PB82-199704
Biological Processes in the Treatment of Municipal Water Supplies
Rip G. Rice, et al
Public Technology, Inc.
Washington, D.C.
 March 1982
                 U.S. DEPARTMENT OF COMMERCE
               National Technical Information Service

-------
                                          PB32-199734

                                       .EPA-600/2-32- C20
                                       January 1982
 BIOLOGICAL PROCESSES IN THE TREATMENT OF

          MUNICIPAL WATER SUPPLIES


                    by
                Rip G. R1ce
         Jacobs Engineering Group

             C. Michael Robson
             Purdue University

              G. Wade Miller
      Public Technology, Incorporated

               John C. Clark
          Temple, Barker & Sloane

               Wolfgang KUhn
           UnlversitSt Karlsruhe
         Grant No. R-804385-01

      Public Technology, Incorporated
          Washington, D.C.  20036
              Project Officer

             J. Keith Carswell
     Drinking Hater Research Division
Municipal Environmental Research Laboratory
          Cincinnati, Ohio  45268
          •  KWOOJCEO er
            NATIONAL TECHNICAL
            INFORMATION SERVICE
               Hi OfPASIMtN! OF COMMERCE
                 Sf BIKGJItlO, V»  22161

MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
     OFFICE OF RESEARCH AND DEVELOPMENT
    U.S. ENVIRONMENTAL PROTECTION AGENCY
          CINCINNATI, OHIO  45268

-------
                 NOTICE






THIS DOCUMENT HAS  BEEN REPRODUCED



FROM THE BEST COPY  FURNISHED  US BY



THE SPONSORING AGENCY.  ALTHOUGH IT



IS RECOGNIZED  THAT CERTAIN PORTIONS



ARE ILLEGIBLE, IT  IS  BEING  RELEASED



IN THE  INTEREST  OF MAKING  AVAILABLE



AS  MUCH INFORMATION AS- POSSIBLE.

-------
                                   TECHNICAL REPORT DATA
                            (Pteae read fattractions on the revert* before completing)
1 REPORT NO
    EPA-600/2-82-020
ORD Report
                                                           3 RECIPIENT ACCSSSIO."»NO.
M3S   19970 A
  -IT.S ANOSUBTITLo
"Biological Processes  In  The Treatment Of Municipal
 Water  Supplies"
                                                           8 REPORT OATS
                                                             1  March
                      6. PERFORMING ORGANIZAT ON COOS
7 AUTHOR(S)
 Rice,  R.G.,  Robson, C.M., Miller, 7.W., Clark,  J.C.
 and Kuhn.  W.	
                                                           8 PERFORMING ORGANIZAT ON REPORT NO
9 PERFORMING ORGANIZATION NAME ANO ADDRESS
                                                           10. PROGRAM S.cMSNT NO.
 Public  Technology, Inc.
 Washington,  D.C. 20036
                       11 CONTRACT/GRANT NO

                          R804385
12. SPONSORING AGENCY NAME ANO ADDRESS
 Municipal  Environmental Research Laboratory - Gin.,  OH
 Office of  Research and Development
 U.S. Environmental Protection Agency
 Cincinnati,  OH 45268
                                                           13, TYPS OF REPORT ANO PSRIOO COVERSC
                      14. SPONSORING AGENCY COOE

                          EPA/600/14
IS SUPPL6M6N-ASY \OTSS
 Project Officer:  J. Keith Carswell  (513) 684-7228
18. ASSTRACT
 The objective of this protect  was to study the use of Biologically  Enhanced
 Granular Activated Carbonjtechnology in European water treatment  plants and
 to determine  its advantages and  disadvantages for  use in the United States.
 Seven European water works were  visited in which chemical preoxidation followed
 by rapid media filtration then GAC adsorption was  being practiced and BEGAC
 process details were gathered.
                               KEY WORDS ANO DOCUMENT ANALYSIS
                  OESCBIP-OflS
                                             b.lOENTIFISBS/OPeN ENOSD T-RMS
                                      CCSATI Field/Group
1*3 OIS-RIS1-TION STATEMENT

   Release to  public.
         19 SECURITY CLASS
           Unclassified
                                                             ThLt R
       I 31 NO OP PAOsS
       I  441	
                                              20 SECURITY CLASS

                                               Unclassified
                                                                        23 PRICE
SPA Form 2220-1 (»-73)

-------
                            	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 of commercial products constitute-endorsement or recommendation
for use.
                                      ii

-------
                                  FOREWORD


     The Environmental Protection Agency was created because 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 testimony to the deterioration of our natural environment.
The complexity of that environment and the interplay between its components
require a concentrated and integrated attack on the problem.

     Research and development is that necessary first step in problem
solution and it involves defining the problem, measuring its impact and
searching for solutions.  The Municipal  Environmental Research Laboratory
develops new and improved technology and systems for the prevention, treatment
and management of wastewater and solid and hazardous waste pollutant dischar-
ges from municipal and community sources, for the prevention and treatment
of 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; a most vital
communications link between the research worker and the user community.  It
is a survey of biological processes used in the treatment of drinking water
supplies.  Particular emphasis is placed on evaluating the use of biologically
enhanced granular activated carbon as it has been developed and is being
practiced 1n certain European drinking water treatment plants today.

     This study was sponsored jointly by the Industrial and Extractive
Processes Division of the EPA Office of Research and Development, by the
Water Supply Research Division of the Municipal Environmental Research
Division and by the Office of Drinking Water in an effort to assess the
performance of advanced treatment techniques for use in the production of
drinking water and in the treatment of wastewaters.  It is hoped that this
report will be interesting and helpful to those active in these fields.
                              Francis T. Mayo, Director
                              Municipal  Environmental  Research Laboratory
                                     111

-------
                                  ABSTRACT

                                                                   /
     This study resulted from a recommendation made during the course of a
previous survey conducted for the U.S. Environmental Protection Agency by
Public Technology, Inc.  During the earlier survey, An Assessment of Ozone
and Chlorine Dioxide Technologies for Treatment of Municipal Water Supplies,
EPA-600/2-78-147, a biological treatment technique was observed in several
European drinking water treatment plants which appeared to be producing high
quality drinking waters and avoiding the synthesis of halogenated organic
materials during the water treatment process.  This treatment technique
involves the sequential application of chemical oxidation (usually by means
of ozone), rapid media filtration, optional reaeration and granular activated
carbon adsorption.

     Because this biological treatment process seemed to have general
applicability to the control of organic chemicals in drinking water, the
objective of the present program was to study the use of this technology in
European water treatment plants and to determine its advantages and disadvan-
tages for use in the United States.

     This biological treatment technique has been called Biological Activated
Carbon, but perhaps should be more properly referred to as Biologically
Enhanced Granular Activated Carbon (BEGAC).  Europeans substitute the BEGAC
process for prechlorination.  Raw waters which contain readily biodegradable
organic materials are pretreated simply by aeration or oxygenation to create
high dissolved oxygen levels in the water.  In the presence of high oxygen
levels and readily biodegradable organic materials, aerobic biological
activity is promoted in the subsequent rapid filter and GAC adsorption
media.  The aerobic microorganisms degrade biodegradable organics, ultimately
converting them to carbon dioxide and water under optimum conditions.

     If the organic materials in the raw waters are blorefractory, then
chemical preoxidation may convert them Into readily biodegradable materials.
If not, then BEGAC will provide no advantages over granular activated
carbon operating purely in its adsorption mode.

     In addition to converting dissolved organic materials to carbon dioxide
and water, the aerobic microorganisms also are capable of converting ammonia
or airmonium ions to nitrate ions by biological nitrification mechanisms.
Most of the nitrification occurs in the rapid filter medium before the GAC
adsorber.  Therefore, oxygenation normally is required after passage through
the filtration unit.
                                     iv

-------
     Advantages of the BEGAC process Include:

1)   Extension of operating lifetimes of the GAC adsorbers  before reacti-
     vation must be conducted.   Swiss water treatment authorities claim a  5-
     to 6-fold Increase in operating lifetimes.   One European  plant has
     utilized the process for over three years without requiring  GAC reactiva-
     tion.  Newer plants are planning to operate at least two  years before
     reactivating.

2)   The process may be substituted for breakpoint chlorinatlon to remove
     ammonia.  This eliminates  the formation of  trlhalomethanes and other
     halogenated organic compounds during the  early stages  of  water treatment
     processes.

3)   The Dohne plant in MUlheim (Federal Republic of Germany)  converted its
     existing breakpoint chlorinat1on/GAC adsorption process to BEGAC in
     late 1977.  This Involved doubling the depth of the GAC contactors and
     installing two-stage ozonatlon.  Because  of savings in operating labor,
     GAC reactivation costs and costs of chemicals, current operating costs
     at the Dohne plant are no  higher than they  were before process conversion
     — and the quality of finished water is higher.

     Bacteria in the GAC adsorber are present  only in the large macropores,
and do not Interfere with the adsorption properties of the  GAC for refractory
organic compounds.  If the raw water contains  both refractory  organics and
readily biodegradable organics, the GAC unit can be sized to maximize bio-
chemical conversion of organics and ammonia without allowing breakthrough  of
these materials.  While these materials are being removed from solution
biochemically, the refractory organics (I.e.,  halogenated organic materials)
are being removed by adsorption.  The GAC then 1s regenerated  only after the
refractory materials break through.

     A major uncertainty exists with respect to  adoption of BEGAC processes
on a broad scale.  This is the concern for endotoxins which may be produced
when specific types of bacteria die.  However, no evidences of endotoxins  as
yet have been found in BEGAC system effluents.  All microorganisms identified
to date as being present in BEGAC filter/adsorption media and  their effluents
are non-pathogenic soil and water bacteria, fungi and yeasts.   £. coli
organisms, which may be present 1n raw waters, cannot survive  in competition
with the non-pathogenic organisms present.

-------
                                  CONTENTS
Foreword	11 f
Abstract	iv
Figures	,  ix
Tables	xv
Acknowledgements	xxi

     1.   Introduction 	   1
               Site visits 	   4
               Literature search and review	4
     2.   Conclusions	7
     3.   Recommendations	12
     4.   Evolution of Drinking Water Treatment Processes	15
               Early methods of treating water 	  15
               Combinations of physical, chemical  and biological
                    processes	16
               Objectives of drinking water treatment	17
               Removal of organic chemicals through biological
                    treatment	24
     5.   A Review of Biological Processes Used in Drinking Water
               Treatment	35
               River sand bank filtration (tlferfiltratlon)	35
               Slow sand filters'	36
               Storage of surface waters 	  37
               Coarse media biological reactors	37
               Fluidized bed nitrification 	  37
               Ground passage of treated water 	  38
               Biologically active filtration media	38
               Biologically active granular activated carbon 	  39
     6.   Microbiological Aspects of Biological Activated Carbon ....  40
               General principles of biological activity 	  40
               Metabolism	40
               Conditions necessary for promotion of aerobic activity.  .  44
               Stoichiometry of biochemical oxidations 	  46
               Studies of operational biologically enhanced GAC systems.  49
     7.   Granular Activated Carbon	60
               Introduction	60
               Granular activated carbon - basic principles	63
               Selection of activated carbons	66
               Operative mechanisms during drinking water treatment
                    with GAC	67
               Applications of GAC in water treatment	71
               Overview of European experience with 3AC	  84
                                     vii
Preceding page blank

-------
     8.    Basic Parameters  of BAG  Systems	86
               Step no.  1 —  Chemical  oxidation,  oxygenation and
                    aeration	86
               Step no.  2 —  Inert media  filtration  plus  biological
                    oxidation	90
               Step no.  3 —  Reaeration/reoxygenation	93
               Step no.  4 --  GAC adsorption  + biological  axidation  ...  93
     9.    European BAC Applications And Studies	97
               Federal Republic of Germany 	  97
               Switzerland	102
               The Netherlands	102
               Case histories	105
                    DUsseldorf pilot plant studies and  operational  data.lCS
                    Bremen  pilot plant study,	126
                    MUlheim	150
                    Rouen-la-Chapelle	156
     10.   European Microbiological Studies in Activated Carbon
                    Filters & Adsorbers	160
               Introduction	160
               European  status as  of 1973	160
               Current European microbiological  studies	166
                    Research  studies at the  KIWA	166
                    Research  studies at the  University  of Saarlands.  .  .179
               Conclusions  regarding microbiological  aspects of
                    BAC  systems	207
     11,   Estimated Costs  For BAC  Systems	211
               Costs for ozonation systems	211
                    Capital costs  for ozonation  systems	212
                    Operating & maintenance  costs of ozonation  systems  .217
               Costs for GAC  systems	220
                    Capital costs  for GAC systems	223
                    Operating & Maintenance  costs of GAC  systems  .  .  .  .232
               Costs for replacement GAC	240
               Costs for GAC  regeneration	240
               Summary of GAC treatment costs	243
               Cost savings with BAC systems 	254
               Conclusion	262
     12.   Engineering Design  Considerations  For BAC  Systems	263
               Introduction	263
               Parameters  affecting biological growth in  GAC media  .  .  .264
               Discussion of  specific design parameters'	265

References	275
Bibliography	291
Appendices (Plant Descriptions)

     A.   Aubergenville plant  (suburbs of  Paris,  France)	324
     B.   Choisy-le-Roi (Edmond PSpin plant), suburbs of Paris,  France.  .328
     C.   DOsseldorf, Federal  Republic of  Germany	332
     D.   Morsang-sur-Seine (suburbs of Paris)	341
     E.   The Donne plant at Mulheim, Federal Republic of  Germany  .  .  .  .366
     F.   Rotterdam, The Netherlands -- Kralingen water treatment  plant  .401

                                     viii

-------
                                   FIGURES
Number                                                                   Page
  1    Examples of typical  systems  to be investigated  	   13
  2   The DOsseldorf process,  Dusseldorf, Federal  Republic  of Germany.  .   18
  3   The Lengg plant,  ZUrich, Switzerland  	   18
  4   The Rouen-la-Chapelle process, Rouen,  France 	   19
  5   The Kralingen process, Rotterdam, The  Netherlands	19
  6   The Oohne process,  MUlheim,  Federal Republic of Germany	20
  7   The Aubergenvllle process, Aubergenvllle,  France 	   21
  8   The Cho1sy-le-Roi process, Paris, France  suburbs.   Preozonation
        and detention reservoir under construction for 1980	22
  9   Treatment process ultimately planned  for  3 Paris suburbs  plants.  .   25
 10   Suggested treatment process  for raw waters containing initially
        non-biodegradable organic  materials	29
 11   Block diagram of the biological activated carbon process  	   33
 12   Biological activated carbon  system 	   34
 13   Enzymatic oxidation of glucose (glycolysls)	42
 14   The tricarboxyllc add (Krebs or citric add) cycle	43
 15   Intermediates in the S-ketoadipate pathway 	   45
 16   Electron micrograph (4,000 x) of GAC  particle showing
        bacterial growths	50
 17   Simplified metabolic pathway used by  Pseudomonas putida  to
        degrade phenol 	   59
 18   Filter/adsorber Installation  	   62
 19   Dual function filter and adsorber	62
                                     IX

-------
20   Pore volume distributions of typical  decolorizing activated carbons  64
21   Pore volume distributions of typical  decolorizing activated carbons  65
22   Examples of chemisorption on GAC surfaces	69
23   Process steps in the regeneration of activated carbon	80
24   GAC regeneration system components 	   81
25   TOC removal upon passage of BASF activated sludge effluent
       through biologically enhanced (with Og) GAC	   92
26   Schematic representation of a GAC particle, showing pore sizes and
       bacterial growths.	94
27   Microbiological loading of GAC -- dependence on adsorptive
       capacity	100
28   Behavior of microbial populations on GAC over 3 years at
       Wiesbaden, Federal Republic of Germany 	  101
29   Efficiency of removal of COD by rapid sand filter and GAC at
       Lengg plant, ZUrich, Switzerland 	  103
30   Efficiency of COD removal by BAG over 3 years at Moos water works,
       Zurich, Switzerland	104
31   General procedural plan for the treatment of drinking water in the
       bank filtration works on the lower Rhine (Dusseldorf area) ...  109
32   Changes in Rhine river water quality parameters during passage
       through the Dusseldorf water treatment process 	  113
33   Changes in organics parameters in Dusseldorf water treatment
       plants	114
34   Performance of various activated carbons at Flehe test plant,
       Dusseldorf	116
35   Quantities of individual substances determined by temperature
       programmed pyrohydrolysis of GAC vs. total org. Cl loading ...  118
36   Organics removal in DUsseldorf GAC post-adsorber by adsorption
       + biological activity	120
37   Distribution profiles for organically bound chlorine and some
       chlorohydrocarbons in Dusseldorf GAC beds	122
38   Distribution of organically bound chlorine in DUsseldorf GAC beds.  123

-------
39   Organics removal  with 2 types of GAC in  2  Dtisseldorf  treatment
       plants	124
40   Change 1n loading of 2 types of GAC with polar  and  non-polar
       organic chlorine compounds at large CUsseldorf treatment  plant  .  125
41   The Auf dem Werder water treatment process,  Bremen, Federal
       Republic of Germany	127
42   Bremen, Germany water works pilot plant	128
43   Schematic of Bremen waterworks pilot plant 3AC  test units	131
44   Comparison of chemical oxygen demand (KMn04) removal  by slow
       sand filtration and BAG at Bremen, Federal Republic of Germany  .  134
45   Variation of permanganate COD and A(CCO/DO)  with time of year.  .  .  134
46   Percent decreases in colony nurbers and  permanganate  oxygen
       consumption (COD) through GAC over 3+  years	138
47   Percent reduction In colony numbers and  permanganate  oxygen
       consumption (COD) values in the Hydraffin LS-GAC  filter,
       1970-1971	140
48   Variation in specific capacity for dissolved organic  carbon (DOC)
       of a GAC filter with operating tine for  normal (BR) and more
       highly activated carbons 	  148
49   The Rouen-la-Chapelle process, Rouen, France 	  157
50   Histogram of agar plate counts, 3 days at  22°C	163
51   Bacterial numbers on GAC, granular non-activated carbon and
       sand, each fed with non-chlorinated tap  water	167
52   Numbers of pseudomonads and coll forms on GAC ROW 0.8  supra  and
       on non-activated carbon ROW 0.8 during aeration at  25°C	  168
53   Numbers of bacteria and actinomycetes on GAC and GNAC; 0?
       consumption with GAC and 3NAC	169
54   Colony counts in GAC, GNAC and sand filters and in  their
       effluents over 340 days	172
55   Colony counts in GAC, GNAC and sand	174
56   Colony counts in effluents from GAC, 3NAC  and sand  filters  .  .  .  .174
57   Colony counts in Norit PKST GAC in 8 slow sand filters at
       Den Haag	175
                                     XI

-------
58   Adsorption isotherms of 4-nitrophenol  and  4-hydroxybenzoate  on
       ROW 0.8 supra GAC, with and without  bacteria  ..........  177

59   Removal of 4-nitrophenol and 4-hydroxybenzoate  on  passage  through
       ROW 0.8 supra 3AC ........................  178

60   Rate of disappearance of 4-nitrophenol and 4-hydroxybenzoate on
       passage through ROW 0.8 supra GAC ................  179

61   Mean colony numbers at various points  in the Schiersteln plant,
       March '73-March '74 .......................  181

62   Development of colony counts in pilot  plant GAC adsorber
       effluents at the Schierstein plant ...............  182

63   Development of colony counts in GAC of pilot plant adsorbers at
       the Schierstein plant ......................  183

64   Colony & cell numbers and free chlorine contents as functions of
       GAC adsorber length .......................  184
65   Oxygen and C0£ contents of GAC effluents vs.  adsorber length  ...  184

66   Organics content of SAC effluents vs.  adsorber depth  .......  185

67   Behavior of microbial populations on GAC over 3 years at
       Wiesbaden, Federal Republic of Germany .............  186

68   Microbiological loading of GAC -- dependence  on adsorptive
       capacity . ,  ..........................  189

69   Dependence of microorganism loading of 'GAC with time  .......  190

70   Dependence of microorganism loading of GAC with ion concentration.  191

71   Fluctuation of colony numbers in GAC adsorber effluents  at  the
       Schierstein plant during 1  year .................  192

72   Development of microbiological activity  in full scale adsorber
       with new GAC at the Schierstein plant ..............  193

73   Changes in colony numbers at each 3AC step and outlet in the
       Schierstein plant ........................  194

74   Development of colony numbers in effluents of new GAC pilot plant
       adsorbers as a function of flow rate at the Schierstein plant.  .  195

75   Development of colony numbers 1n new GAC charges in the
       Schierstein pilot plant adsorbers as a function of  flow rate  .  .  196

76   Development of colony numbers in effluents of pilot plant
       adsorbers with different types of 3AC  at the Schierstein  plant  .  197

                                     xii

-------
77   Bacterial adsorption on GAC	199
78   Effect of GAC on metabolic activity of bacteria	201
79   Effect of GAC on metabolic activity of bacteria	203
80   Capital costs for 9- and !8-m1n.  GAC empty bed contact times  ...  225
81   Material cost for GAC, Including  cost for purchase,  delivery
       and placement	242
82   Production cost for GAC adsorbers vs. period between
       reactivations	255
83   Total production cost vs. reactivation frequency for $0.60, 0.80
       & 1.00/lb GAC for 100 mgd contactor system 	  260
84   Cost of ozone and GAC 1n combination 	  261
85   Aubergenvllle water treatment plant.  Process diagrams 	  326
86   Choisy-le-Ro1 water treatment plant.  Process diagrams 	  331
87   General layout of the Morsang-sur-Se1ne treatment works	342
88   Morsang-sur-Seine water treatment plant -- processes operating
       1n 1977	344
89   Percent reduction 1n levels of organic materials at Morsang.  ...  345
90   Evolution of CHC1, 1n Morsang waters 	  348
91   Process 1 performance parameters  at Morsang plant	351
92   Process 2 performance parameters  at Morsang plant	352
93   Process 3 performance parameters  at Morsang plant	353
94   Process 4 performance parameters  at Morsang plant	354
95   Analytical sampling points for the 4 water treatment processes at
       Morsang-sur-Seine plant, France	355
96   Effect of ozone on GAC adsorption at Morsang plant 	  359
97   Breakthrough curve for the ozonated GAC adsorber without
       biological activity	361
98   Cumulative plots:  GAC adsorption vs. loading;  with and without
       ozone	362
                                    xiii

-------
99   Effect of biological  activity on the cumulative TOC  removal  of
       a GAC adsorber 	  363
100  Comparison of ozonated and non-ozonated SAC adsorbers	364
101  Comparison of actual  vs.  computer-predicted data for ozonized
       water/GAC adsorber 	  365
102  Schematic diagram of Dohne pilot plant process (Mtllheim)	369
103  Influence of prechlorination on the effectiveness of a process
       consisting of flocculation, sedimentation, ozonation,
       filtration and GAC treatment 	  371
104  Performance & loading of BAC adsorbers 	  372
105  Process scheme for surface water treatment at the Dohne water
       utility in MUlheim after introduction of combined  chemical
       and biological oxidation 	  380
106  Pressure filters at the Dohne (MUlheim) plant	382
107  GAC adsorbers at Dohne plant,. MUlheim, Federal Republic of
       Germany	385
108  Initial performance of BAC adsorbers in the Dohne water works. .  .  389
109  Kralingeri (Rotterdam) plant water sources	401
110  Process flow diagram of the Kralingen water treatment plant,
       Rotterdam, The Netherlands 	  404
111  Desorption of trihalomethanes from GAC at Kralingen  plant	  410
112  Removal of detergents by Kralingen plant SAC 	  412
                                    xiv

