&EPA
            United States
            Environmental Protection
            Agency
            Municipal Environmental Research EPA 600 2-80-1 30a
            Laboratory         August 1980
            Cincinnati OH 45268
            Research and Development
Removing Potential
Organic Carcinogens
and Precursors from
Drinking Water
            Volume I
            and
            Appendix A

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was  consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.  Environmental Health  Effects Research
      2.  Environmental Protection Technology
      3.  Ecological Research
      4.  Environmental Monitoring
      5.  Socioeconomic Environmental Studies
      6.  Scientific and Technical Assessment Reports (STAR)
      7.  Interagency Energy-Environment Research and Development
      8.  "Special" Reports
      9.  Miscellaneous Reports

This report has  been assigned  to the  ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental 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.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia  22161.

-------
                                   EPA-600/2-80-130a
                                   August 1980
REMOVING POTENTIAL ORGANIC CARCINOGENS AND
      PRECURSORS FROM DRINKING WATER

          Volume I and Appendix A
                    by

   Paul R. Wood - Principal Investigator
             Daniel F. Jackson
  Drinking Water Quality Research Center
     Florida International University
           Miami, Florida  33199

             James A. Gervers
             Doris H. Waddell
   Miami-Dade Water and Sewer Authority

               Louis Kaplan
  Dade County Department of Public Health
           Miami, Florida  33125
           Grant No. R804521-01
              Project Officer

               Jack DeMarco
     Drinking Water Research Division
Municipal Environmental Research Laboratory
          Cincinnati, Ohio  45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
    OFFICE OF RESEARCH AND DEVELOPMENT
   U.S. ENVIRONMENTAL PROTECTION AGENCY
          CINCINNATI, OHIO  45268

-------
                           DISCLAIMER

     This report has been reviewed by the Municipal Environmental
Research Laboratory, U.S. Environmental Protection Agency,  and
approved for publication.  Approval does not signify that the
contents necessarily reflect the views and policies of the  U.S.
Environmental Protection Agency, nor does mention of trade  names
or commercial products constitute endorsement or recommendation
for use.
                               ii

-------
                           FOREWORD


     The U.S. Environmental Protection Agency was created be-
cause of increasing public and government concern about the
dangers of pollution to the health and welfare of the American
people.  Noxious air, foul water, and spoiled land are tragic
testimonies to the deterioration of our natural environment.
The complexity of that environment and the interplay of its
components require a concentrated and integrated attack on the
problem.

     Research and development is that necessary first step in
problem solution; it involves defining the problem, measuring
its impact, and searching for solutions.  The Municipal Envir-
onmental Research Laboratory develops new and improved tech-
nology and systems to prevent, treat, and manage wastewater and
solid and hazardous waste pollutant discharges from municipal
and community sources, to preserve and treat public drinking
water supplies, and to minimize the adverse economic, social,
health, and aesthetic effects of pollution.  This publication
is one of the products of that research and provides a most
vital communications,link between the researcher and the user
community.

     To protect the consumer of public drinking water, this
study was undertaken to develop feasible and economical metho-
dology for reducing the amount of specific organic contaminants
in drinking water.
                               Francis T. Mayo, Director
                               Municipal Environmental Research
                               Laboratory
                              111

-------
                             ABSTRACT

      The principle objective of the two-year Research Project
 was to devise feasible and economical methodology for removing
 existing organic contaminants from and preventing development of
 potential carcinogens in the public water supplies in Bade
 County, Florida.  Specifically, development of methodology to
 reduce the amount of four trihalomethanes (chloroform, bromo-
 dichloromethane, chlorodibromomethane, and bromoform) present in
 drinking water  was  the prime objective.

      A four-phase study was designed to evaluate the efficiency
 of three adsorbents in removing 19 individual halogenated
 organics and trihalomethane precursors.  These adsorbents were
 XE-34Q—a carbonized polymeric macroreticular resin; IRA-904—
 a strong base cationic resin designed to remove large molecular
 weight substances such as precursors from water; and granular
 activated carbon (GAC).  Adsorbent columns were placed at
 various stages  in the water processing system; i.e., the raw
 water stage, the lime softened stage at the up-flow Hydrotreator
 effluent and the finished water stage.

      Four GAC Filtrasorb 400 columns, each 0.76 meters (2.5 feet)
 deep, arranged  in series on the finished water line were most
 effective in reducing the level of the trihalomethanes present
 in the finished water that would be consumed by the public.

      The Polanyi-Manes Theory of adsorption was applied and
 found helpful in interpreting results.  Preliminary studies were
 made of the bacterial profile of the Preston Water Treatment
 Plant, raw and  finished water, and effluent from four GAC
 columns.  Distribution system samples were also analyzed.

      This report was submitted in fulfillment of Grant No.
 R804521-01 by Dade County Department of Public Health under the
 Sponsorship of  the U.S. Environmental Protection Agency.  This
 report covers a period from June 1976 to June 1980, and work was
„ completed as of May 1980.
                                iv

-------
                            CONTENTS



Disclaimer	    ii



Foreword	   iii



Abstract	    iv



Figures	viii



Tables	xviii



List of Abbreviations	    xx



Acknowledgments 	   xxi



     Section I.    Introduction 	     1



     Section II.   Conclusions	     3



     Section III.  Recommendations	     9



     Section IV.   Plant and equipment description	    11



        Preston Plant water source	    11



        Preston Plant site	    11



        Bench scale adsorption test unit	    13



     Section V.    Methods and procedures	    19



        Operation of bench scale adsorption unit	    19



        GC analytical method	    19



        TOC analysis.	    21



        THM, Terminal THM and THM FP	    21



        Data analysis	    24



     Section VI.  Experimental Plan	    26



        EDI	    26





                                v

-------
   ED1R	    28

   ED2	    28

   EDS	    28

   ED4	    28

Section VII.  Results and discussion 	    32

   Full slcale plant studies	    32

        Specific halogenated organics	    32
             Raw water source	
             Hydrotreator effluent source	    36
             Finished water source 	    38

        TOC and THM FP organics	    40
             Raw water source	    40
             H.T. water source	    40
             Finished water source 	    42

        Other parameters	    48
             Rainfall and chlorides	    48
             Rainfall and TOC	    48
             pH	-	    48
             Turbidity	    49
             Color	    49

   Bench scale studies	    50

        Specific halogenated organics	    50
             Raw water source	    50
             H.T. water source	    76

        THM	    76
             Finished water source 	   105
             Adsorption by XE-340	   187
             Adsorptive capacity and competitive
               adsorption	   188

        TOC and THM FP organics	   196
             Raw water source	   208
             H.T. water source	   220
             Finished water source 	   220

        Other parameters	   246
             Chlorine	   246
             Turbidity	   253
          ^  Color	   253
             pH	   253
                          vx

-------
             Comparison of laboratory and distribution
                  system aging	253

             Total THM growth in adsorbent
                  colunn effluents	256

             Bed life criteria in deep GAG columns	262

             Relationship of TOC and THM FP data	268

             Leaching study on XE-340 resin column	270

             Biological activated carbon (BAG)	281

             Polanyi-Manes Adsorption Theory	283
                  Theory development	283
                  Theory application	290

             GC/MS HOC confirmation data	309

References	311

Appendix A.  Frances Parsons

    Part I.  Microbial Flora of Granulated Activated
             Carbon Columns Used in Water Treatment	314

    Part II. Chlorination of Granulated Activated
             Carbon  (GAG) Column Effluent to Control
             Bacteria	349
                               VI1

-------
                             FIGURES

Number                                                      Page

  1      Flow diagram of John E. Preston Water
              Treatment Plant	   -1-2

  2      Bench Scale Column Adsorption Unit	   16

  3      Plumbing for adsorption column	   17

  4      Detailed view of column fittings	   18

  5      Backwash system for columns	   20

  6      Typical chromatogram of halogenated organics	   22

  7      Flow diagram of Bench Scale Adsorption Unit for
              GAG and XE-340 study in EDI	   27

  8      Flow diagram of Bench Scale Adsorption Unit for
              leaching study in ED2	   29

  9      Flow diagram of Bench Scale Adsorption Unit for GAC
              and IRA-904 resin study in EDS	   30

 10      Flow diagram of Bench Scale Adsorption Unit for deep
              bed study in ED4	   31

 11^13   TOG in raw, Hydrotreater and finished water	 44-46

 14      THM FP in raw water and removal by lime softening..   47

 15-18   cis 1,2-Dichloroethene in raw water and removal
              by 0.76 meter  (2.5 feet)  of  GAC and 0.76
              meters (2.5 feet)  of XE-340	 52-55

 19       Vinyl chloride in raw water and removal by 0.76
              meter  (2.5 feet)  GAC and 0.76 meters
              (2.5  feet)  XE--340	   57

 20       Vinyl chloride in raw water and removal by 0.76
              and 1.52  meters (2.5 and 5 feet)  of IRA-904
              resin	   58
                              Vlll

-------
Number
Page
21      trans 1,2-Dichloroethene in raw water  and  removal
             by 0.76 meter   (2.5 feet) of GAC	    60

22      trans 1,2-Dichloroethene in raw water  and  removal
             by 0.76 meter   (2.5 feet) of XE-340	    61

23      1,1-Dichloroethane in raw water and removal by
             0.76 meter   (2.5 feet) of GAC and 0.76 meters
             (2.5 feet) of XE-340	    63

24      1,1,1-Trichloroethane, 1,2-dichloroethane  and
             carbon tetrachloride in raw water and removal
             by 0.76 meter   (2.5 feet) of GAC	    66

25      1,1,1-Trichloroethane, 1,2-dichloroethane  and
             carbon tetrachloride in raw water and removal
             by 0.76 meter   (2.5 feet) of XE-340	    67

26      Trichloroethylene in raw water and removal by
             0.76 meter   (2.5 feet) of GAC	    69

27      Trichloroethylene in raw water and removal by
             0.76 meter   (2.5 feet) of XE-340	    70

28      Tetrachloroethylene in raw water and removal by
             0.76 meter   (2.5 feet) of GAC and 0.76
             meter   (2.5 feet) of XE-340	    71

29      Chlorobenzene in raw water and removal by  0.76
             meter   (2.5 feet) of GAC and 0.76 meter
             (2.5 feet) of XE-340	    73

30      p-Chlorotoluene in raw water and removal by 0.76
             meter   (2.5 feet) of GAC and 0.76 meter
             (2.5 feet) of XE-340	    74

31      o, m and p-Dichlorobenzene in raw water and removal
             by 0.76  (2.5 feet) of GAC and 0.76 meter
             (2.5 feet) of XE-340	    75

32-33   .Chloroform in H.T. water and removal by 0.76
             meter   (2.5 feet) XE-340	79-80

34-35   Bromodichloromethane in H.T. water and removal
             by 0.76 meter   (2.5 feet) XE-340	 81-82

36-37   Chlorodibromomethane in H.T. water and removal*
             by 0.76 meter   (2.5 feet) XE-340	 83-84

38-39   Bromoform in H.T. water and removal by 0.76
             meter   (2.5 feet) XE-340	 85-86
                               ix

-------
Number

 40-41   cis 1,2-Dichloroethene in H.T. water and
              removal by 0.76 meter  (2.5 feet)  XE-340 ..... 88-89

 42      cis 1,2-Dichloroethene in H.T. water and
              removal by 0.76 meter  (2.5 feet)  IRA-904
                                                               on
              resin ..........................................  au
 43      Vinyl chloride in H.T. water and removal by
              0.76 meter  (2. 5 feet)  of XE-340 ...............  92

 44      trans 1,2-Dichloroethene in H.T. water and
              removal by 0.76 meter  (2.5 feet)  of XE-340 ----  93

 45      1,1-Dichloroe thane in H.T. water and removal by
              0.76 meter  (2. 5 feet)  XE-340 ..................  95

 46      1,1,1-Trichloroethane, 1,2-dichloroethane and
              carbon tetrachloride in H.T. water and
              removal by 0.76 meter  (2.5 feet)  XE-340 ......   97

 47      Trichloroethylene in H.T. water and removal by
              0.76 meter  (2. 5 feet)  of XE-340 ...............  99

 48      Tetrachloroethylene in H.T.  water and removal
              by 0.76 meter  (2.5 feet) of XE-340 ............ IOC

 49      Chlorobenzene in H.T. water and removal by 0.76
              meter  (2.5 feet) of XE-340 .................... 101

 50      p-Chlorotoluene in H.T.  water and removal by
              0.76 meter  (2.5 feet)  of XE-340 ............... 102

 51      o, m and p-Dichlorobenzene in H.T. water and
              removal by 0.76 meter  (2.5 feet)  of XE-340.... 104

 52-54   Chloroform in finished water and removal by
              0.76 meter  (2.5 feet)  of XE-340 ........... 107-109

 55      Chloroform in finished water and finished water thru
              0.76 meter  (2.5 feet)  of IRA-904  resin and
              0.76 meter  (2.5 feet)  of GAC .................. 109

 56      Chloroform in finished water and in the effluent from
              0.76, 1.52, 2.29 and 3.05 meters (2.5, 5, 7.5
              and 10 feet)  of GAC ............................ 11C

 57-59   cis 1,2-Dichloroethene in finished water and
              removal by 0.76 meter  (2.5 feet)  of XE-340. 115-117
                               x

-------

60     cis 1,2-Dichloroethene in finished water and
            removal by 0.76 meter   (2.5 feet) of GAG
            and 0.76  (2.5 feet) of  IRA-904 resin	  118

61     cis 1,2-Dichloroethene in finished water and
            finished water thru 0.76, 1.52, 2.29 and
            3.05 meters  (2.5, 5, 7.5 and 10 feet)  of GAG...  119

62-64  Bromodichloromethane in finished water and
            removal by 0.76 meter   (2.5 feet) of XE-340.121-123

65     Bromodichloromethane in finished water and
            removal by 0.76 meter   (2.5 feet) of GAG and
            0.76 meter   (2.5 feet)  of IRA-904 resin	  124

66     Bromodichloromethane in finished water and removal
            by 0.76,  1.52, 2.29 and 3.05 meters (2.5, 5,
            7.5 and 10 feet) of GAG	  125

67-69  Chlorodibromomethane in finished water and removal
            by 0.76 meter   (2.5 feet) of XE-340	128-130

70     Chlorodibromomethane in finished water and removal
            by 0.76 meter   (2.5 feet) of IRA-904 resin
            and 0.76 meter  (2.5 feet) of GAG	   131

71     Chlorodibromomethane in finished water and removal by
            0.76, 1.52,  2.29 and 3.05 meters (2.5, 5,
            7.5 and 10 feet) of GAG	   132

72-74  Bromoform in finished water  and removal by 0.76
            meter  (2.5  feet)  of XE-340	134-136

75     Bromoform in finished water  and removal by 0.76
            meter  (2.5  feet)  of GAC and 0.76 meters
            (2.5 feet) of IRA-904 resin	   137

76     Bromoform in finished water  and removal by 0.76,
            1.52, 2.29 and 3.05 meters (2.5, 5, 7.5 and
            10 feet)  of  GAC	   138
                                             ;
77-78  Vinyl chloride in finished water and removal by
            0.76 meter   (2.'5 feet)  of XE-340	140-141

79     Vinyl chloride in finished water and removal by
            0.76 meter   (2.5 feet)  of IRA-904 resin and
            0.76 meter   (2.5 feet)  of GAC	   142

80     Vinyl chloride in finished water and removal by
            0.76 and 1.52 meters (2.5 and 5 feet)  of GAC..   143

                              xi

-------
Number                                                     Page

  81     Vinyl chloride in finished water and removal by
              2.29 and 3.05 meters (7.5 and 10 feet) of
              GAG	144

  82-83  trans 1,2^-Dichloroethene in finished water and
              removal by 0.76 meter  (2.5 feet)  of XE-340.147-148

  84     trans 1,2-Dichloroethene in finished water and
              removal by 0.76 meter  (2.5 feet)  of GAG
              and 0.76 (2.5 feet) of IRA-904 resin	149

  85     trans 1,2-Dichloroethene in finished water and
              removal by 0.76, 1.52, 2.29 and 3.05 meters
              (2.5, 5, 7.5 and 10 feet) of GAG	150

  86-87  1,1-Dichloroethane in finished water and removal
              by 0.76 meter  (2.5 feet) of XE-340	153-154

  88     1,1-Dichloroethane in finished water and removal
              by 0.76 meter  (2.5 feet) of GAG and 0.76
              meter   (2.5 feet)  IRA-904 resin	155

  89     1,1-Dichloroethane in finished water and removal
              by 0.76 and 1.52 meters  (2.5 and 5 feet) of
              GAG	 156

  90     1,1-Dichloroethane in finished water and removal
              by 2.29 and 3.05 meters  (7.5 and 10 feet) GAG.. 157

  91-92  1,1,1-Trichloroethane,  1,2-dichloroethane and carbon
              tetrachloride in finished water and removal
              by 0.76 meter  (2.5 feet) of XE-340	159-160

  93     1,1,1-Trichloroethane,  1,2-dichloroethane and
              carbon tetrachloride in finished water and
              removal by  0.76 meter  (2.5 feet)  of GAG and
              0.76  meter   (2.5 feet)  of IRA-904 resin	161

  94     1,1,1-Trichloroethane,  1-2-dichloroethane and
              carbon tetrachloride in finished water and
              removal by  0.76,  1.52, 2.29 and 3.05 meters
              (2.5,  5, 7.5 an.d 10 feet) of GAG	  162

  95-96  Trichloroethylene in finished water and removal
              by 0.76 meter  (2.5 feet) of XE-340	165-166

  97     Trichloroethylene in finished water and removal
              by 0.76 meter  (2.5 feet) of IRA-904 resin....  167
                              XI1

-------
Number                                                      Page

  98-99  Trichloroethylene in  finished water  and  removal
              by 0.76 meter   (2.5  feet) of GAC	168-169

  100    Trichloroethylene in  finished water  and  removal
              by 1.52 meters  (5  feet) of GAC	170

  101    Trichloroethylene in  finished water  and  removal
              by 2.29 meters  (7.5  feet) of GAC	171

  102    Trichloroethylene in  finished water  and  removal
              by 3.05 meters  (10 feet) of GAC	172

  103    Tetrachloroethylene in  finished water and removal
              by 0.76 meter   (2.5  feet) of XE-340	174

 104-105 Chlorobenzene in finished water and  removal by
              0.76 meter   (2.5 feet) of XE-340	176-177

 106     Chlorobenzene in finished water and  removal by 0.76
              meter,  (2.5 feet)  of GAC and 0.76 meter
              (2.5 feet) of IRA-904 resin	178

 107     Chlorobenzene in finished water and  removal by
              0.76, 1.52, 2.29 and 3.05 meters (2.5, 5,
              7.5 and 10 feet) of  GAC	179

 108-109 o, m and p-Dichlorobenzene in finished water and
             .removal by 0.76 meter  (2.5 feet) of XE-340.182-183

 110     o, m and p-Dichlorobenzene in finished water and
              removal by 0.76 meter  (2.5 feet) of IRA-904
              resin	184

 111     p-Dichlorobenzene in  finished water  and  removal
              by 0.76, 1.52, 2.29  and 3.05 meters  (2.5, 5,
              7.5 and 10 feet) of  GAC	 185

 112     o-Dichlorobenzene in  finished water  and  removal
              by 0.76, 1.52, 2.29  and 3.05 meters  (2.5, 5,
              7.5 and 10 feet) of  GAC	 186

 113     Cubic centimeters adsorbed by each GAC column
              for all halogenated  compounds added together... 189

 114     Cubic centimeters of total HOC entering  and
              adsorbed by GAC column #1, 2, 3 and 4 in
              122 days	190

 115     Cubic centimeters of total HOC adsorbed  per
              100 grams of GAC in  GAC columns ttl, 2, 3 and
              4 in 122 days	  191
                               xiii

-------
Number                                                      Page

 116     Adsorption wave front defined by breakthrough
              and saturation time for HOC and Type II and
              Type III substances thru 0.76 meter
              (2.5 feet) of GAG	 193

 117     THM FP in finished water and removal by 0.76, 1.52,
              2.29 and 3.05 meters (2.5, 5, 7.5 and 10
              feet) of GAG	 198

 118     TOG in finished water and removal by 0.76, 1.52, 2.29
              and 3.05 meters (2.5, 5, 7.5 and 10 feet)  of
              GAG	 199

 119     THM FP in finished water and removal by 0.76
              meter  (2.5 feet)  of GAG	 200

 120     THM FP in finished water and removal by 1.52
              meters (  5 feet)  of GAG	 201

 121     THM FP in finished water and removal by 2.29
              meters (7.5 feet)  of GAG	 202

 122     THM FP in finished water and removal by 3.05
              meters (10 feet)  of GAG	 203

 123     THM FP in raw water and removal by 0.76 and
              1.52 meters (2.5  and 5  feet)  of IRA-904 resin. 204

 124     Test extention data for THM  FP removal by 0.76,
              1.52, 2.29 and 3.05 meters (2.5, 5, 7.5 and
              10 feet)  of GAG	 207

 125-126  THM FP  in raw water and removal by 0.76 meter
              (2.5 feet)  of  GAG  and 0.76 meter  (2.5
              feet)  of  XE-340	211-212

 127     THM FP  in raw  water and removal by 0.76 and 1.52
              meters  (2.5 and 5  feet)  of IRA-904 resin	 213

 128     TOG in  raw water and removal by 0.76 meter
              (2.5  feet)  of  GAG  and 0.76 meter  (2.5 feet)
              of XE-340	 218

 129      TOG in  raw water and removal by 0.76 and 1.52
              meters  (2.5 and 5  feet)  of IRA-904 resin	 219

 130      THM FP  in  raw  water and removal by lime softening
              and by  0.76 meter   (2.5 feet)  of XE-340 on
              H.T.  water	 223
                              xiv

-------
Number

 131     THM FP in raw water and removal by  lime
              softening and by  0.76 meter   (2.5 feet)
              XE-340 on H.T. water	224

 132     THM FP in raw water and removal by  lime
              softening and by  0.76 meter   (2.5 feet)
              of IRA-904  resin  on H.T. water	225

 133     TOC in H.T. water and  removal by 0.76 meter
               (2.5 feet)  of XE-340	228

 134     TOC in H.T. water and  removal by 0.76 meter
               (2.5 feet)  of IRA-904 resin	229

 135-136 THM FP in finished water and removal by 0.76
              meter   (2.5 feet) of XE-340	232-233

 137     THM FP in finished water and removal by 0.76
              meter   (2.5 feet) of GAG and 0.76 meter
               (2.5 feet)  of IRA-904 resin	234

 138     THM FP in finished water and removal by 0.76, 1.52
              2.29 and 3.05 meters  (2.5, 5,  7.5 and
              10 feet) of GAG	235

 139     THM FP in finished water and removal by 0.76
              meter   (2.5 feet) of GAG	236

 140     THM FP in finished water and removal by 1.52
              meters  (5 feet) of GAG	237

 141     THM FP in finished water and removal by 2.29
              meters  (7.5 feet) of GAG	238

 142     THM FP in finished water and removal by 3.05
              meters  (10  feet)  of GAG	239

 143     THM FP substances in grams entering and adsorbed
              by each GAG column in 115 days	241

 144     THM FP adsorption by GAG, XE-340 and IRA-904
              resin per column  0.76 meter  deep (2.5 feet)
              at 49 days		243

 145     THM FP adsorption by GAG, XE-340 and IRA-904 resin
              per 100 grams of  adsorbent at  49 days	245

 146     TOC in finished water  and removal by 0.76
              meter   (2.5 feet) of XE-340	249
                               xv

-------
Number

 147     TOG in finished water and removal by 0.76
              meter  (2.5 feet)  of IRA-904 resin and
              by 0.76 meter  (2.5 feet)  of GAG	250

 148     TOG adsorption by GAG,  XE-340 and IRA-904
              resin per column 0.76 meter  (2.5 feet)
              deep at 49 days	251

 149     TOG adsorption by GAG,  XE-340 and IRA-904
              resin per 100 grams of adsorbent at 49 days....252

 150     Comparison of laboratory bottle aged and
              distribution system THM growth	255

 151     Total THM growth in rechlorinated - 2 day
              aged IRA-904 resin column effluent	258

 152     Total THM growth in IRA-904 resin column effluent
              due to THM FP conversion	259

 153     Total THM growth in rechlorinated - 2 day aged
              GAG column effluent	260

 154     Total THM growth in GAG column effluent due to
              THM FP conversion	261

 155     Total THM growth in rechlorinated - 2 day aged
              2,29 and 3.05 meters (7.5 and 10 feet) GAG	263

 156     THM breakthrough and THM FP conversion components
              of Total-THM in two day aged effluent from
              2.29 meters (7.5 feet)  deep GAG column	265

 157     THM breakthrough and THM FP conversion components
              of total-THM in 2  day aged effluent from
              3.05 meters (10 feet)  deep GAG column	266

 158     Inst.  THM in  finished water and finished water
              through  2.29 and 3.05 meters (7.5 and 10
              feet)  of  GAG	267

 159      Relationship of  TOG and THM FP data in raw, H.T.
              and finished water	269

 160      TOG  and THM FP in finished water thru 3.05
              meters (10  feet) of GAG	271

 161      Level  of chlorodibromomethane entering and
              leaving the partially exhausted 0.76
              meter  (2.5 feet)  deep XE-340 column	272

                              xvi

-------
162     Adsorption and leaching of chlorodibromomethane
             on a 0.76 meter   (2.5 feet) deep XE-340 column274

163     Level of Bromodichloromethane entering and
             leaving the partially exhausted 0.76 meter
             (2.5 feet) deep XE-340 column	275

164     Adsorption and leaching of bromodichloromethane
             on a 0.76 meter   (2.5 feet) deep XE-340
             column	  276

165     Level of chloroform entering and leaving the
             partially exhausted 0.76 meter  (2.5 feet)
             deep XE-340 column	277

166     Adsorption and leaching of chloroform on a 0.76
             meter   (2.5 feet) deep XE-340 column	278

167     Level of cis 1,2-Dichloroethene entering and
             leaving the partially exhausted 0.76 meter
             (2.5 feet) deep XE-340 column	279

168     Adsorption and leaching of cis 1,2-dichloroethene
             on a 0.76 meter   (2.5 feet) deep XE-340
             column	280

169     Chloroform adsorption by 0.76 meter  (2.5
             feet) of GAC	286

170     Butane adsorption curves for F-400 GAC and
             XE-340 resin	  287

171-173 Chloroform adsorption by 0.76 meter  (2.5 feet)
             of XE-340	293-294, 296

174     cis ±f 2-rDichloroethene adsorption by 0.76 meter
             f2.S feet) of XE-340	297

175     cis 1,2-Dichloroethene adsorption by 0.76 meter
             (2.5 feet) of GAC	299

176     Bromodichloromethane adsorption by 0.76 meter
             (2.5 feet) of XE-340	302

177     Bromodichloromethane adsorption by 0.76 meter
             (2.5 feet) of GAC	303

178     Chlorodibromomethane adsorption by 0.76 meter
             (2.5 feet) of XE-340	305

179     Chlorodibromomethane adsorption from 0.76
             meter   (2.5 feet) of GAC	 306
                            xvii

-------
                             TABLES
Number                                                      Page

   1     Typical partial analyses, John E. Preston
             Water Treatment Plant	     14

   2     Average chemical application to raw water,
             John E. Preston Water Treatment Plant.  ...     15

   3     Experimental design number and starting
             and ending dates	    26

  4-6    Average concentration of specific halogenated
             organics in raw, H.T. and finished water.  .  . 33-35

  7-8    Percent removal or increase factor for
             specific HOC in H.T. and finished water.  .  .  37, 39

   9     Average total Inst THM and percent of individual
             THM in each experimental design	     40

  10     TOC and THM FP removal by lime softening in
             full scale plant	    41

  11     TOC, terminal THM and THM FP reduction by chlori-
             nation, contact basin and/or sand filtration. .  43

  12     Average pH and free chlorine (ppm) values in
             each experimental design	48

13-18   Specific HOC adsorption data, raw water. . . betw. 51-72

19-24   Specific HOC adsorption data, H.T. water. . .betw. 77-103

25-36   Specific HOC adsorption data? finished water.betw.105-181

  37     Observed adsorptive capacity of 100 grams of GAG
          for'five HOC from finished water compared to
          the Polanyi-Manes predicted value for each
          compound from pure water (adsorbed by 0.76
          meter  [2.5 feet]  of GAG)	    194

38-39   THM FP  and TOC adsorption data,  raw water. . .  . 209,  217

40-41   THM FP  and TOC adsorption data,  H.T. water. .  .  ,221,  227
                              XVlll

-------
Number                                                      page

  42    THM FP adsorption data from finished water	   230

  43    Effect of THM FP influent concentration in raw
          water on adsorption data interpretation	  244

  44    TOC adsorption data from finished water	   247

  45    Sampling procedure to compare laboratory and
          distribution system aging	   254

  46    Total THM growth in adsorbent column effluents	   257

  47    Physical data for Polanyi-Manes calculations	   289

  48    Chloroform adsorption data from finished water	   291

  49    cis 1,2-Dichloroethene adsorption data from raw,
          H.T. , and  finished water	   300

  50    Bromodichloromethane adsorption data from H.T.
          and finished water	   301

  51    Chlorodibromomethane adsorption data from H.T.
          and finished water	   304

  52    Summation of adsorption parameters by GAC	   308

  53    GC/MS HOC confirmation data	   310
                               xix

-------
 GAC




 H.T.




 NORS




 GC




 GC/MS




 TOC




 THM




 HOC




 THM FP




 Total THM




 Inst. THM




 Terminal THM




 BAG




 UV




 m3/s




 mgd




 ED




 L/h




 GPH




MTZ




DOM




EBCT
 LIST  OF ABBREVIATIONS



- granular  activated  carbon



- Hydrotreator  (up-flow  lime  softening unit)



- National  Organics Reconnaissance  Survey



- gas  chromatograph



- gas  chromatograph/mass  spectrograph



- total organic carbon



- trihalomethane(s)



- halogenated organic compound(s)



- trihalomethane  formation potential



• total trihalomethane(s)



• instantaneous trihalomethane(s)



• terminal  trihalomethane(s)



• biologically activated  carbon



- ultraviolet



• cubic meters/second



• million gallons per day



•  experimental design



•  liters/hour



  gallons per hour



  mass transfer zone



  dissolved organic matter



  empty bed contact time
                            xx

-------
                         ACKNOWLEDGMENTS

     Unquestionably, the most valuable help, leadership and
guidance for the Project was provided by the Project Officer,
Mr. Jack DeMarco.  His unselfish and tireless effort in getting
the study organized and keeping it moving in the proper direc-
tion was one of the major reasons for its success.

     The support and cooperation of President Harold Crosby,
Florida International University; Mr. Garrett Sloan, Director
of the Miami-Dade Water and Sewer Authority, and Dr. Richard
Morgan, Director of the Dade County Department of Public Health,
are greatly appreciated.

     The advice of the Technical Advisory Board was an important
factor in evaluating accumulating results of the work.  Their
comments on projected goals are appreciated.

     Advisory Board members were Mr. Anthony Clemente, Dade
County Department of Environmental Resources Management;
Mr. Glenn Dykes, State Department of Environmental Regulation;
Mr. Sidney Berkowitz, Consultant; Dr. John Davies and Dr. Henry
Enos from the University of Miami, as well as Dr. Morgan and
Mr. Sloan.

     The secretarial services of Ms. Phyllis Engles, Ms. Barbara
Weil, Ms. Linda Rountree, Ms. Sarah Bostwick, Ms. Bess Simon,
and Ms. Marlene Blosucci were well performed and appreciated.

     The technical staff of Florida International University
and the Metropolitan Dade County Water and Sewer Authority
contributed much to the success of the program.  These include
Mr. Hunt Harween, Mr. Besteiro Palomeque, Mr. Kenneth Kirkman,
Mr. Russell Lang, Mr. Cesar Ordaz, Mr. William Booth, and
Ms. Laurie Miller.

     Appreciation must be expressed also for the cooperation of
the personnel in the business office and accounting department
of the Dade County Department of Public Health, Florida Inter-
national University, and the Metropolitan Dade County Water and
Sewer Authority.

     The technical assistance of Dr. Milton Manes, Chemistry
Department, Kent State University, Kent, Ohio and Dr. Michael
Rosene, Calgon Corporation, Calgon Center, Pittsburgh, Penn-
sylvania is also acknowledged.

                              xxi

-------
                            SECTION  I

                           INTRODUCTION

     In 1975, the U.S. Environmental Protection Agency announced
by their release of  a report on  "National Organics Reconnais-
sance Survey  for Halogenated Organics in Drinking Water"  (1)
 (NORS) that the drinking water of Dade County, Florida contained
over 300 ppb  chloroform and nearly  6 ppb vinyl chloride. Person-
nel at the Dade County Health Department, Miami-Dade Water and
Sewer Authority, and the Drinking Water Quality Research Center
of Florida International University in cooperation with the U.S.
Environmental Protection Agency  in  Cincinnati, Ohio developed a
research project to  1) develop an effective and economical
method for removing, or substantially reducing, the chemicals of
concern and 2) remove organic solutes  (precursors) from the
water to prevent regrowth  of these  chemicals in the distribution
system where  free chlorine is present.  On June 22, 1976 a one-
year study was approved by EPA and  later extended to September
5, 1978.

     The scope of the Project included 1) a study of the effec-
tiveness of various  adsorbents in removing potential carcinogens
already present in raw water and those generated in the water
treatment process, 2) removal of precursor substances which form
halogenated organics upon  reaction  with chlorine in the treat-
ment plant and distribution system, and 3) effect of the stage
of treatment process on efficiency  for removing contaminants.

     Early in the Research Project  the high levels of halogen-
ated organics reported by  the NORS  (1) study were verified both
in concentration and identification by GC/MS.  These results
were reported at once to local and  state authorities.

     This report describes the results of a study of two adsorb-
ent resins and granular activated carbon for their effectiveness
and efficiency in removing trihalomethane precursors, halogen-
ated organic compounds and total organic carbon from three
locations in the treatment plant, raw, lime softened and fin-
ished water.  The resins were Ambersorb XE-340, a carbonized
polymeric macroreticular resin,  and IRA-904, a strong base
cationic resin for anion exchange,  both manufactured by Rohm and
Haas Company, Philadelphia, Pa.  The Granular Activated Carbon
was Filtrasorb 400, 12 x 40 mesh, manufactured by Calgon Cor-
poration, Pittsburg, Pa.

-------
      The  Miami-Dade Water  and  Sewer Authority  furnishes  water
 for over  one  million people  through three major water plants.
 The John  E. Preston Water  Treatment Plant in Hialeah, Dade
 County, Florida, which operates  at 2.63 m3/s  (60 mgd), draws
 water from the  Biscayne Aquifer  from  seven wells located on the
 plant site.   The wells are approximately 27.4  meters  (90 feet)
 deep. The raw  ground water, which contains an average of 10
 mg/L of Total Organic Carbon,  is treated by lime softening in an
 up-flow Hydrotreator, breakpoint chlorination, and  sand  filtra-
 tion. Before the water leaves the plant the free chlorine level
 is  adjusted to  2.5 ppm.

      The  strata overlying  the  recharge area of the  Biscayne
 Aquifer are predominately  muck,  which accounts for  the rela-
 tively high color present  in the source water.  Thus  organics of
 a natural origin comprise  one problem that cannot be  easily pre-
 vented by attempting to change sources of drinking  water in
 this area.  Other organic  substances  that are  a result of man's
 activities are  also present and  pose  another facet  of the prob-
 lem presented in using the ground water in the area.  Thus
 initial studies of practical methods  of removing organics were
 directed  at attempting to  find a broad based organic  removal
 method.

      The  three  adsorbents  were studied in four experimental
 designs developed by Jack  DeMarco, EPA Project Supervisor.
 Glass columns 2.54 cm  (one inch) in diameter were connected
 directly  to raw, lime softened  (Hydrotreator effluent) and fin-
 ished water lines from the Preston Plant.  Adsorbent  bed depths
 studied were  0.76, 1.52, 2.29 and 3.05 meters  (2.5, 5, 7.5 and
 10  feet)  for  Granular Activated  Carbon, 0.76 and 1.52 meters
 (2.5 and  5  feet) for IRA-904 resin, and 0.76 meter  (2.5 feet)
 for XE-340.   A  flow rate of 122L/min./m2 (3 gal./min./ft.2)  was
 maintained  by rotometers through each column.  Thus empty bed
 contact times were always  6.2 minutes whenever a 0.76 meter
 (2.5 feet) bed  depth of adsorbent was used, 12.4, minutes for a
 1.52 meter  (5 feet) bed, 18.6 minutes for a 2.29 meter (7.5
 feet)  bed  and 24.8 minutes for a 3.05 meter (10 feet) bed.

      The  3.05 meter (10 feet) bed depth of Granular Activated
Carbon (24.8  minutes Empty Bed Contact Time) showed the  most
promise for achieving broad based removal of organics.   Water in
the  normal distribution system was evaluated as a comparison
with  experimental results  obtained by using adsorbents.

-------
                           SECTION II

                          CONCLUSIONS

Full Scale Plant Performance

1.  Over the two-year study, the high levels of halogenated
    organic compounds reported by the national Organics Recon-
    naissance Survey  (1) study were verified both in concen-
    tration and identification by Gas Chromatograph/Mass
    Spectrograph.

2.  The average concentration of halogenated organic compounds
    present in raw water was approximately 15 percent less after
    lime softening  (Hydrotreator effluent).

3.  The average level of trihalomethanes in finished water
    leaving the plant over the two-year study was 67, 43, 28 and
    2 yg/L respectively for chloroform, bromodichloromethane,
    chlorodibromomethane and bromoform.  The average total
    trihalomethane level leaving the plant was 140 yg/L. In the
    distribution system, this level can double in less than two
    days.  The level would rise higher, but the free chlorine is
    exhausted in one to two days.  When, additional free chlorine
    is added at booster stations in the distribution system,
    trihalomethane levels reach their maximum.

4.  The level of some non-trihalomethane halogenated organic
    compounds increased during the plant treatment process.  A
    consistent increase was found for the summed concentration
    of 1,1,1-trichloroethane, 1,2-dichloroethane and carbon
    tetrachloride.  Increases of this summed concentration
    ranged from 1.2 to 77 times the level in raw water.  Since
    the three compounds were summed due to overlapping gas
    chromatograph peaks we do not know if the increase was due
    to one or more of the three substances.  Some other non-
    trihalomethane halogenated organic compounds may have shown
    intermittent increases.

5.  In general, non-trihalomethane halogenated organic sub-
    stances were not consistently well removed by the existing
    full scale treatment plant.

6.  Lime softening removed an average of 28 percent of the
    trihalomethane formation potential and little additional

-------
     actual removal was achieved by the sand filtration process.
     Calcium carbonate floe was believed to provide the mechanism
     for trihalomethane formation potential removal in the full
     scale plant.

 7.   In the Preston Plant, the amount of precursor removal by
     conversion to trihalomethanes by the combined chlorination-
     sand filtration process averaged 23 percent of the Hydro-
     treator effluent level.


 Specific Organic Removal by Adsorbents

 8.   In both raw and finished water, XE-340 has more adsorptive
     capacity in weight of organic substance adsorbed per unit
     weight or volume of adsorbent, for individual halogenated
     organic compounds than granular activated carbon.  While
     the values are different for each halogenated organic com-
     pound and different in raw and finished water, in general,
     XE-340 has approximately three times the adsorptive capacity
     of granular activated carbon.        .

 9.   The adsorptive capacity in weight of organic substances
     adsorbed per unit weight or volume of adsorbent of XE-340
     for halogenated organic compounds is only slightly greater
     when treating raw water than when treating Hydrotreator
     water.  The lower total organic compound concentration in
     the Hydrotreator water (approximately 30 percent lower)  did
     not enhance the ability of the adsorbents for halogenated
     organic compound removal.

 10.  The adsorptive capacity of both XE-340 and granular acti-
     vated carbon for cis 1,2-dichloroethene is less in finished
     water than raw water despite a reduction of 34 percent total
     organic carbon.  The percent of cis 1,2-dichloroethene of
     total halogenated organic compounds in raw and finished
    water is 86.5 and 10 percent respectively.  We attribute the
     30 percent reduction in adsorptive capacity for cis 1,2-
     dichloroethene in finished water to increased competitive
     adsorption from the additional halogenated organics present
     in the finished water.

11. On raw and Hydrotreator water, IRA-904 resin showed no
    removal of any of the halogenated organic compounds present.


12. On finished water,  IRA-904 resin appears to enhance the
    reaction of free chlorine with precursors to form halo-
    genated organic compounds.  The effluent of a 0.76 meter
     (2.5  feet)  deep bed (empty bed contact time of 6.2 minutes)
    contained 1.75 and 1.13 times the influent concentration of

-------
    chloroform and bromodichloromethane respectively.  Increases
    in concentration occurred in some of the non-trihalomethane
    halogenated organic compounds, but the majority showed no
    increase nor decrease in concentration as a result of the
    IRA-904 resin as observed in raw and Hydrotreator water.

13. A 3.05 meter (10 feet) deep bed of granular activated carbon
    with an empty bed contact time of 24.8 minutes was ineffec-
    tive for vinyl chloride removal.

Total Organic Carbon and Trihalomethane Formation Potential
Removal by Adsorbents

14. Total organic carbon data did not consistently correlate
    with trihalomethane formation potential data.  Also, one
    cannot be converted into the other by a single conversion
    factor since total organic carbon analysis measures some
    substances that are not trihalomethane precursors.  As
    expected, total organic carbon analysis is not a precise
    useful indicator of trihalomethane precursors.  However,
    general trends might be noted at a given site.

15. In this report, the shape of the adsorption breakthrough
    curves for total organic carbon and trihalomethane formation
    potential removal by adsorbents are similar to specific
    halogenated organic compound removal curves.

16. A system was devised that could be used on total organic
    carbon and trihalomethane formation potential substances to
    allow a more complete understanding and comparison of
    adsorbent performance.

17. IRA-904 resin removed trihalomethane formation potential
    more efficiently from raw water than did the other two
    adsorbents tested based on percent removal and based on
    weight adsorbed per unit volume of adsorbent.  However,
    granular activated carbon removed trihalomethane formation
    potential more efficiently than the other adsorbents based
    on the weight of trihalomethane formation potential adsorbed
    per unit weight of adsorbent.
18. Although IRA-904 resin was more effective for trihalomethane
    formation potential removal than granular activated carbon
    on a volume basis, and only 20 percent less effective on a
    weight basis, it is important to note that the resin allowed
    about 100 ug/L of trihalomethane formation potential to pass
    through the bed at start-up even with a 1.52  (5 feet) bed
    depth.  The 0.76 meter (2.5 feet) granular activated carbon
    system was able to produce an effluent with about 12 yg/L of
    trihalomethane formation potential at start-up.  Thus, a
    single measure of effectiveness cannot be applied without
    knowledge of performance required of the adsorbent.  The
    breakthrough curve is important in determining the adsorbent

-------
     performance  for  a  specific effluent criteria.

 19.  A direct  comparison of 0.76 meter  (215 feet) beds  (6.2
     minutes empty bed  contact time) of granular activated carbon
     and XE-340 shows that carbon removed more trihalomethane
     formation potential on both an equal adsorbent weight and
     volume basis when  receiving raw water.  Furthermore, the
     breakthrough plot  indicates that carbon maintained a lower
     effluent  concentration than XE-340 for about 17 days.
     During this time period neither adsorbent was very effective
     for trihalomethane formation potential removal at the con-
     ditions tested.

 20.  Lime softening removed an average of 28 percent of trihalo-
     methane formation potential precursors from raw water.  This
     compares  with 29 and 24 percent removal by 0.76 meter   (2.5
     feet), 6.2 minutes empty bed contact time, of granular
     activated carbon and XE-340 over a 119-day test and 46 per-
     cent removal by a bed of IRA-904 resin after 49 days of
     operation.  If all three adsorbents are compared after 49
     days of operation time, removals affected are 26, 24 and 46
     percent for carbon, XE-340 and IRA-904 resin respectively.
     A bed of  IRA-904 resin 1.52 meters (5 feet)  deep, 12.4
     minutes empty bed contact time, removed 55 percent after 49
     days of operation.

 21.  On a weight basis, calcium carbonate floe removed one-third
     as much trihalomethane precursor from raw water as carbon.

 22.  The XE-340 bed removed an average of four percent trihalo-
     methane formation potential from Hydrotreator water as
     compared with 24 percent removed from raw water.

 23.  The IRA-904 resin bed removed an average of 32 percent
     trihalomethane formation potential from Hydrotreator
     effluent as compared to 46 percent from raw water.

 24. An  XE-340 column 0.76 meter  (2.5 feet)  deep removed no
    trihalomethane formation potential from finished water.

25. A  0.76 meter (2.5 feet)  deep bed of IRA-904  resin removed
    13 percent trihalomethane formation potential from finished
    water.   At no time during the test period was the effluent
    trihalomethane formation potential concentration from this
    column below 180 yg/L.

26. Granular activated carbon was more efficient in removing
    trihalomethane formation potential precursors from finished
    water than the other two adsorbents.   For example, in two
    separate runs 0.76 meter  (2.5 feet)  of carbon removed 18
    and 22  percent,  and it was the only adsorbent tested that
    removed enough precursor to keep the trihalomethane form-

-------
ation potential level below 100 yg/L.

Finished Water

27.  Granular activated carbon was chosen for deep bed studies
     because it was the best broad based adsorbent for removal
     of organics in our system.  The deep bed studies were car-
     ried out on finished water which had the lowest level of
     total organic carbon of any location in the plant, and did
     not suffer from execessive calcium carbonate precipitation.
     In our system, at a flow rate of 122L/min./m2 (3 gpm/ft.2),
     an empty bed contact time of approximately 24.8 minutes in
     a 3.05 meter (10 feet) deep granular activated carbon bed
     was necessary to achieve a bed life of 81 days.

28.  Free chlorine residuals were completely removed by 0.76
     meter   (2.5 feet) of granular activated carbon and IRA-904
     resin throughout their respective test periods,  whereas the
     XE-340 completely removed the free chlorine residual for
     about 17 days.

29.  Combined chlorine residuals penetrated all adsorbents
     tested.

30.  Laboratory bottle aging of finished water as a means of
     predicting trihalomethane growth in the distribution system
     produced comparable results.

31.  An XE-340 adsorbent column, partially saturated with halo-
     genated organics in finished water, was treated with halo-
     genated organic-free water to test for halogenated organic
     leaching (desorption).  Desorption of cis 1,2-dichloro*
     ethene, chloroform, bromodichloromethane, and chlorodi-
     bromomethane appeared to follow a curve that was the
     reverse of the adsorption curve.

General

32.  The Polanyi-Manes Theory of adsorption was useful in inter-
     preting and explaining our data.

33.  A bacterial profile study was made on raw and finished
     water at the Preston Plant, and the effluent from the
     granular activated carbon columns.  Two reports  of this
     work are appended to this report.  Clearly, we had a
     Biological Activated Carbon system.  As no additional
     oxygen was added to the water (as in European practice), we
     call our system a partial biological activated carbon
     system.  We cannot speculate on the results of the bacte-
     rial growth were not present.  We do feel, however, that
     despite the massive bacterial growth that eventually pre-
     vented back washing of the columns, adsorptive capacity

-------
of the granular activated carbon for halogenated organic
carbon was not decreased.  Initial breakthrough and
saturation time for each halogenated organic compound
through each column were too consistent to suggest blocking
of active sites by the bacteria.  Bacteria develop large
populations in granular activated carbon columns which
slough off into the column effluents in large numbers,
necessitating disinfection before release into the distri-
bution system.

-------
                           SECTION III

                         RECOMMENDATIONS

1.  As finished water leaves a 3.05 meter  (10 foot) deep granu-
    lar activated carbon bed, it has nil free chlorine and a
    high population of bacteria.  It would have to be rechlori-
    nated to again achieve disinfection and according to pre-
    sent practice in Florida it would have to contain approxi-
    mately 2.5 ppm of free chlorine to provide residual disin-
    fectant before it could enter the distribution system.  As
    some precursors are still present, trihalomethane regrowth
    in the distribution system will occur.  We therefore define
    granular activated carbon exhaustion or bed life as the
    point where trihalomethane regrowth in a sample of the 3.05
    meter (10 foot) granular activated carbon bed effluent
    (after rechlorination to 2.5 ppm of free chlorine and aging
    for two days) reaches the proposed Minimum Concentration
    Level of trihalomethane or 0.1 mg/L (100 ppb).  This level
    was reached in 81 days.  Failure at 81 days, however, was
    not due solely to trihalomethane growth from precursors.
    The column had become saturated to chloroform.  If the
    influent water had not contained such a high concentration
    of trihalomethanes (140 yg/L), the bed life would have been
    somewhat longer.  We can only guess that bed life would have
    been extended another two weeks.  If ozone replaced break-
    point chlorination after the Hydrotreator water, the high
    trihalomethane load would be eliminated, and,  according to
    European reports, precursors would also be greatly reduced
    by subsequent granular activated carbon-biological activated
    carbon treatment.  Bed life then would be greatly prolonged.
    This remains to be confirmed in our system.  We therefore,
    recommend a pilot plant research project located at the
    Preston Water Treatment Plant to study the possibility of
    prolonging granular activated carbon bed life in our system
    by the use of ozone followed by biologically activated
    carbon.

2.  Since a granular activated carbon column effluent contains
    a high population of bacteria and since such a column would
    probably be used at the end of a conventional treatment
    plant we recommend that a bacterial study be made before
    widespread use of such a system is adopted.  Specifically,
    to determine if the disinfection with 2.5 ppm of free
    chlorine at the end of the treatment process is adequate

-------
    with the associated treatment plant contact time which will
    range widely from a few minutes to 24 hours or more before
    discharge into the distribution system.   Our work indicates
    that the standard plate count method is  apparently inade-
    quate in assessing numbers and types of  bacteria that are
    found in a granular activated carbon effluent.  Conditions
    for an optimal bacterial method apparently are still being
    worked out.   We recommend that a committee of bacteriolo-
    gists should be formed to adopt interim  bacteria test
    methods and set up research projects for further study to
    determine optimal methods.

3.  Considerable work is being done in the field of adsorption
    kinetics.  Perhaps more should be done with the Polanyi-
    Manes adsorption theory.  This might include the individual
    adsorption of most of the specific halogenated organic com-
    pounds found in raw and treated water from purified water
    and water containing known amounts of total organic carbon
    and added amounts of other specific halogenated organic
    carbons to study competitive effects.

4.  There are several questions about the source of halogenated
    organic carbon in raw water that should  be answered,  such
    as;

    a.  Source of our high level of cis 1,2-dichloroethene
    b-  Source of vinyl chloride
    c.  Lack of  or very low level of trihalomethanes when other
        volatile halogenated organic carbons are present
    d.  Effect of untraviolet and bacterial  enzyme action on
        specific halogenated organic carbon.
                              10

-------
                           SECTION  IV

                 PLANT AND EQUIPMENT DESCRIPTION

PRESTON PLANT WATER  SOURCE

     The terrain in  the vicinity of Miami and its neighboring
municipalities consists of an outcropping of soft, porous,
Oolite limestone rock.  Water enters this porous rock from local
rainfall, runoff, and from the extensive canal system in south-
ern Florida.  The approximate annual rainfall in the area is
152.4 cm  (60 inches) per year.  The porous water-bearing rock is
the aquifer from which raw water is drawn by the water treatment
plants in the area.

     Seven wells have been drilled  into the surface rock on or
near the Preston Plant site.  This  groundwater has high color
and contains dissolved iron.  The color is attributed in large
part to the organic  matter leached  from decaying vegetation
through which the groundwater percolates.  The water is slightly
basic with an average pH of  7.2.

PRESTON PLANT SITE

     The Miami-Dade  Water and Sewer Authority, through its three
major water plants,  furnished water, either directly or indi-
rectly for over one  million  people.  One of these three major
water plants is the  John E.  Preston Water Treatment Plant (Flow
Diagram in Figure 1), located at 1100 West Second Avenue,
Hialeah, Florida.  The Preston Plant has been in operation
approximately seven  years.

     At present the  Preston  Plant is rated at 2.63 m3/s (60
mgd), and in general is operated near or at maximum capacity.
The wells supplying  the Preston Plant are approximately 27.4 m
(90 feet) deep.  Each well can produce over 34065 m3 (nine
million gallons) per day.  Raw water from these seven wells is
fed to a combination of three upflow Hydrotreator (H.T.) soft-
eners, each rated at 0.88 m3/s (20 mgd).  Silica, activated with
chlorine, is added to the raw water just prior to its entrance
into the upflow softener.  The upflow softener effluent is
channeled into a recarbonation flume.  Sodium silica fluoride is
added at this point.  Chlorine is then added just before the
water enters the chlorine contact basin.  After an average
retention time of 1.25 hours in the chlorine contact basin, the


                               11

-------
     3 O
    »z
                          CHLORINE DIFFUSION
                              IB.O ±PPM
          0.5 ± PPM FLUORIDE
   I 2
 5 MOD RAPIO
SAND FILTERS
  ' 60 M.CD
CLEAR
 WELL
O.6 MOD
                                                 VENTURI
                                                  METER
                      RESERVOIR
                       ».0 H6
       wf

       /
            Q_5) CAPAC1TT  OF  PUMPS IN MOO

            O  WELL
                                                                  -O«6
Figure  1.   Flow diagram  of John  E. Preston Water
              Treatment Plant.
                                 12

-------
water flows into rapid  sand  filters  0.22 m3/s  -  122L/min./m2
 (5 MGD - 3 GPM/ft.2), then to  a  34065 m3  (nine million gallon)
reservior.  From the reservoir the water is  diverted to the
Hialeah Plant  for pumping, or  pumped directly  to high pressure
distribution lines.  Tables  1  and 2  contain  chemical data
related to the Preston  Water Treatment  Plant.

