properties of the water at each stage of the treatment varied only slightly during a 24-hour period, some streams were sampled more frequently if the changes were more rapid or if a special run was being made. The previously described automatic sensing and recording of pH, conductivity, temperature, color, and turbidity provided a constant readout of these properties for any number of sampling points up to twelve. Samples taken each day were analyzed in the research laboratory using the procedures and analytical equipment described in Appendix B. The properties determined in the analyses and the usual frequency of analysis are given in Table 16. In normal operation, we sampled and analyzed seven streams during operation of the bio-carbon sequence, six during the primary-carbon sequence, and eight during the lime-carbon sequence. When the FACET unit was operated, an additional 3 or 4 samples were analyzed. Most of these were analyzed daily but some much less frequently, as indicated in Table 16. 61 ------- Table 16. PROPERTIES OF WATER ROUTINELY MEASURED Property Frequency of analysis per week Color, FPCU (paper filter) 7 MPCU (0.8 micron Millipore filter) 7 CU (standard color) 7 TOG (paper filter) 7 TIC (paper filter) 7 pH 7 Temperature 7 Turbidity 1-7 Conductivity 2-7 Suspended solids 1 Total solids 1 Total dissolved solids 1 BOD-5 1 Sulfate 1 Calcium 7 Volatile neutral compounds 1 Volatile organic acids 1 Foam tendency 1 Adsorption isotherms after each step of treatment 1 per month Metals(Mg, Fe, Na, K, Al, Cr, Ni) about 2 per month Solids in lime sludge 1 Calcium in lime sludge 1 62 ------- Based on pilot plant experience, it was found desirable to determine the parameter of "color" not only by the industry standard APHA-NCASI method but also by modified procedures which measure apparent color. In the APHA-NCASI method, color is measured at 465 nm on a spectrophotometer after adjustment of pH to 7.6 and filtration through a 0.8 micron Millipore filter. All values reported in this report unless otherwise designated were obtained by this method, and are reported as "CU". One modified procedure involved 0.8 micron Millipore filtration prior to adjustment to pH 7.6. (Used after lime treatment, this procedure ensures more complete removal of suspended solids.) Another modified procedure con- sisted of measuring color after filtration through a Whatman No. 2 filter paper and adjustment to pH 7.6, referred to as "7.6 FPCU". A third modified procedure consisted of measuring color after filtration through Whatman No. 2 filter paper without pH adjustment, noted as "FPCU". A fourth modified procedure consisted of measuring color after filtration through an 0.8 micron Millipore filter without any pH adjustment, noted as "MPCU". The 7.6 FPCU for the different treatment sequences were higher than the standard colors by the following factors: bio-carbon, 2.0; primary- carbon, 1.7; lime-carbon, 1.7. Apparent colors (FPCU) measured with pH adjustment (in the pH range of approximately 8 to 11,5) were higher than pH-adjusted colors by a factor of about 2.0. The color of effluents from pulp and paper mills decreases as the pH is reduced from pH 12 to pH 7.6, which is the desired pH for discharge to streams. As the pH is further reduced the color decreases and then drops rapidly at pH 4 and below, at which point the lignin compounds start to precipitate from solution and are most completely precipitated at pH of about 2. The ratios of FPCU/MPCU and also FPCU/TOC were used to indicate the concentration of large colloids and agglomerated color bodies in the water that passed through a paper filter but were retained by the 0.8 micron Millipore filter. MANPOWER REQUIREMENTS The operation of the pilot plant normally required one operating engineer, one operating technician, and one analytical technician for 40 hr per week. Analysis of data, planning, supervision of support- ing laboratory studies required 40 hr per week of an experienced development engineer. Maintenance, especially for the pH controllers and continuous water analyzers, required a craftsman about 8 hr per week. 63 ------- SECTION VII OPERATING RESULTS: PRIMARY CLARIFICATION AND BIO-OXIDATION BIO-OXIDATION The bio-carbon treatment sequence was the first to be evaluated in the pilot plant after shake-down runs were made to check out each treatment stage. The procedure used in operating the aeration basin for biological oxidation was discussed in Section VI. The aeration basin was operated continuously from December 1, 1971 to March 18, 1972, with the primary purpose of providing bio-treated water as feed to the carbon columns that would be typical of bio-treated water from aeration basins used by a large number of pulp and paper mills. The bio-carbon sequence of treatment would provide operating and cost data that could be used by those mills to evaluate the treatment of the water from their aeration basins by carbon adsorption to provide water suitable for reuse in the mill. The rates of removal by bio-oxidation of BOD and TOC in the basin were only slightly less than those obtained by aeration basins in the pulp and paper industry. At the usual flow rate of mill effluent through the basin of 20 gpm, the retention time in the 187,000-gal basin was 6.5 days. The 5-hp surface aerator maintained an average dissolved oxygen content of 6 mg/1. The temperature of the incoming water was about 115°F, and the average temperature in the basin was 74°F. Average results from bio- oxidation are given in Table 17 for the period of February 4, 1972 to March 11, 1972, which was the period used for evaluating the results from the bio-carbon sequence. 64 ------- Table 17. BIO-OXIDATION AND CLARIFICATION FOR PERIOD OF 2/4/72 to 3/11/72 AT FLOW RATE OF 20 GPM Percent To basin From basin removal BOD, mg/1 278 69 75 TOC, mg/1 278 139 50 Color, FPCU 1610 1329 17 MPCU 1330 871 35 CU 950 810 15 pH 10.2 8.6 Temperature, F 116 74 Conductivity, microtnhos 1580 Turbidity, JTU 34 23 32 Suspended solids, mg/1 422 202 52 Methanol mg/1 35 4.1 88 Acetaldehyde, mg/1 9 10 0 During this test period, the average removal of BOD was 75%, the removal of TOC was 50%, and the removal of standard color was 15%. For compar- ison, the removals in the aeration basins treating the total effluent from the Pensacola mill for a seven-month period were 92.5% of the BOD and 28% of the TOC, and the color increased 7%. Most mills find no color reduction in bio-oxidation and less reduction of TOC than was obtained in the pilot plant. No explanation can be given for these differences in pilot plant and mill results, other than pointing out that the pilot plant bio-oxidation basin received raw effluent, while the mill bio-oxidation basin ponds receive clarified effluent. Bio-oxidation and incidental air stripping removed 88% of the methanol but none of the acetaldehyde, on the basis of a single set of samples taken during this period of operation. In this set of samples, there were no significant changes in the concentrations of metal ions. The concentrations of color, TOC, and BOD in the mill effluent as it entered the aeration basin varied widely - as much as four fold from one 24-hr composite sample to another. Such surges were partially dampened by the 6.5-day retention time in the basin, but these surges still caused wide swings in color and TOC concentrations in the bio- treated water fed to the carbon columns, as will be discussed later under carbon column operation. 65 ------- PRIMARY CLARIFICATION The basin was used during the period of April, 1972 to December 22, 1972 as an equalization basin and a primary clarifier to provide water to the primary-carbon and the lime-carbon sequences of operation. Mill effluent water was pumped into the basin at a flow rate about 4 gpm greater than was used for the pilot plant, with the excess over-flowing to waste. The aerator was not used. A low degree of mixing in the basin was provided by jetting the incoming water into the basin. Most of the solids in the incoming water settled in the basin, but some further clarification took place in the clarifier section of the basin. During the period in which the basin was used for equalization and clarification, it was noted that there was anaerobic activity presumably because of the 1-ft to 2-ft layer of sludge (primarily bark fines and fiber) on the bottom of the basin that had accumulated during the bio- oxidation operation. On the basis of 12 analyses, the passage of the water through the equalization basin caused the average methanol concen- tration to decrease from 45 to 33 mg/1, for a reduction of 277o, and the average acetaldehyde concentration to decrease from 22 to 17 mg/1, for a reduction of 23%, the formic acid concentration to decrease from 10 to 7 mg/1, and the lactic acid concentration to decrease from 1.5 to 0.7 mg/1. The glycolic acid content remained at about 10 mg/1, and the acetic acid content increased from 26 to 92 mg/1, for an increase of 254%. Presumably, this large increase of acetic acid resulted from anaerobic biological growth in the basin. The TOG contributed by these volatiles increased from 48 to 79 mg/1, an increase of 65%. As discussed later, these volatile neutral and acidic compounds contributed substantially to the TOG remain- ing in the water after carbon adsorption. There was no significant decrease of color, TOG, or BOD on the basis of eight sets of samples, but accurate values of changes in these parameters were difficult to get be- cause of the 5-day time lag of the water going through the basin. The average amount of suspended solids in the water to the basin was 422 mg/1 during the bio-carbon operation but was only 103 mg/1 during the lime-carbon operation from September through December, 1972. The amount of suspended solids in the water from the settling basin ranged from 48 to 265 and averaged 145 mg/1. The average turbidity from the settling basin during the primary-carbon and lime-carbon operating periods was 22 JTU. 66 ------- SECTION VIII OPERATING RESULTS: LIME TREATMENT AND CARBONATION OBJECTIVES AND DESCRIPTION OF OPERATIONS The lime treatment --carbon adsorption (lime-carbon) sequence was operated from June 30, 1972 to December 22, 1972. A flow rate of 22 gpm was used most of the time and 11 gpm the rest of the time. The lime-treated water was used as feed to the carbon columns at 10 gpm and for FACET carbon adsorption at 10 gpm during most of the period. The lime dosage ranged from 318 to 980 mg/1 CaO. This dosage compares with 750-1800 mg/1 CaO used or practiced by other pulp and paper companies (5,14,18) as the so- called minimum lime treatment. Results obtained in the pilot plant in the first few weeks of operation on this treatment sequence, and related results obtained earlier in laboratory inveatigations, had given indications that a color reduction to the 100 CU range by a lime-carbon treatment sequence might be achieved most economically by using a lime dosage substantially below that required to achieve maximum color removal in the lime treatment step. Indications were that the lime dosage might be reduced to the level where the lime- carbon treated water could be reused in the mill without prior removal of soluble calcium by carbonation, thus reducing lime treatment capital and operating costs. This system was dubbed "micro" lime (microlime) treatment, as compared to the "minimum" lime (minilime) treatment in which carbonation for calcium removal is required. The full results obtained in the pilot plant in fact support the microlime - carbon treat- ment as the treatment of choice to achieve treated effluent quality of 100 CU. Because of the promising results obtained with microlime treatment follow- ed by carbon adsorption without carbonation of the lime-treated water, most of the data obtained on lime treatment was with low lime dosages of 350 to 700 mg/1 of CaO. This microlime period of operation was from July 16, 1972 through December 6, 1972. Carbonation of the microlime treated water was not used because the residual calcium concentration at about 80 mg/1 Ca was considered to be low enough for reuse. In December 1972 the minilime treatment process was evaluated at a dosage of about 970 mg/1 of CaO. Carbonation of the minilime treated water was used since it would be required in mill operations to recover the soluble calcium and to make the water suitable for reuse in the mill. Carbon- ation with lime kiln gas removes soluble calcium in one lime treatment installation (14), while natural carbonation accomplishes calcium removal in another (5,18). 67 ------- In the course of pilot plant operations, it was found that color removal in the carbon colums as well as in lime treatment is directly influenced by the concentration of soluble calcium in the water from the lime treat- er or to the carbon columns, while lime dosage itself exerts only an in- direct influence. However, lack of automatic feed-back control based on soluble calcium dictated continuation of operations based on a predeter- mined lime dosage. This led to swings in the soluble calcium with chang- ing calcium demand of the effluent based on fluctuations in carbonate ion and color concentration. The control of residual soluble lime in the water from microlime treatment was difficult, not only because of variations in concentrations of carbonate ion and color in the feed water to lime treatment, but also because of long time lags in receiving ana- lytical results on the calcium content of the water. About 40% of the lime added in the microlime treatment was used in converting carbonate ion to CaC03, but since the carbonate concentration varied as much as 25% in a 24-hr period, the demand for lime for this reaction also varied by 25%. In general, the color in the feed water increased at the same time as the carbonate increased, which caused additional fluctuations in the demand for lime. Other factors, largely mechanical in nature, in both the lime treater and carbonator caused occasional upsets in the operation. Most of these operating problems were overcome quickly, and changes in equipment or procedure were made to prevent reoccurrence of the same problems. REMOVAL OF COLOR AND TOG IN LIME TREATER Data Analysis As a result of the variations in the composition of the feed water to lime treatment and variations due to mechanical problems, there were few periods for which the effect of independent operating variables could be related to the dependent variables without careful analysis of the data. The dependent variables primarily used in judging perfor- mance were color and TOG concentrations of the treated water. The data were analyzed by several means. For a typical day, 30 to 40 items of primary data were logged from operating log sheets and laboratory analysis sheets onto a summary table. These data were also inserted in a computer program which gave a print-out of all primary as well as many secondary calculated values (see Appendix F). Most of the primary items of data were also plotted each day on elapsed-time graphs. These graphs permitted visual correlations of many of the variables. One method of analyzing the data consisted of selecting periods of operation having similar conditions of operation and grouping these into "runs". There were 15 such runs which lasted for periods of 3 to 20 days each. 68 ------- Another method of data correlation consisted of computer-plotting all values of one property against another property suspected of showing some relationship or of providing more insight to the operation. A third method consisted of obtaining correlation coefficients by computer programs for nine independent versus nine dependent variables and run- ning multiple-regression analyses of these variables. Since many of the dependent variables were largely influenced by the soluble calcium content of the treated water, the data for both lime treatment and carbonation steps were divided into three ranges of calcium concentra- tions. By this means, this important variable (calcium concentration) could be held essentially constant while the effects of other variables were compared. Results of Color and TOG Removals in Lime Treater Alone The results from the treatment of mill effluent with lime in the lime treater without carbonation are given in Table 18. All results, ex- cept for days when there were equipment failures and upsets, are aver- aged under low, medium, and high concentrations of calcium in the product water from the lime treater. The average calcium concentrations in the three ranges were 67, 138, and 228 mg/1. Also, results obtained closer to the preferred microlime level of 80 mg/1 have been averaged from the applicable runs in the low and medium calcium range. As seen in Table 18, the calcium concentration had little relationship to the dosage of lime. This lack of correspondence was caused primarily by variations of Na2CO.j (or TIC) and also by variations in color content of the mill effluent to the lime treater. As indicated in Table 18, the total inorganic or carbonate carbon (TIC) of the feed water happened to be high during the low-calcium runs and low during the high-calcium runs. Lime treatment reduced the average color of the feed water 63% when the product water contained 67 mg/1 of Ca, 68% with 71 mg/1 Ca, 78% with 138 mg/1 Ca, and 81% with 228 mg/1 Ca. Lime treatment also reduced TOC by 34 to 40% on the average for these ranges of soluble calcium. The BOD content was reduced only 9-13%, which shows that the compounds making up much of the BOD are not very susceptible to removal by lime treatment. 69 ------- Table 18. AVERAGE RESULTS FROM LIME TREATMENT Number 'of days averaged Lime dose, as mg/1 CaO as mg/1 Ca Calcium (soluble), mg/1 range, mg/1 Color, FPCU cua removed, CU % TOC, mg/1 removed, 7° TIC, mg/1 BOD, mg/1 removed, % pH Conductivity, micromhos Conductivity increase Turbidity, JTU Volatile organic compounds methanol, mg/1 acetaldehyde, ing/1 acetic acid, mg/1 Sludge density, g/ml Sludge, % dry solids Ca tied up by organics, mg/lc Atoms of C/atom Ca in organics Medium calcium 36 478 34 443 342 feed 9 1446 712 248 64 165 9.7 1225 19 32 30 115 prod. 67 17-83 1071 262 63 163 34 14 151 9 11.8 2245 1020 53 34 28 105 1.40 24 117 2.4 feed 9 1310 761 206 59 187 9.8 1471 18 37 29 70 317 prod. 71 38-103 977 243 68 139 33 12 _ - 11.8 2819 1348 45 . _ - - - 98 2.3 feed 15 1670 956 275 52 261 10.1 1593 24 36 21 85 69 528 377 prod. 138 103-164 975 211 78 169 39 11 276 6e 11.9 3597 2004 49 41 19 59 1.14 25 117 3.0 16 576 412 feed 13 1386 842 246 34 416 10.1 1247 33 12* 3* 46b prod. 228 172-422 770 156 81 146 40 11 348 16 12.0 3749 2502 72 "H 6b 42b 1.27 35 120 2.8 a b c d In the product water, filtration preceded adjustment to pH 7.6 in the standard color measurement. Results from a single sample. Ca added plus Ca in feed water, minus Ca in product water and Ca in CaCO-j formed with TIC. Atoms of TOC removed per atom of Ca tied up by organics. Increase ------- Influence of Soluble Calcium on Removals The influence of calcium concentration on percent removal of color in lime treatment and on color remaining is shown in Figure 9. Each point is for a run which consisted of 3 to 20 days during which conditions were fairly constant. The results show that when the average soluble calcium concentration was 80 mg/1, the color was reduced by about 70% and that the calcium content had to be doubled to increase the removal to 80%. The color was reduced to about 230 CU at 80 mg/1 Ca and to about 100 CU at about 400 mg/1 Ca (which is the calcium content from minilime treatment before carbonation). The TOG data in Table 18 show that removal of TOG in lime treatment was influenced somewhat less than was color removal by the soluble calcium concentration. The influence of calcium concentration on removals of color and TOG are also discussed in Section XI under results from carbon adsorption following lime treatment. In the lime-carbon sequence it was found that the soluble calcium content had to be at least 60 to 80 mg/1 to achieve good removals of color and TOG in the carbon columns. Spruill (14), in full-scale lime treatment of kraft pulp and paper effluent by Continental Can Co., found a similar influence of calcium content on color removal. The Continental Can Co. uses minilime treat- ment and a soluble calcium concentration of about 400 mg/1 Ca to pro- vide reliable operation and satisfactory removals of color. (In this plant the excess calcium in the water is recovered by carbonation.) The conductivity of the water was increased by the addition of lime, primarily from the increased NaOH content from the reaction of Ca(OH)2 with Na2C03 but also from the increased concentration of Ca(OH)2« The solubility of Ca(OH) (about 1500 mg/1) was not exceeded by the lime dosages used in this program, and therefore essentially all of the lime not used in reaction with organic compounds or ^2663 went into solution. Any free Ca(OH)2 found in the sludge at these applied lime dosages is a result of insufficient mixing or time for dissolution. The percent increase of conductivity in lime treatment correlated fairly well with the soluble calcium concentration of the treated water, as shown in Figure 10. When lime was added to give a soluble calcium concentration of 80 mg/1, the average conductivity increased from 1600 to 3200 micro- mhos, for an increase of 1600 micromhos or 100%. The relationship of soluble calcium to conductivity increase might be useful as a rough control of the lime dosage to give a desired concentration of calcium in the lime-treated water. The turbidity of the feed water ranged from 15 to 35 JTU and that of the lime-treated water averaged about 58 JTU (see Table 18). The turbidity of the product water was largely due to slowly settling floe that was not removed in the clarifier section of the lime treater. 71 ------- 100 S3 IJ > Q S O Pn O 30 - 20 - - 400 - 300 P4 3 o CJ - 200 CO - 100 100 200 SOLUBLE CALCIUM, mg/1 300 400 Figure 9. Influence of calcium in lime-treated water on percent of color removed and on color remaining ------- 450 400 MEAN RANGE OF 2/3 OF VALUES 100 200 % INCREASE OF CONDUCTIVITY 300 Figure 10. Soluble calcium in lime-treated water versus increased conductivity due to lime treatment 73 ------- Lime treatment removed very little of the volatile organic compounds, as indicated by the results in Table 18. This result is not surprising since these compounds would not be expected to form insoluble calcium compounds. As discussed in Section VII, it is believed that most of these volatiles were produced by anaerobic activity in the equalization basin. These compounds contributed about 60 out of the 169 mg/1 TOC in the product water from the medium-calcium runs. SLUDGE FROM LIME THEATER The solids from the lime treater were found to settle rapidly and to form a dense sludge. The average sludge, as withdrawn from the lime treater each morning, contained 28% solids. The solids content of the sludge varied almost linearly with the specific gravity of the sludge as shown in Table 19. Table 19. RELATION OF SPECIFIC GRAVITY AND SOLIDS CONCENTRATION OF LIME TREATER SLUDGE Specific gravity, Solids in sludge, g/ml weight % 1.03 5 1.05 9 1.10 17 1.15 24 1.30 42 The amount of sludge produced per unit of water treated as determined from measurement of sludge volume and solids content was quite variable. Therefore the amount of sludge and sludge solids were calculated from the carbonate removed from the water, the organics removed (TOC), and calcium removed in the sludge with the assumption that no solids were carried out in the treated water. These calculations indicated that during the 69 days of operation with the medium calcium concentration range, the average amount of sludge, expressed as concentration in the product water, was 629 mg/1 or 5.2 lb/1000 gal. The sludge solids con- sisted of about 40% Ca of which 54% existed as CaCOo (from removal of carbonates), and 46% as calcium-organics. This rather high content of calcium carbonate is evidently the major factor causing the sludge to settle and to filter readily. In contrast, the sludge from minilime treatment of mill effluents is difficult to filter (3,7,8,14) and special 74 ------- precautions are used to add fiber or lime mud from mill operations to make the color-removal sludge filterable. It is also possible that the color compounds which are removed only when the higher lime dosages of the minilime treatment are used are primarily responsible for the poorer settling and filtering characteristics of the sludge from the minilime treatment. In a filtration test with a representative sample of the sludge from lime treatment, the filtration rate with a vacuum of 24 in. Hg and a final cake thickness of 0.7 in. was 380 Ib dry solids per hour per square foot. The cake had a dry solids content of 74%. It appears that the sludge would be suitable for recycle to the lime kiln by adding it to the lime mud before the mill lime mud filters. The average distribution of calcium in lime treatment is shown in Table 20 for operation at microlime, medium, and high calcium concentrations. The calcium in the product water ranges from a low 71 mg/1 (or 22% of the calcium added plus the calcium in the feed water) in the microlime treat- ment to a high of 228 mg/1 (or 54%) in the high soluble calcium treatment. Similarly, the Ca in Ca-organics ranges from 98 to 120 mg/1, increasing with increasing color removal. The Ca in CaCOo is determined only by the TIC concentration of the feed water since added calcium is consumed in CaCOo formation before forming Ca-organics and raising the soluble calcium concentration. It is expected that with good control of lime dosage to give a constant concentration of calcium in the treated water, good lime-carbon operation will be obtained at about 80 mg/1 Ca in the treated water and that the loss of calcium to the water will be about 25%. The amount of calcium in the sludge as Ca-organics was 98-120 mg/1 for the four groups of runs (Tables 18 and 20). The average ratio of atoms of carbon in the sludge to atoms of calcium associated with the organics (which is the calcium in the feed plus the calcium added less that in the product water and less that tied up as CaC03 by the removal of carbonates) was about the same for the four groups of data from lime treatment and ranged from 2.3 to 3.0. The average ratio for all of the lime treatment operation was 2.74 atoms of carbon per atom of calcium. If it is assumed that the organic material removed is "typical" kraft lignin having 9 carbon atoms and 3.3 oxygen atoms per monomeric group, the above average Ca-organic sludge contained 3.3 calcium atoms per monomeric group. The Ca-organics contained, on the average, 0.82 mg TOC/mg Ca, 5.3 CU/mg Ca, and 6.9 CU/mg TOC. The above Ca-organic atomic composition suggest some difficulty in con- sidering the removal of color from unbleached kraft effluent as a stoi- chiometric reaction of calcium ions arid acidic organic groups such as carboxylic and enolic groups, as Bennett and Dence (52) have shown for kraft bleach effluent. At least a part of the functionality of calcium ions may be explained as coagulation of colloidally suspended substances. No doubt some of the color is originally in colloidal forms, while prod- ucts of stoichiometric reaction with calcium may form additional colloids. 75 ------- Table 20. DISTRIBUTION OF CALCIUM IN LIME TREATMENT Microlime concentr. of soluble calcium, average of 34 days Medium concentr. of soluble calcium, average of 69 days High concentr. of soluble calcium, average of 16 days TIC TOC Calcium in: Ca in feed water Ca added as lime Ca in, total Calcium out: Ca in product water Ca in CaC03 (from TIC removed) Ca in Ca-organics (by difference) Ca out, total Feed water mg/1 , 59 206 9 317 326 ) Prod. water mg/1 % of tot. 12 139 71 22 157 48 98 30 326 100 Feed water mg/1 52 275 15 377 392 Prod. water mg/1 % of tot. 11 169 138 35 137 35 117 30 392 100 Feed water mg/1 32 246 13 412 425 Prod. water mg/1 % of tot. 11 146 228 54 77 18 120 28 425 100 ------- CLARIFICATION IN THE LIME TREATER The rise rate in the clarification zone of the lime treater was 0.31 gpm/ft2 or 450 gpd/ft at the normal flow rate of 22 gpm. At this rise rate some of the ca-organic floe failed to settle even though slude from the reaction-mixing zone settled very rapidly. Several modifications were made to provide increased contact between the feed water and the lime to increase the degree of reaction and utilization of the lime. The dosage of lime, in the range we used (including the minilime range) appeared to have little influence on the settling of the small amount of suspended solids in the water. With the improved conditions of mixing, the average turbidity for the three groups of runs ranged from 19 to 33 JTU in the feed water to the lime treater and from 49-72 JTU in the water from the lime treater. Corresponding values of suspended solids were 170 mg/1 in the feed water and 550 mg/1 in the water from the lime treater. Gould (7,8) also encountered floes that settled slowly in the Georgia Pacific lime treatment of bleach effluent and used synthetic flocculants to remove them. Carry-over of suspended solids from the lime treater was no problem in the operation of the carbon adsorption columns, although it would be prudent to minimize such carry-over. On the other hand, for direct discharge or reuse of micro-or minilime treated effluent, the level of turbidity and suspended solids observed in this study suggests that means for improved suspended solids removal must be further developed, or provided- for in the form of synthetic flocculant, in a commercial installation. One of the changes mentioned above to improve the degree of clarification was to increase the solids concentration in the mixing zone of the lime treater by externally pumping sludge from the bottom of the lime treater to the top of the mixing zone where feed water and lime slurry entered. This modification increased the solids content in the mixing zone from about 4 to 15 g/1. The suspended solids in the water from the lime treater was apparently reduced about 20% by this change. Another modification was made to improve the degree of mixing and to increase the solution of the particles of Ca(OH)2 by increasing the flow of lime-treated water to the lime slurry mixing tank at the lime feeder. The flow was increased from 2 gpm to 5 gpm which was calculated to be sufficient to dissolve about 40% of the lime from the feeder. Undissolved particles of lime were further dispersed by the centrifugal slurry pump from which a portion of the slurry was recycled to the slurry mixing tank. This increased flow of water to the slurry mixing and the recycling of the slurry through the feed pump caused a markedly visible reduction of white lime particles in the sludge from the lime treater and evidently increased the utilization of the lime. 77 ------- The turbidity and filterable solids apparently consisted of slow-settling flocculated color bodies. These did not contribute to the measured color and TOC because the water is filtered before these determinations are made. We observed that, if the pH of the samples were reduced to 7.6 before the sample were filtered, the flocculated color dissolved and increased the color of the water after the normal filtration. OPERATION OF CARBONATOR Most of the operation of the lime treatment section of the pilot plant was with the microlime process which doesn't require carbonation to re- move the excess calcium. When using microlime treatment, the water from the lime treater was allowed to flow through the carbonator as a means of getting the water to the pump that transferred it to FACET carbon adsorption or through the duomedia filter to the carbon columns. The carbonator was used in conjunction with microlime treatment for a period of 11 days to test the effects of low concentrations of calcium (7-23 mg/1) in microlime-treated water on the performance of the carbon columns. Another run was made with the carbonator to provide minilime treatment of water for an 11-day test of carbon adsorption following minilime treatment. In the minilime run, the lime dosage averaged 920 mg/1 CaO with a water flow of 22 gpm. In the lime treater, the color was reduced from 730 to 81 CU for a reduction of 89%, and the TOC was reduced from 212 to 126 mg/1, for a reduction of 40%. The higher lime dosage did not cause noticably improved settling of floe nor reduce the turbidity of the water from the lime treater, as compared to operation at lower lime dosages. Of the calcium going into the system, 63% or 413 mg/1 Ca, ended up in the water from the lime treater, compared to 22 to 54% (71 to 228 mg/1) in earlier operations. This concentration of calcium in the water and the degrees of removal agree closely with those obtained by Spruill (14) in full- scale treatment of kraft mill effluent. The carbonator was operated to reduce the pH to 10.5, the point of least solubility of CaCOo, and to reduce the calcium content from about 400 to 10-20 mg/1. Further reduction of pH to about 8 was obtained with the addition of C02 to the pH-adjust tank. Malfunctioning of the pH sensors and C02 flow controllers caused occasional increases of calcium in the water from the carbonator above the desired range. The concentrations of color and TOC in the water from the carbonator and pH-adjust tank were normally greater than in the water from the lime treater. This increase was caused by the small amount of floe in the lime-treated water that dissolved at the lower pH of the carbonator and pH-adjust tanks. 