-------
                                   TABLES

Number   _                                                                Page
   1  Facilities & Organizations Visited,  June,  1978  	   5
   2  Oxygen Requirements For Conversion of Representative Organic
        Compounds to C02	47
   3  Bacterial Species Found in GAC Adsorber Effluents  at the
        Schierstein (Federal  Republic of Germany)  Water  Treatment P^nt.   52
   4  Types of Fungi and Yeasts Found in GAC Adsorber Effluents  at  the
        Schierstein (Federal  Republic of Germany)  Water  Treatment Plant.   52
   5  Bacteria Isolated from GAC Adsorbers at Philadelphia,  PA  	   54
   6  Bacteria Isolated From GAC Columns and Effluents in Philadelphia  .   55
   7  Enteric Bacteria Isolated From GAC Effluents 	   57
   8  Adsorption Technology Milestones 	   61
   9  GAC Adsorption of Organics 	   68
  10  Comparison of GAC With Sand For Turbidity  Removal  and  Service Life   73
  11  Treatment of Drinking Water by GAC in Sand Filters 	   74
  12  Comparison of GAC Filter/Adsorbers and Post-Adsorbers  at
        Vigneux-sur-Seine Water Treatment  Plant	72
  13  Partial List of European Water Treatment Plants Using  GAC	   76
  14  Partial List of U.S. Wastewater Treatment  Plants Using GAC ....   79
  15  The Main Components of Readily Volatile Compounds  in  the  River
        Rhine at Karlsruhe	108
  16  Comparisons of Organically Bound Sulfur and Chlorine  in GAC
        Control Filters at Different Sampling Stations 	  108
  17  Efficiency of GAC Control Filters for Measuring Dissolved
        Organic Substances 	  110
                                     xv

-------
18  Chlorinated Orgam'cs Removed From GAC  by  Dioxane  Extraction at
      Flehe Test Installation,  Dttsseldorf	119
19  Tests With Ozone and Activated Carbon  Filters  (January  8  to
      February 2, 1970	130
20  Comparative Tests With Different Types of Activated  Carbon for
      Purely Biological  Tests	132
21  Averages of Data Obtained During Last  9 Months  of the 38-Month
      Study With BR Activated Carbon at Bremen,  Federal  Republic
      of Germany	136
22  Extraction of Activated Carbon Used For Different Periods	137
23  Bacterial Data in Raw Water, GAC Filtrate and  RDSFF  Filtrates.  .  .  139
24  Comparison of Various Filtering Rates  (Activated  Carbon LS).  .  .  .  141
25  Comparison of Two Activated Carbon Filters of  the Same  Bed
      Depths with Different Filtering Rates	142
26  Comparison of Different 3AC Grain Sizes (LS  Activated Carbon).  .  .  144
27  Results of Recycling Tests at Bremen Water Works  	  145
28  Tests With B-12 Activated Carbon 	  147
29  Organo-Chlorocompounds Present After Breakpoint Chlorination  ...  150
30  Process parameters at the Dohne Waterworks (Mtilheim) Before
      and After Change of Treatment	152
31  Mean DOC Values After the Different Treatment  Steps  (MUlheim).  .  .  153
32  Geometric Mean Values of Bacterial Counts for  the Mulheim,
      Germany (Dohne) Plant Using Ozone	154
33  Performance of Biological Activated Carbon Filters.   Mean Values
      for 6-Month Operation After a 3-Vonth Starting  Period (Dohne
      Pilot Plant, Millheim, Germany)	154
34  Rouen-la-Chapelle Plant Operational Data  (1976 and 1977)  	  158
35  Removal of Bacteria During Caustic Soda Softening and
      Prechlorination at Colwick, England	162
36  50* Reduction Tiire (Days) of Different Groups  of  Bacteria on
      ROW 0.8 Supra and ROW 0.8 GNAC	168
37  Percentage of Biological Removal of Organic  Substances  During
      Passage Through Granular Activated Carbon	188
                                    xvi

-------
38  Microbiological Share of Oxygen Consumption and C02 Production
      During Passage Through Granular Activated Carbon 	  188
38A Bacterial Species Found in Effluents of Granular Activated
      Carbon Adsorbers at the Schierstein (FRG) Plant	200
39  Types of Fungi and Yeasts Found in Effluents of Granular
      Activated Carbon Adsorbers at the Schierstein (FRG)  Plant.  ...  200
40  Cost Range of Ozonation Systems (From Air)  	  212
41  Construction Cost:  Ozone Generation Systems 	  213
42  Construction Cost:  Concrete Ozone Contact  Chambers	213
43  Capital Costs for Drinking Water Disinfection With Ozone 	  214
44  Construction Costs for Small Scale Ozone Generation Systems.  ...  214
45  Construction Cost for Ozone Contact Chambers (Small Scale)  ....  216
46  Construction Costs for Ozone Generation Systems (Large Scale).  .  .  218
47  Construction Costs for Ozone Contact Chambers (Large Scale).  ...  219
48  Operation and Maintenance Summary for Small Scale Ozone
      Generation Systems 	  221
49. Operation and Maintenance Summary for Large Scale Ozone
      Generation Systems 	  222
50  Conceptual Design Parameters for Package GAC Columns
      (7.5 min. EBCT)	224
51  Construction Costs for Package GAC Columns  (7.5 min. EBCT)  ....  226
52  Construction Costs for Concrete Gravity Carbon Contactors	  227
53  Construction Costs for Concrete Gravity GAC Contactors 	  228
54  Conceptual Design Parameters for Steel Gravity Flow GAC
      Contactors; 20 ft. GAC Depth	230
55  Construction Costs for Steel Gravity Flow GAC Contactors 	  231
56  Construction Costs for Steel Gravity Flow GAC Contactors 	  232
57  Conceptual Design Parameters for Pressure GAC Contactors 	  233
58  Construction Costs for Pressure GAC Contactors 	  234
59  Construction Costs for Pressure GAC Contactors 	  235
                                   xvii

-------
60  Construction Costs for Pressure 6AC Contactors 	  236
61  Operation and Maintenance Summary for Package GAC Columns	237
62  Operation and Maintenance Summary for Gravity GAC Contactors ...  238
63  Operation and Maintenance Summary for Steel  Gravity GAC
      Contactors	239
64  Operation and Maintenance Summary for Pressure GAC Contactors.  .  .  241
65  Construction Costs for Off-Site Regional GAC Regeneration —
      On-S1te Handling and Transportation Facilities 	  243
66  Operation and Maintenance Summary for Off-Site Regional GAC
      Regeneration — Handling and Transportation Only 	  244
67  Conceptual Design Parameters for Multiple Hearth GAC Regeneration.  245
68  Construction Costs for Multiple Hearth GAC Regeneration	246
69  Operation and Maintenance Summary for Multiple Hearth GAC
      Regeneration 	  247
70  Conceptual Design Parameters for Infrared GAC Regeneration
      Furnace	245
71  Construction Costs for Infrared GAC Regeneration Furnaces	  248
72  Operation and Maintenance Cost Summary for Infrared GAC
      Regeneration Furnace 	  249
73  Conceptual Design Parameters for GAC Regeneration	250
74  Construction Costs for GAC Regeneration — Fluidized Bed Process .  251
75  Operation and Maintenance Cost Summary for GAC Regeneration —
      Fluidized Bed Process	252
76  Summary of Estimated Operating Costs « Fluidized Bed GAC
      Regeneration System, Manchester, N.H. Water Works	250
77  Costs of GAC Treatment (18 m1n EBCT), Including Regeneration ...  253
78  Assumptions for 18 Minute EBCT GAC Adsorption	254
79  Amortized Capital and Operating & Maintenance Costs for GAC
      Adsorption (!8-M1nute EBCT), tf/1,000 Gallons 	  254
80  Comparisons Between Total Costs of Various Reclamation Process
      Configurations at National Institute for Water Research,
      Pretoria, South Africa 	  262
                                   xviii

-------
81  Effect of Preozonation (1.25 mg/1)  Before  Storage	330
82  Lower Rhine River Water Treatment Data  for Plants  Utilizing  River
      Bank Filtrat1on/Ozonat1on/Filtration/3AC Adsorption.  Averages
      for 1977	336
83  Plant #4.  Chloro-Organic Material  Data From a  Single  Day's
      Sampling in 1977	337
84  Lower River Rhine Water Treatment Data  for Plants  Using River
      Bank Filtration/KMnO./Polymer Sedimentation/Filtration/GAC.
      Averages for 1977.	338
85  Lower River Rhine Water Treatment Data  for Plant Utilizing River
      Bank Filtration/GAC Adsorption.  Average for  1977	338
86  Upper Danube River Water Treatment  Data for Plant  Utilizing
      Sedimentation/Ozonation/Filtration/GAC.   Average for 1977.  ...  339
87  Comparison of Processes 1, 2, 3 and 4 at Morsang-sur-Seine.
      Data Averaged Over 5 Months of Side-by-Side Operation	356
88  Old Treatment Process at the Dohne  Plant,  MUlheim, Germany,  and
      Median Values of Chlorinated Organics Produced During 1976 and
      1977	367
89  Comparison of Treatment With & Without  Breakpoint  Chiorination  .  .  370
90  Comparison of DOC Removed During First  3 and Second  8  Months of
      Use of Granular Activated Carbon  for  5 Activated Carbons  ....  373
91  Removal of DOC After Treating Ozonized  Water With  A1+3 	  374
92  Biological Activity in Pilot Plant  SAC  Columns, Jan.-Mar.,  1977.  .  375
93  Biological Activity in Pilot GAC Columns,  June-Aug., 1977	376
94  Biological Activity in Pilot GAC Columns.   Mean Values for  a
      6-Month Period (Jan-June, 1977) After a  3-Month Starting  Period.  377
95  Total Bacterial Counts in Biologically  Active GAC Adsorbers. ...  377
96  Biological Oxidation in Fluldized Bed Reactors  	  378
97  Mean DOC and UV-Extinction Values for the  Different Treatment
      Steps at the Dohne Plant 	  384
98  Comparison of Old & New Plant Processes for Removal  of CCC  & NH4+.  386
99  Dohne Plant Performance Data Since  Nov. 1977.  DOC Data,  mg/1.  .  .  388
                                   xix

-------
100  Dohne Plant Performance Data,  Jan.-June,  1978.  Biological
       Production of Inorganic Carbon  in  GAC Contactors  	 390
101  Dohne Plant Performance Data.   Comparison of  DOC Removed versus
       Inorganic Carbon Produced  in Biologically Active  GAC Contactors
       During May - June,  1978	391
102  Bacterial Count Data  at Dohne  Plant  with  BAC  Treatment Process  .  . 394
103  Costs for Ozonation Systems  in MUlheim, Federal Republic of
       Germany. .	398
104  Breakdown of Costs at Dohne  Plant, MUlheim, FRG	 399
105  Description of Treatment Process  Units at Kralingen Plant	 405
106  Analysis of Raw and Finished Waters  at Kralingen, January, 1978.  . 407
107  Haloform Production in Kralingen  Plant, June  1977 - March 1978  .  . 408
108  Representative Colony Count  Data  at  Kralingen Plant	413
109  Capital  Costs of Kralingen Plant  	 414
110  Treatment Components  at the  Kralingen Plant	415
111  Annual Operating Costs of Kralingen  Plant	416
112  Annual Kralingen Water System  Costs	417
113  Rotterdam Water System Revenues,  Water Deliveries and Rates.  .  .  .418
                                    xx

-------
                               ACKNOWLEDGEMENTS


     This report could not have been written without the cooperation of many
people and organizations located 1n Western Europe.  Developing the Informa-
tion necessary to this study required contacting many key Individuals in
Europe who are active 1n water and wastewater treatment, plant superin-
tendents, engineers and chemists, research Institute scientists, academic
research staffs, government scientists and regulatory personnel, executives
of water treatment companies, water and wastewater treatment personnel  at
Industrial chemical manufacturing complexes and representatives of manufactu-
ring companies supplying granular activated carbon and ozonation systems.
Unfortunately, not everyone who contributed to this study can be acknowledged
Individually.

     We are especially grateful to the following Individuals:

Fran$ois Fiesslnger, Soc16t6 Lyonnalse des Eaux et de TEclairage
Drs. Maurice Parl, L. Brener, J. Mignot, Capon, Plcard and A. Moan, Degre"mont
Daniel Versanne, Compagnie GgneVale des Eaux
Jacques Le Pauloufi, Trailigaz
Drs. .Meijers, Krulthof, Poels and Van der Koolj, KIWA
Dr.. J.J. Rook, Dr. van der Helde and Mr. Wiellnga, Rotterdam Waterworks
Drs. E. de Greef and G. P1et, Rijksinstltut
Ewald Hellker and Friedrich Nolte, Dohne Waterworks, Mdlhelm
W. Poggenburg and Dr. C. Engels, Dilsseldorf Waterworks
Prof. Dr. Helnrich Sonthelmer, and Drs. F. Fuchs and M. Jekel, Engler-
     Bunte Institute, University of Karlsruhe
H. Engelhardt and Dr. W. Weisbrodt, BASF, Ludwlgshafen
Prof. Dr. R. Schwelsfurth and Drs. M. Klotz and P. Werner, University
     of the Saarlands
Drs. Roelke, Wlrth and Otto, Lurgl, Frankfurt-am-Ma1n
Wolfgang Kretzschmar, WABAG, Kulmbach
Drs. Eigenwann, KnUsli, Thuer and Dretscher, Ciba-Geigy, Basle
Dr. W1111 Vogler, Roche, Slsseln
Messrs. M. Schalekamp, Valenta, Russell, Naves and Lucy, ZUrlch
     Wasserversorgung
Drs. J. Holgne" and Gujer, EAWAG, DUbendorf
Prof. W. Weber, University of Michigan
Dr. Paul Berger, U.S. EPA, Office of Drinking Water
Prof. Andrew Benedek, McMaster University, Canada
                                     xxi

-------
     Throughout the study we have been appreciative of the-guidance  and
helpful encouragement provided by Paul E.  des  Rosiers  (Industrial  &  Extractive
Processes Division, Office of Research & Development),  Dr. Joseph  A.  Cotruvo
(Office of Drinking Water) and J. Keith Carswell*(Water Supply  Research
Division, Municipal Environmental Research Laboratory),  our  Project  Officer.
                                    xxn

-------
                                 SECTION 1

                                INTRODUCTION


     The Environmental Protection Agency's Water Supply Research Laboratory
in Cincinnati awarded a grant to Public Technology, Inc. in June8 1976 to
conduct a state-of-the-art study of the use of ozone and chlorine dioxide
technologies in municipal water treatment.  The results of this study have
been published as Report Number EPA-600/2-78-147 entitled,, "An Assessment of
Ozone and Chlorine Dioxide Technologies for Treatment of Municipal  Water
Supplies".

     The study revealed that ozonation is a technology currently being
widely used for a range of purposes in at least 1,039 water treatment
plants in 29 countries in 1977.  As part of this study, a team of scientists
and engineers visited 23 water treatment plants in Frances, Belgium,  Gernany
and Switzerland in fay, 1977,,  Twenty of these plants use ozone for one or
more purposes, while chlorine dioxide is used at 14 of these plants.

     During the course of this 4-week on-site survey of European water
treatment facilities, the site visit team observed a biological treatment
technique in use in France, the Federal Republic of Germany and Switzerland
that currently is not practiced in the United States.  This technique involves
the deliberate promotion of aerobic biological growths on filter media
(sand, anthracite) and granular activated carbon media (columns or beds) for
purposes of nitrification and of removing organic chemicals.  The aerobic
biological activity is enhanced by an oxidation step applied prior to activa-
ted carbon treatment.  Such preoxidation steps frequently involve the addition
of ozone.

     Evidence obtained from numerous pilot plant studies and from several
full-scale operating plants in Europe supported the claim that a properly
designed and operated combination of ozone and granular activated carbon
unit processes enhanced organic chemicals removal and reduced the frequency
of regeneration of the activated carbon media.

     Several offices of the U.S. Environmental Protection Agency provided
additional funding to Public Technology, Incorporated in early 1978 to study
this water treatment technique further.  The interest of the Environmental
Protection Agency in this advanced treatrent technique was stimulated by the
need to learn as much as possible about methods for controlling organic
chemical contaminants in drinking water.

-------
     On February 9, 1978, the EPA proposed regulations dealing with the
control of organic chemical contaminants in drinking water.  The proposed
regulations consisted of two major sections, the first of which would set a
maximum contaminant level in finished water for trihalomethanes (THMs).  The
THM regulations since have promulgated by EPA in November 1979.  The second
section of the proposed regulations would require public water supply systems
that serve more than 75,000 persons and that find substantial amounts of
synthetic organic chemicals (SOCs) in their raw water supplies to use granular
activated carbon (GAC), or an equally effective treatment technique, to
reduce the concentrations of these SOCs in their finished waters.

     One of the primary concerns of public water supply systems regarding
the use of GAC, in addition to the high capital cost, is the relatively high
cost associated with frequent regeneration.  The EPA has proposed stringent
performance criteria for the GAC systems.  These may be modified as a result
of coirments received by EPA during the public coirment period.  However, they
were objectives against which water treatment techniques could be gauged as
being cost-effective during the term of this study.  The performance criteria
germane to this study as proposed by EPA on February 9, 1978 are as follows:

(1)  The concentration in the GAC contactor effluent of any volatile halogena-
     ted organic compounds (except for trihalomethanes) determinable by the
     purge-and-trap/gas chromatography method shall not exceed 0.5 yg/1;

(2)  The removal of influent total organic carbon with fresh granular
     activated carbon shall  be at least 50%, and

(3)  The total organic carbon content in the 3AC contactor effluent may not
     exceed the value obtained with fresh granular activated carbon by more
     than 0.5 mg/1.

     Depending upon the type and quantity of synthetic organic chemicals in
the raw water supply of a public water treatment system, under these proposed
performance criteria, a utility might be required to regenerate Its activated
carbon as frequently as 4 to 8 times per year.   Experiences in European
water treatment plants, operating under a different set of performance
criteria for organics breakthrough, have demonstrated that systems using
ozonation followed by granular activated car'bon adsorption can have operatio-
nal lives of their GAC media as long as 2 to 3  years before regeneration 1s
required.

     Some of these experiences were reported on briefly in the earlier EPA
survey report (Miller e_t al., 1978).  Additional details -were required,
however, to substantiate wKether biological treatment in the form of ozonation
followed by GAC is a viable method of enhancing treatment effectiveness
while reducing operating costs.  It has been found that BAC (Biologically
Active Carbon) treatment processes have been -incorporated into at least two
European drinking water treatment plants primarily to remove aimonia biologi-
cally (actually to convert ammonia biologically to nitrate ion).   This.
biological process thereby replaces breakpoint  chlorination and eliminates
the generation of chlorinated organic materials during early stages of the

-------
water treatment processes.  It was desired to learn more about the benefits
of substituting the BAC process for this purpose.
     Therefore, this study was undertaken to acquire Information on the
following specific subject areas:
•    determining design criteria used for ozone/GAC systems 1n Europe;
•    determining mechanisms by which BAC systems operate;
•    determining microbiological aspects of BAC systems;
•    gathering field operational and cost data on ozone/GAC systems;
•    quantification of technical and cost benefits of BAC systems;
•    determining changes 1n U.S. treatment plant designs required for
     retrofitting ozone/GAC systems Into existing plants.
     After consulting with leading European water treatment authorities
during early 1978, the PTI site visit team conducted visits to selected
European facilities during June, 1978.  The primary questions to be answered
were:
(1)  Is BAC effective for removal of organic chemicals, and if so, under
     what conditions?
(2)  Is BAC an effective replacement process for ammonia removal by breakpoint
     chlorination and if so, under what conditions?
(3)  How and why does the BAC process achieve its effectiveness?
(4)  Is a preoxidation step necessary?  If so, must the preoxldant always be
     ozone? "'
(5)  Can the added capital and operating costs of an ozonatlon or other
     preoxldant system be offset by the increase in operating time before
     the GAC must be regenerated?
(6)  Is BAC safe to use for drinking water treatment, from a bacteriological
     point of view?
(7)  What pretreatment and post-treatment steps are made necessary when BAC
     is Incorporated Into a drinking water treatment system?
(8)  Does biological regeneration of the GAC truly occur, and if so, to what
     extent?
     Answers to some of these and other questions were obtained by conducting
a 3-week site visitation of:

-------
 (1)  operational drinking water treatment plants employing granular activated
     carbon facilities designed to promote biological growth,

 (2)  research Institutes and universities conducting studies on the BAG
     process,

 (3)  activated carbon and ozone systems manufacturers 1n Western Europe
     during June, 1978, and

 (4)  by reviewing the published literature on the subject.

     However, the scope of this study could not be confined to ozone/GAC
 treatment systems alone.  Early 1n the study, it became apparent that
 Europeans employ many biological processes 1n the treatment of drinking
water, and that biological activated carbon was simply a more advanced
 treatment system, which was based on earlier operating experiences with
other biological processes applied to the treatment of drinking water
supplies.  Therefore, the scope of this study was extended to Include
discussion of other European biological drinking water treatment methods.
Because of the complexity of the problems of organic chemicals removal
however, our primary emphasis remained on ozone and granular activated
carbon systems.

SITE VISITS

     A complete listing of European facilities visited by the survey team is
shown 1n Table 1.  Pertinent findings from each plant visited are discussed
in Section 9 and a thorough account of the plants visited is included in the
Appendices.

     Several industrial wastewater treatment facilities also were visited,
primarily to learn if the BAG process was being applied to wastewaters, and
if so, to what extent.  If BAG was not being applied to the treatment of
industrial wastewaters, why not and did the process hold any potential for
industrial wastewaters 1n the United States?  The potentials of biological
activated carbon processes applied to industrial wastewaters are reported in
a separate document which has been submitted for review to the EPA Industrial
& Extractive Processes Division, Office of Environmental Engineering and
Technology, Office of Research & Development (P.E.  des Rosiers, Project
Officer).