BENCH SCALE ADSORPTION  TEST  UNIT

     The Bench Scale Adsorption  unit is located  in  the second-
 floor laboratory of the Preston  Plant.  Three  sampling lines
enter the laboratory from the  plant  raw water,  Hydrotreator,
 and  clear well composite lines.   The water is  constantly moni-
 tored by pH chart recorder,  and  samples are  readily available
 at this one location.   Routine testing  is done hourly around-
 the-clock.  A  flow diagram of  the Bench Scale  Adsorption unit is
 shown in Figure  2.  Each glass column is 1.52  meters  (five feet)
 long by 2.54 cm  (one inch) in  diameter.

     A flow rate of  122L/min./m2 (3  gal./min./ft.2) was main-
 tained by rotometers.
  /
     The pumps used  on  the three sample lines  from  the plant to
 the  laboratory were  three-quarter housepower,  water lubricated,
 with Teflon seals.   All lines  were  copper pipe.  Construction
 details of the Bench Scale Adsorption unit are shown in Figures
 3 and  4.
                                13

-------
                  TABLE  1.   TYPICAL PARTIAL ANALYSES,
                JOHN -E.  PRESTON WATER' TREATMENT  PLANT
 Alkalinity (CaCO,)
    Phenolphthalein
    Methyl  Orange
 Hardness  (CaC03)
    Non-Carbonate
    Total
 Carbon Dioxide, Free  (CO,)
 Chlorine Residual (Cl,)
    at plant           •
 Chlorides  (Cl)
 Fluorides  (F)
 Sulfates (SO4)
 Calcium (Ca)
 Iron (Fe)
 Magnesium  (Mg)
 Sodium & Potassium (as Na)
 Silica (Si02J
 turbidity
Total Solids
Electrical Conductivity
    (EC x 10  @ 25°C)
                                                       Treated Water
                                       After Softening,   Entering
                                            Before      Distribution
                             Well Water  Chlorination      System
                                 0.
                               230.
                                20.
                               250.
                                25.
                                 0.

                                40.
                                 0.2
                                24.
                                88.
                                 0.8
                                 7.0
                                29.
                                 8.0
                                Nil
                               350.
     16,
     32.
     21.
     53.
  4.
 40.
 35.
 75.
      0.              0.
      0.05(Combined)  2.0  (Free)
     41.
      0.1
      0.1
Excess 50 units
 55.
  0.7
 24.
 23.
  0.0
  4.2
 32.
  9..0
 Nil
205.
                               580.         	           305.
Color                           50.         25.               6.
pH                               7.3        10.0'- 10.3      8.8
         lAll units 'expressed as mg/Lr  (ppm), .where applicable)
                                 14

-------
      TABLE 2.  AVERAGE CHEMICAL APPLICATION TO
 RAW WATER, JOHN E. PRESTON WATER TREATMENT PLANT
Raw water influent to softening units:

    1.0 to 2.0 ppm chlorine-activated silica —

            chlorine 0.3 to 0.7 ppm.

Softening units:

    160 to 180 ppm CaO as slaked Ca(OH)_

Carbon dioxide addition after softening unit, as

    needed to achieve desired degree of stabilization.

Sodium silica fluoride addition, ±0.5 ppm to bring

    fluoride level to 0.7 ppm in treated water.

Fifteen to 17 ppm chlorine dosage to achieve a chlorine

    residual in treated water leaving plant of 1.5 to

    3.0 ppm free chlorine.  Chlorine contact basin

    average retention time 1.25 hours.  Average minimum

    free chlorine residual at far points in distribution

    system 0.5 ppm.
                          15

-------
        ^s
                                               1.52 meters
                                                   ( 5 feet)
Figure  2.  Bench Scale Column Adsorption Unit
                            16

-------
Figure 3.  Plumbing for adsorption column.
                  17

-------
00
                          Figure 4.  Detailed viev? of column  fittings.

-------
                           SECTION V

                    METHODS AND PROCEDURES

OPERATION OP BENCH SCALE ADSORPTION UNIT

    A flow rate of 122L/min./m2  (3 gal./min.ft.2) was maintained
for this study and is equal to approximately 3.78L per hour
(one gallon per hour), or  89.3 liters per day.  Flow through the
column was maintained at 3.78L/hr.  (1.0 GPH) by adjusting the
rotometer.  This was checked periodically.  The pressure was
adjusted with the regulating valve in order to provide enough
head to maintain the desired flow.  Whenever the increase in
pressure in the column was approximately 7 or 8 psi, the column
was backwashed.  The H.T.  effluent pump, lines, and column were
backwashed every day because of the rapid build-up of calcium
carbonate.  The backwash system for the columns is shown in
Figure 5.  Backwash water was prepared by passing tap water
through a Barnstead Still, an ion-X-changer, and an activated
carbon filter.  This was followed by all glass redistillation.
This water was then boiled and purged with zero-grade helium to
remove volatile halogenated organics.

    Specific data for each Experimental Design not specified
below, such as backwash dates and time, pH, turbidity, color,
and chlorine are presented in Appendix B of this report.

GC ANALYTICAL METHOD

    Purgeable Halogenated Organic Compounds were analyzed
according to the purge and trap method of Bellar and Lichtenberg
(2) with modifications  by Dressman and McFarren (3) for
analysis of vinyl chloride.

    Pure helium was bubbled through a water sample  (0.5 to 7 mL)
at a rate of 20 mL/min. for 11 minutes.  Volatile halogenated
organics were retained in a one-eight inch O.D. by eight inch
stainless still trap.  The first two-thirds of the trap con-
tained Tenax-GC and"the upper one-third contained Silica Gel-15.
Tenax-GC efficiently adsorbes most of the compounds, but Silica
Gel-15 is more efficient in adsorbing the lower molecular weight
highly volatile compounds such as vinyl chloride.

    The trap was backflushed for six minutes at 220?C with
helium (20 ml/min.) onto the GC column (1/8 in. x 7 ft. stain-


                               19

-------
2.3
          Figure  5.   Backwash system  for columns.

-------
less steel packed with Tenax-GC).  Helium carrier gas  (40 mL/
mm.) was then turned on and the oven set for an 18-minute
isothermal hold at 95°C.  The balance of the full 52-minute run
was programmed at 4°C/min. to 220°C.  The Hall Electrolytic
Conductivity Detector reduces all halogenated compounds to
halogen acids, which are then detected.

    A typical chromatogram of a standard sample is shown in
Figure 6.  These are the 19 specific halogenated organic com-
pounds routinely monitored.  Each peak is identified by the
chemical name and number and concentration in micrograms in the
standard.  No significant amount of methyl iodide was found
during this study.  Values for compounds No. 7, 8, and 9 were
summed because the three peaks overlapped and made separate
analysis impractical.  In most of this report values,for the
three isomers of dichlorobenzene were summed.  In ED4 they were
reported separately, because improved chromatographic technique
separated the three isomers.

TOG ANALYSIS

    TOC values were obtained by the EPA Laboratory in Cincinnati,
Ohio on a Dohrmann-Envirotech Organic Analyzer with an Ultra Low
Organics Module.

TRIHALOMETHANES, TERMINAL TRIHALOMETHANES AND TRIHALOMETHANE
FORMATION POTENTIAL

    Four general individual THM compounds were qualified and
quantified in the study as a part of the HOC analytic program.
They were chloroform, bromodichloromethane, dibromochloro-
methane, and bromoform.  In order to facilitate the investiga-
tion of THM's and their control, other parameters were also
utilized.  These parameters are discussed elsewhere  (4) in more
detail and are defined here as they applied to this project.

1.  Total trihalomethane  (total THM) concentration is the sum-
    mation of the concentrations of the individual THMs in a
    sample.

2.  Instantaneous THM  (inst. THM) is the concentration of TTHM
    in the water at the time the sample is collected.

3.  Terminal THM (term. THM) is the sum of TTHM present in the
    water at the moment of sampling and TTHM subsequently formed
    during additional reaction time under defined conditions.
    This value may be used as a general estimate of THM con-
    centrations that the consumer would receive if the water
    from the location sampled were subjected to the pH, temper-
    ature, free chlorine residual and storage time conditions
    that were used for the sample.  During the project, the
                               21

-------
                                                 ,— --•!-- i "••

                                                 C.T.;	r_
                                                 L...U -i.
Figure-6.    Typical chromatogram of  halogenatett organics.
                             22

-------
reaction was driven toward completion by adding chlorine to
exhaust the precursor.  Samples were routinely stored at
the finished water pH of 9.0 and a temperature of 22°C for
six days, i.e., beyond the normal detention time in the
distribution system of the utility, with sufficient free
chlorine added to satisfy demand.  After six days under
storage conditions, a concentration was reached that was
assumed to represent a maximum reaction for Preston plant
distribution water.  Modification of these conditions were
made on additional samples in ED3 and ED4 as described
later.  This value may be used as a general estimate of THM
concentrations that the consumer, would receive if the water
from the location sampled were subjected to the pH, tempera-
ture, free chlorine residual and storage time conditions
that were used for the sample.

Trihalomethane formation potential (THM FP)is the difference
between the terminal TTHM and the instantaneous TTHM (term.
TTHM - inst. TTHM = THM FP), an indirect measure of the
unreacted precursor in the water sampled.  It is the
increase in the TTHM concentration that occurred during the
storage period for the determination of the terminal TTHM
concentration.  The unreacted precursor has the potential to
further increase TTHM concentrations in the presence of free
chlorine.

Handling procedures for HOC including Inst. THM's and Term.
Trihalomethane Samples.  All HOC samples for raw water and
adsorbent column effluents on the raw water line were taken
in septum bottles filled to the top and sealed with no air
entrapment.  No reagents were added.  These are referred to
as the odd number samples 1, 3 and 5.  All HOC samples for
the plant Hydrotreator and clear well effluent, as well as
the adsorbent columns receiving these waters were sampled in
the same way except the septum bottles contained one drop of
10 percent sodium thiosulfate.  The sodium thiosulfate was
added to quench the presence of any free chlorine residual
in the sampled water.  These samples are referred to as odd
numbered samples 7, 9, 11, 13, 15, 17, 19, 21, 23 and 25.

All term. THM samples were taken at the same time as the HOC
samples at each location.  The septum bottles were filled to
about the 3/4 level.  An appropriate amount of buffer solu-
tion and free chlorine solution was added and the septum
bottle then-was quickly filled to the top with water sample.
Care was taken to allow no overflow, but yet to avoid air
entrapment in the sample bottle prior to sealing.  The
samples were delivered to FIU for six-day storage at 22°C.
These samples are designated as even numbered samples 2
through 16, 20, 22, 24 and 26.
                            23

-------
     Additional  samples  were  taken  during  ED3  and ED4  to compare
     the value, of  the laboratory  stored samples  for  estimating
     the THM concentrations that  occurred  in the actual  distri-
     bution system.   Clear well effluent samples (designated as
     11+2)  were  taken  in empty  septum bottles, filled  to the
     top and sealed  without the addit-ion of any  reducing agent or
     additional  chlorine.  These  samples were  stored for two days
     at 22°C and then delivered to  FIU  for analysis.   Free
     chlorine residuals  and pH were determined on these  samples
     at the time of  analysis.  The  samples were  collected for
     comparison  with inst. THM concentrations  found  at the Red
     Road distribution system sampling  station which is  two days
     water travel  time from the plant.   Thus,  a  clear  well water
     sample was  stored in a septum  bottle  for  two days and com-
     pared to a  sample  taken in  the distribution system two days
     later.

     Additional  samples  from  the  Red Road  distribution system
     location were also  taken and buffer plus  chlorine solution
     were added  to these samples  prior  to  bottle storage at pH 9
     and 22°C for  four days.  These samples are  designated as
     17+4 and  represent the THM concentration  that was present
     in a water  that was in the actual  distribution  system for
     two days with the pH and chlorine  residuals normally present
     and four additional days in  bottle storage  with a pH of 9.0,
     temperature of  22°C and  presence of free  chlorine residual.
     The THM concentration of this  six-day stored sample,  two
     distribution  days plus four  bottle storage  days  (sample 17 +
     4)  was compared to  the normally obtained  six-day  bottle
     stored term.  THM concentration of  the clear well  water
     (sample 12).

     Along  with  the  sample of the clear well stored  for  two days
     in a bottle (sample 11 + 2), samples  were taken from the
     effluent of the adsorbent columns.  The column  effluent
     samples were  adjusted as required  to  assure that  a  free
     chlorine residual was present  prior to sealing  and  two-day
     storage.  The adsorbent  effluent samples  were used  to esti-
     mate the THM  concentration that might be  received by a con-
     sumer  two days  from the  plant,  if  a specific adsorbent were
     a  part of the normal treatment system.  The concentrations
     of  these adsorbent  effluent  samples (i.e. samples designated
     13+2,  15+2,  23+2 and 25  +2)  were compared  with clear
     well water  samples  stored for  two  days in a bottle  as well
     as  the inst.  THM samples collected at the Red Road  sampling
     station  two days water travel  time from the plant.

DATA ANALYSIS

     The  format  in this  study was to plot  all  individual data
points  to  form  the  adsorption or breakthrough curve.  A typical
plot of  actual  adsorption data for bromodichloromethane is shown

                               24

-------
in Figure 66, page 125.  Initial breakthrough times for each of
the four columns in series is shown.  Saturation times for the
first two columns are shown.  An extrapolated saturation time
for the third column is shown and the saturation time for the
fourth column cannot be extrapolated because of insufficient
data to establish the slope.  The average influent concentration
in ug/L is determined and each adsorption or breakthrough curve
is integrated to determine the amount of substance entering,
passing and adsorbed by the column at breakthrough, saturation
and at the end of the test period or at some other time period
in common with data in another ED.  The breakthrough point was
sometimes difficult to ascertain from the actual data and plot-
ted curves.  For each HOC, adsorbent type and bed depth studied
there was sometimes intermittent low level leakage before the
time we picked as the breakthrough point.  In general, we define
the breakthrough point as the time required to reach 2 yg/L on
the breakthrough curve.  Actual leakage values during the period
before breakthrough are too low to plot on the yg/L scale used
for the complete breakthrough curves .  In some cases the actual
low level of breakthrough is shown above the plotted data point.

    Throughout the study, 76.2 cm  (30-inch) bed depths were used
for each column.  In all calculations an average weight value
was used for GAC, XE-340 and IRA-904 resin per column.  These
values were, respectively, 176, 215, and 275 grams.  The flow
rate through each column was 122L/min./m2 (3 gal./min./ft.) ,
resulting in a flow of 89.3 L/day and approximately 3.785 L/hr.
 (1 gal./hr.) .

    Interpretation of results includes consideration of Mass
Transfer Zone  (MT?) .  The Mass Transfer Zone for a specific sub-
stance is the minimum bed depth at a given flow rate necessary
to prevent column breakthrough after initial flow.  We used the
following equation:

                  MT  = Ts " Tb
                           s
                   T  = saturation time
                    s
                   T,  = breakthrough time
                    D

    The raw water feed for the Preston plant averaged 10 mg/L
of TOC.  However TOC is merely a representation of the carbon
fraction present.  The concentration of Dissolved Organic Matter
(DOM) is some higher amount depending on the relative percent of
carbon in the total molecular weights of organics present.  We
have extimated that the organics present contain an average of
60 percent carbon.  We calculate approximate DOM values by
dividing TOC values by 0.6.  Our raw water feed thus contained
approximately 17 mg/L of DOM.  The DOM value is helpful when
discussing competitive adsorption.

                               25

-------
                           SECTION VI

                        EXPERIMENTAL PLAN

    The two-year study contained four Experimental Designs  (ED).
These designs are listed below in Table 3 with their starting
and ending dates.  During the conduct of the experimental
designs for the bench scale studies, the samples that were taken
described the influent water to the bench scale experiments are
also designed to describe the operation of the full scale
Preston plant.  Thus a long term comparison of the raw water,
Hydrotreater effluent and finished water for the full scale
treatment plant was a designed phase of the project and desig-
nated as Full Scale Plant Studies.  The full scale plant study
data can also be used with each bench scale study conducted.
For example, during EDI the samples on Figure 7 designated as
1, 2, 7, 8, 11 and 12 describe both the influent to the bench
scale columns as well as the full scale treatment plant water at
the locations noted.

             TABLE 3.  EXPERIMENTAL DESIGN NUMBER
                 AND STARTING AND ENDING DATES
             ED
        Dates
             1

             1R

             2

             3

             4
Aug. 13 - Dec. 7, 1976

Jan. 18 - May 20, 1977

Jun. 3  - Aug. 5, 1977

Aug. 26 - Oct. 18, 1977

Nov. 1, 1977 - Mar. 3, 1978
    The general purpose for each ED was as follows:
EDI
    The flow diagram for EDI is shown in Figure 7.  Two
                               26

-------
NJ
                                                PRESTON PLANT
                               Hydrotreator
            Raw Water from Wells
en f
M  I I

§ « 3
             . G-
             •A-
             .c
'.E
'.3.
>
•V.
        J  I  Sample Point and Number

                                                         2
                                                   I'
                                                     Chlorine
                                                     Contact
                                                      Basin
                                            Bench Scple Adsorption Unit

                                                     8
. E

'."3 •'
 *
  *
                                                     Sand
                                                    Filter
                                                 Clear
                                                 Well
          Figure 7.   Flow diagram of Bench Scale Adsorption  Unit  for  GAG and XE-340  study
                        in EDI.

-------
adsorbents were studied.  Filtrasorb 400, 12 x 40 mesh obtained
from Calgon Corporation, Pittsburgh, Pennsylvania, was chosen  as
the GAC  adsorbent.  The second adsorbent was Ambersorb XE-340
from Rohm and Haas Company, Philadelphia, Pennsylvania.  Amber-
sorb XE-340 is a polymeric carbonaceous adsorbent tailored  for
removal  of low molecular weight organics from water.

     As  shown in Figure 7, both adsorbents were placed in the  raw
water  line.  This enabled a comparison of their abilities to re-
move TOC, HOC and precursor substances from raw water.  Pre-
cursor removal was measured by the THM FP method.  Ambersorb
XE-340 columns were also placed in H.T. and finished water  lines
to study HOC, TOC, and THM FP removal with the respective influ-
ent waters.

ED1R
     As  EDI work progressed, changes in methodology were made  and
the desired complete data base from initial start-up was not
obtained.  This work was thus considered mainly as a shake-down
phase  and EDI was repeated as originally planned.

ED2
     The Flow Diagram for ED2 is shown in Figure 8.  At the end
of ED1R, the partially exhausted XE-340 column on the finished
water  line was selected for a leaching study because this type
of data  had not been previously collected.  As shown in Figure 8,
a fresh  XE-340 column was placed on the finished water line
ahead  of the partially exhausted column.  For a period of time,
essentially all halogenated organics would be removed by the
fresh  column.  The halogenated organic leaching rate for the
second column was determined by analyzing the effluent sample
13A.

ED3
     The Flow Diagram for ED3 is shown in Figure 9.  Rohm and
Haas indicated that IRA-904 resin, avstrong base cationic ad-
sroption resin, was one of the better polymeric adsorbers for
precursor type substances.  We did not select it for halogenated
organic  adsorption.  To study the effect of bed depth, two  IRA-
904 resin columns in series were placed on the raw water line.
One IRA-904 resin column was placed on the H.T. line.  To com-
pare the effectiveness of IRA-904 resin with GAC Filtrasorb 400,
one column of each was placed on the finished water line.

ED4
     The Flow Diagram for ED4 is shown in Figure 10.  Four  GAC
Filtrasorb 400 columns in series were placed in the finished
water  line to study the effect of bed depth and contact time on
halogenated organic and precursor removal.
                                28

-------
                                                     PRESTON PLANT
                      Raw Water
                      from Wells
                                  Hydro-
                                  treator
o
i
Chlorine
Contact
 Basin
 Sand

Filter
Clear

Well
N)
U>
                                             Bench Scale Adsorption Unit
                                                                  In —T-
                                                                  H  19
                                                                  -5) *O 3
                                                                  a S3
                                                                    O r-l
                                                                    It r-t
                                         •f-

                                         :«-:
                                         !o-:
                                        .E'
                                        •;-4
                                                                                        Partially
                                                                                        Exhausted Column
                                                                                        9n Clear Well Water
                                                                                        (Experimental
                                                                                        Design No.1-Repeat)
                                     ["] Sample Points and Numbers
                      Figure  8.   Flow diagram of  Bench Scale Adsorption Unit for
                                    leaching study in ED2.

-------
                                        PRESTON PLANT
CO
O
Raw Water from Wells
y-
'


H

rr

t






_ 1 1,^.
1 r







rn
^•M!

*.9-."

,
."
.
•"
0.'
4;;
. .

."•;






m - - V-

. I. > 1 '
• i


— y


i







'G.'
'.A'.'
•t;' '

* ."
" * '."






                                                                                s
         I   I Sample
Point and Number
        Figure 9.   Flow diagram of Bench  Scale Adsorption Unit  for GAG and IRA-904  resin
                    study in EDS.

-------
                                      PRESTON PLANT
      Hydro-
Raw Water tre
from Wells /"""



j 1 j


^

PI
| 2 j 1 7 1


ator
>

n
|_8 J


Chlorine

















[ll+ll



Clear
Well

Benol1


(Uj
m—f~*
IT3
1—3
1
••
a T.
I "8 1





Scale Adsorption Unit
EH


*'.'•
C.






00









1

E

..'A'.
'•- 9-
.* .* "-"
L


0 EE3
i

J£

'•*'.'
i


D I241
i

JT

.' "G".
"*•


OIE
1 23+21 [25+21
                 Q Sample Point and Number

Figure  10.  Flow diagram of Bench Scale Adsorption Unit for* deep bed  study in
             ED4.

-------
                          SECTION VII

                    RESULTS AND DISCUSSION

    Section VII, Results and Discussion, is divided into two
parts; Full Scale Plant Studies and Bench Scale Studies.  Full
Scale Plant Studies refers to data obtained at sample points
from the Preston Water Treatment Plant during the conduct of
each bench scale experiment.  These Full Scale Plant sample
points include the raw water feed to the plant, effluent from
the lime softening unit  (H.T.), and finished water from the
clear well.  Bench Scale Studies refers to data obtained at
sample points from the Bench Scale Column Adsorption Unit.

FULL SCALE PLANT STUDIES

Specific Halogenated Organics
                                                               >

    Nineteen specific HOC were studied.  The specific compounds
with their chemical identification number for this report are
given in Table 4.  They are presented in their order of elution
from the gas chromatograph using a Tenax GC column (Figure 6).
The average concentration in the full scale raw, H.T. and fin-
ished water during the conduct of each bench scale ED are shown
in Tables 5 and 6 respectively.

Raw Water Source—
    Vinyl chloride levels were 0.8, 6.9, and 12.8 yg/L respec-
tively in ED1R, EDS, and ED4.  This is an insufficient data base
to indicate that vinyl chloride levels might be increasing, but
the possibility should be studied further.

    The level of methylene chloride also increased, 0.08, 0.1,
and 0.45 yg/L.  The level of trans 1,2-dichloroethene varied
erratically from 1.3 to 2.0 yg/L.  1,1-Dichloroethane varied
from 0.3 to 0.6 yg/L.  The compound cis 1,2-dichloroethene was
the highest level HOC in raw water and the four ED varied in no
set pattern from 21.0 to 29.0 yg/L.  The four THM present,
chloroform, bromodichloromethane, chlorodibromoethane, and
bromoform essentially averaged nil concentration in all ED
except ED1R which showed levels of 0.16, 0.11, 0.04, and 0.02
yg/L respectively.   The summed value of 1,1,1-trichloroethane,
1,2-dichloroethane and carbon tetrachloride varied from 0.1 to
0.2 yg/L in ED1R,  EDS and ED4.  Trichloroethylene varied from
0.13 to 0.4 yg/L.   Chlorobenzene varied from 0.19 to 1.3 yg/L.


                               32

-------
TABLE 4).   AVERAGE CONCENTRATION  OF SPECIFIC  HALOGENATED
           ORGANICS IN  RAW WATER
                                   Average concentration in ]_ig/L
Chem.
I.D.#
20
1
3
4
5
6
. 7
8
9
10
11
12
13
14
15
16
: 17
18
19
Chemical Name
Vinyl Chloride
Methylene Chloride
Trans 1 , 2-Dichloroethene
1 , 1-Dichloroethane
Cis 1 , 2-Dichloroethene
Chloroform
1,1,1-Trichloroethane \
1 , 2-Dichloroethane )
Carbon Tetrachloride /
Trichloroethylene
Bromodichlorome thane
Tetrachloroethylene
Chlorodibromomethane
Chlorobenzene
Bromoform
. p-chlorotoluene
m-dichlorobenzene
p-dichlorobenzene
o-dichlorobenzene-
Raw Water
EDI




21
N




N

N

N




ED1R
.8
.08
1.3
.3
29
.16

.11

.13
.11
.06
.04
.19
.02
.11
\
-
/
EDS
6.9
.1
2
.6
26.9
N'

.2

.4
N
N
N
1.3
N
.2
\
)l.O
/
ED4
12.8
.45
1.5
.58
25.6
.08

.1

.34
N
.003
N
1.1
N
.02

.51
.17
                          Total

          N = Nil

         ND = Not Determined
33.51
39.6   43.25
                              33

-------
TABLE 5.   AVERAGE CONCENTRATION OF  SPECIFIC HALOGENATED
           ORGANICS IN H.T.  WATER
                                   Average concentration in yg/L
Chem.
I.D.# Chemical Name
20
1
3
4
5
6
7
8
9
10
11
12
13
14
15
16
; 17
18
19 .
Vinyl Chloride
Methylene Chloride
Trans 1 , 2-Dichloroethene
1 , 1-Dichloroe thane
Cis 1 , 2-Dichloroethene
Chloroform
1,1/1-Trichloroethane \
1, 2-Dichloroethane /
Carbon Tetrachloride /
Trichloroethylene
Bromodichlorome thane
Tetrachloroethylene
Chlorodibromomethane
Chlorobenzene
Bromoform
p-chlorotoluene
m-dichlorobenzene
p-dichlorobenzene
o-dichlorobenzene
H.T. Water
EDI




20
4




1.7

.62

.09




ED1R
.7
ND
.4
.13
25.4
1.1

.2

.07
.26
.003
.24
.03
N
.03
\
).39
/
ED3
6
ND
1.8
.89
24.1
1.43

.09

.44
.6
N
.46
.84
.013
.16
\
) .56
/
ED4
9.7
ND
.95
.45
22.3
1.2

.12

.33
.35
.004
.13
.72
.007
.03
N
.28
.16
                        Total
28.95
28.75   37.38   36.73
          N = Nil

         ND = Not Determined
                             34

-------
TABLE  § .  AVERAGE  CONCENTRATION OF SPECIFIC HALOGENATED
           ORGANICS IN FINISHED WATER
                                Average concentration  pg/L
Chem.
I.D.#
20
1
3
4
5
6
7
8
9
10
11
12
13
14
15
16
; 17
18
19
Chemical Name
Vinyl Chloride
Methylene Chloride
Trans 1 , 2-Dichloroethene
1 , 1-Dichloroethane
Cis 1 , 2-Dichloroethene
Chloroform
1,1, 1-Trichloroethane \
1 , 2-Dichloroethane )
Carbon Tetrachloride /
Trichloroethylene
Bromodichlorome thane
Tetrachloroethylene
Chlorodibromomethane
Chlorobenzene
Bromoform
p-chlorotoluene
m-dichlorobenzene
p-dichlorobenzene
o-dichlorobenzene
Finished Water
EDI




10.9
80.2




37.1

12

.13




ED1R
.6
ND
.18
.2
17.2
71.4

.66

.2
42.7
.02
24.5
.1
1.9
N
\
) .63

ED2
.6
ND
.86
.18
18.4
64

1.47

.57
42.4
N
26.7
.08
1.91
N
\
)"
/
ED3
5.4
ND
1
.3
18.3
57

5.3

.1
39
N
27
.8
2.5
.2
\
•'
/
ED4
6.2
ND
.77
.4
19.9
67.3

7.7

.68
47
.003
33.6
.86
2.5
.1
N
..21
.14
                        Total

        N = Nil

       ND = Not Determined
160.29  159.27   157.2  187.36
                                 35

-------
p-Chlorotoluene varied from 0.02 to 0.2 yg/L.  The summed value
of m, p, and o-dichlorobenzene varied from 1.0 to 1.1 yg/L  in
ED1R and ED3.  In ED4, the three isomers of dichlorobenzene were
reported separately with values of nil, 0.51 and 0.17 yg/L
respectively.

    Thus, in general, the raw water contaminants were fairly
consistent during the project.
                     v
Hydrotreator Effluent Source—
    Three main factors in the lime softening stage of the plant
probably contribute to changed levels of the specific HOC origi-
nally present in raw water.  These factors are, volatile loss,
adsorption on precipitated calcium carbonate (most of which is
removed as sludge) and THM generation by a small amount of
chlorine which is added before the lime to activate the sodium
silicate used as a coagulating aid.

    The percent removal or increase factor (the symbol "X"  used
as "times") based on raw water for the 19 specific HOC are  shown
in Table 7.  Vinyl chloride was reduced by 12, 13, and 24 per-
cent in ED1R, EDS, and ED4 respectively.  Methylene chloride was
not determined on H.T. or finished water since it was used  as an
internal standard for each GC determination.   The average con-
centration of trans 1,2-dichloroethene was reduced by 69, 10,
and 37 percent.  The level of 1,1-dichloroethane was reduced by
57 and 22 percent in ED1R and ED4 respectively.  In ED3, an
increase factor of 1.5X was observed for 1,1-dichloroethane.
This factor is determined by dividing the average concentration
of the compound in the H.T. effluent water by the average con-
centration in raw water.  There were 16 data points in EDS  and
only 5 showed increase factors.  Unless most of the data points
show a consistent increase factor we should probably not put too
much weight on the increase factor as determined.

    The H.T. reduced levels of cis 1-2-dichloroethene by 5, 12,
10, and 13 percent in the four ED.  The four THM, chloroform,
bromodichloromethane, chlorodibromomethane and bromoform
increased from almost nil levels in raw water to average values
of 1.9, 0.7, 0.4, and 0.03 yg/L in the H.T. effluent water.  The
summed value of 1,1,1-trichloroethane, 1,2-dichloroethane and
carbon tetrachloride in EDI and ED4 show increase factors of
1.8X and 1.2X respectively.  The summed value was reduced 55
percent in ED3.   Again, the actual data points do not show  a
consistant increase and partial removal is the usual case.
Trichloroethylene, tetrachloroethylene and p-chlorotoluene
exhibit a similar pattern.  The H.T. reduced levels of chloro-
benzene by 84, 35, and 34 percent in ED1R, ED3, and ED4 respec-
tively.   The isomers of dichlorobenzene are reduced by 65,  44,
and 51 percent.
                               36

-------
TABLE  7.
PERCENT REMOVAL OR INCREASE FACTOR FOR
SPECIFIC HALOGENATED ORGANICS  IN H.T. WATER
                        Percent removal or increased
                            factor based on raw water
Chem.
I.D.t Chemical Name
20
-1
3
4
5
6
7
8
9
10
11
12
13
14
15
16
; 17
18
19
Vinyl Chloride
Methylene Chloride
Trans 1 , 2-Dichloroethene
1 , 1-Dichloroe thane
Cis 1 , 2-Dichloroethene
Chloroform
1,1, 1-Tr ichloroethane \
1 , 2-Dichloroethane /
Carbon Tetrachloride /
Trichloroethylene
Bromodichloromethane
Tetrachloroethylene
Chlorodibromomethane
Chlorobenzene
Bromoform
p-chlorotoluene
m-dichlorobenzene
p-dichlorobenzene
o-«dichlorobenzene
H.T. Water
EDI




5














ED1R
12

69
57
12


1.8X

46

95

84

73
\
65
1
ED3
13

10
1.5X
10


55

1.1X

N

35

10
\
>44
/
ED4
24

37
22
13


1.2X

3

1.3X

34

1.5X
N
45
6
          N = Nil
         ND = Not Determined

          X = Times Factor
                             37

-------
    Generally, the full scale plant H.T. process did not  achieve
significant 'reductions in the concentrations of the specific
organics routinely monitored.

Finished Water Source—
    Four main factors probably contribute to changed levels of
HOC in the finished water of the full scale treatment plant.
Volatile loss, removal of calcium carbonate (turbidity) by sand
filtration which may contain adsorbed HOC, and oxidation  of HOC
by chlorine contribute to the overall HOC reduction.  Break-
point chlorination will greatly increase THM levels.

    In Table 8 the reduction of vinyl chloride in finished
water, based on raw water levels, is 25, 22, and 52 percent in
ED1R, ED3, and ED4 respectively.  The reduction in vinyl
chloride is about the same in the H.T. portion of the plant and
the breakpoint chlorination—chlorine contact basin—sand
filtration stage of the plant.  A pattern of further reduction
in finished water compared to reduction in H.T. effluent  was
observed with trans 1,2-dichloroethene, 1,1-dichloroethane,
cis 1,2-dichloroethene, tetrachloroethylene, chlorobenzene,
p-chlorotoluene and the isomers of dichlorobenzene.

    In Table 8 increase factors of 1.5X and 2X are shown  for
trichloroethylene in ED1R and ED4.  The individual data points
in both these ED show quite a consistent pattern of increase
suggesting that this compound may indeed be increasing in
finished water.  The data in Table 8 for chlorobenzene suggest
that in ED1R and ED4, less of the compound is removed on  a
percentage basis from finished water than H.T. water based on
the original amount present in raw water.  An explanation might
be that chlorobenzene is actually increasing between the  H.T.
and finished water stages, but not enough to indicate an  over-
all increase factor.  Of the non-THM HOC compounds it appears
that one or more of the summed group consisting of 1,1,1-tri-
chloroethane, 1,2-dichloroethane and carbon tetrachloride
increases in the finished water.  The actual individual data
points clearly show that the finished water concentrations were
higher than raw and H.T. water in most of the samples in  each
design phase.  Trichloroethylene and chlorobenzene may also show
some increase.   These increases may be formed by the reaction of
chlorine with precursors or may be introduced with chlorine, or
both.

    The increase in inst.  THM's through the treatment plant
(raw water versus finished water)  are clearly shown by comparing
the chloroform data in Table 4 with the chloroform data in
Table 6.   The average summation of the concentrations of  the
four inst.  THM species are shown for each ED in Table 9 (i.e.
129.4,  etc.).  Also the average percent of the total inst. THM's
are shown for each species (i.e. chloroform was 62 percent of
the 129.4 yg/L concentration for total inst. THM's for EDI).


                               38

-------
TABLE; 8.     PERCENT REMOVAL OR INCREASE FACTOR FOR SPECIFIC
             HALOGENATED  ORGANICS IN FINISHED  WATER
    Chem.
    I.D.#
                                     Percent removal or increase factor
                                           based on raw water
  Chemical Name
                                             Finished Water
                           EDI
ED1R
ED3
ED4
     20
Vinyl Chloride
                                              25
        22
                                                 52
            Methylene Chloride
            Trans 1,2-Dichloroethene
                                  86
        50
        49
            1,1-Dichloroethane
                                  33
        50
        31
            Cis 1,2-Dichloroethene
                           48
41
29
            Chloroform
            1,1/1-Trichloroethane
                     \
            1,2-Dichloroethane
                      I
6X
26.5X
77X
            Carbon Tetrachloride
                     Z
     10
Trichloroethylene
1.5X
75
 2X
     11
Bromodichloromethane
     12
     13
Tetrachloroeth;
Chlorodibromomethane
            m-dichlorobenzene
            p-dichlorobenzene
            o-dxchlorobenzene
            N « Nil
           ND = Not Determined
            X = Times Factor
                                 39

-------
      TABLE 9.    AVERAGE TOTAL INST.  THM AND PERCENT OF
                 INDIVIDUAL THM IN EACH EXPERIMENTAL DESIGN
                          EDI     ED1R     ED2     ED3     ED4


 Total Inst.  THM (yg/L)   129.4    140.5    135.0   125.5   150.4

 TOC (mg/L)                         9.8              8.6     8.3

 Percent of  Individual
     Inst. THM

 chloroform                62       50.8     47.4    45      44.8
 bromodichloromethane     28.6     30.4     31.4    31   .   31.2
 chlorodibromomethane      9.3     17.4     19.8    22      22.3
 bromoform                 0.1      1.4      1.4     2       1.7
      In Table  9 the data show that the average total inst. THM
 varied from  125.5 yg/L to 150.4 yg/L.  TOC values in mg/L for
 ED1R, ED3, and ED4 are also shown in Table 9 and they do not
 correlate  with the average total inst. THM values.  There
 appears to be  a consistent trend in the data for the ratio of
 bromine compounds to increase from EDl through ED4.  The percent
 of  individual  inst. THM data indicates that there was a shift in
 the  composition of the total inst. THM's, whereas chloroform
 comprised 62 percent of the total inst. THM's during EDl, it
 comprised only 44.8 percent during ED4.  Other species increased
 accordingly.   Although the reason is unknown, the possibility of
 slight salt water intrusion could exist.


 TOC  and THM FP Organics

 Raw  Water Source—
     Average THM FP and TOC levels of raw water for EDl, ED1R,
 ED3  and ED4 are shown in Table 10.  There appears to be no
 direct relationship between the concentrations of TOC and THM
 FP.  Comparisons of data in Table 10 show that the highest
 average concentration of TOC was 9.8 mg/L in ED1R with a
 corresponding  average THM FP concentration of 659 yg/L and that
 the  lower average TOC concentration of 8.3 mg/L in ED4 was not
 accompanied by a corresponding lower THM FP concentration.

H.T. Water Source—
     Average THM FP levels of H.T. water for EDl, EDlR, ED3 and
ED4 are shown  in the upper half of Table 10.  The percent of THM
FP removed from raw water by lime softening is also shown.
                               40

-------
TABLE 10.  TOO AND THM FP REMOVAL BY
           LIME SOFTENING IN FULL SCALE
           PLANT

ED
1
1R
3
4
ED
1
1R
3
4
Ave. THM FP
in raw water
>ug/L
816
659
591
662
Ave. TOC in
raw water
mg/L
-
9.8
8.6
8.3
Ave. THM FP
in H.T. water ~
yg/L
573
471
389
531
Ave. TOC in
H.T. water
mg/L
-
6.8
6.0
5.8
Ave . Percent
Removal
%
30
28
34
20
Ave. Percent
Removal
%
-
31
30
31
                   41

-------
 In  404 days of testing over the two-year study, the weighed
 average removed by lime softening was 27 percent.

     Average TOC levels of H.T. water for ED1R, ED3 and ED4
 appear in the lower half of Table 10 with percent removal  data
 from raw water by lime softening.  The weighed average removal
 was 31 percent.  TOC and THM FP removals by lime softening
 appear to correlate quite well with values of 31 percent and
 27  percent respectively.

     Raw water entering the Preston Plant had an average total
 hardness of 245 ppm.  Lime softening reduced the hardness  to
 about 85 ppm.  Non-carbonate hardness averaged 6 ppm.  A
 decrease in carbonate hardness of 154 ppm is equal to 308  mg
 of  calcium carbonate floe per liter.  In all four ED the average
 THM FP removed fro'm raw water by lime softening was 205 yg/L.
 This corresponds to 0.07 gram  of THM FP adsorbed per 100  grams
 of  calcium carbonate floe.

 Finished water Source—
     TOC and THM FP removal data resulting from the combined
 effects of breakpoint chlorination, residence in the chlorine
 contact basin and sand filtration are shown in Table 11.   TOC in
 finished water (lower half of Table 11) is removed an average of
 6 percent and the average THM FP removal is approximately  24
 percent.  However, the THM FP removal was in actuality simply a
 conversion of a part of the THM FP in the H.T. water to a  com-
 bination of actual THM plus remaining THM FP (the sum of inst.
 THM and THM FP is terminal THM) in finished water.  A comparison
 of  the terminal THM values shown in parentheses in Table 11
 shows this result.  For example, in Table 11 the terminal  THM
 concentration for the H.T. effluent during EDI was 586 yg/L and
 the terminal THM concentration for the finished water was  580
 yg/L.  No practical difference exists between these two average
 valuers.  Thus the H.T. water contained a THM FP concentration of
 573'yg/L plus an inst. THM concentration of 13 yg/L while  the
 finished water contained a THM FP concentration of 448 yg/L and
 an  inst. THM concentration of 132 yg/L.  Thus, whereas the com-
 bination of THM FP and inst. THM concentrations for the two
 locations were about equal, the THM's in the finished water
 increased by about 120 yg/L while the THM FP decreased by  about
 the same amount.   Thus the chlorination—contact basin—sand
 filtration step really achieved no removal of precursor but
merely a conversion.   The results are somewhat in line with the
 low TOC removal.

     TOC data for raw, H.T. and finished water are plotted for
ED1R,  ED3 and ED4 in Figures 11, 12, and 13.  THM FP data  for
raw and H.T.  water for ED4 are plotted in Figure 14.  These
plots  are presented at this point, mainly to show the variation
in values of these parameters from sample date to sample date.
Similar plots for the other ED appears later in the report with

                                42

-------
         TABLE  11.   TOG,  TERMINAL THM AND  THM FP
                     REDUCTION BY CHLORINATION,  CONTACT
                     BASIN AND/OR SAND FILTRATION
^HVBHHBIWBB^MB4^HIHHMHallH
ED
1
1R
3
4
ED
1
1R
3
4
^••••^••••'••••••••••••i™ m ^•.^•••i i !• 	 —•.man I 	 ii.iin 	
Ave. THM FP
in H.T. water
yg/L
573 (586)*
471 (476)*
389 (397)*
531 (533)*
Ave . TOC
in H.T. water
mg/L
-
6.8
6.0
5.8
Ave. THM FP
in finished water
yg/L
448 (580)*
349 (495)*
274 (400)*
434 (584)*
Ave. TOC
in finished water
mg/L
-
6.1
5.9
5.4
HMHH^^HMm^MklMMmMMIIMIIIIIIiaHmHVBVBHII^HM^^B
Percent
Removal
%
22 (0)**
26 (0)**
30 (0)**
18 (0)**
Percent
Removal
%
-
10
2
6
 * Terminal THM figures in parentheses

** Percent removal based on Terminal THM values
                                43

-------
   11-
   10-!
    8—
 n
 
-------
en
        13-.

        12-

        11-

        10-

          9-
       5  8-1
7-
       u
       o
5 -

4 -
  v
  v
1 -
     Days
          0
                                                       Av.  8.6
                                               30% reduction
                                               32% reduction
                                                       0   Raw water
                                                      —D— Hydrotreator water
                                                     — A— Finished water
   04    11    18   25   32    39   46   53
          Figure 12. TOC in raw, Hydrotreator and finished water  (EDS).

-------
                                                                               30%  reduction
                                                                                    O
                                                                                     SAv.  5.8
                                                                               35%  reduction
                    Raw water
                O— H.T. water
                    Finished water
Days 0
                                   59 63  70   77    84   91
Figure .13.  TOG in raw, Hydrotreator and finished water  (ED4).

-------
 900
 800
                    Raw water
               0—  Hydrotreator  water
                                                                                                20% reduction
Days
03   7  10
                    1721 2k 28  3135  38  421+5 4952 5659  6366  70  77  8084  8791 9498  101
              Figure 14.  THM FP in raw water and  removal by lime softening (ED4) .

-------
 adsorbent plots.


 Other Parameters
                        I
      As an aid in interpreting data for each ED, plant profile
 information such as color, pH and turbidity of raw, H.T.
 effluent, and finished water, as well as free and combined
 chlorine for H.T. effluent and finished water were collected
 and the raw data is available in Appendix B.  Rainfall data and
 chlorine levels of raw water are also available in Appendix B.

 Rainfall and Chlorides—
      The chloride concentration in raw water generally fe-llows
 a cycle in response to the seasonal rainfall.  Chloride levels
 rise when rainfall is low and decrease when rainfall is high.
 In southern Florida the wet season extends from May through
 October with the heaviest rains usually occurring in May.  The
 dry season occurs between November and April.  March and April
 are usually very dry.  Chlorides reach a low concentration in
 September when rainfall is heavy and remain low until about
 February as the ground water level subsides, then gradually
 climb to a maximum in June.

 Rainfall and TOC—
      TOC levels in raw water may also be influenced by rainfall.
 TOC data were collected only in ED1R, ED3, and ED4.  The long-
 est period of least rainfall in the two-year study period
 occurred during most of ED1R which averaged the highest TOC
 level of 9.8 mg/L.  EDS and ED4 with more rainfall averaged 8.6
 and 8.3 mg/L.

 pH~
      The pH of raw water remains constant at 7.2 ± .01 through-
 out the year.  The average values of the pH in H.T. effluent
 and finished water, and the level of free chlorine in finished
 water is shown in Table 12.

 TABLE 12 .   AVERAGE pH AND FREE CHLORINE (PPM)  VALUES IN EACH
	EXPERIMENTAL DESIGN	

                   EDI


                H.T.   Fin.   H.T.   Fin.   H.T. Fin.  H.T. Fin.
PH

free chlorine
   (ppm)
9.80
9.20 9.92 9.16
2.49
9.80 9.06 9.90
1.91
9.11
2.06
                                48

-------
     There appears to be no apparent  relationship between Total
THM values in Table 9 with THM FP values  in Table 10, nor with
rainfall, chlorides, TOG, pH  and chlorine data.  Perhaps these
two parameters which are controlled by the chemistry of the
reaction of free chlorine with precursors is influenced by
subtle changes in pH, time and temperature which are beyond the
scope of our data.

Turbidity—
     A record of H.T. effluent turbidity  may be important
because with higher levels, additional amounts of precursors may
be carried over to the breakpoint chlorination step.  However,
questions about the turbidity data prevented assessing the
relationship between the turbidity and THM FP.  Tables of H.T.
turbidity results are included in Appendix A.  In EDI and ED1R
the turbidity of the H.T. effluent fluctuated widely from day
to day.  In ED3 and ED4 turbidity levels  appear more uniform.
The average turbidity during  EDI and  ED1R was 10.2 NTU and 9.8
NTU respectively.  In EDS and ED4 it  was  3.4 and 4.6 respec-
tively.  EDS and ED4 data may be misleading because during these
last two phases sampling was  done only when the organic samples
were taken, whereas it was done daily during EDI and ED4.  A
check of the operators' daily turbidity records during ED4
showed a high of 25 NTUs, and an average  of 9.3 NTUs.  These
values are more comparable to values  reported for the first
phases.  Thus, it is possible that the H.T. effluent turbidity
did not change substantially  during the project.

     Turbidity increased in the distribution system.  Values
increased from an average of  0.32 NTU in  finished water to an
average of 1.1 NTU in the distribution system sample.

Color—
     Lime softening removed an average of 55 percent of the
color from raw water.  Another 10 PCU is  removed by chlorination
and sand filtration.
                               49

-------
BENCH SCALE STUDIES

      This portion of the report presents data on the effects  of
the  three adsorbents evaluated on removal of specific HOC  and
other organics as measured by THM FP and TOG in raw, H.T.
effluent and finished water.  The pilot column configuration for
each ED was previously presented in Section VI.

Specific Halogenated Organics

Raw  Water Source —
      The effect's of adsorbents on raw water were studied in
EDI, ED1R, and EDS.  Although the levels of the four THM were  nil
or too low to be evaluated, the adsorption results for the
•specific compounds discussed below show that the XE-340 was more
efficient than GAG..  IRA-904 resin removal, as expected, was
poor and removal data for the raw source is only presented for
cis  1,2-dichloroethene and vinyl chloride to show typical
results with IRA-904 resin.  Appendix A contains additional raw
data tables for all substances if further data are required.

      cis 1,2-Dichloroethene — The HOC occurring in highest con-
centration in raw water was cis 1,2-dichloroethene.  It will be
discussed first.  Its general pattern will aid in interpreting
the  data from some of the substances present in low concentra-
tions.  Adsorption data appears in Table 13.

      The adsorption data in Table 13 were obtained by integra-
ting the actual breakthrough curves which appear in Figures 15,
16,  17, and 18.  The breakthrough point (B) and saturation point
 (S)  are shown on the curves.  Using Table 13, a comparison of
the  effectiveness of GAC versus XE-340 can be made on an equal
volume and equal weight basis at column saturation.  At equal
volumes of adsorbent, XE-340 had 3.8 times and 3.4 times the
adsorptive capacity of GAC in EDI and ED1R respectively.   At
equal weights of adsorbent, XE-340 had 3.2 times and 2.8 times
the  capacity of GAC.  Column breakthrough on GAC occurred  at 21
and  16 days and column saturation at 69 and 73 days respectively.
Breakthrough occurred at 61 and 58 days for XE-340.  Extra-
polated column saturation values of 280 and 242 days were
obtained for XE-340.

      The MTZ for XE-340, 24 and 23 inches, is slightly more
than for GAC, 21 and 23 inches.  It is apparent that GAC
(Figure 15)  and XE-340 (Figure 17) , both allow low level passage
of cis 1,2-dichloroethene long before the value we have recorded
as the breakthrough point.  The actual value in yg/L, which are
too low to plot on the "Y" axis scale, appear above the data
point.   No number above a data point means nil concentration.
Consideration of this low level passage as breakthrough would,
of course, greatly change the recorded MT_ values.
                               50

-------
TABLE 13.  cis 1,2-DICHLOROETHENE ADSORPTION DATA  FROM RAW WATER










ED
1

1

1R
1R

3
3




•P -P



^


fl
»
•0

Feet
2.5

2.5

2.5
2.5

2.5
5










•P
Adsorbe


GAG
XE-

340
GAC
XE-
340
904
904










.p
Average
Influen

ng/i
21

21

29
29

26.9
26.9


2.5
5



A
Di
3
0
H
Column
Break th

Days
21

61

16
58

no
no


feet
feet





e
o
•H
Column
Saturat

Days
69

280

73
242

adsorj
adsorj


= o.:
= l.E







MT
z

Inch
21

24

23
23

tion
tion


6 met
2 met






C!
Test
Duratio

Days
117

117

122
122

53
53


er
ers


Gi
•H
n a -P
0) G m
•P 3 o>
c -i e*
H O
CJ o*
id J3 -H
•P o H
O n) 3

Grams
.219

.219

.316
.316

.127
.127






•O c *J
0) B H)
.Q P 0)

o o
(fl U i -P
f xi id

Grams
.084

.212

.115
.285

0
0





u

w
K^
O rj
4J 0
T3 Id -H
01 -P
x> c 
%
65

61

61
62







C
X
l> 1| tji
0) O M
Di in  C
O U
VI O
-o o -P
•a; H m

Grams
.048

.099

.065
.135

0
0




fj
01
^i ^

-------
en
to
       90
       80
       70
       60
       50
     40
       30
       20
       10
     Dayr
                 Raw water

         — O—  Raw water thru 2.5 ft.  GAC (0.76 meter )


                 Raw water thru 2.5 ft.  XE-340 (0.76 meter )
                                                                                                         Av.21
             Figure  15.   cis  1,2-Dichloroethene in ITAW water and removal by p.70 netor

                          (2.5 feet)  of GAC and 0.76 meter   (2.5 feet) of ::E-340  (EDI).

-------
                   Raw water
                )__Raw water  thru 2.5  feet  GAC  (0.76  meter )
      40
en
                                                                                     94  98
105
03   7 10 14  17 2124  2831  35 38 4245 4952  56    63 6670 73 77    84
Figure 16 . cis 1,2-Oichloroethene  in itaw water and removal by 0.76 meter   (2.5 feet)
           of GAC  (ED1R).
112
                                                                                                           122

-------
              Raw water
              Raw water thru 2.5 feet XE-340  (0.76 meter )
Days
94 98
105   112
122
           Figure 17. cis  1/2-Dichloroethene in raw water and removal by 0.76 meter  (2.5 feet)
                       Of XE-340  (ED1R).

-------
en
ui
                                                             Av.  26.9
                       Raw water  thru 2.5 feet (01:76 meter ) IRA-904

                       Raw water  thru 5  feet (1.52 meters) IRA-904
                7  11 Ik 18 21  25 28  32 35 39 42 46 49 53

               Figure 18.  cis 1,2-D.ichloroethene in »aw water  and removal by   0.76 meter

                           (2.5  feet)  and 1.52 meters (5 feet) of  IRA-904 resin.

-------
      In ED3,  0.76  (2.5 feet) and 1.52  (5.0 feet) meters  of  the
 IRA-904 resin adsorbed no cis 1,2-dichloroethene.  Discussion of
 the other HOC will follow their order of presentation  in Table  8.

      Vinyl chloride— Analysis for vinyl chloride began  on  Test
 Day 94 of ED1R after modifications were made to the present
 equipment.  Since vinyl chloride analysis did not begin  until
 toward the end of the test period, we do not have breakthrough,
 saturation or MTZ information.  The adsorption data obtained are
 plotted in Figure 19.  From Test Day 94 to 122, the average level
 of vinyl chloride in the raw water was 0.80 yg/L and the average
 level through GAC and XE-340 was 0.72 yg/L and 0.77 yg/L respec-
 tively.  Therefore, if the differences in the average  are con-
 sidered significant, from day 94 to 122 there was 10 percent and
 4 percent removal respectively.