78 ------- The sludge from the carbonator, which is essentially all CaCOo, settled and filtered rapidly and was quite similar to the sludge from the lime treater. In commercial plants, this sludge would be pumped back to the lime treater to be combined with lime treater sludge. The turbidity of the water from the carbonator was rather high (about 80 JTU). Only a slight reduction of this turbidity was obtained by duomedia filtration. Three potential sources of this turbidity were examined. The first, i.e., floe carry-over from the lime treater, was ruled out because laboratory tests showed that the carried-over lime- color floe redissolves upon carbonation, resulting in an increase in color and in formation of CaCOo. The second potential source of turbidity is carry-over of CaCO_ formed in the carbonator, particularly in the colloidal form. Acidification of samples to pH 6 to dissolve CaCOo resulted in a turbidity decrease of only about 50%, leaving a white colloidal material in suspension. The nature of this third potential source of turbidity was determined as most likely to be colloidal elemen- tal sulfur formed in the carbonator due to pH reduction of polysulfide- containing effluent, as discussed below. At times the water from the equalization basin contained as much as 20 mg/1 of sodium polysulfide, which caused the lime-carbon treated water to have a bright yellow hue in the absence of carbonation in the carbon- ator. When effluent samples containing the sodium polysulfide were adjusted to pH 6 or lower, the polysulfide liberated colloidal elemental sulfur which had a white colloidal appearance. Several spot tests indica- ted significant increases of polysulfide in the water as it passed through the equalization basin. Polysulfide formation probably involves action of anaerobic bacteria on thiosulfate and/or partial oxidation of sulfide to sulfur. 79 ------- SECTION IX OPERATING RESULTS: BIO-CARBON COLUMNS OBJECTIVES AND DESCRIPTION OF OPERATION Objectives The bio-carbon sequence was the first to be operated in the pilot plant. As discussed in Section VII, mill effluent was bio-oxidized and clarified to provide water for the carbon adsorption operation that was typical of the effluent from pulp mills having secondary treatment. The objectives of the pilot plant operation with the bio-carbon se- quence were to determine how successfully carbon adsorption in columns could remove the color and TOC that remained after typical clarification and bio-oxidation of pulp mill effluent and to obtain data needed to design full-scale plants if this method had technical and economic promise. The needed design data included the amount of fresh carbon that is required per 1000 gal of water treated, the contact time (column length and flow rate) of water in the column needed for remov- als to given levels, pressure drops through the columns, backwash procedures needed, rates of removal of color and TOC, and the effects of anticipated biological activity in the carbon beds. Clarification and.Filtration of Water to Columns The bio-oxidized water was partially clarified in a section of the basin and then filtered in a duomedia filter to insure that the operation of the carbon columns would not be adversely affected by suspended solids in the water. A wooden slat fence used to section off a portion of the basin as a damper was only partially effective in eliminating circulation through this section, hence the water discharged was rather high in suspended solids. Toward the end of the bio-carbon sequence, a plastic film covering of the fenee resulted in achieving effective clarification in this section. The duomedia filter contained 12-in layers of anthracite and sand supported on graded gravel. With the normal flow rate of 15 gpm used in the bio-carbon sequence, the flow velocity in the filter was 1.2 gpm/ft. By trial and error, we found that mud balls would form in the 80 ------- 00 N> H M o o 5| I O CJ o o H 2000 1500 1000 500 300 200 100 MILL EFFLUENT TO BIO-OXIDATION TO COLUMNS FROM COLUMNS III! TO BIO-OXIDATION TO COLUMNS FROM COLUMNSV^ 2/124 68 10 12 14 16 18 20 22 24 26 28 3/1 3 5 79 11 DATE Figure 11. Concentrations of color and TOG during bio-oxidation- carbon adsorption, Feb. 1 through March 11, 1972 ------- to March 11, 1972), the flow was 15 gpm (2.1 gpm/ft2) through four columns in series. Four fresh columns were added resulting in an average dosage rate of fresh carbon of 8 lb/1000 gal. Note that the concentrations of color and TOG in the 24-hour composite samples of mill effluent to bio-oxidation have a close correspondence, and that both varied widely from day to day. These wide swings of color and TOG in the water to bio-oxidation were dampened quite a bit by the average retention time of 6.5 days in the equalization basin. The changes of concentration in the feed to the columns usually caused corresponding changes in the product water. The amount of color and TOG removed remained" fairly constant in spite of the changes of feed concentration to the columns. The average reduc- tions of color and TOG by carbon adsorption were 578 CU and 91 mg/1 IOC. Average results from the bio-carbon operation for the operating period of February 4, 1972 to March 11, 1972 are given in Table 21. More complete operating results are given in Table 22 for the bio-carbon operation. Included in Table 22 are average results for each of the four column positions and for other measured parameters not included in Table 21. A printout of daily summaries of conditions and results is given in Appendix D. This mode of operation at a column feed pH of 8.6 and a product pH of 8.2 gave a water with an average concentration of 58 mg/1 of TOG, which is expected to be acceptable as a general mill feed water. However, the color level of 230 CU may be too high for some areas in the mill. It is estimated that an additional 25 ft of carbon bed would be required to reach a target of 100 CU in the treated effluent. This increased length increases the amount of carbon in use (carbon inventory) in the same proportion, i.e., by 63%, while it increases the average dosage rate only by the 25% for the 25% greater amount of color removed. Both factors are important in the economics of the total system. A rather careful determination must be made by each mill, and by the industry as a whole, of actual water quality requirements in terms of color and TOG in reuse water obtained from effluent. The amount of color and TOG removed in the lead column (Table 22) was greater than in any of the other three columns even though it was fairly well loaded on the average. The amount removed per column became less for each succeeding column. On the other hand, each column removed a nearly equal percentage of its feed. 83 ------- Table 21. OPERATING SUMMARY OF BIOLOGICAL OXIDATION CARBON ADSORPTION SEQUENCE AT 15 GPM LOW CARBON DOSAGE Number of fresh charges of carbon 4 Average temperature of water to columns F 74 Cumulative removal when discharge - CU/g carbon . 646 - TOG, mg/g carbon 98 Average dosage rate, Ib carbon per 1000 gal water 8 Average retention time of water through 4 columns, hr 2.34 Volumetric flow rate, v/(v hr) 0.43 Color removal Range Average Feed to bio-oxidation, CU 460-2700 1210 Feed to carbon columns, CU 500-1200 810 Product from carbon columns, CU 46- 440 232 Removal by bio-oxidation plus filter, % - 33 Removal by carbon, % of feed to carbon - 71 Total removal, % of feed to bio-oxidation - 81 Rate of removal by carbon, CU/(g hr) 0.40-0.87 0.67 TOG removal Feed to bio-oxidation, TOC, mg/1 120- 966 277 Feed to carbon columns, TOC, mg/1 90- 350 149 Product from carbon columns, TOC, mg/1 10- 90 58 Removal by bio-oxidation plus filter, % - 46 Removal by carbon, % of feed to carbon - 61 Total removal, % of feed to bio-oxidation - 79 Rate of removal by carbon, TOC, mg/(g hr) 0.06-0.18 0.11 84 ------- Table 22. AVERAGE RESULTS FOR BIO-CARBON SEQUENCE FOR PERIOD OF FEBRUARY 4, 1972 to MARCH 11, 1972, LOW CARBON DOSAGE Flow rate, gpm Temperature water to basin, °F to columns, °F Color, to basin, FPCU CU to columns, FPCU CU from columns, FPCU CU TOG, to basin, mg/1 to columns, mg/1 from columns, mg/1 BOD, to basin, mg/1 to columns, mg/1 from columns, mg/1 D.O., from basin, mg/1 pH Conductivity, micromhos, Turbidityb, JTU Filterable solidsb, mg/1 Pressure drop, columns0, psi Color removal rate, CU/(g hr) TOC removal rate, mg/(g hr) Color removed, 7<> to each column Color removed, % to lead column TOC removed, % to each column TOC removed, % to lead column Color remaining, % to lead column TOC remaining, 70 to lead column Color removal, CU TOC removal, mg/1 Cumulative color removed, CU/g Cumulative TOC removed, mg/g & Four fresh columns were added during this period, so four individual columns served as the lead column, four as the second, etc. Average of two analyses. c Ap per column just before backwashing. General or overall data 14.8 112 74 1610 1210 1408 810 277 149 278 67 6.0 8.6 16 55 0.67 0.11 71 61 578 91 Lead column 1170 621 109 8.4 1580 3.6 0.87 0.18 23 23 25 24 77 75 189 40 603 87 Second column • 934 445 83 8.3 2.4 0.81 0.12 28 22 23 18 55 57 176 26 394 44 Third column 771 319 70 8.2 2.3 0.56 0.06 28 16 16 9 39 49 126 13 204 20 Fourth column 666 232 58 16 8.2 1580 11 22 2.3 0.40 0.06 27 11 19 9 29 40 87 12 78 8 85 ------- The concentration of color and TOG remaining in the water as a function of length of column is shown in Figure 12. This plot can be used to estimate the length of bed required at these operating conditions to reach any desired color or TOG concentration. For example, under these conditions of operation, the length of bed to reduce the color to 100 CU would be about 65 ft, or about 1.63 times that used in the pilot plant (40 ft) which reduced the color to an average of 230 CU. Alternatively, the color could have been reduced to a lower value if the contact time had been increased by reducing the flow rate, but this would have re- quired a much lower flow velocity in the carbon bed, which causes a reduced rate of adsorption (as discussed later in this section) and would also have required a greater amount of carbon per 1000 gal of treated water. In the case of TOG, Figure 12 shows that the original goal of 50 mg/1 could have been reached if the bed length had been increased to 48 ft. These plots show that, with the bio-carbon sequence, the reduction of color to a suitable level is more difficult than the reduction of TOG. This is true only of bio-carbon adsorption because the bio-oxidation in the aeration basin removed less of the color than TOG. In contrast, a greater percentage of color than TOG was removed by the lime treatment step. The concentration of BOD was reduced by 76% by bio-oxidation to 67 mg/1 and further reduced 76% to 16 mg/1 by the carbon columns, for an overall reduction of 94% by bio-carbon (see Table 22). The pH decreased from 8.6 to 8.2 as the water passed through the columns. As discussed later, this was perhaps caused by bio-activity in the carbon beds. The average turbidity decreased from 16 JTU in the feed water to the columns to 11 JTU in the water from the columns. The average temper- ature of the water in the columns was 74°F. DOSAGE OF CARBON REQUIRED One of the primary objectives of the pilot plant operation was to deter- mine the amount of fresh carbon or regenerated carbon that is needed per 1000 gal of water to provide water suitable for reuse in the mill. This information cannot be obtained from adsorption isotherms with sufficient reliability. Rather it must be obtained in continuous bench-scale or pilot plant columns and is determined from the average amount of water that can be treated per new column of carbon. The dosage rate for the test period of February 4, 1972 to .March 11, 1972, during which time four fresh columns were added, was 8 Ib of carbon per 1000 gal of water, or 940 mg/1. For the period of March 11, 1972 to March 18, 1972 the rate of addition of fresh columns was increased to determine how much the product water quality would be improved by the high dosage rate. On the basis of two column changes, the dosage rate for this period was 26 Ib per 1000 gal, or 3.3 times that of the low- dosage period. The results from these two periods are given in Table 23. 86 ------- 10 20 30 40 50 60 70 LENGTH OF BED, ft 80 Figure 12. Color and TOC remaining versus length of carbon bed during bio-carbon sequence 87 ------- Table 23. RESULTS FROM BIO-CARBON ADSORPTION WITH DIFFERING CARBON DOSAGE RATES Low carbon dosage rate 2/4 to 3/11/72 Conc'n. to aeration basin, CU or mg/1 to carbon columns, CU or mg/1 from carbon columns, CU or mg/1 Removal by carbon adsorption, % of feed to columns Rate of removal, CU/(g hr) or mg TOC/(g hr) No. of fresh charges of carbon (1600 Ib ea.) Cum. removals when discharged, CU/g or mg TOC/g Av. dosage rate, Ib carbon per 1000 gal water Color 1210 810 232 71 0.67 4 ; 646 8.0 TOG 277 149 58 61 0.11 - 98 - High carbon dosage rate 3/11 to 3/18/72 Color TOG 1210 532 716 161 226 49 68 0.46 2 327 26.0 70 0.10 42 Considering, of course, the short duration of the test of high carbon dosage rate, a few points may be noted. Somewhat surprisingly, these results show that the high dosage rate gave only slightly lower concentra- tions of color and TOG, and a mixed response in terms of percentage remov- als, rather than the expected strongly increased percentage removals of color and TOG. Because of these unexpected results, it may be speculated that perhaps a dosage rate even somewhat reduced below the 8 Ib per 1000 gal of water of the low-dosage period might have achieved nearly equal removals. However, this would not be expected on the basis of equilibrium adsorption isotherms. The average cumulative removal in the first column during operation at 8 Ib per 1000 gal was 646 CU/g, which is about 65% of isotherm equilibrium loading and is in line with percentage-of-equilibrium loadings achieved in other carbon column operations. LOADINGS ON CARBON The apparent loadings of color or TOG on the carbon are actually cumula- tive removals per unit weight of carbon. The cumulative removals are greater than the actual loadings of adsorbed impurities because of remov- als by bio-activity and because a substantial amount of the impurities removed from the water is not actually adsorbed within the pores of the carbon but is apparently held on the outer surfaces of the carbon. Much of this adhering impurity is removed during backwashing, as reported later under a discussion of backwashing. 88 ------- The cumulative weight of TOG removed by each column per unit of carbon, expressed as mg TOC/g carbon, is shown in Figure 13 as a function of operating time. The cumulative removals of TOG by four columns, by the time they were removed from operation, ranged from 70 to 130 mg/g of carbon during the period of February 4, 1972 to March 11, 1972, and the average removal was 98 mg/g, or 9.8% based on the original weight of carbon (see Tables 22 and 23). The cumulative removals of color for the same period ranged from 495 to 667 CU/g, for an average of 646 CU/g. Average cumulative removals during the period of March 11, 1972 to March 18, 1972, when the carbon dosage was raised to 26 lb/1000 gal, were expectedly lower, i.e., 42 mg/g TOG and 327 CU/g. RATES OF REMOVAL OF IMPURITIES BY CARBON The rates of adsorption of impurities on carbon have been studied by a number of investigators. The proposed rate equations based on adsorption of single compounds from solution tend not to fit operating data from complex multi-component systems. The rate of adsorption in weight of a given impurity removed per hour is usually expressed as a direct function of weight of carbon, concentration gradient driving force, temperature of the water, and as an inverse function of carbon particle size. Other factors affecting rate are the pore size and pore size distribution of the carbon, the molecular size and polarity of the material being adsorb- ed, the degree of turbulence around the particle, and the diffusivity of the impurity in the water. Development of an applicable rate equation for pulp mill effluent is complicated by the countless compounds in the water and the wide variability of the types of compounds. The concentra- tion gradient driving force used in rate equations is the difference in concentration of the water within the pores of the carbon, or its equilib- rium adsorption concentrations, and the concentration of the bulk solution. Again, the development of an applicable rate equation is complicated by the fact that the equilibrium concentration within the carbon is not known. The equilibrium concentration could be determined from an equilibrium adsorption isotherm for the same water and carbon if one knew the loading on the carbon. But the indicated loading or cumulative removal of the impurity per unit weight of carbon that is measured is not the same as the true loading, as discussed above. Because of these complicating factors, we used a simplified rate expres- sion which is the amount removed per weight of carbon per hour, or CU/(g hr) and mg TOC/(g hr). The rates were calculated (by the computer program) each day for each column from the amounts of impurity removed and the weight of carbon in each column. These rates are of value in comparing the effects of operating variables on rates and for designing full-scale plants. The rates are directly applicable only to water of the same composition and temperature, the same type and size of carbon, the same velocity and retention time of the water in the columns, and the same rate of addition of fresh carbon (dosage rate). 89 ------- td o oo o o H 140 120 8 10 12 14 16 18 20 22 DATE 24 26 28 3/1 11 Figure 13. Cumulative removals of TOG on carbon, Feb. 1 to March 11, 1972 ------- The rates of removal in the pilot plant columns during the bio-carbon sequence were highest in the lead column, in which the concentrations were highest, and were lowest in the last column in which the concentra- tions were lowest (see Table 22). During the period from February 4, 1972 to March 11, 1972, the average rates of removal by column position ranged from 0.87 CU/(g hr) for the first column to 0.40 CU/(g hr) for the fourth column, and the overall average rate was 0.67 CU/(g hr) (see Table 22). The corresponding rates of removal of TOG were 0.18 and 0.06 mg/(g hr), with an average of 0.11 mg/(g hr). These results are shown graphically in Figure 14 as a function of bed length. Note that the highest rates were observed with the lead column that had the most fully loaded carbon and contained the highest concentrations of impurities in the water. The rates of adsorption in a given column as it was moved through the four column positions are also indicated by the slopes of the cumulative removal curves of Figure 13. The rates of removal are plotted versus average concentration in each column in Figure 15. These curves show that rates are very dependent on bulk solution concentration. The effect of concentration would be much greater if the depressing effect of loading on the carbon had not been present. The equations for rates of removal of color and TOG for the curves given in Figure 15 are: Color removal rate = 0.523 In C - 2.55, for 250 to 800 CU TOG removal rate = 0.194 In C-- 0.77, for 70 to 150 mg/1 TOG Where rate is in CU/(g hr) or mg TOC/(g hr), C is average concentra- tion in CU or mg/1. As stated above, backwashing of the columns removed some of the color and TOG that had been loosely held by the carbon. There was no evidence that the rate of adsorption was increased by a higher frequency of back- washing, even though this would be expected. The rates of removal were less when the high dosage rate of carbon was used (March 11, 1972 to March 18, 1972), even though the cumulative removals, or degree of loading of the carbon, were only 50% of those during the low dosage period (February 4, 1972 to March 11, 1972) (see Table 23). It is postulated as one possible explanation, that the external coating of organic matter on the carbon, as discussed earlier, was an overwhelming rate-controlling mechanism, which nullified the effect of fresher carbon from the high-dosage operation. However, the short duration of operations at high dosage rates makes this comparison less than definitive. 91 ------- 00 ^x O *l w I 1.0 0.8 0.6 0.4 0.2 0 0.20 J» 0.15 3 0.10 O O o o H 0.05 603 715 87 125 i 533 394 204 CUM REMOVAL, CU/g 533 382 AVG CONC'N, CU 44 20 CUM REMOVAL, mg/g 95 78 9 AVG CONC'N, mg/1 I I i 10 20 LENGTH OF CARBON BED, ft 30 78 275 8 64 40 Figure 14. Rates of removal of color and TOG by carbon adsorption during bio-carbon sequence, Feb. 4 to March 11, 1972 92 ------- 1000 500 00 s u o H o iJ O w 200 100 50 30 0 I 1 1 1 1- 0.5 1.0 COLOR RATE, CU/ (g hr) 1 0.1 0.2 TOG RATE, mg/ (g hr) 1.5 I I I 1 L. 0.3 Figure 15. Rates of removal of color and TOG by carbon columns as a function of average concentration during bio- carbon operations, Feb. 4 to March 11, 93 ------- EFFECT OF FLOW VELOCITY ON RATES OF ADSORPTION The superficial (empty-column) velocity of water through the columns at 15 gpm was 2.1 gpm/ft2. A lower velocity would give a longer contact time in the four columns for a greater reduction of color, but it was not known whether the lower velocity would cause lower rates of removal per unit of carbon due to an increase in the thickness of the stagnant layer of water around each carbon particle. Special runs were made at 0.7 and 1.75 gpm/ft which indicated that at the higher velocity of 1.75 gpm/ft the rate was only slightly higher than at 0.7 gpm/ft . The results of these tests are discussed more fully later in conjunction with tests made during the primary-carbon sequence of operation (see Section X). BIOLOGICAL ACTIVITY IN CARBON COLUMNS There was evidence of some aerobic and anaerobic bio-activity in the carbon columns, as is more fully discussed under the primary-carbon sequence in Section X. The average dissolved oxygen content of the feed to the columns during the bio-carbon sequence was 6 mg/1 but de- creased rapidly to less than 1 mg/1 in the water from the first column. However, very little removal of TOG (about 3 mg/1) can be attributed to biological consumption of the 6 mg/1 of dissolved oxygen based on stoi- chiometric considerations. Most of the aerobic activity appeared to be concentrated in the top one or two feet of the first column, evidenced by caking of carbon particles and column plugging unless the columns were backwashed every one or two days. Anaerobic activity was also present in the columns, as indicated by bacterial analyses made near the end of the bio-carbon operation. The total bacterial count was low in the mill effluent to the basin and very high from the basin. The concentration was reduced from 3.6 x 10 to 2.1 x 10°/ml by the sand filter. The concentration was further reduced to 1 x 106/ml in passage through the carbon columns. All but about 0.1% of the bacteria were retained on a 0.8 micron filter. Hydrogen sulfide producing bacteria were present in the feed to the columns, but the water from the columns had very low concentrations of these bacteria. Occasionally, there was evidence of H2S odor in the water from the columns. Anaerobes were found in the water to and from the columns. BACKWASHING OF CARBON COLUMNS The columns were backwashed by a high rate of flow of compressed air for about 20 sec followed by an up-flow of 30 gpm of well water for a period of 20 minutes every other day. The backwash water was black and contain- ed about 10% of the color and TOG that had been removed by the columns 94 ------- during the two days of adsorption. When the columns were backwashed at the two-day frequency, the average pressure drop through the four columns in series was about 10 psi before backwashing and about 3 psi after back- washing. CORRELATION OF VARIABLES The interactions of 12 independent variables with 18 dependent variables during the bio-carbon operation of February 4, 1972 to March 11, 1972 at 15 gpm were examined by a computer regression analysis. The results of this analysis generally confirmed the conclusions from the direct analysis of the data. The 12 independent variables included temperature, pH, and concentrations of color and TOG to the columns, and cumulative removals of color and TOG from each of the four columns. (The term cumulative removal is used in preference to loading for reasons stated earlier in this section.) The 18 dependent variables included color and TOC concentrations from each column, removal rates of color and TOC for each column, and percentage removals of color and TOC from the collective columns. This regression analysis of the data showed that the color of the water from the fourth column was largely a direct function of the color to the columns, the cumulative removals of color and TOC by the first two columns, and the cumulative removal of TOC of the fourth column, i.e., the product color increased as the feed color increased and as the cumulative removals increased. The color from the other columns was primarily a function of the color in the feed to the first column. The percentage removal of color by the four columns was almost entirely explained by the cumulative removals of color by the four columns, i.e., the percentage removal de- creased as the cumulative removals increased. The TOC in the water from the fourth column was primarily a direct function of the cumulative removals of TOC by the first and fourth columns and of the TOC concentrations of the feed water to the first column. The rates of removal of color by the columns were influenced as an inverse function of the cumulative removals by the columns. The rate of removal of TOC by the group of four columns was strongly influenced (directly) by the TOC concentration of the feed water and to a lesser degree by the color of the feed water, and was weakly influenced (inversely) by the cumulative removal of color by the first column. The variation in temperature during this period of operation was not very great: 66% of the values fell between 81 and 67°F, with the average being 74°F. In this range, temperature had no significant influence on adsorption performance. 