LITERATURE SEARCH AND REVIEW

     A substantial number of papers were obtained from persons and institu-
tions visited during the June, 1978 survey.  In addition, two technical
conferences, one dealing with Oxidation Techniques  in Water Treatment, held
in Karlsruhe, Federal Republic of Germany during September, 1978, and the
other on Adsorption From the Aqueous Phase, part of the 176th Annual Meeting
of the American Chemical Society held in Miami Beach, Florida, also 1n
September, 1978, contributed several meaningful and timely papers.  A search

-------
TABLE 1.   FACILITIES & ORGANIZATIONS VISITED, JUNE, 1978
Organization
la Chapelle Plant



Morsang-sur-Selne


Degr&nont
KIWA (Keurlngslnstltut voor
Waterle1d1ngsbedr1jven)
(Testing & Research Inst. of
The Netherlands Hater Supply)

R1jks1nst1tuut voor Drlnk-
watervoorzlenlng (National
Institute for Water Supply)

Krallngen Plant
KIWA/R1Jks1nst1tuut
 Water Reuse Facility

Donne Water Treatnent
      Plant

Holthausen Plant
Flehe Plant


Engler-Bunte Irstltut der
 Jnlversltilt Karlsruhe

University of Saarlands


The Lurgi Group


Hardhof Plant
EAWA3 (Eldgenflssissche
Anstalt fllr Wasserversorgung
Abwasserrelnigung und Ge-
wSsserschutz (Swiss Federal
Inst. for Water Resources &
Water Pollution Control
Type
Municipal Owned,
Privately Operated
Waterworks

Privately Owned &
Operated Waterworks

Equipment Supplier
Research Institute
Goverrment Water
  Laboratory


Municipal
Waterworks

Pilot Plant
 Test Unit

Municipal
Waterworks

Municipal
Waterworks

Municipal
Waterworks

Research Inst.-
 Un1vers1ty

University


Carbon Supplier
Municipal
Waterworks

Government Research
 Laboratory
                                                  Location
Rouen, France
         France
Reull Malmalson,
 France

R1jsw1jk,
The Netherlands
Voorburg,
The Netherlands
Rotterdan,
The Netherlands

Dordrecht,
The Netherlands

MUlhelm, Federal
Republic of Germany

Dtlsseldorf, Federal
Republic of Germany

COsseldorf, Federal
Republic of Germany

Karlsruhe, Federal
Republic of Sermany

Homburg/Saar, Federal
Republic of Germany

Frankfurt, Federal
Republic of Germany

Zurich, Switzerland
Dflbendorf, Switzerland

-------
of the published literature yielded a. significant number of papers by Weber,
Packham, Rook, Eberhardt, Sonthelmer, Benedek and others which are applicable
to the subject.  Results of this literature review are Interwoven throughout
this report.

-------
                                 SECTION 2

                                CONCLUSIONS


1)   It 1s universally accepted that the provision of drinking water free
from pathogenic microorganisms 1s the primary responsibility of a drinking
water producer.  Water supply utilities of the United States, 1n achieving
this goal, have sought to preclude the growth of all  types of microorganisms
within the water treatment system.  In other countries, many water utilities
Intentionally Incorporate biological processes Into their water treatment
systems to reduce the levels of dissolved organlcs and still maintain strict
pathogen-free qualities of the finished waters.

2)   The treatment of drinking water by the application of biological
processes 1s not new.  Biological activity 1s one of the processes in the
slow sand filter, which was a key treatment step of early water treatment
facilities, but which is generally considered obsolete in contemporary U.S.
practice.  However, biological treatment, 1n many forms, is an important
process in many European drinking water treatment systems.  Examples of
biological treatment of drinking water include the following:

          .river sand bank filtration
          surface water storage
          gravity clarification
          coarse media biological reactors
          flujdlzed bed nitrification
          biologically active filter media
          biologically enhanced granular activated carbon (BAC)
          ground passage of treated water

3)   The Incorporation of biological treatment steps Into water treatment
processes offers the following prospective benefits 1n water treatment:

     •    reduction of the level of dissolved organic materials
     •    lower oxidant (chlorine, chlorine dioxide or ozone) demand
     •    reduced operational costs
     •    reduced bacterial regrowths in distribution systems.

4)   Biologically enhanced granular activated carbon (BAC) can be defined as
the sequential unit processes consisting of:

     (a) oxygenation by aeration, oxygen injection or chemical oxidation,
     (b) sand, anthracite or multi-media filtration,

-------
(si
               optional reoxygenatlon and
               granular activated carbon (GAC)  adsorption.
      This combination of processes,  chemical  oxidation,  adsorption 'and biochemi-
      cal  oxidation is capable of removing ammonia and some' (but1 not all)  soluble
      organic substances from drinking water.

      5)    Dissolved organic materials in drinking water  can  be classified into
      four generalized categories as  follows:

           (a)   biodegradable, adsorbable by GAC
           (b)   biodegradable, non-adsorbable  by GAC
           (c)   non-biodegradable, adsorbable  by GAC
           (d)   non-biodegradable, non-adsorbable by GAC

      Although  these are simplified generalities stated for the purpose of discus-
      sing treatment of dilute water  streams,  they provide  a  framework  for postula-
      ting mechanisms by which biological  activated carbon  probably functions.
      Weber (1978b) points out that,

           "Bacteria, unless specifically controlled,  will  grow on  virtually  any
           surface, particularly those of fine grained media.   The 'objective  of
           biological treatment schemes  to take advantage of  this phenomenon
           should  be to provide optimum conditions for growth".

      6)    Both the filtration media  and GAC provide supports' for the biomass
      which utilizes soluble organics and ammonia as substrates.  The application
      of strong oxidants,  such as ozone,  to a  raw water stream being treated  can
      change the chemical  nature of the dissolved organic materials.  Strong
      oxidants  can convert some (but  not all)  non-biodegradable  materials  into
      biodegradable materials.   Biochemical  decomposition of  organic nutrients
      adsorbed  by  the high surface area in granular activated  carbon appears  to
      restore a portion of the sites  to again  become available for  adsorption.
      Thus, one objective of preoxidation is to couple adsorption with  biological
      degradation.

      7)    The  porous structure of GAC presents an 'ideal -medium for proliferation
      of attached  biological  growth (fixed film biological  growth).  Both  biomass
      and  substrate are retained by the large  surface  area  of the GAC.   A  relatively
      low  food/microorganism ratio can be maintained'to be  able to  remove  relatively
      low  levels of dissolved organic materials.

      8)    Bacteria become attached to GAC media only  on  the  outer  surface and  in
      the  larger macropores near the  outer surface which  are  sufficiently  large to
      house them.   As a result, only  1 to 2% of, the total surface area  available
      for  adsorption of dissolved organics is  utilized by the  bacteria,  and this
      amount of bacterial  growth 1s not sufficient to  interfere  with normal adsorp-
      tion processes.
\

-------
9)   Bacterial growths build up rapidly in granular activated carbon media.
Those species which consume mainly carbonaceous organic materials attain
their maximum concentrations within 24 to 48 hours after virgin or reactivated
carbon is placed in service.  Nitrogen-converting bacteria take longer to
build up to their equilibrium concentrations (30 to 90 days), but low levels
of ammonia are converted to nitrate within a few days of fresh or reactivated
GAC being placed into service.

10)  Operational water treatment plants utilizing biological activated
carbon processes demonstrate that regeneration cycles of GAC adsorbers can
be extended significantly if a large proportion of the soluble organics
entering the GAC system are biodegradable and if essential conditions, such
as minimum dissolved oxygen levels, are maintained.  While no single BAC
system design 1s as yet universally accepted, sufficient data exist from a
number of plants, to enable the design engineer to commence to establish
design criteria for BAC systems.  It 1s clear that, under certain conditions,
a design can be prepared to take advantage of extended GAC operational life
by enhanced biological activity on both the filter media and on the GAC.

11)  Several microbiological studies have demonstrated that the predominant
microorganisms in the GAC media and in the water leaving the BAC system are
typical, non-pathogenic, soil and water bacteria.  It has been shown that
pathogenic bacteria entering a properly designed and operated BAC system
cannot compete with the predominant microorganisms present, and therefore
the pathogenic species die off.   Further study is required to confirm the
absence of harmful  endotoxins.  It has been demonstrated that only low
dosages of post-disinfectant are necessary to achieve the prerequisite
levels of bacteriological quality of the treated water being discharged to
the water distribution system.

12)  Any decision to install GAC should not be made solely on the benefits
which can be gained from BAC.  Rather, the decision to utilize GAC to
remove specific organic materials should be made first.  Cnce the decision
to install 3AC has been taken, then careful consideration should be given to
extending the operational lifetime and improving the overall organic removal
process performance of the GAC by enhancing biological activity in this
medium.

13)  Reactivation criteria for BAC should be the same as those for GAC, and
should be based upon the particular dissolved organic materials present in
the raw water.

14)  BAC will not provide any significant advantages over GAC when the
dissolved organics to be removed are non-biodegradable and cannot be made
biodegradable even by chemical oxidation with ozone.  Exemplary materials of
this type include many of the halogenated organic compounds produced upon
prechlorination of raw waters.

15)  BAC systems have replaced breakpoint chlorlnation processes in several
new and older European drinking water treatment plants.  This process change
has provided the advantage of not producing halogenated organic materials

-------
during the early stages of the drinking water treatment process (which, once
synthesized, then are difficult to remove).  In addition, replacement of
prechlor!nation with BAG systems also has produced higher quality finished
waters, with respect to dissolved organics, ammonia, lower turbidity levels
and lowered post-disinfectant (chlorine, chlorine dioxide or ozone) demands.

16)  Chemical preoxldation with ozone applied before sand, anthracite or
dual media filtration units followed by GAC (BAG) adsorption in European
water treatment plants has resulted 1n extending the times between back-
washing in each medium by a factor of about 2.

17)  One older European plant replaced breakpoint chlorination with BAG in
1977 at no increase in annual operating costs,  including allowances for
annualized capital costs.

18)  In retrofitting BAG systems into existing  drinking water treatment
plants as post-adsorbers (after sand or other media filtration), provision
should be made to incorporate air scouring Into the backwash cycles of both
the filtration and granular activated carbon media.

19)  Biodegradable organic materials generally  are polar and less tightly
held by granular activated carbon upon adsorption.  Non-biodegradable
organics tend to be non-polar also, for example, many of the halogenated
organic compounds produced upon prechlorination.  Some of these non-polar,
non-biodegradable organic materials can be adsorbed to a higher degree and
be held more tightly by GAC.   Because of these  differences,  some halogenated
organic compounds are able to displace less strongly adsorbed, polar organic
materials from GAC surfaces by the process of desorption.   As a result, even
though a biologically enhanced GAC adsorber may be operating at biological
equilibrium and appears to be saturated with respect to adsorption of biodeg-
radable organic materials, it still can be capable of adsorbing non-polar,
non-biodegradable organics which may be present.  In such instances, reactiva-
tion should be carried out only when the non-polar, non-biodegradable
materials begin to break through the GAC medium.

20)  A suggested screening test can be conducted to determine if a specific
raw water supply will be benefltted by biologically enhancing GAC.  The
amount of biodegradation which can occur in the raw water can be determined,
say by use of a Warburg respirometer.  Next, another sample  of the raw water
can be ozonized with low utilized ozone dosages (1 to 10 mg/1), and the
amount of biodegradable material now present is compared with that of the
non-ozonized raw water.  If oxidation of the organic materials present with
ozone does not increase the rate of biological  activity (the amount of
biodegradable materials present), then it can be concluded that BAC will not
show any advantages over GAC for that water supply.  On the  other hand, if
the amount of biodegradable organic material present Is increased by low
level ozonation, then biological enhancement of GAC should provide performance
advantages.  The extent of such improvements must be determined for each raw
water supply in order to assess whether these process improvements will
justify the increased costs for chemical preoxldation, preoxygenation or
preaeration.


                                     10

-------
21)  For a hypothetical 50 mgd drinking water treatment which has Installed
6AC columns having empty bed contact times of 9 minutes and which must be
reactivated every two months, 1t has been shown that 1f preoxldatlon with 2
mg/1 of applied ozone dosage will extend the GAC reactivation time to six
months, the costs associated with Installing this amount of ozonatlon equip-
ment are balanced by the savings 1n GAC regeneration costs.  Further extension
of the GAC reactivation time (to two and three years as currently occurs 1n
some European drinking water treatment plants using BAC processes) will
provide additional operating cost savings.
                                     11

-------
                                  SECTION 3

                               RECOMMENDATIONS
1.   Investigate the applicability of various biological  water treatment
processes for treatment of drinking water.  This would Include the use of
granular activated carbon, but also other adsorptlve or Inert media.   Such
studies should be conducted on systems not using Initial  breakpoint chlorina-
tion and, Ideally, on systems using no prechlorlnatlon at all.  Examples of
typical treatment systems to be tested are given 1n Figure 1.

2.   Confirm the non-pathogenic nature of bacteria in biologically active
GAC media and in the effluents from such media.

3.   Identify the endotoxins produced by these microorganisms and determine
their toxicological significance.

4.   Conduct studies to confirm the nature of the operative mechanisms
occurring with BAG, i.e., adsorption/desorption versus apparent biological
regeneration.

5.   Obtain more detailed operating Information at selected European plants,
including the Rouen, Dohne, Dusseldorf, Rotterdam and Schierstein plants.
Such information would include characteristics of influent and BAG media
effluents with respect to TOC, COD, DOC, UV absorption, TOC1, ammonia, etc.
Determine the specific parameters used at each operating plant to ascertain
when the BAG must be reactivated.

6.   Determine the operational costs and treatment consequences of doing
away with prechlorination in drinking water treatment plants.  For example,
modification of filter bottoms to allow air scouring and the necessity for
more frequent backwashing.  Operate prototype U.S. plants 1n both modes
(chlorination versus preoxidation by other means) over a (minimum) one year
cycle.

7.   Screen a variety of raw water sources to determine the applicability of
biological treatment processes.  Categorize raw waters according to the
biodegradability of their organic components before and after preoxidation.

8.   Evaluate European water treatment operating practices with particular
emphasis on GAC used without preoxidation.

9.   Demonstrate biological processes for nitrification of ammonia as
possible replacements for breakpoint chlorination.


                                      12

-------
                                                              COARSE MEDIUM,
                                                                  , sand,  etc.)
                            ^-CHEMICAL
                            CLARIFICATION
                                                 air or 0
                                                               AEROBIC
                                                               BIOREACTOR
IMF
preoxidation
                          CHEMICAL
                        CLARIFICATION
GAG
— ^

GAC



,
      Figure  1.  Examples of typical systems  to be  investigated

-------
10.  Study use of oxldants other than ozone for the preoxidation step.
Candidate oxidants other than ozone Include HgO?,  KMn04,  UV (plus air or
oxygen), C102 (free of excess chlorine) and C1NH2  (free of excess chlorine).

11.  Study parameters affecting bacterial  breakthrough 1n biologically
enhanced GAC adsorbers, such as has been reported  in studies  conducted at
the Schierstein, Federal Republic of Germany drinking water treatment plant
after three years of use.  It 1s possible that a bacterial  monitoring para-
meter should be considered for BAG systems.
                                     14

-------
                                 SECTION 4

              EVOLUTION OF DRINKING WATER TREATMENT PROCESSES


EARLY METHODS OF TREATING WATER

     The treatment of water for potable purposes has been conducted 1n a
variety of manners during the history of mankind.  Early practitioners
believed that filtration of seawater through sand would purify it suffi-
ciently for drinking.  Others prescribed the settling of water in copper
containers.  Still other methods called for exposure of water to sunlight
and, interestingly, filtration through charcoal  (Baker, 1949).

     Throughout the centuries, however, techniques used for purifying
drinking water began to evolve a common treatment theme.  The initial
methods used were physical in nature ~ first sedimentation and later
various filtration methods.  Coagulation by chemical addition was used in
ancient Egypt, China and India, but was not practiced widely until  the first
third of the 19th century.  Alum and other chemicals for coagulation were
first used in the United States around 1885 and  became accepted practice in
the early 1900s.

     Likewise, the use'of certain types of chemicals for disinfection has
been practiced for a long time.  Early attempts  at disinfecting were by
boiling.  Chlorine was first used in the United  States in 1908 in a large
scale water works at the Boonton Reservoir of the Jersey City, New Jersey
Water Works.

     The use of bacterial action for purifying water 1s not considered in
the classic U.S. texts on water treatment, although it is common knowledge
that bacteria present in surface waters are the  key actors in purifying
those waters naturally through stabilization or  decomposition of organic
materials.

     Infiltration of polluted surface waters through a soil mantle consti-
tutes an efficient combination of biological and physical treatment processes
which results in purified groundwaters.  Infiltration occurs In two phases:
first, infiltration through the air-saturated (aerobic) zone, then through
an air-deficient (anaerobic) zone.  Organics removal occurs nainly in the
unsaturated zone.

     The upper layers of soil constitute effective biological filters.
Evidence indicates that microbial action for purposes of stabilizing organic
materials does not usually exceed 3 meters of soil depth.  This is because


                                      15

-------
most types of bacteria cannot survive in the highly competitive environment
(Hutchinson & Ridgeway, 1977).  In the saturated zone, pollutant removal is
limited to adsorption. Some microorganisms may be present, but are carried
passively by the flow.  Horizontal microbiological penetration through soils
can be as much as 30 meters (Romero, 1970).

     Biological treatment of wastewater, either through the classical
trickling filter method or by the more recent activated sludge process, has
been practiced successfully for more than half a century.  The so-called
"trickling filter" process was developed in England and was first installed
there in 1893.  In this process, bi .logical processes are promoted on the
surface of the very coarse media which also provide a small degree of filtra-
tion.

     The activated sludge process also was developed in England (1914) and
was so named because it involved the production of an activated mass of
microorganisms capable of aerobically stabilizing a waste stream (Metcalf &
Eddy, Inc., 1972).  The trickling filter process, also developed in England,
is a combination of physical (filtration) and biological  processes.  The
first trickling filter was installed and operated in England in 1893.

     Several other methods of purifying drinking water have been applied
with varying degrees of success over the past century.  Physical methods
(such as microstraining) are used successfully in England and the Federal
Republic of Germany.  Ozone, although rarely used in the U.S.A., is now used
in at least 1,039 drinking water treatment plants throughout the world, as
of 1977 (Miller et al_., 1978).  Many other oxidants also are used; these
include chlorinelfi oxide, chlorine and potassium permanganate.

COMBINATIONS OF PHYSICAL, CHEMICAL AND BIOLOGICAL PROCESSES

     Twentieth century processes for treating water for potable purposes
have changed appreciably in some respects, but not in others.  The methods
that have evolved over the years are either physical, chemical or biological
in nature.  As noted, early methods were mainly physical  in nature.  Chemical
methods came to be practiced much later on.  Today, faced with ever increasing
population growth, mar-made pollution and rising costs of service delivery,
biological treatment processes coupled with physical and chemical methods
are being restudied and employed because of their efficiencies and their
relatively low costs.

     In describing water treatment developments in Europe, Rook (1976) noted
that "national differences have led to a number of (treatment) schools:  the
English school practices purely biological treatment; the French school
adheres to flocculation, often Induced by pulsation, with ozone used as a
disinfectant and color oxidant; many German water engineers swear by the
wholesomeness of groundwater, to such an extent that 1n some instances fully
chemically treated river water is used for artificial replenishment of
groundwater".
                                     16

-------
     This report will focus on biological water treatment processes, primarily
on those occurring in granular activated carbon contactors or adsorbers
enhanced by a preoxidation step.  Before introducing the concept of treating
drinking water biologically, it will be useful to review the purposes for
treating drinking water.

OBJECTIVES OF DRINKING WATER TREATMENT

     The ultimate objective of water treatment is to provide a palatable,
safe supply of water to the public at minimum cost.  Water must be treated
in many cases because (1) groundwater supplies may contain high levels of
dissolved solids, minerals or undesireable contaminants or (2) surface
waters have become polluted by urban and agricultural runoff, industrial
discharges and/or municipal wastewater treatment plant discharges.  Specific
objectives of water treatment are to reduce levels of turbidity to acceptable
levels and to ensure removal of chemical, microbiological and viral contami-
nants.  These objectives are attained by subjecting the raw water to a
series of unit treatment processes.

     Various combinations of physical, chemical and biological treatment
processes have been developed and are practiced widely in Europe.  To
illustrate some of these treatment process combinations, several  simplified
process flow diagrams of processes being practiced during 1978 at selected
European drinking water treatment plants are presented in Figures 2 through
8.

     Each of these plant processes represents a different approach to
drinking water treatment and includes at least one biological treatment
step.  The specific process chosen depends upon (1) local conditions, (2)
the raw water source, (3) plant resources available, (4) local philosophy of
water treatment and (5) standards for finished water quality in the respective
countries.  Yet each process is made up of a combination of physical, chemical
and biological steps.  The physical  and chemical unit processes are readily
recognized, but the biological  process steps require identification.

     River sand bank filtration at the DOsseldorf, Federal Republic of
Germany, plants removes 60 to 75% of the TOC and some ammonia during ground
passage, which requires about three weeks.  (River sand bank filtration is
required at all German plants using the lower Rhine as a raw water source.)
Most of this removal occurs biologically.  Additional biology in the granular
activated carbon columns contributes to organlcs removal and ammonia conver-
sion.

     In the la Chapelle plant at Rouen, France, most of the conversion of
ammonia occurs in the 100 cm deep sand filters placed after preozonatlon.
Most of the biological degradation of dissolved organic chemicals takes
place in the 75 cm deep granular activated carbon beds.  That these organic
materials are biodegradable at Rouen is evidenced by the lack of need to
reactivate the GAC after nearly three years of use.
                                     17

-------
Rhine River
sand bank
filtration



ozondtion
V
J*
filtration

-------
       ozone off-gas recycle
^- •
preozona-
tion of
well water




sand
filtration

V


GAC
adsorption

V


post-
ozonation


^


post-
chlorination

Figure 4.  The Rouen-la-Chapelle Process, Rouen, France
      reservoir
       storage +•
        aeration
     finished water
     • reservoirs.
     post-chlorinatioi
sedimentation
   GAC
adsorption
                               ozonation
                                                                          I
                                filtration
Figure 5.  The Kralingen Process, Rotterdam, The Netherlands

-------
                 off-gas ozone recycle
Ruhr River
water.  Chem.
addition +
preozonation
   ground
  passage
sedimentation
                                     pure
                                   oxygen
     GAC
  adsorption
 ozonation
filtration
                         chlorination
   Figure 6.  The Bonne Process, MUlheim, Federal Republic of Germany
                               20

-------
 well  water,
  coagulation
flocculation,
clarification
                                  phosphorus
                           filtration
                             post-
                          chlorination
cascade
 aeration
                          biological
                           nitrification
Figure 7.   The Aubergenville Process,  Aubergenville,  France.
                                 21

-------
ozone off-
gas recycle
1
1 v
1
1
1
1
1
1
1
1
1
A,
1
1
L.
Seine River raw
water--preozonation
(1.25 mg/1) + 2 days
detention time

post-ozonation
(2.1 ig/1)
v


sedimentation
> r
filtration
(nitrification)
                               post-disinfection

                               (C102 addition)
    Figure 8-  The Choisy-le-Roi  Process,  Paris,  France  suburbs,
               Preozonation and detention  reservoir under
               construction for 1980.
                              22

-------
     At the Oohne plant 1n Mill helm, Federal Republic of Germany, most of the
ammonia conversion occurs 1n the sand filters, after ozonatlon and after
addition of pure oxygen.  Biological degradation of organlcs then occurs 1n
the GAC columns.  The 4 meter deep GAC adsorbers have not required reactiva-
tion during the first year of use, and plant management does not plan to
reactivate the GAC there for at least two years of use.  This decision is
based upon results of a detailed pilot pUnt study using the BAC process.

     Before entering the Kralingen plant in Rotterdam, The Netherlands,
biological activity in the raw water Is promoted by aeration of one of the
two raw water storage reservoirs during an average retention time of 3
months.  Aeration also prevents stratification 1n this reservoir.  Curing
this time, levels of ammonia and TOC are lowered as a result of nitrification
and blodegradatlon, respectively.  Additional  biological activity 1s present
1n the sand filters and GAC contactors of the plant Itself.

     In the Lengg plant in Zurich, Switzerland, some biological  degradation
is known to take place both in the granular activated carbon adsorbers and
in the following slow sand filters.  However,  the raw water  supply from the
Lake of Zurich is of very high quality (less than 2 mg/1 TOC), and the
resulting nutrient accumulation in the GAC and sand media 1s Insufficient to
sustain high levels of bacterial proliferation.