      Results  on 0.76 (2.5 feet) and 1.52 (5.0 feet) meters  of
 IRA-904 resin on raw water appear in Figure 20.  Throughout the
 entire two-year study, IRA-904 resin did not adsorb other HOC
 from  raw, H.T. or finished water.  Therefore we read the
 individual curves and averages in Figure 20 as indicating no
 removal of vinyl chloride.

      trans 1,2-Dichloroethene—The results of adsorbents for
 removal of trans 1,2-dichloroethene from raw water are shown in
 Table 14 and  the breakthrough curves in Figures 21 and 22.

      The HOC, trans 1,2-dichloroethene did not break through the
 XE-340 column (Figure 22) during the 122-day test period, there-
 fore, we cannot compare GAC and XE-340 at column saturation.
 However, it is obvious that XE-340 has greater adsorptive capac-
 ity for trans 1,2-dichloroethene than GAC, both on an equal
 volume or equal weight basis.  The GAC column allowed  low level
 passage (Figure 21) long before the time designated as break-
 through.  At  the end of the test period, the GAC column  had
 adsorbed 82 percent of the entering trans 1,2-dichloroethene and
 73 percent at extrapolated saturation.  XE-340 had adsorbed 100
 percent at the end of the test period.

      1,1-Dichloroethane—Removal results for 1,1-dichloroethane
 by adsorbents from raw water appear in Table 15 and Figure  23
 respectively.

     Table 15 data shows that breakthrough occurred at 21 days
 and 94 days for GAC and XE-340 respectively.  Figure 23  shows
 that the GAC reached saturation at 94 days and that saturation
 did not occur in the XE-340 column.  However, again from the
Table 15 and Figure 23 data it is obvious that XE-340, both on
an equal volume and equal weight basis has greater adsorptive
capacity for 1,1-dichloroethane than GAC.
                               56

-------
      i.o-i
      VI
      a
Ol
-J
       .7-
       .6-
       .5 —
        Raw water


— O—  Raw water  thru 2.5 feet (0.76 meter)  GAC

•••O-"  Raw water  thru 2.5 feet (0.76 meter)  XE-340
       Days
               Figure 19.
                                                                    94 98    105   112

             Vinyl chloride in raw water and removal by 0.76 meter  (2.5 feet)  GAC

             and 0.76 meter (2.5 feet) XE-340  (ED1R).
                                                                                                       122

-------
        20-
        15.
Ul
00
     0)
     4J
     -H
     tr>
10-
         5-
                           Raw water
            — O	  Raw water thru 2.5 feet IRA-904 resin (0.76 meter )

            	D	Raw water thru 5.0 feet IRA-904 resin (1.52 meters)
         0
    Days  0
             T
              7
T
11
                              18
                          21     25   28

Figure 20. Vinyl chloride in raw water and removal by 0.76 and  1.52  meters

            (2.5 and  5  feet)  of IRA-904 resin (EDS).

-------
           TABLE 14.   trans 1,2-DICHLOROETHENE ADSORPTION DATA FROM RAW WATER
10
ED
1R
1R








5
ft
S
TJ
0)
CO
Feet
2.5
2.5








Adsorbent
GAC
XE-
340








Jj Average
j£». Influent
1.3
1.3




2.5



PI Column
to Breakthrough
66
none




feet =



I? Column
*eJ Saturation
142





= 0.7(



MT
z
Inch
16





raete



S? Test
"n Duration
122
122




r



o Total Entering
g Each Column
» During Test
.0142
.0142








" Total Adsorbed
§ by Each Column
01 at End of Test
.0017
.0142








£5 Adsorbed by Each
1 Column at

-------
                                                                  4.9
o\
o
                  Raw water


                  Raw water thru 2.5  feet GAG (0.76 meter )
     Days
         03   7  10 Ik 17 21 2k  2831  35 38 k2 kS  49 52 56
122
           Figure 21.  trans  1,2-Dichloroethene in raw water and removal  by 0.76 meter (2.5 feet)

                       of GAG (ED1R) .

-------
                                                                   4.9
      4-
a\
                  Raw water thru  2.5  feet XE-340

                               (0.76 meter )
                                                                                                              Av.
 o 4*

   o
Days
              7  10 14 17 2124 2831  3538  k2 k5 %9 52  56    6366  70  73  77    84      94  98    105    112     122


               Figure 2.2..  trans 1,2-Dichloroethene in  raw water and removal by 0.76 meter   (2.5  feet)

                           6* XE-340  (ED1R).

-------
TABLE 15.  1,1-DICHLOROETHANE ADSORPTION DATA FROM RAW WATER
ED
1R
1R








5
1
•0
 Column
w Breakthrough
21
94




= 0.



a? Column
"M Saturation
94





76 me



MT
z
Inch
23





ter



% Test
"M Duration
122
122








o Total Entering
g Each Column
w During Test
.0037
.0037








•O C 4J
v § to
xi 3 «j
M rH EH
0 0
in u 
-------
(U
4J
•H
tn
3.
.5
 Days
                           • 0ii  Raw water

                           — O— Raw water thru 2.5 feet GAG (0.76 meter  )

                                  Raw water thru 2.5 feet XE-340 (0.76 meter  )
3  7 10 14 17  21 24  28 31  35 38  42 45 49 52  56

   Figure 23
                                                         63 66 7073  77
                                                                             94  98
                                                                                         105   112
                                                                                                   122
                          1,1-Dichloroethane  in raw water and removal by  0.76 meter   (2.5  feet)

                          of GAC  and"0.76 meter  (2.5 feet) of XE-340 (ED1R).

-------
      l,l/l-Trichloroethane/ 1,2-dichloroethane, carbon  tetra-
chloride—The removal results for the summed value of these
three HOC by adsorbents from raw water appear in Table  16  and
the breakthrough curves appear in Figures 24 and 25.

      With the low average influent concentration of 0.104  yg/L,
and the spread in individual data points, estimation of break-
through and saturation times is difficult.  Table 16, and  Figures
24 and 25 show the estimated time of breakthrough for GAC  and
XE-340 to be about 21 and 77 days respectively.  Reported  sat-
uration times are questionable.  However, since the breakthrough
times for XE-340 is greater than for GAC, we would expect  XE-340
to again have a greater adsorptive capacity than GAC.

      Trichloroethylene—Adsorption data appear in Table 17 and
breakthrough curves in Figures 26 and 27.

      Breakthrough occurred in 77 days and 96 days for the  GAC
and XE-340 respectively.  The saturation times cannot be extrap-
olated  because of insufficient data points after breakthrough
to establish the slope of the curve.  However, it is again
apparent that since XE-340 breakthrough occurred after  GAC break-
through, XE-340 will have a higher adsorptive capacity  than GAC.

      Tetrachloroethylene—Adsorption of tetrachloroethylene by
adsorbents from raw water was studied only in ED1R.  The influent
concentration to GAC and XE-340 columns was very low and erratic,
0.072 yg/L average for the first 31 days of the test and nil to
traces for the balance of the 122-day test.  Influent level and
adsorption curves are shown in Figure 28.  It is interesting to
note  that even at this low concentration, both adsorbents  do
adsorb a high percentage of the compound.  No other conclusions
are drawn.

      Chlorobenzene—Adsorption data appear in Table 18  and
adsorption  curves in Figure 29.  Influent concentration,  plot-
ted in Figure 29, was very erratic.   XE-340 removed all of the
compound for the entire test period.  GAC removed essentially
all of the compound except for the three test dates shown  in
Figure 29 when trace amounts passed.

      p-Chlorotoluene—p-Chlorotoluene was studied in ED1R.  The
erratic level of influent concentration is shown in Figure 30.
The average influent concentration was 0.38 yg/L.  Both GAC and
XE-340 removed all of the compound throughout the test  period.

      o,  ra and p-Dichlorobenzene—Adsorption from raw water by
adsorbents of the summed value of the three isomers of  dichloro-
benzene was studied in ED1R.  The influent concentration curve
appears in Figure 31.  The average concentration was 1.1 yg/L.
Both GAC and XE-340 removed all of the compounds throughout the
test period.

                               64

-------
                   TABLE  16.   1,1,1-TRICHLOROETHANE,  1,2-DICHLOROETHANE,
                    CARBON TETRACHLORIDE ADSORPTION DATA FROM RAW WATER
en
ED
1R
1R








S
8-
Q
•O
&
Feet
2.5
2.5








Adsorbent
GAG
XE-
340





2.5


Jj Average
t* Influent
104
104





feet


Hi Column
n Breakthrough
21
77





= 0.


g Column
"« Saturation
98
98?





'6 me1


MT
z
Inch
24
6?





.er


g Test
"M Duration
122
122








Q Total Entering
§ Each Column
w During Test
.00113
.00113








•o g p
0) | in
ja P i V
f J3 <0
Grams
.00055
.0008








n Adsorbed by Each
| Column at'
in Saturation
.00055
.0008








%Adsorbed at
End of Test
51
71








^ % Adsorbed at
Saturation
56
88








n Adsorption per
g 100 gms. Adsorbent
01 at End of Test
.0003
.00037








S?
| Adsorption per
100 gms. Adsorbent
.0003
.00037








at Saturation
o
o











-------
   .8 -
   .7 -
   .6 _
    Raw water

 •- Raw water thru 2.5 feet GAG
                    (0.76 meter )
Days
      0 3
7 10  14 17 212+  28  31  3538  4245 49 52  56    63 66  70 73  77    84      94  98    105   11
Figure  24,   1,1,1-Trichloroethane, 1,2-dichloroethane and carbon tetrachloride in
             raw water, and removal by 0.76 meter
                                                                                                    122
                                                               (2.5 feet)  of GAC (ED1R).

-------
         O	 Raw water thru  2.5 feet XE-340
                              (0.76  meter )
Days
       17 21 24 28 31  35 38  42 45  49  52  56   63 66 70 73  77    84       94  98    105
Figure 25.  1,1,1-Trichloroethane, 1,2-dichloroethane and carbon tetrachloride in
            raw water and removal by 0.76  meter   (2.5  feet)  of XE-340 (ED1R).
                                                                                                  122

-------
                 TABLE 17.  TRICHLOROETHYLENE  ADSORPTION DATA FROM RAW WATER
oo
ED
1R
1R








5
ft
&
•a
£
Feet
2.5
2.5








Adsorbent
GAC
XE-
340




2.5



5 Average
"t* Influent
.14
.14




feet



m Column
w Breakthrough 1
77
96




= 0.7



p? Column
"a Saturation






i met*



MT
Inch






r



g Test
"a Duration
122
122








o Total Entering
w Each Column
01 During Test
.00153
.00153








£ Total Adsorbed
| by Each Column
01 at End of Test
.00137
.00148









-------
vo
    Days
                 Raw water
            • O— Raw water thru 2.5 feet  GAG  (0.76 meter )
                                                           63 66 70  73  77
                                                                                       112
                                                                                                          122
0  3  7 10  14 17 21 24  28 31  35 38 42 45 49 52 56
 Figure  26.   Trichloroethylene  in raw water and removal  by 0.76  meter  (2.5 feet), of GAG (ED1R)

-------
                      Raw water


                •a---  Raw water thru  2.5 feet XE-340 (0.76 meter  )
-j
o
     CP

     3.
     Dav
-------
 .04 —
 .03-
M
s
•H
 .02-
  .01-
   0

 Days
       .3   .06   .1
       tlltllt
                   .05
                         Average level
                         first 31 days
.072
                                                         Raw water

                                                    O — Raw water thru 2.5 feet GAC  (0.76 meter )

                                                    Q.'.. Raw water thru 2.5 feet XE-340 (0.76 meter )
                                                                                          Av.

                                                                                          .023
0 '3  7 10 Ik 17 2121* 28 31 35 38 ^2  i*5  4952  56    63 66 70 73 77   &t     9^  98    105   112     122
 Figure  28. Tetrachloroethylene in  raw water and removal by -°-76 meter   (2.5 feet) of GAC
            and 0.76 meter  (2.5 feet)  of XE-340  (ED1R).

-------
                  TABLE 18.  CHLOROBENZENE ADSORPTION  DATA FROM RAW WATER
NJ
ED
1R
1R








5
Ot
&
•O
01
m
Feet
2.5
2.5






.

Adsorbent
GAC
XE-
340





2.1


S Average
£t Influent
.19
.19





feel


CD Column
co Breakthrough
none
none





= 0.


I? Column
"M Saturation







76 me


MT
Inch







ter


§? Test
"a Duration
122
122








O Total Entering
| Each Column
w During Test
.0021
.0021








£ Total Adsorbed
H by Each Column
w at End of Test
.0021
.0021








n Adsorbed by Each
g Column at
ca Saturation










% Adsorbed at
End of Test
100
100








^% Adsorbed at
Saturation










0 Adsorption per
r\
| 100 gms. Adsorbent
w at End of Test
.001
.001








o
n
1 Adsorption per
100 gms. Adsorbent










at Saturation
8











-------
                             1.53
U)
                                        Raw water

                                 — 0—Raw water thru 2.5 feet GAC  (0.76  meter )
                                        Raw water thru 2.5 feet XE-340  (0.76 meter )
                                           (all  data points nil)
        0

      Days
0  3  "7 10  14  1721  24  2831  3538 4245 4952  56    63  66  7073  77    84       94  98   105   112
 Figure  29.  Chlorobenzene in raw  water and removal by 0.76 meter  (2.5 feet)  of GAC and
              0.76 meter  (2.5  feet)  of  XE-340 (ED1R).
122

-------
1.
                 Raw water
                 Raw water  thru
                 Raw water  thru
all data points nil
(0.76 meter )
 0
     0  3  7 10  14  17  21 24 28 31  35 38 4245  49 52  56    63 66 70  73  77    84
Days  Figure 30.   p-Chlorotoluene  in raw water and removal by 0.76 meter
                   0.76 meter   (2.5  feet)  of XE-340 (ED1R).
                         94 98    105 108 112 US 119 122
                         (2.5 feet) of  GAG  and

-------
ui
                          Raw water

                          Raw water thru 2.5  feet GAC
                          Raw water thru 2.5  feet XE-3401
                       all  data points nil
                       (0.76 meter )
     Days
31  ?5 38
                                                    £2 56
                Figure 31.
      18 an  35 38  H2 «  k$  &  36    63  66  ?0 ^3 /7   *      g    s   105
o, m and p-Dichlorobenzene in raw water and removal by 0.76 meter
of GAC and 0.76 meter   (2.5 feet)  of XE-340 (ED1R) .
       l'l9'l22
(2.5 feet)

-------
H.T. Water  Source—
    Low  levels of the  four THM,  chloroform, bromodichloro-
methane,  chlorodibromomethane  and bromoform were  present in H.T.
effluent  water and results on  these  compounds will be  discussed
first  in  this section.  The  synthetic organic removal  results
for IRA-904  resin again were poor, as expected, and  only cis
1,2-dichloroethene data is presented to  illustrate the poor
adsorption.  Additional synthetic organic data for the IRA-904
resin  experiment  (EDS) is available  in Appendix B, Raw Data
Tables.

 THM
    Tables  19 and 20 are presented for chloroform and  chloro-
dibromomethane adsorption by XE-340  from H.T. water.   Figures 32
through  39  are presented to  show the influent and XE-340
effluent  concentrations for  the  four THM substances  measured.
The variation in influent and  effluent concentrations  at these
low concentrations makes conclusions relative to  breakthrough
times  and saturation times difficult.  In general the  data  show
that these  substances  appear in  lower concentrations after
treatment of H.T. effluent water using XE-340 columns.   Data
for the THM removal by XE-340  and GAC are presented  later based
on a finished water influent.

    cis  1,2-Dichloroethene—Adsorption data appear in  Table 21.
Breakthrough curves appear in  Figures 40, 41, and 42.

    XE-340 was studied in EDI  and ED1R.  The breakthrough and
saturation  performance of XE-340 on H.T. water in both ED
follows the pattern on raw water (Table  13).  The adsorptive
capacity  for XE-340 at 21 yg/L influent  was calculated from the
H.T. water  tests by using a  log-log plot to compare  directly
with raw  water data at this  influent concentration.  Raw water
and calculated H.T. water values were 0.32 and 0.3 gram   per
column and  0.117 and 0.108 cc  per 100 grams respectively.   From
Table  4 and 5, the total HOC in EDlR for raw and  H.T.  water
were 33.51  and 28.95 yg/L respectively.  The percent of  cis 1,2-
dichloroethene of the  total  HOC were 86.5 percent and  87.7  per-
cent respectively.  Since the  cis 1,2-dichloroethene to  HOC
ratio  is  quite similar, we would expect  little change  in
adsorptive capacity due to competitive HOC.  The  closeness  of
the raw and H.T. water results calculated at a common  influent
level  of  21 yg/L (0.32 and 0.3 gram  per column)  supports this
assumption.

    The effect of competitive  adsorption by TOC on the adsorp-
tive capacity for cis  1,2-dichloroethene can also be compared
using  TOC data from the raw  water and H.T. locations.  The
average TOC concentration in the raw water was 9.8 mg/L  and in
the H.T.  effluent the  concentration was  6.8 mg/L.  Since there
was more  TOC present in the  raw water one might expect that the
adsorptive capacity for cis  1,2-dichloroethene would be

                               76

-------
TABLE 19.  CHLOROFORM ADSORPTION DATA FROM H.T. WATER








ED
1R

1R









•p -P





s
t
•o
stion

152?







= 0.7






MT

Inch
ible

28.6







5 met






Test
Duration

Days
117

122







;r



o*
.3
M C -P
01 § W
4J 3 OJ
(3 H EH
32i
•P U M
o io 3
E-i W Q

Grams
.0355

.013











*o g p
xi i s
o "o *
CO U MH
O -O
rH 10 C
rt W W
O >i -P
EH A «

Grams


.0081










0
10
M
£ c
Adsorbed
Column at
Saturatio

Grains


.0085













•P 4J
a 01
Adsorbed
End of Te
<*>
*


62













*J
10 C
Adsorbed
Saturatio
<*>
%


52









c
A
M H -P
« o w
Qi W 0)
CSH
Adsbrptio
100 gms.
at End of

Grams


.00377









c
a
M M
0) O C
a w o
•0 -H
O U3
01 O
-g o 4J

Grams


. 00395










CC


.0027










-------
              TABLE 20.  CHLORODIBROMOMETHANE ADSORPTION DATA FROM H.T. WATER
oo









ED

1

1R
















Q<
a
•o
s
Feet

2.5

2.5
















Adsorbent

XE-
340
XE-
340

-






•p -P







Average
Influent
ug/L

1.0

.25






2.5





X
Oi
3
Column
Breakthro
Days

83?

?






:eet







c
Column
Saturatio
Days

none

none






= 0.7(








MT
z
Inch










mete








Test
Duration
Days

117

122






r



Oi
K
•H
•M C -P
fl) S in
4-1 3 0)
C H t<
W8 *
rH C
Id JG -H
•P O M
o « 3
fri W Q
Grams

.0047

.0027










•a c 4J
H
•H • O
J-l (0
ft S "O
H & C
O H
CO O
•d O 4J
< rH Id
Grams

.0022

.0013








C
5
fc M
(uoc
Qi 0) O
•0 -rt
c <: -p
Adsorptio
100 gms.
Grams












at Satura
CC













-------
\o
            9 —
            8 —
            7 —
            6 —
            5-
             „
        a.    4 —
             3-
                                                         10.3
                        H.T.  water


                   . O— H.T.  water thru 2.5  feet  XE-340 (0.76 meter  )
                                                                                                        t
                                                                                                       13.4
          Days
 7 9  13    20       32                   6164  6971  7573 8385  9092  97    104        117

Figure 32.  Chloroform in H.T.  water and removal by 0.76 meter   (2.5  feet)  XE-340  (EDI)

-------
  4-
00
O
0)
+J
•i-l
H
\
D"
  2 -
  1 -
                H.T. water

       — O —  H.T. water thru  2.5 feet XE-340  (0.76 meter.)
          03  7 10
                   17 21 2it 28 31  35 38 42 45 49 52  56  63  66 70 73 77 84
94  98
105    112
  ays Figure 33.  Chloroform in H.T. water and  removal by 0.76 meter   (2.5 feet) XE-340  (ED1R) .
                                                                                                     122

-------
                                                                                                  6.5
oo
     Days
                                                                                           Average from 61
                                                                                          to  117  days=2.42
—O—  H.T. water thru 2.5 feet XE-340  (0.76 meter )
           Figure 34.
              20       32                   6164   6971  7678  8385  9092 97    104       117

             Bromodichloromethane in H.T. water and removal by 0.76 meter  (2.5  feet)  XE-340 (EDI)

-------
00
                                       H.T. water thru 2.5 feet
                                            XE-340 (0.76 meter )
      0

     Days
63 66  70 73  77
84
3  7  10  14 17  21  24  28 31 35 38  42 45  49  52  56
Figure 35.  Bromodichloromethane in H.T. water and'removal by 0.76  meter
            (2.5 feet)  XE-340 (ED1R).
94  98
105   112
                              122

-------
                      H.T. water
      
-------
     0)
     -p
     •H
00
    Days
7 10 14 17 21 24  2831  35 38 42451*9 52 56    6366  7073  77   84      91 98    1Q5


   Figure 37. Chlorodibromomethane  in  H.T.  water and  removal by  0.76  meter


                (2.5 feet)  XE-340  (ED1R).
112
122

-------
oo
Cn
                           H.T.  water
                  —-O—  H.T. water  thru 2.5 feet XE-340  (0.76 meter )
                                                              61  64  6971 7678  8385  9092 97
10«t
                                                                                                           117
          «-.    U     /  _/ A vJ    *-V         v**-                    "*  W-r  w_f / 4.  * w ^ i_*  w»y w-^  ^u _/*_  J/     1 U *t        11

             y    Figure 38 . Bromoform  in  H.T.Water  and removal by 0.76 meter   (2.5 feet)  XE-340 (EDI).

-------
00
                                                                                       H.T.  water

                                                                                       H.T.  water thru 2.5
                                                                                       feet XE-340
                                                                                       (0.76 meter )
                                                                                                   122
           Figure  39.  Bromoform in H.T.  water and removal by 0.76 meter   (2.5  feet)  XE-340 (ED1R).

-------
            TABLE 21.  cis 1,2-DICHLOROETHENE ADSORPTION  DATA  FROM H.T,  WATER
oo
-o
ED
1
1R
3







S
Ci
s
1
n
Feet
2.5
2.5
2.5







Adsorbent
XE-
340
XE-
340
904




2.5


g Average
t> Influent
20
25.4
24.1




feet


pi Column
in Breakthrough
32
28
no a




= 0.


$ Column
'Jn Saturation
274
270
Isorpt




'6 mei


MT
z
Inch
27
27
.on




.er


!? Test
"M Duration
3117
122
53







o Total Entering
1 Each Column
<" During Test
.209
< .277
.114







•Q G 4>
a) 1 oi
.0 3 4)
fc i-4 EH
o o
Bl O 
-------
00
00
     0)
     4J
     •H
      Days
                                                                                 H.T. water

                                                                                 H.T. water thru  2.5  feet

                                                                                       XE-340 (0.76 meter )
                                                                     7678  8385  9092  97
117
                Figure 40.  cis  1,2-Dichloroethene in H.T. water and removal by 0.76 meter

                             (2.5 feet) of XE-340  (EDI).

-------
        40 -
                      H.T. water

                      H.T. water thru 2.5
                      feet XE-340 (0.76 meter
        30 -
    M
    0)
    4J
    •H
00
VO
        20
        10
                                               Av.

                                               25.4
      Days
             03  7  10 I1* 17 21 24  28 31  35 38 42  45  49 52 56
63 66 70 73 77
94  98
105   112
119
              Figure 41 .  cis 1,2-Dichloroethene in H.T. water and removal  by 0.76 meter
                           (2.5 feet) of XE-340  (ED1R).

-------
         40
vo
o
         30
      CP
      p.
         20
         10
          0


        Days
                                                                   H.T. water

                                                                   H.T. water  thru 2.5 feet (0.76 meter )
                                                    Av.  24.1
0  4 7  11 lit  1821  2528 32 35  3942 4649  53


  Figure 42.  cis 1,2-Dichloroethene  in H.T.  water and removal by 0.76 meter  (2.5 feet)
              of IRA-904  resin  (ED3).

-------
decreased in the raw water if TOC  competes with the cis 1-2-
dichloroethene and HOC's adsorption  sites.  Since the adsorp-
tive capacity for cis 1,2-dichloroethene by XE-340 receiving
the raw water (0.32 gram  per column) was not  appreciably
different from the adsorptive   capacity for cis 1,2-dichloro-
ethene by XE-340 receiving H.T. water  (0.30 gram  per column)
one might tentatively conclude  that  TOC does not compete with
cis 1,2-dichloroethene  for the  XE-340.  The general lack of
removal of TOC by the XE-340 also  may support  this observation.

    Vinyl chloride—Influent and effluent curves appear in
Figure 43.  The average influent was 0.69 yg/L and the average
effluent was 0.5 yg/L.  Since there  was essentially no adsorp-
tion of vinyl chloride by XE-340 from raw water in the same
time period of 94  to 122 days, these values may not indicate
adsorption.  If they do represent  adsorption,  27 percent was
removed.

    trans 1,2-Dichloroethene—Breakthrough curves appear in
Figure 44.

    The influent concentration  curves in Figure 44 illustrates
a condition which makes data interpretation difficult.  For the
first 84 days of the test, the  influent concentration averaged
0.11 yg/L.  From day 84 to 112  the average was 1.2 yg/L.
Breakthrough occurred at the same  time the influent concentra-
tion increased approximately tenfold.  Research needs to be
done on adsorption of a single  substance from  pure water as well
as a mixture of many substances from pure and  actual plant water.
Such research may indicate that breakthough under a given
condition may occur at  approximately the same  time for a wide
range of concentrations for a specific substance.  If such is
the case, breakthrough might occur at approximately the same
time, but the level of breakthrough  would be much less if the
influent concentration  remained at the 0.11 yg/L level through-
out the test.  In Figure 22, when  the average  raw water influent
concentration of trans  1,2-dichloroethene was  1.3 yg/L, the
same bed depth of XE-340 exhibited no breakthrough throughout
the 122-day test.  Without the  additional research mentioned
above, it is probably best not  to  attempt to explain such
differences obtained on such a  complex system. We can conclude
that XE-340 removes all compound from H.T. water for a period
of 84 days even when the concentration averaged only 0.11 yg/L.

    1,1-Dichloroethane—Adsorption data appear in Table 22.
The breakthrough curve  appears  in  Figure 45.

    The influent concentration, plotted in Figure 45, appears
quite erratic, averaging 0.13 yg/L.  Estimation of the break-
through point from this curve is not too difficult, but the
extrapolated saturation point is questionable  and was based on
adsorption data of this compound through the whole two-year

                               91

-------
    1.0.
0)
4J
•rl
     .9 -
     .8
      "7 IL1LJ
     • /
     .6-
     .5
     .4-
     .3
     .2-
                                                              Av.0.69
                                                             Av. 0.5
                  H.T. water

                  H.T. water  thru 2.5 feet XE-340 (0.76 meter
     .1
     Days
-n     -
 Figure 43.
Ifl
1'05
                          Vinyl  chloride  in  H.T.  water and removal
                          by  0.76  meter  (2.5  feet)  of XE-340 (EDlR)
                                   92

-------
vo
       2   _
       1.9-
       1.8_
       1.7-
       1.6-
    -P  1.4-
    •H
1.2_
1.1 —
1  _
 .9-
 .8-

 .6 —
 .5-
 .4-
 .3-
 .2-
                       H.T.  water
                       H.T.  water thru 2.5  feet XE-340  (0.76 meter )
       n     0 3  7  10  14 17  21 ZH  28  31  35 38 HZ tb <+9 bZ  bb    bi  bb  /U /d  //   B4      yH   y«
         ys   Figure 44 .  trans. 1,2-Dichloroethene in H.T. water and removal by 0.76 meter
                          (2.5 feet)  of XE-340  (ED1R).
                                                                                  98   ,105   112     122

-------
               TABLE 22 .   1,1-DICHLOROETHANE ADSORPTION DATA FROM H.T. WATER
vo
ED
1R









Q.
Q
^D
Q)
fft
Feet
2.5






2.5


Adsorbent
XE-
340






eet


g Average
^i Influent
.13






= 0.7


PI Column
oi Breakthrough
94






6 met


p? Column
"M Saturation
131






ar


MT
z
Inch
9









^ Test
"ra Duration
122









o Total Entering
|j Each Column
w During Test
,00142









gj Total Adsorbed
• § by Each Column
01 at End of Test
.00129









n Adsorbed by Each
g Column at
n Saturation
.00131









^ %Adsorbed at
End of Test
91









^ % Adsorbed at
Saturation
92









o Adsorption per
w 100 gms . Adsorbent
a at End of Test
.0006









O
H
| Adsorption per
100 gms. Adsorbent
.0006









at Saturation
8
.00052










-------
Ul
                   H.T. water
            — O—'  H.T. water thru 2.5 feet XE-340  (0.76 meter )
          0 3  7 10
      Oays  Figure.
 14 17  21 24  28  31  35 38 42 45 49 52 56    63 66 70 73  77    84      94
5.   1,1-Dichloroethane in H.T. water and removal by 0.76 meter
      (2.5 feet) XE-340  (ED1R).

-------
study.  Such estimations are intended to aid overall data inter-
pretation and are not intended to be considered factual.  We
can conclude that XE-340 on H.T. water as on raw water removes
all the compound even when present at the average low level of
0.13 yg/L for a period of time up to approximately 94 days.

    1,1,1-Trichloroethane, 1,2-dichloroethane, carbon tetra-
chloride  (summed concentration)—The average influent concentra-
tion, plotted in Figure 46, was only 0.022 yg/L, which is
probably too low to attempt conclusions.  Periodically through-
out the test, the XE-340 column appears to allow some of the
material to pass while most is adsorbed.

    Trichloroethylene—Adsorption data appear in Table 23.  The
breakthrough curve appears in Figure 47.

    Along with a very low influent concentration averaging 0.066
yg/L, we again have a much lower average of only 0.020 yg/L
entering for the first  84 days with the remaining number of
days averaging 0.240 yg/L.  XE-340 removed all the compound for
the first 84 days except for trace amounts on two sample dates.

    Tetrachloroethylene—The average influent concentration was
only 0.0025 yg/L. The influent and adsorption curves are shown
in Figure 48.  The average concentration was higher during the
first part of the test and the XE-340 column allowed some pas-
sage even on initial start-up.

    Chlorobenzene—The average influent concentration in ED1R
was 0.048 yg/L and individual data points are plotted in Figure
49.  For the first 84 days of the test the average was 0.007
yg/L which was all removed by the XE-340 column.  From day 94 to
the end of the test, the average entering was 0.08 yg/L, of
which a high percentage was removed.

    p-Chlorotoluene—Seven samples of the 30 sampled during the
test period had low levels of p-chlorotoluene ranging from .008
yg/L to .540 yg/L.  The influent data point curve appears in
Figure 50.  All data points for samples through the XE-340
column showed nil concentration except for the sample on day
108, at 1.2 yg/L.  No conclusions are made.

    o, m and p-Dichlorobenzene—Adsorption data appear in
Table 24.   Influent and adsorption curves appear in Figure 51.

    Except for three sample points showing low levels of the
summed values of the compounds, XE-340 removed all the
compounds  from H.T. water.  Removal was essentially complete.
                               96

-------
                                               06
—O—  H.T.  water thru 2.5  feet XE-340

                       (0.76 meter }
                                 #
  0  . ^                 ^     /  \l

Day^5  3   7  10  1417 2124 2831  35  3842  45  4952  56   63 66  7073 77   84      94  98   105
     Figure 46.   1,1,1-Trichloroethane,  1,2-dichloroethane and carbon tetrachloride
                  in H.T. water and removal by 0.76 meter  (2.5 feet) of XE-340 (EDlR).
                                                                                   112
                                                                                          122

-------
                TABLE 23 .   TRICHLOROETHYLENE ADSORPTION DATA  FROM H.T.  WATER
10
CO
ED
1R









P.
5T
Q
*O

Column « Breakthrough 88 ••f = 0.' ° Column "m Saturation 6 me1 MT z Inch er ^ Test "m Duration 122 o Total Entering * Each Column " During Test .0007 •d c -P o g 01 ja 3 o 8-3H 1^° -^-g^ rH (!) C «J W M o s -P EH 43 H) Grams .00057 o Adsorbed by Each g Column at in Saturation ^ %Adsorbed at End of Test 81 ^ %Adsorbed at Saturation o Adsorption per | 100 gms. Adsorbent 01 at End of Test .00027 CD H % Adsorption per 100 gms. Adsorbent at Saturation 8


-------
IO
VD
       M
       (U
       +J
       •H
  3-

 .2-
      .1-

                   H.T.  water

                   H.T.  water thru 2.5  feet XE-340  (0.76 meter.)
                                        45  49 52 56
                                                          63 66  70 73  77
                                                                      84      94 98    105    112
Days Figure  47.  Trichloroethylene in H.T. water  and removal by 0.76 meter  (2.5 feet)

                 of XE-340  (ED1R).
                                                                                                        122

-------
O
O
                      H.T.  water

                      H.T.  water thru 2.5  feet XE-340 (Q.76 meter .)
     Days   03  7 10  Ik 17 2124 2831  35 38  42  4549 52 56   6366  7073  77
94  98
105   112
122
             Figure  48.   Tetrachloroethylene  in H.T.  water and removal by  0.76 meter  (2.5 feet)
                          of XE-340  (ED1R).

-------
 .09.
 .08 —
 .07 —
 .06-
 .05,
 .04-
 .03-
 .02-
 .01 -
Days
H.T. water
H.T. water  thru 2.5 feet XE-340
                j(0.76 meter  )
      03   7  10  1'ft  1721 2^ 2831  36 38 42^5 «t9 52   56   6366 7073  77
                                                              94  98
105   112
                                                                                    122
       Figure 49.  Chlorobenzene  in H.T.  water and removal by 0.76 meter   (2.5  feet)  of XE-340 (EDI)

-------
       .4 -
       .3 -
O
to
     a)
     •P
     en
     p.
       .1 —
              •	H.T.  water
          	O- —t-H.T.  water thru 2.5 feet XE-340  (0.76 meter^
.54
Days
             3  7 10  I«tl7  21  21+28 31  3538  4245 4^ 52  56    6*366 70 73 77    8%       9"+  98    ids 108 U2. l'l5 l'l9122
             Figure 50.  p-Chlorotoluene in H.T. water and removal by  0.76 metex   (2.5  feet) of XE-340 (ED1R)

-------
          TABLE  24.   O,  m,  AND p-DICHLOROBENZENE ADSORPTION DATA FROM H.T. WATER
o
OJ
ED
1R









3
8-
Q.
•8
«
Feet
2.5









Adsorbent
XE-
340





2.5



5 Average
f* Influent
.39





feet



(u Column
"5 Breakthrough
none





= 0."



jj? Column
*g Saturation






6 met



MT
z
Inch






sr



g Test
>oj Duration
122









o Total Entering
g Each Column •
<» During Test
.00425









•o e 4J
M
13 O
o w
1-4 m c
n) W W
O >i 4J
EH f> US
Grams
.00425









n Adsorbed by Each
g Column at
to Saturation










^ *Adsorbed at
End of Test
100









# % Adsorbed at
Saturation










n Adsorption per
g 100 gms. Adsorbent
01 at End of Test
.002









O
H
g Adsorption per
	 100 gms. Adsorbent
at Saturation
8





















-------
 2  —
 1.5-
 1  _
3
•H
  .5 —
H.T. water
H.T. water thru 2.5 feet XE-340  (0.76  meter.)
                                                                                2.57
                                                                                                2.41
 Days 83  7  10  14 17  21 2k  28 31  35 38 42 45 49  52  56    63 66 70 73  77     84      94 98   105
        Figure  51.   o, m and p-Dichlorobenzene in H.T. water and  removal by 0.76 meter
                      (2.5 feet)  of XE-340  (ED1R).
                                                                            112
122

-------
Finished Water Source—
    Chlorination in the plant  process  between  the  H.T.  effluent
sample point and the finished  water  sample  point resulted  in
large increases of the THM.  Chloroform data will  be discussed
first in this section.  A  discussion of cis 1,2-dichloroethane
follows chloroform so the  two  compounds occurring  most  fre-
quently and at the highest concentrations can  be compared.

    Chloroform—Adsorption data  appear in Table 25.  Influent
and adsorption curves appear in  Figure 52,  53, 54,  55,  and 56.
ED4, is shown in Table 25.  Although the average influent  level
varied from 57 for EDS to  67.3 yg/L  for ED4, initial break-
through and saturation times were  quite uniform at approximately
7 and 22 days and 8 and 23.5 days  respectively.  The MTZ also
were quite uniform at 21 and 20  inches respectively.  Thus, The
variation in influent water conditions were such that they did
not greatly affect these parameters.   This  is  more likely  to
occur in ground water sources  which  are generally  subject  to
lesser variations in quality than  river water sources.  Also
maintenance of consistent  contact  times are more easily accomp-
lished in pilot systems as compared  to full scale  plants.

    For these two study periods, a comparison  of the grams of
chloroform adsorbed per 100 grams  of adsorbent at  saturation
 (last column of figures in Table 25) with their respective
average influent level, shows  that the adsorptive  capacity of
GAC for chloroform increased  (0.028  cc and  0.0358  cc) as the
influent concentration increased (57 to 67.3 yg/L), as predicted
by the Polanyi-Manes Theory, which is  discussed beginning  on
page 283. . In ED4, the adsorptive capacity in cc's  per 100  grams
of GAC at saturation appears to  have increased slightly with
increasing bed depth, 0.0449,  0.048, and 0.049 cc  for 1.52
 (5.0 feet), 2.29  (7.5 feet), and 3.05  (10 feet) meters of  bed
depth respectively.  This  is to  be expected because more
strongly adsorbed substances,  both HOC and  non-HOC, are removed
in the shallower bed depths thus reducing competitive adsorptive
effects.  This point will  be discussed later in more detail.

    Chloroform adsorption  from finished water  by 0.76 meter
 (2.5 feet) of XE-340 was studied during EDI, ED1R  and ED2  (Table
25).  Although the average influent  concentration  varied from
80.2, 69.3, and 64 yg/L, initial breakthrough  and  saturation
times for each ED were quite uniform at approximately 3 and 150
days respectively.  The MTZ also were  quite uniform at nearly
the entire column length of 76.2 cm  (30 inches) which was
greater than that found for GAC.   A  comparison of  the grams of
chloroform adsorbed per 100 grams  of adsorbent at  saturation
with their respective average  influent level shows that the
adsorptive capacity of XE-340  for  chloroform decreased  (0.177,
0.148", and 0.134 cc) as the influent concentration decreased.
                               105

-------
TABLE 25.  CHLOROFORM  ADSORPTION DATA FROM FINISHED WATER







ED
1
1R
2
3
3
4
4
4
4
2.!






5
a
Feet
2.5
2.5
2.5
2.5
2.5
2.5
5
7.5
10
feet






Adsorbent

XE-
340
XE-
340
XE--
340
GAC
904
GAC
GAC
GAC
GAC
= 0
JJ 4J





Average
Influent
Vg/L
80.2
69.3
64
57
57
67.3
67.3
67.3
67.3
.76 It




CP
Column
Breakthrou
Days
3
3
0
7

8
29
49
72
eter





Column
Saturation
Days
156
150
150
22

23.!
49
76
98
5





MT
Inch
29
29
30
21

20
24.5
32
31.8
feet '





Test
Duration
Days
117
122
63
53
53
122
122
122
122
= 1.5:


tr»
c
•H
M d JJ
SI "
3 0)
C r-l E-«
3-s-S
o n> 9
tt W Q
Grams
.838
.736
.36
.27
avera<
.733
.733
.733
.733
meters


T3 C 4J
(U n W)
•S Jj £
0 0
in U . 4J
0" JQ it)
Grams
.497
.456
.247
.074
e incree
.094
.235
.376
.511
7.5 f(

u

M
is
Adsorbed b
Column at
Saturation
Grams
.569
.473
.429
.074
se of 1."
.094
.235
.376
.511
Set = 2.




i) 1 1
%Adsorbed a
End of Tes
*
59
62
69
27
5X
13
32
51
70
29 m>




.u
^Adsorbed a
Saturation
*
51
52
50
66

67
80
82
87
sters
C
 
-------
•H
iH
      90_
      80-
      70_
      60-
                                                                                           Finished water
                                                                                           thru 2.5 feet
                                                                                           XE-340
                                                                                            (0.76 meter  )
             Figure 52. Chloroform in finished water and removal  by 0.76 meter  (2.5 feet)  of XE-340
                        IED1).

-------
o
00
        90-
        80-
        70 —
        60-
$50
•H
       40 -I
       30 -
       20-
       10
        0
                                                                                                          150 days
0  - Finished water

     Finished water thru  2.5  feet XE-340  (0.76 meter*)
      Days  03   7 10  14 17 2124  28  31  3538  4245 49 52 56    63 66  7073  77   84      94  98    l"05 "  112'   ' 12~2

             Figure 53. Chloroform in finished water and .removal by 0.76 meter  (2.5 feet) of XE-340 (ED1R)

-------
O
VO
         90
         80
         70
         60
         50
40
         30
         20
         10
                                                                  Finished water

                                                                  Finished water thru 2.5 feet XE-340
                                                                                     (0.76 meter )
               0  «t 7   11 I1*       25    32
                                             5^ 5*6     S~
        Days
                Figure 54. Chloroform in finished water and  removal by Q.76 meter   f2.5feet)  of XE-340 (ED2) .

-------
      M
      0)
O
          200
          175
          150
          125
          100
Finished water

Finished water thru 2.5 feet  IRA-904' resin (0.76 meter )

Finished water thru 2.5 feet  GAC  (0.76  meter )
                                       Av.  100
           25
               0  4 7  11  14  1821  2528 3235 3942 46 49 53

       Days    Figure 55 .  Chloroform in  finished  water  and finished water thru 0.76 meter
                           2.5  feet)  of  IRA-904 resin and Q.76 meter  (2.5 feet) of GAC  (EDS)

-------
120
110
100
                                                                                              3.27
    Finished water
O- 2.5  feet GAG (0.76 meter  )
O" 5  feet GAG (1.52 meters)
  -.7.5  feet GAG (2.29 meters)
    10 feet GAG
                                 35 38  1*21*5
  10
    • • a^^M  VMMHMiri^^^^v
     03  7 10
Days      Figure  56
                                      i   t  r
52  5659 63 66 7073  7780  8487  91  94  98101
                                                                 112
                                                                                                     122
                         Chloroform in finished water and  in the effluent from 0.76, 1.52,  2.29 and
                         3.05 meters  (2.5,  5,  7.5 and 10 feet) of GAG  (ED4).

-------
     Because the adsorptive capacity for GAG and XE-340  is
dependent on influent concentration, the influent concentrations
should be equal to compare the two adsorbents.  If the three
XE-340 data points for adsorptive capacity and influent  con-
centration are plotted on a log-log scale, the resulting plot
is a straight line.  This also applies to our GAC data and to
other HOC on both adsorbents.  This straight line applies to
the concentration range of interest in the particular water
tested, and only when the total HOC profile varies in intensity,
but not when there is a large change in the ratio of specific
HOC.  It appears that the raw, H.T. effluent, and finished water
individually meet these requirements.  However, the finished
water location could not be compared to the H.T. or raw  because
the HOC ratios are not the same.  Therefore, we can predict the
adsorptive capacity at saturation for a specific HOC from the
straight line plot for a given water.  In this way, two  adsor-
bents, or the same adsorbent run at times of different influent
concentrations can be compared at the same concentration.  An
explanation of why these data points form a straight line on a
log-log scale plot and why predictions can be made is presented
in the section on the Polanyi Theory and Manes modifications,
page 283.

     Using the log-log straight line plot for XE-340, we can
calculate the capacity at 67.3 yg/L and compare it directly with
the value for GAC at 67.3 yg/L in ED4.  For XE-340 at 67.3 yg/L,
the grams adsorbed per column is 0.458 and the cc's adsorbed per
100 grams is 0.144.  The GAC data (ED4) in Table 25 shows that
for 0.76 meter  (2.5 feet) of GAC that the grams adsorbed per
column is 0.094 and the cc's adsorbed per 100 grams is 0.0358.
Therefore, XE-340 has 4.9 times (0.458 divided by 0.094)  the
adsorptive capacity for chloroform in our finished water as GAC
per column where the volume of the two adsorbents are the same.
XE-340 had 4 times (0.144 divided by 0.0358) the capacity of GAC,
calculated on an equal weight basis of 100 grams of adsorbent.
This information,  the data in Table 25, and the individual
breakthrough curves give a comprehensive view of chloroform
adsorption in our system.

     As shown in Table 25, the effect of IRA-904 resin on fin-
ished water was studied in ED3.  The level of chloroform leav-
ing the 0.76 meter  (2.5 feet) deep column was 1.75 times the
level entering.  A possible explanation is that the resin was
acting as a phase-transfer catalyst, accelerating the reaction
of free chlorine with precursors to form HOC in the empty bed
contact time of only 6.2 minutes.   A review of phase-transfer
catalysts is available from Aldrich Chemical Company (5).

     During ED1R on H.T. water, the influent concentration to
the XE-340 column for chloroform was 1.2 yg/L and the adsorptive
capacity at saturation of XE-340 was 0.0027 cc per 100 grams
(Table 19).   The log-log straight line plot of XE-340 influent

                               112

-------
concentration and adsorptive  capacity  for  finished water pre-
dicts 0.004 cc per 100 grains  for  finished  water.  When extra-
polated to 1.2 yg/L using the log-log  plot of  finished water
data.  Based on the data in Table 5,   during ED1R, chloroform
was 3.8 percent of the total  HOC  in H.T. water and 45 percent
of the total in finished water,(Table  6).  Because of the higher
ratio of competing HOC in H.T. water we would  expect less
chloroform adsorptive capacity in H.T. water.  Thus, the observed
0.0027 cc value obtained on H.T.  water is, as  expected, less than
the finished water predicted  value of  0.004 cc.  As previously
mentioned, the ratio of specific  HOC is an important factor
in predicting performance under various conditions.

     cis 1,2-Dichloroethene—Adsorption data appear in Table 26.
Influent and adsorption from  finished  water curves appear in
Figures 57, 58, 59, 60 and 61.

     XE-340, 0.76 meter   (2.5 feet) deep,  was  studied in EDI,
ED1R and ED2.  Column breakthrough and saturation time in ED1R
and ED2 are almost identical.  The column  breakthrough reported
in EDI is questionable since  no data points were taken between
day 32 and 61  (Figure 57).

     GAG, 0.76 meter   (2.5 feet)  deep, was studied in ED3 and
ED4.  Breakthrough and saturation times are very close in both
ED, as shown in Table 26 and  Figures 60 and 61.  Using the log-
log plot to compare GAG and XE-340 at  the  same influent concen-
tration, at column saturation, XE-340  has  an average of 3.1
times the capacity of GAG at  equal volumes of  adsorbent  and an
average of 2.5 times per 100  grams.  Calculated at the same
influent level of 21 ug/L, the adsorptive  capacity of both GAC
and XE-340 is approximately 30 percent less in finished water
than it is in raw water.  The TOG values for raw and finished
water average approximately 8.5 and 5.7 mg/L respectively (Table
4).  Again, one might expect  somewhat  increased capacity as the
TOG values decrease.  During  ED1R the  total HOC level in fin-
ished water was 159.6 yg/L and cis 1,2-dichloroethene was only
10 percent of the total (Table 6)  as compared  to the previously
stated 86.5 percent of the total  HOC in the raw water for ED1R.
The 30 percent reduction in capacity in finished water probably
was due to the competitive adsorption  of other HOC.

     As was seen with chloroform,  the  adsorptive capacity of GAC
for cis 1,2-dichloroethene appeared to increase slightly per 100
grams of GAC as the bed depth increased.   The  IRA-904 resin
removed no cis 1,2-dichloroethene from H.T. or finished water
and none was generated.

     Bromodichloromethane—Finished water  adsorption data appear
in Table 27.  Curves appear in Figures 62, 63,  64, 65 and 66.
                                113

-------
TABLE 26 .  cis 1,2-DICHLOROETHENE ADSORPTION DATA FROM FINISHED WATER











ED
1
1R
2
3
3
4
4
4
4
2.5







J"|
+J
CM
s
•a
ID
m
Feet
2.5
2.5
2.5
2.5
2.5
2.5
5
7.5
10
feet







4J
q
3
ti
o

XE-
340
XE-
340
XE-
340
GAC
904
GAC
GAC
GAC
GAC
=0.7(








4J

Averag
Influe
W3/L
10.9
19.4
18.4
18.3
18.3
19.9
19.9
19.9
19.9
met*




£
o<
P
0

JE*
H 0)
8M
m
Days
61?
38
38
18
no c
17
59
101
none
r E






c
Q
•H
4J
Column
Satura
Days
128
191
190
53
dsorpt
66
119
171

feet










MT
z
Inch
16
24
24
20
ion
22
30
37

=1.52








B
O
Test
Durati
Days
117
122
63
53
53
122
122
122
122
mete


tji
a
•H
^1 f* ^J
m § m
•P P 0)
C. H EH
H O

Grams
.114
.206
.1035
,0867
.09
.217
,217
.217
.217
rs 7.5


•O C 4J
IV 1 V)
.0 9 o>
V< rH IH
O O

T3 O
rt fC
O t>
•H Id C
rt W W
O S 4J
EH ft a
Grams
.091
.166
.099
.058
0
.073
,159
.212
,217
Ceet=2 .

•g
IB
W

N
A G
•P O
•a n) -H
a» .p
,Q C fd
SIS
in .H -P
t< o m
rt U w
Grams
.092
.2
.187
.058

.073
.159
.242

29 metei





1 1 [ t
m ui
0)
tj t1*
a)
M O
O
ui -d
•ti C
rt u
%
80
81
96
67
0
34
73
98
100
S 1





*)
m e

*O *H
(U 4J
Adsorb
Satura
*
74
60
60
66

63
75
79

3 fe
£•
JS
V4 H -P
(U O Ul
ft U) 0)
'O EH
c rt
O 
-------
Ul
    M
    +

DayS    Figure  57. cis  1,2-bichloroethene in finished water and  removal by 0.76 meter

                    (2.5 feet) of XE-340  (EDI).
                                                                                                    117
                                                                                                           158

-------
I-1
a\
  0



Days
               0   Finished water



             — O — Finished water  thru 2.5 feet XE-340 (0.76 meter  )
                                                                                                      b


-W-


-9-


9 99 <
/°N ^0-
i 0--0--0 °
| /
M 9 9*+ 28 31  35 38 42 45 49 52 56    63 66  70 73 77   84       94  98    105


            Figure 58.  cis 1,2-Dichloroethene in finished water and removal by  0.76 meter

                        (2.5 feet) of XE-340  (ED1R).
112
119

-------
   40 -
   35
   30 -
VI

-------
                                                                        Finished water
00
                                                                        Finished water thru 2.5  feet
                                                                            (0.76 meter )  IRA-904 resin
                                                                        Finished water thru 2.5  feet GAC
                                                                            (0.76 meter .)
                                                               Av.  19
                                25 28 32 35 39 42  46 49  53
                  Figure  60.   cis  1,2-Dichloroethene in finished water and removal by 0.76 meter
                               (2.5  feet)  of GAC and 0.76 (2.5 feet)  of IRA-904 resin (ED3).

-------
               Fin. water
       — O—  Fin. water thru 2.5' GAC
               Fin. water thru 5' GAC
               Fin. water thru 7.5' GAC
          all data points zero - Fin.
                water thru 10'
                                        (0.76 meter  )
                                        (1.52 meters)
                                          29 meters)
                                       (3.05 meters)
Days
0 3  7 10  141721242831 353842454932 565963 667073 77 80 B>87 9194 981OJ105 112115 122
   Figure 61 .  cis 1,2-Dichloroethene  in finished  water and finished water thru  0.76,  1.52, 2.29
                and 3.05 meters (2.5, 5, 7.5 and 10 feet) of GAC (ED4).

-------
             TABLE 27 .   BROMODICHLOROMETHANE ADSORPTION DATA FROM FINISHED WATER
N)
O







ED
1
1R

2
3
3
4
4
4
4
2.5





1

Feet
2.5
2.5

2.5
2.5
2.5
2.5
5
7.5
10
feet





Adsorbent


XE-
340
XE-
340
XE—
340
GAC
904
GAC
GAC
GAC
GAC
=0.7(





Average
Influent

WfL
37.1
42.7

42.4
39
39
47
47
47
47
met



j3
£
Column
Break thro

Days
20
20

22
16

14
42
73
105
sr !




a
Column
Saturatio

Days
216
210
can't
sxtrap
53

56
98
139
can ' t
extrap
feet





MT
z

Inch
27
27


21

22.5
34.3
42.7

=1.52





Test
Duration

Days
117
122

63
53
53
122
122
122
122
mete


S1

to a 41
O 8 111
sJU
M8*
-A**
JJ O H
O «S 5
E-i W Q

Grams
.388.
.465

.239
.185
averac
.512
.512
.512
.512
CS 7.5.


•d g 4>
41 g W
A 3 «
o"^
$r°
o *o
rH « C
td W w

Grams
.301
.36

.225
.121
e increa
.147
.293
.444
.51
Eeet=2 .