95 ------- SECTION X OPERATING RESULTS: PRIMARY - CARBON COLUMNS OBJECTIVES AND DESCRIPTION OF OPERATION Objectives The objectives of the primary-carbon operation were to determine whether carbon adsorption without pretreatments by bio-oxidation or lime treatment could remove the much larger amounts of color and TOG at high enough loadings on the carbon to give reasonable values of carbon dosage (regeneration rates) and sufficiently low capital and operating costs to make this sequence competitive with the other two sequences. Preliminary economic estimates (see Section V) had indicated that the primary-carbon sequence might be competitive with the lime- carbon sequence. Since the concentrations of color and TOG would be high, the rates of adsorption and the loadings on the carbon were ex- pected to be considerably greater than with the other sequences. Other objectives were to determine the flow velocities and contact times needed for removal of impurities and to determine whether this sequence of operation would give unique problems, such as excessive bio-activity within the columns and greater rates of pressure-drop increase. Operating Periods The primary-carbon sequence was in operation from April through the first week in September, 1972. Four columns were operated in series in downflow, eight weeks at 10 gpm, and four weeks at 5 gpm. Carbon columns were replaced regularly throughout these periods. Two columns were operated in series for the subsequent 11-week period in downflow at 10 gpm without any further carbon changes to observe the long-term capacity of the system for removing color and TOG. PRIMARY CLARIFICATION AND FILTRATION The basin was used for primary clarification and for equalization of impurity concentrations in the mill effluent before it was fed to the carbon columns. Water was pumped into the basin at a rate of 15 gpm which gave a retention time of about 8 days. As pointed out in Section VII on Primary Clarification and Bio-Oxidation, anaerobic activity in the basin caused an increase of low-molecular-weight compounds that are 96 ------- not absorbed by activated carbon. The average TOG content of these compounds was 48 mg/1 in the water to the basin and 79 mg/1 in the water from the basin, for an increase of 65%. Even though the equalization basin provided a retention time of 8 days, the feed to the columns varied from 1550 to 540 CU and 410 to 170 mg/l' TOG during the period covered by these results. The rates of change of concentrations in the feed to the columns were generally less than 100 CU per day and 20 mg/1 TOG per day. These fluctuations in feed concentration introduce a high degree of complexity into the correlation of operating results with operating conditions. In a commercial installation for treating effluent by lime treatment or carbon adsorption, or both, effluent concentrations to the treatment system must be held in a narrower range, and rapid changes in effluent concentrations must be avoided through in-plant control measures or external spill basins and equalization basins. The suspended solids in the water to the basin varied widely - from 37 to 790 mg/1. The water after primary clarification contained from 50 to 265 mg/1 and averaged 168 mg/1 of TSS. The turbidity of the water from primary clarification ranged from 18 to 23 JTU. The water to the columns was not sand-filtered prior to the carbon columns except during the first three weeks. Insufficient data were obtained to determine the amount of suspended solids removed by the duomedia filter when it was being used. The pressure drop across the columns before each backwash remained the same as when the water was filtered. Therefore, we concluded that filtration ahead of the columns is not required if the effluent to the columns contains less than about 200 mg/1 suspended solids and the columns are backwashed on a schedule as discussed later in this section. This suspended solids level may well be achievable, depending on the mill, by grit and floating solids (bark) removal alone or by clarification in a relatively small clarifier at high overflow rate. REMOVALS OF COLOR AND TOG BY CARBON ADSORPTION The results from carbon adsorption following primary clarification (primary-carbon) are summarized in Table 24 for two periods operating at two different flow rates (10 and 5 gpm). Detailed data for each of the four columns are given in Table 25 for the operating period at the higher flow rate (10 gpm). A printout of daily summaries of conditions and results is given in Appendix E. The product water during the 10-gpm run contained, on the average, 185 CU and 83 mg/1 of TOG. These values are somewhat greater than may be desired for reuse in the mill. The relationship of product concentrations to 97 ------- Table 24. SUMMARY OF PRIMARY CLARIFICATION - CARBON ADSORPTION SEQUENCE Dates 4/25-5/20 5/20-6/22 Flow rate, gpm 10 5 Flow rate, gpm/ft2 1.42 0.71 Flow rate, vol water/vol carbon/hr 0.28 0.14 Contact time, hr 3.5 1.8 Number of fresh charges of carbon 4 3 Cummulative removal when discharged CU/g carbon 350 360 TOC, mg/g carbon 61 70 Average dosage rate, Ib carbon per 1000 gal 20.5 28 Color removal Feed to carbon columns, CU 791 997 Product from carbon columns, CU 185 202 Removal by carbon, % 77 80 Rate of removal by carbon, CU/(g hr) 0.69 0.46 TOC removal Feed to carbon, TOC, mg/1 220 310 Product from carbon, TOC, mg/1 83 121 Removal by carbon, % 62 61 Rate of removal by carbon, mg/(g hr) 0.11 0.08 98 ------- Table 25. AVERAGE RESULTS FOR PRIMARY-CARBON SEQUENCE FOR PERIOD OF APRIL 4, 1972 TO MAY 20, 1972 AT FLOW RATE OF 10 GPM Temp, water to basin, F to columns, F Color, to columns, FPCU CU from columns, FPCU CU TOG, to columns, mg/1 from columns, mg/1 pH from columns Cond., micromhos Turbidity, JTU Susp. solids, mg/1, from basin from cols. Pressure drop, psi Color removal rate, CU/(g hr) TOC removal rate, mg/(g hr) Color removal, 70 to ea. col. Color removal, % to lead col. TOC removal, % to ea. col. TOC removal, % to lead col. Color remaining, 7o to lead col. TOC remaining, % to lead col. Color removal, difference, CU TOC removal, difference, mg/1 Cumulative color removal, CU/g Cumulative TOC removal, mg/g General a.v. av. tot. tot. tot. tot. data 114 76 1474 791 220 168 0.69 0.11 77 62 606 137 Lead column 815 483 126 10.0 876 20 1.0 1.66 0.30 39 39 43 43 61 57 308 94 270 45 Second column 588 322 110 10.0 850 18 1.0 0.60 0.05 33 20 12 7 41 50 161 16 101 16 Third column 483 259 99 9.9 840 17 1.0 0.24 0.03 20 8 9 5 33 45 63 11 48 9 Fourth column 368 185 83 9.4 830 17 85 1.0 0.27 0.05 29 9 17 8 23 38 74 16 21 5 99 ------- length of column is given in Figure 16. By extrapolation, the length of column needed to reach 100 CU at 10 gpm and under the same other conditions of operation would be about 60 ft, or about 50% more than was used in the pilot plant runs. It is assumed in this extrapolation that the compounds that would be removed at the lower concentrations would have the same adsorption properties as those already adsorbed. This assumption may not hold for the low-molecular-weight compounds that make up a portion of the TOG. To reach a color of 100 CU, the inventory of carbon would have to be increased by 50% and the carbon dosage rate would have to be increased by about 14% for the increased amount of color removed. The lead column removed a large portion of the total color and TOC removed by the four columns. As shown in Table 25 during the 10-gpm run, the lead column removed 39% of the feed color, the second column removed 20%, the third removed 8%, and the fourth removed 9%. The lead column removed 51% of the total color removed by all columns and 70% of the total TOC removed. The possible mechanisms accounting for this large percentage removal in the first column are given later in a discussion of the ultimate removal capacity and mechanisms in carbon columns. In contrast to the above large removals in the first column during primary- carbon operation, the first of the four columns in the bio-carbon opera- tion removed only 33% of the total color that was removed and 40% of the TOC. The lower removals in the lead column in the bio-carbon sequence might be due to the organic compounds being less bio-degradable after the bio-oxidation treatment and that bio-activity was greater in the primary-carbon columns. The removal of BOD was not determined for the primary-carbon sequence of operation. DOSAGE OF CARBON REQUIRED The most reliable way to determine the carbon dosage rate with fixed-bed columns is to continue a pilot plant operation until the number of fresh columns added are two greater than the number of columns in series, e.g., six for the primary-carbon operation. The dosage rate is then found by dividing the weight of carbon in each column by the average volume of water treated between column changes. In this primary-carbon run at 10 gpm, we added fresh columns six times and discarded the data from the first two column changes because these columns had been used during the bio-carbon sequence and thus the data would not be typical of results from the primary-carbon sequence. The time interval between the next four column changes ranged from 92 to 160 hr and averaged 130 hr. Therefore, the dosage rate was: 1600 Ib carbon per column change x 1000 _ 20.5 Ib 130 hr x 600 gal/hr ~ 1000 gal 100 ------- 1500 10 20 30 40 50 60 LENGTH OF BED, ft 70 80 Figure 16. Color and TOG remaining versus length of bed during primary-carbon sequence 101 ------- During the 5 gpm run, five columns were added and the average time between changes for the last three columns added was 251 hr and the average dosage rate was 28 Ib carbon per 1000 gal. RATES OF ADSORPTION Adsorption Rates During 10 gpm Run The average rates of removal by each column for the 10 gpm run, April 25, 1972 to May 20, 1972, are given in Table 25 and are plotted in Figure 17 as a function of average concentration of the water in each column. The rates during the bio-carbon sequence are also shown. The rates of removal of color and TOC appear to be linear with the log of concentration over most of the concentration range. A discus- sion of factors affecting adsorption rates is given in Section IX on the bio-carbon sequence of operation. In the case of the primary-carbon data, the average rates for the third column were lower than those for the fourth column which is surprising because the fourth column had a lower average concentration. The rates of removal by primary-carbon were lower than those of the bio-carbon sequence at concentrations below 130 mg TOC/1 and 330 CU but were higher at the higher concentrations. The higher rates in the lead column of the primary-carbon sequence were evidently caused by non-adsorptive phenomena discussed later in this section. The equation for the relationship between the rate of removal of color and the concentration during the 10-gpm run with primary-carbon se- quence is as follows: Rate, CU/(g hr) = 1.83 In C - 10.2 where C = average color concentration in the columns, CU This equation is only approximate and is s-trictly applicable only for a color range of 300 to 700 CU, for water with the same properties as was used in the pilot plant, and for the same other operating con- ditions used in the pilot plant run. The equation for TOC removal rates, which is limited to the TOC range of 100-200 mg/1 and to the conditions used in the pilot plant run is: Rate, mg TOC/(g hr) - 0.55 In C - 2.5 where C = average TOC concentration in the columns, mg/1 Effect of Velocity on Adsorption Rates The effect of velocity on adsorption rates was investigated during the bio-carbon and the primary-carbon sequences of operation. The EPA 102 ------- 1000 500 o o H O o u w w 200 O / X BIO-CAKBON BIO-CARBON 100 30 0 1 . . 0.5 COLOR i . 1 i RATE, , , CU/ i 1. (g | 0 hr) i i i 1.5 , 1 , 0.1 0.2 TOC RATE, mg/(g hr) 0.3 Figure 17. Rate of removal of color and TOC by carbon adsorption during primary-carbon operation at 10 gptn 103 ------- Process Design Manual for Carbon Adsorption (19) stated that several studies have found that velocity in the range of 2-10 gpm/ft2 is not a limiting factor in fixed-bed granular carbon adsorption. Other investigators (9,17) found no effect in the 4-10 gpm/ft2 range. The effect of velocity was studied at the end of the bio-carbon oper- ation. The last two columns were used, one with a flow of 5 gpm (0.7 gpm/ft2) and the other with a flow of 12 gpm (1.75 gpm/ft2) with the same water being fed to both. Samples of treated water were removed at different lengths of the carbon beds to give equal retention times of water in the columns. After four days -of operation, the conditions were reversed so that the previous high-velocity column was then the low-velocity column. This switching of conditions should remove the effect of the degree of loading in each column from the comparison. Results showed that the percentage removals and rates of removals were the same for color removal and about 10% greater for TOG removal in the case of the high velocity run. Therefore, this test indicated that an increase of velocity from 0.7 to 1.75 gpm/ft was beneficial by a slight amount. The influence of velocity was also studied during the primary-carbon operation by comparing results from periods of operation at 10 gpm (1.4 gpm/ft2) and 5 gpm (0.7 gpm/ft2), as reported in Table 24. The feed during the 5 gpm run was unfortunately 25% greater in color and 40% greater in TOG concentration. The average product concentrations were also greater for the 5 gpm run, even though the percentage removals of color (77 and 80%) and of TOG (61 and 62%) were about the same for the two periods. Since rates of removal are highly dependent on the concentration of the impurity in the water in the adsorption beds, the rates of removal at the two velocities were compared at the same concentrations on a plot of rate versus average concentration in each column. The rates of removal of color obtained at 10 gpm for April 25, 1972 to May 20, 1972 were two times those at 5 gpm when they are adjusted to the same water concentration. The rates of removal of TOG at 10 gpm were more than three times those at 5 gpm at the same water concentrations. These results indicate that rates of adsorption were substantially greater at a superficial velocity of 1.4 gpm/ft than at 0.7 gpm/ft . This effect of velocity in the primary-carbon sequence was much more pronounced than was observed in the bio-carbon sequence discussed above. We conclude that commercial plants should use a velocity of at least 1.4 gpm/ft2 and perhaps as much as 2-4 gpm/ft2 subject to limitation by the cost of pumping. 104 ------- The pumping energy requirement and pressure drop through the columns increase in direct proportion to the lineal velocity of flow. To maintain the same color removal, the length of column needed'will also by increased in direct proportion to the velocity cut will be decreased by the greater adsorption rate afforded by the greater velocity. The net effect of increased velocity on energy requirements will lie between velocity to the first power and velocity squared. The pumping energy required at 4 gpm/ft2 to remove color from 1000 to 100 CU was estimated to be only about $0.004/kgal and the total pres- sure drop was estimated to be about 40 psi. CORRELATION OF VARIABLES The interactions of 12 independent variables with 18 dependent variables during the primary-carbon operation of April 25, 1972 to May 20, 1972 at 10 gpm were examined by a computer regression analysis. The results of this analysis generally confirmed the conclusions from the direct analysis of the data. The 12 independent variables included temperature, pH, and concentrations of color and TOC to the columns; and cumulative removals of color and TOC from each of the four columns. The 18 depen- dent variables included color and TOC concentrations from each column, removal rates of color and TOC for each column, and percentage removals of color and TOC from the collective columns. This regression analysis of the data showed that the percentage removal of color by the four columns was almost entirely explained by the (inverse) effect of cumulative removals of the columns. The cumulative removal of the first column had the greatest influence, primarily because this column removed more of the total color than the other columns. The TOC removal rates were only weakly influenced by the various conditions (lineal velocity was constant during this operating period). The rate was influenced by color and TOC concentrations of the feed (higher concen- trations caused higher rates) and to a lesser extent by the cumulative removals of the carbon (higher cumulative removals caused lower rates). The overall color removal rates were primarily a (direct) function of color concentration in the feed. In the second and third columns the rate was still largely dependent on feed color concentration but was also directly related to cumulative removals. Therefore in these two columns, the rate increased as the cumulative removals increased, for unknown reasons. In the first and fourth columns, the rate decreased as the cumulative removals increased as would be expected. The rate in the fourth column was primarily due to the cumulative removals of color and TOC of the first column as well as that of the fourth column. 105 ------- The concentration of TOG from the fourth column was primarily a (direct) function of cumulative removals of color and TOG by the first and fourth columns and secondarily a (direct) function of the color and TOG concen- trations in the feed. The color from the fourth column was primarily a (direct) function of the color of the feed and of the cumulative removals of TOG and color by the second and third columns. PRESSURE DROPS AND BACKWASHING OF COLUMNS During essentially all of the primary-carbon runs, from April 4, 1972 to June 22, 1972, the four columns were backwashed every other day, and the water fed to the columns from the equalization basin was not filtered after the first month of operation. During the period of June 22, 1972 to July 8, 1972, backwashing was discontinued to determine how rapidly the pressure drop across the columns would increase and to determine whether a reduced frequency of backwashing could be used. The flow rate during this 9-day test period was 10 gpm through the four columns in series. There was no substantial increase of pressure drop across the four columns; the final pressure drop was only 3 psi. The columns were all backwashed and two columns were then operated at a flow of 10 gpm and were backwashed once per week. After two weeks of operation, the pressure drop increased to 8 psi across the two columns and to 44 psi after four weeks. An inspection of the tops of the beds showed that the top two feet of both carbon beds had become almost solid with particulates removed from the water and with biological growth. The hard layer was broken up with hand tools and with vigorous backwashing. The run was then continued but with a backwashing frequency of once every two days. The pressure drop per column just before backwashing varied from 1 to 30 psi, but averaged about 4 psi. The amount of lost production caused by backwashing amounts to 20 minutes each time or 10 minutes per day, when backwashing every other day, or only 0.7% of the normal daily production. The volume of backwash water for 15 minutes of backwashing at three times the normal flow rate would amount to 23 minutes of production per day, or 1.6% of the daily production. However, it is expected that the water from backwashing would be reused for backwashing after being settled or treated by coagulation to remove most of the material displaced during backwashing. The volume of water needed to be held for backwashing could be further reduced by backwashing several columns in series. 106 ------- DISPLACEMENT OF TOC AND COLOR FROM COLUMNS DURING BACKWASHING The water coming from the columns during backwashing was very black and contained a substantial amount of color and TOC. Tests were run to determine the quantities of color and TOC that were displaced during backwashing of the columns. Four of the tests were during the primary- carbon runs and one during the bio-carbon runs. In these tests the flow of water from the column being backwashed was held constant at 30 gpm and samples of the backwash water were taken about five times during a backwash period of 20-60 minutes. In one of the tests that was continued for 60 minutes, it was estimated that a total backwashing time of 110 minutes would be required before the backwash water would no longer contain an appreciable amount of color or TOC and that approximately half of the total displaceable materials were displaced in the normal backwash period of 20 to 30 min- utes. The backwash from the lead column did not contain much greater concentrations .of color and TOC than that from the second or third columns. The average results from these five tests given in Table 26 show that backwashing displaced about 11% of the color and TOC that had been removed by the carbon since the prior backwashing. The ratio of filter paper color (FPCU) to Millipore color (MPCU) is a measure of the size of the colloids that make up a portion of the color. During the period of maximum color displacement by backwashing, the ratio of FPCU/MPCU was 3.7 but decreased to 1.7 for the average backwash water. The amount of color in the size fraction less than filter paper and greater than 0.8 micron Millipore size was 274% of that in the "true" color range of minus 0.8 micron Millipore filter. During the average backwashing period this percentage decreased to 65%. These results indicate that the color bodies displaced during the first part of the backwashing period are much larger in size than those displaced later in the period and that a large portion of the apparent color (FPCU) is not "true" color (MPCU). The color bodies in the more highly colored initial fraction of backwash water contained 18 FPCU per mg of TOC; whereas the average backwash water contained 9 FPCU per mg of TOC. The feed water to the columns during the primary-carbon operation contained only about 5 FPCU per mg of TOC. This comparison of FPCU/TOC ratios indicates that the color bodies that are deposited during normal operation and displaced during backwashing are not only much larger than the color bodies in the feed water but are much more intense in color per atom of carbon they contain. 107 ------- Table 26. DISPLACEMENT OF COLOR AND TOC DURING BACKWASHING OF COLUMNS Color TOC, Ratio FPCUb MPClF mg/1 FPCU/MPCU FPCU/TOC Average3 max. concentration 3674 983 210 3.7 17.5 Average concentration 1213 734 137 1.7 8.9 Amt. displaced per g carbon 8.3 3.6 0.56 Hrs. of loading removed in 4.1 9.4 backwashing Amt. displaced in backwash- ing as % of amt. removed in normal operation since last backwash 11 11 a Average of results from backwashing of 8 columns in 5 tests for 20 to 60 minutes each. Measured after filtration through Whatman No. 2 paper filter. c Measured after filtration through 0.8 micron Millipore filter. NON-ADSORPTIVE REMOVAL MECHANISMS IN CARBON COLUMNS The objective of the third operating period of the primary-carbon sequence was to determine the residual removal capacity of the two most spent col- umns without further carbon replacement. A potentially significant resid- ual removal capacity of the carbon had been inferred during preceding operations when significant removals were still being obtained from the first or lead column (which is the most nearly spent column) just prior to its replacement as necessitated to maintain desired quality of effluent from the fourth column. This prolonged test of residual removal capacity was continued 15 times the normal length of time between column changes, with backwashing. This run was carried out with the third and fourth columns from the pre- vious primary-carbon runs using primary clarified water from the basin at a flow of 10 gpm to the columns. The columns were backwashed every two days. During this run, no fresh carbon was added. The run was contin- ued for a total of 74 days from June 26, 1972 to September 9, 1972 until the rates of removal appeared to have reached a constant value. After the first 23 days, the color removal rates for both columns had decreased significantly and were about 0.5 CU/(g hr), but the TOC rates had decreased only to about 0.15 mg/(g hr). During the following 21 days, TOC removal rates continued to drop, while color removal remained essentially steady. During the final 30 days of the run, both color and TOC removal were fairly constant and were about the same for the second column as for the first column. The average final equilibrium removal rates were 0.5 CU/ (g hr) and 0.06 mg TOC/(g hr). As shown in Table 27 these rates were 20-30% of those in the first column during the primary-carbon run of April 25, 1972 to May 20, 1972, also at 10 gpm. 108 ------- Table 27. COMPARISON OF REMOVAL RATES OF EXHAUSTED CARBON WITH NORMAL OPERATION Removal Rates a , , Color, CU/(g hr) TOG, mg/(g hr) Final exhaustion run 8/9 to 9/9/72, 10 gpm 0.5 o 06 Adsorption runb 4/25 to 5/20/72, 10 gpm 1^66 o!30 Exhaustion run rate as percent of adsorption run rate 30% 20% a For both columns k First two columns During the final 30 days of this run, fairly wide fluctuations in concen- trations of the water to the columns had very little effect on removal rates, compared to a strong dependence during normal operation. Also, removal in both columns proceeded at equal rates, although water concen- trations decreased from the first to the second column. In normal operation, removal rates in the second column are substantially below those in the first column, as seen in Table 25. By the beginning of the final 30 day period, cumulative removals had increased to levels well beyond equilibrium loadings estimated from a number of isotherms on similar effluent (see Section XIII, Figures 26 and 27). Cumulative removals actually amounted to the following multiples of equilibrium loadings: color, first column, 3.5, second column 3.5; TOG, first column, 1.7, second column, 1.8. Allowing for partial displacement during backwash as discussed earlier in this section would reduce these factors by a maxiumum of 20%. These data and observations suggest that during this extended run, adsorptive TOG removal had essentially ceased by the beginning of the final 30-day period, while adsorptive color removal had ceased prior to that time. Continued removal of both color and TOG therefore requires explanation by other mechanisms. These mechanisms are probably also operative during normal operation. Potential mechanisms include bio- logical activity, filtration, and colloid destabilization. Evidence of bio-activity was found in samples of carbon from the columns containing high concentrations of organisms similar to those found during bio-carbon operation discussed in Section IX. Biological mechanisms are generally believed not to effect color removal from kraft effluent. On the other hand, removal of COD and TOG by bio-activity in carbon columns has been found to amount to an appreciable portion of total COD and TOG removals in work with municipal effluents (6,16,37). The dissolved oxygen in the column feed water during the primary-carbon operation, 109 ------- including the extended run discussed here, was always less than 1 mg/1, hence any bio-activity must have been anaerobic, after consumption of the restricted amount of oxygen in the top one to two feet of the first column. The reported rates of TOG digestion in anaerobic sludge digestors for a large number of installations were equivalent to 0.01 to 0.02 mg/ (g hr) when adjusted to the volumes of solids and the temperatures in the carbon columns. Therefore, anaerobic activity could account for 0.01 to 0.02 mg TOC/(g hr) out of a removal rate of 0.06 mg TOC/(g hr) during the final period, that is 17 to 33% of the total TOC removal. Filtration undoubtedly occurs in the carbon columns (the sand-filter was not in use during the extended run), as evidenced by the reduction in suspended solids (see Table 25). However, it must be borne in mind that TOC is measured on a paper-filtered (about 2 micron) basis, while color is measured on a Millipore-filtered (0.8 micron) basis. Hence, any explanation of color removal by filtration in the column implies the removal of particles less than 0.8 microns in size, which are extremely difficult to filter, or less than 2 microns in the case of TOC. A mechanism of colloid destabilization and coagulation in conjunction with filtration is suggested to explain the unexpected removals of color and TOC. This mechanism is further discussed in Section XIII. Desta- bilization may occur due to removal of stabilizing substances by adsorp- tion or bio-activity, or by a more direct functionality of the carbon surface. The destabilized colloidal particles coagulate and are then more readily trapped by the carbon in the column by a filtration mech- anism. Such filtered substances would then be subject to displacement from the column by backwashing. The displacement by backwashing amounts to about 20% of the color and TOC removed from solution between backwashes during normal operation. During the extended operating phase (final 30 day period), this amount is equivalent to displacing by backwashing 66% of the color and 100% of the total TOC removed from solution between backwashes. During the latter period, the backwash water did contain substantial amounts of black suspended materials, however, no quantitative data on these were obtained. Undoubtedly, a number of different mechanisms are at work simultaneously. Similarly, during normal operation removals are probably increased above strictly adsorptive removal by the non-adsorptive mechanisms in evidence during the extended run. It may be argued that during normal operation the rates of non-adsorptive removal are as high as those shown above, i.e., that 20% of the TOC removal normally observed is non-adsorptive, consist- ing perhaps of up to 7% by bio-activity and 13% by the suggested colloid destabilization - filtration mechanism. At the same time, 30% of the color removal is non-adsorptive and is removed by the colloid destabilization - filtration mechanism if bio-activity may be held (as is generally done) not to affect color. 