     At the Aubergenville plant, located downstream of Paris on  the Seine
River, well water 1s treated by chemical clarification, cascade  aeration,
phosphorus addition, biological nitrification, ozonatlon and post-chlorina-
tlon.  This specific process sequence has been used since 1969,  although the
biological nitrification step was used when the plant was constructed in
1961.

     Finally, recent pilot plant studies at the Cho1sy-le-Ro1 plant on the
Seine River in the Paris, France suburbs has been conducted  on ozonation of
raw Seine River water, followed by chemical addition and two day detention
before Introduction Into the treatment plant Itself (Gerval, 1978, 1979).
This new technique has replaced breakpoint chlorination followed by dechlori-
natlon by sodium bisulfite.  This modified process has produced  significant
savings because of biological action and 1s being added to the full-scale
plant treatment process (Schulhof, 1980).  Biological nitrification now
occurs 1n the rapid sand filters.

     In the detention tank before entering the plant, ammonia and TOC
levels are lowered biologically.  After subsequent processing, the post-
ozonation step now requires about half the ozone dosage that was required
without preozonation.  Overall, the total amount of ozone now used 1n pre-
and post-ozonation at Cholsy has been reduced by 25% to 332.  Furthermore,
the product water now contains less TOC and the chlorine dioxide post-
disinfectant demand also is lowered.

     This recent process modification at the Cho1sy-le-Roi plant was based
on  everal years of successful operation at the East Moscow, Russia water
                                     23

-------
treatment plant, constructed by the French Compagnie G&iSrale des Eaux
(Schulhof, 1979).

     A more recent program to upgrade the three large water treatment plants
in the Paris suburbs has been presented by Schulhof (1980).  At each plant,
plans have been made and construction has been initiated to install  triple-
stage ozonation and biologically enhanced sand filters and GAC adsorbers.
Raw river water will be treated with low levels of ozone (less than  1 mg/1)
and the ozonized water allowed to stand in storage 2 to 3 days, during which
time some settling and biological degradation will occur.

     A second ozonation (0.5 to 0.8 mg/1 dosage) will follow, then chemical
treatment (aluminum polychloride, powdered activated carbon), flocculation/-
sedimentation, then sand filtration.  Most of the nitrification of ammonia
will occur in these sand filters.  The third stage of ozonation will be
next, for disinfection, followed by GAC adsorption.  Finally, chlorination
(0.2 to 0.3 mg/1 residual) will be provided prior to' distribution.  A schema-
tic diagram of this treatment process is shown in Figure 9.  Notice  that
chlorine is added only after the maximum amount of dissolved organics have
been removed.  This treatment approach will minimize the amount of chlorinated
organics which will form in the treated water.

REMOVAL OF ORGANIC CHEMICALS THROUGH BIOLOGICAL TREATMENT

Introduction

     From the time when municipal water supplies first became a reality for
mankind until the early 1900s when chlorine was first used as a water disin-
fectant, primary emphasis has been placed on finding treatment methods to
prevent outbreaks of diseases such as cholera, dysentery and typhoid fever,
all of which are waterborne diseases.  Chlorination, along with improved
methods of filtration, sedimentation and coagulation, served to reduce the
risk of transmission of these diseases by bacteria present in the finished
water supplies to the point that the problem today is almost non-existent in
the U.S.A. and other developed countries.

     During the 1970s, however, a new problem has been presented to water
treatment scientists and practitioners.  This is the problem of potentially
hazardous organic chemicals in finished drinking water supplies.  Some of
the more than 700 organic chemicals identified to date in raw or treated
water supplies are known or suspected carcinogens to animals or humans.  The
precise threshold levels at which they may be carcinogenic to humans currently
are unknown.  In order to reduce the potential risk to the general populace
and to protect the public health, the water supply industry must address
itself to this new problem.  The U.S. Environmental Protection Agency has
been studying treatment techniques for removal of these types of organic
chemicals for several years.

     The majority of these potentially harmful organic chemicals are a
result of industrialization.  More than 4,000,000 different organic  chemicals
are now in existence, more than double the number which existed only a
decade ago.  They find their way into raw water supplies in numerous ways —

                                      24

-------
ro
en
                           O.
 raw
river
water
                           mg/L
raw water
storage
2 3 days
<1 mg/L PjC
t
I chemical
treatment


H
                                                                             flocculatlon/
                                                                             sedimentation
                                               poly AICI3
QAC
Adsorption
•


ozonatlon,
O.4 mg/L.
over 10 mln
                                                                                sand
                                                                              filtration
                                 NaOCI<
                                     Ichlorlnatlon
                                       echlorlnatlon
         .distribution
                                                  NaHSO3
                        Figure 9. Treatment process ultimately planned for 3 Paris
                                   suburbs plants.
                                             (Schulhof, 198O)

-------
through industrial discharges, urban and agricultural runoff and by accidental
spills during transportation from one part of the country to another.

     The so-called "conventional" water treatment process that has served
the U.S. water treatment industry in such good stead for several decades now
is being recognized as inadequate for removal of most of these man-made
organic chemicals.  Many European countries, with no strongly unified conti-
nent-wide control legislation to protect surface water supplies from indust-
rial contamination, have been forced to investigate and put into practice
what to United States specialists are viewed as advanced drinking water
treatment techniques.  Many of these techniques are quite expensive,,especial-
ly in first-time capital cost.  As a result, the Europeans have sought to
offset the expenses associated with the use of ozonation, granular activated
carbon and other high cost techniques by exploiting biological treatment
processes.  Such processes sometimes are viewed in the U.S.A. as involving
risks of biological contamination of the processed waters.  Nevertheless,
they are being utilized on full-scale in many European drinking water treat-
ment plants, where they offer substantial treatment benefits.  When properly
designed and controlled, biological processes, as they are being practiced
in European water treatment plants appear to entail few, if any, public
health risks.  The remainder of this report will be devoted to discussion of
the use of these biological water treatment processes.

Contrasting Water Treatment Philosophies - USA & Western Europe

     European water treatment philosophies and practices are somewhat
different from those in the U.S.A.  The Europeans appear to adopt more of a
systems approach to water treatment.  This means that they determine the raw
water quality, define the qualities and characteristics that the finished
water should possess and then design a new treatment plant or upgrade an
existing plant process to achieve that quality goal.

     In the United States, an inordinate amount of emphasis and faith has
been placed on a "conventional" treatment and, until recent years, there has
been little concern as to the chemical qualities of finished drinking waters.
American water treatment practitioners traditionally have been more concerned
with bacteriological quality of drinking water than with chemical quality,
and they have used the amount of chlorine necessary to ensure (presumably)
achievement of that bacteriological quality without concern for the chemical
quality side effects that decision entailed.

     Europeans cannot understand the American obsession for using large
amounts of chlorine nor why Americans eschew the use of biological processing
steps in drinking water treatment.  On the other hand, American water treat-
ment officials believe that European systems are too costly and they are
concerned about the deliberate promotion of bacterial growths as part(s) of
the water treatment process.

     The major differences in approaches to water treatment appear to be two
in number:
                                     26

-------
 (1)  The Europeans, espejrtally the Dutch and Germans, do not like to add
     chemicals to waters'/preferring physical and biological methods, unless
     chemical use 1s absolutely necessary;

 (2)  The Europeans are not concerned with "a few bacteria" 1n the water (as
     long as they/are non-pathogenic 1n nature), believing that chemical
     safety 1s more Important than protection against non-pathogenic bacteria.
             f*1 / >
     According' to Prof. Dr. He1nr1ch Sonthelmer of the Engler-Bunte Institute
of the University of Karlsruhe, Federal Republic of Germany, American and
European water treatment practices were quite similar until a few years ago
when 1t,w'as discovered that breakpoint chlorlnatlon causes potentially
harmful halogenated organic compounds to be produced.  As a result, the
Germans began to study treatment techniques that could replace breakpoint
chlorlnatlon.  One of the techniques studied, and which now 1s being used on
full scale, Involves a biological treatment step, which can occur on the
surface of sand filters, on the surfaces of anthracite or hydroanthracite
media and on the surface and 1n the macropores of granular activated carbon
adsorbers.

     Prof.  Sonthelmer cautions that biological processes should be taken to
completion in the water treatment plant (i.e., all  of the biodegradable
materials should be allowed to blodegrade in the water treatment plant
filters and adsorbers), otherwise biogrowths can occur 1n the distribution
systems, unless larger than desired amounts of residual disinfectants
(chlorine,  chlorine dioxide or chloramlnes) are applied.  Thus, biological
activated carbon and other biological  treatment steps were tried Initially
because the Germans were studying the use of biological treatment processes
to substitute for chemical addition, specifically prechlorlnation.

     It should not be inferred from the above discussion that European
technology has been far in advance over that of the United States regarding
the use of biological processes 1n water treatment.  References to current
European drinking water literature (Water Research Center, 1977) Indicate
that there 1s no widespread consideration of biological processes as being
commonly used continent-wide.  However, "ammonia removal using biological
practices is practiced on a limited scale 1n France,  Germany and the United
Kingdom" (Water Research Center, 1977).  It would appear, therefore, that
should the United States water treatment industry seriously consider the use
of biological treatment processes 1n drinking water treatment, we would not
be far behind the European level of knowledge on the subject.

Treating A Dilute Water Stream

     The use of microorganisms, particularly bacteria, to treat water can
require adaptation of some of the techniques utilized 1n wastewater treatment.
Raw water supplies normally are far purer than 1s the average treated waste-
water.  Therefore, the pollutants present in raw drinking water normally
will be present in far more dilute concentrations than those present in
wastewaters.  However, the same basic mechanisms apply to biological oxida-
tions conducted in either medium, Including nitrification and den1tr1f1ca-
                                     27

-------
tion.  Basic sanitary engineering textbooks (Metcalf & Eddy, 1972; Water
Pollution Control Federation, 1977) Include discussions of biological proces-
ses which may be used as reference materials as long as the reader remembers
the vast difference In quality between raw drinking water and raw wastewaters.

     Secondary treated municipal wastewater effluents In the United States
normally contain 10 to 30 mg/1 of BODs and a similar concentration of suspen-
ded solids.  Tertiary treatment can reduce these concentrations to less than
10 mg/1.  Raw water supplies, even when obtained from polluted surface water
sources, usually contain a lesser concentration of biodegradable organic
substances than do secondary treated wastewaters.  Thus, when one attempts
to apply biological wastewater treatment process theories to drinking water
treatment, the relatively low concentration of dissolved organlcs must be
taken Into account, as well as a relative level of blodegradablHty.

     Consider a dilute water or wastewater stream containing 10 mg/1  or less
of total organic carbon (TOC).  The organic portion of this water stream can
be divided, theoretically, into four fractions as follows:

                                     r—-1-Biodegradable, Adsorbable
                    Biodegradable
                                    J—2-E
                  Compounds     I—2-Biodegradable, Non-adsorbable

_Raw Water	

                                 I—3-Non-b1odegradable, Adsorbable
               Non-biodegradable
                  Compounds
                                      — 4-Non-biodegradable,  Non-adsorbable

     If a substantial portion of the raw water stream consists of organlcs
which fall into Fractions 1 or 2, the potential exists for blodegradatlon to
occur, either in a sand, anthracite or hydroanthracite filter, or 1n a
granular activated carbon adsorber, without necessarily requiring preoxldatlon.

     The biological oxidation potential of the raw water constituents,
particularly those of the type of Fraction 3, may be increased further by
subjecting the water stream to preoxidatlon.  By oxidizing the components
of Fraction 3 with ozone, for example, high molecular weight  organic molecules
that normally are not readily biodegradable can be converted  into smaller,
lower molecular weight organic molecules which now contain more oxygen than
did the original organic materials and which are now more readily blodegraded.

     In a water treatment process such as that shown 1n simplified form in
Figure 10, the use of ozone followed by filtration then granular activated
carbon adsorption should be effective 1f biodegradable components of the
types comprising Fractions 1 and 2 are relatively low 1n concentration and
the types comprising Fraction 3 are high 1n concentration. Using the same
reasoning, if the stream to be treated consists mainly of biodegradable
organic compounds, neither ozonatlon nor granular activated carbon may be
                                     28

-------
raw
water
i •
^v ^V
chemical
addition
ozone

^

off-gas recycle






-------
required.  Slow sand filtration alone may remove a substantial portion of
Fraction 1 and Fraction 2 materials biologically.

     According to Sontheimer (1978), two general conditions must exist in
order for significant biological activity to occur in granular activated
carbon media.  First, the dissolved organic materials present must be
adsorbable by the carbon; second, they must be of such a molecular nature
that they can be biodegraded at or near the activated carbon surface.

     There are substances that fall Into neither of these categories, and
these will pass through carbon adsorbers and other biologically active media
without being adsorbed and without being biodegraded (Fraction 4).

     One should be cognizant of the fact that segregation of the organic
components of a dilute stream into these four fractions is merely a convenient
way of developing a useful hypothesis of the biological and adsorptive
removal of organic chemicals from solution.  A number of caveatemptors
should be mentioned, however, 1n order to alert the reader to potential
pitfalls that may be encountered during actual drinking water treatment
practice:

(1)  In general, the more biodegradable an organic compound is, the less
     adsorbable (by GAC) it is.   This is due to its increased polarity as a
     result of containing more oxygen atoms.  On the other hand, the less
     polar an organic compound is (the more non-polar the compound) the more
     adsorbable (by GAC) it will be.  The converse also is true:  the more
     polar a compound is, the less readily adsorbed by GAC the compound will
     be.

     Thus8 when ozone oxidizes the rfeadily adsorbed, non-polar organic
     compounds, this will result in more polar, more biodegradable, but less
     easily adsorbable compounds being formed.

(2)  According to several prominent European water treatment scientists who
     have studied the use of ozone for many years, ozonation should be
     followed by either coagulation/filtration or an adsorption step.  In
     quantities normally used in treating drinking water (1 to 5 mg/1),
     ozone will rarely oxidize organic materials completely to carbon dioxide
     and water.  Reduction in concentrations of TOC due to ozonation alone
     is very modest in Maas River water (Rook, 1972).

     Therefore, one of the primary disadvantages of chemical oxidation
     techniques for organics removal is that oxidation often results merely
     in chemical changes in the chemical structures of organic compounds
     present, but it does not, in general, result in their total removal.
     Non-biodegradable organics may be converted to biodegradable compounds.
     This results in a lower COD concentration, but in a commensurate increase
     in BOOc.  Therefore, a biological or adsorption step becomes necessary
     to effect BOD5 removal (Gaunt!ett & Packham, 1973a, 1973b).
                                     30

-------
 (3)  When biological action 1s promoted 1n carbon adsorbers by preoxldatlon,
     chlorine should not be used In breakpoint quantities as a predlsln-
     fectlng or preox1d1z1ng agent.  The use of chlorine results 1n some
     strains of bacteria being destroyed and promotes competition among
     surviving types.  There 1s also evidence to show that bacterial concen-
     trations are much higher 1n solutions leaving activated carbon columns
     than 1n the Influent solutions, even when high levels of chlorine are
     applied (Hutchlnson & Rldgeway, 1977).  In addition, the surviving
     bacteria now are so resistant to chlorine that they may persist through-
     out the distribution system.

(4)  Sonthelmer (1978) notes that the biological treatment step should be
     carried to completion, I.e., all of the biodegradable material should
     be completely converted to C02 and water.  Otherwise, biodegradable
     organic materials will remain that could continue to provide food for
     bacteria 1n the distribution system.

(5)  One of the concerns regarding the use of a biological treatment step,
     for example, granular activated carbon proceeded by ozonation, 1s that
     bacteria will  slough off the carbon column and be present in the column
     filtrates.  However, frequent backwashlng, properly sizing the depth of
     the activated carbon and controlling bacterial growth rate kinetics
     appear to be adequate mechanisms for prevention of bacterial "slough
     off".   With regard to this latter condition,  Characklis (1973) states
     that the growth rate of slime-forming bacteria becomes limited when the
     attaching surface becomes completely covered  with a single layer of
     cells.   At this point the rate of increase in mass of organisms shifts
     from logarithmic to linear.

GAG Adsorption ProceededBy Ozonation As An Organics Removal Technique

     After years of researchi the U.S. Environmental Protection Agency's
Water Supply Research Laboratory 1n Cincinnati has concluded that granular
activated carbon adsorption is the most effective  technique currently
available for broad range removal of organic chemicals from drinking water
(Symons, 1978).  The primary concern regarding GAC is Its high cost, both in
terms of capital costs and the recurring cost of frequent regeneration of
the carbon.   Research conducted at the DUsseldorf, Federal Republic of
Germany, city waterworks has shown that organlcs removal, when ozonation 1s
applied prior to GAC filters, is greater than the  amount of removal that
could be attributed to that of GAC and ozonation acting Independently (Hopf,
1960).   Thus it was concluded that a synergistic effect must exist.  After
further investigation at DUsseldorf, Bremen and Mfllheim 1n subsequent years,
1t has been concluded that additional organlcs removal is the result of
biological  activity in the carbon columns.

     In European pilot studies and in drinking water treatment plants 1t has
been shown by many workers (Scheldtmann, 1975; Schalekamp, 1975; Van L1er e_t
al., 1975;  Sonthelmer, 1975; Eberhardt, 1975; Van  der Kooij, 1975; Kflhn,
Sonthelmer & Kurz, 1978; Gomella & Versanne, 1977; Sonthelmer e_t al., 1978)
that preozonatlon followed by activated carbon adsorption results in:
                                     31

-------
•    Increased capacity of the carbon to remove organics (by a factor of
     about 10),

t    Increased operating life of the carbon columns before having to be
     regenerated (up to 3 years), especially 1f the GAC can be kept free of
     halogenated organics,

•    Biological oxidation of ammonia 1n Inert media and GAC columns, which
     occurs simultaneously with removal of dissolved organics,

9    Use of less ozone or less GAC for removing a given'amount of organics
     than using either process alone (BAC can be cost-effective over ozonatlon
     or GAC 1n removing Dissolved Organic Carbon - DOC).

     Independent studies on physical/chemical treated sewage at the Cleveland
Regional Sewer District (Gu1rgu1s et al., 1976a,b,c,d; 1978) and 1n Israel
(Wachs e^aK, 1977), have confirmed" tKese advantages with respect to removing
organic materials.

     The terminology, Biological Activated Carbon has been applied by R1ce
et al_.  (1977, 1978) and Miller et al. (1978) to the combination of treatment
processes consisting of (1) ozonatlon followed by (2) filtration through an
inert medium, such as sand or anthracite, followed by (3) adsorption in
Granular Activated Carbon columns or beds.  However, chemical  oxldants other
than ozone, and sometimes simply aeration, can be used 1n step (1).  A
reoxygenation step may be desirable before passage of the water through GAC,
in order to maintain high levels of dissolved oxygen.  This 1s because
biological activity in the inert medium also results in conversion of dissol-
ved organic carbon to COg and of ammonia to nitrate, and the CO content of
the inlet water is lowered correspondingly.  Figure 11 shows a schematic of
the BAC subsystem.   Figure 12 expands upon the subsystem shown 1n Figure 11.

     Granular activated carbon and the preceding inert filter media (sand,
anthracite, etc.) are made biologically active by the deliberate Introduction
of sufficient dissolved oxygen (DO) to aqueous streams just before they are
passed through the filter media or GAC columns or beds.   As long as the
influent water contains sufficient DO to maintain aerobicity of the bacteria
and sufficient dissolved organic carbon to provide food,  the aerobic bacteria
will thrive in this environment.  Eberhardt (1975) has likened bacterial
activity 1n such an Ideal environment to a "herd of cows  grazing in a luscious
meadow".

     In Section 9  the available Information on the performances, mechanisms
of action and technological benefits are discussed.  Costs aspects are
discussed in Section n  and in Section 12', design considerations are"1 discussed
which must be considered when Installing biological treatment processes, as
practiced in Europe, Into United States water treatment plants.
                                     32

-------
            water  or
          wastewater
             aeration,
            oxygenation
           or oxidation
          [inert medium
           filtration
         • aeration  or   •
         I oxvgenation  •
         o   *            •
             C4C
           contactor
               Y
          post-treatment
Fl gure 11. Block diagram of the Biological

          Activated Carbon  process.
                  33

-------
  r
  ! pre-treated water-
  t    or wastewater
an
    ozone
  generator
            air or
            oxygen
sand, anthracite
or mixed media
  filtration
                                                 recycle
                                                (optional)
                                                effluent
                                                             aeration
                                                               tank (not
                                                            always needed)
   Figure 12.  Biological  activated  carbon system.
                     granular
                      activated
                       carbon

-------
                                  SECTION 5

        A REVIEW OF BIOLOGICAL PROCESSES USED IN DRINKING HATER TREATMENT


     For purposes of this report, biological processes for treating drinking
water can be divided Into two categories:  (l) those that occur naturally
when only a medium of growth and proper nutrients are provided: and (2)
those that are stimulated or enhanced by a chemical oxidation step.  It
should be recognized that generally the same biological processes occur in
processes listed 1n both categories, and that the pretreatment techniques
involving chemical oxidation merely change the nature of the substrates
utilized by the bacteria.  This section provides a review of the traditional,
well-known natural processes and previews the newer processes that are
commanding more attention currently.

RIVER SAND BANK FILTRATION (UFERFILTRATION)

     Along the lower Rhine River, in the region around DUsseldorf, Feceral
Republic of Germany, the river banks are comprised of diluvial gravel  and
sand sediments.  Advantage is taken of these deposits by utilizing them as
natural filtration media.  All water works located in this region and on the
Rh,1ne take water not from the Rhine directly, but from 10 to 30 m deep,
vertical or horizontal, wells situated 50 to 250 meters from the river
Itself (Poggenburg, 1975).

  '   About two-thirds of the water taken from these wells is Rhine water
which has passed into the wells over an average period of three weeks.   The
balance is groundwater which flows from the land to the Rhine.

     During the three week time of passage from the Rhine to the wells,
about 60 to 70% of the organic materials present 1n the river water 1s
removed by means of biological processes.   Those organlcs removed are  the
easily blodegraded materials originally present 1n the river water.  The
more blorefractory organic compounds, such as halogenated organlcs, pass
through the river sand bank filters unaffected,  and enter the treatment
plant.

     The combined bank filtrate and groundwater 1n the wells is clear  and
almost always bacteriologlcally acceptable (Poggenburg, 1975).  In addition
to removing the easily biodegradable organic materials originally present 1n
the Rhine River, the river sand bank filters reduce some of the nitrate ion
present to nitrite and ammonia (Hopf, 1970b).  Also, iron and manganese are
extracted from the sand and are present 1n the raw water entering the  treat-
ment plant.  All of these contaminants then are treated 1n the plant by
other techniques.        __


                                      35

-------
      Further  details on  the  processes occurring during river sand bank
 filtration  can be found  1n the description of the Dflsseldorf water treatment
 process  in  Appendix C   of this report and 1n recent articles by Kussmaul
 (1979),  Piet  (1979), P1et &  Zoeteman (1980) and Sontheimer (1980).