•g
(8
H

Z C
Adsorbed
Column at
Saturatio

Grams
.391
.439


.121
se of 1.1
.147
.293
.444

29 metei




JJ 4J
IB 10
Adsorbed
End of Te
df
%
78
77

94
65
3X
29
57
87
99
S 1




4J
at a
Adsorbed
Saturatio
Of
*
55
54


65

63
71
76

0 fe
G
Jl
H tl *J
(DOW
d w a>
dS^
O  C
O H
0! O
-d o 4J
rtl ^1  (8
O W
01 O
•o o v
rtj f— ( 
-------
      70 -i
      60
ro
           Finished water
           Finished water thru  2.5  feet XE--340  (0.76 meter )
     Days
0    7 9  13   20       32                   61  6^  6971  76  78  8385  9092 97   107        117
Figure 62.  Bromodichloromethane in finished water and removal by 0.76 meter  (2.5 feet)
           of XE-340  (EDI).

-------
to
to
 40






 30





 20






 10





  0

Days
                   Finished water


                   Finished water thru 2.5 feet XE-340  (0.76 meter )
                                 i  i  'i  i—i     i  i—IT
 03  7 10 14 17 21 24  2831  35  38  4245  49  52 55    63  66  7073 77

Figure 63.. Bromodichloromethane in finished water and removal by 0.76 meter

           of XE-340  (ED1R).
                                                                                        (2.5 feet)

-------
                                 52
to
to
     40-
     30-

     r-l
P.

20 _
      10-
                                 t
                                                             Av. 42.4
                                                                          Finished water

                                                                          Finished water  thru 2.5 feet XE-340
                                                                                        (0.76 meter  )
                         B
          Figure 64.  Bromodichloromethane in finished water and removal by 0.76 meter  (2.5 feet)
                     of XE-340 (ED2).

-------
to
                                                                       Finished water
                                                               	O— Finished water thru 2.51 IRA-904 resin
                                                                                      (0.76 meter )
                                                               ......Q.™.. Finished water thru 2.5' GAC
                                                                                      (0.76 meter )
           18 21 25 28 32 35  39  42 46 49  53
Figure 65. Bromodichloromethane in finished water and  removal  by  0.76  meter
           of GAC  and 0.76  meter  (2.5 feet)  of IRA-904 resin  (ED3).
                                                                                           (2.5 feet)

-------
tO
cn
      10
       0
                       Finished water
                    — Finished water thru 2.5 feet GAC  (0.76 meter ,)
                       Finished water thru 5.0 feet GAC  (1.52 meters)
                       Finished water thru 7.5 feet GAC  (2.29 meters)
                       Finished water thru 10.0 feet GAC (3.05 meters)
           03  7  10  14 17 2124  2831  3538 42 45  49 52 56 59 6366 70 73 77 80 84 87 91 94  98 101     112
      Days   Figure  66.  Bromodichloromethane in finished water and removal by0.76,  1.52,  2.29 and
                          3.05 meters (2.5, 5, 7.5 and 10 feet) of GAC  (ED4).
122

-------
     On  finished water,  calculated  at  the  same  influent concen-
 tration  of  39  yg/L, XE-340 had  3.4  times the  capacity of GAG
 per  column  and 2.8 times per  100  grams.  It appears  that for
 both GAG and XE-340 the  tests with  higher  influent concentra-
 tions  yielded  higher  amounts  of bromodichloromethane adsorbed
 than the tests with lower influent  concentrations  as expected
 and  demonstrated throughout for all substances.  The adsorptive
 capacity per 100 grams of GAG did not  appear  to  change as the
 GAG  bed  depth  increased.  Initial breakthrough of  bromodichloro-
 methane  occurred in all  runs  on GAG and XE-340,  but  saturation
 was  not  always reached.  The  IRA-904 resin caused  an increase of
 1.13 times  the influent  level.

     Chlorodibromomethane—Adsorption  data appear  in Table 28.
 Curves appear  in Figures 67,  68,  69, 70, and  71.

     On  finished water at calculated equal influent  concentra-
 tions, XE-340  had 3.0 times the capacity of GAG  per  column and
 2.5  times per  100 grams.  There appeared to be no  change in
 capacity of GAG at 0.76  meter  (2.5 feet)  and 1.52 meters (5.0
 feet)  of depth.  The  IRA-904  resin  did not reduce  or increase
 the  level of chlorodibromomethane in finished water.

     Bromoform—Adsorption data appear in  Table  29.   Curves
 appear in Figures 72, 73, 74, 75, and  76.

     Saturation by bromoform  from finished water was not reached
 consistently on XE-340.  Saturation was reached  in 0.76  meters
 (2.5 feet) of  GAG, but not in deeper GAG beds.   There was no
 removal  and no increase  of bromoform by the IRA-904  resin, in
 finished water.

     Vinyl chloride—Adsorption data appear in Table 30.   Curves
 appear in Figures 77, 78, 79, 80, and  81.

     The curves for ED1R appear in  Figure  77.  The influent
 average  from day 94 to day 122  was  0.55 yg/L.  The average
 effluent from  the 0.76 meter  (2.5  feet) deep XE-340 column was
 0.4  yg/L.  As  discussed  on H.T. water, it  is questionable
 whether  these  figures represent adsorption.  If  adsorption did
 occur, 27 percent was removed.

     In  ED2, the curves  in Figure 78 show  that column break-
 through  occurred very early.  Since samples were taken at 0 days
 and  4  days, we  show a figure  of 2 days for breakthrough.

     In  ED3, the curves  in Figure 79 show  that the vinyl chloride
 effluent from  a 0.76 meter  (2.5  feet)  deep IRA-904  resin column
 averaged 3.9 yg/L over the test period.  The influent average
was  5.4  yg/L.  Based on our other IRA-904  resin  data with other
 HOC, we  do not believe that this  represents any  adsorption.   In
 Figure 79, the average level  of effluent from a  0.76 meter

                                126

-------
             TABLE 28.   CHLORODIBROMOMETHANE ADSORPTION DATA FROM FINISHED WATER
K)
-4
P P










ED
1
1R
2
3
3
4
4
4
4
2.5







j3
Di
S
•o
(1)
03
Feet
2.5
2.5
2.5
2.5
2.5
2.5
5
7.5
10
feet







4*
(U
J)
in
S

XE-
340
XE-
340
XE-
340
GAC
904
GAC
GAC
GAC
GAC
=0.7€








P
4) C
0> 0)
Id 3
to •-!
 a
ft H
ug/L
12
24.5
26.7
27
27
33.6
33.6
33.6
33.6
met




tf*
&
3
o
il
H 01
Sto
m
Days
47
45
25
21
no <
21
59
98
none
:r S






c
o
•H
4J
9 3
rH +J
8 id
' W
Days
260
260
can't
extraj
76
dsorpt
101
174
can't
extrai

feet









MT
Inch
25
25

21.7
ion
23.8
39.7


=1.52








§
•H
P id
in to
1
»Q fl
P O
•0 id M
a) P
01 H P
•O O id
< CJ w
Grams
.165
.334

.117

.183
.349
can't
extrap

29 metei





P P
It) VI
a)
t» EH
•Q m
M O
O
W -O
•O fi
< w
%
87
89
99
82
0
50
95
99
100
S 1





£j
id c
o
•O -H
ID P
•P 1s
M b
O 3
W P
*O id
< U]
%
59
59

64

60
67


D fe<
C

ti t 1 Ij
ID o w
ft w 
-------
       
       -H
NJ
00
25-,
20-
     Days
                      Finished water

                      Finished water  thru 2.5 feet XE-340  (0.76 meter .)
                                                       61 6*t  6971  7678  8385 90 92  97
                                                                                  104
117
          Figure 67. Chlorodibromomethane in  finished water and removal by  0.76  meter  (2.5 feet)
                      of XE-340  (EDI).

-------
to
\£>
      10
          03   7  10  14  172124 28 31  3538  4245 4952 56
63 66 70 73 77
94 98    105
ft 5
          Figure 68. Chlorodibromomethane in finished water and removal by 0.76 meter (2.5 feet)
                      of XE-340  (ED1R).

-------
                  Finished water

                  Finished water thru 2.5 feet XE-340 (0.76 meter )
U
O
      3.

     35 —
     30 _
     25 J
     20 -
     15 -
     10 _
      5 -
                                                B.
          Figure  69.  Chlorodibromomethane  in finished water  and removal  by 0.76 meter (2.5 feet)
                      of  XE-340  (ED2).

-------
(U
+J
•H
 3.


50
20
10
               Finished water


               Finished water  thru 2.5  feet IRA-904  resin (0.76 meter )
               Finished water  thru 2.5  feet GAG  (0.76 meter )
    0  47  11 14 1821  25 28 3235 3942  4649  53
                                                            76
ayS   Figure 70. Chlorodibromomethane in finished water and removal by 0.76 meter (2.5 feet)
                 of  IRA-904  resin and 0.76  meter (2.5 feet)  of GAC (EDS).

-------
     40
     30
U)
NJ
     M

     fl
     -H
     Cn
     3.
     20
     10
                     Finished water
                     Finished water  thru 2.5 feet GAG
                                   (0.76 meterJ
                     Finished water  thru  5.0 feet GAG
                                   (1.52 meters)
                     Finished water  thru  7.5 feet GAG
                           mftt-.ftr.si
Finished water thru 10 feet

  GAG - all  data points nil

          (3.05 meters)
 0                _  ^ ^ ^ ^  __

     T'T' TTo T4 17  21 T4 28 31  35 ^SjT 42 45  4952 5659 63 66 7073 7780  8487
Days • Figure 71. Chlorodibromomethane in finished water and removal by

                  3.05 meters  (2.5, 5, 7.5 and 10 feet)  of GAG  (ED4).
                                                                  91 94 98 101     112
                                                                  0.76, 1.52-,  2.29 and
                                                                                                        122

-------
                  TABLE 29.  BROMOFORM ADSORPTION DATA  FROM FINISHED WATER
co
co










ED
1
1R

2
3
3
4
4

4
4
2.







,rj
£
•d
0)

Feet
2.5
2.5

2.5
2.5
2.5
2.5
5

7.5
10
> fee







4.)
Adsorben


XE-
340
XE-
340
XE-
340
GAG
904
GAG
GAG

GAG
GAG
.=0.'








Average
Influent

U9/L
.1?
1.9

1.91
2.5
2.5
2.5
2.5

2.5
2.5
6 met


•

,£*
tn
3
O
Column
Break thr

Days

?

63
none

42
91

none
none
er






G
O
Column
Saturati

Days



can't
extra]


94
can't
extra^


5 fee








MT

Inch






16.6

'


t= 1.








Test
Duration

Days
117
122

63
53
53
122
122

122
122
>2 me


O*
G
•H

 feet=

•g
fO
w

^1
n fl
4J O
M r-l 4J
rt U to

Grams






.013




2.29 met





JJ 4J
Id (A
(U
Adsorbed
End of T
*
%



100
100
0
56
98

100
100
ers





4J
n) C
O
Adsorbed
Saturati
*
*






72




10 J
C

M M P
u o n
Oi U) HI
TJ EH
G <
O MH
•H • O
P in
& §>*§
o w
(0 O
SO -P
in id

Grams



.005
.007

.0071
.0075

.005
.004
eet=3 .
(5

-------
   .9 -.
.8. -
.7-
   .6_
                    Finished water
            	O	Finished water thru 2.5 feet XE-340  (0.76 meter,)
   .5-
 3
 •H
   .4-
                                                                                  average from day 92
                                                                                  to  day 117 =0.47
Days
 7 g  13   20        32                   61  64  6971   7678  8385 9092
Figure 72.  Bromoform in finished water and removal by 0.76 meter
            XE-340  (EDI).
                                                                            97
                                                                             (2.5  feat)  of

-------
     
-------
U)
     M
     
-------
U)
                             0   Finished water
                           —-O-- Finished water thru 2.5 feet  IRA-904 resin (0.76 meter )
                           —A— Finished water thru 2.5 feet  GAC (0.76 meter )
                                      all data points nil
     Days
                                                         » Av.  2.7
                                                      — -O Av.  2.5
                                   3235  39
0  if 7  11 lit  IB 21 25 28 32 3S 39 42 US ?9 5~3
   Figure 75.  Bromoform in finished water and removal by 0.76 meter  (2.5  feet)
               of GAC and 0.76 meter  (2.5 feet) of IRA-904 resin  (ED3).

-------
00
                           Finished water
                   _O — Finished water thru  2.5  feet GAC (0.76  meter )
                         -- Finished water thru  5.0  feet GAC (1.52  meters)
                    Finished water thru 7.5  feet  GAC - all data points nil (2.29 meters)
                    Finished water thru 10.0 feet GAC - all data points nil (3.05 meters)
17  21 24  28 31  35 38 42 45 49 52 56 59 6366 70 73  77 80  84 87  91 94 98 101
                                                                                               112
          03   7  10
             Figure 76.  Bromoform in finished water and removal by 0.76, 1.52, 2.29 and 3.05 meters
                         (2.5, 5, 7.5 and 10 feet) of GAC (ED4) .
122

-------
               TABLE 30.  VINYL CHLORIDE ADSORPTION DATA FROM FINISHED WATER
u>
VD











ED
1R

2

3
3
4
4
4
4

2.5








,£•
ft
g
•a
«

Feet
2.5

2.5

2.5
2.5
2.5
5
7.5
10

feet








.P
Adsorben


XE-
340
XE-
340
GAC
904
GAC
GAC
GAC
GAC

•-o.it









Average
Influent

pg/L

.55
.6

5.7
5.7
8.4
8.4
8.4
8.4

mete





x
Cn
p
o
Column
Break thr

Days


2


no
3
10
17
35

r







a
0
Column
Saturati

Days


7


emovai
10
21
45
87

feet









MT
z

Inch


21



21
31
56
72

=1.52









Test
Duration

Days

122
63

53
53
122
122
122
122

mete:



Cn
c
•rl

0) 6 W
4J 9 0)
C H H
"is JB -S
•POM
o id P
EH W 5

Grams

;.0014
.0034

.027
.027
.092
.092
.092
.092

:s 7.5



•o g -P
» a in

V4 r-i EH
0 0
in U MH
id H M
O ^t -P
EH .Q id

Grams





0
.0049
.0117
.019
.045

:eet=2 .

A
o
id
w

^i
XI C
-P 0
•0 fl) -H
(U -P
^3 C «
si^
in H 4J
•O o id
< U 03

Grams


.00025



.0049
.0117
.019
.045

9 meter






•P -P
id co
IV
Adsorbed
End of T
<*>
%





0
5
13
21
49

S 1






•P
Id d
o
Adsorbed
Saturati
dP
%


34



65
74
56
69

fe<
•C

f^
rl M -P

-------
0)
-M
•H
tn
3.
    .8 -i
    .7 -
    .6 -
    .5 -
    .4 -
    .3 -
    .2 -
    .1 -
     0  .
   Days
     Finished water
          —O-— Finished water thru 2.5 feet XE-340 (0.76 meter.)
                                                           Av. .55
    \ *•         i   j     i   i   i  i   i  r
                9*t  gS   105   112     122
Figure 77.  Vinyl  chloride in finished water and removal by 0.76 meter (2.5 feet)
            of  XE-340  (ED1R).

-------
0)
-p
-H
01
       0

     Days
                                                               Av.  .6
                                                                         Finished water


                                                                         Finished water thru

                                                                           2.5 feet XE-340
                                                                            (0.76 meter )
0  k 7  11 1\       25   32              53 56

 Figure 78-  Vinyl Chloride in finished water and removal by 0.76 meter  (2.5 feet)

           of XE-340 (ED2).

-------
          —•—  Finished water
          -O—  Finished water thru 2.5 feet  IRA-904 resin (0.76 meter )
          ...p...  Finished water thru 2.5 feet  GAC   (0.76 meter )
10 -
                           14.2
                                     14.1
                                                      Av. 5.7
                                                      Av.  3.9
                                                     Av.  2.2
0
    off  11 ll  1*8  \\  25 Ye h ^5 & «fe 4*6 *f9 5$
      Figure 79.. Vinyl chloride  in finished water end removal by
                 0.76 meter  (2.5 feet)  of IRA- 904 resin and
                 0.76 meter  (2.5 feet)  of GAC  (ED3).
                                 142

-------
                                                                                                34.7
CO
     25
       n. 0    Finished water
      — Q — Finished water thru 2.5  feet GAC  (0.76 meter )
            . Finished water thru 5.0  feet GAC  (1.52 meters)
      Days
3  7 10 l
-------
0)
-p
•H
Cn
3.
                      Finished water

             — D—  Finished water thru 7.5 feet GAC  (2.29 meters)

             	A	  Finished water thru 10.0 feet GAC (3.05 meters)
    15
    10
Days
                 o  1  17 21 3t  2831  35~38 4245  4952  5659 63 66  7073 7780  8487  91  9498101       112
          Figure 81.   Vinyl chloride  in finished water and removal by  2.29  and 3.05 meters

                        (7.5 and 10 feet) of GAC  (ED4).
                                                                                                    122

-------
(2.5 feet) deep GAG column was  2.2  yg/L.   Based on  our  other
work with GAG, with other HOC,  this probably  represents some
adsorption, but it is probably  incorrect  to divide  the  GAG
effluent by the influent to  get a percentage  figure of  59 per-
cent removal.  With vinyl chloride  data,  many more  data points
over the test period would have to  be  taken to  determine the
exact nature of adsorption and  possible desorption  (roll-over)
that may be occurring with this substance.

     The curves in Figures 80 and 81 show the influent  and
effluent concentration  for four bed depths of GAG for ED4.
Breakthrough and  saturation  times for  the four  bed  depths  (as
seen from the curves and reported in Table 30)  show a steady
increase with bed depth indicating  that adsorption  does appear
to be taking place.  It is possible that  after  initial  saturation
is reached on each column that  roll-over  occurs.  That  is, some
of the previously adsorbed vinyl chloride is  desorbed.   After
roll-over it appears that another period  of adsorption  may occur.
Considering the average influent and effluent from  each bed depth
over the 122 day  test period has questionable merit but may show
a trend.  The average influent  was  8.4 yg/L while the effluent
was 4.8 yg/L, 3.1 yg/L, 2.9  yg/L, and  2.8 yg/L respectively for
0.76  (2.5 feet),  1.52  (5.0 feet), 2.29' (7.5 feet),  and  3.05
 (10 feet) meters  of GAG bed.  This  represents removal of 43 per-
cent, 63 percent, 64 percent and 67 percent.

     trans 1,2-Dichloroethene—Adsorption data  appear in Table
31.  Curves appear in Figures 82, 83,  84, and 85.

     Columns 0.76 meter  (2.5 feet)  deep  of XE-340  were studied
in ED1R and ED2.  In Figure  82, breakthrough  was reported at 84
days and extrapolated saturation at 134 days.  The  same bed
depth of XE-340 was studied  in  ED2.  In Figure  83,  no break-
through occurred, which is as expected since  the test duration
for ED2 was only  63 days, which was considerably less than the
breakthrough time of 84 days in ED1R.

     In ED3, the  0.76 meter   (2.5 feet) deep  IRA-904 resin
column  removed none of the  compound (Figure  82).   GAG  columns,
0.76 meter   (2.5 feet) deep,  were studied  in ED3 and ED4 (curves
in Figures 84 and 85) .  The  adsorption curve  for GAG in Figure
84 except for two data  points,  indicates  essentially complete
removal for the 53-day  test  period.  No breakthrough is recorded
for EDS.  We are  probably justified in ignoring the two low
level passages at days  14 and 35 in Figure 84 since the adsorp-
tion curve for 0.76 meter   (2.5 feet)  of  GAG  in ED4 (Figure 85)
shows no breakthrough at all (all sample  points nil) up to day 56.

     Comparing XE-340 and GAG,  both in 0.76 meter   (2.5 feet)
deep columns, at  the same influent  concentration  (using log-log
method) XE-340 had 2.0  times the adsorptive capacity of GAG at
equal volumes of  adsorbent,  and 1.3 times at  equal  weights.

                                145

-------
TABLE 31.  trans 1,2-DICHLOROETHENE ADSORPTION DATA FROM FINISHED WATER








ED
1R

2

3
3
4
4
4
4

2.5







ti
O,
s
•d
0!
Feet
2.5

2.5

2.5
2.5
2.5
5
7.5
10

feet=







Adsorbent

XE-
340
XE-
340
GAC
904
GAC
GAC
GAC
GAC

0.7€







Average
Influent
W9/L
.54

.86

1.04
1.04
.77
.77
.77
.77

met«




x:
c
3
Column
Breakthro
Days
84

none

none
no
52
none
none
none

r






B
Column
Saturatio
Days
134




dsorpl
63?




feet
| i _^j






MT
z
Inch
11




ion
5




=1.52






Test
Duration
Days
122

63

53
53
122
122
122
122

mete


O1
•rl
H fi *J
01 G 01
•p 3 oi
B H EH
W O
id X! -H
Grams
.0059

.00484

.0049
.0049
.0084
.0084
.0084
.0084

:s 7.5


tJ fi 4->
O B B)
xi 9 > 4J
B X» i«
Grams
.0052

.00484

.0049
0
.004
.0084
.0084
.0084

:eet=2 .

•s
tf
W
rN
•Q a
Adsorbed
Column at
Saturatio
Grains
.0053





.004




29 metei




i) i)
id 01
^Adsorbed
End of Te
%
88

100

100
0
48
100
100
100

S 1




^j
id B
fe Adsorbed
Saturatio
%
90





93




) fe<
B

Ij ^J J t
von
t3 EH
B ri!
O UH
•rH « O
4-1 I/I
O W
01 O
t) 0 4J
ie£ rH Id
Grams
.003

.0023

.0028
0
.0023
.0034
.0016
.0012

!t=3.05
c
JS
M U
01 O B
a 01 o
•a -H
B < -P
O id
•H • to
4J 10 3
&&S
O ui
01 O
•a o .P
Grams
.003





.0023




meters
CC
.00238





.00183






-------
                             Finished water thru 2.5
                                feet XE-340  (0.76  meter )
        99  90
         7  10  li+ 17 21 24 28 31  3T5 38 42  45 49 52  56
63 66 70 73 77
94  98  10S
112
122-
DaySFigure 82 . trans 1,2-Qichloroethene in finished water and removal  by 0.76 meter
               (2.5  feet) of XE-340  (EDlR).

-------
                       Finished water

                       Finished water thru 2.5 feet XE-340  (0.76 meter )
        1.5-
        1  _
CO
      0)
      -P
      •H
      i-l
         .5-,
                                                                     Av.  .86
               ? Y ll It       °5    8               ?3 &    63
          Y  Figure 83. trans 1,2-BLchloroethene in finished water and removal by 0.76 meter
t
                        (2.5 feet) of XE-340  (ED2).

-------
4  -.
    o
    -P
    -H
    •H

    tn
vo
                     Finished water




             —"O—  Finished water thru 2.5 feet IRA-904 resin (0.76 meter  )



                     Finished water thru 2.5 feet GAG (0.76 meter )
                                                             Av.  1.04
      Days
             lilt  1821  25  28  3235  3942  4649 53


       Figure  84.  trans 1,2-Dichloroethene in finished water and removal by  0.76 meter

                   (2.5  feet) of GAG and 0.76 meter   (2.5  feet) of  IRA-904  resin (EDS).

-------
Ul
o
                           Finished water
                           Finished water thru 2.5 feet GAC
                                       (0.76 meter )
                           Finished water thru 5.0 feet GAC, all data points  nil
                                       (1.52 meters)
                           Finished water thru 2.29 and -3.05 metprs
                             (7.5  and  10  feet)  GAC, all data points nil
            03   7 10 1<4 17 2124  28 31 3538  4245  4952 5659 63 66 7073  7780  84 8791 94  98101      112      122
     Days     Figure 85.  trans 1,2-Dichloroethene in finished  water and removal by 0.76,  1.52, 2.29 and
                          3.05  meters (2.5,  5,  7.5 and 10 feet)  of GAC  (ED4).

-------
     The adsorption curve  for  the  1.52 meters  (5.0  feet)  deep
GAG column in ED4, Figure  85,  shows  three  low  level passages of
the compound at three widely separated sample  dates (days  28, 91,
and 122).  Since  all the other data  points showed nil  concen-
tration, these three points are not  considered as breakthrough.
All data points for the 2.29  (7.5  feet)  and 3.05  (10 feet) meters
deep GAC columns were nil.  Adsorption was complete over the full
test range for these two bed depths.

     1,1-Dichloroethane—Adsorption  data appear in  Table  32.
Curves appear in  Figures 86, 87, 88,  89, and 90.

     XE-340, 0.76 meter   (2.5  feet)  deep,  was  studied  in ED1R
and ED2.  Taking  both adsorption curves  into account,  Figures 86
and 87, we record breakthrough at  84  days  in ED1R and  63 days in
ED2.  In Figure 86 we also could consider  the  extremely low
level passage  (.002 yg/L)  on days  49, 63,  and  66 as part of the
breakthrough curve, thus setting the  breakthrough point at 49
days which is still in fair agreement with the 63 days in ED2.
The adsorption curve in Figure 88  for the  IRA-904 resin, 0.76
meter   (2.5 feet) deep indicates no  removal of the  compound in
EDS.  GAC, 0.76 meter   (2.5 feet)  deep,  was studied in ED3 and
ED4^  Considering the data as  a whole, the breakthrough and
saturation times  recorded  in Table 32 are  fairly close for ED3
and ED4.  In ED4, Table 32, we observe a steady increase in
breakthrough and  saturation time as  GAC  bed depth increases
despite the spread of data points  in  the adsorption curves in
Figures 89 and 90.

     1,1,1-Trichloroethane, 1,2-dichloroethane, carbon tetra-
chlorlde  (summed  value)—Adsorption  data appear in  Table 33.
Curves appear in  Figures 91, 92, 93,  and 94.

     XE-340, 0.76 meter   (2.5  feet)  deep,  was  studied  in ED1R
and ED2.  In Table 33, breakthrough  at 98  days was  reported for
ED1R.  The adsorption curve in Figure 91 indicates  complete
removal up to day 49, at which time  a very low passage occurred
up to day 98 when passage  increased  sharply.   On raw water,
Figure 25, a similar low level passage occurred from day  3 to
day 84, at which  time the  passage  increased sharply.   On H.T.
water, Figure 46, the low  level passage  started on  day 7.  It
appears that the  trend for XE-340  is  to  allow  some  low level
passage of this summed group of substances from very early after
initial flow has  begun, then to reach a  period of increased
breakthrough varying from  98 days, 84 days and 66 days respec-
tively for ED1R,  ED2 and ED4.

     IRA-904 resin, 0.76 meter  (2.5  feet)  deep, was studied in
ED3, and the adsorption curve  in Figure  93 indicates no removal.
GAC, 0.76 meter   (2.5 feet) deep,  was studied  in EDS and  ED4.
If the data point at day 3 (Figure 94) for the 0.76 meter
(2.5 feet) deep GAC column is  not  considered,  both  0.76 meter

                                151

-------
              TABLE 32.   1,1-DICHLOROETHANE ADSORPTION DATA FROM FINISHED WATER
Ul
K)






ED
1R

2
3
2
4
4
4
4

2.5




I
Q

Feet
2.5

2.5
2.5
2.5
2.5
5
."
10

feet




Adsorbent


XE-
340
XE-
340
GAC
904
GAC
GAC
GAC
GAC

=0.7*




Average
Influent

yg/L
.18

.8
.32
.32
.4
.4
.4
.4

mete




Column
Breakthrough

Days
84

none
16
no
28
42
52
77

•C I




Column
Saturation

Days
105


21
idsorp
49
56
59
87

feet




MT
z

Inch
6


7
:ion
13
15
11
14

=1.52




Test
Duration

Days
122

63
53
53
122
122
122
122

mete:



S1
Total Enteri
Each Column
During Test

Grains
.002

.001
.0015
.0015
.0044
.0044
.0044
.0044

rs 7.5



•a g -P
v i ia
A 9 <»
83*
01 U 
%
90


88

79
88
94
94

) fe<
c

J3

Adsorption p
100 gms. Ads
at End of Te

Grams
.0009

.00047
.00025
0
.0008
.0005
.0004
.0004

!t=3.05
c
4)
A

01 .3 .3
d rtl JJ
0 n)
•H • M
4J 01 3
a E 4->
vj a> m
o ui
in o
•O O 4J
< I-H n)

Grams
.0009


.00025

.0008
.0005
.0004
.0004

meters

cc
.00077


.00021

.00068
.00042
.00034
.00034



-------
                                                                                        1.1
Ul

w
                                Finished water



                                Finished water  thru 2.5 feet XE-340 (0.76 meter )

                                                                                     I


                                                                                  11
                                                                                  I  I
                                                                                  I  I
                                                                                  I   I
                                                                                  I    I
      Days
03  7 10
                       17  21  24 28 31  35 38 42 45 49 52
                                                                        94  98
105   112
                                                                                             122
           Figure  86.  1,1-Dichloroethane in finished water and removal by 0.76 meter   (2.5  feet)


                       of XE-340  (ED1R).

-------
tn
                                                                         Finished water

                                                                         Finished water thru 2.5 feet XE-340
                                                                                     (0.76 meter )
          Figure  87.   1,1-Dichloroethane  in finished  water and removal  by 0.76 meter (2.5 feet)
                       of  XE-340  (ED2).

-------
tn
ui
                  11; 14 10  21  25 28 32 33 39
                                                                           Finished water
— Q _  Finished water thru 2.5'
            IRA-904 resin  (0.76 meter )
...Q...  Finished water thru 2.5' GAG
               (0.76 meter  )
        Days
             Figure 88.   1,1-Dichloroethane  in  finished water  and  removal  by 0.76 meter (2.5 feet)
                          of GAC and  0.76 meter  (2.5  feet)  IRA-904  resin  (ED3).

-------
         2.0
        1.5
    a)
    •P
    •H

Ul

(Tl
        1.0
         .5
                             • Finished water

                          O —Finished water  thru  2.5  feet GAC  (0.76 meter )

                            ...Finished water  thru  5.0  feet GAC  (1.52 meters)
       Days
                                                                     9194  98101      112

Figure 89.  1,1-Dichloroethane in finished water and removal by 0.76  and 1.52 meters

            (2.5 and  5  feet)  of GAC (ED4).
                                                                                                            122

-------
                     Finished water
                     Finished water thru 7.5 feet GAG
                                  (2.29  meters)
                     Finished water thru 10 feet GAC
                                  (3.05  meters)
Days
                  28        U2         56         73                9H         108
Figure 90.  1,1-Dichloroethane in finished water  and removal by 2.29 and  3.05  meters
            (7.5 and 10 feet)  of GAC (ED4).
                                                                                                    122

-------
                    TABLE 33.  1,1,1-TRICHLOROETHANE,  1,2-DICHLOROETHANE AND
                    CARBON TETRACHLORIDE ADSORPTION DATA FROM FINISHED WATER
oo
4J 4-*













ED
1R

2

3
3
4
4
4
4

2.5








jCj
^3
Q<
S
•O
0)
n

Feet
2.5

2.5

2.5
2.5
2.5
5
7.5
10

feet








4J
fj
dsorbe
rij


XE-
340
XE-
340
GAC
904
GAC
GAC
GAC
GAC

=0.71









4>
verage
nfluen
< H

ug/L
.66

1.47

5.3
5.3
7.7
7.7
7.7
7.7

3 met





.a
Cn
3


olumn
reakth
u «

Days
98

none

28

38
87
none
none

sr 1







a
Q
•rH
C n)
o *$
o w

Days




58

73
199



i feel










MT
z

Inch




16

14
34



.=1.52









c
0
4-> n)
0) H
 3
EH Q

Days
122

63

53
53
122
122
122
122

mete



&*
C
•rl

o S w
4J 3 fl)
f* f^ C>|
O «3 3


Grams
.0072

.0083

.025
average
.084
.084
.084
.084

rs 7.5



*O C 4-*
0) B O
•Q 3 (U
M •-< EH
0 0

*o o
•H S fl
n) W W
O >i 4J
EH ^J id

Grams
.0067

.0083

.02
increas
.038
.078
.084
.084

f eet=2 .

g*
o
16
W

s-
ja c
4J O
13 tj *H
0) 4J
XI g 10
(0 r-l 4J
^ O ifl
< U UJ

Grams




.0204
i of 1.5X
.038
.098



29 meteJ






-M -P
ID !/>
QJ
r^ g^
XI 
%
93

100

80

45
93
100
100

•s ]






•u

%




74

76
72



0 fe
c
V
.q
H tl 4J
(!) O Ul
Q, w a)

e <
O  (0 3
&&S
o w
in o
T3 O 4J
rtl rH nj

Grams




.0116

.022
.0278



meters

cc













-------
Ul
10
     Days
                   Finished water

            — O —• Finished water  thru 2.5 feet XE-340  (0.76 meter )
                                     3538, 42  45 49  52  56
                                                                                 94  98
                                                                  105
112
122
           Figure 91.
1,1,1-Trichloroethane, 1,2-dichloroethane and carbon tetrachloride in finished water
and removal by 0.76 meter (2.5  feet)  of  XE-340 (ED1R).

-------
                                  Finished water

                                  Finished water thru 2.5 feet XE-340  (0.76 meter )
•H
H
\

3-
                                                                  Av.  1.47
Days
      Figure  92.   1,1,1-Trichloroethane,  1,2-dichloroethane and carbon tetrachloride in finished
                   water and removal by 0.76 meter (2.5 feet)  of XE-340 (ED2).

-------
                      A
                           11.7
                           . Finished water
                           — O — Finished water thru 2.5'  IRA-904 resin
                                               (0.76  meter. )
                                 Finished water thru 2.5'  GAC
                                               (0.76  meters)
                                                    Av. 7.9
                                                    Av. 5.3
 0
Days
                                            58
Figure  93.  1,1,1-Trichloroethane,  1,2-dichloroethane and carbon
            tetrachloride in finished water and removal by
            0.76 meter   (2.5  feet)  of GAC  and  0.76  meter
            (2.5 feet)of  IRA-904  resin  (EDS).
                                161

-------
               Finished water
          O—-Fin.  water thru 2.5' GAC
                  (0.76 meter  )
              -Fin.  water thru 5.0' GAC
                  (1.52 meters)
       Fin. water thru 7.5' GAC -
              all data points nil
                  (2.29
       Fin. water thru 10.0' - all data
              points nil
                  (3.05 meters)
     03   7 10 14 17 2124  28 31 35  38  42 45 4952 56 59 63 66  70 73 7780 84 87 91 9498  101

Days   Figure 94.
                                                                     112
122
1,1,1-Trichloroethane, 1-2-dichloroethane  and carbon tetrachloride in
 finished water and removal by 0.76, 1.52,  2.29 and 3.05 me-ters  (2.5,
 5,  7.5  and 10 feet)  of GAC (ED4).

-------
(2.5 feet) deep GAG curves  (Figures  93  and 94)  are  similar
because breakthrough times  are  28  days  and 38  days  and  satura-
tion times are 58 days  and  73 days respectively.  As  expected,
as the GAG bed depth increases,  the  time  to breakthrough
increases (Figure 94).

     Trichloroethylene—Adsorption data appear in Table 34.
Curves appear in Figures  95, 96, 97,  98,  99, 100, 101 and 102.

     XE-340, 0.76 meter  (2.5 feet)  deep,  was  studied in ED1R
and ED2.  A breakthrough  time of 94  days  was determined from the
adsorption curve in Figure  95.   This  compares  closely with a
reported breakthrough on  raw water of 96  days  and 88  days for
H.T. water.  Saturation was not reached and at the  end of the
test, 122 days, the adsorbent removed 96  percent of the compound
entering.  The adsorption curve in Figure 96 for ED2  appears to
be the first contradictory  data of the  whole project.  Tri-
chloroethylene breakthrough occurred on initial start-up and
XE-340 effluent from test day 14 to  day 32 was higher than the
influent.  All other results on trichloroethylene adsorption
from raw, H.T. and finished water  in the  two-year study were in
line and as expected with related  substances and concentrations.

     Results with a 0.76  meter   (2.5  feet)  deep column of IRA-
904 resin in ED3 are shown  in Figure  97.   The  IRA-904 resin
effluent average concentration  over  the test period was 10 times
the influent concentration. It appears that trichloroethylene
is being generated by the column.  The  adsorption curve for
0.76 meter   (2.5 feet)  deep of  GAG in ED3 is shown  in Figure 98.
The average influent concentration was  very low, 0.075 ug/L.
The GAG column effluent contained  trichloroethylene on two
sample dates.  From test  day 28 to the  end of  the test at 53
days, the influent and  effluent concentration  was nil.  Because
of the very low average influent,  establishment of  a  break-
through or saturation time  was  not considered.

     In ED4, the average  influent  concentration was 0.68 yg/L
and adsorption curves for 0.76  '(2.5  feet), 1.52 (5.0  feet),
2.29  (7.5 feet) and 3.05  (10 feet) meters of GAG are  shown in
Figures 99, 100, 101 and  102 respectively.  The reported break-
through and saturation  times are open to  question,  but when con-
sidering all four bed depths, overall removals were 39 percent,
80 percent, essentially 100 percent,  and  essentially  100 percent
respectively in the 0.76  (2.5 feet),1.52  (5.0  feet),  2.29  (7.5
feet) and 3.05  (10 feet)  meters deep columns.   The  respective
adsorption curves clearly show  increased  adsorption with
increasing bed depth.   The  effluent  concentration for all data
points fox the 2.29  (7.5  feet)  and 3.05 (10 feet) meters deep
columns was nil except  for  trace passage  for two points and
points and three points respectively.
                               163

-------
K
H


I

Q
W
W
W
H
&
H
I
IS
o
H
EH


O
W
^
!H
W
EH
H
O
K
O
U
H
8
uoT^BJm^es 4«
^uaqjospv 'sui6 QOT
xad uoT^dJospv |
h
o
}ssi jo pus ^« a
HiaqjospY *sui6 QOT 3
jsd uoT^clzosptf u
uoT^Bjn^pg
^B psqaospv% *
3ssj 30 pug
5B paq;tosp\f% *
UOT^Bjn^BS 01
^B UUJtlXOD 1
qoeg Aq peqjospv u
^sai 50 pua ^B w
UBintoo tpea ^q §
psqaospv T^OI g
^ssi BUT ana 

CM
CM
O
O
•
ro
CM
O
o
*
CM
CM
i-l

can't
extrat
•<*
en
rH
CM
•
1 0
as
in
CM
«
H







CM
en
O
O
m
VD
9)
jj
can't
alcula
0
r«
in
1 0
%%
in
*
CM
CM



0)
4J
14
^
(d
S
s
r-l
9
•ation t<
concent:
co
in



m
r-
o
•
%
in
.
CM
n





X
O
H
M
o
iM
co
0)
4J
1
CTI
CN
01
11
4J
0)
0)

-------
en
       0

     Days
63  66  7073 77
9+   98
                                                                                  IDS
03   7  10  14 17 21 24  28 31  35 38  42 45 49 52  56
Figure 95. Trichloroethylene in finished water  and removal by  0.76 meter   (2.5  feet>
            of XE-340  (ED1R).
112
122

-------
          2.CU
cn
                                                  0   Finished water

                                                  O — Finished water thru 2.5 feet XE-340
                                                                 (0.76  meter )
                                                                      Av. .57
       ays    Figure 96.  Trichloroethylene in finished water and removal'by 0.75 meter   (2.5  feet)
                         of XE-340  (ED2).

-------
                  Finished water
               - Finished water thru 2.5 feet  IRA-904 resin  (0.76 meter )
     2 -
     1 -
Days
0  4  7  1114  18  21 2528  3235  39 42  46  49  53
 Figure 97 . Trichloroethylene in finished water and removal by  0.76 meter  (2.5 feet)
             of IRA-904 resin  (ED3).

-------
        0)
        4J.
CTl
00
        i.o—
                       Finished water


                       Finished water  through  2.5  feet GAG   (0.76 meter )
                                                            Av. 0.075
                         ^Hft£HHM^B^^HH&*^fl^^^RM«HB^^ftM^^B^H^^^^^^^^BB«M^^
                         ww  vv 
-------
  0  .  Finished water

—-O-  -  Finished water through 2.5 feet GAC  (0.76 meter )
                                                                         Av.  .68
    -^Q  99
5 59 63 66 70 73  77 ffO
                                                  91
        /\
       /  \

9^ i b i iTsi'oei? 21*15 1*1
                                                                           122
Days

Figure  99.  Trichloroethylene in finished water and removal by  0.76 meter  (2.5 feet) of GAC  (ED4).

-------


-------
1 -
 S-l
 0)
 +J
                Finished water


                Finished water through  7.5  feet GAG (2.29  meters)
                                                                                           Av.
                                                                                        -• .68
                                              oo
                                              o
0
Days
• +  f y  99  99  99  99  » r"*N>  ? O  f ?  »  P  9 f>  *»  r^f^f »  f » » »
7 IT) 1U 17  212«t2831 35384245  4952 5659  6366  7073 77 ffO 8487 91 9*t 98 TOlltTSlOai 12115119122
Figure 101. Trichloroethylene in finished water and removal by 2.29 meters  (7.5 feet) of GAG  (ED4).

-------
        4J
        -H
to
      1  —
                   0    Finished water


                  •O — Finished water thru 10 feet GAG  (3.05 meters)



                           Y-
                                                                                          CM
                                                                                          O
                           9999  99  P 0
           03   7  10  14  17  21 2i+ 2831  35  38 W 45 4952  5659  6266 69 73 7780  8487 9194  98 101 105 108 112 115119"l22
      Days    Figure 102.Trichloroethylene in finished water and removal by 3.05 meters  (10 feet) of  GAG (ED4).

-------
     Tetrachloroethylene—The  influent concentration  in  all  ED
was very low.  The  average  in  ED1R was 0.016  yg/L.  The  concen-
tration in ED2 and  ED3 was  nil for all test points.   In  ED4,  all
points were nil to  test  day 108 and from day  108  to the  end  of
the test at 122 days, the average concentration was only 0.02
pg/L.  The influent concentration and adsorption  curves  for
XE-340, 0.76 meter   .(2.5 feet)  deep,  in ED1R  are  plotted in
Figure 103.  The column  allowed trace passage at  test day 10,
14, 17 and 98.  In  EDS,  while  the influent concentration was nil
for all test points, the effluent from the IRA-904 resin column,
0.76 meter   (2.5 feet) deep, had low levels on six of the six-
teen test points.   In ED3,  all effluent test  points showed nil
concentration through a  parallel GAG column 0.76  meter  (2.5
feet) deep.  In ED4, all effluent test points showed  nil con-
centration through  all four bed depths of GAG.

     Chlorobenzene—Finished water adsorption data appear in
Table 35.  Curves appear in Figures 104, 105, 106, and 107.

     The influent concentration to the 0.76 meter  (2.5  feet)
deep XE-340 column  in ED1R  was erratic as shown by the curve in
Figure 104.  The influent concentration was nil on more  than
half of the test days.   The column effluent contained no chloro-
benzene during any  of the test dates.   In ED2, the average
influent concentration was  lower, Figure 105. The XE^340 column
removed all chlorobenzene up to day 56, when  a trace  level was
noted in the effluent occurring after a peak  in the influent
concentration.  On  test  day 63, the column again  had  no  chloro-
benzene in the effluent. In ED3, Figure 106, the IRA-904 resin
column had more chlorobenzene  in the effluent than in the
influent, the average increase being 1.4 times.   GAG,  0.76
meter   (2.5 feet) deep,  was studied in EDS and ED4.   In  EDS,
Figure 106, except  for one  data point, test day 4, all the
chlorobenzene was removed for  the entire test period  of  53 days.
When repeated in ED4, Figure 107, all the chlorobenzene  was
removed to the breakthrough time of 84 days.   It  is questionable
whether saturation  occurred as indicated in Figure 107,  or if
the data point at day 115 was  merely a spike  in the effluent.
In deeper GAG columns, 1.52(5.0 feet), 2.29 (7.5  feet),  and
3.05  (10 feet) meters, no trace of chlorobenzene  was  found in
the column effluents throughout the test period of 122 days.

     p-Chlorotoluene—The influent concentration  in ED1R, ED2,
EDS, and ED4 was essentially nil and the effluent from all
adsorbent columns was essentially nil.

     o, m and p-Dichlorobenzene—Adsorption data  appear  in Table
36.  Curves appear  in Figures  108, 109, 110,  111  and  112.

     The influent and adsorption curves for the summed value of
the three isomers in ED1R are  shown in Figure 108.  The  XE-340
column, 0.76 meter   (2.5 feet)  deep,  removed  the  isomers to

                               173

-------
  .05
                       t
                      .0
                  Average first 7 days
                  of test .072
                                                              .Finished water
                                                           O —Finished water thru 2.5 feet
                                                               XE-340  (0.76  meter )
                                                                                 I \
                                                                                 I   I
                                                                                 /   I
                                                                                 /   \
                                                                                I
                                                                                       \
Days
3  7 10  14  17 21     2'8 31 35 38 4!> 4"5  49 52  56    63  6&  70 73 77   84      95   §8   1
 Figure  103.  Tetrachloroethylene in  finished water and removal by 0.76  meter  (2.5 feet)
             of XE-340 (ED1R)
                                                                                                    1T2

-------
                TABLE 35.   CHLOROBENZENE ADSORPTION DATA FROM FINISHED WATER
en









ED
1R

2

3
3
4
4
4
4

2.E






f .
•o
 a
< H

Ug/L
, .14

.08

.8
.8
.86
.86
.86
.86

j met



t*
o>
3
lumn
eakthro
u m

Days
none

none

none

84
none
none
none

5r





c
lumn
turatio
O id
O' w

Days






115?




5 fee







MT

Inch






8




XL. 5:






§
•H
4i
P 10
W fc
 o to
Q, in d)

C ^
O M-<
•rl • O
P B)
0 W
in o
TJ O -P
< rH id

Grams
.0007

.00021

.0022

.0043
.0027
.0018
.0013

3t=3'.OI
a
(V
0! O C
a w o
•O -H
a < -P
o id
P (IJ 3
& § -P
o w
ra o
•a o -P
ft >-t 10

Grams






.0043




meters

CC






.0039






-------
              Finished water

              Finished water  thru 2.5 feet XE-340
                                   (0.86 meter  )
- all data points nil
    0  3  7 10 14 17 21 24  28 31 35
Days   Figure 104: Chlorobenzene in finished water and removal by 0.76 meter
                         94
 98   105  112     122
(2.5 feet)of XE-340 (ED1R).

-------
                                                     Finished water

                                                     Finished water through 2.5 feet XE-340
                                                                              (0.76 meter  )
                                                         Av. .08
                                      53 56    
-------
00
                                                Finished water
                                                Finished water through  2.5  feet IRA-904 resin  (0.76  meter )
                                                Finished water through  2.5  feet GAC (0.76 meter  )
             0  h  7   1114  1821  2528 3235 39 42 4649 53
       DayS     Figure 106..Chlorobenzene in finished water  and removal by 0.76 meter  (2.5 feet)  of
                            GAC and 0.76 meter (2.5 feet) of IRA-904 resin  (ED3).

-------
         2.5
vo
                          Finished water
                          Finished water thru 2.5 feet GAC
                                   (0.76 meter )
                  Finished water thru 5  feet GAC
                                (1.52  meters)
                  Finished water thru 7.5  feet GAC
                                (2.29  meters)
                  Finished water thru 10 feet GAC
                                (3.05  meters)
all points nil
         0.5
              03  7 10  14 17 21 24 28 31 3538 4245  49 52 5659  5355  70 73 77  80  8487  91 94 98  101
         Days Figure 107. Chlorobenzene in finished water and removal by 0.76, 1.52,  2.29  and 3.05
                           meters (2.5, 5, 7.5, and 10 faet) of GAC (ED4).

-------
                        TABLE 36.   o, m,  AND p-DICHLOROBENZENE ADISORPTION

                                    DATA FROM FINISHED WATER
oo
o







ED

1R


2

3

4
4
4
4







5
S1
Q
•a
4)
Feet
THRt
2.5


2.5

2.5
m-Dl
2.5
5
7.5
10







Adsorbent

E ISO
XE-
340
YT?—

340
904
CHLOF
GAC
GAC
GAC
GAC

|
I
i
I
A
\ ^^
I
 fi 0 M
< H 1 CJ ffl
pg/L Days
1ERS SUMMED
.63 1 none

I
2.1 1 none

.3 1 inc
OBENZBNF.
nii conce
during er
nii conce
duifing er
nil cone*
during er
nil conce
during er
1






Column
Saturation
Days






reasec

ntrati
tire t
ntrati
tire t
ntrati
tire t
ntrati
tire t







KT
z
Inch






2.7X

on
est
on
est
on
est
on
est







Test
Duration
Days

122


63



122
122
122
122




0>
•H
^J f^ ^3
O S 01
4J 3 01
fi H EH
w o
Grams

.0069


.0068











«a g 4J
OEM
f> 3 i -P
EH J3 rt
Grams

.0069


.0068









J3
O
n)
H
^H
Adsorbed b
Column at
Saturation
Grams


















-U 4-1
^Adsorbed a
End of Tes
%

100


100













Jj
^Adsorbed a
Saturation
%













-M
q)
o
M M 4->
4) O in
O, in 
-------
                                          TABLE  36.   (CONT.)
CO
4J 4J









ED

4
4
4
4

4
4
4
4






5
I
4)
0)

Feet
P-D3
2.5
5
7.5
10
o-Dl
2.5
5
7.5
10






orbent
<


CHLOF
GAC
GAC
GAC
GAC
CHLOI
GAC
GAC
GAC
GAC






if
M rH
&  C
< H

yg/L
3BENZE
.24
.24
.24
.24
DBENZE
.14
.14
.14
.14




JS
0>
umn
akthrou
•H 
•rl
M C JJ
(D B I/I
JJ 3 0>
"8*
id tC *H
-POM
O it) 3
EH W Q

Grams

.0026
.0026
.0026
.0026

.0015
.0015
.0015
.0015



*O C JJ
•838
rl H EH
O O
W U MH
*o o
rt J3
O TJ
rH Id C
id w w
JJ
O S JJ
EH J3 id

Grams

.0026
.0026
.0026
.0026

.0015
.0015
.0015
.0015

A
o
n>
H
^i
13 id -H
a) JJ
X) d id
10 rH JJ
•o o as
< O W

Grams















JJ JJ
id 10
•8s
!°
KC H
OP
%

100
100
100
100

100
100
100
100





JJ
orbed a
uration
01 JJ
•O id
< CO
dP
%










a
0)
f%
M M JJ
ao to
n a)

O M-l
•H • 0
JJ Ul
O W
m o
•o o JJ
rtl rH Id

Grams

.0015
.00074
.00049
.00037

.00085
.00043
.00028
.00021
c

lt\
t^ ^J
2.8S
•a -n
•rl • rl
JJ W 3
a & JJ
o w
0) O
•O O JJ

Grams











CC

.0012




.00065



            2.5 feet=0.76 meter   5 feet=1.52 meters  7.5  feet=2.29 meters  10 feet=3.05 meters

-------
00
to
                                             Finished water

                                    — O —   Finished water thru  2.5  feet XE-340 (0.76 meter )
     Days
          03  7 10 14 17 21 2H  2831  35 38  42,^5
94-  98
105
115
122
          Fiqure 108 o,  m and p-Dichlorobenzene  in finished water and r.emoval by   0.76 meter
                      (2.5 feet)  of XE-340  (EDlR).

-------
00
CO
                                                                     Av.  2.1
                                                                            Finished water

                                                                            Finished water thru 2.5 feet
                                                                                 (0.76 meter )    XE-340
              Figure ]_09  °' m ant^ p-Dichlorobenzene in finished water and removal by 0.76 meter
                          (2.5 feet)  of XE-340  (ED2).

-------
00
*>.
-
4-






3-



^H


2 -



••

1 -
«•


Days





















^
\
Lx4
4
1 1 7 l'l
P-i m-iv*^
? 4'8
1
1
1
1
! 0 Finished water
1
1 __Q-_ Finished water thru 2.5 f eet IRA-904 resin
1 (0.76 meter )
1
1
I
I
1
1
1
1
1
1
1
1
! A
! 0 \ 	 o AV. .8
I / \

VAy ^V^L, * v
m 18 21 25 28, 32 35 39 42 k& 49 52
Tin n^ m and r>— ni nVil nvnhpnT'.pne in f'inishfid wati<=r anrl i-oTn/->Tr=> 1 K™ n 7ft moi-eiT
                            (2.5 feet) of  IRA-904 resin  (ED3).

-------
CO
en
                       Finished water
                                                         (0.76 meter )
       Finished water thru  2.5  feet GAG
Finished water thru 5 feet  GAC
       (1.52 meters)
Finished water thru 7.5  feet GAC
       (2.29 meters)
Finished water thru 10 feet GAC
       (3.05 meters)
                                                      all  points nil
       0
     Days  03   71
           Figure  111.
            2«t 2831  35 38 "»2iT5 49 52 56 59 63 66 7077 80 8>4 87 91 94 98 101     112       122
        p-Dichlorobenzene in finished water and removal  by 0.76, 1.52, 2.29 and 3.05 meters
                            (2.5, 5,  7.5 and 10 feet)  of GAC (ED4).

-------
00
                    Finished water

                    Finished water  thru 2.5 feet GAG (0.76 meter )
                                             Finished water thru 5 feet GAG
                                                     (1.52 meters)
                                             Finished water thru 7.5 feet GAC
                                                     (2.29 meters)
                                             Finished water thru 10 feet GAC
                                                     (3.05 meters)
all poxnts ni
          "03  7  10  It 17  21  24 28  31  3538 42 45 4952  56  5963  66  7073 7780  8487  91  9498  101      112     122
           Figure112. o-Dichlorobenzene in finished Water and removal by -0.76,  1'.52,  2.29 and 3..05  meters
                       (2.5, 5, 7.5 and  10  feet)  of  GAC (ED4).