110 ------- SECTION XI OPERATING RESULTS: LIME - CARBON COLUMNS OBJECTIVES AND DESCRIPTION OF OPERATION The sequence of lime treatment followed by carbon adsorption (lime-carbon) was evaluated while treating mill effluent that was passed through the equalization basin and lime treatment at a flow of 22 gpm. The operation of lime treatment was described in Section VIII. The lime-treated water was fed to two carbon columns in series at a flow of 10 gpm, or 1.4 gpm/ft The total height of carbon bed was 20 ft, which provided a superficial contact time of 1.8 hr and a bed volume rate of 0.55 volume of water per volume of carbon per hour. The columns were backwashed every two days. The major objective of this lime-carbon sequence of operation was to determine the operating performance of the carbon columns with mill efflu- ent treated by the microlime process, as discussed in Section VIII. The goal of the microlime treatment was to use dosages of lime sufficiently low to provide treated water with such low dissolved calcium content that there would be no need for carbonation. This level of dissolved calcium was defined as about 80 mg/1 Ca, or a range of 60-100 mg/1. On the other hand, the microlime treatment was expected to be sufficient to make any remaining color and TOG readily removable by carbon column adsorption. REMOVALS OF COLOR AND TOG; EFFECT OF SOLUBLE CALCIUM CONCENTRATION The microlime-carbon treatment was shown to be capable of reducing color to less than 100 CU and TOG to about 100 mg/1, when the level of soluble calcium was in the 50-100 mg/1 range. The percentage removals of color and TOC by carbon adsorption increased and the concentrations of color and TOC in both the feed and product water from the columns decreased as the soluble calcium concentration of the water from lime treatment to the carbon columns increased. Removals of color and TOC appeared to be independent of pH in the range of 7 to 12. Final color and TOC levels obtained in the combined lime-carbon treatment are shown in Figure 18 as a function of dissolved Ca in the column feed. Each point represents a run consisting of consecutive daily results with- in a fairly narrow range of dissolved calcium. Runs immediately after column changes are identified as squares in this figure (three runs). Ill ------- 1-4 O CJ I en § 200 r-l "{If 150 • 100 u 8 50 0 400 300 M O E-! ^ o o 200 100 0 CK O i Q o •o- o CT 20 40 60 80 100 120 140 160 'SOLUBLE CALCIUM IN WATER TO COLUMNS, mg/1 Figure 18. Effect of calcium concentration on removal of color and TOG by carbon adsorption (averages from operating periods) 112 ------- The data in this figure include all runs regardless of whether or not carbonation was used after lime treatment, using for correlation in all cases the dissolved calcium in the column feed. As this figure shows, both color and TOG tend to level off when more than a certain dissolved calcium concentration is maintained, i.e., the color was about 60 CU at calcium concentrations above 80 mg/1, and TOC was about 100 mg/1 with calcium concentrations above 40 mg/1. Other factors such as service age of the carbon and concentrations of the feed to the lime treater caused some fluctuations from the levels stated. Since the soluble calcium content of the water to the columns (and the related concentrations of color and TOC) was found to have a major influ- ence on the adsorption of color and TOC from the water, data from all of the lime-carbon operation were grouped into ranges of soluble calcium in the water to the columns. The operating conditions and results of the lime-carbon operation are given by this method of grouping in Tables 28 and 29. Additional results from the carbon adsorption step are given in Table 30. A printout of daily summaries and conditions is given in Appendix G. In the low-calcium range of 10-50 mg/1, which averaged 42 mg/1, carbon adsorption reduced the color from 276 to 214 CU, or by only 22%. At the medium-calcium range of 50-100 mg/1, which averaged 86 mg/1, good removals of color were obtained in the carbon columns, with the feed color reduced from 252 to 76 CU. This removal was 71% of the color to the columns (Table 30) and was 21% of the color to the lime treater. The total removal by the lime-carbon sequence in the medium lime range was 91%. Color measured after filtration through 2-micron paper filters was found to give valuable insight into the colloidal nature and size of the color bodies. A comparison of colors of product water after being filtered through filter paper and through 0.8 micron Millipore filters shows that a large portion of the apparent (unfiltered) color of the treated water at low calcium levels was caused by colloidal color bodies or colloidal calcium carbonate smaller than about 2 microns and larger than 0.8 micron. TOC removal was influenced less than color removal by the calcium concen- tration, probably because a larger portion of the TOC is made up of com- pounds that do not react with calcium. With the medium-calcium runs, the TOC was reduced to 100 mg/1 TOC to give a reduction of 44% of the TOC in the feed to the columns and 28% of the feed to lime treatment. The total removal of TOC in lime-carbon treatment was 63%. The variation of color and TOC concentrations of water with length of carbon adsorption bed are shown in Figure 19 for the three groups of runs. The slope of these lines is steeper as length of column increases, thus> in contrast to the results with the primary-carbon runs, the last (second) 113 ------- Table 28. LIME-CARBON SEQUENCE OF TREATMENT: REMOVALS OF COLOR AND TOC Low calcium Med. calcium High calcium Soluble Ca to cols., mg/1 range 10-50 50-100 100-200 avg. 42 86 138 Lime dosage, Ca(OH)2, mg/1 592 692 709 CaO, mg/1 448 523 536 pH of water to columns 9.7 11.3 11.6 Color CU to lime treatment, 861 852 937 to carbon columns 276 252 200 from carbon columns 214 76 61 TOC, filtered, mg/1 to lime treatment 225 272 275 to carbon columns 165 177 165 from carbon columns 118 100 98 % removal of color to lime trt. in lime treatment 68 70 79 in carbon columns 7 21 15 total 75 91 94 7o removal of color to columns 22 70 70 % removal of TOC to lime trt. in lime treatment 27 35 40 in carbon columns 21 28 24 total 48 63 64 % removal of TOC to columns 29 44 41 114 ------- Table 29. SUMMARY OF CARBON ADSORPTION FOLLOWING LIME TREATMENT: REMOVALS OF BOD AND TURBIDITY Soluble Ca to cols., mg/1 range avg. Lime dosage, Ca(OH)25 mg/1 pH of water to columns BOD, mg/1 to lime treatment to carbon columns from carbon columns 70 removal of BOD to lime trt. in lime treatment in carbon columns total Turbidity, JTU to lime treatment to carbon columns from carbon columns Low calcium Med. calcium High calcium 10-50 42 592 9.7 234 212 128 9 36 45 26 52 30 50-100 86 692 11.3 166 150 130 10 12 22 24 30 12 100-200 138 709 11.6 254 235 171 8 25 33 24 50 7 ------- Table 30. CARBON ADSORPTION TREATMENT OF LIME-TREATED EFFLUENT FLOW = 10 GPM Low calcium No. of days Soluble Ca Temperature Hours since pH Color, FPCU CU TOC, mg/1, BOD, mg/1 averaged to cols., mg/1 range avg. from cols. °F fresh col., 1st col 2nd col , 1st col. 2nd col. , 1st col. 2nd col. 1st col. 2nd col. Turbidity, JTU Conductivity, micromhos 7» removals Color, % of in columns color to 1st col. total TOC, % of TOC to 1st col. BOD removal total , total, %a 21 10-50 » . To 1000 276 165 212 52 2245 1st 8 16 42 81 1390 796 9.7 From 897 649 253 214 139 118 128 30 2nd 14 22 13 29 40 Med . calcium 30 50-100 To 1092 252 177 150 30 3597 1st 29 19 56 86 1056 447 11.3 From 724 254 180 76 144 100 130 12 2nd 41 70 25 44 13 High calcium 64 100-200 To 914 200 165 235 50 3749 1st 28 20 138 81 1103 396 11.6 From 455 188 145 61 133 98 171 7 2nd 42 70 21 41 27 Rates of removal Color, CU/g hr avg. TOC, mg/g hr avg. 0.18 0.30 0.25 0.08 0.07 0.075 0.41 0.66 0.54 0.10 0.14 0.12 0.40 0.44 0.42 0.10 0.11 0.11 Some values are averages of only 2 or 3 determinations 116 ------- 300 200 q 100 - O u 300 o H 10 20 CARBON BED LENGTH, ft 30 Figure 19. Concentrations of color and TOG remaining versus length of carbon bed 117 ------- column in lime-carbon adsorption had a greater adsorption rate than the lead column, especially in the case of color adsorption. These curves show that with the medium level of soluble calcium and with the other operating conditions used in the pilot plant, the length of bed needed to reach 100 CU is 17 ft and the length needed to reach 100 mg/1 TOC is 20 ft. The color remaining in the water from each column as a function of oper- ting time is given in Figure 20. The color is expressed as a percentage of the feed color to each column to normalize to some extent the variation of feed concentration. A similar plot of data for TOC concentrations is given in Figure 21. These curves show that there is an immediate break- through of nearly 50% of the TOC, and a rapid breakthrough of some of the color as well, in the second column after it is added as the fresh column. In the case of TOC, the water from the fresh column jumped with- in one day to 40-50 mg/1 TOC, which was caused by the non-adsorbable low-molecular-weight compounds in the water. The color and TOC concentrations after each step in the lime-carbon sequence are shown in Figure 22 for the medium-calcium runs. Note that the color drops sharply in lime treatment as compared to a smaller per- centage drop in TOC concentration. The slight increase of color and TOC in the carbonator-filter step was caused by the use of carbonation during one run of this four-run medium-calcium group. As discussed previously, carbonation caused redissolution of floe not removed in the lime-treater clarifier. DOSAGE OF CARBON REQUIRED The lime-carbon sequence was operated for a relatively long time primarily to provide reliable information on how much water could be treated by carbon adsorption per unit weight of fresh carbon. Since the lime-treated feed water to the columns had low concentrations of color, it was necessary to operate an average of 44 days before the product water increased in color and TOC to the point that a fresh column was needed. The lime-carbon sequence was operated for a total of 160 days, and four fresh columns of carbon were added. On the basis of the last three column changes, the average dosage of fresh carbon used was 2.5 Ib per 1000 gal. The dosage requirements would be lower if the total bed length were increased, the contact time increased, or the feed water contained less color and TOC. 118 ------- 100 en o CO C_> So o w u w i-*^ SJ 21 M Crf O O Z i-J M O a u DATE Figure 20. Remaining color in water from columns during lime-carbon sequence ------- 100 to o cn w o PL, O CO O O O O W o So 7/26 DATE Figure 21. Remaining TOC in water from columns during lime-carbon sequence ------- 1000 60 a o o H o v^- I o 800 600 400 200 U BASIN LIME CARBONATOR FIRST SECOND TREATER & FILTER COLUMN COLUMN Figure 22. Color and TOG versus stage of lime-carbon treatment for medium calcium runs 121 ------- RATES OF REMOVAL OF COLOR AND TOG The average rates of removal of color and TOG are given in Table 30 for each column. The rates were almost doubled in going from the low-calcium runs to the medium-calcium runs. The rates for medium-calcium runs of 0.54 CU/g hr and 0.12 mg TOC/g hr were slightly greater than those of the high-calcium runs, but this increase was probably due to the higher feed concentrations during the medium-calcium runs. At approximately equal concentrations, the rates of color removal in the lime-carbon column sequence were much greater than those of the bio-carbon and primary-carbon sequence, but the rates of TOG removal were much less than those for the bio-carbon and primary-carbon runs. However, in the comparison with TOG removal during bio-carbon, an adjustment of the reference concentration of TOG in the lime-carbon sequence by deducting the non-adsorbable organic solution (volatiles discussed later) results in more nearly equal TOG removal rates at equal adjusted concentrations. The rates of removal of color and TOG were generally greater in the second column than in the first column even though the concentrations were much greater in the first column. It is possible that the carbon in the first column was subject to blinding by Ca-organics or residual high molecular weight color bodies, the latter also being suspected of that effect in the primary-carbon operations. EFFECT OF OPERATING VARIABLES ON REMOVAL OF COLOR AND TOG Effect of pH The effect of pH at low and high levels of soluble calcium was investi- gated in a series of runs, each lasting several days. The purpose was to determine whether the beneficial effects of calcium on the lime- carbon sequence were enhanced or not by the high pH that was provided by lime treatment. It is generally recognized that the color intensity and the solubility of the color bodies of kraft mill effluents are increased by adjusting the pH from the normal level of 8-10 to a pH of 12. Therefore, one might expect that adsorption of color bodies by carbon at a high pH of 12 to 12.5 would be poorer than at a lower pH. Also, carbon adsorption of organics in general is generally thought to be better at lower pH levels. These pilot plant tests showed that good carbon adsorption performance was obtained at pH 11.7, or greater, when the calcium content was great- er than about 60 mg/1 and that adsorption remained good when the pH was reduced to 8-9 if the calcium content remained above about 60 mg/1. However, if the calcium content was reduced to the range of 7 to 23 mg/1 (the pH was 7-9), the carbon adsorption performance immediately became very poor. 122 ------- A regression analysis of the data from 153 days of operation at a fairly wide range of conditions showed that color and TOG removals were influenced only slightly by changes in pH and that an increase of pH tended to increase the percentage removal of color and TOG. It therefore is apparent that pH is not a major determining factor of adsorption performance. Particularly the high pH normally associated with the preferred minimum dissolved calcium level of 60 mg/1 (as discussed earlier for the microlime treatment) is not detrimental to good performance of the carbon columns, as might have been expected. Rather, the higher pH tends to improve column performance. Hence, if it is desirable to reduce the pH of the water before reuse or dis- charge, the water should be carbonated after carbon adsorption, rather than before. To avoid precipitation of CaC03, such pH adjustment should be taken to pH 9 or lower. Effect of Feed Concentrations The concentrations of color and TOG, as well as soluble calcium, in the feed to the columns were major determinants of product color and TOG and of rates of removal. The influences of independent variables on the dependent variables were analyzed by means of plots of all conditions and results as functions of operating days. Further insight on the effects of one parameter on another were obtained from a regression analysis which utilized a computer program. The independent variables included in the regression analysis were concen- trations in the feed of Ca, total inorganic carbon (TIC), pH, FPCU, standard color, TOG, the cumulative hours of operation of the second column since it was recharged with fresh carbon, and the cumulative removals of color in each column. The dependent variables were color, FPCU, and TOG from the columns, rates of removal of color and TOG, turbidity of product water, percentage removals of color and TOG, and the ratio of FPCU/TOCo This regression analysis indicated that the parameters affecting color and TOG of the product water in the order of their contribution were color, FPCU, TOG and Ca of the feed, and hours of operation of the second column. This analysis showed that pH and cumulative removals were also factors, but the contributions of these variables were already explained by the related variables of calcium concentration and hours of operation since a new column was added. This analysis indicated that the rates of removal were primarily functions of the concentrations in the feed, the hours of operation since a fresh column was added, and the related variable of cumulative removals by the columns. The percentage removals were primarily influenced by hours operated (and cumulative removals) and by the calcium content of the water. The TIC (carbonate) concentration had an almost negligible influence on column performance. 123 ------- CUMULATIVE REMOVALS BY CARBON As discussed in Section X, it was estimated during the primary-carbon sequence of operation that about 30% of the color and TOC removed by the carbon was in turn eliminated from the carbon columns by the combination of backwashing of the columns and bio-activity in the columns. These measurements were not repeated during the lime-carbon operation, but it is believed that a similar portion of the color and TOC was removed by non-adsorptive mechanisms during lime-carbon operation. Therefore, the term cumulative "removals" is used rather than "loadings" to indicate the degree of saturation of the carbon. The cumulative removals of color and TOC for each column of carbon are shown as a function of operating time in Figures 23 and 24. The average cumulative removals of the columns when discharged were 660 CU/g and 175 mg TOC/g. The slopes of these curves are directly proportional to rates of removal. As pointed out previously, the rate of removal at any given cumulative removal is almost entirely dependent on the concentrations of color and TOC in the feed to that column. (Other factors affecting rates are fixed in this case, such as flow velocity, amount of carbon used, and temperature.) On several days the columns actually released more color and TOC than they adsorbed, as indicated by the downturn of the curves of cumulative removal in Figures 23 and 24. These releases of color and TOC were caused by upsets in the lime treatment, such as a failure of the lime feeder during the night. It is normally expected that loadings on carbon columns will reach about 60% of equilibrium loadings established in batch isotherms. The average loading expected in the first column on the basis of isotherms with lime-treated water (see Section XIII) hence would have been 60% of 350 CU/g, or 210 CU/g, at the average concentration of about 220 CU in the first column in the pilot plant. However, at 500 to 900 CU/g (Figure 23) the cumulative removals of the carbon columns when discharged were 140 to 260% of the isotherm loadings of 350 CU/g, or 240 to 430% of the load- ing that would normally have been expected. This difference between isotherm loading and cumulative removal is evidently due to removals of color by non-adsorptive mechanisms, discussed above, and in more detail in Section X. The magnitude of the apparent non-adsorptive removals is much greater in the lime-carbon sequence than was observed in the normal operation of the primary-carbon sequence. The average isotherm loading for TOC was 130 mg/1 at the average pilot plant feed concentration of 170 mg/1, and this value was exceeded by two of the last three columns in the lime-carbon operation (see Figure 24), 124 ------- 1000 - 800 - ro Cn O U I o M 53 (-1 600 - 400 200 0 Col I I 1 J_ 1 10 20 JULY 1 10 20 AUG 1 10 20 1 10 20 1 10 SEPT OCT NOV 20 1 10 DEC 20 Figure 23. Cumulative removal of color versus time for lime-carbon operation ------- I-1 NJ o 250 200 oo 00 S 8 15° S a 100 50 - Col Col. B, j_ 28 10 20 JUNE JULY 1 10 AUG 20 1 10 SEPT 20 1 10 OCT 20 1 10 NOV 20 1 DEC 10 20 Figure 24. Cumulative removal of TOC versus time for lime-carbon operation ------- These cumulative removal curves show that the lead column when taken off- line was removing color and TOG at a fairly good rate. Therefore, the lead column could have been left on line for a longer period except that a fresh column was needed to maintain low values of color and TOC in the discharged water. In designing a commercial plant, the savings in regeneration costs by using more columns in series must be balanced against the greater fixed costs in providing the additional length of carbon bed. REMOVAL OF LOW MOLECULAR WEIGHT COMPOUNDS A large portion of the TOC remaining after carbon adsorption was made up of low molecular weight organic compounds, which are not readily adsorbed by activated carbon. The average concentrations of identifiable compounds during the lime-carbon operation are given in Table 31. Table 31. ORGANIC COMPOUNDS IN LIME-CARBON OPERATIONS To carbon From carbon columns columns Methanol, mg/1 37 37 Ethanol, mg/1 trace trace Acetaldehyde, mg/1 19 18 Acetone, mg/1 10 14 Formic acid, mg/1 9 6 Acetic acid, mg/1 70 79 Total identified volatiles, mg/1 145 154 TOC of total identified volatiles, mg/1 60 65 Each value is an average from 8-12 analyses. The concentrations of some of these volatiles varied with time. For example, the acetaldehyde concentrations were 30-45 mg/1 in August and September, 1972, were 0-11 in October and November, 1972 and 11-14 later in November, 1972. As discussed in Section VII on Clarification and Bio-oxidation, it is believed that much of the acetic acid was formed by anaerobic activity in the equalization basin. These volatiles apparently contributed about 65 mg/1 of the 100 mg/1 of TOC in the lime-carbon product water. REMOVAL OF BOD AND TURBIDITY As computed from the data in Table 29, the removals of BOD were generally low in the carbon columns, ranging from 13 to 40% of the BOD in the feed to the columns. The overall removals of BOD in the lime-carbon treatment 127 ------- ranged from 22 to 45% for the periods grouped by calcium levels. The low removals of BOD can be readily understood by the relatively high concentrations of hard-to-remove, low molecular weight compounds discussed above, which are all bio-degradable. Turbidity was contributed mainly by insoluble calcium-organic compounds. As shown in Table 29, the turbidity of the water to the carbon columns was lower in the medium-calcium runs than in the high- and low-calcium runs. The columns removed 60% of the turbidity at the medium-calcium content and 86% of the turbidity at the high-calcium content. A regres- sion analysis of the lime-carbon data showed that the turbidity of the carbon-treated water was directly related to hours of operation of the second column since it was added as a fresh column, indirectly related to calcium concentration (and the associated parameter of pH), and directly related to the filter paper color of the feed water. REMOVAL OF METAL IONS On the basis of 16 sets of samples, the average soluble calcium content of the water to the columns was reduced in passage through the columns by 3 mg/1 from 80 mg/1 to 77 mg/1. The amount of total calcium removed, including that in colloidal calcium-organic compounds or calcium carbonate, was probably greater, but it was not determined. Other metal ions to and from the columns were determined by analysis of from one to six samples by atomic absorption after filtration of the sample through a 0.8 micron Millipore filter. The average results of these analyses are given in Table 32. Most metal ion concentrations were affected very little by carbon adsorption. Only manganese and potassium were removed to an appreciable extent. 128 ------- Table 32. METAL ION CONCENTRATIONS IN LIME-CARBON SEQUENCE to lime treater 13 1.6 2.0 0.36 0.28 0.13 12 309 0.03 0.5 0.5 7.6 ConcentrationT to columns 80 0.3 1.5 0.30 0.14 0.07 12 327 0.04 0.6 0.4 7.3 mg/1 from columns 77 0.36 2.6 0.34 0.03 0.06 3.5 _ 0.03 0.6 _ _ No. values Metal averaged Ca 16 Mg 6 Al 5 Fe 6 Mn 6 Zn 5 K 1 Na 1 Cu 4 Ni 4 Ti 1 Si 1 PRESSURE DROP AND BACKWASHING OF COLUMNS The lime-treated water that was used in the carbon columns was filtered through the duomedia filter to remove as much of the suspended solids as possible. The turbidity of lime-treated water was reduced by the filter from about 60 to 45 JTU or by 15 JTU. The turbidity was further reduced by the columns to 12 JTU in the medium-calcium runs and to 7 JTU in the high calcium runs. The columns were backwashed every two days throughout the lime-carbon operation. The pressure drop across the first column just prior to backwashing averaged 4 psi and across the second column averaged 2 psi. The average pressure drop across both columns (20 ft of carbon) during operation was about 2 psi. The pressure drop did not increase with the length of time a given column was in service. There was no evidence of calcium sludge deposits in the column after backwashing, as determined by inspection of the carbon when discharged from a spent column. During the first several weeks of the lime-carbon sequence, the columns were backwashed with well water that had very little hardness, high bicarbonate content, and a pH of 4-5. When the first spent column was removed, we found that the lower two feet of the carbon was cemented together with deposits of calcium carbonate. Evidently the high carbonate content of the wash water formed the calcium carbonate from the calcium in the lime-treated water which was at a pH of about 12. Thereafter, product water was used for backwashing and no further trouble was experi- enced with carbonate deposits. 129 ------- SECTION XII OPERATING RESULTS: LIME-FACET ADSORPTION OBJECTIVES AND DESCRIPTION OF OPERATION The FACET carbon adsorption system was under development in short-term runs in the pilot plant prior to May, 1972. Continuous operation of the unit was started on May 15, 1972, with a feed of lime-treated water. After two weeks of operation, the FACET unit was shut down for the next four months to give more attention to the lime-carbon column operation. The FACET unit was then operated seven days per week from September 28, 1972 to December 7, 1972, with lime-treated water at 10 gpm. During these periods of operation, techniques were perfected to give good control of slurry concentrations in the FACET tanks and to give uniform rates of transfer of carbon as a slurry of 10-15% solids from one tank to the next and of discharge of spent carbon. The carbon feeder gave reliable and consistent delivery of fresh carbon from the feeder storage hopper to the third FACET tank (third in the direction of water flow). A flow diagram for the FACET system is given in Figure 6. RESULTS FROM FACET OPERATION WITH LIME-TREATED WATER There were five operating periods of three to five days each in which the feed water to FACET remained fairly constant and which appeared to give results that are representative of the FACET operation. The condi- tions and results from these periods are summarized in Table 33. The five periods are designated as Runs A through E and are presented in chronological order. The average temperature of the water was 85°F. The conditions of operation that have a major effect on removals of color and TOG include the concentration of color and TOC in the feed, concentration of soluble calcium in the feed, carbon slurry concentration, carbon feed (or dosage) rate, and turbidity and suspended solids of the feedwater. The concentration of suspended solids and color in the feed- water varied greatly as a result of unstable conditions in the lime treat- ment step. Quite often, a color increase in the water from lime treat- ment caused a much greater percentage increase in the color from the FACET treatment. The carbon transfer rates given in Table 33 under "Feed" are the rates to the third tank, and the rates under the tank numbers are the rates of transfer from each tank. The rate of discharge of spent carbon from the system is that given as discharge from the No. 1 tank. All weights of carbon, including the feed, are on an oven-dried basis. 130 ------- Table 33. RESULTS FROM FACET OPERATION WITH LIME-TREATED WATER AT 10 GPM Run no . Daces Length, days A 5/15-5/18/72 3.3 Feed or Tank no.a Carbon rate Ib/hr Carbon in tank, g/100 ml Ib Lime to lime treater mg CaO/1 Soluble Ca to FACET, mg/1 Color, FPCU CU Color removed, ?„ of feed Color removal rate, CU/(g hr) TOC, mg/1 TOC removed, % of feed TOC removal rate, mg/(g hr) pH Suspended solids, mg/1 Turbidity, JTU BOD, mg/1 Total 16.3 10.3 543 600 101 1090 250 81 1.8 151 54 0.75 11.8 370 1 23 19 258 807 88 65 3.1 83 45 1.3 2 16 11 150 691 63 10 0.8 70 9 0.4 3 14 10 135 746 48 6 0.5 70 0 0 11.7 411 B 5/23-5/27/72 5 Feed or Total 9.2 6.8 274 610 140 885 266 79 3.9 162 45 1.3 12.1 1 8.6 6.2 84 783 163 39 6.2 130 20 1.9 2 8.1 7.0 173 846 111 19 2.8 96 21 1.8 3 8.3 7.3 98 617 55 21 2.9 90 4 0.3 12.0 162 C 9/28-9/30/72 3 Feed or Total 2.1 12.2 517 490 77 1086 264 78 2.1 168 51 0.78 12.0 25 242 1 3.6 14.5 204 952 196 26 1.7 116 30 1.23 2 3.2 12.7 179 1317 126 26 2.0 98 11 0.50 3 2.4 9.5 134 993 58 26 2.5 82 10 0.60 12.1 D 10/3-10/6/72 4 Feed or Total 1.8 11.7 493 470 136 581 160 54 0.86 140 38 0.53 12.3 25 1 2.5 12.6 178 1055 152 5 0.23 131 6 0.25 2 2.0 12.7 179 820 95 36 1.59 101 22 0.84 3 1.4 9.7 137 . 922 74 13 0.77 87 10 0.51 11.9 250 E 10/26-11/6/72 12 Feed or Total 2.7 14.3 605 670 153 872 157 53 0.71 158 36 0.46 12.1 40 266 1 1.7 14.0 197 1022 166 -6 -0.23 154 2 0.10 2 1.8 13.5 190 1055 100 42 1.74 132 14 0.57 3 2.4 15.5 218 741 73 17 0.62 101 20 0.71 12.0 416 146 Avg •C, D, E 19 Feed or Total 2.42 13.4 567 600 137 844 174 59 0.96 156 39 0.54 12.0 1 2.17 13.7 194 1018 167 4 0.17 143 8 0.33 2 2.06 13.2 186 1047 103 37 1.75 120 15 0,62 3 2.19 13.3 188 819 71 18 0.95 95 16 0.66 12.0 280 Ratio, FPCU/CU 4.4 9.2 11 15 3.3 4.8 11 4.1 4.9 10 17 3.6 6.8 8.6 12 5.6 6.2 11 10 4.9 6.1 10 12 The results for tank 3 are for the water before it was filtered in the duomedia filter. The averages are weighted by the days of operation of each run. ------- SECTION XII OPERATING RESULTS: LIME-FACET ADSORPTION OBJECTIVES AND DESCRIPTION OF OPERATION The FACET carbon adsorption system was under development in short-term runs in the pilot plant prior to May, 1972. Continuous operation of the unit was started on May 15, 1972, with a feed of lime-treated water. After two weeks of operation, the FACET unit was shut down for the next four months to give more attention to the lime-carbon column operation. The FACET unit was then operated seven days per week from September 28, 1972 to December 7, 1972, with lime-treated water at 10 gpm. During these periods of operation, techniques were perfected to give good control of slurry concentrations in the FACET tanks and to give uniform rates of transfer of carbon as a slurry of 10-15% solids from one tank to the next and of discharge of spent carbon. The carbon feeder gave reliable and consistent delivery of fresh carbon from the feeder storage hopper to the third FACET tank (third in the direction of water flow). A flow diagram for the FACET system is given in Figure 6. RESULTS FROM FACET OPERATION WITH LIME-TREATED WATER There were five operating periods of three to five days each in which the feed water to FACET remained fairly constant and which appeared to give results that are representative of the FACET operation. The condi- tions and results from these periods are summarized in Table 33. The five periods are designated as Runs A through E and are presented in chronological order. The average temperature of the water was 85°F. The conditions of operation that have a major effect on removals of color and TOG include the concentration of color and TOG in the feed, concentration of soluble calcium in the feed, carbon slurry concentration, carbon feed (or dosage) rate, and turbidity and suspended solids of the feedwater. The concentration of suspended solids and color in the feed- water varied greatly as a result of unstable conditions in the lime treat- ment step. Quite often, a color increase in the water from lime treat- ment caused a much greater percentage increase in the color from the FACET treatment. The carbon transfer rates given in Table 33 under "Feed" are the rates to the third tank, and the rates under the tank numbers are the rates of transfer from each tank. The rate of discharge of spent carbon from the system is that given as discharge from the No. 1 tank. All weights of carbon, including the feed, are on an oven-dried basis. 130 ------- Table 33. RESULTS FROM FACET OPERATION WITH LIME-TREATED WATER AT 10 GPM Run no . Dates Length, days A 5/15-5/18/72 3.3 Feed or Tank no.a Carbon rate Ib/hr Carbon in tank, g/100 ml Ib Lime to lime treater mg CaO/1 Soluble Ca to FACET, rng/1 Color, FPCU CU Color removed, 70 of feed Color removal rate, CU/(g hr) TOC, mg/1 TOC removed, 7, of feed TOC removal rate, mg/(g hr) PH Suspended solids, mg/1 Turbidity, JTU BOD, mg/1 Total 16.3 10.3 543 600 101 1090 250 81 1.8 151 54 0.75 11.8 370 1 23 19 258 807 88 65 3.1 83 45 1.3 2 16 11 150 691 63 10 0.8 70 9 0.4 3 14 10 135 746 48 6 0.5 70 0 0 11.7 411 B 5/23-5/27/72 5 Feed or Total 9.2 6.8 274 610 140 885 266 79 3.9 162 45 1.3 12.1 1 8.6 6.2 84 783 163 39 6.2 130 20 1.9 2 8.1 7.0 173 846 111 19 2.8 96 21 1.8 3 8.3 7.3 98 617 55 21 2.9 90 4 0.3 12.0 162 C 9/28-9/30/72 3 Feed or Total 2.1 12.2 517 490 77 1086 264 78 2.1 168 51 0.78 12.0 25 242 1 3.6 14.5 204 952 196 26 1.7 116 30 1.23 2 3.2 12.7 179 1317 126 26 2.0 98 11 0.50 3 2.4 9.5 134 993 58 26 2.5 82 10 0.60 12.1 D 10/3-10/6/72 4 Feed or Total 1.8 11.7 493 470 136 581 160 54 0.86 140 38 0.53 12.3 25 1 2.5 12.6 178 1055 152 5 0.23 131 6 0.25 2 2.0 12.7 179 820 95 36 1.59 101 22 0.84 3 1.4 9.7 137 * 922 74 13 0.77 87 10 0.51 11.9 25.0 E 10/26-11/6/72 12 Feed or Total 2.7 14.3 605 670 153 872 157 53 0.71 158 36 0.46 12.1 40 266 1 1.7 14.0 197 1022 166 -6 -0.23 154 2 0.10 2 1.8 13.5 190 1055 100 42 1.74 132 14 0.57 3 2.4 15.5 218 741 73 17 0.62 101 20 0.71 12.0 416 146 Avg .C, D, E 19 Feed or Total 2.42 13.4 567 600 137 844 174 59 0.96 156 39 0.54 12.0 1 2 2.17 2.06 13.7 13.2 194 186 1018 1047 167 103 4 37 0.17 1.75 143 120 8 15 0.33 0.62 3 2.19 13.3 188 819 71 18 0.95 95 16 0.66 12.0 280 Ratio, FPCU/CU 4.4 9.2 11 15 3,3 4.8 7.6 11 4.1 4.9 10 17 3.6 6.8 8.6 12 5.6 6.2 11 10 4.9 6.1 10 12 j* The results for tank 3 are for the water before it was filtered in the duomedia filter. The averages are weighted by the days of operation of each run. ------- REMOVALS OF COLOR AMD TOG Unusually low average values of color (48 and 55 CU) and TOG (70 and 90 mg/1) were obtained during Runs A and B. The good removals were probably mainly due to the high carbon feed rates of 9.2 and 16.3 Ib/hr. At a much lower average carbon feed rate of 2.4 Ib/hr (4.0 lb/1000 gal) during Runs C, D, and E (19 days), the color was reduced from 171 to 71 CU for a reduction of 59%, and the TOG was reduced from 156 to 95 mg/1, for a reduction of 39%. Also included in Table 33 are filter-paper colors (FPCU) because this provides a measure of the amounts of colloidal color bodies resulting from the FACET treatment that are removed by the 0.8 micron filter used for standard color measurement. Most of the total (unfiltered) color of the FACET product water was removable by filtration through Whatman No. 2 paper or through the duomedia filter in the pilot plant. During Runs A and B, the FACET product water after the duomedia filter had filter-paper colors as low as 128 to 150 CU and standard colors in the range of 30 to 60 CU. During most of the remaining periods, the product water contained substantial amounts of colloidal color bodies that were not removed by the duomedia filter or by filter paper. It was noted that when the low filter-paper colors were obtained, the black color bodies normally in the water from the last FACET tank were sufficiently large that they would settle from the water when a sampe was left undisrupted in a beaker for about 20 minutes. We were unable to pinpoint the condi- tions of operation that would insure that the color bodies would always be of the size that would be removed by the duomedia filter. A high calcium content is believed to favor the growth or coagulation of the larger color bodies during FACET treatment. Even though the data of Table 33 do not show a very strong correlation between soluble calcium concentration and product water colors, other pilot plant data indicate that, in general, best removals of color were obtained with high concen- trations of soluble calcium in the feedwater to FACET. It was found that the ratio of filter-paper color (FPCU) to standard color (CU) is a useful indicator of the size distribution of the color bodies, and the average values of this ratio are given in Table 33. The values of FPCU/CU increased from about 2.0 in the feed to lime treatment to a range of 3.3-5.4 in the feedwater to FACET and to a maximum in the third tank in the range of 10-17. This increase indicates that most of the apparent color was in the range larger than the 0.8 micron size by the time the water was discharged from the third tank. An experimental program, discussed in Section XIII, indicates that some of the color bodies in the product water that were larger than 0.8 micron were peptized carbon, while some were coagulation products of color bodies originally in the water. The exact contribution of each was not determined. 