 SLOW  SAND FILTERS

      Although all treatment  processes may have some effect on the micro-
 biological  quality of water, only certain of these processes are biological
 in nature.  A classical  biological action in water treatment occurs 1n slow
 sand  filters.  Removal of impurities by slow sand filtration is effected by
 a combination of different processes.  The most Important of these are
 mechanical  straining, sedimentation and some adsorption, chemical and biologi-
 cal activity.  Through chemical and biological oxidation, ammonia is converted
 to nitrate and soluble ferrous and manganous compounds compounds are oxidized
 Into  insoluble ferric and manganic compounds.

      The magnitude of the bacterial activities necessary to effect these
 biological changes are most pronounced in the upper part of the filters, but
 biological activity continues to a depth of 0.4 to 0.7 meter.   Several types
 of predatory organisms thrive in this upper layer of the filter bed.   However,
 in the lower part of the bed, the level  of concentration of the organic
matter has been reduced to the point that few bacteria are able to survive.
Similar conditions arise 1n granular activated media.

      Depending upon whether the slow sand filter i's constructed open or
closed, a significant reduction 1n concentration of organic matter can
occur.  This is brought about by the presence of a thin, shiny matting on
 top of the sand, which consists of threadlike algae and numerous other forms
of aquatic life.

     When water Is filtered at a slow rate through fine sand,  a biologically
active filter mat forms in the top few centimeters of the sand.  This mat,
termed a "schmutzdecke", removes suspended matter, bacteria and modifies the
character of the water biochemically.  There 1s evidence that the ability of
slow  sand filters to remove microorganisms is due to microblal  competition
and predation of protozoa and Invertebrates (Hutchinson & Ridgeway,  1977).

     According to Hulsman (1978), the most Important purifying action of a
slow  sand filter 1s the removal of bacteria,  including £. coll, and  pathogens
when  they are present in the raw water.   The total bacterial  count may be
reduced by a factor of 1,000 to 10,000 and the E..  coll  count by a factor of
100 to 1,000.

     According to Rook (1974), for many lake waters and some Impounded
 surface waters in which the contents of blorefractory organics remain low,
slow  sand filtration is the best technical means of providing  a single step,
efficient reduction in levels of pathogenic and col 1form organisms and
producing biooxldation of ammonia and biodegradable organic matter.

      Recently Schmidt (1978)  has described the removal  of micropollutants
 (dissolved organic compounds) in slow sand filters.

                                      36

-------
 STORAGE OF SURFACE WATERS

     Although principally used for reasons other than storage, many valuable
 chemical and biological changes 1n the characteristics of the raw water can
 occur when surface waters are stored.  Specific changes Include reduction 1n
 the concentrations of organic materials, reduction 1n bacterial levels 1n
 general and in virus, pathogenic bacteria and fecal organism levels 1n
 particular.  Mechanisms by which these actions occur Include ultraviolet
 radiation, natural flocculatlon and sedimentation, competition for available
 nutrients, production of blocldal  waste products by various microorganisms,
 and predation by protozoa and possible parisltic bacteria of the genus
 Bdenovibrio (Fry & Staples, 1976; Hutchlnson & Rldgeway, 1977).

     British research scientists have reported biological sedimentation
methods for removal of ammonia, which can occur during reservoir storage.
This is achieved by controlling the water flow closely in horizontal  and
upflow tanks and thereby holding a biofloc seeded on river silt, sand or
alum floe in suspension.  The ammonia can be removed if a large enough
surface area of biologically active floe is maintained (Milliner et al.,
 1972).

COARSE MEDIA BIOLOGICAL REACTORS

     Coarse media reactors have been applied successfully 1n drinking water
treatment plants for biological conversion of ammonia-nitrogen to nitrate-
nitrogen and have the potential for biological denltrlfication.   A survey of
European water treatment practice  (Water Research Center, 1977)  states that
biological  conversion of ammonia 1s practiced on groundwaters In France at
Croissy, Aubergenville and Villeneuve.

     The 7,000 cubic meter per hour (44 mgd) Aubergenville water treatment
plant which supplies water to the  suburbs of Paris, France has included a
biological  nitrification process since 1961 (Blue, 1979).  Water drawn from
wells in the vicinity of the Seine River contains ammonia levels of approxima-
tely 3 mg/1.   Treatment includes chemical clarification,  cascade aeration,
nitrification, filtration, ozonation and post-chlorlnation.   Ammonia  concent-
rations are reduced from 3 mg/1 as the water enters the nitrification reactor
to a level  of 0.1 to 0.2 mg/1 leaving the reactor.  This  level  Is further
reduced to only trace quantities after sand filtration.   A dosage of 0.1 to
0.2 mg/1 of phosphorus (as P20s) is added to the water before It passes
upward through the 2.5 meter (8.2  feet) deep bed.  Supplemental  air is
provided by blowers on a cycle of 2 munites on and 6 minutes off.  The
reactor medium 1s 0.15 cm to 0.5 cm pozzolanlc (volcanic  stone)  material.
Further information on the Aubergenville plant Is provided in the Appendix.

 FLUIDIZED BED NITRIFICATION

     A fluidized bed bioreactor for intensified nitrification of ammonia has
 been developed at the Water Research Centre in the United Kingdom (Jekel,
 1978a).  This bioreactor uses fine sand grains as a support medium for the
 nitrifying bacteria.  Several investigators have reported the feasibility of
 carrying the reaction to completion by denitrifying the nitrate ion to


                                      37

-------
nitrogen gas in an aerobic phase of biological sedimentation.  This can be
achieved by the addition of methanol.  Rook (1972) notes that this same type
of reactor can be used for biological denitrification of river water.  It
has been demonstrated that fluidlzation of river solids and attached bacteria
can reduce nitrate ions to nitrogen effectively in feed waters made anaerobic
by dosing with ammonia and subsequently dosing with methanol.

GROUND PASSAGE OF TREATED WATER

     This technique sometimes is used for further biochemical polishing of
treated waters.  The Dohne plant at MUlheim, Federal  Republic of Germany
subjects treated water to ground passage with a retention time of 12 to 50
hours.  Prior to installation of the new BAG process  at Dohne, dissolved
organic chlorine (COC1) concentrations were lowered from 151 vg/1 to 92 ug/1
during ground passage.  In addition, total bacteria plate count numbers were
lowered from 3,700 counts/ml  in the GAC effluents to  27 counts/ml after
ground passage.  DOC levels were lowered on the average from 3.4 mg/1 before
ground passage to 2.0 mg/1 after ground passage during 1975 and 1976 (Sont-
heimer, et_a_L, 1978).

     When ground passage was first incorporated at the Dohne.treatment
plant, it was a necessary part of the treatment process, providing a desired
degree of polishing.  Since installing the new BAC process, however, Dohne1s
plant management is using ground passage only as a convenient method of
treated water storage.  At the two other plants in MUlheim which were instal-
ling the BAC process 1n 1978, there will be no need for ground passage
because of the higher quality of water produced (Heilker, 1978).

BIOLOGICALLY ACTIVE FILTRATION MEDIA

     Biological activity is known to occur in filtration media such as
anthracite, hydroanthracite and granular activated carbon.   Anthracite and
hydroanthracite are used In dual media filters placed before GAC adsorbers
in water treatment processing.

     Anthracite frequently is used as an upper layer  in a dual media filter.
The lower layer can be a sand filter having.grain sizes of 0.5 to 2.0 mm.
In between, there can be a layer of coarser anthracite to form a graded dual
medium anthracite prefiltration unit which prevents blinding'Of the sand
filter.  Anthracite also has a lower specific gravity than sand,  so that
after backwashing, this coarser medium will settle on top of the sand filter.

     Anthracite also is used as an upper layer in post-filtration activated
carbon adsorbers of the pressurized, down-flow type,  found in the Cflsseldorf,
Federal Republic of Germany,  plants.  While its purpose in these  plants is
to provide a medium for filtering insoluble ferric and manganic compounds,
anthracite also can provide a medium for bacterial growth.

     Although no specific applications of this type have been noted in
European drinking water treatment plants, European water research scientists
have shown that both anthracite and hydroanthracite can support bacterial
growth.  The primary difference between anthracite and granular activated


                                     38

-------
carbon, with respect to supporting bacterial growth, Is that anthracite
behaves similarly to a trickling filter medium, whereas activated carbon can
accumulate nutrients through adsorption of organic molecules into macro-,
meso- and micropores.

BIOLOGICALLY ACTIVE GRANULAR ACTIVATED CARBON

     Bacterial growths in granular activated carbon contactors have been
reported by many investigators.  These growths can occur under either
anaerobic or aerobic conditions, provided that sufficient substrate is
available and provided that toxic inhibitors are absent (Presecan, 1978).
In fact, addition of toxic inhibitors, such as chlorine, heavy metals,
silver or heat in the form of steam can prevent biological  growths from
occurring.  Until recently, these biological growths, especially in wastewater
treatment applications, have been viewed as nuisances,  causing fouling,
Increased head losses and production of hydrogen sulflde when the activated
carbon adsorbers were operated under anaerobic conditions.

     In recent years, water and wastewater research scientists have begun to
recognize the beneficial effects of biological growths  1n granular activated
carbon adsorbers.  Studies by Weber & Ying (1977), Guirguis et al. (1978)
and Argo (1978) in wastewater treatment applications in the DTSTA". and by
Jekel (1977, 1978), Sonthelmer et al. (1978) and by Schweisfurth and his
coworkers in the Federal Republic" oT Germany (Klotz, Werner & Schweisfurth,
1975; Werner, Klotz & Schweisfurth, 1978) have demonstrated that biological
action can contribute significantly to enhancing treated water quality and
reducing the frequency of granular activated carbon regeneration.

     These effects can best be achieved by (1) using water streams containing
a substantial portion of biodegradable organic materials or substances which
can be converted into biodegradable materials by preoxidatlon, (2) elimination
of chlorination as an Initial treatment step (I.e., breakpoint chlorination)
and (3) use of empty bed contact times (EBCT) of more than 10 minutes.
                                     39

-------
                                  SECTION 6

           MICROBIOLOGICAL ASPECTS OF BIOLOGICAL ACTIVATED CARBON


     In this section those fundamental principles of biological activity
which are considered most germane to understanding the various phenomena
which are believed to occur in operating BAC systems will be discussed.
These biological processes take place during passage of the aqueous solution
through GAC in conjunction with the physical processes of adsorption/desorp-
tion.  Both the biological and adsorption processes can be strongly affected
by the prior treatment step of chemical oxidation.

     In addition, microbiological data will be discussed which have been
obtained from operating pilot plant or full scale plant BAC media.

GENERAL PRINCIPLES OF BIOLOGICAL ACTIVITY

     Background for this discussion has been taken from Clark, Vlessman &
Hammer (1971), Rafelson, Brinkley & Hayashi (1971) and Moat, 1979).

     Several factors affect the growth and proliferation of aerobic microorga-
nisms, the most important of which include temperature, pH, availability of
nutrients, oxygen supply, the presence, of toxins and type of substrate.
Psychrophilic (cold-loving) bacteria grow best at temperatures slightly
above freezing (4°C to 10°C).  Thermophilic (heat-loving) bacteria function
best at an optimum temperature range of 50°C to 55°C.   Mesophilic (moderation-
loving) bacteria grow best 1n a temperature range of 20°C to 40°C.

     Most biological treatment systems operate best 1n'a neutral  pH environ-
ment.  For example, the optimum pH range for operation of activated sludge
systems 1s between pH 6.5 and 8.5.  At pH 9.0 and above, microbial activity
is inhibited.  Below pH 6.5, fungi are favored over bacteria 1n the competi-
tion for food.

     Bacterial toxins commonly present 1n water and wastewater Include
certain heavy metals, such as copper, mercury, lead, zinc, cadmium, chromium;
polychlorlnated organic materials and systemic poisons, such as cyanides.

METABOLISM (Rafelson et^aK, 1971)

     Heterotrophic bacteria (those which use organic compounds as an energy
and carbon source) include the aerobes, anaerobes and  facultative bacteria
(which use free oxygen, when available, but can also respire and  multiply 1n
its absence).  In heterotrophlc metabolism, organic matter is the substrate
(food).  However, the majority of organic molecules present in wastewater

                                      40

-------
are  in the form of large (high molecular weight) molecules which cannot
penetrate bacterial cell membranes and be assimilated.  In order to metabolize
high molecular weight substances, the bacteria must be capable of hydrolyzing
these large molecules Into diffusible fractions for assimilation into their
cells.  Bacteria perform this function by generating oxidation-reduction
enzymes, which are organic catalysts which aid in performing biochemical
reactions at temperatures and chemical conditions compatible with biological
life.  Co-enzyme components of the enzyme systems determine what chemical
reactions will occur.

     The primary product of aerobic metabolism 1s energy, according to
equation 1:

                         bacteria
     Organics + oxygen  	^CQ? + H^O + energy      eq.  1

     Microorganisms use the energy thus generated to process organic matter
to create new cells.   During heterotrophic metabolism of wastewater organics,
and for a given population of microorganisms, the maximum rate of removal of
organic matter occurs during periods of maximum biological growth.   Conversely,
the lowest rate of removal  of organic matter occurs when microbiological
growth ceases.  In aerobic metabolism, assuming an abundance of dissolved
oxygen, biologically available carbon is the limiting factor.  The  supply of
substrate is rapidly exhausted through respiration of C02 and synthesis into
new cells.  An aerobic process results in complete metabolism and synthesis
of the substrate into a large quantity of biological growth.

     If the dissolved oxygen supply becomes depleted during metabolism,
anaerobes will develop.  On the other hand, If the dissolved organic supply
becomes depleted,  then the viable aerobic microorganisms will compete for
the small amount of substrate still in solution.   The rate of metabolism
thus decreases at an Increasing rate, resulting in a rapid decrease in the
number of viable microorganisms.   Starvation then occurs, such that the rate
of microbiological deaths exceeds the rate of cell production.  The total
irass of mlcrobial  protoplasm decreases as the cells utilize their own proto-
plasm as an energy source.   Under these circumstances, mkroblal  activity 1s
termed endogenous.  Cells become old, die and lyse (rupture), releasing
nutrients back into solution.  Cell lysis decreases both the number and the
mass of microorganisms.  Consequently, 1t 1s not possible to convert all
organic carbonaceous material present to C02 by means of biological processes.

Chemical Reactions Involved With Metabolism

     Metabolism involves a large number of individual biochemical reactions,
which can be grouped into the categories of the tricarboxyllc add  cycle,
l.ipid metabolism,  carbohydrate metabolism, amino acid metabolism, or any
permutation of these.  It is the nature of metabolism that there is no
beginning or end of these multiple processes.  All of the reactions are
occurring at the same time, in a very carefully controlled and coordinated
manner.  For the purposes of this and subsequent discussions of metabolic
reactions occurring during BAC processing, we shall consider only some of


                                     41

-------
the reactions which are known to occur with carbohydrates  and trlcarboxyllc
acids.                                                           (

Glycolysis--
     The term glycolysls usually is restricted to the series  of enzymatic
reactions involved in the biological transformation of either glucose  or
glycogen to lactic acid.  This series of reactions 1s also known as  the
Embden-Meyerhof pathway.  A simplified summation of the reactions which
occur and the structures of organic intermediates and products involved
which are pertinent to discussion of pretreatment by chemical  oxidation
(given in Section 8) is given in Figure 13.  One mole of glucose (a  6-carbon
molecule) is converted to two moles of trioses (3-carbon molecules)  by a
series of 4 enzyme-catalyzed reactions, which then are converted into  two
moles of pyruvic acid by a series of 5 additional enzymatic reactions.
Pyruvic acid can be converted, reversibly,  Into lactic acid by reduction or
to acetaldehyde by continued enzymatic oxidation.  Under aerobic metabolic
conditions, pyruvic acid is not reduced to  lactic acid,  but is further
oxidized to COg and water through the trlcarboxyllc acid cycle (see  next
section).
           CHO

         H-C-OH

        HO-C-H

         H-C-OH
         H-C-OH
           CH2OH
steps
  CHO

H-C-OH

  CH2OH

   two
 trioses
                                 CO2
steps
                                    CH3
                                    c=o
                                    CO2H
pyruvic
 acid
                                         acetaldehyde
                                                               lactic
                                                                acid
         Figure 13.  Enzymatic oxidation of glucose (glycolysls).


                     (Rafelson,  Blnkley &  Hayashl, 1971).
THcarboxyllc Acid Cycle (Krebs Cycle; Citric Acid Cycle)—
     This cycle includes the reactions which occur during  glycolysis,  but
also involves reactions of other organic acids,  specifically  oxalacetic,
malic, fumarlc, succinic, citric, aconitic,  isocitric,  oxalosuccinic and cc-
ketoglutaric.  Figure 14 shows a simplified  representation for  the  relation-
ships between these compounds during the tricarboxylic  acid cycle.  The
structures of these organic acids become of  greater significance when  it is
                                      42

-------
realized that many of these same adds are produced when many of the  organic
pollutants  present 1n water and wastewater are ozonized.  This point  will be
discussed 1n  greater detail 1n Section 8  of this report.
                                                        glucose
                      OsC-COjH
oxalacetlc
  acid
 malic   HO-C-C02H
  acid
                It
 fumarlc       HC-COjH
  acld   HOOC-CH
                1!
succinic   H2C-CO2H
  8cld    H2C-C02H
              CO2
                                              CO,
                                        2 acetyl
                                         co-enz.-A
                                                       2 trloses
                                                          tl
                                                     -2 pyruvic acids
                                                       2 lactic acids
                                  HO-C-CO2H
                                  H2C-CO2H
                                               citric  acid
                                          H2C-C02H  aconltlcacld
                                             C-CO2H
                                            HC-CO2H
                         H2C-C02H
                          HCH
                         O=C-CO2H
                   o-ketoglutarlc acid
                      H2C-CO2H
                    —  HC-CO2H 	'
                      OsC-CO2H
                    oxalosucclnic acid
                                                                isocltr.c
                                                      HC-CO2H    acid
                                                    HO-C-CO2H
2CO2
    FlguroU   The trlcarboxyllc acid (Krebs or  citric acid)  cycle.
                (Rafelson, Binkley & Hayashl, 1971).
                                     43

-------
The g-Ketoadipate Pathway-
     Figure 15 shows the structures of chemical Intermediates which are
formed then oxidized further during the degradation of benzole and p-hydroxy-
benzolc acids by reaction sequences catalyzed by a variety of enzymes to
produce succinic add.  Once formed by these mechanisms, sucdnlc add can
be further oxidized according to the Krebs cycle.  Many of the chemical
intermediates produced during the 8-ketoadipate cycle also have been identi-
fied during chemical oxidation of aromatic compounds (see Section 8).

CONDITIONS NECESSARY FOR PROMOTION OF AEROBIC ACTIVITY

     In the balance of this Section, only the promotion of aerobic biological
growth will be considered, because BAC processing, as currently practiced,
involves aerobic microorganisms.  In order to maximize the aerobic bloactivl-
ties reguired, it 1s necessary to first maintain a minimum level  of dissolved
oxygen (DO) 1n the aqueous medium being treated.  The total amount of CO
considered minimal to convert the maximum amount of dissolved organic carbon
(DOC) to C02 will be a function of the concentration of biologically oxidiz-
able materials present.  Raw waters being processed into drinking water
contain relatively small amounts of DOC and ammonia, thus the maintenance of
3 to 6 mg/1 of DO, sufficient to guarantee aerobic conditions, can be accom-
plished by addition of relatively small amounts of oxygen'or air.  On the
other hand, industrial wastewaters usually contain much higher concentrations
of organic substrate materials.   Therefore, the amount of oxygen  required to
maintain an aerobic biomass will be considerably higher.

Solubility of Air and Oxygen In Water

     Air contains only 21% oxygen, most of the balance being nitrogen.   As
a consequence, when water is saturated with air, the-maximum concentration
of dissolved oxygen which can be attained Is on the order of 6 to 9 mg/1,
depending upon the temperature.   The solubility of oxygen and nitrogen in
water increases as the water temperature decreases, so that the maximum
solubility of each of these gases occurs near 0°C.

     On the other hand, when water 1s saturated with nominally pure oxygen
rather than air, the maximum concentration of dissolved oxygen which can be
attained rises to 40 mg/1 (at 20° to 25°C) and to 69 mg/1 (at' 0°C).  Therefore,
use of pure oxygen will allow 7 to 8 times the DO level to be attained in
aqueous systems than will the use of air.  This point 1s of significance
when the initial loading of biodegradable material 1s high.

     The point is also significant when considering the rates of  aerobic
microbiological reactions, which depend upon the DO concentrations.  The
faster the rate of microbial oxidation in an operating BAC system, the lower
will be the detention time required for the aqueous medium to be  1n contact
with the microbial species performing the oxidations.   Thus, the  higher the
rate of biological oxidation, the smaller can be the biological reactor
and/or the faster can be the flow rate of water or wastewater to  be put
through the specific treatment system being employed.
                                     44

-------
  protocatechuate
       branch
 p-hydroxy-
    benzoate
    proto-
  catechuate
              OH
               COOH

               OH
               COOH
3-carboxy-
 muconate
                 COOH
       HooclOcOOH
3-carboxy-
 mucono-
 lactone
                               eatechol  branch

                                      COOH

                                          benzoate
                                HOOC^OH   1-2 dlhydro-1,2-
                                             dihydroxy-
                                             benzoate
                                     OH
                                         OH eatechol
                                               muconate
                                              (+)-mucono-
                                                 ,actone
                                          OOH

                              SCOOH  ^-ketoadlpate
                              C-O     enol-lactone

                            \
                  0-ketoadipyl-coenzyme- A

                            J
                  succlnate  +  acetyl-coenzyme-A
  Figure 15. Intermediates In the  0- ketoadipate
             pathway. (Moat, 1979)
                            45

-------
STOICHIOMETRY OF BIOCHEMICAL OXIDATIONS

Dissolved Organic Carbon

     Organic compounds have basic carbon-carbon backbone chains.  Those
organic materials which contain only carbon and hydrogen atoms are called
"hydrocarbons".  Examples include methane, butane, benzene, gasoline, kerosene
and phenanthrene.  When the organic molecule also contains oxygen atoms, in
addition to simply carbon and hydrogen, the material is called generically a
"carbohydrate".  Examples include alcohols, acids, aldehydes, ketones,
phenols and sugars.  In a sense, one can consider carbohydrates as partially
oxidized hydrocarbons, on the way to being oxidized completely to C0£ and
water.

     One can generalize the chemistry involved in the total oxidation of
hydrocarbons as follows:
          CH  + (x + 0.25y)02 - *-xC02 + 0.5yH20      eq. 2

     The amount of oxygen theoretically required (x + 0.25y) for complete
oxidation of all dissolved organic carbon to carbon dioxide and for converting
all hydrogen to water will depend upon the relative amounts of these two
elements which are present initially.  If adequate oxygen is present for the
reaction to proceed to completion, the products can be as shown by equation
2.  If insufficient oxygen is available however, only partially oxidized
intermediate organic products, will be obtained.  In addition, once the
dissolved oxygen supply has bean depleted, the predominant microbial species
may become anaerobes, which can draw their required oxygen from the partially
oxidized intermediates (carbohydrates) and regenerate reduced organic ccm-
pounds as end products, such as in the reversible conversion of pyruvic acid
to lactic acid.