-------
test day 122 when a low level of passage  occurred.  A  large
spike in the influent concentration on  test  day  122 was noted.
XE-340, 0.76 meter  (2.5 feet) deep,  gave similar results  in
ED2, as indicated by the curves in Figure 109.   In ED3, Figure
110, the IRA-904 resin, 0.76 meter   (2.5  feet) deep, appears to
have a higher level of isomers in the column effluent  (2.7 times)
than in the influent.

     In ED4, each isomer was reported separately.  The influent
concentration and effluent  concentration  of  m-dichlorobenzene
was nil for all GAC bed depths.  The  average influent  concen-
tration, Figure 111, for p-dichlorobenzene was 0.24 yg/L.  The
effluent concentration for  all bed depths was nil for  all  test
points.  The average influent concentration, Figure 112, for
o-dichlorobenzene was 0.14  yg/L.  The effluent concentration for
all bed depths was nil for  all test points.

Adsorption by XE-340—

     We have presented much data showing  that XE-340 resin has
approximately three times the adsorptive  capacity for  individual
HOC as GAC.  The adsorptive capacity  of adsorbents is  usually
compared by measuring their capacity  to adsorb butane  gas.  The
adsorbent which adsorbs the most butane is usually considered to
have more adsorptive capacity for substances like HOC  than an
adsorbent which adsorbs less butane.  Butane gas phase adsorp-
tion data for the GAC and XE-340 used in  this study appear in
Figure 170.  The method of  obtaining  these curves and  their
interpretations are discussed in the  section of  the report in
which they appear.  For the moment we will say only that the
curves show that based on butane adsorption  data GAC should have
about seven times the adsorptive capacity of XE-340.   This
figure is arrived at by projecting a  vertical line, from 20 for
example on the "X" axis to  the Butane Gas Phase  (XE-340) and
Butane Gas Phase  (GAC) curves, and then at the intersect points,
reading the corresponding cc adsorbed per 100 grams of adsorbent
on the "Y" axis.  Values of 0.7 cc and  0.1 cc are obtained for
GAC and XE-340 respectively.  Thus GAC  adsorbs seven times as
much butane as an equal weight of XE-340. It has been shown by
Neely  (6) that XE-340 adsorption does not follow the usual pat-
tern of physical adsorption on GAC because in addition to
adsorption in micropores, substances  like HOC are taken into the
polymer matrix of the resin.  The resin matrix swells  as a
result of this incorporation.  Thus,  in our  tests, XE-340
exhibits approximately three times the  adsorptive capacity for
HOC as GAC because of adsorption into the polymer matrix.  We
will show later in the report, that for substances which are not
so readily taken up by the  polymer matrix, GAC has more adsorp-
tive capacity than XE-340,  as predicted by the order of butane
adsorption data.  These are the fulvic  acid  degradation sub-
stances which make up the bulk of substances measured  by TOC and
THM FP analysis, generally  known as precursors.

                               187

-------
Adsorptive Capacity and Competitive Adsorption—

     Figure 113 is a plot of the total volume in cc's  of  all  the
purgeable HOC in the finished water entering and leaving  each
of the four GAG beds in the 122-day test, ED4.  Integration of
these curves produces the three curves in Figure 114.  Curve  I
indicates the total volume in cc's of HOC entering each column.
Curve II indicates the total volume in cc's of HOC adsorbed by
each column and the cc's adsorbed per 100 grams of GAG. Curve
III indicates the total cc's adsorbed by 0.76 (2.5 feet),  1.52
 (5.0 feet), 2.29 (7.5 feet) and 3.05  (10 feet) meters  of  GAC.
In 122 days, the finished water entering column 1 contained
1.235 cc of HOC, the entire first column adsorbed 0.382 cc or
0.217 cc per 100 grams of GAC.  The next three columns each
received less and adsorbed less HOC than the preceding column.
When Curve II is projected to the "Y" axis the total adsorptive
capacity in cc's of HOC, 0.27 cc per 100 grams of GAC, for our
particular system is indicated.  Adsorption per column decreased
as the concentration of adsorbate decreased.  Therefore,  for
maximum adsorbent use, the greatest amount of adsorbent possible
should be in contact with the highest possible concentration  of
adsorbate.  These data support the generally accepted view on
carbon use.  The most efficient adsorbent usage would be  a
continuous in-out GAC system.  According to ,our data such  a
system would have approximately 43 percent greater adsorptive
capacity than a single 3.05 meters (10 feet) deep bed. Multiple
beds in series also offer advantages in better carbon usage.
General practice is to operate the series system until the last
bed in series reaches the effluent criteria.  The first bed is
then replaced with regenerated carbon and becomes the last bed in
the series arrangement.  The new lead bed is the one that  was
previously second in the series.  In this way, four columns,
each 0.76 meter  (2.5 feet) deep, arranged in series would have
about 31 percent more adsorptive capacity than one single  column
3.05 meters (10 feet) deep.  However, the actual design con-
figuration must take into consideration the higher capital costs
of these systems, as well as, the reduced operating costs.  The
lowest total cost system is the one desired.  Data described  in
this study are important in achieving the most practical design
(7).

     The curves in Figure 115 were obtained by integrating the
curves in Figure 113 at various sampling dates and more clearly
show the rates of saturation.  It is evident from the top  curve
in Figure 115 that column 1 will probably reach saturation at
the 0.27 cc value discussed above.

     We have already presented data suggesting that the adsorp-
tive capacity of GAC and XE-340 for cis 1,2-dichloroethene is
30 percent less in finished water than in raw water, probably
due to competitive HOC adsorption.  To aid the discussion  of
competitive HOC adsorption, Figure 116 shows the adsorption wave

                              188

-------
        125
oo
vo
                                                                                           Finished water enter-
                                                                                           ing column  1
                                                                                     — O — Column 1 effluent
                                                                                     •••D"" Column 2 effluent
                                                                                       A   Column 3 effluent
                                                                                     — •—Column 4 effluent
       Days
03  7 10 14 17  21 24 2831  35 38  4245 49 52  56 59 63 66  7073  7780  84  87  91 94  93 J  105   112
Figure 113.  Cubic  centimeters  adsorbed by each GAC column for all halogenated compounds
             added  together (ED4).
                                                                                                           122

-------
I-1
vo
o
        w
85
K +J
rH 
-------
                                                                                                          217
•a
o
0)
"0
u
u
                     GAG Column  #1



                     GAC Column  #2
                     GAC Column  #3




                     GAC Column  #4
            3   7  10 It 17  2124 2831  3538  4245  4952 5659 6366  70 737780  84 87 9194 98101    112      122


             Figure 115. Cubic centimeters of total HOC adsorbed per 100 grams of GAC in GAC columns

                        #1,  2, 3 and 4 in 122  days (ED4).

-------
fronts defined by breakthrough and saturation time for  several
HOC and Type II and III substances on a 0.76 meter   (2.5  feet)
deep bed of GAG.  The vertical height at the end of  each  HOC
curve represents the concentration in yg/L of each HOC.   The  '
concentration is read on the "Y" axis scale.  The "X" intercept
for each curve is the days until breakthrough occurs and  the
vertical projection of the end of each line to the "X"  axis is
the time until saturation for that substance.  The first  number
in the parenthesis also gives the breakthrough time  in  days and
the second number gives the time to saturation.  The Type II  and
III substances shown in Figure 116 are defined in the  discussion
of precursor removal that follows on page 192.  Finished  water
entering the first GAG column contains 7.8 mg/L Dissolved
Organic Matter  (DOM) of Type II substances and 0.58 mg/L  of Type
III, which corresponds to 4.69 and 0.36 mg/L respectively of  TOG.
The curve for the Type II substances in Figure 116 merely repre-
sents the initial breakthrough and saturation time of 0 and 16
days respectively.  The arrow at the end of the curve indicates
that the mg/L concentration cannot be shown on the yg/L scale
on the "Y" axis.  The dash-line curve representing the Type III
substances in Figure 116 merely indicates that these strongly
adsorbed substances have an initial breakthrough and saturation
time of unknown values which are much beyond the breakthrough
and saturation time for all the HOC studied in this work.  We
have calculated an average MTz for Type III substances of about
three inches.  Type III substances, which are adsorbed  by the
top portion of the GAG column, do not offer competition to
adsorption of all the HOC throughout most of the column.  On  the
other hand, Type II substances compete with HOC throughout the
column.

     Of the five HOC shown in Figure 116, chloroform encounters
the least competitive HOG adsorption and bromoform encounters
the most competition.  The results of this competitive adsorp-
tion are shown in Table 37.  These values show the predicted
adsorptive capacity of each of five HOC from pure water at
saturation for the GAG used in this study.  These values  are
based on calculations using the Polanyi Theory and modifications
by Manes and Hofer which are described beginning on page  283.

     Chloroform at the concentration in our finished water is
adsorbed 5 percent of its predicted capacity from pure water.
Bromodichloromethane, chlorodibromomethane and bromoform  which
are present in progressively decreasing concentration, never-
theless, have higher predicted capacities than chloroform.  How-
ever, the percent of predicted values steadily decreases (3.6,
1.8 and 0.12 percent) because of increasing competitive HOC
adsorption as indicated by the order of their wave front  curves
in Figure 116.   Cis 1,2-dichloroethene is present at a  lower
concentration than chloroform and has a predicted adsorptive
capacity, as shown in Table 37, less than chloroform (0.46 com-
pared to 0.68).   The observed adsorptive capacity is 6.5  percent

                               192

-------
                                DOM - Dissolved Organic Matter
                                TOC - Total Organic Carbon
                                          Time in days to breakthrough
                                               Time in days to saturation
                                               64
80
96
Figure 116.  Adsorption wave front defined by breakthrough and saturation time for HOC and
             Type II and Type III substances thru 0.76 meter   (2.5 feet) of GAC (ED4).

-------
of the predicted value.  These data appear to present no problem
until the wave front curve for cis 1,2-dichloroethene is con-
sidered.  The Manes-Hofer scale factor calculated from the
refractive index of a compound (described on page 284 ) in general
predicts the order of elution of the HOC compounds in our water
both from the GC Tenax column used for their analysis and from
the bench scale GAG columns on the water lines.  On the GC Tenax
analysis column, cis 1,2-dichloroethene elutes before chloroform.
On the GAG columns on the water lines, cis 1,2-dichloroethene
elutes after bromodichloromethane (Figure 116).  Because of its
wave front position in Figure 116, cis 1,2-dichloroethene
encounters more competitive HOC adsorption than chloroform, yet
in Table 37 the percent of predicted adsorption is higher than
for chloroform.  Apparently the refractive index scale factor
does not predict the stronger than predicted adsorption shown by
this compound.  Manes (private communication)  has indicated that
carbon tetrachloride exhibits less adsorption than predicted by
its scale factor.  Molecular geometry and other physical chemical
properties such as dipole moment can result in divergence from
the scale factor predicted value.  Carbon tetrachloride has a
dipole moment of zero and due to its molecular structure presents
a small surface area to the carbon surface compared to its molar
volume.  This results in less adsorption than predicted by the
Manes scale factor based on refractive index.   Perhaps the d6uble
bond in cis 1,2-dichloroethene has greater affinity for the
carbon surface than a single bond resulting in greater adsorp-
tion than predicted.  Our data also indicate that trans 1,2-
dichloroethene and trichloroethylene exhibit more adsorption
than predicted and a shift in the wave front as found with
cis 1,2-dichloroethene.

TABLE 37.'  OBSERVED ADSORPTIVE CAPACITY OF 100 GRAMS OF GAC FOR
           FIVE HOC FROM FINISHED WATER COMPARED TO THE POLANYI-
           MANES PREDICTED VALUE FOR EACH COMPOUND FROM PURE
           WATER (adsorbed by 0.76 meter  [2.5 feet] of GAC)

                       Polanyi-Manes  Observed
                         Predicted    Capacity       Percent
                       Capacity from    from           of
                        Pure Water   Finished Water Predicted
                            cc           cc
cis 1,2-Dichloroethene
Chloroform
Bromodichloromethane
Chlorodibromome thane
Bromoform
0.46
0.68
1.14
2.2
1.7
0.029
0.032
0.04
0.04
0.002
6.5
5.0
3.6
1.8
0.12
                               194

-------
     The percent of predicted values in Table  37 and their
relationship to the wave fronts shown  in Figure 116 apply only
for our finished water as it appeared  during this study.  Any
treatment plant modification that  changed the  ratios of the
component HOC would change all values  in Table 37.  We have
indicated that the maximum adsorption  capacity of 0.76 meter
(2.5 feet) of Filtrasorb 400 GAG for total HOC in our finished
water was approximately 0.27 cc .per 100 grams  (Figure 114).  We
expect that if the concentration of the four THM were reversed,
the maximum adsorptive capacity would  rise considerably above
the present 0.27 cc capacity.  All observed capacities and per-
cent of predicted values would change  in Table 37.

     In Table 37, the observed adsorptive capacity of chloroform
per 100 grams of GAC is 0.032 cc,  only 5 percent of the pre-
dicted value for pure water.  It would be interesting to know
how much of this reduction is due  to the competitive adsorption
of DOM  ( or TOC) and how much is due to other  HOC.  We can
probably determine this from our existing data.  Chloroform made
up approximately 98 percent of the entire volume of all the HOC
entering GAC column 4 during ED4 for up to 98  days.  We there-
fore can conclude that this column was receiving only chloroform
from finished water.  Integration  of the 3rd and 4th GAC column
curves in Figure 56 shows that the 4th GAC column adsorbed
0.093 cc of chloroform at saturation  (98 days), which corres-
ponds to 0.053 cc per 100 grams of GAC.  This  value is 8 percent
of the value that the Polanyi-Manes Theory predicts should be
adsorbed from pure water.  Therefore,  DOM  (or  TOC) substances
accounted for 92 percent of the reduction in adsorptive capacity
of chloroform from finished water  compared with its capacity
from pure water.  The competitive  effect of other HOC in fin-
ished water further reduces the capacity to 95.3 percent of the
capacity for pure water  (Table 37).  GAC column 1, had an adsorp-
tive capacity for chloroform of 0.0358 cc/100  grams of GAC.
Column 4 had a capacity of 0.053 cc/100 grams  of GAC, which is
32 percent higher than column 1.   This percentage value is very
close to the 30 percent increased  capacity found for cis 1,2-
dichloroethene in raw water compared with finished water, which
we feel is due to the absence of competitive adsorption of HOC.

     As shown in Table 25, the four bed depths of GAC in ED4
show increasing values of adsorptive capacity  for chloroform per
100 grams of GAC; 0.0358, 0.0449,  0.048, and 0.049 cc respec-
tively for each column.  This was  because the  adsorptive capac-
ity for chloroform increased in each consecutive column as the
HOC competitive adsorption decreased.

     The percent of predicted adsorptive capacity decreased as
the number of bromine atoms in the adsorbed molecular increased
 (Table 37).  The adsorptive capacity of these  bromine-containing
THM on GAC from pure water has not  been determined experimentally.
The Polanyi-Manes predicted values given in Table 37  for these

                               195

-------
three compounds are thus not confirmed values.  Therefore, we  do
not know if the decrease in percent of predicted value  for these
three compounds as shown in Table 37 is caused by only  increas-
ing competitive HOC adsorption or if part is due to an  error in
the predicted value itself.  For example, it is possible that  the
predicted values are too high because of steric exclusion result-
ing from the greater bulk of bromine atoms compared to  chlorine
atoms.  Further work is needed to determine the adsorptive capac-
ity of these molecules from pure water.

     The amount of total cubic centimeters of HOC adsorbed in  a
given column was always less than in a preceding column  (Figure
114).  It might first appear that the explanation for this is
simply that, as expected with a single HOC in water, the adsorp-
tive capacity decreases as the concentration in the column
influent decreases.  It is not that simple.  We have already
shown that if a single HOC in our system reached saturation in
all four columns (as chloroform did), the adsorptive capacity
actually increased from column 1 to column 4.  To explain the
results shown in Figure 114, one must consider the contribution
of each individual HOC.  Column 1 adsorbed more HOC than the
other columns because it was receiving more of certain  individual
HOC that had higher adsorptive capacities (Table 37).

     Throughout this study, we did not observe much roll-over
(displacement of an adsorbed substance by a more strongly
adsorbed substance).  There are several possible explanations
that may apply either alone or collectively.  Unknowns such as
this make prediction of adsorptive capacity difficult if the
ratios of HOC in our system were to change greatly.  We are not
without predictive capabilities, but we also recognize the
limitations of present methods and the direction further research
should take to improve this capability.


TOC and THM FP Organics

     Precursors in this study were measured by TOC and THM FP
analysis.  Both methods measure a complex family of compounds
instead of a specific substance as is the case for analysis of
individual HOC.  TOC includes some substances that are not THM
precursors at all.   It is likely that only a small portion of
the TOC represents precursors for THM.  If a single test method
measured the whole family of organics present, the resulting
adsorption breaktheough curve would look quite different from
that of an individual organic breakthrough curve.  This effect
is shown in Figure 113 where the total cubic centimeters of the
HOC mixture entering and adsorbed by 0.76 (2.5 feet), 1.52
(5.0 feet), 2.29 (7.5 feet) and 3.05 (10 feet) meters of GAC bed
are plotted.  After initial breakthrough the frontal adsorption-
zone has a steep slope, similar to that observed by a specific
HOC.  However, the total HOC curve then changes to a gradual

                               196

-------
slope .that approaches a plateau in  some cases.   During the  122-
day test period, none of  the  total  HOC  curves  reached the
influent level curve.  As the GAG bed depth increased,  the
difference between effluent and influent curves  increased.  This
was because the more strongly adsorbed  compounds were adsorbed
in various degrees up to  complete adsorption.  The  total HOC
curves in Figure 113 representing 19  specific  compounds are
similar to those obtained when TOG  and  THM FP  data, which
measure a larger number of substances,  are plotted  except that
the plateaus of the TOC and THM FP  curves are  more  pronounced.
Examples of such breakthrough are shown in Figures  117  and  118.
It is possible that the plateaus exhibited on  TOC and THM FP
breakthrough curves are the result  of biological degradation
occurring along with adsorption. Thus  the apparent plateaus on
the curvets for the TOC and THM FP may have partially  different
explanations than those for the total HOC.

     Figure 117 shows the THM FP in yg/L entering the first GAC
column and in the effluent from each of the four 0.76 meter
 (2.5 feet) deep columns connected in series.   Figure  118 shows
TOC data in mg/L of effluent  from the same four  GAC columns.
In Figure 118, the TOC breakthrough curves show  three distinct
areas.  There is some TOC breakthrough  from all  four  columns
right from the beginning  of initial flow.  This  breakthrough
appears to be about equal through all four columns.   It is
possible that this represents a nonadsorbable  fraction of TOC.
It appears that after this base line breakthrough we  observe a
rapid rise in the curves. This rapid rise begins further from
time zero as the bed depth increases.  These rapid  rise or
steep slope regions then  change into apparent  plateaus  for each
column depth.

     The curves in Figure 117 are replotted individually for
each column in Figures 119, 120, 121 and 122.  These  individ-
ually plotted THM FP curves more clearly show  the same patterns
shown by the TOC data in  Figure 118.  In Figures 119, 120,  121
and 122 we again have three distinct zones of  the breakthrough
curves.  Starting from a  base line  of a consistent  low level
THM FP passing through all four columns, there is an  initial
breakthrough period from  start of flow  that increases in time
as the bed depth increases.   The breakthrough  then  expands  into
a rapid rise or steep slope zone to a plateau.   The plateau can
be seen to occur at lower concentrations with  increase in column
bed depth.  These differences will  be discussed  later.

     The adsorption curve for THM FP adsorbed  from  raw water by
0.76  (2.5 feet) and 1.52  (5.0 feet)  meters of  IRA-904 resin is
shown in Figure 123.  The level of  THM  FP passing through the
1.52 meter  (5.0 feet) deep bed of  IRA-904 resin from start of
initial flow is about equal to the  value from  the 0.76 meter
 (2.5 feet) deep bed.  However, in the 1.52 meter  (5.0 feet)
deep bed, the steep slope portion of the curve begins four  days

                               197

-------
900-
800-
700-
600-
        Finished water
  ._O—  Finished water thru 2.5  feet GAG  (0.76 meter  )
 — A ^ Finished water thru 5  feet GAG    (1.52 meters)
 ...Jk"- Finished water thru 7.5  feet GAG  (2.29 meters)
        Finished water thru 10 feet GAG   (3.05 meters)
Days
                28         42      "   56           73           91        105      115
Figure 117. THM FP in finished water  and removal by 0.76, -1.52, 2.29 and 3.05 meters
            (2.5, 5,  7.5,  and 10  feet) of. GAG  (ED4) .

-------
vo
V£>
     a
                   Finished water
                   Finished water thru 2.5 feet GAG  (0.76 meter )
                   Finished water thru 5 feet GAG    (1.52 meters)

                   Finishes water thru 7.5 feet GAG  (2.29 meters)
                   Finished water thru 10 feet GAG   (3.05 meters)
       Days
                                 u>  ~.3
                                                                                                   Av. 3.21*
                                          it appears that TOG data on these two sampling
                                          dates are probably in error (too low)
                                                                                   *for plateau segment
7   m    21   29    35   42    H9     59  63  70   77   84  91   98   105
 Figure 118. TOG in finished water and removal by 0.76, 1.52, 2.29 and 3.05 meters
             (2.5, 5,  7.5,  and 10 feet) of GAG  (ED4).

-------
     0)
     -p
     -H
to
o
o
p.



04
       700
       600
       500
   400
       300
       200
       100
     Days
                                              Finished water


                                              Finished water thru 2.5 feet GAC  (0.76 meter )
                                                                                   *for plateau segment
            II   II  II	1  II—I	IT	I—I	T—I	IT

                                                70         84

Figure  119. THM FP  in finished water and removal  by 0.76 meter

            of  GAC  (ED4).
                                                                                98

                                                                            (2.5 feet)
                                                                                                112

-------
 
-------
to
o
        700'
        600-
        200 -
       100 -
	/V___  -M-----^MIZJLipi..._ '...-._ Jr.-" f-

/ V   *  \
                                  \    /*     \
        r\       ^  /f           \^&        A    A
        /  \  ^-A           ^               A

B      /  X^
 3     /  /             9   Finished water
1  A-Ar
•^ /  >^              —A— Finished water thru 7.5 feet GAC (2.29 meters)
                                                                                 \l
                                                                                 A
                                                                                                    250*
                                                                                 * for plateau segment
      Days
                                 28         42    '    56         70
                                                                                    98
                              112
             Figure  121.  THM FP in finished water and removal by 2.29 meters  (7.5 feet of GAC  (ED4) .

-------
                       Finished water
              —— O~—  Finished water thru 10 feet GAC  (3.05 meters)
to
o
u>
   TOO—.
   600 __
   500 —
   400 _
   300 —
   200_
   100 —
    Days
                                                                                                Av.

                                                                                             H.T.  water

                                                                                                531
                                                                                                Type III

                                                                                                substances
                                                     partial  saturation plateau
                                                                                 Type  II

                                                                                 substances
                             I	 Type I
                             T   substances
                                                                                                        229
          11 2&         1*2        56           73            93        105     115

Figure 122. THM FP in finished water and removal by 3.05 meters (10 feet)  of GAC (ED4)

-------
  soo
              Raw water
              Raw water thru 2.5 feet
        —D~" Raw water thru 5 feet
                                                     Av- 600
Days
                                                     Av.377*
                                 *for plateau segment

  7 10 Ik IB 21 25 2*8 32 '   \2
Figure 123.  THM FP in raw water  and  removal  by  0.76 and
             1.52 meters (2.5 and 5 feet) of IRA-904 resin  (ED3)
                                  204

-------
 after  a  period of initial breakthrough.  Again, partial satu-
 ration plateaus are reached and in this case they appear to be
 equal  for both bed depths, but the plateau for the 1.52 meter
 (5.0  feet)  deep bed occurs 14 days later than for the 0.76 meter
 (2.5  feet)  deep bed.

      Adsorption data tables were prepared for HOC adsorption
 curves.   A problem arises when this is attempted with TOC and
 THM FP curves.  The three zones of the curves and perhaps some
 idea  of  what produces these three zones should be taken into
 account.  Although the predominant substance that acts as pre-
 cursors  for THM formation have been cited as being humic acid
 or fulvic acid, it is reasonable to assume that such complex
 substances may adsorb like a family of different compounds
 possessing various adsorption affinities with a given adsorbent.
 Likewise, TOC can more easily be seen to comprise a family of
 different substances with varying adsorption affinities.

      THM FP data will be discussed by assuming that three types
 of substances comprise the precursors for THM in the Florida
 water tested.  The broad definitions for each type of substance
 is as follows:

      TYPE I   -   Substances that are not initially adsorbed
                   under the conditions tested.

      TYPE II  -   Substances that are initially adsorbed and/
                   or biodegraded within the adsorbent column,
                   but that eventually breaks through in
                   increasing concentrations under the
                   conditions tested.

      TYPE III -   Substances that are completely adsorbed and/
                   or biodegraded within the adsorbent column
                   under the conditions tested.

      Figure 122 shows the relationship of the three types of
 substances to the three zones of adsorption.  In this figure,
' the Type I substances are shown as the fraction that passes
 through  3.05 meters  (10 feet) of GAC, with a contact time of
 24.8  minutes from the start of initial flow of water through
 the column.  This could be a nonadsorbable fraction or a frac-
 tion  nonadsorbable at that concentration with the contact time,
 type  of  GAC used and other conditions of the test.

      Type II are substances that show a rather classic pattern
 of complete adsorption initially, followed by breakthrough and
 increasing effluent concentrations up to an apparent plateau
 value below equilibrium with the influent concentration.  This
 could be either a substance with a given adsorption affinity or
 a family of substances that have various adsorption affinities
 that  yield the effluent curve as shown.

                                205

-------
    Type III substances might be those that have a very strong
adsorption affinity such that they are completely adsorbed
within the adsorbent column for the duration that the study was
conducted, a biodegradable fraction of the THM FP, or a combi-
nation of both.  Either assumption could be used to provide a
possible explanation of the plateau shown on the curve.  We
have limited data to suggest that in our system as studied, that
strong asorption of the Type III substances  accounts for more
of the plateau level than could be accounted by biodegredation.
These data are shown in Figure 124.  The average values for the
plateau levels in Figures 119, 120, 121 and 122 which extend to
the end of the test period of 115 days, are presented again in
Figure 124.  At the end of the test period in ED4 , the columns
were allowed to operate further and two additional samples were
taken on days 176 and 177.  These two data points are plotted
in Figure 124 and their average value is shown.  For each
carbon bed depth, it is seen that the plateau level rises.  For
example, for 0.76 meter   (2.5 feet) of GAC, at 115 days, the
plateau level was 84 percent of the influent level and at 177
days it was 94 percent of the influent level.  This suggests
that the Type III substances from test day 115 to test day 177
are showing a typical breakthrough pattern.  From these limited
data we cannot determine how closely the breakthrough curve at
Type III saturation will approach the influent level.  However,
even at test day 177, it is apparent that biodegradation can
only account for a maximum of 6 percent of the influent level if
the Type III breakthrough curve levels off at the 177 day level.
We suspect that the breakthrough will continue further.  Since
the test was primarily designed for a four month period, the
frequency of testing beyond that time does not allow more than
speculation that the apparent plateau may be caused more by
adsorption than biodegradation.  It is obvious that we do not
know the true reason for the apparent plateau at this time.  It
is likely that both factors play a role.  Future studies will
collect data relative to factors influencing the apparent
plateau and ways, to enhance it.

    In evaluating data in Figures 119, 120, 121 and 122, using
the assumptions regarding the three types of substances, it is
apparent that the difference between the influent THM FP con-
centration and apparent plateau, which defines Type III,
becomes greater as the bed depth and contact time are increased
from 0.76 (2.5 feet) to 3.05 (10 feet) meters (6.2 minutes EBCT
to 24.8 minutes EBCT).  Concurrently the Type II substances
defined by the difference between the plateau concentration and
breakthrough concentration decreases with increased bed depth
and contact time.  The general definitions state that each
class of substance is defined at a given set of conditions and
each figure present a different condition.  However, the
changes in the amounts of Type II and III substances with
different contact times may be explained as the effect of
increased carbon volume and contact times that provide more

                             206

-------
     -H
         500-4
10
o
400-
300-
         200 _
                             Av. 365(84% of influent)  2.5 feet GAG
                                                    (0.76 meter )
                             Av. 312(72% of influent)  5 feet GAC
                                                    (1.52 meters)
                                                              00

                                                              Q
Av. 434  influent
Av. 407 (94% of  influent)
Av. 380(88% gf  influent)

Av. 338 (78% of  influent)
Av. 316(73% of  influent)
                             Av.  250(58% of influent)7.5 feet GAC (2.29 meters)
                             Av.  229(53% of influent} 10 feet GAC  p!o5 meters)
         100-
           DAYS
                   Figure 124.  Test extention data for THM FP removal by 0.76, 1.52, 2.29 and 3.05
                                meters  (2.5, 5, 7.5 and 10 feet) of GAC  (ED4).

-------
effective adsorption of Type II substances with adsorption
affinities closer to Type III substances while also more effec-
tively removing substances of weaker adsorption affinities.
One cannot dismiss the possibility that biodegradation is the
cause for the different plateaus.  As previously stated, it is
likely a combined effect that causes the plateau and we cannot
as yet discern a primary factor.

    Using this devised system of data presentation we have
assumed that Type I substances continuously break through at
constant levels and the Type III substances are completely
adsorbed by a specific bed depth and contact time at the same
level during the test period under the conditions of the study.

    This assumption allows analysis of the TOC and THM FP data
and presentation in tabular form such as presented for the halo-
genated organic compounds.

Raw Water Source—
    THM FP adsorption data appear in Table 38.  Adsorption
curves appear in Figures 125, 126 and 127.

    EDI data are plotted in Figure 125.  Since collection of
data began on test day 61, complete analysis cannot be made on
this ED.  Comparison of the average values from day 61 to the
end of the test period can be made from Figure 125.  The
average influent THM FP was 816 yg/L.  The average effluent
from the 0.76 meter  (2.5 feet) deep GAG column was 757 yg/L,
a reduction of seven percent.  The effluent from the 0.76
meters  (2.5 feet) deep XE-340 column was 833 yg/L, about two
percent higher than the influent.

    Column 0.76 meter  (2.5 feet) deep of GAC and XE-340 were
studied in ED1R.  Adsorption curves appear in Figure 126.  In
Figure 126, average values are shown for the influent and for
the column effluents from a point after Type II saturation on
the plateau portion of the adsorption curves.  Considering
these average values, from test day 17 to the end of the test
period, GAC and XE-340 removed 24 percent and 20 percent
respectively of the influent THM FP.  Integration of the curves
for the whole test period produced the results in Table 38
(ED1R).  During ED1R 0.76 meter  (2.5 feet) of GAC adsorbed 29
percent of the THM FP precursors from entering raw water over
the entire test period.  Type I substances represented two
percent; Type II, 72 percent; and Type III, 26 percent of the
total precursors entering.  Also during ED1R, 0.76 meter
(2.5 feet)  of XE-340 adsorbed 24 percent of the precursors from
entering raw water over the entire test period as compared with
29 percent adsorbed by the same bed-depth of GAC.  One hundred
grams of GAC adsorbed 1.5 times as much precursor substances as
100 grams of XE-340 during the 119-day test period.  When equal
volumes of the two adsorbents are compared, at 119 days, GAC

                              208

-------
                        TABLE 38.   THM FP ADSORPTION DATA FROM RAW WATER
O
10
O\ *O 13




ED
1

1
1R

1R
3
3






•O ft
33
Feet
2.5

2.5
2.5

2.5
2.5
5






Adsorbent

GAG
XE-
340
GAG
XE-
340
904
904




S
J)
3
H
>H
a
H

% of Total
Enter in
%
ot ca


29

24
(34)
46
55



^
O^g
CO < -rl
Sg«
sIB
SUM
HAH
0 H
H id
(d w CD
•M Oi
P >»>*
6 ffl £
5rams
culate

"
.37

No Sat
.59
1.14




04
fa
0*
o
IIP
*



5


22
43


conti

Type III
Substances

Adsorbed
per Column
Grams



1.85

1.71
(.76)
.98
.99


ued)

3 g
S Of -r
fa >•
M-I a
o § -P
frj j
dp P t
%



26

24
(26)
37
37



           2.5 feet=6.76 meter   5 feet=1.52 meters

-------
                                 TABLE 38%(continued)







ED
1
1

1R
1R

3
3








•O ft

Feet
2.5
2.5

2.5
2.5

2.5
5








sorbent
S

GAG
XE-
340
GAG
XE-
340
904
904




_4
&
i_l
t-l O
O cn
*S
%



4
0

45
44



£
t.
rtj
rH 4J
ro G
| w
* 1
%



72
74

(72)
45
45



Type I
Substances

*o






Type II
turation
•P 10
m w
Grams



.21
No Sat

.22
.21






at
!
%
Grams



.56
.33

.44
.26




Per
Col.



w
!
%
Grams



.99
.71

1.21
1.43



2.5 feet=0.76 meter   5 feet=1.52 meters

-------
to
       1100-
       100CL.
        900-
        800-
        700_
      0)
      4J
      tn
        60CH
        500_
        400-
        300-
         Raw water  thru  2.5  feet GAC  (0.76  meter )

  — A— Raw water  thru  2.5  feet XE-340 (0.76  meters)

  "" 9 	•  Raw water
        Days
                                                                                         reduction
                                            _ 61 64  69717678  83   9092  97

Figure 125.  THM FP in raw water and removal by 0.76 meter (2.5 feet) of GAC and

             0.76 meters (2.5 feet)  of XE-340 (EDI).
117

-------
   900
                                                                                                  20%
                                                                                              reduction
                                                                                           : 527   •
                                                                                                 24%
                                                                                             reduction
                                •    Raw water

                                     Raw water thru 2.5 feet GAC  (0,76 meter )
                                 	 Raw water thru 2.5 feet XE-340  (0.76 meters)
Davs   0          14        28         42         56         70          84         98         112

       Figure 126.  THM FP in raw water and removal by 0,76 meter   (2.5 feet) of GAC and

                    0.76 meter  (2.5 feet) of XE-340  (EDlR) .
                                                                                                11 9

-------
900
800
                Raw water
                Raw water thru 2.5 feet
                Raw water thru 5 feet
                                                         38% reduction


                                                         45% reduction
Davs    Figure  127
                          28         "*2    "»9
                    THM FP  in raw water  and removal by  0.76  and  1.52 meters  (2.5 and  5  feet)  of
                    IRA-904 resin (ED3).

-------
adsorbed 1.2 times as much precursor substance as XE-340.   XE-
340 had about three times the capacity of GAG for HOC adsorption.
If Type II substances have an initial breakthrough and  satura-
tion time through XE-340, the values are less than three days
when the first datum point after initial flow was obtained.

     In EDS, 0.76 meter   (2.5 feet) of IRA-904 resin was evalua-
ted for 49 days on raw water compared with a test period of 119
days for adsorbents used in EDlR.  Since the THM FP influent
levels of the EDlR and EDS are quite close, 659 yg/L and 600 yg/L
respectively, the two sets of data can be compared.  For pur-
poses of comparison, integration values for 0.76 meters  (2.5
feet) of GAG at 49 days in EDlR are shown in parentheses in
Table 38.  At 49 days 0.76 meter   (2.5 feet) of GAG adsorbed 34
percent and 0.76 meters (2.5 feet) of IRA-904 resin adsorbed 46
percent of the precursors entering each column.  A bed of 1.52
meters  (5.0 feet) of IRA-904 resin adsorbed 55 percent of pre-
cursors from raw water during EDS.  The adsorption curves for
both IRA-904 resin bed depths are shown in Figure 127.

     We have no way of knowing the status of the IRA-904 resin
plateau after the end of the 49-day test period.  If it remained
at the same level for the same number (119) of test days as in
EDlR, 41 percent would have been adsorbed compared with 29 per-
cent for GAG.  In Figure 127 the plateau levels are equal for
both bed depths of IRA-904 resin.  We can speculate, but have no
explanation for the two levels being equal.

     Initial breakthrough and saturation times for the Type II
substances were longer in the 1.52 meter   (5.0 feet) deep IRA-
904 resin bed than in the 0.76 meter  (2.5 feet) deep bed.  The
level of Type I substances was the same through both bed depths
of IRA-904 resin.

     The last four columns of data in Table 38 present the ad-
sorption of total precursors per 100 grams of adsorbent at the
ned of the test period, at Type II saturation, at a common point
in time of 49 days, and adsorption per column at 49 days which
was the shortest test duration.  If 0.76 meter  (2.5 feet) of
GAG, 0.76 meter  (2.5 feet)  of IRA-904 resin, and 1.52 meters
(5.0 feet)  meters of IRA-904 resin columns were all regenerated
at their respective Type II saturation times of 10, 18, and 34
days, adsorption per 100 grams of adsorbent would be similar;
i.e., 0.21, 0.22, and 0.21 grams respectively per 100 grams of
adsorbent.   If all adsorbents tested (0.76 meter  [2.5 feet] of
GAG, XE-340, and IRA-904 resin, and 1.52 meters (5.0 feet) of
IRA-904 resin)  were generated at the same time, 49 days, adsorp-
tion in grams per 100 grams of adsorbent would be 0.56, 0.33,
0.44, and 0.26 respectively.  GAG removed the most total precur-
sor per 100 grams of adsorbent.  The last column of values  in
Table 38 are amounts of precursors adsorbed by the entire column
of each adsorbent containing equal volumes of the adsorbents.

                              214

-------
When equal volumes of the  three  adsorbents  (0.76 meter   [2.5
feet] of GAC, XE-340 and IRA-904 resin)  are  considered,  the ad-
sorption from raw water at 49 days  is  0.99,  0.71,  and 1.21 grams
respectively.  When compared on  a volume basis, IRA-904  resin
adsorbes more precursors than GAC.   However,  the initial break-
through concentration of precursors was  the  highest  for  the IRA-
904 resin at about 100 yg/L at start-up, and the lowest  at start-
up was obtained by GAC.  At this point,  no meaningful conclusions
can be made regarding the  total  merits of the three  adsorbents
applied to raw water without a specific  objective  in mind and
research on regeneration of each adsorbent and the accompanying
cost.

     In 404 days of testing, we  found  that lime softening of raw
water removed an average of 27 percent of the THM  FP.  This com-
pares favorably with THM FP removal of 29 percent  by GAC and 24
percent by XE-340 for 119  days by 0.76 meter  (2.5 feet) deep
beds.  A 0.76 meters  (2.5  feet)  deep bed of  IRA-904  resin remov-
ed 46 percent of the precursor from raw  water for  49 days when
the test ended.  If all three adsorbents were assessed at the
end of 49 days, the expected results would be removal of 34 per-
cent by GAC, 24 percent by XE-340 and  46 percent by  IRA-904
resin.  As is the case with HOC  results, the shape of the adsorp-
tion curves affect data interpretation.  Choosing  a  reference
point for comparison of adsorbent capacity is important.

     The flocculated calcium carbonate in the H.T. can be con-
sidered as still another adsorbent  and compared with our column
adsorbents.  Earlier in the report, it was shown that adsorption
occurring on calcium carbonate floe in the H.T., under normal
operating conditions of the Preston Plant, removed 0.07  grams of
THM FP per 100 grams of floe.  The  0.76  meter  (2.5  feet) deep
GAC column in EDlR, Table  38, at Type  II saturation  removed 0.21
grams of THM FP per 100 grams of GAC.  It is interesting to note
that the type of adsorption occuring on  calcium carbonate floe
removed about 33 percent of the  precursors,  as measured  by THM
FP _as was removed by GAC on a weight basis.

     TOC adsorption data from raw water  appears in Table 39.
Adsorption curves.appear in Figures 128  and  129.

     In Figure 128, from the start  of  initial flow through the
XE-340 column, essentially no TOC was  removed.  Over the entire
test period an average of  all test  data  indicates  that the XE-
340 column removed only two percent of the influent  TOC.  The
GAC column removed 8 percent of  the influent TOC,  calculated
after Type II saturation.   In Table 39,  the  results  obtained by
integration of the entire  adsorption curves  are presented, GAC
and XE-340 removed 12 percent and two percent of the influent
TOC respectively.  The same columns removed  29 percent and 24
percent respectively of THM FP.   The GAC column removed  10.2
times and 8.4 times as much TOC  as  the XE-340 column on  an equal


                               215

-------
                                 TABLE  39.  TOC ADSORPTION DATA FROM  RAW WATER
ro










ED
1R
1R
3
3













,f
W S
Feet
2.5
2.5
2.5
5











iiJ
C!
fi
o
01
a

GAG
XE-
340
904
904






•P
C
0)
3
H
tl_j
£
{•^

g>
2 „

35
C M
Q

CJ
y B 4J
C^ 9 01
S H fl)
H
g ^
Grams
111.1
111.1
40.7
40.7





•d
2
M
O -P
01 <;
»=«; s -p
5 M

(Hi C^
80
O *H
f o
43
o -d
H Rj C
nJ W H
4J
S&
Srams
13.2
2.3
10.5
15.4






8
EH

ty*
3C
*rH
j^l
0 (U
EH -P
c
o

*
12
2
26
38





1
S-P O
(4! -H
K to
3 Ej
H JJ
U O <0
^^ CJ Ul
£H
/3 H
O M
H ID
flj |JF^ (JJ
S &6*
Srams
5.1
No
Sat.
9.2
14.6






0
EH

Cn


•P M
O (U
O
dP
*
5

23
36




(cont

Type III
Substances



S
i^
•d 3
^3
O
CO M
a &
Grams
9.1
2.3
2.4
2.8




.nued)



r-|
id
-P O
WO)
* s i
*
8
2
6
7





            2.5 feet=0.76 meter   5 feet=1.52 meters

-------
                                  TABLE  39.  (continued)
to
ED
1R
1R
3
3


^••^••••••••••••••••••H

•O Oi
ss
Feet
2.5
2.5
2.5
5


••••••••^^••^•^••••WBBI

Adsorbent
GAC
XE-
340
904
904




4J
1
H
 niaav/a.j^«=
Per 100 grams
of Adsorbent
+J
(0
e
m
o
•s
0)
8
Grains
7.5
1.1
3.8
2.8


•***
**
8
Grams
4.3
.42
3.8
2.8


"•^•MW^^^HVOH


Per
Col.
(0
1
o»
<*
8
Grams
7.6
.9
10.5
15.4


•Aq^^^^^^^^^^^

               2.5 feet=0.76 meter    5  feet=1.52 meters

-------
Days
                                                                 2% reduction
                                                                 8% reduction
                       O — Raw water thru 2.5 feet GAC   (0.76 meter  )

                            Raw water thru 2.5 feet XE-340  (0.76 meter  )
to
m
yg/liter
1
/
/
/
O
i
\ 1
V
      037
Figure 128.
                                                                          98   105   112   119   127
                   TOC in raw water and removal by 0.76 meter   (2.5 feet) of GAG and
                   0.76 meter   (2.5 feet) of XE-340  (ED1R) .

-------
M
0)
•M
•H
     11-
    10-
    9 -
    8 -
7-
     6-
      5-
      4-
      3-
      2-
      1-
                                                        Av.  8.6
                                                          Raw water

                                                   — D — Raw water  thru  2.5  feet IRA-904 resin
                                                                          (0.76 meter)
                                                          Raw water  thru  5  feet  IRA-904 resin
                                                                          (1.52 meters)
     Days
                11
               18
25
32
3
-------
volume and equal weight basis respectively.

     In EDS, Table 39 and Figure 129, 0.76  (2.5  feet) and  1.52
(5.0 feet) meters of IRA-904 resin removed  26 percent and  38 per-
cent of the influent TOG respectively.  Comparing 0.76 meter
(2.5 feet) of IRA-904 resin with 0.76 meter   (2.5 feet) of GAG
at an equal time of 49 days, the resin removed 0.9 times and 1.4
times as much TOG as GAG on an equal weight and  equal volume
basis.

H.T. Water Source—
     THM FP adsorption data appear in Table 40.  Adsorption
curves appear in Figures 130, 131, and 132.

     XE-340, 0.76 meter  (2.5 feet) deep, was studied in EDl and
adsorption data shown in Figure 130.  Since collection of  data
began on day 61, complete analysis is not possible.  From  day 61
to the end of the test period, the XE-340 column removed four
percent of the influent THM FP.  On raw water, the XE-340  column
removed seven percent during the same time period.  However,
while we will continue to compare THM FP and TOG adsorption
across water sources, raw, H.T. and finished, it really is not a
valid way to interpret these data.  We will show in the discus-
sion on Finished Water Source, which follows later, that one
cannot compare directly across water sources when the influent
levels of THM FP and TOG change.

     In ED1R, Figure 131, 0.76 meter  (2.5 feet) of XE-340 was
studied from initial column start-up.  The XE-340 column removed
two percent of the influent THM FP compared to 24 percent  remov-
al (Table 38) from raw water.

     In ED3, 0.76 meter  (2.5 feet) of IRA-904 resin was studied,
Figure 132 and Table 40.  The IRA-904 resin removed 32 percent
of the influent THM FP compared to 46 percent from raw water.
In Figure 132, the adsorption curve suggest that additional Type
II substances are being removed from H.T. water but no Type III
(at Type II saturation the influent and effluent curves coin-
cide) .

     TOG adsorption data appear in Table 41.  Adsorption curves
appear in Figures 133 and 134.

     In ED1R, 0.76 meter  (2.5 feet) of XE-340 removed six per-
cent of the in-fluent TOG compared to 24 percent  from raw water.
In ED3, 0.76 meter  (2.5 feet) of IRA-904 resin  removed 41 per-
cent of the influent TOG compared to 46 percent  from raw water.

Finished Water Source—
     THM FP adsorption data appear in Table 42.  Adsorption
curves appear in Figures 135, 136, 137, 138, 139, 140, 141, and
142.


                              220

-------
                         TABLE  40.  THM FP ADSORPTION DATA FROM H.T. WATER
ED
1
1R
3






•O ft
SS
Feet
2.5
2.5
2.5






Adsorbent
XE-
340
XE-
340
904






g Average Influent
j> THM FP
580
474
394






Type II
Substances
jj? Column
"^ Breakthrough

<3
0






fi? Column
'to Saturation
Par
<3
39






MT
z
Inch
:ial ru
30+
30






p? Test
"m Duration
i onlj
119
49






$ Total THM FP Entering
| Each Column During
5 Test
- car
5.0
1.72






ft Total THM FP Adsorbed
§ By Each Column At
01 End of Test
not c<
.12
.55






S
H£
$V
o M W
O

%
Iculal
2
32






$ Total THM FP Adsorbed
| By Each Column At
» Type II Saturation
Q
No Sat
.55






&
tp
•S-S
•p M
gs
c

-------
                                        TABLE 40.(continued)
ED
1
1R
3






i!
Feet
2.5
2.5
2.5






Adsorbent
XE-
340
XE-
340
904






jg Average Influent
£> THM FP
580
474
394






Type II
Substances
H Adsorbed
§ per Column

0
.55






i? Passed
I Each Column

4.88
.83






i? Total Entering
p Each Column

4.88
1.38






% of Type II
* Adsorbed

0
40






% of Total
THM FP Entering

98
80






Type I
Substances
tl

-------
N)
to
U)
    1000-




     900-




     800-



 0)
.tJ   700-i
rH
\
3.

fc   600.




     500,




     400
                                      1047
                                                                                          4% reduction
                                                     Hydrotreator water
                                                     Hydrotreator water thru 2.5 feet XE-340
                                                                       (0.76 meter .)
                                  6^71 7*678  8*385  ^0^2  9*7
                                                               U7
                 Figure  130.  THM FP in raw water,  and removal by lime softening and by
                               0.76  meter  (2.5 feet) of XE-340 on H.T. water  (EDl).

-------
M
(O
     M
     (0
     4-1
     •H
     Cn
     CM
     En
           900
           800
           700
           600
          500
          400
          300
          200
          100
            Raw water

            H.T. water

            H.T. water thru 2.5 feet of  XE-340  (0.76 meter )
               0  3
              Days
7 10 1417  212"+ 28313538  42 »»5, 4952  56    6366 7073  77   84      9'4  98  105  108112115119

 Figure 131.  THM FP in raw water and removal by lime softening  and by
               0.76 meter   (2.5 feet) of XE-340 on H.T. water  (EDI).

-------
to
to
tn
        10
Raw water


H.T. water


H.T. water thru 2.5 feet

IRA-904 resin (0.76 meter )
            047  1114 1821  2528  3235 3942 4649 53

            Figure  132.   TKM FP in  raw water and removal by lime softening and by 0.76 meter
                          of IRA-904 resin on H.T. water (EDS).
             (2.5 feet)

-------
TABLE .41.  TOC ADSORPTION DATA FROM H.T.  WATER
ED
1R
3







$
•O D.
3S
Feet
2.5
2.5







Adsorbent
XE-
340
904







|j Average Influent
^ TOC
6.8
6.0







Type II
Substances
jj? Column
*Jj; Breakthrough
?
0



•



p> Column
"iw Saturation
?
46







MT
z
Inch
•)
•
30







tu Test
to Duration
127
53







j^ Total TOC Entering
H Each Column During
to Test
77.1
28.4







•a
.8
n
O 4J
(0 <
•O
•4! S +J
1 co
3 i
EH ffl
Srams
3.1
11.3







% of Total TOC
* Entering
4
40







ft Total TOC Adsorbed
fi By Each Column At
™ Type II Saturation
No
Sat.
10.9







o
% of Total TO
* Entering

38






(conti
2.5 feet=0.76 meter
Type III
Substances
O
g Adsorbed
3 per Column
ul
3.1
3.1






nued)
% of Total
*> TOC
Entering
4
11









-------
                                   TABLE 41.   (continued)
to
N)
ED
1R
3







•d ft
0) 0)
n a
Feet
2.5
2.5







Adsorbent
XE-
340
904







•M
C
Q)
5

TOC Entering

65







Type I
Substances
$
§ Total Passed

6.6







% of Total
* TOC Entering

14







j.ww fluawj.M'c;
Per 100 grams
of Adsorbent
•P
to
u
IH
o
•s
0)
^
Grams
1.4
4.1







n At Type II
§ Saturation

2.9







01
1
0\
•&
%
Grams
.6
4








Per
Col.;
(0
1
en
^
$
Grams
1.2
11.1







               2.5 feet=0.76 meter

-------
       9 -
NJ
N>
00
       6 _
     0) c
     u ->
I
     84 -
       3 -
       1 _
       0
                                                                                                          4%
                                                                                                       reduction
                                   H.T. water

                                   H.T. water through 2.5  feet  XE-340  (0.76 meter  )
            37     11+21  28     35   42               64   71   78   84    91    93   105   112  119
       Days       Figure 133. TOC  in H.T. water and removal  by 0.76 meter   (2.5 feet)  of  XE-340 (ED1R) .
                                                                                                      127

-------
     in
     0)
     +J
     •H
to
to
                                                                  Av. 6
                                                                   H.T. water


                                                                   H.T. water through 2.5 feet

                                                                   IRA-904 resin (0.76 meter)
          Days     Figure 134.  TOC in H.T. water and removal  by 0.76 meter (2.5 feet)

                                of IRA-904 resin  (ED3).

-------
                           TABLE 42.  THM FP ADSORPTION DATA FROM FINISHED WATER
Ul
o












ED

1


1R

3
3
4
4
4
4


•







rC
•^
•o a
S3
Feet

2.5


2.5

2.5
2.5
2.5
5
7.5
10







+J
a
0)
.Q
M
O
CO
S

XE-

340
XE-

340
GAC
904
GAC
GAC
GAC
GAC


t!
2
3

IH
a
H

a)
Cn CM
fd p4
M
<1) §
^^ nQ
ug/L

451


355

274
274
434
434
434
434


Type II
Substances


fi
0^
3
O

S +*
K f^
3 (O
rH 
C
MH W
O
%

.ate


3

18
13
22
39
53
60

13 	
JS
Vt C
O 4J O
M ft -H
*d -P

C 1 1
CM 9 0
i4 «— 1 -M
O rd
2
EH X H
O H
H Id
(d PQ Q)
S&&
Grams






.14
No Sat
.36
.84
1.15
1.41


PM
S
hM
£-4
tn
^-S
•P M
£$
C
M-l FT1
O
*






12

8
19
26
32
(conti

Type III
Substances




S
'0 3

M U
o
CO M
Grams




.138

.11
.15
.71
1.25
1.89
2.11
iued)



H
id
•P tr
o c
EH CM -H

m a)
O 2 4J
<*> EH W
*




3

9
13
16
28
43
48

               2.5  feet=0.76 meter   5 feet=1.52 meters  7.5 feet=2.29 meters  10 feet=3.05 meters

-------
                                           TABLE 42. (continued)
to
co












ED
1

1R

3
3
4
4
4
4










fj
•C ft
0) 0)
PQ Q
Feet
2.5

2.5

2.5
2.5
2.5
5
7.5
10







.p

A
fc
o
CO
s

XE-
340
XE-
340
GAG
904
GAG
GAG
GAG
GAG


-P
0)
3
H
M-l

H

9
cr> PJ
jd fa
4) g
rt! EH
ug/t
451

355

274
274
434
434
434
434

Type II
Substances



£4
B
•d 9
as
ou
(0 N
"O flj
ri! ft
Grams




.11
p
.25
.49
.49
.57



rt
c
2
H
O
tJ CJ
0)
W JS
to U
« id
cu w
Grams




.89
?
3.39
2.61
1.97
1.67

C
•H

® I

rt i-«|
w g
H
id ^:

O id
EH W
Grams




1.0
?
3.64
3.1
2.46
2.24


H
H

(U

S a)
IH O
o S
%




11
'
7
16
20
26

H1
•H
I.

iHI Jj
dlj ^
4-* ri
EH P.
M-l
O 21
* 1
%




83
?
82
70
55
50

Type I
Substances

•o
0)



PM
H
id

S
Grams




.09
?
.11
.11
.11
.11

£
-H

d)
3jj
*^
.T2
g o,
h
O 51
* 1
%




8
?
2
2
2
2


Per 100 grams
of Adsorbent
W
 rd
ft, CO
Grams




.08
No Sat
.2
.24
.22
.2




CO
^1
(d
-0
5

<5
Grams


.032

.13
.07
.3
.29
.25
.21

Per
Col.