132 ------- The average TOG concentrations of the FACET product water during Runs A-E ranged from 70 to 101 mg/1, and the percent removals of TOG from the feedwater to FACET ranged from 36 to 53%. As would be expected, the best removals of TOG occurred in Runs A and B which had high carbon feed rates. The values of TOG concentrations in the product water are believed to be much higher than they would have been if it had not been for the low molecular-weight compounds formed in the equalization basin by anaerobic activity, which was discussed in the section on the primary-carbon sequence (Section X). The percentage removals of TOG and color by each stage varied widely from one run to another. During Runs A and B with a high feed rate of carbon, most of the color and TOG was removed in the first tank. These results at the high carbon feed rate indicate that the color and TOG remaining after the first stage was poorly adsorbed, even with relatively fresh carbon. At the lower carbon feed rates of 1.8 to 2.7 Ib/hr during Runs C, D, and E, a large portion of the total reduction of color and TOG occurred in the second and third stages, which indicates that the carbon feed rate was about the lowest that could be used. At the average carbon feed rate of 2.2 Ib/hr for these runs, the dosage was 4.0 lb/1000 gal or 480 mg/1. During Run E, on the basis of three sets of samples, the BOD was reduced from 266 mg/1 to 146 mg/1, for a reduction of 45%. ADSORPTION PERFORMANCE BY LIME-FACET COMPARED TO THAT BY LIME-CARBON COLUMNS The rates of adsorption per unit weight of carbon in the system for FACET operation during Runs C, D, and E, are compared in Table 34 to those from microlime-carbon column operation at medium calcium concentrations. The rate of removal of color in lime-FACET was 0.96 CU/(g hr) versus 0.59 CU/(g hr) for the lime-carbon column operation, or 1.6 times as great. The rate of removal for TOG was 0.54 mg/(g hr) versus 0.14 mg/(g hr) for lime-carbon columns, or 4 times as great. The results from these two modes of operation are not strictly comparable. For instance, the soluble calcium content of the FACET feed was higher (137 versus 86 mg/1). The feed concentrations and the amounts of color and TOG removed for FACET were less. The cumulative removals on the spent carbon were less for the FACET operation, i.e., 213 versus 668 CU/g, and 126 versus 263 mg TOC/g. The carbon dosage rate was greater for FACET (4.0 versus 2.2 lb/1000 gal). The feed to the lime-carbon columns was filtered through the duomedia filter, whereas the feed to FACET was not, which put an additional load of suspended organics on the FACET system. If the conditions in the FACET operation had been the same as 133 ------- Table 34. COMPARISON OF LIME-FACET AND LIME-CARBON COLUMN OPERATION AT A WATER FLOW RATE OF 10 GPM Lime-carbon Lime-FACET Period used Med. calcium Runs C,D,E Reference table 28a 33 Days of operation 30 19 Avg. soluble Ca, mg/1 86 137 PH 11.4 12.0 Carbon dose rate Ib DWB/1000 gal 2.2 4.0 Contact time of carbon, days 97 10 Contact time of water, hr 1.77 0.84 Carbon in system, Ib 2880 567 Mean particle size, microns 670 345 Color, feed, CU 252 174 product, CU 76 71 removed, CU 176 103 % removed 70 59 cum. removed, CU/g 668 213 rate, CU/(g hr) 0.59 0.96 TOC, feed, mg/1 177 156 product, mg/1 100 95 removed, mg/1 77 61 % removed 44 39 cum. removed, mg/g 263 126 rate, mg/(g hr) 0.14 0.54 Values involving weight of carbon have been adjusted to dry-carbon basis. 134 ------- in the carbon-column runs, it is believed that the FACET rates would have been lower but still substantially greater than those of the carbon- column runs. This comparison at equal water flow rates show that the rates of removal with FACET were much greater, that the size of the contacting zone was only 11% of that for the columns, and the inventory of carbon was only 18% of that for the columns. The advantages of FACET over column adsorp- tion are that it permits changing of dose rate of carbon to meet changing feed concentrations, it has ability to handle a feed with a high concen- tration of filterable solids, it requires no backwashing, it will require smaller and lower-cost equipment, it requires a smaller inventory of carbon, and it provides a continuous flow of carbon to regeneration. Potential disadvantages of FACET include: it might have a greater attri- tion loss of carbon, it requires coagulation or filtration of the product water, and it requires more automatic controls to regulate slurry concen- trations. In a commercial unit, a greater number of stages would be used which would permit a closer approach to true counter-current contacting, high cumulative removals by the carbon, and lower dose rates. ATTRITION OF CARBON IN FACET OPERATION One of the major concerns in considering FACET as a viable commercial process for treatment of water is whether attrition of the carbon particles in stirred contactors would cause excessive loss of carbon as fines in the product water. Data obtained to date indicate that loss of carbon will constitute a minor operating cost. Ten tests were made in the laboratory to determine attrition losses by stirring slurries of Darco XPT in effluent water and in deionized water for periods of 6 hr to 5 days. In most of the tests the carbon sample was wet-screened through a 200-mesh sieve before and after the test, and the loss in weight of +200 mesh particles was usually defined as the attrition loss. Since it was found that the Darco XPT carbon has a high water-soluble content of 1.2 to 3.5%, the carbon sample was first water extracted to remove solubles. The attrition loss in the above tests was generally less than 1% when the carbon was stirred 4-5 days and the results calculated on weight of water-washed oven-dry carbon. We attempted to determine attrition losses by measuring total losses of carbon from the FACET system during the pilot plant runs by means of weight balances over periods of about a week, but it was found that it was very difficult to get a reliable weight balance on this system. To get reliable data, the weight of carbon in the system at the start and end of the test peripd was needed but could not be measured with sufficient precision. 135 ------- Another inaccuracy in making a weight balance was in measuring the rate of removal of spent carbon with sufficient precision. The amount of filterable solids in the product water could not be used as a measure of attrition losses because the value of filterable solids in the feed water (about 300 mg/1) was as great as that of the product water, and both were difficult to determine precisely. Some carbon was lost to the product water during the pilot plant runs due to some of the feed carbon to the third tank not being completely wetted which caused some particles to float and to be carried out in the overflowing product water. The clarification zone of this tank was later modified to prevent carry-over of fresh carbon in the product water. 136 ------- SECTION XIII SUPPORTING LABORATORY STUDIES A large amount of laboratory work was carried out during the construction and operation of the pilot plant to provide information and guidance on changes in the pilot plant and to provide answers to problems that arose. A problem requiring the major effort was that of color increase in the treated water during certain conditions of carbon adsorption. Other laboratory studies included providing carbon adsorption isotherms of various streams throughout the pilot plant operation, determining best conditions of operation for lime treatment and carbon adsorption, and special analyses. The findings from these studies are summarized below. COLOR INCREASE IN PRESENCE OF ACTIVATED CARBON Laboratory evaluations and pilot plant runs of the FACET stirred-carbon adsorption system showed that there was generally an increase of color of unfiltered product water even though the TOC of the water decreased. In most cases, this apparent color was removed by filtering the water through a No. 2 Whatman filter paper or a sand filter in the pilot plant and was greatly reduced by filtering the sample through the 0.8 micron filter used in determining standard color. Since the "color" produced during contact of the water with stirred carbon was removed by the 0.8 micron filter, it was not "true" color but colloidal color. This increase of apparent color was also noticed in preparing adsorption isotherms, especially at low dosages of carbon of 0.2 to 0.5 g/1. The color that was removed from the carbon columns in the pilot plant during backwashing had much the same physical and chemical properties as that produced in the FACET tanks. This increase of color was studied extensively in the laboratory because the formation of the colloidal color bodies would be expected to coat the external surface of the carbon. This would greatly reduce the rate of diffusion to the carbon pores and thus the rate of adsorption of the smaller molecules on the inner surfaces of the carbon in both the FACET and fixed-bed adsorption processes. It was desirable to learn how to control the formation of the colloidal color in the FACET system so that the colloids could be easily removed from the product water by sand filters or by coagulation. It was also hoped that knowledge of the mechanism of formation and the properties of the colloidal color would suggest improved modes of operation of the fixed-bed and FACET carbon adsorption systems or might lead to novel less-costly processes for coagulation and removal of color from pulp mill effluents. 137 ------- The increase of color occurred whenever powdered or granular carbon was stirred for 1-4 hours with pulp mill effluent or synthetic effluent (dilute black liquor) if the ratio of initial color to external surface area of the particles was in the range of about 5,000 to 30,000 CU per square meter. Results of a representative experiment are shown in Figure 25. The color increase also occurred, at a much lower rate, when no carbon was present and occurred in presence of non-carbon granular solids and powders such as alumina, silica, diatomaceous earth, and pumice. The increase of color occurred with each type of carbon tested: Darco S-51 powder, Darco XPT 40 x 140, Darco 20 x 40, and Westvaco WVL granular. Most other brands of carbon exhibited the color increase at low dosages in the standard isotherm tests. Temperatures up to 90°C increased the rate of formation several fold over that at 25°C. The rate of color increase was accelerated by increasing the pH over the range of 8 to 12. Heavy metal ions (which are present in the ash of the activated carbon and in the mill effluent) did not seem to affect the rate of color increase. The rate of color production decreased with time of stirring but was still significant after 4-24 hours. An increase of solid carbon concentration in the water during stirring over the range of 0.3 to 100 g/1 generally increased the production of color, but the rate of increase per weight of solid was greatest at low concentrations of solids. If the ratio of initial color to surface areas of the solid was outside of the range of 5,000 to 30,000, there was little increase of color. The test conditions for maximum increase of color of pulp mill effluent were: 5-10 g/1 of Darco 40 x 140 at 50°C for 1-4 hours. These conditions gave increases of about 700 unfiltered CU/g of carbon and 100 paper filtered CU/g. The mechanisms causing this observed color increase were not pinpointed. The results of the various studies indicate that, as might be expected, some of the color is produced by carbon particles from the activated carbon that were originally present as dust or that were removed by attrition or by peptization in the presence of mill effluent. The remain- ing increased color was evidently produced from substances in the effluent itself. The chemical and physical properties of the colloidal color bodies were studied by filtering the samples through filters of 2-0.05 microns, TOC analysis, microscopic observations, electron micrographs, ultraviolet absorption, infrared absorption, elemental analyses, and action of solvents, bleaching agents, and caustic. These studies showed that the color in- crease was due primarily to formation of black color bodies larger than 1 micron. As time of stirring increased, the concentration of "true" color passing 0.05 and 0.8 micron filters decreased, therefore there was 138 ------- 50,000 CONDITIONS 100 g/1 40 x 140 MESH CARBON 25°C, pH = 10-11 COLOR MEAS. AT pH 10-11 SETTLED 20 MIN., UNFILTERED rt 30,000 PAPER FILTER 0.8 MICRON FILTER SETTLED 20 MIN. PAPER FILTER 0.8 MICRON FILTER I I 123 TIME OF STIRRING, hr Figure 25. Changes in color of dilute black liquor with time of stirring with activated carbon 139 ------- a shift from the small color bodies to the large colloidal bodies. The fraction of color removed by filter paper and 0.8 micron Millipore filters had very large concentrations of color per unit -of TOG 25 to 50 times as great as that of mill effluent. The absence or presence of oxygen in a stirred test had no apparent influence on the degree of color increase. The color intensity of the color bodies from a color increase test was not reduced by treatment with hypochlorite or hot caustic, and the color was not extracted with methanol. Carbon stirred in water alone or with detergent present did not give appreciable color increases which indicated that attrition or peptization of the carbon were not major contributors to the color increase. However, digestion of carbon in dilute HC1 at pH 2 caused significant color increase due to suspended black particles that were removable by an 0.8 micron filter and suggested that peptization of the carbon occurs under some conditions. Suspensions of activated carbon of up to about 2 micron size were found to give a 70% detection as TOG of the carbon actually present. Taken together with the observed high CUrTOC ratio, this point tends to suggest that the color increase was not from activated carbon particles. A study of this problem by the Institute of Paper Chemistry (unpublished report) indicated that a portion of the color from the backwashing of carbon columns could be metal sulfides, that the color bodies after methanol extraction had "infrared spectra" (actually IR scatter, i.e., no distinguishing IR adsorption bands were seen in either sample) similar to those of activated carbon, and therefore the IPC concluded that colloi- dal carbon from the activated carbon was the primary source of the color of the backwash water. Our studies have indicated that a substantial portion of the color in- crease is due to color bodies formed from the organic matter of the water in addition to that possibly contributed by the activated carbon. Possible mechanisms include (1) destabilization of organic colloids by contact with the carbon surfaces or after adsorption on the carbon (it appears reasonable to assume that at least a portion of the effluent color bodies is present in colloidal form, similar to lignin color in natural waters (1,2), particularly in view of the high molecular weight determined recently (43)), (2) formation of metal sulfides and metal-organic complexes at the surface of the carbon that become filterable colloids, (3) oxidative condensations to form insoluble compounds of lignin, as hypothesized by Ganczarczyk (39) in bio-oxidation of pulp mill effluents, (4) free-radical polymerization of lignin degradation products (42) and (5) other conden- sation products from hydrogen bonding and other types of chemical bonds. This study of color increase of pulp mill effluents in contact with carbon under certain conditions has shown that the mechanisms causing it are very complex and that methods of controlling it for improving color removal by carbon are elusive. The study of this phenomenon is being continued by St. Regis Paper Company. 140 ------- ADSORPTION ISOTHERMS Equilibrium isotherms were prepared using Darco S-51, Darco 20 x 40, and Darco XPT 40 x 140 (all pulverized to less than 325 mesh) with samples of water from each step of treatment used in the pilot plant. Color and TOC isotherms were prepared using 43 samples of water. Average isotherms for each treatment step are given in Figure 26 for color and in Figure 27 for TOC. Average loadings from the isotherms for selected remaining concentrations of color and TOC are given in Table 35. The procedure used in making the isotherms is given in the Appendix B. There was a considerable spread of loadings at a given color for a given stream, probably because of differences in the chemicals present from one sample to the next. The average loadings shown in Table 35 and Figure 26 indicate that with Darco carbon the greatest equilibrium loadings (at given concentrations) of color were obtained with water from lime treatment, lime-carbon, and bio-oxidation, but much lower color loadings were obtained with water that had passed through the retention basin and primary-carbon adsorption. The highest loadings of TOC at concentra- tions above 80 mg/1 were with bio-oxidized water, retention basin water, and lime treated water. Table 35. AVERAGE EQUILIBRIUM LOADINGS ON CARBON FROM ISOTHERMS PREPARED DURING PILOT PLANT OPERATION4 Water Source No. isotherms averaged Loading on carbon, CU/g at 100 CU 500 CU 1000 CU From mill ditch 7 From reten. basin 14 From P-C cols. 4 From lime trt. 7 From L-C cols. 5 From bio-oxid'n 2 From B-C cols. 1 186 87 87 179 260 120 230 345 290 207 318C 480 857 450 2600 TOC loading on carbon,mg/g at 80mg/l 29 30 21 27 16 58 20 lOOmg/1 41 47 33 63 29 73 - 150mg/l 61 144 70 116 60 112 - a Isotherms all with powdered Darco S-51, pulverized -325 mesh Darco 20 x 40, or pulverized Darco XPT 40 x 140. Abreviations: carbon. At 200 CU. P-C = primary-carbon, L-C = lime carbon, B-C = bio- 141 ------- 1000 60 o s O u 100 10 BIO-CARBON LIME-CARBON—>> - LIME TRTR BIO-OXID'N A MILL DITCH RETENTION BASIN PRIMARY-CARBON I i i i 11 10 100 1000 COLOR REMAINING, CU Figure 26. Average color isotherms after various treatments during pilot plant operations 142 ------- 00 1? o o I-J u 8 100 10 LIME TRTR BIO-CARBON 10 BIO-OXID'N RETENTION BASIN -MILI DITCH LIME-CARBON PRIMARY CARBON ' i 1 - 1 1 ll 100 TOG REMAINING, mg/1 1000 Figure 27. Average TOG isotherms after various treatments during pilot plant operations 143 ------- The lowest TOG loadings were with lime-carbon and primary-carbon treated water. These average isotherms indicate that the color remaining after bio-carbon, lime-carbon, and lime treatments is more readily adsorbed than the TOC and could have been removed to a greater extent with greater retention time in the carbon columns. The color isotherms generally have two slopes. At higher concentrations remaining after treatment the color loadings decrease rapidly with concen- tration, but at lower concentrations the loadings decrease less rapidly. In contrast, the TOC loadings are very sensitive to concentration remain- ing after treatment, with the bio-treated water being the least sensitive. The steep slopes of the TOC isotherms show that high loadings (above 30- 80 mg/g) of TOC are achieved at concentrations greater than 100 mg/1 but the loadings at less than 50 mg/1 are very low, which indicates that it would be difficult to reduce TOC of mill effluents to concentrations of less than 50 mg/1, except after bio-oxidation. The steep slopes and the very low loadings of TOC obtainable at low residual TOC concentrations can be taken as an indication of the presence of very-hard-to-adsorb organic molecules. Actual chemical analyses presented in Sections X and XI show hard-to-adsorb, low molecular weight organics in the 40 to 80 mg/1 of TOC range. For all three treatment sequences, TOC isotherms obtained on effluent already treated with carbon show lower loadings (at given concentrations), particularly in the higher solution concentration range, than do isotherms of the corresponding sequences before carbon treatment. This observation is an indication that more readily adsorbed substances are adsorbed first, hence the solution remaining after carbon column adsorption are of a different qualitative composition, rather than being simply equally lower in concentration of each of the many compounds representing the measured quantities of TOC. This observation is of interest in conjunction with the discussions of rates of adsorption in Sections IX, X and XI. In the bio-carbon and primary-carbon sequences, adsorption rates decrease with decreasing concentration due to a reduction of driving force both from the reduction in concentration in solution and from the reduction in equilibrium loading at given concentration. On the other hand, in the lime-carbon sequence, rates in the second column, at lower effluent concentration, are actually higher than in the first column. This is contrary to expectations and is perhaps an indication of extensive blinding of the carbon surfaces in the first column from calcium-organic deposits. The color adsorption isotherms before and after carbon column adsorption show an opposite relationship to that observed for TOC. A possible explanation for this phenomenon may be found in the known (43) wide distribution of molecular weight (up to 64,000 MW) of color bodies in kraft effluent before carbon treatment. Removal of color bodies in the 144 ------- high molecular weight ranges from the original effluent may be predominant- ly in the very large pores and on the external surface of the carbon, making the majority of the smaller pores inaccessible. Once the high molecular weight fraction is removed and fresh carbon is brought into contact with the solution, the lower molecular weight materials (perhaps in the 500 to 2,000 MW range) have ready access to the full range of carbon pore sizes and adsorptive area. This hypothesis is perhaps support- ed by the higher loadings obtained in lime treated water as compared to non-lime treated water. Lime treatment of unbleached kraft effluent has been found (43) to remove high molecular weight color bodies in preference to lower molecular weight color bodies. Isotherms were used in checking the adsorption characteristics of several lots of Darco 20 x 40 and XPT 40 x 140 received for use in the pilot plant. Isotherms made with pulp mill effluent showed that the granular forms of Darco achieved slightly greater loadings of color and TOG at given concen- trations than S-51 when all were pulverized to pass a 325 mesh sieve. DYNAMIC VERSUS EQUILIBRIUM CONCENTRATIONS IN PILOT PLANT COLUMNS Several tests were made in the pilot plant to determine how close the color and TOG concentrations in the carbon columns determined during normal operations (dynamic concentrations) approached the equilibrium concentrations for possible use in an evaluation of measured adsorption rates. These tests were made by shutting off the flow of water through the columns and then removing samples of water from the base of each column at intervals of time up to four days. After about 8 hours, the concentra- tions had almost reached equilibrium values. These tests indicated that the columns were operating at color concentrations 30 to 100% higher than the equilibrium values and at TOG concentrations 30 to 80% higher. FOAMING TENDENCY OF TREATED EFFLUENT The effect of the various steps of treatment on the foaming of the water by a standard foaming test was checked on 10 samples over a 6-month period. These tests indicated that the foaming of the mill effluent increased 50 to 100% during passage through the retention basin, without aeration. Lime treatment reduced the foaming of the feed water about 50%. Carbon adsorption by columns or FACET removed almost all of the foaming tendency. 145 ------- EFFECT OF OPERATING CONDITIONS ON ADSORPTION OF COLOR AND TOG ON CARBON Effect of pH The effect of pH of the feed water on adsorption was investigated in several isotherm tests with mill effluent adjusted to pH values of 3 to 12. Within the range of 5-12, the results were inconclusive. Loadings varied as much as two-fold between different pH levels for specific samples, evaluated at a given color level. Results from test to test and color level to color level, however, were largely contradic- tory to each other. The limited amount of testing leaves the question of pH effect unresolved. At pH levels of less than 4, the lignin compounds become much less soluble and "loadings" (including precipita- tion) in the carbon increase sharply. Effect of Temperature A limited number of isotherms prepared at temperatures of 75° and 130°F with Darco S-51 and Westvaco AN and mill effluent indicated that the higher temperature gave about twice the equilibrium loadings of color at a color concentration of 300 CU and 1 to 2.6 times as much loading of TOG at a concentration of 100 mg/1. No comparative isotherms were obtained to observe a possible temperature effect on bio or lime treated mill effluent. In the pilot plant, the operating temperature range was very narrow (74-85°F). Hence, no temperature effect was observed. The temperature effect deserves further study, particularly in view of the fact that mill effluent can be available as high as 130-150°F, thus requiring no added cost for heating if the temperature effect is to be exploited in a large installation. The plant design and cost studies presented in Section XIV are based on the pilot plant results obtained at 74-85°F. 146 ------- SECTION XIV TREATMENT PLANT DESIGN STUDIES AND COST ESTIMATES OVERALL DESIGN CRITERIA Engineering design studies were made for full-scale plants for treating unbleached kraft mill effluents by the lime-carbon and bio-carbon sequences to provide water suitable for reuse in the mill. These studies were carried out to provide a basis for comparative cost estimates and to illustrate the methods of equipment design and operating procedures that appear best on the basis of our pilot plant studies and results from other effluent treatment studies. The design and cost studies were based on the conditions given in Table 36 and water compositions given in Table 37. The treatment sequences discussed in this section and resultant water qualities are also present- ed in Figure 28. These conditions were chosen as typical of new and older unbleached kraft pulp and paper mills in the south. It was assumed that the effluents will be from mills that have catch basins for spills and sudden discharges of highly contaminated water. With such control, it was assumed that the variation was such that for 90% of the time the maximum color would be less than 1320 CU and that the minimum color would be greater than 680 CU. With this degree of variation, the TOG concen- tration would remain in the range of 175 to 325 mg/1 90% of the time. It was assumed that the effluents from the older and new mills would have the same concentrations of color, TOG, and BOD, but that the older mills would have a 50% greater volume of effluent per ton of pulp produc- tion. The compositions of effluent after the various treatment steps and sequences in Table 37 are based on results achieved in the pilot plant, particularly in terms of microlime and carbon treatment, and on results in mill scale practice of bio-oxidation and minilime treatment. For the sequences involving activated carbon, the target of reusable water quality was defined as 100 CU and 100 mg/1 TOG, although better results were achieved in the pilot plant. For the microlime-bio and minilime-bio sequences, the final color is based on that achieved after lime treatment, although in both commercial operations of minilime-bio the color increases to substantially higher levels in the bio-oxidation ponds, resulting in levels of 250 to 500 CU. Color increases were also seen in the laboratory as discussed in Section IV. No cost estimate is included for the lime-FACET sequence because it was felt that the status of development achieved in the FACET system was not sufficient to warrant an estimate of comparable precision to those 147 ------- Table 36. BASES USED FOR DESIGN AND COST STUDIES Case no. 1 2 3 4 4a 5 5a 6 6a 7 8 9 10 11 Age of mill new new older older older older older older older new new older new older Mill pulp capacity, tons/day 800 1600 800 800 800 800 800 800 800 800 800 800 800 800 Effluent vo lume , gal/ton 12,000 12,000 18,000 18,000 18,000 18,000 18,000 18,000 18,000 12,000 12,000 18,000 12,000 18,000 mgd 9-6 19.2 14.4 14.4 14.4 14.4 14.4 14.4 14.4 9.6 9.6 14.4 9.6 14.4 Treatment sequence microlime-carbon microlime-carbon microlime-carbon prim-bio-carbon prim-bio-carbon^ microlime-bio microlime-biob minilime-bio^ minilime-bio minilime-carbon minilime only minilime only microlime only microlime only Microlime and minilime treatments have no separate primary clarification. Estimate excludes capital cost of bio-oxidation facilities. Table 37. AVERAGE COMPOSITION OF WATER TO AND FROM TREATMENT3 From mill to treatment From microlime From minilime From bio-oxidation From microlime-carbon From minilime-carbon From bio-carbon From microlime-bio From minilime-bio Color, CU 1000 300 150 1000 100 100 100 300 150 TOC mg/1 250 150 135 175 100 100 100 70 50 BOD mg/1 250 225 187 25 150 117 20 20 20 _pJL 10.5 11.5 8 8 8 8 8 8 8 Other Properties: Temperature: lime-carbon = 100°F., bio-carbon and bio-lime = 90°F. Conductivity: minilime-carbon = 1200 micromhos, microlime-carbon = 3600 micromhos. 