     The amount of oxygen required for complete conversion of an organic
material to C02 and HgO also depends upon the composition and the degree of
oxidation of the starting material.  For example, Table 2 shows the arrount
of oxygen required to attain 100% oxidation to C02 and ^0 starting with the
2-carbon-containing hydrocarbons ethane (C2He), ethylene (C2H4) and acetylene
(C2H2), the 2-carbon-contain1ng carbohydrates ethanol (Cl^CHgOH). acetaldehyde
(CHsCHO), acetic acid (CHjCOOH), oxalic acid (HOOC-COOH) and the 6-carbon-
containing compound phenol (CeHsOH).  The two-carbon compound containing the
most hydrogen (ethane) requires the most oxygen, in the 2-carbon-containing
series.  On the other hand, oxalic acid, which contains four atoms of oxygen
and only two hydrogen atoms, requires the smallest amount of additional
oxygen to carry the oxidation to completion.
                                     46

-------
             TABLE 2.  OXYGEN  REQUIREMENTS  FOR
                        CONVERSION OF  REPRESENTATIVE
                        ORGANIC  COMPOUNDS  TO CO2
CH3-CH3 +
CH2=CH2 +•
CH5CH +
CH3CH2OH H
CH3CHO +•
CH3CO2H +
HO2C-CO2H 4
C6H5OH +
3.5 O2 	
3 Q2 	
2.5 O2 	
• 3 O2 — -
2.5 02 	
2 O2 —
0.5 O 2 —
7 02 	
-*-2CO2 •«
-^2CO2 4
-»-2CO2 ^
~^2CO2 H
-*-2CO2 H
— »-2CO2 ^
-*-2CO2 •(
-*-6CO2
• 3H2O
• 2H2O
H2O
h 3H2O
• 2H2O
h 2H2O
1- H2O
4- 3H2O
     For comparison, the corresponding oxygen requirement of the six-carbon
compound, phenol (CgHsOH) to go to completion 1s 7 moles  of oxygen  per mole
of phenol.  If only 0.5 mole of oxygen is available,  the  maximum oxidation
which phenol can undergo would be to a d1hydroxyaromat1c  compound,  either
catechol, resorcinol or hydroqulnone:
                                                                         • 3
        phenol
catechol
resorcinol
                                                               H
hydroqulnone
     When organic materials contain nitrogen or sulfur,  even  more  oxygen may
be required to convert these elements to their fully oxidized states.   For
example, amine groups (-NH2) will  require 1.5 equivalents  of  oxygen  per
equivalent of amine-nitrogen to produce a nltro group, and 2  equivalents of
oxygen to produce an equivalent of nitrate ion.   Sulfur-containing compounds
can require 2 equivalents of DO per equivalent of sulfur atom to produce one
equivalent of sulfate ion, depending upon Its degree of  oxidation  in the
original organic compound.

     From these considerations, 1t 1s clear that 1n order  to  operate BAG
processes at maximum efficiency, sufficient oxygen should  be  employed  to
convert the maximum amount of dissolved organic carbonaceous  material  into
C02 and water.  It must be remembered, however, that biological processes
cannot remove 10056 of the dissolved organic carbonaceous material.  This 1s
because microbial processes inevitably involve cell lysis  and release  of
bacterial cell contents to solution.
                                     47

-------
Nitrification

     AutotropMc bacteria use C02 as a carbon source and oxidize Inorganic
compounds for energy.  Autotrophs of greatest significance 1n sanitary
engineering are the nitrifying, sulfur and Iron bacteria.  Free ammonia
and/or ammonium Ions react with oxygen biochemically 1n the presence of
nitrifying organisms to produce nitrite ions and, ultimately, nitrate ions:

                         Nitrosomonas
               NH3 + 02 	*~ N02~ + energy       eq. 4

                          Nitrobacter
               N02~ + 02	*"N03~ + energy       eq. 5

overall;

          NH3 + 02	*-N03~ + [H,0, H* or (OH)"]  +  energy    eq. 6

     The actual stoichiometry Involved will depend upon the pH; of the
medium.  On the alkaline side of pH 7,  ammoniacal hydrogen will be converted
to hydroxyl Ions, some of which will not proceed to water molecules.  On the
acid side of pH 7, some hydrogen atoms will be converted to hydrogen ions,
which will not continue to react further to become water.

     Cornelia & Versanne (1977) have stated that one equivalent of ammoniacal
nitrogen should require 4.57 equivalents of oxygen to produce one1 equivalent
of nitrate ion.  On the other hand, these authors found during full-scale
drinking water treatment plant studies  at the Rouen-la-chapel1e plant 1n
France, that only 3.2 equivalents of oxygen actually were required to elimi-
nate ammoniacal ammonia from the raw water.  It 1s possible that the nitrifi-
cation is incomplete at Rouen (see Section 9) and that some nitrite is
present after GAC treatment along with  nitrate.  This would explain the
lower-than-theoretlcal  amount of oxygen apparently required a£ Rouen.

     On the other hand, Jekel (1978) considers, the stoichtometry of nitrifica-
tion which occurs at the Dohne plant in Mfllheim, Federal Republic of Germany,
to require 3.56 equivalents of oxygen per equivalent of ammonium-nitrogen,
according to the following equation:

               m^ + 202-	>-N03~ + 2H* + H20           eq. 7

     From the above considerations, 1t 1s clear that for maximum conversion
of dissolved organic compounds Into C0£ and water and for the maximum amount
of nitrification to occur, a considerable amount of DO can be required.
This amount depends upon the components of the original organic material
present originally.  In most cases, however,  the equivalent amount of oxygen
necessary is some multiple (3 to 10) of the equivalents of ammonia-nitrogen
and dissolved organic carbon and hydrogen present originally.  The presence
of other oxygen-demanding elements, such as sulfur, can Increase the requisite
amount of 00 even further.
                                      48

-------
     There 1s one more Important point to consider with nitrifying bacteria.
 In the presence cf large quantities of heterotrophlc microorganisms, the
 growth and proliferation of nltriflers 1s suppressed.  Therefore, when
 treating industrial wastewaters containing high levels of readily biodegra-
 dable pollutants, nitrification may not be observed.  Cn the other hand, 1n
 treating raw drinking waters (which contain much lower levels of COC), the
 density of heterotrophlc microorganisms will be much lower.  Therefore,
 nitrlfiers will have a better environment in which to proliferate and function.

 STUDIES OF OPERATIONAL BIOLOGICALLY ENHANCED GAC SYSTEMS

     Van der Kooij (1975) investigated the presence and behavior of bacteria
 on granular activated carbon filters treating drinking waters.  Tests were
 conducted on three experimental filters:  GAC, granular non-activated carbon
 and sand.  Each filter was fed with tap water at 13° to 17°C and 3.5 m/hr (3
minutes contact't1 ire) flow-through rates over a 10 month period (side-by-
 side experiments).  Bacterial  counts were made at regular intervals by the
 colony count technique on diluted agar (0.35 g/1 beef extract, 0.65 g/1
 peptone, 10 g/1 agar) after incubation 10 days at 25°C.  The number of
 bacteria on the GAC was found to be 2 to 3 times the number found on inert
media (van der Kooij, 1978).

     From these experiments,  van der Kooij (1975) calculated the average
 surface area occupied by a single bacterial  cell on activated carbon to
 be 40 square microns when the colony count was 108/cu cm.   Since the surface
area of the activated carbon used for these tests was much greater than 40
 sq cm/cu cm, 1t was concluded that the density of bacteria on the carbon
 surface was very low.  This was confirmed by electron microscopy.  It was
 also concluded that normal adsorption processes of organic solutes on activa-
 ted carbon were not hindered by the presence of the bacteria on the carbon,
 using a dilute substrate solution such as drinking water.

     In support of this work of van der Kooij, it has been shown by Weber,
 Pirbazari & Melson (1978) in the USA that individual bacteria are small
 enough to be able to fit Into the macropores (50 to 100 microns in size) of
 the GAC, where they are protected from shear forces.  Cn the other hand,
 bacteria are too large (1 to 16 microns) to fit Into the GAC mlcropores,
 some of which are less than 0.05 micron 1n diameter.  Since the micropores
 make up about 99% of the available carbon surface area, the fact that bacteria
 occupy only 1» of the total GAC surface area is consistent with this ratio.

     Figure 16 is a 4,000 times magnification scanning electron micrograph
 of a GAC particle taken from a GAC column operating at the Philadelphia
 Torresdale drinking water treatment pilot plant, and shows a typical bacterial
 population (Cairo et al_., 1979}.   Microorganisms are found to be scattered
 over the carbon surface, with heavier concentrations in the areas of the
 macropores.  Bacteria were not observed to form a uniform biofUm over the
 entire surface area of the GAC particle, as has been hypothesized by some
 investigators.  Few higher forms of life, such as protozoa, were observed en
 the 3AC particle surface.  Although such an advanced ecosystem was observed
 in wastewater GAC systems by Weber, Pirbazari & Melson (1978), lower concen-
                                     49

-------
Figure 16. Electron  micrograph (4,000 ») of GAC  particle
           snowing  bacterial growths.
           (Cairo. McElhaney &  Suffet, 1979)
                        50

-------
     There 1s one more Important point to consider with nitrifying bacteria.
 In the presence of large quantities of heterotrophlc microorganisms, the
 growth and proliferation of nltHflers 1s suppressed.  Therefore, when
 treating industrial wastewaters containing high levels of readily biodegra-
 dable pollutants, nitrification may not be observed.  On the other hand, 1n
 treating raw drinking waters (which contain much lower levels of DCC), the
 density of heterotrophic microorganisms will be much lower.  Therefore,
 nltrlfiers will have a better environment in which to proliferate and function.

 STUDIES OF OPERATIONAL BIOLOGICALLY ENHANCED GAC SYSTEMS

     Van der Kooij (1975) investigated the presence and behavior of bacteria
 on granular activated carbon filters treating drinking waters.  Tests were
 conducted on three experimental filters:  GAC, granular non-activated carbon
 and sand.  Each filter was fed with tap water at 13° to 17°C and 3.5 m/hr (3
minutes contact time) flow-through rates over a 10 month period (side-by-
 slde experiments).  Bacterial  counts were made at regular Intervals by the
 colony count technique on diluted agar (0.35 g/1 beef extract, 0.65 g/1
 peptone, 10 g/1  agar) after Incubation 10 days at 25°C.  The number of
 bacteria on the GAC was found to be 2 to 3 times the number found on inert
media (van der Kooij, 1978).

     From these experiments,  van der Kooij (1975) calculated the average
 surface area occupied by a single bacterial  cell on activated carbon to
 be 40'square microns when the colony count was lOS/cu cm.   Since the surface
area of the activated carbon used for these tests was much greater than 40
 sq cm/cu cm, 1t was concluded that the density of bacteria on the carbon
 surface was very low.  This was confirmed by electron microscopy.  It was
 also concluded that normal adsorption processes of organic solutes on activa-
 ted carbon were not hindered by the presence of the bacteria on the carbon,
 using a dilute substrate solution such as drinking water.

     In support of this work of van der Kooij, it has been shown by Weber,
 Pirbazari & Melson (1978) in the USA that Individual bacteria are small
 enough to be able to fit Into the macropores (50 to 100 microns in size) of
 the GAC, where they are protected from shear forces.  On the other hand,
 bacteria are too large (1 to 16 microns) to fit Into the GAC micropores,
 some of which are less than 0.05 micron 1n diameter.  Since the micropores
make up about 99% of the available carbon surface area, the fact that bacteria
 occupy only IX of the total GAC surface area 1s consistent with this ratio.

     Figure 16 is a 4,000 times magnification scanning electron micrograph
 of a GAC particle taken from a GAC column operating at the Philadelphia
Torresdale drinking water treatment pilot plant, and shows a typical bacterial
 population (Cairo et al., 1979}.   Microorganisms are found to be scattered
 over the carbon surface, with heavier concentrations 1n the areas of the
 macropores.  Bacteria were not observed to form a uniform b1of1lm over the
 entire surface area of the GAC particle, as has been hypothesized by some
 Investigators.  Few higher forms of life, such as protozoa, were observed en
 the GAC particle- surface.  Although such an advanced ecosystem was observed
 in wastewater GAC systems by Weber, Pirbazari & Melson (1978), lower concen-
                                     49

-------
Figure 16. Electron micrograph (4.000 x) of GAC  particle
           showing bacterial growth*.

           (Cairo. McElhaney & Suffet, 1979)
                        50

-------
 trations of nutrients  in the potable water GAC system are considered by
 Cairo et al_.  (1979)  to be responsible for the limited biological colonization
 on the GAC particles.

     On the other hand, soluble organic compounds generally are small
 enough to be able to fit inside and be adsorbed by the GAC micropores.
 Therefore, since the adsorption of organics occurs mostly where the bacteria
 are not present, it would be predicted that bacterial presence will rot
 interfere with the normal GAC adsorption processes, unless the nuirber of
 bacteria become so high so as to physically encapsulate the carbon particles,
 thereby preventing the diffusion of soluble organics to the micropores
 (Weber & Van Vliet, 1978).   However, such excessive bacterial growth is
 dependent upon the amount of soluble, biodegradable organics (food) present,
 and raw drinking water supplies simply do not contain this much DOC.  Even
 physical-chemical treated sewage containing as much as 25 mg/1  of BOD5 does
 not cause biological blinding of the BAC columns (Guirguis et_al., 1978),
 provided that backwashing is conducted in the proper manner anBat proper
 intervals.

     Section 9 includes a discussion of research studies of Eberhardt,
Madsen & Sontheimer (1974)  conducted at the Bremen (Federal  Republic of
 Germany) water treatment plant which showed that GAC which had a higher
fraction of macropores (2%) than did normal GAC (1%) removed higher amounts
of DOC (about 50% as opposed to 20 to 25%) after attaining biological equili-
brium.  Macropores are capable of housing bacteria, therefore the Improved
removal  of DOC was explained by the higher fraction of macropore area avail-
able to house more bacteria.   These results suggest that if GAC can be
prepared with even higher macropore contents, even higher COC removals might
be attainable with BAC systems.

     However, Tzesos & Benedek (1980) compared the performance of two types
of GAC (Filtrasorb 400 and  Special A) treating phenol and operated under
 biologically enhanced, fluidized bed conditions.   After establishing biologi-
cal equilibrium, the TOC concentration 1n the effluents from the 2 GAC
 coluirns was about the same.  Thus 1t was concluded that Special A GAC (which
 has a surface area about 10% of that of Filtrasorb 400, produced by transitio-
 nal pores near 70 A in size)  adsorbs similarly to Filtrasorb 400, once
 corrections for surface area differences are made.   This appears to indicate
 no special  benefit of larger pores.  Hence, Tzesos & Benedek (1980) also
 concluded that the biologically generated residual  organic molecules (from
microbial metabolism of phenol) are likely to be smaller than 70 A in molecu-
 lar size.

     Studies by Werner, Klotz & Schweisfurth (1978) have been conducted on
 biologically active GAC columns used for more than three years at the
 Schierstein water treatment plant at Wiesbaden, Federal Republic of Germany.
 Microbiological species listed in Tables 3 and 4 have been extracted and
 identified from these GAC materials.  All of the species identified to date
 are non-pathogenic 1n nature, and are found naturally occurring in soils and
 waters.
                                     51

-------
TABLE 3.  BACTERIAL SPECIES FOUND IN GAC ADSORBER EFFLUENTS  AT  THE  SCHIERSTEIN
  	(FEDERAL REPUBLIC OF GERMANY) WATER TREATMENT PLANT	
     Psuedomonas alcallgenes            Chromobacterlum vlolaceum
     Pseudomonas cepacla                Ne1sser1a slcca
     Pseudomonas facllls                Aclnetobacter calcoacetlcum
     Pseudomonas fluorescens            Micrococcus  luteus
     Pseudomonas lemolgnel              Staphylococcus saprophyticus
     Pseudomonas mendodna              Bacillus  cereus
     Pseudomonas ruhland11              Bacillus  clrculans
     Pseudomonas stutzerl               Bacillus  Hchenlformls
     Pseudomonas spec.                  Bacillus  megaterlum
     Gluconobacter oxldans              Bacillus  pumulls
     Azomonas agH1s                    Bacillus  thur1ngens1s
     Azomonas Inslgnis                  Corynebacterlurn spec.
     Azomonas macrocytogenes            Mlcromonospora spec.

          Source;  Werner, Klotz & Schwelsfurth,  1978
TABLE 4.  TYPES OF FUNGI AND YEASTS FOUND IN GAC ADSORBER  EFFLUENTS AT  THE
  	SCHIERSTEIN (FEDERAL REPUBLIC OF GERMANY)  WATER TREATMENT  PLANT
     Filamentous fungi             Phlalophora  hoffmann11
                                   PhUlophora  mutabllls
                                   Taphrlna spec.

          Yeasts                   Rhodotorula  mlnuta var.  texensls
                                   Cryptococcus uniguttulatus
                                   Candida gulllermondll var.  gulllemiondli
                                   Hansenula anomala  var.  anomala

          Source:  Werner, Klotz & Schwelsfurth, 1978
                                      52

-------
     Bacteria listed 1n Table 5 are those which have been Isolated and
 Identified from the GAC pilot plant columns or from their effluents operating
 at Philadelphia's Torresdale drinking water treatment plant during initial
 studies by Cairo, McElhaney & Suffet (1979).  The API system and other
 biochemical tests for enterobacteriaceae and other gram-negative bacteria
were used.  The most predominant bacteria that were isolated and identified
 belonged to the genus Pseudomonas.  Members of the genus Bacillus also were
 found in large numbers"!  Bacteria that were identified in this study were
 predominantly so.il or water saprophytic chemo-organotrophs and have no known
 pathogenic significance to man.  However, some species of Flayobacterium,
 Pseudomonas and Alcallgenes are known to be opportunistic pathogens, chlorine
resistant, or suppressors of total coliform.

     Table 6 (McElhaney et al., 1980) lists results of further bacterial
studies conducted by the~PhTTadelphia scientists.   One objective of this
later study was to determine the effects of varying linear flow velocity  of
water being treated through GAC columns on the distribution of microbial
species in GAC media and in their effluents.  Two GAC columns were studied,
one operated at 55.6 ml/min and the second at 166.7 ml/m1n.   Both columns
were designed so as to have empty bed contact times of 9 minutes.  In the
column treated at the slower flow rate, the predominant bacteria in the
initial stages of the microbial community were Pseudononas cepacla and two
species of Acinetobacter.   As the community became more established, the
Acinetobacter group was replaced by Bacillus brevis, however Pseudomonas
cepacia remained stable throughout the study.

     In the column operated at the faster flow rate, the microbial community
was dominated from the beginning by members of the genus Flavobacterium and
Ps. cepacia.   As the community developed, a member of the family Actinomycetes
aTso became established.  Others less common, bacteria also appeared and
there was a shift in the composition of the communities with depth of GAC
column.

     Although the species composition was different in both columns, the
composition of bacteria 1n the effluents of both columns was similar, being
composed of Flavobacterium and Bacillus brevis throughout the 7C day study,
with Ps_. cepacia and Actinomycetes developing during the last 20 and 30
days, respectively.  There was no difference found in the bacterial densities
either on the GAC or in the effluents.   In addition, there was no difference
 in TOC removal at the 2 different flow rates studied.

     Even though high levels of microbiological species exist in the biologi-
cally active GAC media and even though high levels can pass Into the GAC
effluents, only small amounts of chlorine or chlorine dioxide are used in
 European drinking water treatment plants following BAC treatment to provide
satisfactory disinfection.  Maximum dosage levels of 0.3 to 0.6 mg/1 of
chlorine are prescribed by German drinking water laws (Sontheimer, 1977).
At the French Rouen water treatment plant, 0.5 to 0.6 mg/1 of chlorine
 follows post-ozonation, which, itself follows BAC passage (Gomella & Versanne,
 1977).
                                      53

-------
TABLE 5.  BACTERIA ISOLATED FROM GAC ADSORBERS AT PHILADELPHIA. PA
Identification
Ps. putida

Ps. fluorescens
Ps. mal tophi lia

Ps. pseudoal call genes
Pseudomonas spp. other

Pseudomonas spp. other

Pseudomonas spp other

Pseudomonas spp other

Pseudomonas spp other

Pseudomonas spp other

Pseudomonas spp other

Achromobacter spp
Al call genes odorans
Alcaligenes spp
Flavobacterium odoratum
CDC IV E
CDC II F

API Code
2 200 004 41
2 201 004 41
2 005 006
1 000 000 41
1 202 000 51
0 000 004 50
2 000 004 50

0 000 004 50

0 000 OC4 51

0 002 004 41

0 002 004 40

0 001 004 50

2 200 004 51

1 010 004 41
0 001 004 50
0 000 004 40
0 002 004 01
0 010 004 01
0 202 004 00
0 242 004 00
Comment
acceptable
acceptable
excellent
acceptable
very good
best fit
good likelihood but
low selectivity
good likelihood but
low selectivity
good likelihood but
low selectivity
good likelihood but
low selectivity
good likelihood but
low selectivity
good likelihood but
low selectivity
good likelihood but
low selectivity
best fit
best fit
best fit
best fit
best fit
best fit
excellent
Source; Cairo e_t ajL , 1979
                                      54

-------
TABLE 6.  BACTERIA ISOLATED FROM GAC COLUMNS AND EFFLUENTS IN PHILADELPHIA
Identification
Ps. cepacia
Ps. acldovorans
Ps. paucimobiUs
Ps. pltuda
Adnetobacter - 3 species
Azotobacter
Actlnomycetes - 2 species
Flavobacterlum - 3 species
Cytophaga
II B
Al call genes odorans
Alcallgenes denltrlf leans
Alcaligenes SP. other
Bacillus brevls

Relative Abundance
common
common
occasional
common
common
common
common
common
rare
common
common
occasional
common
common
Source; McElhaney et_ al_. , 1980

     It is significant that BAC processing never is the final  step in
European drinking water treatment plants.  A small amount (0.3 to 0.6 mg/1)
of chlorine or chlorine dioxide always is added to disinfect the GAC effluents
and to maintain an adequate residual in the distribution system.

     McElhaney &. McKeon (1978) reported on studies of the biology 1n GAC
media at the Torresdale Filter Plant of the City of Philadelphia.  Column
influents (after passage through sand filters) had an average pH of 8.6 and
a free chlorine concentration of 2 mg/1.  There was no preoxidatlon (other
than prechlorination) or preaeration of column Influent waters.   Bacterial
levels in the GAC column effluents increased to 103 to 104 bacteria/ml
within the first month of continuous use, then decreased and remained at
slightly lower levels for the balance of the study.  Bacterial levels 1n the
GAC effluents at all times were significantly higher than those  of the sand
filter effluents.  The most predominant species of bacteria identified
belonged to the genus Pseudomonas.  The following bacteria were  isolated and
                                      55

-------
identified in addition to those isolated and identified by Werner, Klotz &
Schweisfurth, (1979):  Ps_. pituda. Ps. maltophila, Achromobacter sp.. Alcall-
genes odorans, Alcaligenes sp. and FTavobacterium odoratum, CDC Iv E and CDC
trrr

     Latoszek & Benedek (1979) isolated and identified bacterial species
present in biologically active GAC columns used for treating coagulated,
settled and sand-filtered domestic wastewater.  Extensive growths of bacteria,
up to 109 viable cells/g of wet drained GAC, were detected.  The majority of
isolates were classified as belonging to the genus Pseudomonas and to the
Flayobacterium cytophaga group.  A high percentage of the bacteria exhibited
denitrifying ability.  These authors concluded that the microblal life in
activated carbon adsorbers used to treat wastewaters appears to be similar
in nature and numbers to that found 1n activated sludge mixed Hquor.