CO
^i
id
•O
5

<
Grams


.057

.23
.19
.53
1.02
1.32
1.48

             2.5 feet=0.76 meter   5 feet=1.52 meters  7.5 feet=2.29 meters  10 feet=3.05 meters

-------
to
00
to
       800  -I
        700  -
        600 -
     U
     0)
     4J
     •rl
        500 -
 p.


 PH
        400 -
        300 -
        200 -
                                   Finished water


                                   Finished water  thru 2.5 feet XE-340  (0.76 meter 0
                                                                                                     Av. 521
        100 -
Days w  '  "                                           61 6\  6971  7^788385  90 ^2 97     Tl
           Figure 135. THM FP in finished water and removal by 0.76 meter  (2.5 feet)

                        of XE-340 (EDI).
                                                                                                    117

-------
       800
       700
to
CJ
                                                                                                      Av.
                                                                                                   H.T. water
                                                                                                      471
                                                                                                  23% reduction
                                                                                                     362
                                                                                                           349
                                                                                                     3% reduction
                        Finished water

                        Finished water thru 2.5 feet XE-340  (0.76 meter )
              03  7 10 l«t 17  2124  28  31  3538 4245 4952 56   63 66 7073  77    84       94  98         112   119

                 Figure  136. THM FP in finished water and removal by0.76 meter   (2.5 feet)  of XE-340  (ED1R)

-------
       800
       700
       600
       500
    Finished water

    Finished water thru 2.5 feet GAC (0.76 meter: )
    Finished water thru 2.5 feet IRA-904 resin  (0.76 meter )
     tn
to
       400
       300
       200
       100
      Days
                                                                Av.  274

                                                                Av.  230
                                                                16%  reduction
                      T
                    11 14
-i—i—n—r~i—n—r—r
 18 2125 28  3235  3942  4649
T
53
                Figure 137.  THM FP in finished water  and removal  by 0.76 meter
                             and  0.76 meter  (2.5 feet)  of IRA-904 resin (&D3).
                                                         (2.5  feet)  of GAC

-------
                     Finished water
                     Finished water  thru  2.5  feet GAC (0.76 meter  )

                 _  Finished wate*  thru  5  feet GAC (1.52 meters)

             ».A-'<  Finished water  thru  7.5  feet GAC (2.29 meters)

             _.Q._  Finished water  thru  10 feet GAC (3.05 meters)
700
                                                                                              Aye.
Days o         1*          28         «t2         56            73            91         105     115
    Figure  138. THM FP in finished water and removal  by 0.76, 1.52, 2.29 and 3.05 meters
                  (2.5, 5, 7.5, and 10 feet) of GAC (ED4).

-------
                           Finished water
to
U)
en
                                           thru 2.5 feet GAC (0.76 meter' )
        Days 0          I't       ~  28  "   ~  ~  
-------
                          Finished water

                          Finished water thru  5  feet GAC (1.52 meters)
to
10
       600-
     M
     
-------
                        Finished water

                    r... Finished water thru 7.5 feet GAG (2.29 meters)
        7001
to
oj
00
                                                                                               Av.

                                                                                            H.T. water

                                                                                               531
            0
        ll         28        42         56           73    '        93        1051  'll5

Figure  141. THM FP in finished water  and removal  by 2.29 meters (7.5 feet) of GAG  (ED4).

-------
NJ
U)
       700-.
                  9   Finished water

                —O— Finished water thru 10 feet GAC (3.05 meters)
                                                                      Type III Substances
                                                          Saturation
                                      Type I Substances
                                                                                              *for plateau
                                                                                               setmen
      Days
0         14        28         42         56         70          at         93          112
  Figure 142. THM FP in finished water  and  removal  by  3.05 meters  (10 feet)  of  GAC (ED4) .

-------
     In EDI, Figure 135, from day 61 to the end  of  the  test per-
iod, the average THM FP effluent from the  0.76 meter   (2.5  feet)
deep XE-340 column was greater than the influent.   In ED1R,
Figure 136, the average THM FP effluent from the 0.76 meters
(2.5 feet) deep XE-340 column was three percent  below the influ-
ent, indicating essentially nil removal from finished water.  Ad-
sorption curves from ED3 for 0.76 meter   (2.5 feet) of  GAG  and
IRA-904 resin appear in Figure 137.  It is interesting  to note
that the GAC column is removing additional Type  II  substances,
as evidenced by the 0 and 7 day test point portion  of the adsorp-
tion curve.  From time 0, this portion of  the adsorption curve
is absent from the IRA-904 resin column.  After  Type II satura-
tion, the average THM FP reduction was 10 percent for GAC and 16
percent for IRA-904 resin (calculated from time  0).  Because of
the Type II portion of the adsorption curve, the GAC column  at
the end of the test period removed 18 percent of the influent
THM FP compared to 13 percent for IRA-904 resin  (Table  42).  At
a common point in time of 49 days, GAC removed 1.9  times and 1.2
times as much THM FP as IRA-904 resin on an equal weight and
equal volume basis respectively.

     Unlike the 0.76 meter  (2.5 feet) deep bed  of  IRA-904 resin
evaluated in ED3 on finished water, the carbon bed  showed a
typical Type II substance removal zone.  The time of Type II
saturation was only 11 days.  However, of the three adsorbents
tested on finished water, GAC was the only one to exhibit low
Type I bleed, some Type II adsorption and greater total precur-
sor removal.  Because of this, GAC was selected  for study in ED4.
Four 0.76 meter   (2.5 feet)  deep columns were connected in
series providing carbon bed depths of 0.76 (2.5  feet),  1.52  (5.0
feet), 2.29 (7.5 feet) and 3.05 (10 feet) meters.  THM  FP ad-
sorption curves for all four bed depths for ED4  appear  in Figure
138.  The curves for individual bed depths appear in Figures 139,
140, 141 and 142.  The data from the 0.76 meter   (2.5 feet)   deep
bed in ED4 can be compared with the data from the same  bed depth
in ED3.  The average influent level of THM FP was 274 yg/L in
ED3 and 434 yg/L in ED4.  Total precursor adsorption at Type II
saturation was 0.08 grams per 100 grams of carbon in ED3 and 0.2
grams per 100 grams of carbon in ED4,  This indicates that ad-
sorptive capacity for precursor substances increases as adsor-
bate concentration increases, as was found with  HOC adsorption.
GAC beds, deeper than 0.76 meter  (2.5 feet), ED4 showed in-
creased Type II breakthrough and saturation time; i.e., 14 and
46 days respectively in the 3.05 meters  (10 feet) deep  bed.
Total THM FP adsorption data for the four columns are summarized
in Figure 143.  Column 1 received 4.46 grams of  THM FP substances
and adsorbed 0.97 grams.  The other columns received and adsorb-
ed less.  However, about 20 percent of the influent THM FP to
each column was uniformly removed.  The THM FP adsorbed per  100
grams of carbon is also shown in Figure 143 for  each 0.76 meter
(2.5 feet) column on Curve II and the 0.76 (2.5  feet),  1.52
meters (5 feet), 2.29 meters (7.5 feet) and 3.05 meters (10 feet)


                              240

-------
to
      4 -
       2 -
CM

B
       1 -
                                 4.46
             1.2
      .68
                                  .55*
             *Per 100 grams of GAC

                                                                                                          .3
                                                                                                          -fc
                                                                                                          17*
          GAC  Column               1                        2                        3-4

          Figure 143.  THM FP substances in grams entering and adsorbed by each GAC column in 115 days  (ED4) .

-------
dolumns on Curve III.  Extension of the bottom curve to the  "Y"
axis may provide a rough estimate of the.maximum adsorptive
capacity of GAC at 1.2 grams per column and 0.68 grams per 100
grams.  Further work is needed to verify the usefulness of such
an approach.

     It was mentioned earlier that comparing THM FP adsorption
by adsorbents across water sources was not really a valid method
of interpreting these data because of the change of THM FP in-
fluent across water sources.  Some of the THM FP adsorption data
from Tables 38, 40 and 42 appear plotted in Figures 144 and 145
for further consideration.  Figure 144 presents THM FP adsorp-
tion per column (0.76 meter  [2.5 feet] deep) by GAC, XE-340 and
IRA-904 resin at a common point in time of 49 days.  The three
data points from raw, H.T. and finished water for GAC and IRA-
904 resin fall on a straight line.  The three data points for
XE-340 do not fall on a straight line.  The XE-340, H.T. and
finished water data points are small numbers, 0.053 grams and.
0.057 grams respectively.  These values were obtained by inte-
grating the adsorption curves, and it is possible that the minor
adsorption over the test period of only two percent and three
percent respectively is itself not a very accurate base.  In
Figure 144, it is probably-not very important which XE-340 line
is accepted, A-B through the H.T. water point, A-D through the
finished water point, or an average line A-C.  For our first
discussion, we can eliminate the XE-340 curve.  The GAC and IRA-
904 resin curves, both containing three data points, appear to
be a straight line.  This might indicate that the change in THM
FP influent across water sources, decreasing from raw to H.T. to
finished water, is the predominant cause of decreased adsorption
per column. To compare the effectiveness of different adsorb-
ents across water sources, the curves in Figure 144 may lead to
a more accurate interpretation of data.  Raw water adsorption
data for the three adsorbents is compared in Table 43, with
varying THM FP Influent levels and at a constant THM FP level.


     In Figure 144, the horizontal line of 600 yg/L may indicate
what the adsorption would be at that uniform THM FP level.  Line
A-C was chosen for the XE-340 curve for this discussion.  The
data in Table 43 show that the effectiveness of IRA-904 resin
compared to GAC and XE-340 varies when influent concentration is
considered.  The slopes of the GAC and IRA-904 resin curves are
different, and as the influent THM FP level decreases the curves
cross.  At low influent levels, GAC becomes more effective than
IRA-904 resin.  Comparison of adsorbents would be different if a
different common time point was chosen.  The results are also
different at 49 days when compared at an equal adsorbent weight
basis, Figure 145.  In Figure 145, GAC is more effective than
IRA-904 resin in all three water sources.  In both Figures 144
and 145, XE-340 has considerably less adsorptive capacity than
GAC or IRA-904 resin.

                              242

-------
to
>t»
U)
     700-1
     600
     500,
400 -
     300 .
     200 .
     100 -
                                                                            GAC  data points

                                                                       •O—  IRA-904 resin data points

                                                                          - XE-340 data points
          0
                                                                             .9
                                                                               1.0
                                                                                           1.1
                                                                                              1.2
                                        .5      .6     .7      .8
                            Grams of THM FP adsorbed per column

Figure 144. THM FP adsorption by GAC, XE-340 and IRA-904 resin per column 0.76 meter deep (2.5 feet)
            at 49 days.

-------
                           TABLE 4 3.
     EFFECT OF THM FP INFLUENT
     CONCENTRATION IN RAW WATER ON
     ADSORPTION DATA INTERPRETATION
tO
*»
*»
Adsorption
per column
at 49 days
at varying
THM FP
influent
levels
(from
Table 38 )
                                              Adsorptive
                                              capacity of
                                              904  resin
                                              campaiced to
                                              GAC  and
                                              XE-340
Adsorption
per column
at 49 days
at 600 yg/
of THM FP
influent
level (from
Figure 144 )
Adsorptive
capacity
of 904
resin
compared
to GAC
and XE-340
adsorbent grams of THM FP grams of THM FP
2.5 feet 904 resin
2.5 feet GAC
2.5 feet XE-340
1.21 1.21
.99 1.22 times .88 1.38 times
.71 1.70 times .57 2.12 times

              2.5 feet=0.76 meter

-------
800 _
                                                                                   GAC data points

                                                                                 " IRA-904 resin data
                                                                                         points
                                                                                 — XE-340 data points
   0
71
.i
                            .3      .k      5     .       .7      .6
                          Grains of THM FP adsorbed per 100 grams of  adsorbent
   Figure 145.  THM FP adsorption by GAC, XE-340 and  IRA-904 resin per 100 grams of adsorbent at 49 days.

-------
     TOG adsorption data  from  finished water  appear  in  Table  44.
Adsorption curves appear  in Figures  146,  147  and  118.

     XE-340, 0.76 meter   (2.5  feet)  deep, was studied in  EDlR,
Table 44 and Figure 146.  TOC  removal averaged three percent  of
the influent.  Removal was also very low  in raw and  H.T.  water,
five percent and four percent  respectively.

     GAC and IRA-904 resin, 0.76 meter   (2.5  feet),  were  studied
in EDS.  In Figure 147, both GAC and IRA-904  resin appear to  re-
move some Type II TOC substances.  In Figure  137, the IRA-904
resin did not appear to remove Type  II THM FP substances.   In
EDS, the IRA-904 resin removed 1.3 times  as much  TOC from finish-
ed water as GAC on an equal volume basis.  On an  equal  weight
basis, GAC removed 1.2 times as much as IRA-904 resin.  GAC, 0.76
meter   (2.5 feet) deep, was studied  in both EDS and  ED4.   As  the
influent TOC level decreased,  the adsorptive  capacity decreased.
In Figure 118, it is clear that increased GAC bed depth resulted
in larger periods of time to breakthrough and saturation.   I«
Figure 118, it appears that the TOC  data collected on test days
29 and 35 are too low and should probably be  discarded.   When
adsorptive data for the three  adsorbents are  compared across
water .sources some problems arise.   TOC adsorption data from
Tables 44,  46 (see page 257),  and 49 (see page 300)   are plotted
in Figures 148 and 149.  Straight lines were  not obtained with the
three data points for each adsorbent. In both Figures 148  and 149,
there appears to be a trend to increased adsorption  for each ad-
sorbent as the influent TOC level decreases, which does not appear
reasonable. Perhaps additional work with TOC  data in future re-
search is necessary before conclusions can be reached. In Figure
148, the adsorptive capacity (equal volume basis)  of IRA-904 resin
was always better than GAC and GAC always better  than XE-340.
In Figure 149, on an equal weight basis, GAC  was  generally bet-
ter vthan IRA-904 resin and IRA-904 resin was  always  better than
XE-340.

Other Parameters

Chlorine—
     The effect of free and combined chlorine in  finished water
as it passes through adsorbent columns is included in this  sec-
tion with the chlorine data on finished water.  XE-340  removes
essentially all free chlorine  for 17 days.  Free  chlorine in the
effluent than steadily rose and by the end of the test  (122 days)
was about one-third the influent level.  Throughout  the test
period the level of combined chlorine in XE-340 effluent  remain-
ed about one-half of the influent.   IRA-904 resin removed all
free chlorine and 93 percent of the  combined  chlorine throughout
the 53-day test.  In the same  test period, GAC  removed  all free
chlorine and initially 90 percent of the combined chlorine.  Com-
bined chlorine gradually increased in the effluent,  reaching  72
percent removal at the end of  the test.   Increased amounts of

                               246

-------
                   TABLE 4 4 .  TOC ADSORPTION DATA FROM FINISHED WATER











ED
1R
^^^^MM^^^H^^Hm
3
3
4
4
4
4











s
*w Qi
SS
Feet
2.5
^^H^^^^^^^^M
2.5
2.5
2.5
5
7.5
10









•P
C
d)
8
a

XE-
J340
GAC
904
GAC
GAC
GAC
GAC




£
Q)
3
rH
H-l
C
H
0)
(^
2
Q) CJ
£ 8
mg/L
6.5
^^^^^^^HBBHHBIIIH
5.9
5.9
5.4
5.4
5.4
5.4



Type II
Substances


A

a
fl 5
I*
rH 0)
O rl
O ffl
Days
ca
	 sa
0
0
0
0
7
14





0
C rt
I 3
rH +J
0 Hi
u w
Days
i't
L£U 	
32
39
24
32
43
60







MT

Inch
BBVVriHWriBBHIHBBBBBBHBIBBVBBBHB
30
30
30
30
25
23









C
O
•H
-P
(0 rl

Days
127
53
53
127
127
127
127


-S
a! t?
c n
w 3


p*
o||
O EH
O
H
£ W
Grams
73.7
IHHIHHIHVVI^HIBB
rt -H
O 0)
EH -P
i K
EH ffl EH
Grams

8.6
11.3
5.9
9.7
14
21.2




0
Q
F~i

tn
i f+
(0 -rl
-P rl
S5
•H S
O

%
HBIBBBflBBBBHHBflBBH^v^pvv
31
41
10
16
23
35

iconti

Type III
Substances




'0 1
fl) rH
X) O
W rl
•O 0)
<; ft
Grams
2.2
2.8
4.7
4
10.9
18.9
25.1

ued)



rH
j fr 1
EH -'
M
4-1 {
dp EH &
%
3
10
17
7
18
31
41


2.5 feet=0.76 meter   5 feet=1.52 meters  7.5 feet=2.29 meters  10 feet=3.05 meters

-------
                               TABLE  44*   (continued)
to
it*.
CO
1









ED
1R

3
3
4
4
4
4











•d ft
Feet
2.5

2.5
2.5
2.5
5
7.5
10








^
(3
2
Adsor

XE-
340
GAC
904
GAC
GAC
GAC
GAC



4J


3
r-l
MH
C
H

(1)
nt
aj u
mg/L
6.5

5.9
5.9
5.4
5.4
5.4
5.4


Type II
Substances




§
•d 3

M U
0
01 M
•d a)
(=< ft
Grains


6.9
7.8
5.1
6.6
7.6
9.4





fl
1
rH
•d o
Passe
Each
Grains


15.8
13.4
48.4
40
31
23



*J*
•S
M a
 <
%


44
37
10
14
20
29



C*
•H
m
H "
id g

g
O U
%


81
76
87
76
63
53


Type I
Substances


•d
0)
0)
01
id
ft
Total
Grams


2.4
2
4
4
4
4


CT>
g
•H
(U
•a J
4J r *i
s

%


9
7
6
6
6
6



Per 100 grams
of Adsorbent

^J
0)
(U
E-i

-------
to
*>.
6 -
      M
      0>

      •P
      •H
        4 -
        2 -
        1 _
                                                                                                       Av.  6.5
                                                                                                            6.3

                                                                                                            3%

                                                                                                    reduction
                               Finished water


                               Finished water thru 2.5  feet  XE-340 (0.76  meter )
        Days
      37    1"+21  28     35    42              64    71    78  84    91   98   105  112   119    122


         Figure  146.   TOG  in finished water and removal by  0.76 meter (2.5 feet)  of XE-340  (ED1R) .

-------
      0)
      -p
      •H
      rH
      u
(Jl
O
           8 -
           7 _
           6 -
           5 -
           4  -
           3  -
                                                                 Av. 5.9
                                                   Finished water

                                                   Finished water thru 2.5'IRA-905
                                                        resin (0.76 meter )
                                                   Finished water thru 2.5' GAG
                                                              (0.76 meters)
           Days
                       11
                                 25
     18

Figure 147.
32
                    39
                                                       53
                                     TOC in finished water and removal by 0.76 meter  (2.5 feet) of
                                     IRA-904 resin and by 0.76 meter (2.5 feet) of GAG  (ED3).

-------
                       GAC data points

              	Q	  "1RA-904 resin data points

              — Q—  XE-340 data points
U1
10 —


•P 9 -
I
H
IM
•S 8-
Ir
7 „
^

6 .
5 .
Raw
D ED1R
l\
l\
1\
\
\
H.T. « \
EDlRg \ Fin.
Q ED1R


                                                                                           Raw
                                                                                           ED3
                                                                                           O
                                                                                                o
                                                                                              H.T.
                                                                                              ED3
                •o
               Fin.
               ED3
•^i
1 -

%
Illllllillll'l
1234 56 78 9 10 11 12 13
                                  Grains of TOC adsorbed per column
         Figure 148. TOC adsorption by GAC, XE-340 and IRA-904 resin per  column-0.76 meter
                     at  49  days.
(2.5 feet)  deep

-------
   10 -i
•M

§  
-------
combined chlorine were removed with  increased  GAC bed depth.

Turbidity—
     The effect on turbidity when  raw,  H.T.  and  finished water
pass through adsorbents and turbidity  in  the distribution system
are included in Appendix A.

     Adsorbents removed turbidity  from H.T.  effluent but turbid-
ity of the raw water  increased considerably  after passing through
XE-340.  Turbidity also increased  as raw  water passed through
IRA-904 resin.  This  same  resin  removed turbidity from finished
water and H.T. effluent.

Color—
     No color was removed  from raw water  by  XE-340 or GAC.

pH—
     The effect on pH of raw, H.T. and finished  water as they
pass through adsorbent columns is  included in  this section with
other pH data.  The average pH of  raw  water  through GAC, XE-340
and IRA-904 resin increased by approximately 0.1.  The average
pH through XE-340 was one-tenth  lower  than H.T.  water; through
IRA-904 resin it was  three-tenths  lower.  There  was no change
through GAC.  The average  pH of  finished  water decreased 0.1
when passed through XE-340, and  0.2  through  GAC  and IRA-904
resin.

Comparison of Laboratory and Distribution
   System Aging
     It is important  to know if  bottle aging in  the laboratory
to determine total THM, terminal THM or THM  FP correlates with
actual distribution system formation of these  parameters.  To
avoid confusion, we will refer to  the  THM growth of two-day aged
samples as total THM  growth.  The  parameter, terminal THM will
be reserved for THM growth occurring in six-day  aged samples
which is the time factor used throughout  this  study in deter-
mining THM FP  (THM FP = term.  THM - inst THM) .  Based on mea-
surements of the length of time  it took for  abrupt changes in
fluoride concentrations made at  the  plant to reach certain points
in the distribution system, we estimated  that  it takes two days
for finished water leaving the Preston Plant to  reach the Red
Road sampling point.  A sampling procedure to  determine THM
growth was established as  part of  ED3  to  compare laboratory
bottle aged samples with samples taken directly  from the distri-
bution system sampling point at  Red  Road.  The sampling proce-
dure appears in Table 45.

     The data obtained are presented in Figure 150.  During the
53-day test period of ED3, the Inst. THM  levels  of finished
water at the Preston  Plant are shown by the  lower curve in Fig-
ure 150.  Finished water leaving the Preston Plant had an aver-
age Inst. THM level of 128 yg/L.  Total THM  levels of finished

                              253

-------
       TABLE 45.  SAMPLING PROCEDURE TO COMPARE LABORATORY AND DISTRIBUTION
                  SYSTEM AGING
                 WATER SAMPLE
             TREATMENT
      Finished water at Preston Plant
aged 2 days in bottle with no additional
chlorine
KJ
Ul
      Finished water at Preston Plant
pH 9 buffered and excess free chlorine
added then aged 6 days in bottle
      Red Road distribution system sample
has been aged approximately 2 days in
distribution system
      Red Road distribution system sample
sample taken and treated
after approximately 2 days in distri-
bution system with pH 9 buffer and excess
free chlorine and then stored in bottle
in laboratory for 4 days before analysis
(represents a total of 6 days aging)

-------
500
                                A.

                                o-
   Pinished water at Preston Plant
   bottle aged 6 days
   Red Road water (buffer + chlorine)
  'aged 4 days in bottle.
                                                       Av.436

                                                       Av.410
                                                      Av.261
                                                ——Q Av.255
                                                ——O Av.128
                              0    Red Road water
                           _-O-« Finished water at Preston  Plant

                           — -D*— Finished water at Preston  Plant
                                   aged 2  days in bottle .
39 4*2
       4  }  f i 14 il h  h is  ^2 a'h  39 42 w  49
       Figure 150 . Comparison of laboratory  bottle  aged  and
                   distribution system THM growth  (EDS).
                               255

-------
water at the Preston Plant aged two days in a bottle also are
plotted in Figure 150, indicating an average total THM level of
255 yg/L, a growth of 1.99 times.  The Red Road distribution sys-
tem sample had an average total THM level of 261 yg/L, a growth
of 2.04 times, which is very close to the average value of the
bottle aged growth.  Although the specific two-day bottle and
actual distribution sample values vary more widely than .the aver-
age values, the general magnitude of the distribution water
values can be roughly estimated by laboratory aged samples.
Curves II and III in Figure 155, show additional data on the com-
parison of bottle aged and distribution system samples.

     Finished water from the Preston Plant aged six days in a
bottle with a pH 9 buffer and excess free chlorine (data plotted
in Figure 150) had an average terminal THM level of 410 yg/L, a
growth of 3.2 times.  The Red Road sample, aged in a bottle for
four days with a pH 9 buffer and excess free chlorine  (a total
of six days from leaving the Preston Plant when including the
two days travel in the distribution system plus four days bottle
storage) had an average terminal THM level of 436 yg/L (data
plotted in Figure 150), a growth of 3.4 times, which again is
very close to the six-day bottle aged sample.  Therefore, labora-
tory bottle aging compares fairly closely with actual distribu-
tion system aging.  It should be noted that the finished water
sample from the Preston Plant that was aged two days and the
Red Road sample that had been in the actual distribution system
for two days, had no  free chlorine left at the end of two days,
and would have had higher total THM values if free chlorine had
been maintained.

Total THM Growth in Adsorbent
   Column Effluents
     As finished water passes through an adsorbent column, it
loses all of its free chlorine and various amounts of its com-
bined chlorine.  The column effluent would have to be rechlori-
nated to a free chlorine level of approximately 2.5 ppm before
entering the distribution system.  To study the effect of total
THM growth with two days of chlorine contact on such effluent
samples from 0.76 meter  (2.5 feet) deep columns of GAC and IRA-
904 resin, a sampling procedure was established as part of ED3.
The sampling procedure appears in Table 46.
                               256

-------
   TABLE 46.  TOTAL THM GROWTH  IN  ADSORBENT COLUMN EFFLUENTS
           WATER SAMPLE

Finished water through
0.76 (2.5 feet) meter  of
IRA-904 resin
         TREATMENT
                •*
buffered at pH 9 and free
chlorine added to 2.5 ppm,
bottle aged 2 days
Finished water through
0.76 (2.5 feet)  meter
of GAC
buffered at pH 9 and free
chlorine added to 2.5 ppm,
bottle aged 2 days
     The data obtained are presented in Figures 151, 152, 153
and 154.  In Figure 151, as a reference base, curves are presen-
ted showing the inst. THM in finished water at the Preston Plant
and the total THM growth occurring after two days of bottle ag-
ing.  The average inst. THm level in finished water was 128 yg/L
and the total THM after two days of bottle aging averaged 255
yg/L.  The inst. THM levels in the 0.76 meters (2.5 feet) deep
IRA-904 resin column effluent is also presented in Figure 151.
The average level was 176 yg/L.  This level is higher than in
the finished water entering the IRA-904 resin column due to the
catalytic generation of THM in the column as previously discuss-
ed.  The average THM growth in the column was 48 yg/L (176 yg/L-
128 yg/L).  Total THM levels in the IRA-904 resin column efflu-
ent, buffered to pH 9, rechlorinated to 2.5 ppm of free chlorine
and bottle aged two days are presented in Figure 151.  The aver-
age value of THM growth resulting from catalytic generation, 48
yg/L, is subtracted from the 309 yg/L average above, an average
value of 261 yg/L is obtained.  This average value is very close
to the 255 yg/L average value (Figure 151) for the Total THM
growth obtained by aging finished water for two days.  These
data indicate that the IRA-904 resin column did not remove much
THM FP from finished water in ED3.  This separate information
confirms the information reported in Table 42, that the IRA-904
resin column in ED3 0.76 meter  (2.5 feet) deep removed only 13
percent of the influent THM FP.

     The curve shown in Figure 152 was obtained by subtracting
Curve III from Curve IV in Figure 151.  It represents the total
THM growth in the IRA-904 resin column effluent due only to THM
FP conversion.  It clearly shows that throughout the test period
even from initial startup, the IRA-904 resin never removed
enough THM FP to keep THM regrowth below 100 yg/L.

     In Figure 153, the total THM growth in a 0.76 meter   (2.5
feet) deep GAC column effluent which had been buffered, rechlor-

                               257

-------
to
en
CO
                       Curve I - Inst, THM in finished water at Preston  Plant

                       Curve II - Total THM in finished water at Preston Plant bottle aged 2 days

                       Curve III - Inst. THM in finished water thru  2.5  feet of IRA-904 resin (0.76 meter)

                       Curve IV - Total THM in finished water thru 2.5 feet IRA-904 resin, pH 9 buffer
                                  2.5 ppm free chlorine, bottle aged 2 days (0.76 meter )
                   7  li  lV  18 21 25 28  32  35 39  42  46 49  53
                 Figure 151.  Total THM growth in rechlorinated - 2 day aged IRA-904 resin  column effluent(EDS).

-------
      M
      0)
      •P
      O>
      I
                                                       Total THM growth due to THM PP  conversion in 0.76 meter
                                                        (2.5 feet) deep IRA-904 resin" column effluent

                                                       buffered at pH 9, rechlorinated to  2.5  ppm
                                                       free chlorine and aged 2 days
     300-i
to
en
ID
     200 -
     100 -
     Days
                                                      Av.  167
                                             4649  53
            Figure 152. Total THM  growth  in  IRA-904 resin column effluent due to THM FP conversion (ED3).

-------
       400 —,
       300 —
     
-------
  400 .
  300
0)
-P
•H
  200
rti
•M
O
  100
           Total THM growth due to THM FP conversion in

           2.5  feet deep GAC column effluent buffered at

           pH 9, rechlorinated to 2.5 ppm free chlorine

           and  aged 2 days  (0.76 meter )
 Days
0  4711
                    18 21 25 28  32 35  39  42  46 49 53
       Figure 154. Total THM growth in GAC column effluent due to

                   THM FP conversion (EDS).
                                261

-------
inated and aged two days is shown.  Curve IV represents  the
total THM present after two days aging.  Unlike the  similarly
treated IRA-904 resin column effluent, Curve IV in Figure  151,
the GAC curve indicates that for a period of time after  initial
column flow, the GAC removed sufficient THM FP precursors  to
keep the total THM below 100 yg/L.  The curve shown  in Figure
154 was obtained by subtracting Curve III from Curve IV  in
Figure 153.  It represents the total THM growth in the GAC
column effluent due only to THM FP conversion.  It indicates
that up to some point between 7 and 14 days, the GAC column
removes sufficient THM FP precursor to keep the total THM  growth
below 100 yg/L.  Of the three adsorbents tested, GAC, XE-340  and
IRA-904 resin, GAC was the only adsorbent to show this charac-
teristic.  This was the basic reason for studying deeper GAC
columns in ED4.

Bed Life Criteria in Deep GAC Columns

    In ED4, the effluents from the 2.29 meters (7.5  feet)  and
3.05 meters (10 feet) deep GAC columns were buffered to pH 9,
rechlorinated to 2.5 ppm of free chlorine and aged two days to
study total THM growth.  The results should give some idea of
the bed life one could expect from such columns.  GAC exhaustion
criteria at this point can be expressed in various ways.   The
EPA tentatively has proposed basing exhaustion time  of GAC to
assure maximum protection for the consumer (8).
                                /
    In this report we are determining GAC exhaustion based only
on the proposed MCL regulations that the total THM should  not
exceed 100 yg/L.  As mentioned previously, finished  water  pass-
ing through a GAC column will have nil free chlorine.  It  would
therefore have to be rechlorinated to approximately  2.5 ppm of
free chlorine before it could enter the distribution system.
Since precursors are still present, the addition of  free
chlorine would cause THM regrowth.  We therefore define GAC
exhaustion or bed life at the point where THM regrowth in  a
sample of GAC column effluent after rechlorination to 2.5  ppm
of free chlorine and aging for two days, reaches the THM MCL
level of 0.1 mg/L (100 ppb).  The results obtained in ED4  on  the
2.29 meters (7.5 feet) and 3.05 meters (10 feet)  deep GAC
columns are shown in Figure 155.  Instantaneous THM  level  varia-
tions in Preston Plant finished water (carbon column influent)
are shown in Curve I.  The average value was 147 yg/L.  When
Preston Plant finished water.was bottle laged two days, addi-
tional THM growth occurred and the Total THM present is shown
by Curve II.   The average was 243 yg/L, an increase  of 1.7
times.   Bottle aging and distribution system aging was compared
in ED4.   Samples were again taken at the Red Road sampling point
in the distribution system.  Total THM levels are represented by
Curve III.   The average value was 218 yg/L, an increase of 1.5
times.   This demonstrates again that the laboratory  bottle aging
approximates Total THM growth in the distribution system.

                              262

-------
       400
    M
    
U)
20
                  Curve I - Finished water at the Preston Plant

               — Curve II -Preston Plant, finished water aged 2 days

               — Curve III - Red  Road sample point

             • A--- Curve IV - 7.5 feet  deep GAG column effluent - rechlorinated  -  aged 2 days (2.29

             •O — Curve V - 10  feet deep GAC column effluent - rechlorinated  -  aged 2 days  (3.05
                                                                                         meters)
       100
       100
      Days
1 /^. ,A' ^
/V ' V-*' ,
) k ............. * ... . i
/
N>--°''
65
Days
1
81
Days
               14          28         42         56         70      80          94         112

         Figure 155.  Total  THM growth in rechlorinated - 2 day aged 2,29 and 3.05 meters

                      (7.5 and 10 feet) GAG-
                                                                                                        122

-------
    In Figure 155, data points prior to  day  42 were  not deter-
mined for the 2.29 meters  (7,5 feet) deep GAG column.   In Figure
155, GAG bed life  (based on when the rechlorinated GAG  effluent
reaches 100 yg/L) is shown to be 65 days for 2.29 meters (7.5
feet) and 81 days for 3.05 meters  (10  feet).  The total THM in
the two-day aged 2.29 meters  (7.5  feet)  and  3.05 meters (10
feet) deep GAG column effluents are the  result of two condi-
tions, 1) increasing amounts of inst.  THM in the effluent due
to column breakthrough and 2) increasing amounts of  THM growth
due to the THM FP breakthrough and conversion.  Curve I in
Figure 156, shown the total THM in the rechlorinated two-day
aged 2.29 meters (7.5 feet) deep GAG column  effluent and Curve
II, is obtained by subtracting the inst. THM levels  in  the
column effluent, breakthrough and  indicates  the contribution of
total THM resulting from THM FP breakthrough and conversion.
On test day 56, inst. THM began to breakthrough the  column.  It
is seen that at day 65 when Curve  I reached  the 100  yg/L level,
approximately 63 percent of the THM was due  to THM FP and 37
percent due to inst. THM breakthrough.  From test day 65 to the
end of the test period the average of  the data points for Curve
II is 93 yg/L, which is below the  100  yg/L limit.  It would
appear that the 2.29 meters  (7.5 feet) deep  GAG column  removes
sufficient THM FP precursor to keep the average THM  regrowth
below 100 yg/L at least up to the  115  day test period.   Thus,
one might attribute column failure at  65 days to inst.  THM
breakthrough, since if the THM did not break through, the
column might last over 115 days.   In Figure  158, it  is  apparent
that THM breakthrough alone would  cause bed  life failure (reach-
ing 100 yg/L) in 94 days.

    Curve I in Figure 157, shows the total THM in the rechlori-
nated two-day aged 3.05 meters (10 feet) deep GAG column efflu-
ent.  Curve II, obtained by subtracting the  inst. THM levels in
the column effluent, indicates the contribution of total THM
resulting from THM FP breakthrough and conversion.   On  test day
70, inst. THM began to break through the column.  It is  seen
that at day 81, when Curve I reached the 100 yg/L level,  only  a
very small amount of the total THM was due to THM breakthrough,
approximately 8 percent.  Thus one might attribute bed  failure
at 81 days to THM FP breakthrough.  However, from test  day 81  to
the end of the test period the average of the data points for
Curve II is 95 yg/L, which is below the 100  yg/L limit.   As  with
the 2.29 meters (7.5 feet) deep GAC column,  it appears  that the
3.05 meters (10 feet) deep column  removes sufficient THM FP
precursor to keep the average THM  growth below 100 yg/L at least
up to the 115 day test period.  Again, one might attribute
column failure at 81 days and beyond for the 3.05 meters (10
feet) deep GAC bed to inst. THM breakthrough.  In Figure 158,
it is apparent that THM breakthrough alone would cause  bed life
failure in 119 days.  GAC bed failure  is thus caused by a com--
bination of THM and THM FP breakthrough and  the ratio of the
two components at bed failure probably varies with bed  depth and

                              264

-------
                   --D-" Curve
      I - total  THM in rechlorinated and

          two  day aged 7.5 feet deep GAC column effluent (2.29  meters)

Curve II - Curve I minus inst-THM in the 7.5  feet

          deep GAC column effluent (2.29 meters)
         250 -,
to
cr\
01
          Days
7 10  14 17 21 24  28 31  35 38 42 45  49 52 56 59 63 66  70 73 77 80 84 87  91 94  98
                                                                                               105
                                                                           112
122
                  Figure 156. THM breakthrough and THM FP conversion components of  Total-THM in

                              two day aged effluent from  2.29 meters (7.5 feet)  deep GAC column  (ED4) .

-------
     0)
     -p
ON
        200 -
        150 -
Curve I - total THM  in rechlorinated and

two day aged  10 feet deep  GAC column effluent


Curve II - Curve I minus  inst-THM in the  10  feet

deep GAC column effluent  (3.05 meters)


                                           'p.-
        100
         50 -
        Days
                   7  10  14 17  21 2H  2831 35 38 1*245 4952  5659  63  66 7073  7780 84 87  9194 98101
                                                                 112     122
                Figure 157.  THM breakthrough and THM FP  conversion components of total-THM
                             in 2 day aged effluent from  3.05 meters  (10  feet)  deep GAC column  (ED4)

-------
     200
to
                                Inst. THM xn finished water
                                Inst. THM in finished water thru 7.5 feet of GAG
                                                               (2.29 meters)
                                Inst. THM in finished water thru 10 feet of GAC
                                                               (3.05 meters)
   r  T T i
52  56 59 63 66
                                                               77 80 8k 87
   0

Days
q q? qp qqgqi qg gp-^
 3  7 10  14 17  21 24  28 31 35 38 42 45  49 5
                                                             9194 98101105108112115119122

Figure  158. Inst.  THM in finished water and finished water through2.29 and 3.05 meters
            (7.5 and 10 feet)  of GAC (ED4).

-------
 influent concentrations of both.

    The curves in Figure 143, show that the  first  column,
 which received the highest concentration of  adsorbate,  adsorbed
 the greatest amount over the test period.  Therefore, for  great-
 est GAG efficiency, as much of the carbon in a bed as possible
 should at some time be exposed to the highest possible  concen-
 tration of adsorbate.  If 3.05 meters  (10 feet) of bed  is
 necessary, by using four beds 0.76 meter  (2.5 feet) deep  in
 series, and removing the first bed at saturation while  placing
 a new bed at the end of the series, we expect 31 percent
 increased adsorptive capacity.  This would increase the bed life
 from 81 days to 106 days. A continuous in-out bed  would increase
 bed life by 43 percent to 116 days.

    In Figure 155, it is apparent that if a  MCL below 100  yg/L
 were chosen, bed life would seriously be reduced.   At 50 yg/L
 and 25 yg/L, bed life for a 3.05 meters (10  feet)  deep  GAG bed
 would be approximately 18 days and 4 days respectively.

Relationship of TOC and THM FP Data

    Determination of the THM FP of water is  a new  analytical
 method.  When a new method is introduced the question of its
 relationship or correlation with an existing method or  methods
 usually arises.  In this case, the relationship to the  deter-
 mination of TOC arises.  From our research work, we can divide
 a discussion on this subject into two parts,  1) relationship
 in the treatment plant and 2) relationship in GAG  column
 effluents.  TOC and THM FP data were simultaneously collected
 in ED1R, ED3 and ED4.  TOC and THM FP data from Tables  10  and
 11 are plotted in Figure 159.  ED1R data points for raw, H.T.
 and finished water are connected by line segments  A-B and  B-C.
 As one might expect, the slope of these two  segments are differ-
 ent.  A-B is the result of adsorption on precipitated calcium
 carbonate, and B-C is the result of conversion by  chlorination,
 oxidation and sand filtration.  The two segments in EDS, A1- B1
 and B1- C', are quite similar to those in ED1R.  The length of
 the segments differ in proportion to the percent removal data in
 Tables 14 and.15.  If only these two sets of data  were  available,
 ED1R and ED3, one might conclude that correlation  between  TOC
 and THM FP results are quite close.  However, the  ED4 data
 points show a considerable shift to the right.  The two segments,
 in ED4, A"- B" and B"- C", are relatively parallel to the  seg-
 ments in ED1R and EDS.  If an unknown point  "X" were chosen
 within the triangle formed by the three raw  water  data  points
 and parallels were drawn to all segments in  the three ED,  the
 shaded area represents the predicted zone of results from  data
 point "X".  While there may be some degree of confidence in this
 prediction, the value of it is unknown.  If  point  "X" were
 chosen outside the triangle, the degree of confidence might be
 less.  In our treatment plant data, one could say  that  there is

                              268

-------
to

CT»

VO
      10-
       9-1
       8.
        7-
        6-
        5-
A,A1,A" = Raw water data points from EDlR, 3 and 4
          respectively  *

B,B',B" = H.T. water,data points from EDlR, 3 and 4
          respectively   *

C,C',C" = Finished water data points from EDlR, 3 and 4
          respectively       *
                                                                    —0~
EDlR data points


ED3 data points

ED4 data points


Shaded area - predicted result

      from data point X
                                                           B1
         200           300           400            500            600            700

                                                THM FP ygAiter

          Figure 159.  Relationship of TOC and THM FP data in raw,  H.T. and finished water.

-------
a relationship between TOC and THM FP data, but one cannot be
converted into the other by a single or simple conversion
factor.

    TOC and THM FP data from the 3.05 meters  (10 feet) GAG
effluent in ED4 are plotted in Figure 160.  The average level of
TOC for the first 14 days was 0.37 mg/L.  The average THM FP
for this same period was 17 yg/L.  From day 14 to day 49, both
curves show a steady rise.  The THM FP curve has a steeper
slope.  From day 49 to the end of the test period, when the
plateau portion of both curves was attained, the TOC level
averaged 3.0 mg/L and the THM FP level averaged 240 ug/L.  The
TOC concentration increased 8.1 fold, from 0.37 to 3.0 mg/L.
The THM FP concentration increased 14.1 fold from 17 yg/L to
240 yg/L.  There is a relationship in the segments of the two
curves, but, again, no single or simple conversion factor for
converting one to the other.  As GAG bed depth changes, a seg-
ment relationship exists, but is different for each bed depth.

    We can probably conclude that while there is some degree of
predictable relationship between TOC and THM FP data in both our
treatment plant and a separate degree of predictable relation-
ship in our GAG effluent waters, both tests yield separate and
valuable information which are not convertible one into the
other by a single or simple conversion factor.  It is unlikely
that this situation will change as more data are collected,
especially when different geographical relationships are con-
sidered.

Leaching Study  on XE-340  Resin Column

    The experimental design of the leaching study on a 0.76
meter   (2.5 feet) deep XE-340 column is discussed under ED2.
In ED2, a fresh XE-340 column was installed, preceding the
partially exhausted XE-340 column on the finished water line of
ED1R, to supply halogenated organic free water for the leaching
study on the partially exhausted column.  In this discussion we
will first present data on chlorodibromomethane.  The level of
chlorodibromomethane entering and leaving the leaching study
column in the 63-day test period is plotted in Figure 161.  A
weakness in the experimental design is apparent.  The lower
curve indicates the chlorodibromomethane in the fresh XE-340
column effluent and entering the leaching study column.  Ideally,
to obtain a mass balance of.leached substances, we should have
replaced the fresh XE-340 column with a new fresh XE-340 column
before breakthrough of HOC occurred.  Actually, it also would
have been better to use a deeper XE-340 column, because as seen
on the chloroform data to follow, the MTZ for chloroform is
greater than 0.76 meter  (2.5 feet).  Nevertheless, despite the
breakthrough of chlorodibromomethane we can still draw some
conclusion on leaching from Figure 161.  If we consider the data
up to test day 53, the level of chlorodibromomethane entering

                                270

-------
      400-
      300-
     M
     
-------
          Curve I - Chlorodibromomethane leaving the
          exhausted column
— — O— —  Curve II - Chlorodibromomethane entering the
          exhausted column
:
:
Days ° * 7
\r
__^~~-i
11 14 25 32 53 56 63
Figure 161. Level of Chlorodibromomethane entering and leaving
            the partially exhausted 0.76 meter   (2.5 feet)
            deep XE-340 column (ED2).
                        272

-------
the leaching study column  for  most of  the  period was below  1
Ug/L.  Of this 53 day period,  the  data point  at day 53 on Curve
I is in greatest error.  It would not be scientifically correct
merely to subtract Curve II  from Curve I to obtain the true
leaching curve without  the interference of entering chloro-
dibromomethane.  We  can at this point  only be aware that the
test point on Curve  I at test  day  53 is considerably lower  than
plotted.  Curve I is replotted in  Figure 162, with the full
chlorodibromomethane adsorption curve  indicating the level  of
breakthrough leaving the XE-340 column at  the end of the ED1R
study.  In Figure 162,  it  is seen  that the breakthorugh level of
the column on test day  122 of  ED1R and the leaching from the
column on test day 0 in ED2  are approximately equal at about
8 ug/L-  As the leaching study continued (Curve II) it is
apparent that the leaching curve is just the  reverse of the
adsorption curve  (Curve I).

    The level of bromodichloromethane  entering and leaving  the
leaching study column is plotted in Figure 163.  Leaching data
points to test day 32 are  plotted  in Figure 164.  Again, it is
apparent that  the breakthrough level of the column on test  day
122 of ED1R is approximately equal to  the  leaching level on test
day 0 of ED2  (20 yg/L) . Again, the leaching  curve appears  to
be the reverse of the adsorption curve. Similar results are
shown for chloroform in Figures 165 and 166 and for cis 1,2-
dichloroethane in Figures  167  and  168.  It is not surprising to
find that on XE-340, desorption of the four HOC discussed above
appears to be  the reverse  curve of adsorption. It is well  known
that adsorption and  desorption on  GAG  follows this pattern  with
some substances while other  substances may exhibit some hyster-
esis on desorption.
                               273

-------
                                                                     Curve I
                                                                 0  adsorption from finished
                                                                     water (ED1R)

                                                                 -O—Curve II
                                                                     leaching (ED2)
   -P
   -H
Days
7 10  14 17 2i 2k 28 31  35 3   4245   49  2   6   6§ 66 7b  3  7 8   8  87
                                                                                   8   105    112     122

                                                                            I      I
        Figure  162.   Adsorption and leaching of chlorodibromomethand on a 0.76 meter
                      (2.5 feet) deep XE-340 column  (ED1R and ED2).

-------
         25-*
         20-
K)
>»J
Ul
      N
      
3.
         15-
   10 _
          5 -
          0
       Days
                                                                             Curve I- Dichlorobromo-
                                                                          fr  methane leaving the
                                                                             exhausted column
                                                                       _ Q  fit rye II-Dichlorobromo-
                                                                          w  methane leaving the
                                                                             exhausted column
                                                  o-o
                                                      --o
                                    X
                                            X
                                      X
                                X
       M-q-rr""
          i+  7   11 14      25   32             53 56   63
           Figure 163. Level of Bromodichloromethane entering and leaving the partially
                      exhausted 0.76 meter  (2.5 feet)  deep XE-340 column (ED2).

-------
to
     _. Curve I - adsorption from finished water

                                                                (ED1R)




                                                         ... Curve II - leaching (ED2)
     Days 03  7 10 14 17 21 24 28 31 35 38 4245 49 52  56 59 63 66 7073 77 80 84 87 91 94 98 101105108112115119122

     Curve I

                                                       6^   56 5*332    ?5

     Figure 164. Adsorption and leaching of bromodichloromethane on a 0.76 meter  (2.5 feet) deep

                 XE-340 column (ED1R and ED2).
Tf-ri—r—i—i Days
1411  74   0 Curve
              II

-------
         60 -i
to
          10 -
             •   Curve I -  Chloroform
                 leaving the  column
           —O — Curve II - Chloroform
                 entering the column
           0

        Days
63
%  7 11 14       25    32             53 56

Figure 165. Level of chloroform entering and leaving the partially

            exhausted 0.76 meter   (2.5 feet) deep XE-340 column  (ED2)

-------
00
      60  -
      50  _
       40-
        30_
     M
     <0
     •P
     -H
                           Curve I - adsorption from finished water (ED1R)
                    —O—Curve II - leaching (ED2)
Days Curve I
                                                                               32   25        1411  740

          Figure 166. Adsorption and leaching of chloroform on a  0.76  meter  (2.5 feet)  deep XE-340

                     column  (ED1R and ED2) .
                                                                                                        , Days

-------
       15—i
to
-j
vo
3
                                                                               Curve I - cis l,2-0ichloroethene
                                                                           0   leaving the column


                                                                        —o—Curve II- cis 1,2-oichloroethene
                                                                               entering the column
                7  11 1^      25   30             53 56   63

             Figure  167. Level of cis 1,2-Dichloroethene entering and leaving the partially exhausted

                         0.76 meter   (2.5 feet) deep XE-340 column  (ED2).

-------
                                                             Curve I - adsorption from finished water
                                                                         (ED1R)
                                                         _O-Curve XI " leaching  (ED2)
to
00
o
       15
          0 3
     Days Curve I
           Figure 168.  Adsorption and leaching of cis 1,2-dichloroethene on a 0.76 meter  (2.5 feet)
                        deep  XE-340 column (ED1R and ED2).

-------
Biological Activated  Carbon  (BAG)

    An  aquatic microbiology laboratory has been established in
 the Drinking Water Quality Research Center at Florida Inter-
 national University.   At the beginning of ED4, Dr. Frances
 Parsons began a bacteria profile study of raw and finished water
 at the  Preston Plant and the effluent from each of the four GAC
 columns.  Samples of water from the distribution system were
.also  analyzed.  This work was contributed to this project by
 Florida International University to demonstrate their capability
 and  interest in this area.  Two reports by Dr. Parsons are
 available in Appendix A. It is obvious from this work that we
 had  a BAG system.  Since no additional oxygen was added to the
 water,  we call it a partial BAG system to differentiate it from
 the  oxygenated European system.  We do not know how our adsorp-
 tion results would differ if the bacterial growth had not
 developed.  Despite massive bacterial growth that hindered and
 finally prevented backwashing of the columns, the adsorptive
 capacity of the GAC for HOC did not seem to be affected.  The
 initial breakthrough and saturation times for each HOC through
 each column were too consistent to suggest blocking of active
 sites or the pore openings in the carbon by the bacteria.

     The two reports in Appendix A indicate the inadequacy of
 the  standard plate count method in bacterial analysis of drink-
 ing  water.  It stresses the need for longer incubation times at
 various temperatures and with different media for specific
 species as first indicated by Van Der Kooij (9).

     Conclusion and tentative recommendations from the two
 reports by Dr. Frances Parsons are as follows:

     Report I:

     Bacteria that occur in small numbers in raw water survive
 treatment and colonize granular activated carbon  (GAC) columns
 used to remove organic solutes from treated water.  The bacteria
 multiply, form slime that interferes with column maintenance by
 preventing backwashing, and slough off in large numbers into
 the  water passing through the columns.

     Based on influent and effluent sampling, the size and com-
 position of the microbial populations in GAC columns appear to
 change  with time.  The composition of the microbial population
 of the  raw water apparently influenced the population in the
 columns.  Each column had a somewhat different population com-
 position and size on each sample date.

     Some of the organisms that multiply in the GAC columns may
 pose a  health hazard because of the vast numbers Pfesent if the
 column  effluent is ingested or comes in contact with suscept-
 ible body surfaces such as the otic canal or the naso-pharyngeal

                                281

-------
mucosa.  The possibility of a consumer incurring enteritis,
intoxication, and/or an opportunistic infection should be
studied.  Because of the large numbers of Gram-negative organ-
isms that colonize GAC columns, endotoxin should be assayed
using the LAL method.  Staphylococci sp. sometimes present in
finished water should be tested for coagulase.

    The large numbers of noncoliform bacteria found in column
effluents will suppress coliform growth and interfere with
interpretation of the standard coliform detection test.

    Effects of rechlorination of column effluents on the sub-
sequent bacterial population of this water over a period of
time is being studied.  Preliminary results indicate that enter-
ing organisms that survive treatment plant processes become a
major component of the microbial population in GAC columns.  A
count of 300/100 mL of sample of Enterobacter agglomerans was
obtained in a GAC column effluent .sample two days after
rechlorination to 3 ppm.  Two colonies of Enterobacter
agglomerans per 100 mL of sample were isolated from a sample of
effluent that was chlorinated to 10 ppm and held at 25°C for six
days.  These results suggest that small numbers of bacteria can
survive chlorination probably inside of cell aggregates.

    Report II:

    Chlorination of the effluent from granular activated carbon
(GAC) columns apparently kills bacteria that grow on the carbon
granules and slough off into the effluent, but the initial dose
of chlorine must be adequate to combine with the bacteria and
leave sufficient free chlorine to prevent regrowth.  The con-
centration of chlorine necessary would vary with the bacterial
biomass and chlorine demand due to all constituents of the
water.