148 ------- COST ESTIMATE 1,2,3 EFFLUENT FLOW DIAGRAM 4,4a 5,5a 6,6a 8,9 10,11 LIME-i COo-i T Z 1 MICRO- LIME CARBON pH * ADJUST * \ SLUDGE NUTRIENTS AIR ^ • PRIMARY BIO »m. PflDDflM ^ "• UMKtiUN »*- » SLUDGE NUTRIENTS LIME-i AIR-n MICRO- LIME *_, BIO — T SLUDGE LIME-n CO? -r I T . MINI- LIME T SLUDGE LIME-r MINI- LIME *^ T SLUDGE LIME-i MINI- " LIME { SLUDGE LIMEn MICRO- " LIME CARBO- NATOR *• BIO ~*~ | SLUDGE C02 -i C02 -i CARBO- NATOR pH *" ADJUST " CARCON | SLUDGE C02 -T C02 -i CARBO- NATOR I pH ADJUST pH ^ * ADJUST SLUDGE SLUDGE EFFLUENT Color TOC CU mg/1 100 100 100 100 300 70 150 50 100 100 300 150 Figure 28. Flow diagrams and water quality for Cost Estimates 1-11 149 ------- presented. Substantial savings are expected, as indicated in Section V, in equipment and carbon inventory cost, partially off-set (to a lesser degree than indicated in the preliminary estimates of Section V) by increased carbon dosages. The lime-FACET system operating costs, including fixed costs, are expected to be lower by 10-15% or several cents per 1000 gal. No estimate is included for the primary-carbon system. Rough estimates made on the basis of pilot plant results indicated costs to be well out of line with the other potential systems, primarily because of the high capital cost requirement as a consequence of long contact time require- ments. The influence of excluding the capital cost of pre-existing bio-oxidation facilities was determined for each estimate that included bio-oxidation. The influence of price of activated carbon ($0.27 and 0.10 per pound) was determined for most estimates that used carbon adsorption. The comparative costs for microlime alone and minilime alone were determined as were the incremental costs of carbon treatment following lime treat- ment. DESIGN METHODS AND ASSUMPTIONS In this part of this section, the design methods and assumptions are discussed as they apply to each treatment step. Further information applying to design can be found in the appendix. Detailed design calcu- lations for Estimate No. 1 are given in Appendix J. Primary Clarification The design for primary clarification includes a mechanical clarifier in which the rise rate is 700 gpd/ft , in which no coagulants are added, and which produces a clarified water containing 50 mg/1 of suspended solids. Also included were water lift pumps, sludge pumps, a sludge settling tank, and sludge filters. The dewatered solids are hauled to landfill. Bio-Oxidation Bio-oxidation was assumed to be provided by aerated basins that are typical of those used by St. Regis and other southern kraft pulp mills. It was assumed that bio-oxidation removes 90% of the BOD, 30% of the TOC, and none of the color of the water. Nutrients are added in the ratio of 5 Ib of N and. 2.5 Ib of P for each 100 Ib of BOD loading. Aerators are used at a ratio of 32 pounds of BOD in the water to bio- oxidation per hp-day. 150 ------- Microlime Treatment The design for the microlime treatment facilities is based primarily on the results from the pilot plant operation but also on the results from the Interstate Paper (18) and Continental Can (14) lime treatment plants. The lime treatment system includes all auxiliary equipment: conveyors from the mill's lime storage, a lime slaker, and a lime slurry feeding system that is controlled by instrumentation to provide 60-90 mg/1 of soluble calcium in the lime-treated water when adding about 520 mg/1 of CaO. Sensors are used in the lime-water mixing zone of the reactor-clarifier to measure the soluble calcium content and to provide a control signal to the lime-slurry feeder. Lime slurry at 10% solids is introduced into the mill effluent just ahead of the lift pump to the lime treater. The lime treater is a standard reactor-clarifier with a central agitated mixing zone to provide a long retention time of 30 minutes. A solids concentration of 0.5 to 1.0% solids is maintained in the mixing zone. A reaction time of 10 minutes is adequate, but the 30-minute retention time and high solids concentration are used to dampen short-term (less than 2 hr) fluctuations of color load. Longer-term variations in con- centration and flow rate are handled by varying the lime dosage rate to maintain the desired concentration of soluble calcium. The lime treater has heavy-duty sludge rakes to handle the dense sludge. The sludge is pumped at about 15% solids to a 12,000 gal sludge settling and surge tank in which the sludge is concentrated to about 30%. The sludge is dewatered with a wire filter at a rate of 100 Ib/hr dry solids per sq ft of filter area to provide a cake of about 70% solids which is conveyed to the mill's lime kiln feeder. The sludge was estimated to contain, on dry basis, about equal amounts of CaCO., and calcium-organics and a total of 43% calcium determined as CaO. The loss of lime as soluble calcium in the treated water was estimated to be 21% of the lime added to the water in microlime treatment but only 2% in minilime treatment. An additional loss of 5% of the lime added was assumed in the handling of lime sludge and filtration. This total loss of lime was assumed to be made up by purchased lime; all other lime used was assumed to be recalcined lime from the lime kiln. It was assumed that the lime cake from effluent treatment is recalcined in an existing lime kiln. The percentage of added load to the kiln from microlime treatment was 8% and from minilime treatment was 21% in the case of new mills having 12,000 gal effluent per ton of pulp. As discussed below, the capital costs for lime treatment include a proportionate share of the lime kiln cost. 151 ------- Minilime Treatment The design for minilime treatment is based primarily on the results reported by Spruill (14) for the Continental Can Company's minilime treatment of 9 mgd of unbleached kraft total mill effluent. Also used in the design calculation were results from the St. Regis - EPA pilot plant program, from the Interstate Paper Company minilime treatment plant at Riceboro, Georgia (18) and from the Georgia Pacific Corporation plant for lime treatment of bleach plant effluent at Woodland, Maine (7,8). The minilime treatment plant includes a lime feeder, lift pumps, sludge pumps, lime treater-clarifier, carbonator (with kiln gas), carbonator- clarifier, sludge dewatering, and conveyor to return lime sludge cake to the lime kiln. The lime treater and carbonator clarifier use a rise rate of 700 gpd/ft2. Additional information is given under microlime treatment (last para- graph) . Carbon Adsorption The design for the carbon adsorption plants was based primarily on the results from the several modes of operation in the pilot plant. The arrangement of adsorption contactors or columns, the transfer of carbon as a water slurry, and the carbon regeneration unit were chosen to be similar to those used in the EPA report prepared by M. W. Kellogg Co. (17). In that report detailed designs and cost estimates were presented for treating by carbon adsorption 1, 10 and 100 mgd of municipal wastewater from biological treatment. These cost data are apparently the most reliable of the available costs on carbon adsorp- tion and were chosen as the basis of the costs used in this report for carbon adsorption. A process flow diagram for that plant is given in Figure 29. The design procedure is discussed below. Further information for the desgin of carbon adsorption plants is given in Appendix H, relation- ships useful in the designs of downflow carbon columns are given in Appendix I, while a detailed example is provided in Appendix J as part of Estimate No. 1, microlime-carbon treatment. The total volume of carbon for a system was found by dividing the volumetric flow rate (volume of water per volume of carbon per hour) by the volume of water to be treated per hour. The volumetric flow 152 ------- Process Water Oi NOTES 1. Tank A is operating as Primary Adsorber. 2. Tank B is operating as Secondary Adsorber. 3. Tank C is operating on Backwash. 4. Tank D is Idle. 5, Tank E is operating on . Cycle to Remove Carbon. 6. Tank F is operating on ; Cycle to Refill with Activated Carbon. Figure 29. Process flow diagram for FWPCA-Kellogg study (17) ------- rate is the inverse of contact time in hours. The contact time and volumetric flow rates were determined from pilot plant results for the feed and product concentration of each estimate. The superficial velocity and the cross-sectional areas of the columns were selected from pressure-drop considerations. The combined height of columns in series stems directly from the contact time required. It was assumed that the rate of adsorption would be the same as that in the pilot plant even though the rate is expected to be greater in the plant because of the greater flow velocity selected for the full- scale plant. The number of columns in series was selected to be at least three (to provide adequate counter-current contacting) and to maintain a reasonable height/diameter ratio and to keep the height at less than about 60 ft. The number of parallel trains was determined by the need to keep the height/diameter ratio between 1 And 2. There- fore, the diameter of the columns was fixed by the number of parallel trains used. In general, one additional column was provided for about every other train to handle excessive flows and concentrations of the feed. Two additional columns were provided, one being emptied of spent carbon and the other being refilled with regenerated carbon. In operation, the lead (more nearly spent) column is removed from a train of columns when the quality of the water from the last column exceeds a tolerable level. A column of regenerated carbon is then added as the last column, and the carbon from the removed spent column is transferred as a slurry to the regeneration furnace. The regenerated carbon from the furnace is quenched in water and transferred to an empty column. When filled, this column is held in reserve to be used behind the last column temporarily whenever an overload to the system causes the product water to rise above a set tolerable level. This reserve column is put into regular service when the color or TOG concentration in the product water increases above the selected value and the first column of that train is removed from service for regeneration of the carbon. Since the function of each column progresses through the series of columns, adequate piping and valving is provided to give the required flexibility. The total inventory of carbon was determined from the number and dimensions of columns. The total inventory also included a 30-day supply of make-up carbon. Backwashing of the columns requires a large flow of product water (5300 gpm in Estimate No. 1) for 15 minutes every day for the lead columns and every other day for the other columns. To minimize the loss of production due to this flow, the same water is used repeatedly for backwashing columns. In the lime-carbon sequences, the water is recycled to lime treatment after use for backwashing six columns. 154 ------- Two clarifiers are provided to remove the suspended solids from the backwash water before each reuse. With this arrangement the loss of product water was only 0.8%. The off-time for backwashing reduced production by 3%, making the total loss of production 3.8%. An equivalent procedure limits the lost production due to backwashing in the bio-carbon treatment sequence. The carbon regeneration unit must provide regenerated carbon at the dosage rate, which was 1000 Ib/hr in Estimate No. 1. With an expected loss of 5% per regeneration cycle, the make-up rate with fresh carbon was 1200 Ib/day. At the average regeneration rate in Estimate No. 1, one column is regenerated each 10.6 days. The major design assumptions used in the cost studies for carbon adsorption are given below: Carbon Columns Carbon Type Flow Velocity Excess Capacity- Backwash ing Filtration Carbon Make-up Regeneration down-flow, fixed bed, 50% free-board (referred to carbon bed height). 20 x 40 mesh, equivalent to Granular Darco with bulk density of 25 lb/ft3. with lime-treated feed water =4.2 gpm/ft , contact time = 1.8 hr, volumetric flow = 0.56 v/(v hr) with bio-treated feed water =4.0 gpm/ft , contact time = 4.4 hr, volumetric rate = 0.23 v/(v hr) based on average feed rates after allowance for off-time for backwashing = 20 to 30% of normal flow rate or normal contaminant loadings in the feed water. with air for 1 min and with water for 20 min at 10 gpm/ft . Lead columns are backwashed each day, others every other day. Backwash water is collected, clarified, and reused in backwashing more-contaminated columns, and then returned to lime treater. Water used is 1% of feed and off- time is 3% for a total loss of production of 4% for backwashing. no filtration is required for feed to columns. total loss from handling and regeneration of carbon is assumed to be 5% per regeneration. This rate of make-up was achieved in large-scale treatment of municipal effluent. In treatment of pulping effluents, it is reasonable to expect that some of the adsorbed organics will form activated carbon, which might cause a lower loss rate when regenerated. spent carbon is transferred as a slurry, dewatered, and fed to a multiple-hearth furnace. A regeneration furnace capacity of 100 lb of carbon per day per square foot of hearth area is assumed and 25% excess capacity is provided. 155 ------- COST ESTIMATES Cost estimates were prepared for plants treating the effluent from un- bleached kraft pulp and paper mills to show which items of capital cost and operating cost are the most critical or contribute the greatest amount to the overall operating costs, to show the effect of price of carbon on the costs, to show the effect of including in the capital cost of bio-carbon systems the cost of a new bio-oxidation unit or omitting this cost (for mills that have already amortized the unit), to compare the costs of microlime and minilime treatments, and to compare the costs of treating effluent from new mills and older mills (larger water usage per ton of pulp production). Precision of Estimates These estimates are believed to be within 20 to 30% of the actual costs. The costs of one estimate relative to another for the assumed conditions are expected to be much closer - probably within _ 10%. Since any par- ticular mill will have a different set of conditions, the cost estimates would have to be revised to suit those conditions before a reliable cost for that mill can be obtained. Bases of Estimates The bases used in the design of the plants were discussed earlier in this section and are presented, in part, in Tables 36 and 37. Details of the assumptions used, the design procedures used, and method of estimating plant capital and operating costs are given in Appendix J for Estimate No. 1. Capital Costs - Capital costs are based on total plant costs for existing plants or on published costs from detailed cost studies. All costs were adjust- ed to January 1973 by use of the Engineering News Record Construction Cost Index. The capital costs and some of the operating costs for minilime treat- ment are based on those of the 9 mgd lime treatment plant installed by Continental Can Co. in 1972 at Hodge, La., which was supported by an EPA grant (35). The costs are the actual costs for the complete installation including engineering and contractor's overhead and profit. Their costs were adjusted for a lower clarifier rise rate used in our estimates and for larger flow rates by use of 0.7 as the exponent of the ratio of flows. 156 ------- The capital cost for microlime treatment was derived from the Continental Can cost data with adjustments made for eliminating the carbonation and final clarification step, for lower lime dosage, and for adding a final pH adjustment step. No pH adjustment step was added for the microlime- bio sequence. The capital costs for carbon adsorption were derived from those pre- pared by M. W. Kellogg for EPA in 1969 (17) for 1, 10 and 100 mgd carbon adsorption plants for treating municipal effluent. Costs of each category of equipment of the Kellogg estimate were adjusted for the increased volumes of carbon needed and for other differences in our estimates, as illustrated in Appendix J. The costs for primary clarification (for those estimates using bio- oxidation) were based on the costs for complete units for St. Regis mills. The capital costs for aeration basins were based on the average costs for three installations; two St. Regis mills and a Crown-Zellerbach mill (36). These costs were adjusted for differences in flow rate by use of an exponent of 0.7. Operating Costs - The bases for operating costs are as follows: Amortization - 6.25% of total cost of investment (TCI) per year Repair and Maintenance - 3% TCI/yr. Taxes and Insurance - 2% TCI/yr. Labor - On expected manpower requirements at $5.85/hr. Plant Overhead - 75% of labor costs. Steam - $0.80/1000 Ib. Fuel - $0.60/mil. Btu. Electricity - $0.01/kWh. Lime - Reburned - $12.40/ton CaO Make-up - $24.23/ton CaO Carbon make-up - $0.27/lb and $0.10/lb Credit for treated water to replace mill make-up water = $0.06/kgal RESULTS OF COST ESTIMATES The capital and operating costs for new and older mills and for various treatment sequences are given in Table 38. The new mill was assumed to produce 12,000 gal of effluent per ton of pulp production and the older mill was assumed to produce 18,000 gal/ton. The quality of the water from both mills to the effluent treatment plant was assumed to be the same. The credit for treated water is based on typical mill water cost of $0.10/kgal less $0.04/kgal for pumping the treated water to the mill supply tank. 157 ------- Table 38. COMPARISON OF COSTS FOR TREATING UNBLEACHED KRAFT MILL EFFLUENT IN NEW AND OLDER MILLS BY VARIOUS TREATMENT SEQUENCES WITH CARBON AT $0.27/lb Ul 00 Estimate No. Pulp tons per day Effluent, gal/ton Effluent, mil.' gal/day Treatment Product water quality Color, CU TOC, mg/1 BOD, mg/1 Capital cost, $ million Lime trt. & pH adjustment Carbon adsorption Clarifier & bio-oxid'n Total Operating cost, $/kgal Fixed Labor, o.h.,util. Lime and chemicals Carbon, at $0.27/lb Total op. cost Operating cost, $/ton pulp Operating cost with credit of $0.06/kgal for mill make-up water Net, $/kgal 0.299 Net, $/ton pulp 3.58 1 800 12,000 9.6 microlime -carbon 100 100 150 1.85 4.90 6.75 0.217 0.075 0.033 0.034 0.359 4.30 2 1,600 12,000 19.2 microlime -carbon 100 100 150 2.90 9.00 11.90 0.191 0.064 0.033 0.034 0.322 3.85 3 800 18,000 14.4 microlime -carbon 100 100 150 2.36 6.98 9.34 0.200 0.069 0.033 0.034 0.336 6.04 4 4aa 800 18,000 14.4 primary-bio- carbon 100 100 20 11.92 11.92 5.09 2.39a 17.01 14.31 0.364 0.306 0.127 0.127 0.003 0.003 0.146 0.146 0.640 0.582 11.52 10.48 5 5a 800 18,000 14.4 microlime -bio 300 70 20 2.14 2.14 2.43 0.0a 4.57 2.14 0.098 0.046 0.032 0.032 0.036 0.036 0.166 0.114 2.98 2.04 6 6a 800 18,000 14.4 minilime -bio 150 50 20 3.68 3.68 1.82 0.0a 5.50 3.68 0.118 0.079 0.049 0.049 0.059 0.059 0.226 0.187 4.07 3.37 0.262 3.13 0.276 4.96 0.580 10.44 0.522 9.40 0.106 1.90 0.054 0.96 0.166 2.99 a Capital cost of bio-oxidation not included (for mills with existing bio-oxidation facilities) b 12,000 gal/ton is for new mill, 18,000 gal/ton is for older mill 0.127 2.29 ------- Microlime-Carbon, Large versus Smaller New Mills Microlime-carbon treatment of effluent from a new 800-ton mill (Estimate No. 1) was estimated to require a capital investment of $6.75 million, of which $1.85 million is for lime treatment and pH adjustment and $4.9 million is for carbon adsorption. The operating costs were $0.36/kgal, or $4.30/pulp-ton, or $0.30/kgal and $3.58/pulp-ton including credit for water. The fixed costs for amortization, taxes, insurance, and maintenance amount to $0.22/kgal, or 60% of the total operating costs. When the mill capacity and effluent volume is doubled (Estimate No. 2), the capital cost is increased by 67% to $11.90 million and the operating cost is reduced to $0.32/kgal and $3.85/pulp-ton, or $0.26/kgal and $3.13/pulp-ton with credit for the water. Microlime-Carbon, New versus Older Mill The capital cost for an older mill (18,000 gal of effluent per pulp-ton) is $9.35 million (Estimate No. 3), or 38% greater than for a new mill (Estimate No. 1) having 12,000 gal/pulp-ton. The operating costs (before water credit) per 1000 gal for the older mill are slightly lower than for the new mill but are 40% greater per pulp-ton. Primary-Bio-Carbon Sequence The primary-bio-carbon treatment plant costs considerably more than the plant using microlime-carbon ($17.01 million for Estimate No. 4 versus $9.35 million for Estimate No. 3). The major share of this cost differential is caused by the substantially higher cost of the carbon adsorption system ($11.92 million versus $6.98 million), the remainder of the differential is accounted for by the bio-oxidation system. Estimate No. 4a is the same as Estimate No. 4 except that the capital cost of the bio-oxidation unit is not included in Estimate No. 4a. Even with no capital costs for bio-oxidation, the primary-bio-carbon sequence is 53% greater in capital cost and 74% greater in operating cost than the microlime-carbon sequence (Estimate No. 3). One advantage of sequences using bio-oxidation is that the BOD content of the product water is only about 20 mg/1 as compared to about 150 mg/1 for water from microlime treatment. It appears then for treating mill effluent for reuse that the bio-carbon sequences will be considerably greater in capital and operating costs than the lime-carbon sequence. The cost of the primary clarifier is included in both Estimates Nos. 4 and 4a. To exclude the effect of an existing clarifier, the clarifier capital cost and its share of the fixed cost must be deducted from 159 ------- Estimate No. 4a. It can also be seen that the capital cost for the primary clarifier in Estimate No. 4a is the same as that for the micro- lime treatment in Estimate No. 3. An existing clarifier may be used as part of a microlime treatment system and the costs adjusted accord- ingly. Primary-Carbon Sequence The pilot plant results indicated that the direct use of carbon adsorp- tion after primary clarification would require a very long contact time (about 5 hr) in the carbon adsorbers to reduce the color from 1000 to 100 CU. Because of the indicated high cost of a plant using this sequence, an estimate for it was not included in these studies. However, rough approximations indicated that the capital cost for a 9.6 mgd plant would be about $15 to $20 million and that the operating costs would be about $0.70 per 1000 gal and $13 per pulp-ton. These costs indicate that other sequences involving carbon adsorption would be more economical. Lime-Bio versus Lime-Carbon Sequence Estimates Nos. 5 and 6 were prepared to determine whether effluent could be treated at lower cost by microlime or minilime followed by bio-oxida- tion, than by microlime- or minilime-carbon treatment. Microlime-bio would produce water having a color of 300 CU (or perhaps greater if there is an increase of color during bio-oxidation) which might make it unsuited for reuse. However, the capital cost (including bio-oxidation capital cost) is only $4.57 million, or 49% of that for microlime-carbon (Estimate No. 3) and the operating cost is $0.17/kgal, or also 49% of that for microlime-carbon. If a mill were to add microlime treatment ahead of an existing bio-oxidation unit at no capital charge for bio-oxidation (Estimate No. 5a), the capital costs are reduced to $2.14 million and the operating cost is reduced to $0.11/kgal and $2.04/pulp-ton, or $0.054/kgal and $.0.96/pulp-ton with water credit. If minilime is used instead of microlime ahead of bio-oxidation (Estimate No. 6), the color would be 150 CU, if we assume there will be no color increase during bio-oxidation. (The two mills using minilime-bio- oxidation, Interstate Paper and Continental Can, have experienced large color increases during bio-oxidation of lime-treated water.) The capital costs (including that for bio-oxidation) for minilime-bio are 20% great- er than for microlime-bio and the operating costs are 36% greater. These comparisons indicate that microlime-bio has significantly lower capital and operating costs than minilime-bio but that the color of 160 ------- the water from microlime-bio will be greater and probably limit its suitability for reuse applications. These comparisons also indicate that, if water that has higher color (150 - 300 CU) but lower BOD concentrations (35-40 mg/1) can be reused in the mill, either micro- lime-bio or minilime-bio can provide the water at significantly lower capital and operating costs than microlime-carbon. Microlime versus Minilime The microlime and minilime processes without subsequent treatment steps are compared in Table 39 when treating effluent from an older mill (Estimates Nos. 9 and 11) and from a new mill (Estimates Nos. 8 and 10). Before credit for reused water, the minilime process in both cases has a 57% greater capital cost and a 71% greater operating costs. In the older mill, microlime treatment would cost $1.82/pulp-ton compared to $3.11 for the minilime treatment. The microlime process has an addition- al advantage in that it could be operated when the lime kiln is off since the treated water does not require carbonation and the lime dosage is much smaller. These advantages are partially offset by some dis- advantages: microlime provides treated water having 300 CU as compared to 150 CU for water from the minilime process, the TOG concentration is slightly greater, the control of lime addition to maintain the soluble calcium in a narrow range of about 70 to 90 mg/1 is more demanding than maintaining a fixed dosage, and the greater hardness value of the product water (about 200 mg/1 CaCOs) might limit its use in some parts of the mill. The combinations of microlime and minilime with carbon adsorption are compared in Estimates Nos. 1 and 7 for an older mill. The capital cost for microlime-carbon is 57o less and the total operating cost is 107o less ($0.50/pulp-ton), which gives microlime a slight economic margin when used in the combined sequence to give water meeting the selected standards for reuse. Effect of Carbon Price on Treatment Costs The price of carbon used in all estimates discussed thus far was $0.27 per pound, which was typical for bulk quantities of commercial granular carbons (January 1973). Lower-cost granular carbons are now available that have higher ash contents, and new sources of carbon, such as from the St. Regis process of producing granular carbon from a hydropyrolysis pulp-chemicals recovery process that is now being developed under an EPA grant (Part II of No. 12040 EJU), indicate that lower-cost carbons might be available in the future. To show the effect of carbon price, 161 ------- Table 39. COMPARISON OF COSTS FOR MICROLIME AND MINILIME PROCESSES Estimate No. 11. 9. 10 8. I 7_ Pulp tons per day 800 800 800 800 800 800 Effluent, gal/ton 18,000 18,000 12,000 12,000 12,000 12,000 Effluent, mil. gal/day 14.4 14.4 9.6 9.6 9.6 9.6 Treatment microlime minilime microlime minilime microlime minilime -carbon -carbon Product water quality Color, CU 300 150 300 150 100 100 TOG, mg/1 150 135 150 135 100 100 BOD, mg/1 225 187 225 187 150 117 Capital costs, $ million Lime trt. & pH adjust. 2.36 3.68 1.85 2.91 1.85 2.91 Carbon adsorption 4.90 4.20 o\ Clarifier and bio-oxidation M Total 2.36 3.68 1.85 2.91 6.75 7.11 Operating costs, $/kgal Fixed 0.050 0.079 0.059 0.093 0.217 0.228 Labor, o.h., util. 0.018 0.035 0.018 0.035 0.075 0.087 Lime and chemicals 0.033 0.059 0.033 0.059 0.033 0.059 Carbon, at $0.27/lb 0.034 0.026 Total operating cost 0.101 0.173 0.110 0.187 0.359 0.400 Operating cost, $/ton pulp 1.82 3.11 1.32 2.25 4.30 4.79 Operating costs with water credit3 $/kgal 0.041 0.113 0.050 0.127 0.300 0.340 $/ton pulp 0.74 2.03 0.60 1.53 3.58 4.07 a With credit of $0.06/kgal for mill make-up water ------- it was assumed that carbon having the same properties as the $0.27/lb carbon would be available at $0.10/lb if produced and used on-site. The influence of this lower-priced carbon on the costs of lime-carbon and bio-carbon sequences is shown in Table 40. For microlime-carbon and minilime-carbon plants (Estimates Nos. 1, 2 and 7), the lower price of carbon reduced the capital costs 6 - 9% and the operating costs by 8 - 12% before credit for the reused water and 9-15% after credit for reused water. In the case of the primary- bio-carbon plant (Estimate No. 4) the lower priced carbon reduced the capital costs by 12% and the operating costs by 21%. The greater influence of carbon price on costs for primary-bio-carbon was due to the much greater dosage and size of carbon adsorption units to remove the higher concentration of color after bio treatment. CONCLUSIONS FROM THE COST ESTIMATES 1. The microlime-carbon sequence is the lowest cost process for meeting the product water quality standards of 100 CU and 100 mg/1 TOG for mill reuse. The capital costs for treating 9.6 mgd from a new 800 ton/day mill are estimated to be $6.75 million, and the total operating costs are estimated to be $0.30/kgal and $3.58/pulp-ton with credit for the reused water. 2. Doubling the plant pulping capacity increases the capital cost by 67% and reduces the operating costs about 12% to $0.26/kgal and $3.13/pulp- ton with credit for water. 3. Microlime-carbon treatment for an older mill producing 50% more water per pulp-ton but having the same pulp capacity requires a 38% greater capital cost, and the operating costs are about the same per 1000 gal ($0.28/kgal) but the cost per pulp-ton was greater ($4.96), both after credits for the reused water. 4. A reduction in price of carbon from $0.27/lb to $0.10/lb reduces the capital cost for lime-carbon plants about 8%, and operating cost about 13%. 5. The capital and operating costs for primary-bio-carbon are 1.5 to 2 times those for microlime-or minilime-carbon, and the costs for primary- carbon sequence are about 15% greater than those for primary-bio-carbon sequence. 6. The minilime-carbon process has about 5% greater capital and 147» greater operating costs than the microlime-carbon process. 7. For reuse of water in a new 800 ton/day mill, for applications that can tolerate a color of 300 CU and a hardness of 200 mg/1 CaCO^, micro- lime treatment could be used at a capital cost of only $1.85 million and an operating cost of only $0.11/kgal or $1.32/pulp-ton, or $0.05/ kgal and $0.60/pulp-ton after credit for the water. The costs for microlime treatment are about 40% below those for minilime alone before credit for the water. 163 ------- Table 40. EFFECT OF CARBON PRICE ON TREATMENT COSTS Estimate No. I 2 4 _7 Pulp tons per day 800 1600 800 800 Effluent, gal/ton 12,000 12,000 18,000 12,000 Effluent, mil. gal/day 9.6 19.2 14.4 9.6 Treatment microlime-carbon microlime-carbon bio carbon minilime-carbon Carbon price, $/lb 0.27 0.10 0.27 0.10 0.27 0.10 0.27 0.10 Capital costs, $ million Lime trt. & pH adjust. 1.85 1.85 2.90 2.90 2.91 2.91 Carbon adsorption 4.90 4.35 9.00 7.91 11.92 9.95 4.20 3.75 Clarifier & bio-oxid'n 5.09a 5.09a Total 6.75 6.20 11.90 10.81 17.01 15.04 7.11 6.66 Operating costs, $/kgal Fixed Labor, o.h., util. Lime and chemicals Carbon Total operating cost With water credit*3 a Including capital cost of bio-oxidation b With credit of $0.06/kgal for mill make-up water 0.217 0.075 0.033 0.034 0.359 0.299 0.199 0.075 0.033 0.012 0.319 0.259 0.191 0.064 0.033 0.034 0.322 0.262 0.173 0.064 0.033 0.012 0.282 0.222 0.364 0.127 0.003 0.146 0.640 0.580 0.322 0.127 0.003 0.054 0.506 0.446 0.228 0.087 0.059 0.026 0.400 0.340 0.214 0.087 0.059 0.010 0.370 0.310 ------- Microlime-blo and minilime-bio processes are significantly lower in capital costs and 33 to 50% lower in operating costs than microlime- carbon, but the color of the product water will be greater, particularly if a color increase during bio-oxidation occurs as it does in the two mills using lime-bio treatment. 165 ------- SECTION XV REFERENCES 1. Black, A. P., and Christman, R. F. "Characteristics of Colored Surface Waters", J. Am. Water Works Assoc. 55 (6):753(1963). 2. Black, A. P., and Christman, R. F. "Chemical Characteristics of Fulvic Acids", J. Am. Water Works Assoc. 55 (7):897(1963). 3. Berger, H. F. and Brown, R. I. "The Surface Reaction Method for Color Removal from Kraft Bleaching Effluents", Tappi 4.2 (3) :245(1959). 4. Berger, H. F., and Smith, D. R.. "Waste Water Renovation", Tappi 51 (10):37A(1968). 5. Davis, C. L. Jr. "Tertiary Treatment of Kraft Mill Effluent Including Chemical Coagulation for Color Removal", Tappi 52 (11):2132(1969). 6. English, J. N., et_ al_. "Removal of Organics from Waste Water by Activated Carbon", Water-1970. AlChE Symposium Series 67 (107):147 (1971). 7. Gould, M. "Lime-Based Process Helps Decolor Kraft Wastewater", Chem. Eng. 78 (3) :55(1971). 8. Gould, M. "Color Removal from Kraft Mill Effluent by an Improved Lime Process", Tappi 56 (3):79(1973). 9- Masse, A. N. "Removal of Organics by Activated Carbon", presented at 156th National A.C.S. meeting (Sept. 1968). 