     Brewer & Carmichael (1979) also studied and characterized microblal
populations of GAC adsorber systems used to treat drinking water following
ozonation.  GAC samples were taken from two 6-foot deep columns operated at
the U.S. Environmental Protection Agency laboratories in Cincinnati, Ohio,
which were treating Ohio River water.  One column received ozonated water
during the six week study, and the second column received untreated water.
In most cases,  only about 10% of the bacteria present in the influent was
recovered in the effluents.   Counts in the effluents ranged from 1.10 x 102
to 1.30 x 103,  with a mean of 4.4 x 102 colony-forming units (CFU)/ml.
Fungi did not appear in the effluents until the third week, after which
approximately 80% of the fungi was recovered 1n the effluents.

     Counts of enteric bacteria in the influents ranged from 50 to 1,000
CFU/mL, but only about 25% was recovered in the effluents.  Col 1form bacteria
in effluents were detected or.iy during the first week of sampling, the
number of colonies being similar to those present in the influents.  No
fecal coliforms were detected 1n the ozonated waters.  Pseusomonads were
detected in all  influent samples,  but only in one effluent sample.  Table 7
shows the results of bacterial characterization of gram-negative populations
determined by Brewer & Carmichael  (1979).   Enterobacter agglomerans was the
most prevalent bacterium isolated  (54.3% of the total Isolates).Species of
the genus Proteus also were prominent (P.  mlrabHis, 29% and £. yulgaris.
5%).  No enteric pathogens were Isolated",  and E. coll represented only 2.6%
of all  Isolates.  These Investigators noted a shift In population, however,
during the first 3 weeks of the study.  Initially, Proteus was the significant
genus isolated,  but Enterobacter dominated all  plates during the last 3
weeks.

     Brewer & Carmichael (1979) also estimated the potential endotoxin
concentrations present in the GAC column effluents by correlating with
enteric bacterial populations using the Limulus lysate method of Evans,
Schillinger & Stuart (1978).   Endotoxin concentrations averaged 0.011 and
0.006 mg/1 in the effluents from the ozonated and untreated samples, respec-
tively.  These averages represent 28% and 13% of endotoxin recovery 1n the
effluent from each water sample, respectively.   These authors cautioned that
these data indicate the retention  of endotoxins by the GAC adsorbers, and
                                                                     v

                                     56

-------
that subsequent release of such material into water passing through the
adsorbers might produce endotoxln levels sufficiently high to cause illness
in human beings.

     TABLE 7.  ENTERIC BACTERIA ISOLATED FROM GAC EFFLUENTS
Organism Isolated
Enterobacter agglomerans
Enterobacter cloacae
Proteus mirabilis
Proteus vulgaris
Escherichia coli
Citrobacter freundii
Serretia liquifaciens
Unknown
Percentage of
Total Isolated
54.3%
0.9
29.0
5.2
2.6
1.7
1.7
4.4
(Brewer & Carmichael, 1979)

     On the other hand, Wolf, Camp & Hawkins (1979) studied the pyrogenic
activity of GAC-filtered water supplies and wastewaters treated by advanced
waste treatment techniques, including GAC containing aerobic microbial
activity.  One of the main public health concerns about biological growth in
GAC adsorbers, particularly for preparing potable water supplies, is the
potential for generating unwanted microorganisms and/or their toxic by-
products.  Among the toxic by-products to be aware of are the gram-negative
endotoxins, which are lipopolysaccharides which are all thought to cause a
pyrogenic response (fever) when injected into animals.  The rabbit is used
in the standard United States Pharrracopeia (USP) pyrogen test, but man
reportedly can be ICO-fold more sensitive to endotoxins than is the rabbit.

     Major conclusions drawn by Wolf e_t al_. (1979) are as follows:

(1)  The numbers of bacteria found to be present in GAC effluent waters from
     water treatment plants in which the GAC had been operated as long as
     110 months were "very few and certainly would constitute no public
     health concern as long as a disinfection process is subsequently applied"

(2)  All four disinfection processes employed subsequent to GAC treatment
     (ultraviolet, chlorination, high pH and ozonation) appeared to decrease
     endotoxin content.
                                     57

-------
 (3)  Free endotoxin appears to be not adsorbed by GAC.  Bound endotoxin
     appears to be removed via a filtration mechanism.  Hence, total endotoxin
     is removed during passage through GAC only by the amount of bound
     endotoxin filtered out.

 (4)  Using 2 GAC columns in series with a filtered, nitrified activated
     sludge effluent, the first column removed 20% of the COD and total,
     bound and free endotoxlns by 64%, 77% and 41%, respectively.  Additional
     removals in COD, total and bound endotoxins of 35%, 20% and 42%, respec-
     tively, were observed in the second GAC column (which had a longer
     empty bed contact time — 45 minutes versus 30 minutes), but free
     endotoxin content increased by 12%.

 (5)  Pyrogenicity (in rabbits) was demonstrated for advanced wastewater
     treatment (AWT) effluents but not for GAC-processed drinking waters.
     Although endotoxin levels generally were lower for drinking waters,
     irany samples possessed higher endotoxin activity than some of the AWT
     samples which resulted in pyrogenlc responses.  This suggests some
     basic differences in water quality, which must be studied further.

     Finally, Scott et al.  (1975) have pointed out that the biological
degradation of phenoTTc^Type compounds has been studied extensively.
Various types of Pseudomonas bacteria (which have been identified in biologi-
cally operating 3AC media)  appear to be best adapted for degradation of
phenol.  An example of such a metabolic degradation pathway was determined
for Pseudomonas pituda by Sala-Trepot, Murray & Williams (1972), and is
reproduced in Figure 17.  Notice that this pathway involves Initial  addition
of oxygen to the aromatic ring, producing catechol, after which ring rupture
occurs, followed by further oxidative degradation to produce acetaldehyde
and pyruvic acid.  Thpse are three of the same products which have been
identified as oxidation products isolated upon ozone oxidation of phenol
 (Rice & Miller, 1977).

     Heilker (1979) has presented a detailed description of the BAC process
which has been operating on full scale at the Donne drinking water treatment
plant at MUlheim, Federal Republic of Germany.  Discussion of this plant and
the process employed is presented in Section 9 and in  Appendix E.    One
observation made during the summer of 1978 will be pointed out here.  When
the process first was installed, 1t was found that backwashlng frequency of
the rapid sand and GAC media could be decreased to periods greater than
every 7 days.  During the summer of 1978 "a population explosion of nematodes11
was observed in both of these media.  Because the reproduction cycle of
these nematodes is 3 to 4 days, simply increasing the backwashing frequency
to 3 days has eliminated this problem.

     On the basis of studies conducted to date, McElhaney et_ al_. (1980)
conclude that when GAC adsorbers are first installed or when vTrgin or
regenerated GAC is first utilized, microblal populations initially increase
exponentially for the first 2 to 3 weeks, then enter a phase of negative
growth acceleration where populations level off and finally develop into a
pseudo-steady state.  Populations in the top levels of the GAC column reached
                                     58

-------
                 g
approximately 10  bacteria/g of wet GAC 1n the pseudo-steady state.   This
growth curve  1s  similar to the sigmoldal type of growth  curve which  Is
typical  of most  species during their initial stages of growth, except for
the absence of an observed Initial lag phase.
                     ?H hydroxy,-  P OH
                         _a»eL_ r^V°?	  f^^COOH

                                      'VST  k^°  2hydTy
                                              -V^-T    muconlc
                    phenol       eatechol               aemlaldehyde
                                               I hydrolase
                 4-hydroxy»2-    ,
                 ketopentanolc  HO
                    •cld       ^..  ,           2 oxopent 4 enole
                             ^xiidola.e             .e|d
                         *r
                  CH3-C-COOH   pyruvlc  acid
                      O

                    CH3CHO     aeetcldehyde

                  Figure 17  Simplified metibollc  pathway used by
                           P»uedomona» putlda to degrade phenol
                           (Sala Trepot,  Murray & Williams,  1972)

     The predictability  of microbial  growth curves  in  a GAC  system is
important when evaluating  the effect  of environmental  factors,  such as
linear velocity, on  that growth.   McElhaney et_al_.  (1980)  and  Klotz et aJL
(1976) showed that bacterial  densities decreased with  increasing  depth~1n
the GAC column,  which  might be attributed to a decrease in food supply.
McElhaney e_t al_. (1980)  showed that TOC levels in the  GAC also  decrease with
increasing depth.  TOC levels are a good indicator  of  the mlcrobial  food
supply because carbon  is the major source of energy for all  of  the bacteria
identified in their  study.

     With these  few  fundamental  principles regarding microbiological and
biochemical oxidations 1n  hand,  we can now address  the basic parameters of
BAC systems (see Section 8)  so as to  combine the best  features  of three
treatment processes  (chemical oxidation, GAC adsorption and  biological
oxidation).  Successful  design of BAC systems appears  to provide  treatment
systems which can be more  efficient in removing organic pollutants than
either of the three  processes taken individually.   In  addition, savings in
the operational  costs  of GAC adsorption systems can be effected,  under
certain conditions,  by Incorporating  chemical preoxidatlon of the water or
wastewater to be treated and by promoting aerobic mlcrobial  activity in the
filter and GAC adsorption  media.
                                       59

-------
                                 SECTION 7

                         GRANULAR ACTIVATED CARBON
INTRODUCTION
     Granular activated carbon (GAC) has a long history of successful
applications in the purification of liquids and gases.   This section will
address the applications of granular activated carbon to the treatment of
drinking water.  Mechanisms of pollutant removal from water by GAC treatment,
historical GAC applications and modern GAC applications in European drinking
water treatment plants will be discussed.

     The first application of GAC for drinking water treatment was described
by Batchelor (1977) to have occurred in 1910 at Reading, England, where
drinking water was superchlorinated to oxidize organic iraterlals and then
was dechlorinated by passage through activated carbon.   Subsequent applica-
tions of GAC to municipal water treatment were at Hamm, Germany in 1929 and
at Bay City, Michigan in 1930 (Hasler, 1974).  Economic analyses, however,
indicated that powdered activated carbon (PAC)  was more cost-effective at
the dosages normally found necessary at that time.  The first municipal
application of PAC was at the Hackensack Water Company at New Mil ford, New
Jersey (Kornegay, 1978).  Subsequently, nore than 1,000 water treatment
plants have adopted the use of powdered activated carbon for taste and odor
control.

     Deteriorating raw water quality caused ever-increasing dosages of PAC
and this renewed interest in the use of granular activated carbon for taste
and odor control.  An organization which has made major investments in the
application of GAC for this purpose commencing at Hopewell, Virginia in 1961
is the American Water Works Service Company (Blanck, 1978).     ,

     Ten filters of the Nitro, West Virginia water treatment plant were
converted to 3AC in 1965 and were operated successfully until 1973, at which
time the plant was abandoned in favor of a new facility using a different,
higher quality raw water supply which did not require the use of GAC.   A
multiple hearth furnace was used at the original Nitro plant for GAC regenera-
tion.  The Nitro plant was the subject of considerable study during its
operational life (Woodward, et. ^1_., 1964; Dostal e_t al., 1965; Flentje &
Hager, 1964).

     Based on successful applications at the Hopewell and Nitro water
treatment plants, the American Water Works Service Company has converted
sand filters to filter/adsorbers by replacing a portion of the sand filter
media with GAC at other of the company-owned water treatment plants (Blanck,

                                      60

-------
1978).  They currently have 20 plants operating in this  mode (Moser,  1978).
This type of installation is illustrated by Figure 18.   Overall,  there  are
more than 60 municipal water treatment plants in the United States  currently
using GAC, primarily in the sand replacement mode (Froelich, 1978).   Table 8
summarizes important milestones attained with carbon adsorption technology.

               TABLE 8.  ADSORPTION TECHNOLOGY MILESTONES
   Year

 1550 B.C.   Early recorded use of charcoal
 1811 A.D.   Bone char used for sugar processing
 1828        First char regeneration instituted
 1852        First granular charcoal filter,  Elizabeth,  N.J.
 1889        Hershoff multiple hearth furnace introduced
 1906        First commercial production of activated carbon  (Eponite,
              Europe)
 1910        First application of 3AC in drinking water  treatment (Reading,
              England)
 1913        First commercial production of activated carbon  in the U.S.
 1928        First use of powdered activated carbon for  taste and odor
              control (Chicago meat packers)
 1929        First GAC filter installed (Hamm, Genrany)
 1930        First municipal use of powdered activated carbon for taste and
              odor control (Hackensack Water Company)
 1961        GAC filters installed at Hopewell, VA water treatment plant
 1965        First advanced wastewater treatment plant incorporating GAC
              (South Lake Tahoe, California)
 1978        First fluidized bed GAC regeneration furnace installed in the
              U.S.

     Source;  Kornegay, 1978
     Granular activated carbon pressure post-adsorbers were installed in a
nurrber of water works in the lower Rhine River area of the Federal  Republic
of Germany in the 1950s and early 1960s.  These units incorporated  both
pressure filtration and pressure adsorption in a single vertical  steel
vessel, as illustrated in Figure 19.  The upper compartment utilizes an
inert medium for the removal of turbidity as well as oxidized iron  and
manganese compounds.  The lower compartment utilizes GAC as a post-adsorption
irediun.  The pressure units were installed originally for taste and odor
removal as well as dechlorination, but now serve as biological reactors and
adsorbers due to the elimination of prechlorination and addition of a preozo-
nation step ahead of filtration.

     Pressurized post-adsorber 3AC units have been in operation in  the
Foxcote Treatment Works of the Bucks Water Board of England since 1960.  GAC
was installed at this plant for taste and odor control.
                                     61

-------
                                       backwash troughs
                                     -* GAC
                                        sand
                                     ^ filter bottom
    Figure 18.   Filter/adsorber installation.
                BACK      FILTER IN
               WASHING    OPERATION
                                  ___ACTIVATED CARBON/
                                        GRAVEL

                                  — NOZZLES
                                   GRANULAR CARBON/
                                        GRAVEL

                                   NOZZLES
                                    OZONIZED
                                     WATER
                               TREATED WATER
            WASH
           WATER
           CITY OF DUSSELDORF, GERMANY


Figure 19.  Dual function  filter and adsorber.

            (Source:   Hagar & Fulker, 1968)
                         62

-------
GRANULAR ACTIVATED CARBON

     Charcoal long has been applied for the purification of liquids such as
drinking water.  The "activation" of charcoal  during manufacture or treatment
to enhance its adsorptive power has broadened  its applications in the purifi-
cation of liquids and gases.   Activated carbon 1s manufactured from a number
of carbonaceous materials which include the following:

                    Bituminous coal
                    Coconut shells
                    Lignite
                    Wood
                    Pulp mill residues

     Activation is carried out by treatment of the carbon to achieve a high
degree of porosity and an associated high surface area.  The high surface
area renders activated carbon an excellent adsorbent for a wide range of
substances from both liquid and gaseous phases.  Activation of carbon is
accomplished by one of two processes, gas activation and chemical activation.

     Manufacture of activated carbon by gas activation 1s essentially a two-
step process, carbonization and activation.  Carbonization usually is achieved
by burning of the carbonaceous raw material In the absence of oxygen, a
process known as pyrolysis.  The carbonized intermediate product is activated
by being brought into contact with stean or carbon dioxide to *orm the
necessary porous structure.

     Manufacture of activated carbon by chemical  activation is a process in
which uncarbonized carbonaceous materials are  mixed with chemicals which are
capable of dehydrating and oxidizing.  The chemically treated material then
is heated to create the necessary pore structure.  The activating agents,
such as zinc chloride, phosphoric acid, sulfuric acid, etc., are washed out
and recovered.

     In conventional practice, activated carbon is used as an adsorbent.
Its capacity for adsorption is governed by:

(a)  the adsorbent surface area actually available to the adsorbate molecules
     and

(b)  the affinity of the adsorbate molecules for the available adsorbent
     surface areas.

     The following Information is  important, therefore, when evaluating
activated carbon as an adsorbent for contaminants present in a liquid
phase:

          a.   Surface area
          b.   Pore volume distribution
          c.   Chemical nature of  the activated carbon surface area
          d.   Chemical structure of the adsorbate molecule(s)
                                     63

-------
     Activated carbons in common use have total  surface  areas  (sum of the
outside surface area plus available internal  surface  area  inside  the  pores)
ranging from 450 to 1,500 square meters per gram.   The  total  surface  area
normally is determined by the Brunauer, Emmett & Teller (BET)  nethod, in
which the total volume of nitrogen gas adsorbed at liquid  nitrogen temperature
(minus  195°C) is measured at various pressures.  The plot of volure  of
adsorbed nitrogen versus pressure at constant temperature  is  called an
adsorption isotherm, and this enables calculation of  the volume (and  hence
the surface area) of a layer of nitrogen one molecule thick.   However, the
molecules encountered in liquid phase operations usually are  nuch larger
than the nitrogen molecule.  As a result, the actual  area  available for
adsorption normally will be considerably less than the  total  surface  area  as
measured by the BET method.

     Another measure of activated carbon capacity is  pore  size distribution
within the carbon particles.  One method for measuring  pore size  distri-
bution is by using a range of adsorbate molecules of  varying  size as  molecular
probes (Calgon Corporation literature, undated).  Adsorbate probes include
nercury, nitrogen and water.  Edwards (1979) points out that pore volume
distributions may be a function of both raw material  (Figure 20)  and  activa-
tion conditions  (Figure 21).
          o
          o
          o
          i
          c:
          O
         LLJ
         CO
         ^
          >S
         «j°

         Ib
         o
         u.
         al
0.14


0.12


0.10


0.08


0.06


0.04


0.02


0.00
                              i
     CARBONS OF HIGH, AVERAGE AND
  3   LOW ACTIVITY FROM SAME RAW
   1  MATERIAL BY THERMAL OXIDATION
                                   /  -—HIGH ACTIVITY
10       100       1000     10,000
  r = PCRE RADIUS  N ANGSTROMS
                                                                100,000
        Figure 20.  Pore volume distributions of typical decolorizing
                    activated carbons.

                         (Edwards, 1979)
                                      64

-------
           O  0-14
           3

           2  0.12

           <
           g  0.10
            CO
          to
              0.08
              0.06
              0.04
          2b
              0.02
            o
        CARBONS FROM DIFFERENT
             RAW MATERIALS
10       100       1000      10,000
   r = PORE RADIUS IN ANGSTROMS
                                                                100,000
         Figure 21.   Pore volume distributions of typical  decolorizing
                     activated carbons.

                               (Edwards,  1979)

     The nomenclature used in designating pore volume size ranges is not
universal.  United States practice appears to refer to pore sizes larger
than 1,000 angstroms (A) diameter as "macropores" and pore sizes from 1C to
1,000 A in diameter as "micropores" (Weber, 1972; Calgon Corporation litera-
ture, undated).  Reference to European literature, however, indicates that
such categories are by no means universal.  Juntgen (1975) uses the following
classification:  macropores >_ 500 A diameter, mesopores <_ 5CO to 20 A in
diameter, micropores < 20 to 8 A diameter and submicropores < 8 A in diameter.
A Lurgi publication (T976) uses the following definitions:  iracropores >_ 2CO
A diameter, transitional pores < 200 A and > 20 A diameter, micropores £ 20
A diameter.  This point is  emphasized to ensure care in transposing European
experiences to the United States.  Pore size distribution is particularly
important in considering the use of biological activated carbon (BAC),
discussed in Section 8.

     The nature of the carbon surface is important in determining the
capacity of activated carbons for a specific adsorbate molecule.  Greater
knowledge is developing in this area as the use of activated carbon for
water treatment increases.  The necessity of removing trace quantities of
toxic organic compounds demonstrates that need.

     The nature of the adsorbate affects the efficiency of adsorption.  Some
adsorbates are strongly adsorbed while others are weakly adsorbed.  Liquid
                                      65

-------
streams containing a number of chemical  compounds further confuse  the issue
because of the chemisorption phenomenon, which is discussed in a following
subsection.

SELECTION OF ACTIVATED CARBONS

     Selection of the type of activated carbon to be used depends  upon a
number of factors.  Powdered activated carbon (PAC) is used predominately
for routine applications in the United States nainly on the basis  of economics
(AWWA, 1971).  The use of both PAC and 3AC is growing in the European
countries (Water Research Center, 1977)  for the following reasons:

"(a) It is more economic to use a granular carbon bed if large doses (720
     mg/1) of powdered carbon are normally required, and

(b)  a granular carbon filter is held to be more effective in the  removal of
     micropollutants."  (Micropollutants are generally understood  to be
     those soluble materials which cannot be reiroved by simple filtration.)

     The remainder of this discussion will be limited to GAC since the
evaluation of Biological Activated Carbon (BAC) for drinking water purposes
has been limited to this form of activated carbon.  Primary properties to be
considered in SAC selection include the following:

                    Capacity
                    Hardness
                    Permeability
                    Solubility

     The capacity of GAC used for a specific water treatment objective can
be determined by pilot plant evaluations.  Standardized equipment and pro-
cedures are available, but some insight is required to determine periods of
study and the number of studies required for a specific water source.  The
effects of biological activity and desorption further complicate the situation.

     The hardness of the GAC is a factor in determining the amount of
carbon loss and associated cost of nakeup GAC per each adsorption-regeneration
cycle.  Other factors which influence GAC losses during use are handling and
burnoff losses during the reactivation process.  It is not clear which of
these factors is of greatest significance, but since losses can range from
3.5 to 10% or more, they are of obvious importance.

     Permeability of the activated carbon is of obvious importance in terms
of head loss in downflow contactors and of bed expansion in upflow units.
Permeability is a function both of particle size and gradation.  The smallest
particle size of GAC normally would be expected to be specified so as to
maximize the available total surface area.  This intent would be tempered by
the available hydraulic head or the sensitivity to plugging due to biological
activity in the contactor.  Solubility of adsorbate in water is of signifi-
cance also, since the more polar the nolecule is, the greater is its solubi-
lity in water and the less readily removed from water it is by activated
carbon adsorption.  Surface characteristics (outer surface and inner pore

                                      66

-------
surfaces) also affect adsorption of water soluble molecules, and these
characteristics depend upon the raw material used to produce the carbon, the
activation process and the solutes to be removed from aqueous solution.

OPERATIVE MECHANISMS DURING DRINKING WATER TREATMENT WITH GAC

     Successful use of granular activated carbon 1n the treatment of drinking
water results from a number of mechanisms and processes, including adsorption,
filtration and biological activity.  These factors are discussed in the
following subsections.

Adsorption

     This is the most commonly recognized function of GAC.  It consists of
removal of solutes from solution and their concentration on the surfaces
(external and internal) of the adsorbent.  The three types of adsorption are
(Weber, 1974):

(1)  exchange adsorption (electrical attraction of the solute to the adsorbent)

(2)  physical or ideal attraction (Involving weak van der Waals forces)

(3)  Chemisorption or chemical adsorption (chemical reaction of the adsorbate
     with the carbon surface).

In water treatment by means of GAC, the primary mechanisms are physical
adsorption (which 1s reversible) and chemisorption, which is generally
considered to be irreversible.

     Physical adsorption is a function of a number of factors, including the
following (Edwards, 1979):

(1)  solubility of the adsorbate (pollutant)

(2)  affinity of adsorbate for the adsorbent (GAC) surface

(3)  properties of the adsorbent (GAC).