    This study was of a cursory nature and was only intended to
suggest a more complete study.  Shorter sampling intervals
(daily), for a period of time longer than six days (end point
determination), with more than these two concentrations
(especially less than 3 ppm free chlorine) of several disin-
fectants (chlorine, chloramines, chlorine dioxide, ozone,
ferrates)  should be examined.  Certainly, the minimum level of
chlorine needed and the time that it is effective for several
bacterial population sizes should be determined.  All of these
factors; i.e., dose size, contact time, regrowth rate and size
and composition of the bacterial population should be studied
and compared with parallel determination of the bacteriology of
the distribution system.
                              282

-------
Polanyi-Manes Adsorption Theory

Theory Development—
     Theories  of adsorption from solution were developed by
Polanyi.  (10)   They were modified by Hansen and Fackler. (11)
Manes and Hofer (12) made modifications for predicting relative
adsorption potentials from the refractive index of a substance.
With these modifications one can estimate the adsorptive capa-
ity at saturation of a variety of miscible organic liquids by
activated carbon over a wide concentration range.

     Deviations from the Polanyi Theory and its modifications
are ascribable to specific chemical interactions or steric
effects.   Work on adsorption of miscible liquids from water has
been done by Wohleber and Manes.  (13,14)  Their work has been
expanded by Chiou and Manes  (15) to include solids from solu-
tion.  Adsorption of Binary Organic Liquids was studied by
Schenz  and Manes.  (16)  The most recent work on competitive
adsorption was reported by Rosene and  Manes.  (17,18,19,20)

     Application of the theory usually starts with adsorption
 in the gas phase.  The mechanisms involved are illustrated in
 the following drawing of a micro pore  in a carbon particle.
                             0   molecules in the gas phase
                             O
                            more concentrated gas phase
                           liquified
                                283

-------
     Two molecules in the gas phase  are  attracted by
Van der Waals  forces.  The distance  between  two  molecules is

"r" and the attracting force decreases by  -T-.   As the mole-
                                           r
cules enter the pore they become more concentrated,  and when
"r" becomes small enough they will condense  into a liquid.
Molecules of both gas and liquid are held  to the carbon surface
molecules by Van der Waals forces.

     There are two parts to the driving  force responsible for a
substance going from the gas to the  liquid phase.   One is the
energy part of the attractive force.  It is  related to the
polarizability.  The other part of the driving force is entropy,
which is related to solubility.  The less  soluble  a substance
is, the easier it is to bring to saturation  (condensation).   The
driving force  is expressed as e.

                                    P_
                          e = RTln   s

             R = ideal gas constant  1.987 cal/deg/mole

             T = absolute temperature

             P = saturated pressure
              S

             P = equilibrium pressure

     The polarizability of a substance is ps~


                           s    ni
                          P  = 	T	
     where n. = refractive index.

     The refractive index for all hydrocarbons is about the
same, therefore pb, where b stands for butane , equals 0.236.
To determine the adsorption isotherm curve for heptane adsorbed
on a particular GAC (gas phase) , the volume of heptane adsorbed
(cc per 100 gm of GAC) is measured at various concentrations of
heptane in the gas phase.  This is usually plotted with concen-
tration expressed as   e   on the horizontal axis.
                     4.6V
     The natural log is converted to base 10 log by:
                      In X = 2.3 log X
                                  P               P
             e = 1.987 T 2.3 log -- =4.6 T log

             molar volume = V molecular weight
                                   density
                               284

-------
               divide both  sides  of  equation by V

               divide both  sides  of  equation by 4.6
     The adsorption isotherm (gas  phase)  for butane on GAC
Filtrasorb 400 is shown  in  Figure  169.   On  the horizontal axis,
"0" is adsorption on GAC from pure butane in the gas phase.  The
higher numbers on the  axis  represent lower  concentrations of
butane (usually in nitrogen gas) .   The  seven data points making
up the Butane Gas |>hase  curve in Figure 169 (and the four data
points for the Butane  Gas Phase  on XE-340 in Figure 170) were
supplied by Rosene at  the Calgon Corporation.  His data extended
to the 0.1 cc range on the  "Y" axis. Later he indicated that
subsequent work on GAC resulted  in data points falling on a
tangential straight line extension of the curve below 0.1 cc.
This extension is drawn  in  Figure  169 as a  dash line segment.
Later in this discussion we will extend this straight line still
further for both GAC and XE-340  to get  some idea of the behavior
of very low concentrations .

     To determine how  another gas  substance, other than hydro-
carbon, will adsorb on the  same  GAC (if it  is a substance that
follows the modified Polanyi Theory), one can calculate a theo-
retical adsorption isotherm using  a scale factor YS/ based on the
ratio of the polarizability of a given substance p  to the polariz-
ability of butane
     The theoretical  curve is obtained by multiplying  any point
on the butane curve by ye •
                         O

     Using the Polanyi Theory modifications of Hansen  and
Fackler, as well  as the modifications of Manes and Hofer, one
can calculate a theoretical adsorption isotherm curve  for a
substance from a  solution, in our case water,  by the following:

where C  = solubility of the substance in water (gm/lOOcc)
water) .s
     C = concentration of the substance in the influent water.
     The scale  factor for a liquid substance (s)  adsorbed from
 another liquid  (in  our case from water) is Ysl-
                               285

-------
100

                           4.6V
    Figure 169.   Chloroform  adsorption by 0.76 meter  (2.5 feet)of GAC.
                                  286

-------
     100.
 SemMxifiarUhiSc
4 Cycles z 10 to the Inch
         Figure 170,   Butane adsorption curves for K-400  GAG and XE-340  resin.
                                            287

-------
                          Ysl = Ys -
            = scale factor for substance to be adsorbed

            = scale factor for water
Y  »
 e
                          (1.33281) +2
                                      = 0.206
                         Y , = EL - 0.206
                          S    P

For chloroform the calculated y , is:
                    ni2-!

                    ni2 + 2
(1.44643)2+2
                       = .2669
                        (refractive index
                         from Table 4 7 )
                          ^2669.
                          . ^ Jo
Using this scale factor (.93), take as many points on the butane
curve (Figure 169 )  as needed to produce the predicted curve for
adsorption of chloroform from pure water on Filtrasorb 400.  For
example, take the lowest data point "A" on the Butane Gas Phase
(GAG) curve having coordinates 0.1 cc and 26.2 for . gv.  The
horizontal coordinate corresponding to 0.1 cc on the 'predicted
chloroform curve is .93 (26.2) = 24.4.  This is point "B".
Continuing in this manner generates the predicted curve for
chloroform shown in Figure 169.  This curve indicates the cc of
chloroform that will be adsorbed from purified water per 100
grams of Filtrasorb 400.  Physical data for the halogenated
organic compounds required for application of the Polanyi-Manes
adsorption theory are provided in Table 47.  As seen in Table 47,
the solubility of chloroform is 0.82 grams/100 cc of water.

                           e   _ T     Cs
                         4.6V  - v 10g C~

            T = 295

            V = 80 (from Table 47 )
                               288

-------
                                                    TABLE 47.   PHYSICAL DATA FOR POLANYI-MANES CALCULATIONS
N3
00
VO
Cpd.
No. Compound
1
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20



Methylene chloride
Trans 1 , 2-dichleroethene
1 , 1-dichloroethane
Cis-1 , 2-dichloroethene
Chloroform
1,1, 1-trichloroe thane.
1,2-dichloroethane sum
Carbon tetrachloride
Trichloroethylene
Bromodichloromethane
Tetrachloroethylene
Chlorodibromome thane
Chlorobenzene
Bromoform
p-chlorotoluene
m-dichlorobenzene
p-dich lorobenzene
o-dichlorobenzene
Vinyl chloride

Total

ED4
Cone,
in fin.
water
ug/L
SD
.77
.4
24.1
67.3

7.7

.68
47
.003
33.6
.36
2.5
.1
nil
.21
.14
6.2

191.6

ED4
Cone, of
cpds.5,6, Ysl
11 S 13 Scale
pg/L Factor
.875
.931

24.1 .937
67.3 .93
.906
.928
.961
.993
47 1.033
1.052
33.6 1.14
1.092
1.24
1.081
1.135
1.084
1.148
.752

172
90* of Total
Solubility
e in water
4.6V gra/100ce
22°
22.68 .63

20.17 .35
18.75 .8222*
.8215
.86920
.082°
20.86 .1
18.45 .60622*

29*
18.80 .519
16.68 -MS%1°
21.00 *,g30

.012325
13.88 .007925
15.75 .014525
24.06 	 .2825
17.7 	 009


Molecular
wt.
84.
96.
98.
96.
119.
133.
98.
153.
131.
163.
165.
208.
112.
252.
126.
147.
147.
147.
62.



93
94
97
94
38
42
96
82
39
83
85
29
56
75
58
01
01
01
6 lig.
gas


Density
gm/cc
1.
1.
1.
1.
1.
1.
1.
1.
1.
2.
1.
2.
1.
325
26
177
284
492
338
256
594
464
006
623
451
106
2.89
1.
1.
1.
1.
.
m


07
288
241
305
9013
00279


Molar
Volume Dipole
cc/mole Moment
64.1 1.54
76.
84.
75.
80.
99.
78.
96.
89.
81.
102.
85.
101.
87.
118.
114.
118.
112.
69.



9
1
5 1.9
0 1.02
7 1.79
8 1.19
5 0
8 1.22
7
2
0
8 1.7(1.55)
5 1.8
3
1 1.72
5 0
7 2.52
34



Refractive
Index
1.424
1.4490

1.4519
1.44643
1.4377
1.4448
1.46305
1.4777
1.4964
1.5055
1.5482
1.52479
1.5980
1.5193
1.54570
1.52104
1.5518
1.370



                  ND = not determined

                  *0ur analysis in tap water at 22°C

-------
            V -82
            C = concentration of chloroform in water

              = 67.3 ug/liter from ED4

              =6.73 ug/100 cc

              = .00000673 gm/100 cc
              £  	 295 i _ _. .82	 -I Q -jj-
            T76V ~ ~W i0g .00000673 ~ 18-75

This means that at 18.75 on the horizontal axis  (Figure 169)
which corresponds to a concentration of 67.3 ug/L of chloro-
form in purified water, we expect Filtrasorb 400 to adsorb about
0.68 cc  (1.015 grams) per 100 grams of GAC.  Manes and his asso-
ciates have studied the adsorption of several halogenated organic
compounds, including chloroform, from pure water on GAC. Actual
data points on chloroform coincided very well with predicted
values.

Theory Application—

     Application of the Polanyi-Manes Theory to interpretation of
data in this report first considers chloroform adsorption from
finished water by 0.76  (2.5 feet) meter  deep columns of GAC.
Two runs were made, EDS and ED4, Table 48 .  In these two ED,
average influent levels were 57 yg/L and 67.3 yg/L respectively.
Adsorption per 100 grams of adsorbent at saturation were
0.0280 cc and 0.0358 cc respectively.  The adsorptive capacity
increased as the concentration increased, as predicted by the
Polanyi-Manes Theory.  These two finished water data points are
plotted in Figure 169.  The two vertical lines from the "X" axis
were drawn from the appropriate  k >v  values shown in Table 48.
These vertical lines were then projected from their intercept
points on the Chloroform Predicted  (GAC) curve to the "Y" axis.
For ED3 and ED4, respective predicted adsorption values were
0.64cc and 0.68 cc from pure water compared to actual adsorption
values of 0.028 cc and 0.0358 cc.  Respective actual values are
4.4 percent and 5.3 percent of the predicted values.  These data
appear in Table 48.  This reduction is the result of competitive
adsorption by DOM, including other HOC.  To construct an actual
adsorption curve through the two finished water data points it
would be better if the two points were further apart.  However,
working with what is available, we can calculate the actual YS!
value for these two points from our actual water.  Projections
horizontally from the two data points to the Butane Gas Phase
(GAC)  curve indicate   z  values of 28.8 and 28.4 respectively.
                     4. bv
                               290

-------
       TABLE 48.   CHLOROFORM ADSORPTION DATA
                    FROM FINISHED WATER
ED
1
1R
'2

3
3
4
4
4
4
£
t3 a
<]> (U
ffl T3
Feet
2.5
2.5
2.5

2.5
2.5
2.5
5
7.5
10
Adsorbent
XE-
340
XE-
340
XE-
340
GAG
904
GAG
GAG
GAG
GAG
r Average
*x influent
f
80.2
69.3
64

57
57
67.3
67.3
67.3
67.3
«
1 Adsorption per
w 100 grams adsorbent
at saturation
.265
.22
.2

.042

.0534
.067
.071
.073
e
CC
.177
.148
.134

.028

.0358
.049
.048
.049
4.6V
18
18
18

19

18
18
18
18
.45
.71
.83

.02

.75
.75
.75
.75
Predicted Polanyi-
o Manes adsorption
for GAG


.64

.68
.68
.68
.68
Percent adsorption
* of predicted value


4.

5.
7.
7.
7.
§
M
0

-------
Respective actual YS! values are 0.660 and 0.660.  Using  this
Ysl value, the actual adsorption curve, shown in Figure  169, was
generated, and it passed through the two data points.  We predict
that for our water, this generated actual chloroform adsorption
curve can be used to predict chloroform results on our water over
the entire chloroform concentration range we will experience.  To
generate this curve, only one actual data point is required to
calculate the actual Ysl value.  Since there was nil chloroform
in our raw water, we cannot test this prediction from our data on
chloroform.  In the discussion on XE-340 which follows we will
show that it appears to work down to a very low chloroform level.
The curves in Figure 169 are a log-log plot since the equation
for _!•— contains a log function.  In our water, the concentra-
tions'of HOC are within the tangential straight line portion of
the adsorption curve, therefore the log-log plot of data  points
mentioned earlier in this report, fall on a straight line and the
line can be used to predict adsorption values at different con-
centrations .

     Chloroform adsorption from finished water by 0.76  (2.5 feet)
meter  deep columns of XE-340 were studied in three runs,  EDI,
ED1R, and ED2, Table 48 .  The Butane Gas Phase (XE-340) curve
and the generated Chloroform Predicted (XE-340) curve for pure
water appear in Figure 170-  The predicted curve was again
generated from the gas phase curve by using the 0.93 Ysl  scale
factor  (Table 47) for chloroform.  The average influent level
varied from 80.2, 69.3 to 64 yg/L.  For these three runs,  compare
the grams of chloroform adsorbed per 100 grams of adsorbent at
saturation with their respective average influent level.   The
adsorptive capacity of XE-340 for chloroform decreases  (0.177 cc,
0.148 cc and 0.134 cc) as the influent concentration decreases,
as predicted by the Polanyi-Manes Theory.  Using  £   values
                                                 4.6V
(from Table 48) corresponding to the three chloroform concentra-
tions of 80.2, 69.3 and 64 yg/L, the three levels of cc's
adsorbed per 100 grams of adsorbent at saturation (also shown in
Table 48) are plotted in Figure 171.  The three vertical  lines
from "X" axis correspond to the three   e   values.  The  three
finished water data points are replotted'in Figure 172  on  an
expanded "X" and "Y" axis scale for greater accuracy to show that
they fall approximately on a straight line.  To minimize  drawing
error the slope of this was transferred to the scale in Figure
171 and the actual adsorption curve drawn as shown through the
three finished water data points.  It is apparent that the actual
adsorption curve is not parallel to the Butane Gas Phase  (XE-340)
curve nor to the Chloroform Predicted (XE-340) curve.  The
predicted curve predicts adsorption from pure water of 0.105 cc,
0.094 cc and0.09cc instead of the 0.177 cc, 0.148 cc and
0.134 cc actually adsorbed from our finished water.  Obviously,
the Polanyi-Manes Theory does not apply to XE-340, an adsorbent
                               292

-------
9_
8..
7_
....	 " "	 !    ' "~~"  i    "1!     I——|-
                                                         _water_dat;a_Eoints;i'... _l
                         "

.01
                                                      0.76 me
                       (2.5 feet)of XE-340  (EDI,  ED1R and ED2).
                                    293

-------
I
CO
=3

O
01
O

O
i-H
a
V

•a
O
(0
•d
(8

u
O
                                                                          19.0
                                     4.6V

        Figure 172.  Chloroform adsorption by  0.76 meter   (2.5 feet)

                     of XE-340.
                                    294

-------
that not only allows micropore  surface  adsorption  but  adsorption
into the polymer matrix.   However,  if at  least  two actual ad-
sorption data points are  obtained on a  water  system, the actual
adsroption curve could be drawn to  predict  adsorption  at some
other concentration on the straight line  portion of the curve.
We do not have finished water experimental  data to test this
possibility but we do have a data point on  H.T. water  in EDlR,
Table 19.  Competitive adsorption by HOC  and  TOC will'be differ-
ent in H.T. and finished  water  so one cannot  expect too much
from this comparison.  The H.T.  data point  is plotted  in Figure
173 at the appropriate    e  value  of 25.2.   From  the  prediction
                        4.6V
curve we predict an adsorptive capacity of 0.0023 cc and we reported
an observed capacity of 0.0027  cc.   Granted,  error possibilities
are great, but at least even at this very low concentration
 (1.2 yg/L of chloroform)  adsorption does  occur, can be measured
and the adsorptive capacity predicted with  some degree of
success.

     Adsorption of cis 1,2-dichloroethene on  0.76  (2.5 feet)
meter  of XE-340 was studied in EDI, EDlR and ED2.  The Butane
Gas Phase  (XE-340) curve  appears in Figure  174.  From  this curve,
the cis 1,2-dichloroethene predicted (XE-340) curve was genera-
ted by using the scale factor ys^ of 0.937  appearing in Table
47.  As with chloroform adsorption  on XE-340, Figure 171, plot-
ted cis 1,2-dichloroethene adsorption points  in Figure 174 lie
above the predicted curve and,  in the case  of raw  and  H.T. data
points, lie above the Butane Gas Phase  (XE-340) curve.  Lines
drawn through the data points are not parallel  to  the  predicted
curve.  As with chloroform XE-340 data, the Polanyi-Manes Theory
does not apply to adsorption of cis 1,2-dichloroethene by XE-
340.  Again however, two  actual data points on  raw, H.T. and
finished water should be  sufficient to  generate an actual ad-
sorption curve for our samples.   In Figure  174, notice how ad-
sorption data points from raw and H.T.  water  fall  nearly on the
same straight line.  Notice also how adsorption points for fin-
ished water fall on a displaced straight  line,  indicating con-
siderably less adsorptive capacity  per  100  grams of adsorbent.
This reduction in adsorption of about 30  percent is due, we
believe, to increased competitive HOC adsorption,  as was dis-
cussed for chloroform on  page 112 and for cis 1,2-dichloroethene
on pages 113 and  1S3.  The adsorptive capacity  of  XE-340 is
about the same from raw and H.T. water, indicated  by the actual
data points falling on almost the same  curve.   The capacity of
XE-340 to adsorb from finished  water was  less.  The competition
of TOC adsorption decreased from raw to H.T.  to finished water,
corresponding to TOC values of  8.3, 5.8,  and  5.4 mg/L.  This
would indicate possibly greater adsorptive  capacity of cis 1,2-
dichloroethene from finished, water.  The  level  of  HOC  in finish-
ed water is approximately 150 yg/L  compared to  cis 1,2-dichloro-
ethene levels 25 yg/L in  raw and H.T. water.  The  competitive
adsorption of other HOC would indicate  less adsorption capacity

                              295

-------
.00*-
                                         A
     18.83                             25.2
Figure 173.  Chloroform adsorption by  0.76  meter  (2
             XE-340  (EDI, ED1R and ED2).
                                                         5 feet) of
                                   296

-------
1..0..



                       points:: (EDl .and :EDlR)_;r ^-p:tr
        .".Finished v/ater_data_^Epints_(EDl
                  	'	|	
                                     17    18   19   20    21
     Figure 174.  cis  1,2-Dichloroethene adsorption by 0.76 meter
                   (2.5 feet) of  XE-340.(ED1, ED1R and ED2).
                                        297

-------
in finished water.  The data showed less adsorption; therefore,
this may suggest that competitive adsorption of HOC has a  great-
er influence than TOC competitive adsorption from finished water.

     Adsorption of cis 1,2-dichloroethene on 0.76 (2.5 feet)
meter  of GAG was studied in ED3 and ED4.  Plotted curves  and
actual data points appear in Figure 175.  Raw and finished water
data points, as with XE-340 in Figure 174, fall on displaced
lines, indicating again the reduction in adsorptive capacity due
to increased competitive HOC adsorption.  The lines through the
actual data points in Figure 175 were drawn through the points
by sight and by calculating the ysl value for each point and
generating the line from Butane Gas Phase (GAG) curve.  Since
the two methods produced the same lines, on GAG columns, one
actual data point is enough to generate an actual adsorption
curve.  The predicted Polanyi-Manes adsorptive capacities  for
GAG in EDI, EDlR, EDS and ED4 appear in Table 49.  The percent
of actual adsorption in our water is also shown.  In two separ-
ate runs, columns 0.76 (2.5 feet) meter  deep on raw water both
adsorbed 8.8 percent of the predicted adsorption value from pure
water.  On finished water two runs at the same bed depth adsorb-
ed 6.4 and 6.5 percent of the predicted adsorption value from
pure water.  In ED4, as bed depth increased, the adsorptive
capacity increased as expected due to less competitive HOC ad-
sorption with increasing bed depth.

     Adsorption data from Table 50 for bromodichloromethane by
0.76  (2.5 feet) meter  of XE-340 from H.T. and finished water
are plotted in Figure 176.  Since the H.T. water data point is
probably not on the same straight line as the finished water
data points, the dashed line drawn in Figure 176 represents only
an approximate actual adsorption curve for this HOC in our sys-
tem.  Adsorption data from Table 50 for bromodichloromethane by
0.76  (2.5 feet) meter  of GAG from finished water are plotted in
Figure 177.  In this case the bromodichloromethane predicted
(GAC) curve lies above the Butane Gas Phase (GAG) curve because
the YS! value for this HOC in Table 47 is 1.033.  The calculated
Ysi value for both data points, obtained from the curves,  was
0.652.  This indicated that the two points fall on a straight
line parallel to the Butane Gas Phase (GAC)  curve.  Thus,  one
data point would have been sufficient to generate the actual
adsorption curve as shown.  The actual adsorption of this  HOC
in our water is only about 3.6 percent of the predicted value
from pure water.

     Adsorption data from Table 51 for chlorodibromomethane by
0.76 (2.5 feet) meter  of XE-340 from H.T. and finished water
are plotted in Figure 178.  The approximate actual adsorption
curve is shown.  Adsorption data from Table 51 for adsorption
by 0.76 (2.5 feet) meter  of GAC from finished water are plotted
in Figure 179.   The calculated ysl value for both data points,
obtained from the curves, were 0.602 and 0.598, averaging  0.6.

                              298

-------
 _O _Baw._water. data, points. ;_P l.U.a. I|\-!_M.. -l-)-|-(
 ; •         I           !       ; •  \ X  :.....:

 D Finished water data points  - •.  \ ^
Figure 175.  cis 1,2-Dichloroethene adsorption by 0.76 meter

               (2.5 feet)  of GAX:  (ED3 'and ED4) .
                                   299

-------
     TABLE 49.   CIS 1,2-DICHLOROETHENE ADSORPTION DATA  FROM
                 RAW,  H.T., AND  FINISHED WATER
ED
Bed
depth
feet
Adsorption per
100 grams Predicted Percent
Average adsorbent at Polanyi adsorption
influent saturation adsorption of
E predicted
Adsorl ant
yg/L Grams
CC 4.6V CC
value
HAW WATER
1
1R
1R
1R
2.5
2.5
2.5
2.5
GAC
XE-340
GAC
XE-340
21
21
29
29
.048
.15
.065
.181
.037 20.4 .42
.117 20.4
.0509 19.36 .58
.141 19.86
8.g

8.8

H.T. WATER
1
1R
2.5
2.5
XE-340
XE-340
?0
25.4
.134
.157
.104 20.49
.122 20.08


FINISHED WATER
1
1R
2
3
4
4
4
4
2.5
2.5
2.5
2.5
2.5
5
7.5
10
XE-340
XE-340
XE-340
GAC
GAC
GAC
GAC
GAC
10.9
19.4
18.4
18.3
19.9
19.9
19.9
19.9
.043
.093
.087
.033
.039
.045
.049

.003
.0724
.068
.0257 20.57 .4
.03 20.17 .46
.035 20.17 .46
.0357 20.17 .46
20.17



6.4
6.5
7.6
7.8

2.5 feet=0.76 meter
5 feet=1.52 meters
7.5 feet=2.29 meters
10 feet=3.05 meters
                                300

-------
    TABLE 50.  BROMODICHLOROMETHANE ADSORPTION DATA FROM
                H.T. AND  FINISHED WATER


ED


1

1R
Bed
depth
feet


2.5

2.5
Adsorption per
100 grams
Average adsorbent at
influent saturation

Adsorbent


XE-340

XE-340


Vg/L Grams

variable
see Pig.
erratic
see Fig.
H.T.

39

40

CC
WATER

.0028


Predicted Percent
Polanyi adsorption
adsorption of
e
4.6V CC


23.1

26.9
predicted
value





FINISHED WATER
1
1R
2
3
4
4
4
4
2.5
2.5
2.5
2.5
2.5
5
7.5
10
XE-340
XE-340
XE-340
GAC
GAC
GAC
GAC
GAC
37.1
42.7
42.4
39
47
47
47
47
.182
.204

.069
.084
.083
.084

.0907
.1017

.0344
.0419
.0414
.0419

18.82
18.60

18.74 1.13
18.45 1.21
18.75 1.14
18.75 1.14
18.75



3.04
3.46
3.63
3.68

2.5 feet=0.76 meter
5 feet=1.52 meters
7.5 feet=2.29 meters
10 feet=3.05 meters
                                 301

-------
              SpiffliiSi
liijjjffr -^-jjg|^^f£y£L:l3y^5i3 ^T^rtt^
^J^S^^I^S^¥S^':^^:^^^^^
j^^^i^-ff^iBK^^^^^^f^^
      302

-------
• 1-!
                                        ; • ' I-TJ j si" f =• 11 = ;

    4^-^-r~4i44^4^u44^                      ^r  •':  ;:;1
                                                    -•—i--H' . —i	1—^ *..j .j-


                                                    ?}^f^E!^

                                                     '•—'-tH- :4-t-(-tJ:F
                                                  iv  - ----- --•-

                                                -4-V-H-Hh
                                                -pj-ye^-t—r-i-
.01
  18
  Figure 177.  Bromodichloromethane adsorption by 0.76 meter

              (2.5 feet) of GAC (ED3 and.ED4).
                               303

-------
          TABLE  51.  CHLORODIBROMOMETHANE ADSORPTION DATA
                      FROM H.T. AND FINISHED WATER
    Bed
    depth
ED  feet
Adsorbent
                    Adsorption per
                    100 grams
          Average   adsorbent at
          influent  saturation
Grams   CC
      Predicted   Percent
      Polanyi     adsorption
      adsorption   of
_£	              predicted
4.6V      CC      value
1   2.5     XE-340

1R  2.5     XE-340
                    H.T.  WATER

             1.0           .0009    23.31

              .25          .0005    25.39

                    FINISHED WATER
1
1R
2
3
4
4
4
4
2.5
2.5
2.5
2.5
2.5
5
7.5
10
XE-340
XE-340
XE-340
GAC
GAC
GAC
GAC
GAC
12 .077 .031 19.56
24.5 .155 .063 18.48
26.7
27 .067 .027 18.34 2.2
33.6 .104 .042 18.01 2.3
33.6 .1 .041 18.01
33.6
33.6



1.2
1.8



 2.5 feet=0.76 meter
 5 feet=1.52 meters
 7.5 feet=2.29 meters
 10 feet=3.05 meters
                                   304

-------
ts
a)
•8
o
CQ
O
O
•a

I
.001
  .0001
        Figure 178.  Chlorodibromomethane adsorption by 0.76 meter
                      (2.5 feet) of XB-340  (EDI  and ED1R).
                                   305

-------

                                      en data points
                         _IjEi)3.!and'jED4 T.TI
18   19    20  21   22   23   24   25   26  27    28  29    30   31   32   33
                                 4.&V
Figure  17.9.   Chlorodibromomethane  adsorption from 0.76 meter
              (2.5 feet) of GAC  (ED3  and ED4).
                                  306

-------
Once again, this indicated  that  both  points  lie  on  the same
straight line parallel  to the  predicted curve.   The actual ad-
sorption of this HOC  in our water is  only  about  two percent of
the predicted value from pure  water.

     Adsorption data  in this part of  the report  for the HOC dis-
cussed above and for  additional  HOC appear in Table 52.  The
chemical identification number given  in Table 8,  for each HOC
studied correspond to their elution time on  a GC through a
Tenax column  (except  for vinyl chloride, Chem. No.  20, which
elutes first).  Chemical No. 3 is eluted first  (after vinyl
chloride), and  17, 18,  and  19  are eluted last.   In  Table 52, it
is seen that  the Ysl  values in most cases  predict this order of
elution from  a  Tenax  column.  This table supports and adds to
the data in Table  37, which is discussed beginning  on page 193.

     We have  found the Polanyi-Manes  Adsorption  Theory useful in
interpreting  and explaining the  data  we have obtained.  It is
obvious that  more  fundamental  research is  needed in the labora-
tory to determine  the actual adsorption of all our  HOC from
pure water.   This  will indicate  whether or not our  observed
reduction  in  adsorption as  the molecular weight  of  the HOC in-
creased,  is due to steric  exclusion or largely to competitive
HOC adsorption.  We  found  the Theory  useful in predicting re-
sults  in  a given  system where the concentration  of  dissolved
substances change  in magnitude but not in  ratio.
                               307

-------
                                          TABLE 52.  SUMMATION OF ADSORPTION PARAMETERS BY GAC
u>

Cpd.
No. Chemical
3 Trans 1, 2-dichloroethene
5 Cis-1 , 2-dichloroethene
6 Chloroform
7>v 1 f 1 / 1- trichloroethane
\ \
8 ^>l,2-dichloroe thane sum
y Carbon tetrachloride
10 Trichloroethylene
11 Bromodichloromethane
13 Chlorodibromomethane
15 Bromoform
17v m-dichlorobenzene
18 ^ p-dichlorobenzene sum
1» o-dichlorobenzene
Column
Bleed
Time
Days
52
18
7


29

38
15
21
42

none

Column
Saturation
Time z
Days Inches
? 16+?
65 22
23 21


76 18

56 9.6
54 21
89 22
94 16.6

>122 <10

Ysl
.931
.937
.93
.906

.928
.961
.993
1.003
1.14
1.24
1.135
1.084
1.148
Polanyi-Manes
Predicted
Capacity
cc
.03
.46
.68




.65
1.14
2.2
1.7

4.9

Observed
Capacity Percent of
Predicted
eg
.002 6.7
.029 6.5
032 5.0




.0012 .19
.04 3.6
.04 1.8
.002 .12



                                                                                            Total     .1462

-------
GC/MS HOC  Confirmation Data

     Periodically during the  two-year  study, GC/MS determinations
were made on both raw  and  finished water to confirm GC peaks of
HOC.  All nineteen HOC have been  confirmed.  Sample dates and
results are summarized in  Table 53.  Analyses on sample dates
August and November  4, 1976;  and  May 3 and May  20, 1977 were
determined on a Hewlett-Packard Mass Spectrograph, Model 5981,
with a Model 5933 Dual Disc Inter-Active Data System.  The data
on October 11, 1977  were obtained by EPA Laboratories in
Cincinnati.  Analyses  on sample dates  March 11  and April 17, 1978
were made on a Varian  MAT  Model 112S Magnetic Sector, double-
focusing, high resolution  mass spectrograph coupled to a MAT 166
data system.
                                309

-------
                                                     TABLE  53.  GC/MS HOC CONFUTATION DATA
U)
M
O
I 4/17/78
Cham Raw
No. vg/
20 14.2
. 1 .45
3 j 1.6
4 | .6
5 J24.3
6 j nil
7 £
8 jj.13
9
10
11
12
13
14
15
16
17-W
ben-
^^ns
tol-
uene
I
.37
nil
.003
nil
1.2
nil
.03
nil
.57.19
NR
MR ,
Wtr.
KS*
Y
Y
Y
Y
Y
ND
ND
Y
Y
Y
Y
Y
ND
Y
ND
Y
ND
J2L
Y
Y
Pin.
pc/
11.3
int
_st<3
.87
.5
22.1
69

a.i

.63
46
.003
34.3
.81
2.7
.11
sy*
NR
NR
Wtr.
MS
Y
NR
Y
Y
Y
' Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y/Y
Y
Y
3/11/78 . .
Raw
\ig/
12.8
.52
1.3
.7
27.4
nil

j.17

.41
nil
.003
nil
1.3
nil
.04
JB&
NR
NR
Wtr.
MS
Y
Y
Y
Y
Y
ND
ND
Y
Y
Y
ND
Y
ND


ND
Y
Y^Y
Y
Y
Fin.
pg/
11.7
int.
stcU
.78
.61
25.2
73

7.8

.72
43.4
.002
37.2


2.9
.13
JV-19
NR
NR
Wtr.
MS
Y
NR
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y


Y
Y
!ND
1X/Y
Y
Y
10/11/77.**
Raw
V<3/
10.8
.13
2.1
X.O
19.8
nil

.16

.58
nil
nil
nil


nil
.87
1.2
ND

Wtr.
MS
Y
Y
Y
Y
Y
ND
ND
Y?
ND
Y
ND
N0
ND


ND
A
A
Y

•Fin.
v,
-------
                          REFERENCES

1.  Symons, J.M., et al.  National Organics Reconnaissance
    Survey for Halogenated Organics in Drinking Water. JAWWA,
    67(11) :634-647, 1975.  Update, 67 (12):708-709, 1975.

2.  Bellar, T.A. and J.J. Lichtenberg.  Determining Volatile
    Organics at Microgram per Liter Levels in Water by Gas
    Chromatograph.  JAWWA 66 (12):739-744, 1974.

3.  Dressman, R.C. and E.F. McFarren.  Sample bottle purging
    method for determination of  vinyl chloride in water at sub-
    microgram per liter  levels.  J. of Chrom. Science, Vol. '15,
    pp. 69-72, 1977.

4.  Stevens, A.A. and J.M. Symons.  "Trihalomethanes and Precur-
    sor Concentration Changes Occurring During Water Treatment
    and Distribution."   JAWWA 69(10) :546  (Oct. 1977).

5.  Aldrich Chemical Co., Inc.   Phase-Transfer Catalysis in
    Organic Synthesis.   Aldrichemica Acta, Vol. 9, No. 3 (Final
    Issue), 904 W. St. Paul Ave., Milwaukee, WI 53233, 1976.

6.  Neely, J.W.  A model for the removal of trihalomethanes from
    water at Ambersorb XE-340.   Rohm and Haas Company, Research
    Laboratories, Spring House,  PA 19477.  IN:  Proceedings of
    the ACS Annual Meeting, Environmental Section:  Activated
    Carbon Adsorption of Organics from the Aqueous Phase,  Miami
    Beach, FL, 1978.

7.  DeMarco, J. and P. Wood.  Design Data for Organics Removal
    by Carbon Beds.  IN:  Proceedings of National Conference on
    Environmental Engineering, Research Development and Design,
    Kansas City, MO, 1978, pp. 149-156.

8.  Symons, James M.  Interim  treatment guide for controlling
    organic contaminants in drinking water using granular acti-
    vated carbon.  Water Supply  Research Division, Municipal
    Environmental Research Laboratory, Office of Research and
    Development, Cincinnati, Ohio, January 1978.

9.  Van Der Kooij , D.  Some investigations into the presence
    and behavior of bacteria in  activated carbon filters.   IN:
    Translation of reports on special problems of water tech-
    nology, H. Sontheimer, ed.  Vol. 9, Adsorption, op. cit.,
    MERL,  USEPA, Cincinnati, OH, 1976, pp. 348-354.

                              311

-------
10.  Polanyi, M.,  Verb. dent. Physik.  Ges., 18, 55, 1976 and
     M. Polanyi, Physik, 2, 111, 1920.

11.  Hansen, R. and W. Fackler, Jr.  A Generalization of the
     Polanyi Theory of Adsorption from Solution.  The Journal of
     Physical Chemistry, Vol. 57:634-637, 1953.

12.  Manes, Milton and L. Hofer.  Application of the Polanyi
     Adsorption Potential Theory to Adsorption from Solution on
     Activated Carbon.  J. Phys. Chem., 73(2):584-590, 1969.

13.  Wohleber, D.  and M. Manes.  Application of the Polanyi Ad-
     sorption Potential Theory to Adsorption from Solution on
     Activated Carbon. II.  Adsorption of Partially Miscible
     Organic Liquids from Water Solution.  J. Phys. Chem., 75(1):
     61-64, 1971.

14.  Wohleber, D.  and M. Manes.  Application of the Polanyi Ad-
     sorption Potential Theory to Adsorption from Solution on
     Activated Carbon III.  Adsorption of Miscible Organic
     Liquids from Water Solution.  J. Phys.  Chem., 75(24) :3720-
     3723, 1971.

15.  Chiou, C. and M. Manes.  Application of the Polanyi Adsorp-
     tion Potential Theory to Adsorption from Solution on Acti-
     vated Carbon.  V.  Adsorption from Water of Some Solids and
     Their Melts,  and a Comparison of Bulk and Adsorbate Melting
     Points.  J. Phys. Chem., 78(6):622-626, 1975.

16.  Schenz, T. and M. Manes.  Application of the Polanyi Adsorp-
     tion Theory to Adsorption from Solution on Activated Carbon.
     VI.  Adsorption of Some Binary Organic Liquid Mixtures.  J.
     Phys. Chem.,  79(6):604-609, 1975.

17.  Rosene, M. and M. Manes.  Application of the Polanyi Adsorp-
     tion Potential Theory to Adsorption from Solution of Activa-
     ted Carbon.  VII.  Competitive Adsorption of Solids from
     Water Solution.  J. Phys. Chem., 8 (9) :953-959, 1976.

18.  Rosene, M. and M. Manes.  Application of the Polanyi Adsorp-
     tion Potential Theory to Adsorption from Solution on Activa-
     ted Carbon.  VIII.  Ideal, Non-ideal, and Competitive Ad-
     sorption of Some Solids from Water Solution.  J. Phys. Chem,
     80(23):2586,  1976.

19.  Rosene, M. and M. Manes.  Application of the Polanyi Adsorp-
     tion Potential Theory to Adsorption from Solution on Activa-
     ted Carbon.  IX.  Competitive Adsorption of Ternary Solid
     Solutes from  Water Solution.  J. Phys.  Chem., 81:1646, 1977.
                              312

-------
20.  Rosene, M. and M. Manes.  Application of the Polanyi Adsorp-
     tion Potential Theory to Adsorption from Solution on Activa-
     ted Carbon.  X.  pH Effects and Hydrolytic Adsorption in
     Aqueous Mixtures of Organic Acids and Their Salts.  J. Phys.
     Chem., 81:1651, 1977.
                                313

-------
        MICROBIAL FLORA OF GRANULATED ACTIVATED CARBON
                COLUMNS USED IN WATER TREATMENT
                             (Part I)

                              by
                      Frances Parsons
           Drinking Water Quality Research Center
               Florida International University
                        Tamiami Campus
                        Miami, Florida
                           ABSTRACT

     Differential bacteria counts were made on samples of efflu-
ents of granulated activated carbon (GAC) columns used to remove
dissolved organic material from drinking water.  The membrane
filter procedure, four primary media,  and incubation at 25°C for
six days were used to isolate colonies.  Identification was done
using Roche Diagnostics systems and additional diagnostic tests.
Most of the growth occurred on tryptone glucose extract agar and
Czapek Dox agar after four days incubation at 25°C.  Most bacte-
rial growth was not detected when standard methods were used.

     Raw water organisms, which apparently can survive existing
treatment plant processes, colonized the initially bacteria-free
GAC columns, and released vast numbers of bacteria into the
water flowing through the columns.  Some of the organisms,
though innocuous in small numbers, may pose a threat to human
health when they are present in drinking water in large numbers.
These organisms include Pseudomonas-like bacteria, Acinetobac-
ter, Alcaligenes faecalis, Moraxella,  Enterobacter agglomerans,
and Flavobacterium (probably aquatile).

     The development of bacterial growth in the GAC columns
interfered with backflushing the columns.  Preliminary results
indicate that the GAC system provides an ecological advantage
for entering organisms that survive treatment plant processes.
Enterobacter agglomerans in GAC column effluent survived expo-
sure to 3ppm chlorine.  This suggests that at least in our sub-
tropical environment a careful study is required to insure prop-
er bacterial control before installing GAC adsorbers in treat-
ment plants.
                               314

-------
           CONCLUSION  AND RECOMMENDATIONS  (TENTATIVE)


     Bacteria  that  occur in small numbers  in  raw water survive
treatment and  colonize granulated activated carbon  (GAG) columns
used to remove organic solutes from treated water.  The bacteria
multiply, form slime that interferes with  column maintenance by
preventing backflushing, and slough off  in large numbers into
the water passing through the columns.

     The size  and composition of  the microbial populations in
GAG columns changed with time.  The composition of  the microbial
population of  the raw  water apparently influenced the population
in the columns.  Each  column had  a somewhat different population
composition and size on each sample date.

     Some of the organisms that multiply in the GAG columns may
pose a health  hazard because of the vast numbers present if the
column effluent is  ingested or comes in  contact with susceptible
bbdy surfaces  such  as  the otic canal or  the naso-pharyngeal mu-
cosa.  The possibility of a consumer incurring enteritis, in-
toxication, and/or  an  opportunistic infection should be studied.
Because of the large numbers of Gram-negative organisms that
colonize GAC columns,  endotoxin should be  assayed using the LAL
method.  Staphylococci sp.  sometimes present  in finished water
should be tested for coagulase.

     The large numbers of noncoliform bacteria found in column
effluents will suppress coliform  growth  and interfere with in-
terpretation of the standard coliform detection test.

     Effects of rechlorination of column effluents on the sub-
sequent bacterial population of this water over a period of time
is being studied.   Preliminary results indicate that entering
organisms that survive treatment  plant processes become a major
component of the microbial population in GAC  columns.  A count
of 300/100 ml  of sample of Enterobacter  agglomerans was obtained
in a GAC column effluent sample two days after rechlorination to
3 ppm.  When the concentration of chlorine was increased to 10
ppm, two colonies of Enterobacter agglomerans were recovered
from 100 ml of sample  held for six days  at 25°C. This experiment
was repeated and supplementary survival  tests using organisms
isolated from  chlorinated column  effluent  were done to verify
these results.  Cursory examination of results of these experi-
ments indicate that small numbers of bacteria can survive in

                               315

-------
water containing 3 ppm chlorine for as long as six days, probably
inside of cell aggregates.

     The Standard Plate Count method (APHA, AWWA, WPCF 1975) is
inadequate for enumerating these aquatic bacteria.  Longer incu-
bation time and lower temperatures than specified by Standard
Methods are needed.  New media that would support more kinds of
heterotrophic organisms should be developed and tested.  Better
methods for identification of these organisms need to be devised.
                              316

-------
                          INTRODUCTION


     Granulated  activated carbon (GAG)  columns  that  are capable
of retaining  organic material and removing chlorine from water can be
expected to diminish the bacteriocidal  property of treated water
passed through them and to provide  metabolic  substrate for
microorganisms that survive chlorination.   Controlling bacteri-
al populations in  treated water is  important  because bacteria
that may be harmless in small numbers may be  capable of causing
disease under certain conditions (Geldreich 1973, Peterson and
Favero 1975).  Wallis et al.  (1974)  pointed out that charcoal
filters used  in  domestic water supplies released large numbers
of bacteria to water flowing through them.  Allen et al.  (1977)
demonstrated  that  excessive bacterial populations mask coliform
growth in the  standard method for  determining  potability of
water.  Fiore and  Babineau (1977) stated that activated carbon
filters in household use had no effect  on bacteria counts in
water passed  through them.   They used the pour  plate method and
incubated the cultures at 30. *€ for a 48-hour period.  Klotz et
al.  (1976) demonstrated that 48 hours was inadequate for the
slow-growing micro flora that developed on  activated carbon fil-
ters, and incubated their cultures  at 27°C  for  seven days.  Our
work supports that of Klotz.

     Although American workers in the field of  water quality
are concerned about the increase in numbers of  bacteria in
water filtered through GAC columns,  European  workers encourage
bacterial growth in carbon filters  used at  various points in the
treatment process  (Eberhart 1976).   By  doing  so, adsorbed carbon
material is mineralized and soluble inorganic ions are immobi-
lized in bacterial biomass.   This can be an attractive feature
to American designers of water systems.  Van  der Kooij (1976),
Klotz et al.  (1976),  and Eberhardt  (1976) described the develop-
ment of bacterial  populations on carbon filters in Europe and
their activity in  mineralizing organic  substances from the water
flow.

     This study  is being done to determine  the  changes in micro-
bial population  composition and size in treated water subjected
to filtration through GAC beds of various depths.  It is neces-
sary to determine  successional changes  in bacterial populations
to determine if  a  potential health  hazard can result 1) from
development of massive populations  of ordinarily harmless bacte-
ria, and 2) from a change in  kind of bacteria multiplying in
the carbon filters from innocuous species to  chlorine resistant

                               317

-------
pathogenic species that ordinarily may be present in small num-
bers in raw water.

     It will also become increasingly important to understand
how bacterial growth on carbon filter material can be controlled
to facilitate maintenance of the filters.  Bacterial growth
tends to develop slime within the carbon granules, which makes
back-flushing difficult, and after a time impossible.
                              318

-------
                     METHODS AND MATERIALS


     The bench  model column adsorption system used  in Experimen-
tal Design No.  4  of EPA Grant Project R804521-01  (Wood et al.
1979)  is shown  diagrammatically in Figure 1.   Sample dates and
bed depths are  shown also.   The four 1" ID columns  were connect-
ed in  series  and  packed with 2.5 feet of GAG  Filtrasorb 400 to
give bed depths of 2.5, 5,  7.5, and 10 feet.   Sample ports were
located at the  effluent end of each column.   Two-liter samples
were collected  on the dates shown and analyzed within two hours.
The membrane  filter technique was used to isolate bacteria from
water  samples (APHA,  AWWA,  WPCF 1975).  Figure 2 is a flow dia-
gram of the procedure used.  Table 1 is a list of diagnostic
tests  and media used in identification.  Membranes  used to fil-
ter water samples were placed on several primary media, incuba-
ted at 25°C,  and  read at 1-, 2-, 3-,  and 6-day intervals as this
temperature and these incubation times were shown previously to
yield  higher  counts and greater diversity of  bacteria than those
specified by  Standard Methods (APHA,  AWWA, WPCF 1975).

     When the primary cultures were examined,  every recogniz-
ably different  colony type  on each primary medium was described,
assigned a number, and counted.  At least two colonies of each
type were picked  and each was streaked on a new plate of medium
to insure isolation.   Dissimilar colonies were expected to be
sometimes identical as bacteria express different morphologies
on different  media.  Gram stains were made of each  colony type
and examined  for  purity of  cell morphology.   When the purity of
the isolates  was  assured, they were inoculated into differential
media  (Figure 2),  incubated at 25°C,  and examined daily for six
days.  The pattern of biochemical reactions for each colony type
was compared  with those listed in Sergey's Manual  (1974), and
names  were assigned.

     To determine the bacterial flora of the  carbon granules
used to fill  the  filter columns, a 250 ml volume of new carbon
granules in 500 ml sterile, buffered,  deionized water was shaken
for one hour  and  then allowed to settle.   Two hundred ml of the
rinse  water was passed through each of two membranes, which were
then placed on  TGE and Endo's media.

     When it  became apparent that large numbers of  bacteria were
being  generated in the granulated activated carbon  columns, it
was necessary to  determine  the effect of rechlorination of the
effluent to control this bacterial growth.  One-liter samples of

                               319

-------
effluent from the end of the series of columns (10 feet of gran-
ulated activated carbon)  were collected and treated as follows:

     1)  effluent plated on primary media following membrane
        filtration;

     2)  effluent plus 3 ppm chlorine plated after one hour con-
        tact time;

     3)  effluent plus 3 ppm chlorine aged for two days before
        plating to  simulate residence time of water in a dis-
        tribution system with the possibility of depletion of
        residual chlorine;

     4)  effluent plus 10  ppm chlorine aged for six days to simu-
        late a condition  of overabundance of chlorine, assured
        residual chlorine,  and a lengthy residence time in a
        distribution system.
                              320

-------
                             RESULTS


     Table 2 lists the numbers  of  different kinds of microorga-
nisms isolated from raw  and  finished water from the treatment
plant prior to sampling  the  GAG columns.  Table 3 gives total
colony counts on the different  primary media obtained from raw,
finished, and filtered water samples on  11/21/77 after the GAC
columns had been in use  for  19  days.  Counts are reported for
100 ml of filtered sample  because  the low numbers of bacteria
isolated would result in fractional values if they were ex-
pressed per ml as Standard Methods (APHA, AWWA, WPCF 1975) spec-
ifies.  No bacteria were isolated  from water in which new carbon
granules had been shaken.  Table 4 lists all the different colo-
nies picked for identification,  shows the primary medium and
system location where each colony  was found, and the colony num-
ber assigned for reference during  the identification process.
Table 5 is a list of groups of colonies with similar biochemical
reactions that resulted  from growth in differential media.
Table 6 lists the different  identified organisms and their popu-
lation size in raw and finished water and column effluents.
Values given in Table 6  are  the counts obtained from the most
favorable medium.  Each  medium  favors a  different population of
microorganisms and no single medium can  yield an accurate esti-
mate of population size.

     Table 7 gives total counts on different primary media ob-
tained from raw, finished, and  effluent  samples taken on 12/5/77.
Table 8 is the initial description of colonies isolated from
samples taken on 12/5/77.  The  original  11 colonies were mixed
cultures; they were streaked on agar plates to separate the co-
habitants, and resulted  in 29 colonies that were subjected to
identification procedures.   Table  9 shows the identification
process, including colony  number assigned during the identifica-
tion process, colony description,  identity, and population size
in each part of the GAC  system.  Table 10 shows the numbers of
identified organisms isolated from samples taken on 12/5/77 in
each part of the GAC system.  Table 11 is the initial descrip-
tion of colonies isolated  from  samples taken on 1/6/78, their
numbers and location in  the  GAC system.  Table 12 shows the num-
ber of identified organisms  isolated on  1/6/78 from each part of
the GAC system.

     Group (genus) names only were assigned in the tables show-
ing organism identification  (Tables 6, 10, 12) as many of the
organisms isolated fit no  single taxonomic category with the


                               321

-------
diagnostic tests used.  Where species epithets are used in the
tables, identification is reasonably certain.  Several organisms
were identified to genus only and some to group; e.g., "Pseudo-
monas-like," "pseudomonad," and Alcaligenes-like."

     Orange colonies that constituted a majority of the popula-
tion on 12/5/77 (Table 8) and a large proportion of the popula-
tion on 1/6/78 (Table 12), were composed of Flavobacterium sp.
(red colonies)  and Enterobacter agglomerans (yellow colonies).
The orange colonies were streaked repeatedly to separate the  co-
habitants in order to identify them.  The values given for popu-
lation size of each of these two organisms is the same as the
number of orange colonies that they originally formed.  Entero-
bacter agglomerans also appeared by itself.

     Results of one test for regeneration of chlorine resistant
bacteria are given in Tables 17, 18, and 19.  Table 17 shows
residual total and free chlorine concentrations after a one-hour
contact time, two days aging, and six days aging of rechlori-
nated column effluent.  Table 18 is a descriptive list of colo-
nies isolated from unchlorinated and chlorinated column effluent
samples.  Table 19 shows the distribution of identified orga-
nisms .
                              322

-------
                           DISCUSSION


m   0*;9anisms  found in the effluent from GAG  columns  (Tables 6,
10, and 12) are those normal to raw water sources  and  finished
water degraded  by standing in distribution systems as  described
by Geldreich  (1973).  The bacteria survived the treatment
plant process and colonized the
initially bacteria-free activated carbon granules.  Their numbers
in raw water  and in  treated drinking water are often too few to
detect when relatively small volumes (100-200  ml)  of water are
examined by membrane filter techniques  (Tables 3,  6, 7, and 12),
but they multiply on the carbon granules in the columns and many
then slough off into the water stream.   Some of the organisms
isolated may  have health significance because  of kind  and/or
number  (Geldreich 1973,  Wallis 1974).   Pseudomonas aeruginosa,
a possible health threat (Hoadley 1977),  was present in numbers
too few to be detected in raw and finished water,  but  multiplied
to give counts  of 25 per 100 ml of column effluent on  11/21/77
(Table 6).  Whether  this concentration  constitutes an  infectious
dose to people  would depend on the circumstances of exposure.
Other Pseudomonas species that may have clinical significance
(von Graevenitz and  Grehn 1977)   are present in greater numbers
(120/100 ml sample on 11/21/77).   Acinetobacter, Alcaligenes,
Moraxella, Flavobacterium and Enterobacter,  which  constituted a
great part of the population in column  effluent on 11/21/77,
with more than  one million colonies per 100  ml of  sample, were
present in raw  and finished water in small numbers when 100-200
ml samples were filtered (Table 14).

     Population size in each column and in raw water is not
static.  Although the total colony counts obtained from column
effluents increased  within 19 days after the columns were put
into use (11/1/77 to 11/21/77),  counts  in those columns de-
creased in the  following 14 days (Table 13).   Klotz et al.(1976)
reported a rapid increase in bacterial  populations followed by a
decline to a  lower and stable level in  carbon  beds they studied.
They suggested  that  the decrease in bacterial  numbers  may have
been due to an  increase in numbers of bacteria that do not con-
tribute to the  plate count.  This may be more  a reflection of
culture technique that favor some portion of the population than
an actual shift in proportion of the population held by any one
species in the  population.