10. McGlasson, W. G. Thibodeaux, L. J., and Berger, H. F., "Potential Uses of Activated Carbon for Wastewater Renovation", Tappi 49 (12): 521(1966). ~ 11. McGlasson, W. G. "Treatment of Pulp Mill Effluents with Activated Carbon", NCASI Tech. Bull. 199(1967). 12. Miller, C.O.M., and Clump, C. W. "A Liquid Phase Adsorption Study of the Rate of Diffusion of Phenol from Aqueous Solution into Activated Carbon", AlChE Journal 16 (2):169(1969). 13. Morris, J. C., and Weber, W. J. "Adsorption of Biochemically Resistant Materials from Solution", U.S.P.H.S., AWTR-9(1964). 166 ------- 14. Spruill, E. L. "Color Removal and Sludge Recovery from Total Mill Effluent", Tappi 56 (4):98(1973). 15. Timpe, W. G., Lang, E. W., .et .al. "The Use of Activated Carbon for Water Renovation in Kraft Pulp and Paper Mills", presented at 7th TAPPI Water and Air Conf., Minneapolis (June 1970). 16. Weber, W. J., Hopkins, C. B., and Bloom, R. "Physicochemical Treat- ment of Wastewater", J. Water Pollution Control Federation 42 (!)• 83(1970). ' ~~ ~ 17. Cover, A. E., and Pieroni, L. J. "Appraisal of Granular Carbon Contacting", Report Numbers TWRC-11 and TWRC-12. FWPCA, U.S. Depart- ment of the Interior (May 1969). 18. Anon. "Color Removal from Kraft Pulping Effluents by Lime Additon", Interstate Paper Corp. for EPA, Final Report, Program #12040 ENC, Grant #WPRD 183-01-68 (Dec. 1971). 19. Anon. "Process Design Manual for Carbon Adsorption", EPA Technology Transfer (October 1973). 20. Bloom, R., .££ _§!.. "New Technique Cuts Carbon Regeneration Costs", Enviro. Sci. and Tech. 3 (3):214(1969). 21. Berg, E. L., je_C a±. "Thermal Reactivation of Spent Powdered Carbon Using Fluidized Bed and Transport Type Reactors", Chem. Eng. Prog. Symposium Series 67 (107):154(1970). 22. Herbet, A. J. "A Process for Removal of Color from Bleached Kraft Effluents Through Modification of the Chemical Recovery System", NCASI Tech. Bull. 157 (June 1962). 23. Anon. "Design Workshop for Water Pollution Control", Environmental Science Services (November 12, 1969). 24. Smith, R. "Cost of Conventional and Advanced Treatment of Wastewater", J. Water Pollution Control Federation 40 (9):1546(1968). 25. DiGregorio, D. "Cost of Wastewater Treatment Processes", TWRC-6. FWPCA. U.S. Department of the Interior (Dec. 1968). 26. Anon. "Cost of Clean Water IV-- Projected Wastewater Treatment Cost in the Organic Chemicals Industry", FWPCA. U.S. Department of the Interior (Jan. 1969). 167 ------- 27. Anon. "Cost of Clean Water III - Industrial Waste Profiles No. 3 - Paper Mills".FWQA Publication No. IWP-3. U.S. Department of the Interior (Nov. 1967). 28. Eckenfelder, W. W. "Economics of Waste Treatment", Chemical Engineer- ing 67, (18):109(1969). 29. Page, J. S. "Estimator's Manual of Equipment and Installation Costs", Gulf Publishing Co., Houston,(1963). 30. Guthrie, K. M. "Capital Cost Estimating", Chemical Engineering 76 (6):114(1969). 31. Berger, H. E. "Development of an Effective Technology for Pulp and Bleaching Effluent Color Reduction", presented at NCASI Meeting (Feb. 1969). 32. Davies, D. S. "Activated Carbon Eliminates Organics", Chemical Engineering Progress 60 (12):46(1964). 33. Hager, D. G., and Rizzo, J. L. "A Chemical and Physical Wastewater Treatment Process", presented at AlChE Meeting, Atlanta (Feb. 1970). 34. Cooper, J. C., and Hager, D. G. "Water Reclamation with Activated Carbon, Chemical Engineering Progress 62 (10):85(1966). 35. Spruill, E. L, Private Communication Concerning Costs for Continental Can Co. Lime Treatment Plant (July 1973). 36. Amberg, H. R. "Water Pollution Control in Pulp and Paper Industry", Industrial Water Engineering 7 (11):26(1970). 37. Bishop, D. F., et aLt "Physical-Chemical Treatment of Municipal Wastewater", J. Water Pollution Control Federation 44 (3):361(1972). 38. "Standard Methods for Examination of Water and Wastewater", 13th Edition, APHA, New York(1971). 39. Ganczarczyk, J. "Fate of Lignin in Kraft Effluent Activated Sludge Treatment", J. Water Pollution Control Federation 45 (9):1898(1973). 40. Timpe, W. G., Lange, E. W., and Miller, R. L. "Kraft Pulping Effluent Treatment and Reuse-State of the Art", EPA R2-73-164. Program 12040 EJU (Feb. 1973). 168 ------- 41. Timpe, W. G., and Lange, E. W. "Processes for Reducing the Organic Carbon Content of Water Contaminated with Organic Compounds by Continuous Counter Current Multistage Treatment with Activated Carbon" U.S. Patent 3,763,040 (October 2, 1973). 42. Allan, G. G., Mauranen, P., Neogi, A.N. "Fiber Surface Modification, Part IX, Removal of Phenolic Pollutants from Water by Oxydative Coupling to Lignocellulosic Substrates", Paperi ja Fuu No. 6 (1971); see also "Pollution Abatement by Fiber Modification", EPA Final Report Program 12040-EFC (Jan. 1971). 43. Swanson, J. W., Dugal, H. S., et al. "Kraft Effluent Color Character- ization Before and After Lime Treatment", EPA R2-73-141, Project 12040 DKD (Feb. 1973). 44. Gulp, R. L., Gulp, G. L. "Advanced Wastewater Treatment", Van Nostrand Reinhold Co., New York, p.253(1971). 45. Timpe, W. G., and Lang, E. W. "Activated Carbon Treatment of Kraft Mill Effluent for Reuse", Water-1973, AIChE Symposium Series 70(136): 579(1974). 46. Timpe, W. G., and Lang, E. W. "Activated Carbon and Other Techniques for Color Removal from Kraft Mill Effluents", Procedings of the XV EUCEPA Conference, Rome, ATICEljCA in coop. w. FAO, Rome, p.535(1973). 47. Timpe, W. G., and Lang, E. W. "Activated Carbon Treatment of Unbleach- ed Kraft Effluent for Reuse-Pilot Plant Results", presented at TAPPI Environmental Conference (May 1973). 48. Thibodeaux, L. J., and Berger, H. F. "Laboratory and Pilot Plant Studies of Water Reclamation", NCASI Tech. Bull. 203 (June 1967). 49. Schoeffel, E. W., and Zimmermann, F. J. "Wet Air Oxidation of Combustible Materials Adsorbed on Carbonaceous Adsorbent", U.S. Patent 3,386,922 (June 4, 1968). 50. Koches, C. F. "Transport Reactor with a Venturi Tube Connection to a Combustion Chamber for Producing Activated Carbon", U.S. Patent 3,647,716 (March 7, 1972). 51. Koches, C. F., and Smith, S. B. "Reactivate Powdered Carbon", Chem. Eng. _79 (9):46(1972). 52. Bennett, D. J., Dence, C. W., £t al. "The Mechanism of Color Removal in the Treatment of Spent Liquors with Lime", TAPPI 54 (12)-.2019(1971). 169 ------- SECTION XVI PUBLICATIONS AND PATENTS PUBLICATIONS AND PAPERS PRESENTED Timpe, W. G., Lang, E. W., et al., "The Use of Activated Carbon for Water Renovation in Kraft Pulp and Paper MillsV, presented at 7th TAPPI Water and Air Conf., Minneapolis (June 1970). Timpe, W. G., "Removal of Color from Kraft Mill Effluents by Activated Carbon and Combined Treatments", presented at the NCASI Southern-South Central Regional Meeting, Atlanta, Ga., (July 1970). Timpe, W. G., and Lang, E. W., "The Removal of Color from Kraft Mill Effluents by Activated Carbon and Combined Treatments", presented at the 1971 Annual Meeting, Gulf Coast Section, TAPPI, Biloxi, Miss. (May 1971). Timpe, W. G., "A Chemical Engineering Approach to a Closed Water Cycle in the Kraft Industry", presented at the Ninth API - TAPPI Research Conference, Seattle, Wash., (Oct. 1971) (presented by G. R. Webster of EPA). Timpe, W. G., Lang, E. W., and Miller, R. L., "Kraft Pulping Effluent Treatment and Reuse-State of the Art", EPA R2-73-164, Program 12040 EJU (Feb. 1973). Timpe, W. G., and Lang, E. W., "Activated Carbon Treatment of Unbleached Kraft Effluent for Reuse-Pilot Plant Results", presented at TAPPI Environ- mental Conference (May 1973). Timpe, W. G., and Lang, E. W., "Activated Carbon and Other Techniques for Color Removal from Kraft Mill Effluents", Procedings of the XV EUCEPA Conference, Rome ATICELCA in coop. w. FAO, Rome, p.535(1973). Timpe, W. G., and Lang, E. W., "Activated Carbon Treatment of Kraft Mill Effluent for Reuse", Water-1973, AIChE Symposium Series 70(136);579(1974). PATENTS AND PATENTS APPLIED Timpe, W. G., and Lang, E. W., "Processes for Reducing the Organic Carbon Content of Water Contaminated with Organic Compounds by Continuous Counter Current Multi-stage Treatment with Activated Carbon", U.S. Patent 3,763,040 (October 2, 1973) (This is the so-called FACET process). Timpe, W. G., "Process for Reducing the Organic Carbon Content and Improving the Color of Aqueous Plant Effluents", U.S. Patent Application Serial No. 301,400 (Nov. 1, 1972). 170 ------- Timpe, W. G. , "Separation of Organic Compounds in Aqueous Condensates from Kraft Pulping", U.S. Patent Application Serial No. 301,401 (Nov. 1, 1972). 171 ------- SECTION XVII GLOSSARY BOD - 5-day biological oxygen demand. CU - Color units measured according to APHA-NCASI standard method (see Appendix B). FPCU - Color after filtration through 2 micron paper filter, no pH adjustment (see Appendix B). MPCU - Color after filtration through 0.8 micron Millipore filter, no pH adjustment (see Appendix B). TIC - Total inorganic carbon. TOC - Total organic carbon. Lime treatment - Effluent treatment based on the use of lime - see mini- lime and microlime treatment. Lime-carbon Primary/lime-carbon - Terms used interchangeably for an effluent treat- ment system involving lime treatment without prior or separate primary clarification followed by carbon adsorption. Minilime treatment - Effluent treatment with the minimum lime dose that reaches near-leveling off of color at about 85% removal; dosage approx. 1000 mg/1 CaO, leaving about 400 mg/1 dissolved Ca in solution which must then be removed by carbonation (see Section VIII). Microlime treatment - Effluent treatment achieving about 70% color removal; dosage adjusted to achieve 80 mg/1 dissolved Ca, avoiding need for subsequent removal of Ca by carbonation; dosage approx. 500 mg/1 CaO (see Section VIII). Primary-carbon - Effluent treatment sequence involving primary clarification followed by activated carbon adsorption. Bio-carbon Primary- b io - c a rbo n - Terms used interchangeably for an effluent treatment system involving primary clarification, biological oxidation and carbon adsorption. Adsorption isotherm - Relates the quantity adsorbed per unit of adsorbent to the concentration of adsorbate at equilibrium. 172 ------- SECTION XVIII APPENDICES Page A. Water Quality Standards for Specific Uses 175 Table 41: Potential Water Quality Standards for Specific 176 Water Uses B. Procedures for Isotherms, Laboratory Lime Treatment, 179 Laboratory Bio-Oxidation and Analyses C. Detailed Information on Pilot Plant Equipment 183 D. Printout of Daily Summaries of Conditions and Results for 185 Bio-Carbon Sequence E. Printout of Daily Summaries of Conditions and Results for 186 Primary-Carbon Sequence F. Printout of Daily Summaries of Conditions and Results for 187 Lime Treatment G. Printout of Daily Summaries of Conditions and Results for 188 Lime-Carbon Sequence H. Generalized Procedure for Designing Plant for Treating Kraft 189 Effluent in Granular Activated Carbon Columns I. Relationships Useful in Design of Downflow Carbon Adsorption 193 Columns J. Cost Estimate No. 1 - Micro lime-Carbon Treatment of 9.6 MGD 194 of Effluent from a New 800 TPD Unbleached Kraft Pulp and Paper Mill Figure 30: Flow Diagram for Microlime-Carbon Treatment of 195 Unbleached Kraft Effluent for Reuse Major Conditions and Assumptions Design of Lime Treatment Unit •I QQ Design of Carbon Adsorption Units 173 ------- Page Capital Cost of Lime Treatment Unit 202 Capital Costs for Carbon Adsorption Unit 203 Capital Cost of pH Adjustment after Lime Treatment 204 Capital Cost of Backwash Clarification and Surge Tanks 204 Operating Costs for Lime Treatment 204 Operating Costs for Carbon Treatment 205 Total Capital and Operating Costs for Estimate No. 1 206 K. English to Metric Units Conversion Table 207 174 ------- APPENDIX A WATER QUALITY STANDARDS FOR SPECIFIC USES The objective of maximizing water reuse economically very likely will require avoidance of non-essential treatment by more closely matching treated water quality to actual use requirements. At present, no comprehensive listing appears to be available on water quality standards for specific water uses throughout the pulp and paper mill. Although certain standard use and reuse patterns exist through- out the industry, no standards have been developed for the water input for many individual process steps. A substantial amount of reuse is at present handled to fit specific mill circumstances, and disagree- ment is found as to what represents suitable water. The situation is made more complex by the fact that certain contaminant levels cannot be defined in terms of concentration in the water feed stream alone. If the receiving system is a water recycling system, consideration of maximum concentration in the receiving recycle system together with the feed and blowdown volumes is required. In other cases, a system downstream must be considered together with the direct receiving system. Temperature requirements must also be considered. A beginning was made in October, 1969, to collect quality standards data and guidelines, resulting in the data in Table 41 and notes to this table. 175 ------- Table 41. POTENTIAL WATER QUALITY STANDARDS FOR SPECIFIC WATER USES USE Paper machine Headbox & wire showers low volume, high pressure, make-up Felt showers Pulp mill Brown stock washer) fresh showers 1250-1500 last drum ) contain, showers 1250-1500 White liquor system White liquor filter (Eimco System) 100-200 Mud washer Smolt dissolving 23) Bleaching system Fresh water in counter-current system 19) 19) Woodyard 17) Showers 0-2000gpm 30) <105 18) Cooling and sealing water 20) Boiler feed water Cooling water S T A N D A R D Temp. °F 135(board)3)* 103(paper) 160 160 125-200 125-200 BOD Organics 2-5 4) 3, 4) 2-5 12) 13) 13) 21) 21) 21) Total Color Turbid, solids CU JTU mg/1 50-100 10-15 <250 5) 50-100 10-15 100 10) 13) 50-100 50-100 50-100 Susp. Diss. solids solids pH 0-5 6) 5) 2) 6-10 8) fiber- 9) 11) free 9) 11) 16) 16) Hardness Cl SO*" Na+ mg/1 mg/1 mg/1 mg/l 7) 2000 1,2) 7) 250 250 9) 14) 15) 22) Turbine and evaporator 24) Small heat exch.; seal water on 26) pumps, bearings; air cond.; air compr. <95 0-10 26) <50 6-9 6 or higher 18) No Ca 25) 24) .2-.5 * See notes at end of table ------- Notes to Table 41. WATER QUALITY STANDARDS FOR SPECIFIC WATER USES 1. Corrosion is problem. 2. 2000 ppm of S0^= in white water is probably the TAPPI established limit in conjunction with wet strength resin use. Concentration in total white water, hence level in this make-up stream, is function of make-up flow and S04= (incoming with pulp and A12(S04)3 and H2S04 added for pH adjustment. 3. Refining adds 5°F. 4. No rosins or resins should be present to avoid plugging up machine showers; also, keep slime down by slimicides and residual Cl of 0.2 mg/1 (this is done anyway). ^ 5. Watch in conjunction with chloride and sulfate build-up. 6. Related to turbidity specification. 7. Possibly need a Ca spec., a minimum being required for proper functioning of sizing or wet strength additives; 40-50 ppm Ca suggested. Ca(ECO^)2 hardness undesirable. 8. Adjust to maintain stability of water. 9. Water must be free of contaminants which would form scale in heat exchanger or deposit on filter/washer wires. Low Ca and Fe desired. 10. Lower limits apply if pulp is to be bleached. 11. Probably 7-12. Essential to avoid lignin precipitation at this point. 12. Low molecular weight organics pose no direct problems except as nutrients in paper machine water system, resulting in slime growth. Could be high as "contaminated condensate" without carry-over (i.e., clean evaporator condensate). 13. Can be contaminated up to perhaps 1% BLS. There is a concentration of about 2% to 4% BLS in collection system under drum. Firmer data need to be established through tests and calculations. 14. A low Na content is desired. A certain Na limit may be calculated from washing requirements: washing done on 1270 Na in liquid; 0.570 to 2.0% Na desired in dry mud, at 60% to 70% solids in cake. 15. Since this is the final washing stage for the lime mud, low Na content in wash water may be even more critical than as discussed in note 14. 16. Low pH acceptable if alkali consumption can be limited. 17. Only concern is to avoid excessive foaming. Otherwise, a closed recycle system is now contemplated for Mill B. 18. Uncertain. Consider effect of higher temperatures on extraction of wood acids and on C0£ absorption from air. Corrosion due to low pH is to be avoided. 19. Consider that there are three water systems to a bleach plant (batch): a. Cold or ambient temperature water b. 115°F water c. 160°F water 20. In Mill A water is first used as cooling water before other uses. 21. Presence of low molecular weight organics considered not presenting a problem. 177 ------- 22. As low as possible. Chloride build-up in liquor system to be avoided. 23. Use of weak wash water is standard. Contaminated condensate (such as clean evaporator condensate) can be used. 24. Relatively clear water, stabilized, non corrosive, non scaling. 25. Either no suspended solids, or otherwise chemically controllable. 26. Stabilized, non corrosive, non scale forming, non slime forming. 178 ------- APPENDIX B PROCEDURES FOR ISOTHERMS, LABORATORY LIME TREATMENT, LABORATORY BIO-OXIDATION AND ANALYSES PROCEDURE FOR MAKING CARBON ADSORPTION ISOTHERMS(44) 1. Pulverize a representative sample of the carbon (10-20g) and screen it through a 325-mesh sieve. Oven-dry the carbon for 3 hr at 150°C. 2. Obtain a representive sample of the wastewater to be tested, making sure that the pH is that of the original source of the sample. In the laboratory program (Section 4), samples were adjusted to the average pH for a given stream if the samples were within 0.7 pH units of the average, or not used if outside of this range. 3. Weigh out 5 amounts of carbon in disposable weighing dishes. For pulp mill effluents, use 0.3, 0.5, 1.0, 2.5, and 5.0 g per 500 ml of sample (0.6 - 10.0 g/1). 4. Add 500 ml of the water to be tested to a 1-liter beaker, heat to 40°C, add one of the weighed dosages of carbon, stir vigorously (100 rpm on a Phipps and Bird gang-stirrer) for 15 minutes while maintaining the water at 40°C. Filter about 100 ml of the water through Whatman No. 2 filter paper using a pressure-type lab filter (to prevent loss of volatile organics). (Isotherms, except those on condensate streams, prepared prior to 6/71 were made at 95°C and 2 minutes, but were equivalent to those made at 40° and 15 minutes.) 5. Repeat for the other dosages of carbon with separate 500-ml quantities of water to be tested. 6. Analyze the filtrate from each dosage of carbon for the impurity of interest and express in mg/1 or APHA color units. Normally, isotherms were prepared for both TOG and color. 7. For each carbon dosage, subtract the final from the initial impurity concentration, and divide this difference by the dosage of carbon in g/1. This number is the loading of impurity per g of carbon used. On 3 x 3 cycle log-log graph paper, plot loading on the y-axis versus remaining concentration on the x-axis. The isotherm is completed by drawing a straight line through the plotted values. (Sometimes the line has two slopes and has a dog-leg shape.) Extrapolate the line to the initial concentration. The loading at this intercept is the ultimate capacity of the carbon for that effluent. 179 ------- LABORATORY BIO-OXIDATION Bio-oxidation was standardized so that approximately a 70% reduction in BOD would be obtained. The following conditions were used: Aeration for 40 hours with filtered air; ammonium'nitrate addition of 3 mg/1 per 100 mg/1 BOD in the feed; phosporic acid addition of 0.6 mg/1 per 100 mg/1 BOD; strained sewage addition of 25 mg/1 per 100 mg/1 BOD. LABORATORY LIME TREATMENT The standard laboratory lime treatment consists of lime addition, separation of precipitate, carbonation to pH 10.5 to remove excess dis- solved lime, separation of precipiate, and finally carbonation to pH 7.0. A lime dosage of 5000 mg/1 CaO was used unless otherwise specified. This is a compromise dosage and is well above the so-called minimum lime (mini- lime) dosage of 1-2,000 mg/1 (5,14) (also called stoichiometric), but below the so-called massive lime dosage of 20,000 mg/1 (22). The lime dosage was slaked in 500 ml of effluent for treatment of 6-10 liter of effluent. The slurry was poured into the effluent and mixed in a 5 gal tank. The mixture was held under slow stirring for 5 minutes. The mix- ture was allowed to settle and the supernatant was then filtered through a Whatman No. 2 filter. The filtrate was placed into another tank and bottled C02 was bubbled into the sample using a gas dispersion tube. The sample was slowly stirred during carbonation which continued until the pH reached 10.5. The slurry was again allowed to settle and the supernatant was filtered through a Whatman No. 2 filter. Carbonation was then continued down to a pH of 7.0. This treatment follows the NCASI procedure (22). LABORATORY PRIMARY CLARIFICATION All samples were filtered through a Whatman No. 2 paper filter to simulate the result of primary clarification. ANALYTICAL METHODS The methods used for analyses of samples in the work covered by this report are listed below. The standard method number given refers to the method in the 13th edition of Standard Methods (38). Methods modified by St. Regis Paper Company are available on request. Color - Std Method 206A in which the pH of the sample is adjusted to 7.6. and the sample filtered through a 0.8 micron Millipore filter and light transmittance is measured at 465 nm on a Spectronic 70 spectro- photometer, as adopted by the NCASI. Prior to 2/4/71, the NCASI adopted 180 ------- method called for filtration through 0.45 instead of 0.8 micron Millipore filter, and light transmittance measurement at 430 nm instead of 465 nm. In both cases, color is determined from a calibration curve using a cobalt chloroplatinate standard, where equivalent mg/1 of cobalt chloroplatinate is termed color units or "CU". Based on pilot plant experience, it was found desirable to determine the parameter of "color" not only by the industry standard APHA-NCASI method but also by modified procedures which measure apparent color. One modified procedure involved 0.8 micron Millipore filtration prior to adjustment to pH 7.6. (Used after lime treatment, this procedure ensures more complete removal of suspended solids.) Another modified procedure consisted of measuring color after filtration through a Whatman Ho. 2 filter paper and adjust- ment to pH 7.6, referred to as "7.6 FPCU". A third modified procedure consisted of measuring color after filtration through Whatman No. 2 filter paper without pH adjustment, noted as "FPCU". A fourth modified procedure consisted of measuring color after filtration through an 0.8 micron Millipore filter without any pH adjustment, noted as "MPCU". TOG and TIC - Std Method 138 using a Beckman 915 total carbon analyzer; sample filtered through Whatman No. 2 paper filter. jjH - Std Method 144A Turbidity - Std Method 163A using a Hach Model 2100 turbidimeter. Total Suspended Solids (TSS) - Std Method 224C using Whatman GF/C glass fiber filter discs. Conductivity^ - Std Method 154 using Yellow Springs Instrument Co. Model 31 conductivity bridge. BOD-5 - Std Method 219 Dissolved Oxygen (DOl - Std Method 218F using a Weston and Stack dis- solved oxygen analyzer Model 300. Na2S - Mercuric chloride titration method according to E. Bilberg, Norsk Skogindustri. 11/58 470 (1958), modified by St. Regis Paper Company. Na2Sx and Na2S203 - by unpublished methods developed by St. Regis Paper Co. Sulfate - Std Method 156B, gravimetric method. Chloride - Orion Instrument Company selective ion electrode method. Metal ions - Std Method 129A using atomic adsorption with the Perkin- Elmer Model 403 spectrophotometer. Volatile Acids - by gas chromatographic method of S. M. Aronovic, et al, TAPPI 54 1963 (1971) and modified by St. Regis Paper Company. 181 ------- Neutral Volatiles - by gas chromatographic method of B. G. Turner and I. T. Van Horn, presented at TAPPI National Meeting, February 1970, and modified by St. Regis Paper Company. 182 ------- APPENDIX C DETAILED INFORMATION ON PILOT PLANT EQUIPMENT 1. Basin volume - 25,000 ft3 or 187,000 gal, retention time at 30 gpm = 4.3 days. aerator - 5 hp Chemineer clarifier - area 510 ft2, rise rate at 30 gpm = 86 gpd/ft2 = 0.06 gpm/ft2 2. Surge tank, 10,000 gal, for backwash water (normally) or surge capacity between stages of processing 3. Lime treatment - by Potter and Rayfield lime treater, dimensions - 10'd x ll'-6"h, 4900 gal. rise rate at 30 gpm - 550 gpd/ft2 =0.38 gpm/ft2 retention time in mixing zone - 20 min. retention time total = 2.7 hr. lime feed rate for 30 gpm and 1300 mg Ca(OH)2/l = 20/lb hr. lime feeder - Wallace and Tiernan Series A-728 volumetric feeder with 41 ft-* hopper and 50 gal mixing tank for lime slurry. carbonator, dimensions - 7'-6"d x ll'-6"h, 2920 gal, rise rate at 30 gpm = 980 gpd/ft2 =0.68 gpm/ft* C02 from cylinders, pH automatically controlled (normally at 10.5) pH adjust tank, 200 gal, C0? through controller to reduce pH to about 8.5 4. Carbon columns - by Potter and Rayfield four columns piped to permit any combination of series and parallel flow up-flow or down-flow. dimensions - 3'd x 15'h, 430 gal in bed, 795 gal total carbon bed - 10" supported on sand-gravel plus 5' free-board at top. charge - 1600 Ib ICI America (Atlas Chemical) granular Darco 20 x 40 mesh. retention time in bed at 15 gpm = 0.6 hr = 2.4 hr total. flow yel. at 15 gpm = 2 gpm/ft2 =1.7 bed volumes/hr each col. backwash = 30 gpm, normally every 48 hr, after air backwash. carbon density = 25 lb/ft. 5. FACET, three stirred tanks, gravity flow of water, carbon slurry pumped counter-current to water flow, with sloping baffle to provide settling zone at top of tank. 183 ------- 5. FACET (continued) 3'd x 5'h (straight side) 170 gal in mixing zone, 248 gal total retention time in mixing zone at 10 gpm=17 min ea = 51 min total at 15 gpm=ll min ea = 33 min total rise rate at 10 gpm=1.4 gpm/ft at 15 gpm=2.1 gpm/ftz weight of carbon in each tank at 10% solids = 170 Ib. operating densities = 10-20% in mixing zone. Carbon - ICI America, Darco XPT 40 x 140 mesh. carbon feeder - Wallace and Tiernan, Series A-728 volumetric feeder and 90 ft^ hopper. 6. Duo-media filters - by Potter and Rayfield before carbon columns - 4'd, 12.5 ft2 area, I1 of 0.4 x 0.6mm sand plus I1 of 1 x 3 mm anthracite, flow velocity =2.1 gpm/ft2 at 15 gpm. after FACET - 3'd, 7.05 ft2, I1 of sand olus I1 of anthracite, flow velocity =2.1 gpm/ft^ at 15 gpm. 7. Instrumentation - by Foxboro recorders - a. 10 temperatures b. pH of controlled streams - carbonator and pH adjust flow rates - to lime system, to carbon columns, and to FACET c. measured properties of sample streams = pH, color, turbidity, conductivity. 12 sample streams that can be monitored instrumentally, Tenor drum programmer switches flow of sample stream to measurement sensors and controls recording of properties on recorder. Controlled parameters = 3 water flow rates and 2 pH values (carbonator and pH adjust). timer controlled blow-down of sludge from lime treater and carbonator. 8. Parameters determined in lab analyses of samples from carbon treat- ments - color, pH, total inorganic carbon, total organic carbon, soluble calcium (at least daily). volatiles and low molecular weight organics, suspended solids, BOD (every few days). 184 ------- APPENDIX D PRINTOUT OF DAILY SUMMARIES OF CONDITIONS AND RESULTS FOR BIO-CARBON SEQUENCE DATE=08/15/72 TIME=16:26s 8 RR012 - WATER TR DATE TIME OF DAY ATMENT PILOT PLANT FILE DUMP -OLUMN F CU TO BASIN PAPR ICU/LJ MP CU FM BASIN PAPR MP PAPR MP HAPR MP CU PCRT-!> HAPR CU PORT-0 CU PCRT-2 TBC TO BASN MG/L TOC FM BASIN TOC PORT-0 TOC PORT-2 po BoTTTD BASN M7L Ui BOO FM BASIN BOO FM LAST CCL DO FM BASIN MG/L PH TO COLUMNS CDND MICPC MHOS TURB JTU TO BASN FM BASN PRESS DROP COLS PRESS DROP F-l CHARGE LB CUM MRS OPTO TOTAL REMOVAL * TOC REMOVAL * TOTAL 'REMOVAL * BOO REMOV bASN X OC KtM X SYS IN COLOR % OF FEED TQC X OF FEED OPERATING DAY /> COLOR RATE TQT*C TOC LOG,RATE T*C DISK FILE LOCN I045. 00 48 0.0 00 0.0 00 0.0 00 0.0 001000.00 00 570.00 0 0.0 _.0 0.0 1000.00 750.00 65. 910. 750. 0. 5. 0.0 0.0 lt 0.0 0.0 0.0 0.0 750.00 585.00 338.00 233.00 0.0 " ' 0.0 0.0 >.0 0.0 0.0 585.00 435.00 126.00 0.0 151.00 99.00 0.0 0. 0.0 0.0 84.00 68.00 0.0 0.0 1600.001600.001600.001760 417.50 345.50 188.50 38 .0 .0 .0 .0 00 ~0 82.91 34.44 62.25 0.0 0.0 15.15 0.0 0.0 0.0 0.0 19.05 16.18 0.0 0.0 0.0 0.0 65.5- 57.00 71.91 9.88 145 146 ...60 28.24 45.03 0.0 71.91 9.88 147 "I* 7.2B 17.09 37.75 0.0 71.91 9.88 148 13FB 0800 15.60 0.0 ... 8:8- 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0. 1165.00 825.. 715.00 525.01 0.0 0.0 825.0' 0.0 0.0 ~-8t«—- 8:8- 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 675.00 675.00 540.00 345.00 259.00 0.0 0.0 0.0 0.0 540.00 428.00 0.0 0.0 0.0 0.0 119.00 85.00 60.00 53.00 0.0 0.0 0.0 0.0 0.0 0.0 0.0 7.80 0.0 0.0 0.0 0.0 iI 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 __2fl__aQ ___ i*. IQ ___ jL o.o o7o o.o tr. 0.0 0.0 12.02 18. 1C 0.0 0.0 2.00 0.0 0.0 0.0 3.00 0.0 0.0 0.0 0.0 0.0 1600.001600.001600.001-760.00 463.50 391.50 234.50 84.50 520.59 271.74 132.42 35.72 479.54 232.85 113.84 20 537.73 288.38 138.47 40.49 UM COLOR LOG 63.82 29.05 12.84 2.90 14.35 4.86 74.59 35.96 16.37 4.40 74.83 28.57 63.S7 0.0 0.0 29.41 0.0 0.0 28.57 73.43 71.43 59.00 0.0 11.67- 0.0 0.0 21.C_ 48.25 50.42 0.0 5.88 36.22 44.54 0.0 0.0 18.87 0.0 0.0 8.40 25.17 36.13 0.0 120.34 120.34 120.34 120.34 17.10 17.10 17.10 17.10 14FB 0800 565.72 510.00 825.00 845 0.0 0.0 0.0 0.0 0 675.0 5<>5.00 .00 675.00 620.00 450.00 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 675.00 525.00 0.0 0.0 0.0 0.0 525.00 450.00 "• ,00 225.00 0.0 0.0 0.0 0.0 450.00 390.00 28 92.00 0.0 0.0 0.6 99.00 65.00 50.00 30.00 0.0 0.0 0.0 0.0 0.0 0.0 0.0 8.10 — 1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 6.0 0.0 0.0 0.0 0.0 4.00 2.02 3.00 0.0 30.09 0.0 0.0 0.0 1600.001600.001600.001760.00 486.50 414.50 25.7.. 50 107.50 149 150 151 152 153 0.0 23.08 0.0 0.0 -Zfc£r 15.15 45.97 50.51 0.0 46.92 6.72 154 0.0 0.0 40.00 -16.67 0.0 0.0 0.0 0.0 •*&-»* 36.29 30.30 0.0 46.92 6.72 155 156 24 00 00 00 0.0 0.0 0.0 O.C 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1090.; 900. C 825.C 600.C 825.C 640. C O.C O.C 640.00 480.00 465.00 383.00 .00 640.00 480.00 465.00 .00 435.00 285.00 210.00 .0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 101.00 105.00 0.0 0.0 0.0 72.00 55.00 0.0 0.0 49.00 0.0 0.0 0.0 0.0 0.0 8.00 0.0 0.0 0.0 3.00 37.09 0.0 0.0 0.0 0.0 0.0 0.0 2.00 0.0 78.04 37.74 16.89 72.97 31.43 60.00 61.11 0.0 23.61 0.0 0.0 0.0 9.09 0.0 0 67.97 68.57 61.00 50.81 6.85 157 0.0 50.81 6.85 158 32.81 47.62 0.0 50.81 6.85 159 0.0 50.00 0.0 0. 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.02 0.0 0.0 0.0 "S'i* 304.'05 146:31 44loO •Hi 5.16 0.0 16.00 0.0 0.0 r.62 27.03 40.00 0.0 50.81 6.85 160 ------- APPENDIX E PRINTOUT OF DAILY SUMMARIES OF CONDITIONS AND RESULTS FOR PRIMARY-CARBON SEQUENCE 00 0\ , ,0*T6»UAO/T2 2 • WATER TR ATNENT PlCOT PLANT FILE DUMP 15JU ~~257o"0"?375o-~217oo~~25750' _fl.£___(U(L. (jro o.o 0.0 405.00 293.00 975.00 0.0 0.0 1383.001125.00 450.00 .050.00 525.JO 3^800 453.00 428.-00 3^5.00 300.00 0 .0 0 .00 450.00 405.00 0.130.aO_291*QO 0. 0 0.0 0. 1125.00 450.00 408.00 525.00 338.00 30O.QO 191.00 0.0 C.O 3.0 198.00 123.00 97.00 9*.00 .0 0.0 0. 228.00 0.0 0.0 232.00 104.00 J09.00 0.0 0.0 0.9 217.00 0.0 0. 2Z1.00 137.00 121.00 0.0 0.0 0.0 0.0 aro oa 0.0 0. 