     The decrease in bacterial numbers  (Table  13)  from 11/21/77
to 12/5/77 occurred  when sampling was done four days after back-

                               323

-------
flushing of the columns.  Backflushing during this period was
routinely done twice a week.  As time passed, backflushing be-
came impossible because of formation of bacterial slime in the
carbon columns.  The columns were last backflushed on 12/8/77.
Increased numbers of bacteria were isolated from samples col-
lected on 1/6/78 when backflushing had not been done for 29 days
prior to sampling.  It has been assumed that backflushing only
removed surficial deposits of calcium carbonate; its effect,
if any, on the resident bacterial culture is unknown.

     The populations sampled on different dates were not com-
posed of the same organisms (Table 14).  There was an apparent
increase in population size with increasing length of carbon bed,
but the organisms isolated from samples at different points along
the length of the bed were not always the same (Table 15}.
Klotz et al. (1976)  stated that changes in the bacterial popula-
tion composition of the raw water influences the character of
the bacterial populations in the carbon beds they studied.  The
bacterial composition of the raw water was not always the same
as that of the GAC columns in this study (Table 15).  In most
cases the bacterial population composition in the GAC columns
did change with time and with the bacterial composition of the
raw water.  Flavobacterium, Staphylococcus, Moraxella, Alcali-
genes, Citrobacter,  Klebsiella, Pseudomonas sp. were found in
the effluents from the columns when they appeared in the raw
water.  Pseudomonas aeruginosa was found in raw water once on
11/2/77 and was always found in column effluent.  Erwinia was
found in column effluents, but not in raw water.  Aeromonas was
found in raw water,  but not in column effluents.  Enterobacter
agglomerans was found in column effluents when it was not found
in raw water on 12/5/77, but on 1/6/78 it was found in raw water
as well as in column effluent.  Enterobacter cloacae was found
in raw water on 12/5/77 and on 1/6/78; it was found in column
effluent on 12/5/77, but noton 1/6/78.  Acinetobacter was found
in raw water and in column effluents on 11/21/77.It was not
found at all on 12/5/77.  On 1/6/78 Acinetobacter was not found
in raw water, but was present in column effluents.  Raw water
bacteria apparently do affect the composition of populations in
the columns, and these resident bacteria apparently determine
which of the incoming organisms can colonize and coexist with
them.  In most cases the bacteria count in raw water was smaller
than in column effluent.  Finished water had detectable bacteria
only on 11/2/77 and 12/5/77 (Table 15).

     The population size of Acinetobacter, Moraxella, Pseudo-
monas, Alcaligenes,  Enterobacter agglomerans, and Flavobacterium
increased with increasing bed length  (from Column 1 to Column 4).
This suggests that increasing distance from the point of chlori-
nation may affect the numbers and kinds of bacteria able to
colonize the carbon bed.  Table 16 gives residual chlorine values
in parts-per-million of water passed through the columns for the
period of this study.  Dates on which samples were taken for

                              324

-------
bacterial analysis  are shown inserted in Table 16.   The  finished
water entering  the  first column had from 0.2  to 3.3  ppm  free
chlorine, which was taken up by the first column.  Combined
chlorine, however,  passed through the first column in consistent
quantities.   The amount of combined chlorine  entering the  first
column ranged from  0.5 to 3.2 ppm.  Effluent  from the first col-
umn, after the  initial 10 days of negligible  quantities  (0.05
ppm), had from  0.15 to 0.20 ppm combined chlorine.   The  column
retained from 0.3 to 3.0 ppm total chlorine,  which may account
for the lower bacteria counts in the effluent from Column  1.
This suggests that  combined chlorine may be used to  control bac-
terial growth in carbon columns.  Column 1 on occasion did have
greater numbers of  several bacteria species than subsequent col-
umns; these  were Staphylococcus on 12/5/77, Flavobacterium on
1/6/78, aud  Erwinia on 1/6/78 (Table 15).

     The microorganisms exhibit succession of species as do most
dynamic plant communities.  The conditions supporting this suc-
cession have not yet been studied.  It probably depends  on over-
growth and death of a pioneer species, which  supplies necessary
metabolites  for the succeeding species.

     Only Enterobacter agglomerans was found  in chlorinated col-
umn effluent (Table 19).  Column effluent chlorinated to 3 ppm
had its population  of Enterobacter agglomerans greatly reduced
within one hour. These regenerated within two days  (300/100 ml
sample), as  residual chlorine was depleted within 24 hours
 (Table 17).   Column effluent with 10 ppm chlorine added  had 4.5
ppm free and 6.3 ppm total chlorine remaining at the end of six
days  (Table  17). Enterobacter agglomerans was found in  this
highly chlorinated  effluent at the end of six days;  however, only
three were isolated from 100 ml of sample. The rechlorination
experiment is being repeated to determine if  the isolated  colo-
nies were indeed survivors or if they were merely contaminants.
Additionally, a survival experiment is being  done to determine
if selection and adaptation is taking place in this  species.  In
this experiment chlorinated finished water is seeded with  orga-
nisms to give a concentration of about 10,000/ml.  The suspension
is held at room temperature (25°C) for various periods of  time
and filtered to isolate and enumerate the surviving  bacteria.

     Enterobacter aqglomerans was not found in finished  water
durini this  study even though 200 ml volumes  (twice  the  volume
suggested by Standard Methods) were passed through membrane
filters.
                                325

-------
                           REFERENCES
Allen, M. J., R. H. Taylor, and E. E. Geldreich.   1977.   The im-
     pact of excessive bacterial populations on coliform method-
     ology.  Microbiological Treatment Branch, Water  Supply Re-
     search Division, MERL, USEPA, Cincinnati, Ohio.   In Press.

APHA, AWWA, WPCF.  1975.  Standard Methods for the examination of
     water and wastewater, 14th edition.  Am. Publ. Hlth.Assoc,,
     Washington, DC.  1193 pp.

Sergey's Manual of Determinative Bacteriology, 8th edition.
     1974.  Williams and Wilkins, Baltimore.  1246 pp.
                      ••\

Eberhardt, M.  1976.  Experience with the use of biologically
     effective activated carbon.  Pages 331-347.  IN H. Sontheijner,
     ed.  Translation of reports on special problems  of  water
     technology, MERL, USEPA, Cincinnati, Ohio.  453  pp.

Fiore, J. V. and R. A. Babineau.  1977.  Effect of an activated
     carbon filter on the microbial quality of water.  Appl.  and
     Env. Microbiol.  34 (5):541-546.

Geldreich, E. E. 1973.  Is the total count necessary? Presented
     at the AWWA First Water Quality Technology Conference.
     December 1973, Cincinnati, Ohio.  11 pp.

Hoadley, A. W.  1977.  Pseudomonas aeruginosa in surface waters.
     Pages 31-57.  IN Viola Mae Young, ed.  Pseudomonas aeruginosa;
     Ecological aspects and patient colonization.  Raven Press,
     NY.  137 pp.

Klotz, M., P. Werner, and R. Schweisfurth.  1976.  Investigations
     concerning the microbiology of activated carbon  filters.
     Pages 312-330.  IN H. Sontheimer, ed.  Translation  of  reports
     on special problems of water technology, MERL, USEPA,
     Cincinnati, Ohio.  453 pp.

Peterson, N. and M. Favero.  1975.  Significance of  Gram-negative
     bacteria in water supplies.  Presented at the Third AWWA
     Water Quality Conference.  December 1975.  Cincinnati,  Ohio
     10 pp.

van der Kooij, D.  1976.  Some investigations into the presence
     and behavior of bacteria in activated carbon  filters.   Pages

                               326

-------
     348-354.  IN H. Sontheiraer, ed.  Translation of reports on
     special problems of water  technology,- MERL, USEPA, Cincin-
     nati, Ohio.  453 pp.

von Graeveniz, A. and M. Grehn.  1977.  Clinical microbiology
     of unusual Pseudomonas  species.  Pages  50-134.  IN ASM
     Committee on Continuing Education.  Unusual organisms of
     clinical significance.   ASM,  Washington, DC.  184 pp.

Wallis, D.,  C. H. Stagg, and J.  L. Melnick.  1974.  The hazards
     of incorporating charcoal  filters  into  domestic water
     systems.  Water Res.   8:111-113.

Wood,  P.,  L. Kaplan, J.  A.  Gervers, D.  Waddell, andD.F. Jackson.
     IN preparation. Removing  potential organic carcinogens and
     precursors  from drinking water.  Report to EPA, Grant Pro-
     ject R804521-01-2,  Florida International University, Miami,
     Florida.
                                327

-------
                                           PRESTON PLANT
to
00
WELLS "
D CLEAR
1 WELL

WATER FINISHED
WATER



T
30"



1 1
_L

c
o
L.
1


G
A
C













~D











(
L

•»
^
0
L.
2


(
i


•*
j
v
c
t
1 23
Sample Point 1 I 1 	 _










•••••MM
MBMBM










1




c
o
L.
3




G
A
C
1
C
*

Raw Finished 1st. Col. 2nd. 3z
Effluent
Sampled:
11/2/77 Raw and Finished Water (Columns installed and flow begun)
11/21/77 All Points
12/5/77
1/6/78



















~1











c
0
L.
4


G
A
C
1
6

•d. 4th.






Figure 1 . Bench scale column adsorption unit (Experimental Design No. 4) .
                    Sampling points and date of sampling at  each point are indicated,

-------
         WATER SAMPLE (RAW,  FINISHED, AND COLUMN EFFLUENTS)
           MEMBRANE FILTER (1-, 10-, 100-, 200-ml volumes)

                                 t
                       PRIMARY MEDIA (Table 1)

—•f	—•t	v	»	,	r
  TGE        BHI              ENDO'S          DBS      Ps       Cz

Colony #1    Colony 1,  etc.   Colony 1, etc.

Colony #2
              1)   Differential colony counts made on each
                  primary culture.


              2)   Each colony type (numbered) picked for
                  isolation and identification.
Colony #23
Diagnostic Tests and Media
  i
Colony Identification:  Name applied to each colony type isolated
                       from each medium;  duplicates combined.
             Figure  2 .  Flow diagram of  procedure
                         used to  isolate  and identify
                         organisms from raw, finished,
                         and filtered water.
                                 329

-------
TABLE 1 .  PRIMARY MEDIA AND  DIAGNOSTIC TESTS  USED TO ISOLATE AND
           IDENTIFY BACTERIA  IN RAW,  FINISHED,  AND FILTERED  WATER
Primary media*

        Endo's broth (Endos)                     Pseudosal agar  (Ps)

        Desoxycholate lactose                    Brain heart  infusion
           agar (DES)                               broth (BHI)

        Tryptone glucose yeast                   Czapeck agar (Cz)
           extract agar (TGE)


Diagnostic tests

        Roche Oxiferm system                     Hanging drop/motility

        Gram stain                              Motility agar

        Triple Sugar Iron agar                   Flagella stain

        Oxidase                                 Casein digestion

        Nitrate reduction                       Catalase

        Roche Enterotube system                  Starch digestion
*TGE and Cz gave highest counts; other media used with poor results were
 Simmons citrate agar,  potato dextrose agar, nutrient agar, PA agar, and
 milk agar.


TABLE  2   BACTERIAL POPULATION COUNTS  (COLONIES/100ml  SAMPLE)
           OBTAINED FROM RAW AND FINISHED WATER SAMPLES COLLECTED
           ON  11/2/77

Name
Ac inetobac ter
Moraxella
Pseudomonas aeruginosa
Pseudomonas sp«
(other than P. aeruginosa)
Penicillium
Raw
120
4
2
22
0
Finished
12
2
3
0
85
                                 330

-------
     TABLE  3 .   BACTERIAL POPULATION COUNTS (COLONIES/lOOml  SAMPLE) OBTAINED FROM SAMPLES
                OF RAW AND FINISHED WATER  AND COLUMN EFFLUENTS COLLECTED ON 11/21/77
U)
to

Medium
TGE
Endo's coliform
non-coliform
Ps
Raw
96
0
41
27
Finished
0
0
0
0
Col. 1 Col. 2
6*
972 10
0 0
27 100
0 520
Col. 3
6*
10
0
120
25
Col. 4
6*
10
0
131
171
      Cz
None was evident in 24,  48, 72 hours, but when growth developed in four to
five days, the plates were overgrown with minute yellow colonies.  Finished
water had none.
      DES
58
92
208
420
240
      BHI
62
                                                          120
        480
           640
          483
      *estimated by counting 10cm

-------
    TABLE 4 .   BACTERIAL  POPULATION  (COLONIES/lOOml SAMPLE) OF RAW

               AND FINISHED WATER* AND COLUMN  EFFLUENTS  COLLECTED

               ON 11/21/77,
Colony type
Endo ' s
2mm, dark red
0.7mm, dark red
<0.3mm, It. red

2mm, It. red
DES
3mm, gray
2mm, tan
2mm, black
10 >10 >10
1 2

12 2
5 25
10 11
7 520 19 160
5 6

*No growth was obtained from 200ml volumes of finished water on any medium.

 Blanks opposite colony description indicate no growth.
                         2
**Estimated by counting 10cm
                                 332

-------
TABLE 5 .  GROUPING  OF  COLONIES WITH  SIMILAR BIOCHEMICAL
          REACTIONS, ISOLATED FROM SAMPLES TAKEN 11/21/77

Group number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Identity
Acinetobacter
Aeromonas
Alcaligenes
Citrobacter
Enterobacter
Klebsiella
Moraxella-like
Pleisomonas
Proteus
Pseudomonas
aeruginosa
Pseudomonas sp-
Pseudomonas
maltophiTa
Pseudomonas
stutzeri
Pseudomonas-like
arouo 5E-1
Includes colony number
2, 16,
If 2
13
11
29, 32
10
14, 17
22, 24
23
0
4, 5,
3, 6,
28
7, 8
15
27
, 19, 20, 21,
, 24A, 26, 31
12, 18, 25, 30
9
                             333

-------
TABLE 6.   DISTRIBUTION OF IDENTIFIED BACTERIA (COLONIES/10Oml
          SAMPLE)  ISOLATED FROM WATER SAMPLES TAKEN ON 11/21/77

Group name
Acinetobacter
Aeromonas
Alcaligenes
Citrobacter
Enterobacter
agglomerans
Klebsiella
Moraxella
Pseudomonas
aeruginosa
Pseudomonas spp.
(other than
Ps. aeruginosa)


Raw
19
37
54
54
25
4
67
0
2

Finished
0
0
0
0
0
0
0
0
0
Sample
Col. 1 Col. 2
2 1000
0 0
0 80
0 80
0 0
0 2
970 >106
0 0
27 100

Col. 3 Col. 4
670 630
0 0
0 2
0 2
6 0
0 0
>io6 >io6
1 25
120 110

TABLE 7 BACTERIAL
OBTAINED
EFFLUENT
POPULATION COUNTS (COLONIES/lOOml SAMPLE)
FROM RAW AND FINISHED WATER AND COLUMN
SAMPLES COLLECTED ON 12/5/77

Medium
TGE
Endo's coliforms
Endo's noncoliform
Ps
Cz

Raw
179
0
10
3
40

Finished
5
0
0
0
0
Sample
Col. 1 Col. 2
15,180 20,000
0 0
0 0
0 0
2,100 8,160

Col. 3 Col. 4
25,000 25,000
0 0
0 8
4 17
6,950 8,880
                              334

-------
     TABLE 8
INITIAL DESCRIPTION  OF COLONIES COUNTED AND
SELECTED FOR ISOLATION AND  IDENTIFICATION
FROM SAMPLES TAKEN 12/5/77
Colony no.
1
2
3
4
5
6
7
7a
8
8a
9
10
11
12
a,b
a,b,c

5b
a,b,c,d,e
a,b,c
(a) (b)
a,b
(a) (b)
a,c
a,b,c
a,b,
,13,14
Source Primary
sample medium
1
1
5
5
1
1
1
1
5
6
1
6
6

TGE
TGE
TGE
TGE
Endos
Endos
Ps
Ps
Ps
Ps
Cz
TGE
TGE

Colony
description
2mm, yellow
1mm, white
<. 3mm, yellow
5mm, light yellow
3mm red, mucoid
5mm, light red,
wrinkled
3mm, cream-yellow
3mm , creamy ,
mucoid, stinks
5mm, green
3mm, green son. ,
fruity odor
3-5mm, white
3-5mm, (greenish,
yellow)
<.3mm yellow

-— "•- 	 '•• '• • -— 	
Gram stains
G-
mixed sizes
G+,staph
G- mixed?
G-
G-
G-
G-
G-
G-
G-
G-
G-

mixed
mixed
mixed
mixed
mixed
mixed
mixed
mixed

sizes
sizes
sizes
sizes
sizes
sizes
sizes
sizes
mixed sizes &




&
&
&
&
&
&
&
&

shapes
shapes
shapes
shapes
shapes
shapes
shapes
shapes
shapes


Tiny Orange=
  Yellow +
    Red
       Cz     <.3mm orange,
              developed 6 days
              after plating
mixed sizes & shapes
 *Colony  numbers followed by letters were suspected of being composed of
  more than one species when the Gram stain was examined.   They were
  streaked on TGE and SMA to effect separation and subcultures were made
  of several isolated colonies.
                                   335

-------
  TABLE 9.   POPULATION DISTRIBUTION (COLONIES/lOOml SAMPLE)* AND  COLONIES SELECTED
             FOR ISOLATION AND  IDENTIFICATION FROM SAMPLES TAKEN 12/5/77
u>
U>
CT>
Colony
old no.
TGE 1
v,
2
3
4
Endo's
5
6
Ps 7
8
C£ 9
10
11
Tiny Orange ~")
(6 days later!
yellow f
red J
Includes
new no.
1
2
4
(5al
bbj
H
L6bJ
f7(SMA)l
{8 (7TSM)\
(9<7A) j
10(asbf|
il J
12
13
14
Sample
source
Raw
Raw
Col. 3
Col. 3
Raw
Col. 4
Col. 3
Col. 3
Raw
Col. 4
Col. 4
Col. 4
Colony
description
2mm, yellow
3mm, white
3mm, yellow-
green
3mm, yellow-
green
•3mm, It. red
2mm,med.red
2mm, white
3mm, green
2mm , wh ite
3-5mm, green
<0.3mm,
yellow
< 1mm, orange
Identity
Group, 2K-1
Pseu.-like
Staph . (sapro-
phyticus)
Enterobacter
agglomerans
Ps . aeruginosa
Ps .putida [
Ent. cloacae!
Serratia -'
marcescens
pilcaligenes~^
\ faecalis /
/Enterobacter 1
\ cloacae ?
Enterobacter 1
cloacae J
"PseudomonasJ
, aeruginosa \
Ps.cepacia J
Enterbacter
cloacae
Ps . aeruginosa
Enterobacter
agglomerans
Enterobacter ~"J
agglomerans 1
Flavobacteriure/
(aquatile) _J
Raw
14
144
4
6
40
Fin.
2
3
1
Col. 1
180
15,000
3
8
2,140
Col. 2
30
20,000
57
8,100
Col. 3
40
25,000
1
/>
3
1
250
6,000
Col. 4
60
25,000
3
8
17
3
380
8,500
*Blank spaces indicate none was isolated

-------
TABLE 10.   BACTERIAL  POPULATION DISTRIBUTION (COLONIES/lOOml)
             SAMPLE OF  ALL ORGANISMS  ISOLATED  FROM RAW AND
             FINISHED WATER AND COLUMN EFFLUENTS COLLECTED ON
             JLfL/ J/ / /
                                      '	"	      ' "    •	        'I   	   „„...,.
Group name/specie*         Raw   Finished   Col. 1  Col.  2  Col. 3   Col. 4



Alcaligenes faecalis                                           3

Enterobacter
  agglomerans                               15,000  20,000  25,000   25,000

Enterobacter
  cloacae                  40                   133

Pseudomonas
  aeruginosa                                                   1      17

Pseudomonas
  cepacia                                                      1      17

P_. putida                   4

Pseudomonas-like,
  Group 2K-1               14       2

Serratia marcescens         6                                          8

Staphylococcus
  (saprophyticus)          144       3         180      30      30      60


Flavobacterium
  (aguatile) ~             20               2,140   8,100  6,700     8,500
*Species names given are accurate within the  limits of the Roche
 system and additional media and tests listed in Table 1.
 Blank spaces in columns opposite organism name indicates  none was isolated.
                                  337

-------
U!
CO
00
          TABLE 11.   INITIAL DESCRIPTION OF  COLONIES COUNTED AND SELECTED FOR
                      ISOLATION AND IDENTIFICATION ON 1/6/78

Colony Colony Sample Primary
number description source med.(vol.)
1 3mm, light red Raw water Endos(lO)
2 3mm, rose beige " TGE (100)
3 1mm, yellow " TGE (10)
4 1mm, white " TGE (10)
5 4mm, white " Cz (100)
6 1mm, yellow Column 1 TGE (10)
7 
-------
TABLE 12.   POPULATION DISTRIBUTION (COLONIES/100ml  SAMPLE)* OP
             rmrmif ™J™  ISOLATED FROM RAW AND FINISHED WATER
             COLUMN  EFFLUENTS TAKEN ON 1/6/78
Organism isolated
                           Raw   Finished  Col.  1  Col. 2  Col. 3    Col. 4
Enterobacter agglomerans    780
Enterobacter cloacae        100
Ac inetobacter
Erwina (Stewartii)
Pseudomonas aeruginosa
Group, 5A-2,
  Pseudomonas-like
Pseudomonas (syringae)       20
Flavobacterium
  (aquatile)                90
 1,730     910    610     8,000

        2,800  80,000   850,000
60,000           2,040
                   i
                            2

 3,000  70,800


 1,730     910    610     8,000
 *Blank spaces opposite organism name indicates none was isolated.
TABLE 13    BACTERIAL  POPULATION  (COLONIES/lOOml SAMPLE)  GROWN
             ON MOST FAVORABLE MEDIUM IN EACH COLUMN EFFLUENT,
             RAW,  AND FINISHED WATER (TOTALS  OF TABLES  2, 6, 12)
Sample date
11/2/77


11/21/77
12/5/77
l/6/78b
Raw
148


96
179
990
Finished Col. 1
17 (+85
moulds) a

0 999
5 15,180
0 66,460
Col. 2 Col. 3 Col. 4

a a a
6 6 6
>10 >10 >10
20,000 25,000 25,000
75,820 83,260 866,000
 aNo growth was obtained from 200ml of water taken from 500ml water shaken
 for one hour with a 250ml volume of dry unused carbon granules.
 bBackflushed 30 days prior to sampling.  Other samples were taken four days
 after backflushing.
                                   339

-------
TABLE 14 .  BACTERIAL POPULATION (COLONIES/lOOral SAMPLE)  OF  THE
           SAMPLE POINT GIVING THE HIGHEST VALUE FOR THE LISTED
           SPECIES ON THE DIFFERENT SAMPLE DATES COMPARED WITH
           POPULATION OF INCOMING RAW WATER*
Date sampled
Organism 11/2/77
Acinetobacter (120) 12
Moraxella (4) 2
Ps. aeruginosa (2) 3
Pseudomonas sp. (22) 0
Penicillium (0) 85
Aeromonas
Alcaligenes
Citrobacter
Klebsiella
Enterobacter
agglomerans
Enterobacter
cloacae
Pseudomonas-like
Serratia xnarcescens
Staphylococcus
Flavobacterium
Erwinia

11/21/77 12/5/77 1/6/78
(90) 1,000 (0) 850,000
(67) >106
(0) 25 (0) 17 (0) 2
(2) 120 (4) 0 (20) 0
(37) 0
(54) 80 (0) 3
(54) 80
(4) 2
(25) 6 (0) 25,000 (780) 8,000
(40) 3 (100) 0
(14) 3 (0) 70,000
(6) 8
(144) 180
(20) 8,500 (90) 8,000
(0) 60,000
 *Raw water values given in parentheses
                              340

-------
TABLE 15.  DISTRIBUTION OF ORGANISMS ISOLATED FROM
           COLUMN EFFLUENTS

Organism

Acinetobacter




Moraxella




Pseudomonas

aeruginosa


Pseudomonas
species



Aeromonas



Alcaligenes




Citrobacter



S
m
P
1
e
R
1
2
3
4
R
1
2
3
4
R
1

3
4
R
1
2
•j
j
4
p
*x
1
2
4
R
1

3
4
R
1
2
3
4
Sample date
11/2/77 11/21/77 12/5/77
120{12)* 19
2
1000
670
630
4(2)* 67
1 970
>io6
>io6
>io6
2(3)* 0
0
o
™* ^/
1
25
22(0)* 2
27
100
120
110
0 37
fl
w \J
0
0
0 54
0
80
0
-)
^ £t
0 54
0
80
^^ ^f\f
o
2
^^ &»
0
0
0
0
0
0
0
0
0
0
0
0
0
1
17
4
0
0
1
17
0
0
0
0
0
0
0
0
3
0

0
0
0
0
0

1/6/78
0
0
2800
80,000
850,000
0
0

0
0
0
0

0
2
20
0
0
0
0
0
0
0
0
0
0
0
0
0
0

0
0
0
0
0
(continued)
                          341

-------
TABLE 15.   (continued)


Organism



Klebsiella




Enterobacter
agglomerans



Enterobacter
cloacae



Pseudomonas-like




Serratia
marcescens



Staphylococcus




Plavobacterium



S
a
m
P
1
e
R
1
2
3
4
R
1
2
3
4
R
1
2
3
4
R
1
2
3
4
R
1
2
3
4
R
1
2
3
4
R
1
2
3
4
Sample date


11/2/77
0
-
-
-
-
0
-
-
-
-
0
-
-
-
-
0
-
-
-
-
0
-
-
-
-
0
-
-
-
-
0
-
-
-
-


11/21/77
4
0
2
0
0
25
0
0
6
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


12/5/77
0
0
0
0
0
0
15,000
20,000
25,000
25,000
40
1
3
3
0
14(2)*
0
0
0
0
6
0
0
0
8
144(3)*
180 t
30
40
60
20
2140
8100
6700
8500


1/6/78
0
0
0
0
0
780
1730t
910
610
8000
100
0
0
0
0
0
3000
70,800
0
0
0
0
0
0
0
0
0
0
0
0
90
1730t
910
610
8000
                                   (continued)
         342

-------
                       TABLE  15.   (continued)
s
m
Organism
	 p
1
e
j^
Erwinia stewarti ,
1 	 J-
2
3
4



11/2/77
0
0
-
—
Sample


11/21/77
0
0
0
0
0
date


12/5/77
0
0
0
0
0



1/6/78
0
60,000t
0
2040
0

*Values in parentheses are for finished water, all other values for
 finished water were less than 1.   Zero values indicate less than 1,
 - indicates not done, R = raw water,  1 through 4 are column numbers.

tColumn 1 effluent had higher counts than following columns.
                                  343

-------
TABLE 16.  CONCENTRATIONS  OF  RESIDUAL FREE AND TOTAL CHLORINE IN FINISHED
           WATER AND COLUMN EFFLUENT FOR THE STUDY PERIOD

11
Finish
Date
ll/l/77a
ll/2/77b
11/3
11/4
11/73
11/8
n/ioa
11/11
11/14
11/15
ll/17a
11/18
11/25
11/29
Free
2.25


1.60

1.80

1.60
1.20

2.00
1.35
1.75
Total
3.25


2.30

2.60

2.40
2.10

2.65
1.85
2.40
19
Col. 1
Free
0


0

0

0
0

0
0
0
Total
0.05


0.05

0.05

0.10
0.10

0.10
0.15
0.15
21 23 25
Col. 2 Col. 3 Col. 4
Free
0


0

0

0
0

0
0
0
Total Free
0.05 0


0.05 0

0.05 0

0.05 0
0.05 0

0.05 0
0.05 0
0.05 0
Total Free
0.05 0


0

0.05 0

0.05 0
0.05 0

0.05 0
0.05 0
0.05 0
Total
0.05


0.05

0.05

0.05
0.05

0.05
0.05
0.05
17
M. Lakes
Free Total


0.10 0.55



0.05 0.50

0.05 0.40



                                                                        (continued)

-------
                             TABLE 16.   (continued)

11
Finish
Date Free Total
12/1 a
12/2 2.00 2.65
12/5b 1.90 2.50
12/8a
12/9 1.40 2.10
12/13 1.95 2.40
12/15
12/16 1.70 2.10
12/20 3.30 6.55
12/22
12/23 2.40 2.90
12/27/77 0.20 2.70
J^
I/O
19 21 23 25 17
Col. 1 Col. 2 Col. 3 Col. 4 M. Lakes
Free Total Free Total Free Total Free Total Free Total
0.05 0.45
0 0.15 0 0.05 0 0.05 0 0.05
0 0.20 0 0.05 0 0.05 0 0.05
0.10 0.50
0 0.20 0 0.05 0 0.05 0 0.05
0 0.15 0 0.05 0 0.05 0 0.05
.05 .40
0 0.15 0 0.05 0 .05 0 .05
0 0.15 0 .05 0 .05 0 .05
.05 .50
0 .20 0 .05 0 .05 0 .05
0 0.15 0 0.05 0 0.05 0 .05
sackflushed .columns
Samples taken for bacterial analysis

-------
TABLE  17 .  RESIDUAL FREE AND TOTAL CHLORINE CONCENTRATIONS
            IN  GAG COLUMN EFFLUENT AFTER  AGING
Sample
Rechlorination
concentration
                Age of  sample
                (Contact time)
Chlorine,  ppm
Free    Total
Col. 4
  effluent

Col. 4
  effluent

Col. 4
  effluent
 3ppm


 3ppm


lOppm
                     24 hours
                     48 hours
                      6 days
0.05
0.05
4.5
0.2
0.3
6.3
                                346

-------
TABLE 18.   INITIAL  COLONY  DESCRIPTION AND IDENTITY OF BACTERIAL
             ISOLATES FROM RECHLORINATED, AGED  GAG COLUMN  EFFLUENT
Colony
Medium number
Endo ' s

Endo ' s
Endo ' s
Endo ' s
Endo ' s

TGE


TGE
TGE

TGE
TGE
TGE
TGE
Cz
\f1U


Cz
Cz
El

E2
E3
E4
4

1


2
3

5
6
10
11
7
/


8
9
Sample*
4+0 (0)

4+0 (0)
4+0 (0)
4+0 (0)
4+2 (3)

4+0 (0)
4+3 (3)
4+6 (10)
duplicate
4+2 (3)
4+6 (10)
4+0 (3)
4+0 (0)
4+0 (0)
4+0 (0)
4+0 (0)
-» • ^^ \ ** /
4+0 (3)
4+2 (3)
4+6 (10)
4+0 (3)
4+0 (0)
— 	 — 	 — — 	 	
Colony Colonies/
description 100ml sample
3mm, white,
moist
2mm, red, domed
2mm, white/halo
<0. 3mm, red
2mm, red


-------
TABLE 19 .  BACTERIAL POPULATION  (COLONIES/100 ml  SAMPLE)  OF GAC
            COLUMN EFFLUENT  CHLORINATED AND AGED TO SIMULATE
            DISTRIBUTION SYSTEM CONDITIONS

Sample

Organism
4+Oa
(0)
4+Ob
(3ppm)
4+2C
(3ppm)
4+6d
(lOppm)
Acinetobacter
   Clwoffi)                 600

Enterobacter
agglomerans
Moraxella
Pseudomonas-like

59,000 4 300
48,000 0 0
500
3
0
0

Effluent from Column
collection .

4, no chlorine added, analyzed one hour after



 collection.

 Effluent  from Column 4,  3ppm chlorine added, aged 2 days before  analysis.

 Effluent  from Column 4,  lOppm chlorine added, aged 6 days before analysis.

eMore than one-half (55,800) of the colonies identified as Enterobacter
 agglomerans were orange.  Previously orange colonies yielded Flavobacterium
 sp.  as well, but Flavobacterium was not isolated from orange colonies that
 developed from these samples.
                                  348

-------
        CHLORINATION OF GRANULATED ACTIVATED CARBON  (GAG)
                COLUMN EFFLUENT TO CONTROL BACTERIA
                              (Part II)

                                 by


                         Frances Parsons
             Drinking Water Quality  Research Center
                 Florida International University
                          Tamiami Campus
                          Miami, Florida
                             ABSTRACT
     Granulated activated carbon (GAG)  columns  used in water
treatment produce effluents with bacteria counts  up to 10,000/ml.
A cursory examination was made of the effect of chlorination of
effluents on  these bacteria.  Samples of GAG column effluent
chlorinated to  3 ppm and 10 ppm were held at 25°C and plated at
two-day intervals for six days to test for survival and regrowth
of bacteria.  At 3 ppm,  chlorine killed most of the bacteria
(none recovered initially)  and controlled regrowth to less than
500/ml for up to five days.  When the 3 ppm chlorine initially
added to GAC  column effluent was depleted during  aging for six
days, bacterial regrowth reached 36,000/ml or greater than half
that of six-day-old unchlorinated GAC column effluent, which was
62,000/ml.  When GAC column effluent had 10 ppm chlorine added,
residual free chlorine was 3 ppm throughout the aging period
(1.0 ppm in one sample on the sixth day), and no  bacteria were
recovered.

     Finished water to which 8,000 bacteria per ml and 10 ppm
chlorine was  added,  which retained 3 ppm residual free chlorine,
had three colonies per ml after six days.  Finished water to
which 8300 bacteria per  ml and 10 ppm chlorine  was added, which
had less than 0.4 ppm residual free chlorine, had 83,000 colonies
per ml after  five days aging.
                               349

-------
                           CONCLUSION
     Chlorination of the effluent from granulated activated
carbon  (GAC) columns apparently kills bacteria that grow on  the
carbon granules and slough off into the effluent, but the  initial
dose of chlorine must be adequate to combine with the bacteria
and leave sufficient free chlorine to prevent regrowth.  The
concentration of chlorine necessary would vary with the bacterial
biomass and chlorine demand due to all constituents of the water.

     This study was of a cursory nature and was only intended to
suggest a more complete study.  Shorter sampling intervals
(daily), for a period of time longer than six days  (end point
determination), with more than these two concentrations  (espe-
cially less than 3 ppm free chlorine) of several disinfectants
(chlorine, chloramines, chlorine dioxide, ozone, ferrates)
should be examined.  Certainly the minimum level of chlorine
needed and the time that it is effective for several bacterial
population sizes should be determined.  All of these factors;
i.e., dose size, contact time, regrowth rate and size and compo-
sition of the bacterial population should be studied and compared
with parallel determinations of the bacteriology of the dis-
tribution system.
                               350

-------
                           INTRODUCTION
     Granulated  activated carbon (GAG)  columns  used  to remove
organic substances  from water are colonized by  bacteria that
survive water  treatment plant processes.   Large populations of
these bacteria develop on the carbon granules and  slough off into
the water  flowing through them.   Some of  these  bacterial popula-
tions have been  characterized (Parsons 1978) .   In  an attempt to
find a way to  control bacterial  numbers,  GAG column  effluent was
chlorinated  and  analyzed for bacterial survivors and for regrowth

     Finding Enterobacter agglomerans in  highly chlorinated  (10
ppm) GAG column  effluent prompted this study to determine  if the
few colonies isolated represented selection of  a resistant strain
or survived  because of protection offered by cell  aggregation
or slime.
                                351

-------
                      METHODS AND MATERIALS
     The following samples were taken from the water treatment
plant and from the bench scale model GAG system described earlier
 (Wood    et al. 1979) :

Finished water                  referred to as:  Finished
   (containing 3 mg/1 chlorine)

Column 4 effluent               referred to as:  Column  4  (0)
   (with no added chlorine)

Column 4 effluent               referred to as:  Column  4  (3)
   (with 3 mg/1 chlorine added)

Column 4 effluent               referred to as:  Column  4  (10)
   (with 10 mg/1 chlorine added)

Finished water                  referred to as:  Finished Water
   (with 10 mg/1 chlorine added)                  + Bacteria  (10)
   (bacteria were added to this
   sample in the laboratory)

     The samples were taken to the laboratory where bacteria  were
added to the sample of finished water containing 10 mg/1 chlorine
to give a concentration of approximately 10,000 cells per ml.
A culture of Enterobacter agglomerans isolated from chlorinated
Column 4 effluent was used to seed the chlorinated  (10 ppm) fin-
ished water on 2/2/78.  An Acinetobacter isolated from the 2/2/78
samples was used on 2/17/78 when the experiment was repeated.
Chlorine contact time for this sample was approximately  ten
minutes as plating was begun shortly.after addition of the bac-
teria to the water.  The time interval (contact time) for the
other samples on 2/2/78 was approximately three hours after col-
lection.  The remainders of the samples, after initial plating,
were stored at room temperature (25°C) to simulate conditions in
a distribution system, and plated for bacteria at intervals dur-
ing the following six days.  The samples were aged to determine
bacterial regrowth potential of chlorinated GAG column effluent.

     Enumeration and isolation of bacteria from the water samples
were done by the membrane filter technique  (APHA 1976).  Volumes
from 0.01 to 200 ml were passed through Gelman GN6  (0.45 ym pore
size) filter membranes, which were then placed on tryptone glu-
cose extract (TGE) agar in petri dishes.  Volumes less than  100

                               352

-------
ml were diluted with 100 ml sterile,  buffered,  glass-distilled
water to assure uniform distribution  of the organisms on the
surface of  the membrane.  The cultures were incubated at 25°C
and examined daily.   Cumulative colony counts were  made daily for
six days of all different types of colonies observed.  Different
colony types were described, assigned numbers,  and  picked  for
identification.  The API system (Analytab Products, Division of
Ayerst Laboratories, Inc., Plainview, NY), and  supplementary
media and  tests (Parsons 1978) were used in identification.  Gram
positive  cocci were planted in glucose OF medium.   Staphylococci
were  tested for coagulase.

      Residual chlorine was determined at the time of plating
using a  Taylor Basic 2000 test kit (DPD).
                                 353

-------
                            RESULTS
     Table 1 is a summary table of total bacteria counts obtain-
ed from samples taken on 2/2/78.  Concentrations of residual free
chlorine are also given.  The distribution of the different kinds
of bacteria isolated in the samples is shown in Table 2.  Table
3 lists the original colony description, the sample from which
it was isolated, the estimated population size, and the identity
of the bacterial isolates from samples taken 2/2/78.

     Table 4 is a summary table of bacteria counts and free
chlorine concentrations for samples taken 2/17/78.  Table 5
shows" the different kinds of bacteria isolated and the sample
from which they came.  Table 6 is a list of colonies, with their
description, chosen for isolation.  Their final identity is also
given.

     Comparison of large numbers and small numbers precludes
uniform rounding.  Small numbers  (less than 100)  are not
rounded; large ones are.
                              354

-------
                            DISCUSSION


     Three mg/1  added chlorine  was  sufficient to reduce the bac-
teria count  in Column 4(3)   effluent to  the  level found in treat-
ment plant Finished Water when  both were plated shortly after
chlonnation.  Table  1 shows that both samples had 0 colonies per
ml isolated  on 2/2/78.   Table 4 shows that Finished Water had one
colony per ml and  Column 4(3)   effluent  had  none on 2/17/78.
After four days  aging of the samples taken on 2/2/78, Column 4(3)
had 330 colonies per  ml and Finished Water had 22 colonies per ml.
After five days  aging of samples taken on 2/17/78, Column 4(3)-
effluent had 250 colonies per ml and Finished Water had four
colonies per ml.  These numbers have little  health significance
as there is  no standard for noncoliform  bacteria, but do indicate
regrowth potential when chlorine is depleted by initially high
bacteria counts.  The initial bacteria counts for Column 4 efflu-
ent with no  added  chlorine, 4(0), were 1100  colonies per ml on
2/2/78  (Table 1) and  8300 on 2/17/78  (Table  4) .

     On 2/2/78 Column 4 effluent with  3  ppm  chlorine added ini-
tially had less  than  0.4 mg/1 free  chlorine  three hours later
when initial plating  was done.   Finished water, which had 3 ppm
chlorine added by  the treatment plant process, had 0.5 mg/1
(Table 1) .   At four days age, Column  4(3) had ten times more
bacteria than finished water, but the count  was only 330/ml
(Table 1) .   On 2/17/78 when finished water had less than 0.4 mg/1
free chlorine at the  time of initial plating, it had, a higher
bacterial count  (though still a small number:  42/ml) at the end
of three days age  than did  Column 4(3) effluent  (Table 4).
Column 4 effluent  without added chlorine had high counts (83,000/ml
initially that became larger (120,000/ml) with aging.  When the
concentration of chlorine added to  Column  4  effluent was in-
creased to 10 ppm, it effectively killed bacteria; one per ml was
recovered after  a  contact time  of 3 hours, none after six days
aging.  There was  sufficient chlorine  at this-level  (10 ppm) to
sustain a residual of 1 ppm at  the  end of  six  days after sample
preparation  on 2/2/78 (Table 1) .  Only one  colony was recovered
from this sample initially.  Chlorine  residual was  3 ppm at the
end of five  days after sample preparation  on 2/16/78  (Table 4).
No bacteria  were recovered  from this sample.  No  attempt was made
to determine the products  formed by the  action of  chlorine  on
bacteria.

     Bacteria added to finished water (Finished Water  + Bacteria
                               355

-------
(10) to determine if chlorine-resistant bacteria were developing
in GAC columns bathed in chlorinated finished water were effec-
tively controlled when 3 ppm residual free chlorine was present.
Three colonies/ml were recovered after six days  (Table 1).  When
free chlorine was initially depleted in column effluent rechlor-
inated to 10 ppm on 2/17/78, regrowth occurred and reached 83,000
colonies per ml in five days (Table 4).

     Staphylococcus sp. and Acinetobacter sp. grew in Finished
Water + Bacteria (10) although only Enterobacter agglomerans was
added to the sample  (Tables 2,  3), which suggests that these
organisms, survivors of the treatment plant processes, survived
the subsequent rechlorination to 10 ppm because the added Entero-
bacter agglomerans cells (approximately 10,000 cells/ml were
added to the Finished Water) combined with and depleted the avail-
able free chlorine in the sample.  On 2/17/78 only Acinetobacter
sp. was added to the Finished Water + Bacteria (10) sample and
only Acinetobacter sp. was recovered (Tables 5, 6).

     Bacteria in GAC column effluent could deplete chlorine,
which in turn could allow pathogenic survivors of the treatment
process to grow in the distribution system.

     The results do not follow uniform patterns.  Cell aggregates
examined briefly by microscope, make it exceedingly difficult to
obtain ten-fold counts from Log.Q dilutions.

     There seems to be a tendency among water bacteria to form
mixed-culture colonies that appear well isolated on solid medium
and for them to have similar cellular morphologies.  For example,
colonies that first appear as small, white, and entire become
yellow after two or three days.  Staining only discloses Gram-
negative rods of sizes and shapes within the range of individual
variability.  When diagnostic media are inoculated, the results
often indicate a mixed culture.  Upon re-isolation by streaking
from a diagnostic medium and from the original colony, a mixed
culture often results.  Succession of species, as often occurs
in higher plants, is suspected.
                               356

-------
                            REFERENCES
APHA, AWWA, WPCF.   1975.   Standard Methods  for  the examination of
     water and wastewater,  14th edition.  Am. Publ. Hlth. Assoc.,
     Washington,  B.C.   1193 pp.

Wood, P., L.  Kaplan,  J. A.  Gervers, D.  Waddell,  and D. F. Jack-
     son.  1978.   Removing potential organic  carcinogens and
     precursors  from drinking water.  Report  to EPA, Grant
     Project  R804521-01.   Florida International University,
     Miami, Florida.

Parsons,  F.   1978.  Microbial flora of granulated activated car-
     bon  columns used in  water treatment.   IN P. Wood et al.
     Report to  EPA, Grant Project R804521-01.  Florida Interna-
     tional University, Miami, Florida.
                                357

-------
U)
en
oo
       TABLE 1.  BACTERIAL POPULATIONS  (COLONIES  ISOLATED/ml SAMPLE)  OBTAINED  IN SAMPLES

                  OF FINISHED WATER AND  GRANULATED ACTIVATED  CARBON  (GAG)  COLUMN EFFLUENT

                  TREATED IN  SEVERAL WAYS.  SAMPLES TAKEN  2/2/78

Finished 4 (3)
water (3)
Initial
Residual Cl2d 0.5 <0.4
Colonies/ml 0 0
2 days
Residual Cl_ ND6 ND
2
Colonies/ml 0 29
4 days
Residual Cl <0.4 <0.4
Colonies/ml 22 330
6 days
Residual Cl ND ND
Colonies/ml 330 36000
4 (10) Bact. +
fin. (10)°

3 3
1 8000

ND ND

0 130

3.0 <0.4
0 1

1.0 ND
0 3
4 (0)

0
1100



1600


19000


62000

       lumber in parenthesis is ppm (mg/1) Cl  added initially.


        Contact time about 3 hours.  Plating was done two hours after sample collection.

       f"
        Contact time, 10 minutes.  Bacteria were added in the laboratory just prior to plating.


        Initial residual chlorine values, mg/1, determined at plating time.

       @
        ND = not done

-------
TABLE 2.   DISTRIBUTION OF  BACTERIA BY  TYPES  IN SAMPLES TAKEN
            2/2/78.*  ONLY COUNTS  GREATER THAN I/ml ARE GIVEN
                        Finished  (3)  Col.4  (3)  Col.4 (10)  Bact-+ Col.4(10)
                                                           Pin.
Day 1

   Acinetobacter

   Enterobacter
     agglomerans

Aged 2 days

   Ac inetobacter

   Enterobacter
     agglomerans

   Pseudomonas-1ike

   Staphylococcus

Aged 4 days

   Acinetobacter

   Enterobacter
     agglomerans

Aged 6 days

   Acinetobacter

   Moraxella
               28
 22**
330
            36000
                                  8000
                                    10
                     120
 940



 140




1300



 300


 260
                                           19000
                                           62000
330
 *Blank spaces in table  indicate no growth.
**^ese colonies were initially white and developed a yellow&
  two days.  It is suspected that the Knterobacter aqqlomerans overgrew and

  replaced another organism.
                                    359

-------
TABLE 3.  COLONIES CHOSEN FROM ISOLATION MEDIUM FOR
          IDENTIFICATION.   SAMPLES. COLLECTED 2/2/78

Colony
number
la
2a
3a
4a
5a
6a
7a
8a
1
2
3
4
5
6
7
8
9
Description
1mm, yellow
1mm, golden
1mm, lemon
1mm, white
< 1mm, white
malodorous
5mm, white
slimey,
malodorous
1mm, yellow
< 1mm, white turns
yellow, malodorous
5mm , white , mucoid
1mm, yellow

-------
TABLE 3.   (continued)
woj.ony
number
•• '••' 1 11
10
11
12

13
14


15



16



17

18

19



20

21
22
23
24

Description
_
1mm, golden
1mm, white
5mm, white,
mucoid
1mm, yellow

-------
                         TABLE  3.    (continued)
Colony                      Sample
number   Description        source          Colonies/ml        Identity

 25      1mm,white,
         mucoid,malodorous   Fin.+Bact.            3      Acinetobacter

 26      2mm,white,
         mucoid,malodorous   Fin.+Bact.        10000      Acinetobacter

 27      2mm,white,
         mucoid,malodorous   Fin.+Bact.        62000      Acinetobacter
                                  362

-------
U)
        TABLE  4.   BACTERIAL POPULATIONS  (COLONIES  ISOLATED/ml SAMPLE)  OBTAINED IN SAMPLES
                   OF FINISHED WATER AND  GRANULATED ACTIVATED  CARBON  (GAG) COLUMN EFFLUENT
                   TREATED IN SEVERAL WAYS.   SAMPLES TAKEN 2/17/78

Finished 4 (3)
water (3)a
Initial13
Residual Cl d <0.4 <0.4
Colonies/ml 1 0
Aged 3 days
Residual Cl NDe ND
Colonies/ml 42 0
Aged 5 days
Residual Cl,, ND ND
2
Colonies/ml 4 250
4 (10)

3
0

3
0

3

0
Bact. + 4 (0)
fin. (10)

<0.4
8300 8300

ND
2 25000

ND

83000 12000

         Number in parenthesis is ppm (mg/1) C12 added initially.
         Contact time about 3 hours.  Plating was done two hours after sample collection.
         CContact time, 10 minutes.  Bacteria were added in the laboratory just prior to plating.
         Initial residual chlorine values, mg/1, determined at plating time.
         ND
not done

-------
TABLE  5.   DISTRIBUTION  OF BACTERIA (COLONIES/ml)  BY TYPES IN
           SAMPLES TAKEN 2/7/78.*   ONLY COUNTS GREATER THAN
           I/ml  ARE GIVEN
                                                      Bact.+
                    Finished (3)   Col.4  (3)  Col.4  (10)  Fin.  (10)  Col.4 (0)
Day 1

   Ac inetobacter

Aged 3 days

   Acinetobacter

Aged 5 days

   Acinetobacter

Enterobacter
   agglomerans

Staphylococci

II K 2
42
 2

 2
           250
                                8300
83000
           8300
          25000
7500
                                          4150
*Blank spaces in table  indicate no growth.
                                  364

-------
TABLE 6.  QOLONIES CHOSEN FROM ISOLATION MEDIUM FOR
          IDENTIFICATION.  SAMPLES COLLECTED 2/17/78
Colony
number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Description
2mm , white , mucoid
1mm, white , entire
2mm, golden , entire
1mm, lemon , entire
< 1mm, orange

-------
                         TABLE 6.  (Continued)
Colony
number Description
Sample
source
Colonies/ml
sample
Identity
15   
-------
  REPORT NO.
   EPA-600/2-80-130a
                                         TECHNICAL REPORT DATA
                                (flease read Instructions on the reverse before completing)
2.
                                  3. RECIPIENT'S ACCESSION-NO.
  riTLE AND SUBTITLE
   REMOVING POTENTIAL ORGANIC CARCINOGENS  AND PRECURSORS
   FROM DRINKING WATER
   Volume  I and Appendix A
                                  S. REPORT DATE
                                     August 1980 (Issuing  Date)
                                  6. PERFORMING ORGANIZATION CODE
   Paul R.  Wood, Daniel F.  Jackson,  James  A. Gervers,
   Doris  H. Waddell, and Louis  Kaplan
                                                                    8. PERFORMING ORGANIZATION REPORT NO
  'ERFORMING ORGANIZATION NAME AND ADDRESS
   Drinking Water  Quality Research Center
   Florida International University
   Miami,  Florida   33199
                                  10. PROGRAM ELEMENT NO.
                                    61C1C,  SOS #1,  TASK  42
                                  11. CONTRACT/GRANT NO.
                                     R-804521
  . SPONSORING AGENCY NAME AND ADDRESS
   Municipal Environmental Research  Laboratory—Gin.,OH
   Office of Research and  Development
   U.S.  Environmental Protection Agency
   Cincinnati, Ohio   45268
                                  13. TYPE OF REPORT AND PERIOD COVERED
                                     Final  6/22/76-6/30/80	
                                  14. SPONSORING AGENCY CODE

                                     EPA/600/14
15. SUPPLEMENTARY NOTES
    See  also  Volume II,  EPA-600/2-80-130b
    Project Officer:   Jack  DeMarco   (513)  684-7282
16. ABSTRACT
       Feasible and economical methodologies were needed to remove existing organic contaminants—specifically,
       four trihalomethanes  (chloroform, bromodichloromethane, chlorodibromomethane, and bromoform)—from and
       prevent development of potential carcinogens in the public water supplies in Dade County, Florida. A
       four-phase study was  designed to evaluate the efficiency of three adsorbents in removing 19 individual
       halogenated organics  and trihalomethane precursors.  These adsorbents were XE-340—a carbonized polymeric
       macroreticular resin; IRS-904—a strong base catonic resin designed to  remove large molecular weight sub-
       stances such as precursors from water; and granular activated carbon (GAG).  Adsorbent columns were
       placed at various stages in the water processing system: the raw-water, the lime-softened at the up-flow
       Hydrotreator effluent, and the finished water stage.  Four 0.76-meter-deep GAC Filtrasorb 400 columns,
       arranged in series on the finished water line, were most effective in reducing the level of the trihalo-
       methanes present in the finished water.  The Polanyi-Manes theory of adsorption was applied and found
       helpful in interpreting results.

       Appendix A contains the preliminary studies made of the bacterial profile of raw and finished water and
       effluent from four GAC columns from the Preston Water Treatment Plant.  Raw water organisms, which appar-
       ently, can survive existing treatment plant processes, colonized the initially bacteria-free GAC columns
       and released vast numbers of bacteria into the water flowing through the columns.  The development of
       bacterial growth in the GAC columns interfered with backflushing the columns.

       Appendix B  (Volume II of this report) contains the supporting data for  the study.
17.
                                     KEY WORDS AND DOCUMENT ANALYSIS
                     DESCRIPTORS
                                                     b.lDENTIFIERS/OPEN ENDED TERM
                                                                                    c.  COS AT I Field/Group
  Adsorbents,  Granular  Activated Carbon
  Treatment,  Synthetic  Resin Treatment,
  Potable  Water, Organics Control
                    Adsorption,  Specific
                    Organic compounds,
                    General Organic  Para-
                                                                                          13B
                                                      meters
13. DISTRIBUTION STATEMENT

  Release  to Public


  •MWMM^HMMMMMMWMMMM
EPA Form 2220-1 (9-73)
                                                      19. SECl
                                                        389
                   20. SECURITY CLASS (Thispage)
                    Unclassified
                                                  22. PRICE
                 367
«, U S. GOWWMEm PRINTING OFFICE: 1990-657-165/0114

-------