5 320.75 162.00 rfeft-fefr ^7^~I^M"I 3T8?; 25.56 49.79 0.0 5 26.04 19.38 37.93 30.1 0.0 0.0 0.3 22.22 35.19 0.0 0.00 32.19 61.99 54.75 144.00 0. ! ------- APPENDIX F PRINTOUT OF DAILY SUMMARIES OF CONDITIONS AND RESULTS FOR LIME TREATMENT oo T W.ANT - UME TREATMENT R CARBR PHA+F BASIN S.ft B-0 1 9 BASIN TUTOR CARBR 9.04 OiS ft*fi_ 2—8 0.0 12.25 o.q ^—8:8—8t8 WATER FLOW GPM 1 0.0 12.25 0.0 0.02^.?5 Q.n 0.0 12.25 0.0 0 29.00 Q.O COLOR FPC AT PH MPC AT PH 750.00 690.00 608.00 396.00 1125.00 600.00 743.00 410.00 1050.001275.00 713.00 428.00 QO *£JL QUO HICROMHOS TURBIDITY JTU SUU.ATJ.L.E 25.00 90.00 -fcg—Mt—M- 0.0 WQiO 0.0 .0 1193.001193.00 0.0 7632.00 7858.00 0.0 53.64 0. 11*53 p.0 *u_ i.i_ M LIME U CU7.6 REMOVD *BA TOD REMOVED tBA IME H57 PROD WAT MG/L SLUD JULIAN DISC FILE LOCN 0.0 0.0 ------- APPENDIX G PRINTOUT OF DAILY SUMMARIES OF CONDITIONS AND RESULTS FOR LIME-CARBON SEQUENCE RR01ZB - WATER TREATMENT PILOT PLANT - LIME DATE _ 8.23 5.0 8.i* 0.< Q ^.n , MRS OPER FLOW RATE GPM mftomtn: MPC oo oo CU TO COL FPC JiR£_ CU FM COL FPC TOC"FM-§ASJI.N£— - TO COL FM COL CA FM BASIN TQ/FROW CQLS ' TIC~FM~~BASIN~ TO/FM COLS BOD TO/FM COLS PH F.M BASIN.. . TO/FM COLS CONO TO/FM MMHOS Tl«B TO/FM JTU PflESS DROP .CQL.S. C.AQa2_DQS£_c!G/L_ VOLATILES MEUH ACETALD..__ ...Jpf-MMfiAbzi CUM LOG' LOG RATE /HR RATE TOT *C ^ ^f 1 LOG RA" 516.18 248.48 0.29 0.28 57.59 0.0 -fl:l*-W LOG RATE /HR RATE TOT #C —REM % SYS IN TOT REM * IN aE&l_!_I0._£uJ.. .-JA.QJ Tt«_.|ji- 522.02 253.20 0.26 0.21 46.95 0.0 .28,33 22,87 Si.19 0.0 .1.44 ,.11 59., 0.03 O.ov 11.58 0.0 8.57 26.6? 35.24 0." -tti1 -W-MY" pTCE'Lor 0.0 O.Q —8^—8rf 0.0 0.0 0.0 0.0 _._.^j_0 OjO 0.0 0.0 §:§ 8:8 -mt88—fcH -52577r26_ 0.15 0.34 49.14 0.0 ,. 18.92 _4U70 60.62 0.0 hnt.-ft-Ttf 0.28 O.?0 47.89 0.0 -&ff-W 538.4 0.19 51.64 -HtH-1 ^ 540.70 279.94 0.15 0.21 36.62 0.0 19.A4 26.9 46.43 0.0 542.03 283.6? 0.06 0.17 22.54 0.0 7.54 21.03 28 .54 21.03 .57 0.0 0.0 0.0 O."0~~* 0.0 0.0 0.0 0/0 0.0 0.0 0.0 O.D h-H-^i- 1?.83 0.0 12.70 19.84 32,5-4 0.0 —i|Ip__Z2t»ia 0.0 0.0 238.00 61.00'^ ------- APPENDIX H GENERALIZED PROCEDURE FOR DESIGNING PLANT FOR TREATING KRAFT PULP MILL EFFLUENT BY ADSORPTION IN GRANULAR ACTIVATED CARBON COLUMNS 1. Determine the flow and quality of the water to be treated. Normally, the plant is designed to treat the average of the daily maximum flow rates and to treat water with the average maximum color and TOC concentrations. Other properties of the effluent should also be deter- mined: BOD, conductivity, settleable solids, suspended solids, pH, temperature, etc. Determine what concentrations of color and TOC are to be tolerated in the treated water. The design of the carbon adsorption plant for a given flow rate is greatly affected by the prior treatment of the effluent water (clarification alone, lime treatment, or bio-oxidation) and by the permissible concentrations of color and TOC in the product water from the carbon treatment. 2. Select type of carbon. The type of carbon is determined from extensive laboratory isotherm tests and preferably from bench-scale column adsorption tests after a prelim- inary selection based on previous experience, price, service provided by manufacturer, resistance to abrasion, and other factors. The mesh size should probably be about 20 x 40 mesh, since this size minimizes the retention time required and yet does not cause appreciable pressure loss at flow velocities of 1 - 4 gpm/ft . 2. Contact time and flow velocity. Determine from pilot-plant or bench-scale column adsorption tests the contact time that is required at flow velocities of 1 - 4 gpm/ft of column area to meet the desired properties of the treated water. The value of retention time dictates the volume of the carbon needed for the plant and is the major determinant of capital costs of the plant. 4. Dosage of carbon per unit of water treated. The dosage is the pounds of regenerated carbon that must be added to the adsorbers (and loaded carbon that is removed) per 1000 gal of water treated to maintain the desired quality of product water. The dosage rate is a major determinant of operating costs since losses (make-up) 189 ------- of carbon and regeneration costs are directly related to the dosage. The dosage is best determined from bench-scale or pilot plant adsorption tests with at least three columns in series operated at the expected flow-rate (gpm/ft2) until it has been necessary to add 5 fresh columns while maintaining the desired product water concentration, or more gener- ally 2 more columns than the number of columns in the adsorption train. dosage rate in Ib/kgal = (Ib carbon per column) x (total fresh columns added) total kgal of water treated The rate can also be determined (with less reliability) from lab isotherms with at least 5 representative samples of the water to be treated. From the isotherms, determine average loading for the average feed concentra- tion. dosage rate = (feed concentration - product concentration) 0.65 (loading at feed concentration) where: dosage is g carbon per 1 water concentrations are in mg/1 loading is in mg/g carbon. 5. Volumetric flow rate and volume of carbon needed. Volumetric flow rate, vol. water 1 hr x vol. carbon ~ C.T. where C.T. = contact time, hr volume of carbon needed, ft = (ft3 water/hr) x C.T. = 0.125 gpm x C.T, weight of carbon, Ib = 25 lb/ft3 x vol. of carbon 6. Design of carbon columns. (See also EPA "Process Design Manual for Carbon Adsorption" October 1973(19)) Total cross sectional area of the columns, ft2 = (water flow in gpm)/(ve- locity in gpm/ft2). Diameter of columns /4(area of columns) V 3714 - Use multiple trains of columns in parallel flow to give height/diameter ratio between 1 and 3. 190 ------- Total length of carbon bed, ft = 8(gpm/ft2) x C.T. or length = (vol. carbon)/(total area) Length of bed in each column = total length/number of columns —(minimum of three columns) in series to provide adequate counter-current contacting. Freeboard in columns for backwash bed expansion = 50% of carbon bed. Extra columns for regeneration - a minimum of 2 columns is required to function as feed and receiving column for the regeneration furnace. Two columns can serve several trains if provided with proper piping arrangements. Extra columns for peak load - - provide 10 to 30% extra columns, depending on anticipated variability of flow and concentration. Consider separate system for pre-dampening fluctions by storage and/or spill control. Down-flow or up-flow of water in columns. With proper backwashing, down- flow appears preferable. Backwash - use velocity that expands the carbon bed 20 to 50%. This value is about 10 gpm/ft2 for 20 x 40 mesh Darco carbon. For backwash inlet distributor, graded gravel and sand is recommended, but other types of carbon support are available and probably satisfactory. The period of backwashing for each column should be about 20 minutes every 48 hours. Provide air at start of backwash to break up any clumps that might have formed. Weight of carbon per column, Ib = (area, ft2) x (bed height, ft) x (density, lb/ft3) Frequency of column change per train, days = (wt. of carbon per column, Ib) (dosage, Ib/kgal) (flow per train, kgal/day) Retention time of carbon in train, days = (wt. of carbon per train, Ib) (dosage, lb/kgal)(flow per train, kgal/day) 7. Total inventory of carbon. Total carbon = (no. of columns) x (volume of each) x (density of carbon)= weight of carbon needed. Warehousing will require additonal carbon; the amount being dependent on the purchase requirements. 191 ------- 8. Make-up carbon. The make-up carbon is equal to the loss of carbon during removal from columns and in regeneration. This loss runs 5 - 10% of the carbon dosage or weight of carbon that is regenerated. It is expected that the loss will probably be near 5% for large plants treating pulp mill effluents. 9. Carbon regeneration plant. a. Capacity, Ib/day = (carbon dosage, Ib/kgal) x (water flow, kgal/day) b. Type of furnace - a multiple-hearth furnace is generally used. There are three major manufacturers. c. Operating conditions - selected by consultants or through trials by furnace manufacturer. 192 ------- APPENDIX I RELATIONSHIPS USEFUL IN DESIGN OF DOWNFLOW CARBON ADSORPTION COLUMNS Superficial (empty column) velocity, ft/hr = 8(gpm/ft2) Total cross-sectional area of columns = gpm/(gpm/ft ) Bed Height, ft = (lineal velocity, ft/hr) (contact time, hr) = 8(gpm/ft2) (contact time, hr) Volumetric flow rate,by/hr = (vol water/hr)/(vol carbon) = 1 contact time,hr - 8(gpm/ft2)/(bed height) Bed height or length = 0.02 x flow x AC rate Where: Flow - water flow rate, gpm/ft2 AC = inlet - outlet concentrations in mg/1 rate = rate of adsorption in mg g x hr This relationship is generally applicable, but the rate must apply at the flow rate used and over the concentration range used. 193 ------- APPENDIX J ESTIMATE NO. 1. - MICRQLIME-CARBON TREATMENT OF 9.6 MGD OF EFFLUENT FROM A NEW 800 TON/DAY UNBLEACHED KRAFT PULP AND PAPER MILL 1AJOR CONDITIONS AND ASSUMPTIONS Flow Sheet - See Figure 30 Lime Treatment Microlime treatment is used Dosage = 520 mg/1 CaO (approximate) = ^ Rise rate = 700 Water quality: Feed Product % Reduction Color, CU 1000 300 70 TOG, mg/1 250 150 40 BOD, mg/1 250 225 10 pH 10.5 11.5 Conductivity, micromhos 1200 3600 Susp. solids, mg/1 200 250 Temperature, °F 100 100 Lime sludge dewatered to 70% solids and returned to lime kiln. Includes capital cost of new lime kiln for the proportion of load added by effluent treatment sludge. Desgin basis for lime treatment includes data from Interstate Paper Company (5,18) and Continental Can Company (14) on lime treatment of kraft mill effluents. Capital costs for lime treatment are derived from costs of the Continental Can Company lime treatment plant at Hodge, La. Carbon Adsorption Water quality, average composition: Reduction Feed Product by carbon overall Color, CU 300 100 67 90 TOC, mg/1 150 100 33 60 BO0* tag/1 225 150 33 40 pH H.5 8.0* Temperature, °F 100 100 Susp. solids, mg/1 250 15 *Product water is carbonated to pH 8 194 ------- LIME SLAKER MILL EFFLUENT KILN GAS VO UT SLUDGE FILTRATE SLUDGE -<- SOLIDS TO KILN FEED r-o-J SLUDGE FILTER SLUDGE SETTLING TO BACKWASH 1i CARBON CONTACTORS BACKWASH WATER RETURN r^ PH _ ADJUST TANK TREATED WATER TO MILL REUSE MAKE-UP WATER SETTLING TANKS FOR BACKWASH WATER Figure 30. Flow diagram for microlime-carbon treatment of unbleached kraft effluent for reuse ------- Designed for average flow of 9-6 mgd and maximum flow of 30% greater, or 12.5 mgd. Designed for average concentrations above, but with excess capacity for greater concentration of color and TOG. Down-flow, fixed-bed, pressure carbon contactors (columns). Air and water backwashing of columns. Feedwater to columns is not filtered. Product water is adjusted to pH 8 by addition of kiln gas. Carbon used: 20 x 40 mesh, equivalent to that used in pilot plant. Make-up (loss) of carbon per regeneration = 5%. Contact time, volume of carbon needed, and dosage rate are based on pilot plant results. Capital costs for carbon adsorption plant are based on cost data given in FWPCA Report TWRC-11 (17) prepared by M. W. Kellogg Company and in EPA Process Design Manual for Carbon Adsorption, October, 1973 (19). Cost of make-up carbon = $0.27/lb and $0.10/lb. Costs for Complete Plant Capital costs include equipment, installation, instrumentation, buildings, engineering, contractor overhead and profit. Time basis for cost = January, 1973 adjusted by Engineering News Record Construction Cost Index. Amortization = 16 years, straight line = 6.25% of total capital invest- ment per year. Repairs and maintenance = 3% of total capital investment per year. Taxes and Insurance = 2% of total capital investment per year. Plant overhead = 75% of direct labor cost. Steam = $0.80/1000 Ib, electricity = $0.01/kwh, gas = $0.60/106 Btu. DESIGN OF LIME TREATMENT UNIT Lime Dosage - to maintain 80 mg/1 dissolved Ca in water leaving lime treater. On basis of pilot plant treatment of unbleached kraft total mill effluent, average dosage is 520 mg/1 CaO, or 0.52 x 8.34 = 4.33 lb/1000 gal, or 4.33 x 9,600/2000 = 20.8 tons CaO. Lime Slaker and Feeder Mill lime is fed to a lime slaking tank to maintain a constant 10% solids slurry which is then fed through a controlled slurry feeder to the effluent line going to the lime treater. The soluble calcium concentration in the lime treated water is controlled by a sensor in the lime treater which sends a signal to the slurry which is then fed through a controlled slurry feeder to the effluent line going to the 196 ------- lime treater. The soluble calcium concentration in the lime treated water is controlled by a sensor in the lime treater which sends a signal to the slurry feeder to maintain the calcium concentration at 80* 10 mg/1 Ca. Lime Treater A conventional reactor-clarifler is used that has an enlarged center- well for the reaction of lime with the organic compounds of the effluent and for reaction and growth of CaCOo crystals. Volume of reaction zone to give 30 min. retention = 30 x 9.600.000 ,nn nnn 1 144Q - 200>000 gal. Top area of reaction zone that has enlarged top section is estimated to be 670 ft2, or 29 ft in diameter. Area of clarifier section at rise rate of 700 gpd/ft2 = 9.600.000 _ 9 -2- - ~ ft Total area of lime treater » 14,370 x 4/3.14 = 135 ft. Lime Balance for Lime-Carbon Treatment Conc'n. in water Ca. mg/1 CaO. mg/1 In feed water 12 17 Lime added to lime treater 372 520 Soluble in product water 80 112 In suspended solids from lime treater3 50 70 Loss in sludge handlingb 19 26 Net recovered in sludge to kilnc 286 400 Overall recovery of CaO = 100 x 400/520 = 77% a At 20% Ca in suspended solids, or 0.20 x 250 = 50 mg/1 Ca. This calcium is recovered as sludge from backwashing of carbon column which is added to the lime-treater sludge. b Estimated loss - 5% of added lime, or 0.05 x 520 - 26 mg/1 CaO c 520 + 17- 112 -25 = 400 mg/1 CaO 197 ------- Sludge Composition On the basis of pilot-plant results, the sludge will consist of these materials : Conc'n in water mg/1 Ib/kgal CaC03 from 003 in feed water 343 2.85 Ca-organics 278 2.32 Susp. solids in feed water3 200 Total sludge solids produced 821 6.84 a Mainly inorganic solids CaO in lime sludge solids to kiln = 400 mg/1 of water treated (see above) or 100 x 400/821 = 49% CaO Total sludge solids produced = 6.84 x 9600/2000 =32.8 tons/day CaO in sludge = 32.8 x 0.49 = 16.1 tons/day Volume of sludge produced at 15% solids from lime treater 6.84 x 9600 x — L_ x — Loc _ L_ = 48,200 gal/day=33.5 gpm 0.15 8.34 1.09 sp. gr. Sludge Tank for Settling - at 12-hour retention and 30% solids = 48.200 x J.5 = 12,000 gal. 2 30 Sludge Filters - on the basis of pilot plant experience, the sludge filters readily, presumably because microlime gives a high percentage of CaC03 in the sludge, as compared to minilime treatment. On the basis of the pilot-plant results, assume sludge from sludge tank will contain 25% solids and will filter at a rate of 100 Ib/hr ft2 on wire-drum filter. Area of filter = 6.84 Ib/kgal x 9.600 kgal/day = 27.4 ft2 24 hr/day x 100 Ib/hr ft2 (Possibly, the settled sludge could be added to the mill's lime mud filter). 198 ------- DESIGN OF CARBON ADSORPTION UNITS Contact Time Required Average pilot-plant results show that when removing color from 300 to 100 CU when the calcium content of the water was about 80 mg/1 Ca. the length of adsorption bed at a superficial flow rate of 1.4 gpra/ft* was 19 ft and the contact time was 1.7 hr. The data showed that a slightly shorter length of bed was required to reduce the TOG from 150 to 100 mg/1. Therefore, use a contact time of 1.7 hr plus 7% excess or 1.8 hr. (An additional 20% excess capacity is allowed later as spare columns.) Selections of Velocity, Pressure Drop, and Length of Carbon Bed. Pilot plant data showed that it is desirable to use a superficial velocity greater than 2 gpm/ft2 to provide increased turbulence and adsorption rates. However, as velocity is increased, the length required increased in direct relation to velocity (to maintain constant contact time) and the pressure drop is increased by both the increased length and greater velocity. The following tabulation shows the effect of flow rate on pressure drop, as determined from Fig. 3-4 of the EPA Design Manual for Carbon Adsorption (18) for the desired contact time of 1.8 hr (0.56 V/V hr) when using 20 x 40 mesh carbon. Flow rate, Length of bed Pressure gpm/ft2 required, ft drop, psi 1.4 20 1.1 2.8 40 4.2 4.2 60 9.7 5.6 80 18.1 To provide a fairly low pressure drop of 9.7 psi, a flow rate of 4.2 gpm/ft is selected, which requires a total length of bed of carbon of 60 ft. The effect of velocity on adsorption rates was not studied in the pilot plant for the microlime-carbon sequence, while operations were carried out at 1.4 gpm/ft2. Based on increased adsorption rates, or decreased retention times, observed for increased flow rate in the primary-carbon pilot plant work, it is possible that the design flow rate of 4.2 gpm/ ft2 chosen above in conjunction with contact time requirements based on 1.4 gpm/ft2 in the pilot plant, could represent a sizeable effective excess contact time. 199 ------- Dimensions of Adsorption Columns Area = 6670 gpm/4.2 gpm/ft2 = 1590 ft2 total. No. of parallel trains - use 3 to keep diameter of columns at reason- able size. 1590 x 4 Diameter of each column - = 26 ft. 3 x 3.14 No. of columns in series per train - use 3 to provide adequate counter- current contacting and to keep the ratio of height to diameter about 1. Height of bed in each adsorber = 6£ = 20 ft. 3 Total height of each column, allowing 50% free board for bed expansion during backwashing = 20 x 1.5 = 30 ft. Arrangement of Columns feed- rOOO V Active Columns Total columns Active columns Spare columns Ready to have carbon discharged Ready to receive regenerated carbon Total columns with carbon Spare columns for surge loads inactive - being refilled with carbon inactive - being discharged to regeneration = 13 = 9 = 2 » 1 = 1 = 12 Volume of carbon per column = 3.14 x 262 x 20 = 10,600 ft3. 4 Weight of carbon per column = 10,600 x 25 lb/ft3 - 265,000 Ib. Total weight of carbon in system = 12 x 265,000 « 3,180,000 Ib. Excess capacity available for surge loads = 2 columns out of 9 active columns = 22% excess. 200 ------- Total excess capacity for surge loads 22% excess columns 77, extra contact time (above) 4% lost capacity during backwash (below) 25% net excess capacity. Column Features Columns are field erected steel columns, tested to hold 70 psi, dished heads, with 8-mil interior coating of epoxy-tar, or equivalent, with gravel support for carbon, and with piping connections for backwashing for changing the flow sequence through the columns, and for movement of carbon as slurry to and from regeneration. Backwashing of Columns On the basis of pilot plant experience and results from commercial carbon adsorption for treating municipal effluents, the 3 lead columns are backwashed once each day and the remaining 6 columns are backwashed every two days. Total number of columns backwashed per day = 6. During each backwash, each column is agitated with a short backwash of compress- ed air and then with a flow of water at 10 gpm/ft2 (5300 gpm) for 15 min- utes. Volume of water used per backwashing = 15 x 5300 = 80,000 gal. Void volume of each column at 50% voids in carbon = 10,600 ft3 carbon x 0.5 + 5300 ft3 freeboard = 10,600 ft3 = 10,600 x 7.5 gal/ft3 = 80,000 gal. So backwashing displaces an equal volume of water from the column. Backwash Schedule A schedule is arranged to meet above backwashing requirements by back- washing the No. 3 and No. 1 columns on one day and No. 2 and No. 1 columns on alternate days. One quantity of 80,000 gal of carbon-treat- ed water is used each day; first for backwashing the No. 3 columns or No. 2 columns, then for No. 1 columns, and then pumped to the lime treater. The water from backwashing each column is discharged to a second 80,000-gal tank where a coagulant is added and the suspended solids removed by settling and the sludge is pumped to the lime sludge tank. The clarified water is used for backwashing the next column and is then discharged to the other settling tank. Thus two settling tanks are needed to permit partial reuse of the backwash water. Columns backwashed per day = 6, or 2 each shift. Product water used = 80,000 gal/day - 100 x 80.000 - 0.8% of daily pro- 9,600,000 duction. 201 ------- Total off-time of columns for backwashing at 20 minutes per set of columns x 2 = 40 minutes/day = 3% of time. Total loss of production for backwashing = 3.0 + 0.8 = 3.8%. Carbon Dosage Rate The pilot plant runs for microlime-carbon required 2.5 Ib/kgal to reduce color 230 to 70 CU and TOG 170 to 110 mg/1. Since this design calls for about the same removals, use a dosage of 2.5 Ib/kgal. Carbon added = 2.5 x 9,600 kgal/day = 24,000 Ib/day * 1000 Ib/hr. Carbon Make-Up Rate Most commercial carbon adsorption units experience a loss of about 5% per generation, so use this. 0.05 x 24,000 = 1200 Ib/day Regeneration of Carbon Average rate » 24,000 Ib/day = 1000 Ib/hr. Hours of operation per day = 24 Furnace - use2multiple-hearth regeneration furnace with capacity of 100 Ib/day/ft , which is based on full-scale plants for treating municipal wastewater. Total hearth area with allowance for 20% down time (including 5% make up) = 24,000 x 1.2 - 290 ft2 100 No. of hearths - use 6 Area per hearth = 290 ft2/6 =48.3 ft2 Diam. of furnace = v^48.3 x 4/3.14 = 8 ft. CAPITAL COST OF LIME TREATMENT UNIT The most reliable cost data on lime treatment is that for the Continental Can Company plant (35) which uses minilime treatment (1000 mg/1 CaO) for 9 mgd of total mill effluent from a pulp and paper mill producing 750 tons/day- Their cost include all associated costs including capital cost for additional lime kiln capacity, engineering, contractor's over- head and profit, lime feeder and slaker, sludge dewatering, carbonation, and instrumentation. Total cost of Continental Can lime treatment plant $1,762,000 Adjusted to microlime system (lower lime rate, no carbonation, etc.) 1,182,000 202 ------- Adjusted for lower rise rate in lime clarifier (use 700 rather than 980 gpd/ft2) Adjusted from 9-0 to 9.6 mgd Adjusted from 1972 to Jan. 1973 costs by ENR Construction Index, multiply by 1.09 Add 87=, of cost of lime kiln for sludge lime added to mill lime kiln, adjusted to Jan. 1973 costs Total cost for microlime treatment system $1,261,000 1,328,000 1,448,000 232,000 1,680,000 CAPITAL COSTS FOR CARBON ADSORPTION UNIT The cost for the carbon adsorption unit is based on the study by M. W. Kellogg Co. for FWPCA in 1969 (17). In the Kellogg-FWPCA report the total installed plant cost was broken down into categories. In preparing this cost estimate, each of Kellogg costs were adjusted for the conditions used in this estimate. For example, the cost of carbon adsorbers (44% of the total Kellogg cost) was multiplied by 2.2, which is the ratio of volume of carbon in this estimate to that in the Kellogg estimate. The table below lists the categories of costs for both estimates and the multiplying factor used in adjusting the Kellogg costs to this estimate. The multiplying factors include, where applicable, a ratio of sizes to selected exponents. Adjusted Costs for Carbon Adsorption Unit 1969 costs, not including first charge of carbon Kellogg- FWPCA costs,$ Foundations, etc. $ 42,000 Carbon adsorbers 670,750 Tanks, carbon transport, 103,880 carbon storage Structural steel 20,900 Pumps compressors 65,070 Office buildings 5,400 Regeneration system 128,240 Valves, piping 268,990 Electrical, instrumentation 188,260 Paint, utilities 9,660 Conveying equipment 9,360 $1,512,510 Adjust- ment factor 1.5 2.2 1.5 Adjusted to this estimate, $ 63,000 1,476,000 155,800 % of total 55 '.7 5.9 31,400 78,100 5,400 192,400 403,500 225,900 12,600 14,000 1.2 2.9 0.2 7.2 15.2 8.5 0.5 0.5 $2,658,100 ( 100.2) Adjust total cost above to Jan., 1973 by ENR Equipment Index (1812/1216) = $3,961,000 203 ------- Cost of First Charge of Carbon First Charge = 264,000 Ib/column x 12 columns - 3,170,000 Ib Warehouse carbon for 30 days of requirements « 1200 Ib/day total losses x 30 days = 36,000 Ib. Total inventory of carbon = 3,206,000 Ib Cost at $0.27/lb = 0.27 x 3,206,000 = $865,600 Cost at $0.10/lb - 0.10 x 3,206,000 = $321,000 CAPITAL COST OF pH ADJUSTMENT AFTER LIME TREATMENT Amount of C02 needed to decrease pH from 11.5 to 8.0, as determined from acid titrations of samples from pilot plant runs = 0.018 eq/1 or 3.3 Ib C02/kgal. Kiln gas needed at 8.4 ft3/lb and 25 volume-% C02 in kiln gas = 3.3 x 8.4/0.25 = 111 ft3 kiln gas per kgal. Kiln gas needed per day = 111 x 9,600 = 1,080,000 ft3 (STP)/day. Kiln gas at 907o utilization of C02 = 1,080,000/0.9 = 1,200,000 SCFD = 833 SCEM. Retention time for neutralization - allow 10 minutes on basis of results at other plants. Volume of pH adjust tank = 10 x 9,600,000/1440 = 67,000 gal = 8,900 ft3. Diameter of tank =^8900 x 4/3.14 = 28 ft. Agitation on basis of carbonators at Continental Can plant = 3 at 40 hp ea. Cost of pH adjust tank, including agitators, gas piping, installation,etc., (29,30,24) - $169,000. CAPITAL COST OF BACKWASH CLARIFICATION AND SURGE TANKS Need two 80,000-gal tanks with stirrers and conical bottoms to clarify the washwater after each column is backwashed. Cost including sludge pumps, installation, etc. = $70,000 OPERATING COSTS FOR LIME TREATMENT Operating costs for recalcining lime sludge are included in the cost of recalcined lime. <- Make-up lime at mill purchase price =4.7 tons/day x $24.23/ton = $114/day Calcination of lime in lime sludge at mill operating cost = 16.1 tons/day x $12.40/ton=$200/day Total cost of lime = $114 + 200 = $3l4/day, or $319/9600 kgal = $0.0328/kgal 204 ------- Labor - use 3 man-hours per shift at $5.85/hr, including fringe benefits= 3x3 shifts x $5.85/9600 kgal - $0.0055/kgal Electricity - average costs for lime treatment (5,35) = $0.0080/kgal OPERATING COSTS FOR CARBON TREATMENT Labor - (maintenance labor is included in repair and maintenance factor based on total plant investment) use 1 operator per shift and 1 lab tachnician per day = 4 man-shifts/day x 8 hr = 32 x $5.85/9600 kgal = $0.0195/kgal Fuel at 5,700 Btu/lb carbon regenerated (19) and gas at $0.60/mil Btu = 5,600 x ID"6 x $0.60 = $0.0034/lb carbon or 2.5 Ib/kgal x $0.0034= $0.0085/kgal Steam - Kellogg report used 1 Ib steam/lb carbon regenerated. cost @ $0.80/1000 Ib = 2.5 Ib/kgal x $0.80/1000 = $0.0020/kgal Electricity - experience at large plants for treating municipal effluent averages $0.012/kgal. Use this. 205 ------- TOTAL CAPITAL AND OPERATING COSTS FOR ESTIMATE NO. 1 Costs At carbon cost of $0.27/lb. At carbon cost of $0.10/lb. Capital Costs. $ Million Lime treatment Carbon adsorption Backwash tank pH adjust tanks Carbon inventory Total cost of investment, TCI 1.680 3.961 0.070 0.169 0.866 6.746 1.680 3.961 0.070 0.169 0.321 6.201 Operating Costs. $/kgal Amortization @ 0.0625 TCI/yr Repairs and maint. @0.03 TCI/yr Taxes and insurance (§0.02 TCI/yr Labor Plant overhead at 75% of labor Steam Gfls Electricity Carbon make-up Total operating costs, $/kgal 0.1202 0.0577 0.0387 0.0255 0.0191 0.0020 0.0085 0.0200 0.0328 Q.Q338 0.3583 Operating costs/ton pulp (@ 12 kgal/ton) $4.30 0.1105 0.0530 0.0355 0.0255 0.0191 0.0020 0.0085 0.0200 0.0328 0.0125 0.3194 $3.83 206 ------- NJ O English Unit APPENDIX K ENGLISH TO METRIC UNITS CONVERSION TABLE Abbreviation Conversion Abbreviation British Thermal Unit Cubic foot Cubic foot Degree Fahrenheit Foot Gallon (U.S.) Gallons per day per square foot Gallons per minute Gallons per minute per square foot Gallons per ton (of production) Horsepower Inch Million gallons per day Pound (mass) Pounds per cubic foot Pounds per square inch Pounds per thousand gallons Ton, short BTU cu ft cu ft °F ft gal gpd/sq ft gpm gpm/sq ft gal /ton hp in mgd Ib Ib/cu ft psi lb/1000 gal ton X X X 0. X X X X X X X X .X X X X X X 0.2520 = 0.02832 = 28.32 555(°F-32) = 0.3048 * = 3.785 0.04074 = 0.06308 = 40.7 4.172 0.7457 25.4 * 3,785 = 0.4536 16.02 703.1 119.84 0.9072 kg cal m3 1 C° m 1 m3/m2 day I/sec l/min/m2 1/kkg kw mm m3/day kg kg/m3 kgf/m2 g/m3 kkg Metric Unit Kilogram - calories Cubic meters Liters Degree Centigrade Meters Liters Cubic meter per square meter day Liters per second Liters per minute per square meter Liters per metric ton (1000 kg) Kilowatts Millimeters Cubic meters per day Kilograms Kilograms per cubic meter Kilograms force per square meter Grams per cubic meter Metric ton (1000 kilograms) * Indicates exact conversion factor ------- TECHNICAL REPORT DATA (f lease read Instructions on the reverse before completing) .REPORT NO. EPA-660/2-75-004 2. 3. RECIPIENT'S ACCESSION>NO. 4. TITLE AND SUBTITLE ACTIVATED CARBON TREATMENT OF UNBLEACHED KRAFT EFFLUENT FOR REUSE 5. REPORT DATE 6. PERFORMING ORGANIZATION CODE 7.AUT E. W. Lang, W. G. Timpe, R. L. Miller 8. PERFORMING ORGANIZATION REPORT NO EPA-660/2-75-004 9. PERFORMING ORGANIZATION NAME AND ADDRESS St. Regis Paper Company Research and Development Center Pensacola, Florida 10. PROGRAM ELEMENT NO. 1BB037 11. CONTRACT/GRANT NO. 12040 EJU 12. SPONSORING AGENCY NAME AND ADDRESS National Environmental Research Center Pacific Northwest Environmental Research Laboratory Corvallis,Oregon 97330 13. TYPE OF REPORT AND PERIOD COVERED Part I - Final Report 14. SPONSORING AGENCY CODE 15. SUPPLEMENTARY NOTES is.ABSTRACT A four-year pilot plant program was carried out to investigate the techni- cal and economic feasibility of treating unbleached kraft pulp and paper mill effluent for reuse. Preliminary laboratory studies and cost estimates indicated that the following treatment sequences should be investigated in the pilot plant: 1) primary clarification, carbon adsorption; 2) lime treatment, carbon adsorption; 3) primary clarification, bio-oxidation, carbon adsorption. Water of reusable quality can be provided from unbleached kraft effluent by several combinations of treatment utilizing activated carbon. Unbleached pulping effluents typically contain about 1000 color units, 250 mg/1 TOC, and 250 mg/1 BOD. Reusable water quality as defined in this study is 100 color units and 100 mg/1 TOC. The most economical treatment is the microlime-carbon process that utilizes low dosages of lime and clarification followed by carbon adsorption in down-flow granular carbon beds. Capital cost for treatment by this process of 9.6 mgd of unbleached kraft effluent from an 800-ton-per-day mill was estimated to be approximately $6.7 millions. Operating costs, inclusive of capital depreciation, were estimated to be $0.30 per 1000 gal and $3.58 per pulp-ton, including credit for the reused water. Carbon adsorption in continuous counter-current stirred contactors was found to have promise of lower operating cost and substantially lower capital costs as compared to adsorption in fixed beds. 17. KEY WORDS AND DOCUMENT ANALYSIS a. DESCRIPTORS b.lDENTIFIERS/OPEN ENDED TERMS COSATI Field/Group Industrial Waste Treatment, Activated Carbon Treatment, Chemical Removal, Bio- chemical Oxygen Demand, Industrial Wastes, Pulp Mills, Cost Estimates, Pilot Plants, Effluents, Calcium Carbonate Pulp Mill Effluent, Color Removal, Tertiary Treatment, Water Reuse, Unbleached Kraft Waste- water Treatment, Micro- lime-Carbon Process, Wastewater Treatment, Tnhal Organic 05D Ores iS (TTtli 8. DISTRIBUTION STATEMENT Release Unlimited 19. SECURITY CLASS (ThisReport/ Unclassified 21. NO. OF PAGES 207 20. SECURITY CLASS (This page) Unclassified 22. PRICE Form 2220-1 (9-73) U.S. GOVERNMENT PRINTING OFFICE: I975-698-576 I\S2 REGION 10 ------- |