     An adsorbate of low solubility in water will be nore adsorbable by GAC
and vice-versa.  Chemical compounds vary in their affinity for GAC surfaces
due to a number of factors, including the presence in the adsorbate molecules
of specific  functional groups as well as the molecular shapes and structures
of the molecules themselves.  Table 9 lists a number of organic compounds
and their relative adsorbabilities onto GAC.  Additional information on the
adsorption of organic chemical compounds by GAC is provided in Section 8.
The primary  desired 3AC properties for use  in water treatment are high
available surface area and pore size distribution, which were described in
an earlier subsection.

     Chenisorption involves chemical reactions of  the solute with the  GAC
surface.  This can occur, for example, by carbonyl or hydroxyl groups
(present on  the activated carbon surface) condensing with amine, hydroxyl or

                                      67

-------
activated hydrogen groups in the adsorbate to form very strong covalent
bonds, as indicated in Figure 22.  Restoration of the GAC surface with
respect to chemisorbed solutes results in the physical  destruction of the
carbon surface involved in the chemical  reaction.  Because the amount of
chemisorption involved with removal of organic pollutants generally is very
small compared with physical adsorption, chemisorption  will  not be discussed
further in this section.

          TABLE 9.  GAC ADSORPTION OF ORGANICS
          Strongly Adsorbed

 • Aromatic Solvents: benzene, toluene,
    nitrobenzenes, etc.

 • Chlorinated Aromatics: PCBs, chloro-
    benzene, chloronaphthalene

 • Phenol & Chlorophenols

 • Polynuclear Aromatics: acenaphthene,
    benzopyrenes, etc.

 • Pesticides & Herbicides: DDT, Aldrin,
    Chlordane, BHCs, Heptachlor, etc.

 9 Chlorinated Non-Aroma tics: CCI^
    chloroalkylethers, heaxchlorobuta-
    dier.e, etc.

 • High Molecular Weight Hydrocarbons:
    dyes, gasoline, amines, hunvic acids
   Weakly Adsorbed

   alcohols
• low molecular weight
   ketones, acids & aldehydes

• sugars and starches

t very high molecular weight
   or colloidal organlcs

• low rolecular weight
   aliphatics
     A major concern with physical adsorption is competitive adsorption
between two or more solutes present in the solution being purified.  A
weakly adsorbed pollutant will be displaced from the activated carbon
surface by a more strongly adsorbed pollutant.  The displaced, weakly
adsorbed material then will be readsorbed further down in the activated
carbon medium.  This repeated adsorption, desorption and readsorption
seqeuer.ce is known as "the chromatographic effect".

     Desorption of an adsorbed pollutant also may occur if the concentration
of the pollutant in the liquid feed to the GAC medium becomes significantly
lower and the adsorbed pollutant 1s not strongly adsorbed.  This behavior is
of particular concern when removal of certain toxic organic compounds, such
as chloroform, is required.

     Each adsorbable dissolved organic compound is held on the GAC surface
or in its pores usually by physical forces, the strengths of which depend
upon a nutrber of factors related to the irolecular structure and polarity of
the adsorbed molecule.  If the GAC is loaded with a weakly adsorbed compound

                                      68

-------
and the influent solution now is spiked with  a  second  dissolved organic
compound which is nore strongly adsorbed  than the  first,  then the first
compound will be displaced, or desorbed,  from the  surface and/or the pores
by the second, which now becomes the adsorbed solute.   For example,  phenol
is readily adsorbed by GAC, but is readily  desorbed  by p-chlorophenol.
                       granular activated carbon surface
                                   I
                 Figure 22, Exairples of crerlsorptlon on 3AC surfaces.
     If the volume of  influent  solution  is  adjusted so that phenol  is
adsorbed only in the initial few  centineters  of the GAC column and then
a dilute solution of p-chlorophenol  is passed through the column, desorption
of the phenol will occur.   In the next lower  section of the GAC, which has
not yet been exposed to  p-chlorophenol,  the phenol  desorbed from the higher
levels will be readsorbed.  When  ir.ore p-chlorophenol reaches the second
layer of GAC containing  adsorbed  phenol,  desorption again occurs.  This
process of phenol adsorption, followed by desorption, continues until
finally, phenol is removed  completely from the GAC  and is eluted into the
GAC effluent.

     This chromatographic effect  has been noted recently at the Kralingen
plant in Rotterdam, The  Netherlands, with respect to adsorption of chloroform
by GAC (Rook, 1978).   The Kralingen  plant began operating a process including
ozonation, sand filtration  and  granular  activated carbon adsorption in early
1977.  Raw water is taken from  two reservoirs located 20 kilometers away
from the plant and through  which  water flows  in series to the plant.  One of
these reservoirs is aerated to  control algae, but this practice also promotes
                                       69

-------
aerobic biodegradation of organlcs and nitrification of ammonia during the
three months average residence time of the water in these reservoirs.

     During summer, biological growth occurs in the 20 km long raw water
transport pipe from the reservoir to the KraHngen plant, and low level
prechlorination (0.5 to 2.0 mg/1  dosage — called "transport chlorination")
then is employed to eliminate this growth.  Use of prechlorination produces
chloroform and other trihalomethanes, which are not removed during the
subsequent coagulation, flocculation, sedimentation or ozonatlon processlrg
steps, but are adsorbed by the granular activated carbon.

     On the other hand, chloroform appears rapidly in the effluents of the
Kralingen plant biologically active GAC columns, and its concentration in
the effluent approaches that in the influent after about 10,000 to 12.CCO
bed volumes of water have been passed through,  indicating that the GAC is
not very efficient for the adsorption of chloroform.

     When prechlorination is halted, however, chloroform continues to  be
found 1n the GAC column effluents, and its concentration decreases continually
during the next few months.  No  "transport chlorination" is necessary
during winter months at Kralingen, yet chloroform continues to be eluted
from the GAC, probably by desorption caused by more strongly adsorbed  organic
compounds.

Filtration

     Water filtration incorporates both physical and chemical processes for
the removal of suspended and colloidal materials from water.  While sand and
crushed anthracite coal currently are the most commonly used iredia, experience
has shown that GAC can serve as a substitute filter medium.  The physical
characteristics of GAC, when used as a filter Tedium, would lead to the
definition of GAC filtration as deep bed filtration or depth filtration.
Weber (1972) described depth filtration as a combination of a number of
mechanisms, including first a transport mechanism which brings a solute
particle from the bulk of the liquid within the void space close to the
surface of the filtration nedium.  Secondly, an attachment mechanism retains
the solute particle as it contacts the filter iredium or a previously retained
particle.  Both surface and depth filtration mechanisms are thought to occur
in deep bed filters, but it is expected that deep bed mechanisms would
predominate in GAC filtration, similar to those in anthracite coal media.

Biological Activity

     In granular activated carbon adsorbers, biological activity is a  well
recognized phenomenon.  Evidence of the benefits of biological activity 1s
manifested by operational full scale GAC systems in water and wastewater
treatment applications which achieve removals of organic carbon in excess of
those which could be predicted by adsorption alone.

     GAC provides an excellent surface for biomass attachment and a favorable
environment for growth of organisms (Weber & Ying, 1977).  The blonass may
be aerobic, anoxic or even anaerobic, depending upon the concentrations of

                                      70

-------
biodegradable soluble organic materials present and the oxygen resources of
the liquid stream entering the GAC bed.  Operational practice with GAC in
U.S.A. wastewater treatment applications normally has been to attempt to
inhibit biological activity, as it frequently results in the formation of
hydrogen sulfide odor problems.  In contrast, a number of European water
treatment facilities -nake efforts to enhance biological activity in 3AC
post-adsorbers by providing supplemental oxygenation.

     Weber & Ying (1977) point out that bacteria will grow on any fine media
surface.  The rough GAC surface provides many good places for attachment,
while the adsorption capability of the medium is unimpaired.  The activity
of the biomass will be enhanced by moderate rates of air scour, but the
biomass itself will not be removed without vigorous air and water backwashing.
The adsorptive capacity of the GAC is believed by Weber & Ying to concentrate
organic substances on the surface of the carbon, thereby providing a more
concentrated substrate for the biomass and enhancing biological growth and
substrate removal.  However, this hypothesis currently is in dispute (Benedek
e_t al_., 1979, 1980).

APPLICATIONS OF GAC IN WATER TREATMENT

     There are a number of operational examples of the various GAC mechanisms
and processes discussed previously.  Therefore, those considering the applica-
tion of GAC can base their decisions on data and experiences of full scale
operational facilities in both the drinking water and wastewater treatnent
fields.

     For the purpose of this discussion, operational GAC facilities are
div'ded into three categories, which are:

                         Filter/adsorbers
                         Post-adsorbers
                         Biological reactors

     Filter/adsorbers coirbine the function of both filters and adsorbers as
they are not preceeded in the process by a filtration step.  On the other
hand, the post-adsorber unit is preceeded by a filter for the removal of
particulates.  The major function of the post-adsorber is to adsorb organic
materials, in contrast to the dual function of the filter/adsorber.

     Both filter/adsorbers and post-adsorbers can function as biological
reactors in that the 3AC -nedia serve to retain the biomass.  The filter/-
adsorber and post-adsorber can continue to serve as biological adsorbers
even after the GAC adsorption capacity appears to be exhausted.

Filter/Adsorber Applications of GAC

     The use of coal as the upper layer of dual media  filtration units  is
common practice in modern water treatment.  Replacing  the upper portion of a
sand filter with coal provides greater solids storage capacity, longer
filter runs, lower backwash rates and shorter backwash runs.  GAC can and
has been used in place of coal in filters where it serves as filter medium,


                                      71

-------
adsorbent, and for dechlorination within its capacity limit.   Even after
exhaustion of its adsorption capacity, the GAC continues to serve as an
effective filtration medium.  Table 10 compares sand versus GAC as filter
media 1n terms of turbidity removal and length of filter runs.  Table 11
provides information on GAC in sand filters for removal of THMs and TOC.

     Blanck (1979a) describes design criteria successfully used in conver-
sion of existing sand filters to dual media units.  The upper 18 to 30
inches of existing sand filters was reiroved and replaced with 12 x 40 mesh
GAC.  The dual media (GAC and sand) filters were operated at a rate of 2.0
to 3.0 gal/min/sq ft.  Operating backwash rate was 10.5 gal/min/sq ft.  The
GAC/sand filters served a dual purpose, both to remove taste and odor as
well as to filter out particulate matter.

     Richard (1973) described French experiences at the V1gneux-sur-Se1ne
water treatment plant near Paris.  Filter/adsorbers were compared with post-
adsorbers on a plant scale operation.  While a major part of the test was
intended to compare different granular activated carbons, several conclusions
can be drawn, including the practical operating conditions which are summari-
zed in Table 12:

TABLE 12.  COMPARISON OF GAC FILTER/ADSORBERS AND POST-ADSORBERS AT
            VIGNEUX-SUR-SEINE WATER TREATMENT PLANT
Parameter
Bed Depth
Filtration Rate
Empty Bed
Contact Time
Backwash
Air Scour
Water Wash
Units
m
ra3/m2/hr
minutes

m3/m2/hr
m3/m2/hr
Sand
Filter
0.8
5.0
—

55
20
GAC Filter/-
Adsorber
0.8
5.0
9.6

55
20
GAC Post-Adsorber
2.5
13.5
9.6

55
*
* varied, depending upon type of GAC used
Source: Richard, 1973
 (1)  No special operational difficulties were encountered using GAC in
     either node of operation.

 (2)  The cost of SAC was less than that of PAC to maintain a similar water
     quality.
                                     72

-------
TABLE 10.  COMPARISON OF GAC WITH SAND FOR TURBIDITY REMOVAL  AND  SERVICE  LIFE

Site

Pennsylvania


Canada

Pennsylvania

Pennsylvania

Wisconsin

Turbidity (JTU)
Influent

±2.1+0.9*


±9.1+5.0*

±4.9+1.1*

±13+7.2**

±1.5-0.75**

Effluent
Sahci
±0.18+0.05
anthracite

±0.33+0.03

±0.48+0.16

NO***

±0.38+0.14

GAC
±0.2-0.06


±0.35-0.08

+0. 45-0. 19

±0.38-0.18

±0.36-0.12

Bed Depth
(feet)
Sand
2.0
anthra-
cite
3.5

2.5

2.0

3.0

GAC
FS-300
4.0

FS-300
2.5
FS-400
2.5
FS-400
2.0
FS-400
2.9
Flow Rate
(gpm/sq ft)

2.0


3.6

1.4

2.0-3.0

1.8

Service Life
(
Sand
32


±14+2

ND***

ND***

±73+

ours)
GAC
65


±23+3

53

ND***

±62+10

* after pretreatment ** raw water data (no treatment) *** operators noted no difference
Source; Froellch, 1978

-------
TABLE 11.  TREATMENT OF DRINKING WATER BY GAC IN SAND FILTERS
Plant
Location
California
West Virginia
Canada
Pennsylvania
Pennsylvania
Contact Time
(minutes)
10
7.5
8.5
11.0
7.5
THM — Time To Reach
20%
Breakthrough
6 wks
4 wks
2 wks
2 wks
100%
Breakthrough
18 wks
14 wks
10 wks
14 wks
0.5 mq/1
fiver
Initial
4 wks
6 wks
2 wks
2 wks
4 wks
TOO. — Time To
50%
Breakthrough
12 wks
3 wks
8 wks
6 wks
4 wks
Reach
0.5 mg/1
Over 50%
12 wks
10 wks
12 wks
7 wks
8 wks
Source: Froelich, 1978




-------
(3)  Use of GAC as the post-adsorber resulted in better water quality than
     when used as a filter/adsorber.

(4)  Backwash cycles in the filter/adsorber would be dependent on headloss
     or some other operational  parameter.   It was found that better effluent
     quality resulted from use  of the GAC  as a post-adsorber with backwashing
     once per week.

Post-Adsorber Applications of GAC

     The use of GAC as a "post-adsorber" (after a filtration step) has been
practiced frequently in the United States  (Hager & Flentje,  1965; Hager &
Fulker, 1968).  However, the majority of experience with this mode of opera-
tion in drinking water treatment currently exists in Europe.  Table 13
provides a partial listing of European plants using granular activated
carbon.  Of these, it is known  that water  treatment plants of the Federal
Republic of Germany on the lower Rhine River (DUsseldorf, Duisburg and
Wuppertal) placed GAC post-adsorber pressure units in operation in the late
195Cs and early 1960s.  The Foxcote Treatment Works of the Bucks Water Board
(England) has operated GAC units after rapid gravity sand filtration since
1960 (Ford, 1973).  At the Foxcote plant,  filtrate is pumped to 2.44 meter
high and 2.4 meter diameter steel pressure vessels with 0.91 meter of 3AC
providing a contact time of 323 seconds.

     Several water treatment facilities are operated by the  Sol eta Water
District of California, in which pressure  GAC post-adsorbers are used
following diatorraceous earth filtration (Hager & Fulker, 1968; Lawrence,
1968).  Several continuous backwash GAC adsorbers also are in operation in
the New England area.

     Much can be learned regarding 3AC post-adsorber design  and operation
from U.S. municipal wastewater  treatment facilities incorporating this mode
of treatment (Culp, 198C).  A number of wastewater treatment systems are in
operation or are under construction.  Most of these facilities include on-
site GAC reactivation systems.   Table 14 provides a partial  listing of U.S.
wastewater treatment plants which incorporate GAC adsorption.  Guidance on
the design of GAC systems is available from an EPA Design Manual (U.S. EPA,
1973) as well as from technical articles (Culp, 1980; Carnes, 1979).  Techni-
cal representatives of GAC material suppliers also are good sources of
design infornation.

Biological Reactors

     The intentional enhancement of biological activity in GAC filter/-
adsorbers and post-adsorbers is not common practice in United States water
or wastewater treatment plants, although it is practiced in a number of
European water treatment facilities.  The phenomenon of biological activated
carbon (BAC) is described in detail in Sections 8 and 9.  It is recognized
that 3AC will continue to serve as a biomass retention tedium long after the
exhaustion of its adsorption capacity.
                                      75

-------
TABLE 13.  PARTIAL LIST OF EUROPEAN WATER TREATMENT PLANTS  USING  3AC
Location
BELGIUM
Wounren
FEDERAL REPUB-
LIC OF GERMANY
Albstadt
Cologne
Cologne
Duisburg
DUren
DUsseldorf
DQsseldorf
Dusseldorf
Frankfurt
Friedrichs-
hafen
Koblenz
Ktinigswinter
Langenau
Leverkusen
Mainz
Mfllhelm
Schwabisch
Hall
Siegburg
Solingen
Plant

de Blankaart

—
--
--
Wittlaer III
—
Holthausen
Flehe
Am Staad
—
--
—
--
--
--
—
Dohne
—
—
--
Type of GAC
Contactor

beds

—
—
—
columns
—
columns (MH)
columns
columns
—
«
—
—
beds
—
—
columns
—
—
—
Plant
Capacity
cu m/day

43.2CO

18,000
—
~
72,000
36,000
192,000
88,000
119,000
—
30, COO
—
4,800
198,700
--
—
48,000
—
70,000
2,000
Preozonation

yes

yes
yes
—
yes '
yes
yes
yes
yes
—
yes
—
yes
yes
—
—
yes
—
yes
no
                                      76

-------
TABLE 13 (CONTINUED)
Location
Stolberg
Stuttgart
SOsel
Vallendar
Wiesbaden
Wuppertal
FRANCE
Dunkirk
Nancy
Paris
Paris
Paris
Rouen
ITALY
Turin
NETHERLANDS
Amsterdam
Rotterdam
The Hague
SWITZERLAND
Arbon
Biel
Plant
Stolberger
Berg
--
Vallena
Schierstein
Benrath

Vouelle
--
Morsang-sur-
Seine
Vigneux-sur-
Seine
V1ry
Chantillon
la Chapel! e

River Po

--
Kralingen
—

—
--
Type of GAC
Contactor
—
columns
--
--
columns
columns (FB)

beds
—
beds
beds
beds
beds

beds

--
columns
--

—
--
Plant
Capacity
cu m/day
30,000
90, COO
30,000
30.0CO
24.4CO
150, OCO

50,000
—
150,000
36, COO
120, OCO
50,000

86,400

82,200
210,000
—

30,000
36,000
Preozonation
no
no
yes
yes
no
yes

—
—
...
yes
no
yes

no

...
yes
—

yes
yes
            77

-------
                         TABLE 13 (CONTINUED)
Location
Kreuzlingen
St. Sallen
Zurich
ZUrich
SWEDEN
Gtiteborg
Gtiteborg
UNITED KINGDOM
Church Milne
Foxcote
Plant
—
Riet
Lengg
Moos

Alelyckan
Lackareback

—
--
Type of GAC
Contactor
beds
«
beds (FB)
beds

beds (MH)
beds

(MH)
columns
Plant
Capacity
cu m/day
32,400
60, CCO
250,000
—

—
—

—
9,100
Preozonation
yes
yes
yes
yes

no
no

no
no
NOTES: (MH) = on-site Multiple Hearth GAC regeneration facilities
(FB) = on-site Fluidized Bed GAC regeneration facilities
Regeneration Methods and European Costs

     Granular activated carbon loses its adsorptive capacity as the surface
area of SAC becomes covered with adsorbed molecules of pollutants.   This
adsorbed material must be removed if the activated carbon surfaces  are to be
reused.  GAC regeneration is the process of reactivation or removal of the
adsorbed materials from the activated carbon surfaces.

     There are many means for regeneration of GAC, including steam, chemicals,
desorption, and even biological activity.  However, for water and wastewater
treatment applications, thermal regeneration using high terrperatures and
oxidizing atmospheres is the most commonly utilized method.  Thermal regenera-
tion is similar to the original activation process and consists of three
phases:  drying, pyrolysis and regeneration.  Figure 23 illustrates the
three process steps, while cigure 24 provides a diagram of the steps involved
in on-site activated carbon regeneration.
                                     78

-------
TABLE 14.  PARTIAL LIST OF U.S.  WASTEWATER TREATMENT  PLANTS  USING GAC


Plant Location
Upper Occoquon, VA
Tahoe Truckee, CA
S. Lake Tahoe, NV
Orange County, CA
Rocky River, OH
Fitchburg, MA
Piscataway, MD
Derry Township, PA
Vallejo, CA
N. Tonawanda, NY
Ocean Beach, NY
Alexandria, VA
Tra, TX
Garland, TX
Niagara Falls, NY
Arlington, VA
Fairfax County, VA
Pomona, CA
Cortland, NY
Leroy, NY
Cedar Creek, NY
Colorado Springs, CO
Clvld (Westerly), OH
Rosemount, MN
WatervWe Vly, NH
St. Charles, fO
Sylvan Beach, NY
Aquia, VA
Seneca Creek, MD
Horsepen Branch, MD
Lower Anacostia, MD
Santa Clara, CA
Ely, MN
Damascus, MD
lype
of
Contactor
U, P/S
D, P/S
3, P/S
D, P/S
D, P/S
D, P/S
D, P/S
U, G/C
U, P/S
U, G/C
U, G/C
U, G/C
U, G/C
U/D, G/C
C, G/C
D, G/C
D, G/C
D, G/C
D, P/S
D, P/S
D, P/S
D, P/S
D, P/S
D, P/S
D, P/S
D, G/C
U, G/C
U, G/C
U, G/C
U, G/C
U, G/C
U, P/S
D, P/S
U, S/C
Operational
Status
Sept. 1980
1.0.
1.0.
1.0.
1.0.
O.S.
1.0.
O.S.
1.0.
1.0.
i.C.
1.0.
I.C.
I.C.
I.C.
I.C.
I.C.
I.C.
1.0.
I.C.
1.0.
I.C.
1.0.
I.C.
1.0.
1.0.
1.0.
I.C.
I.C.
1.0.
1.0.
O.S.
—
O.S.
1.0.
mgd
Plant
Flow
15
5
7.5
15
10
15.3
5
5
13
14
0.7
54
100
30
48
30
36
10
10
1
5.5
3
50
0.6
0.5
5.5
2
3
5
1
2
4
0.6
0.75

Type 0£
Treatment
tertiary
tertiary
tertiary
tertiary
P/C
P/C
tertiary
P/C
tertiary
P/C
P/C
tertiary
tertiary
P/C
P/C
tertiary
tertiary
dechlorination
P/C
P/C
tertiary
P/C
P/C
P/C
P/C & de-Cl2
tertiary
P/C
P/C
dechlorination
dechlorination
dechlorination
tertiary
P/C
dechlorinatlon
Abbreviations: P/S = pressurized steel contactors
G/C = gravity concrete contactors
U = upflow contactor
D = downflow contactor
1.0. = believed to be In Operation
O.S. - believed to be Out of Service
I.C. = believed to be In Construction
P/C = Physical/Chemical treatment
tertiary = following conventional secondary treatment
dechlorination = functions primarily as a dechlorination unit
• — *
                                    79

-------
drying

                                    heat                              i
      CARBON* ADSORBATE + H2O '  •' . j.»  CA RBON« ADSORB AT E +H2OT
pyrolysls

                           heat
        >N» ADSORBATE QAft-'-uno-c" CARBONIC' + VOLAT
                                                         ADSORBATE
°                 CARBON* ADSORBATE nnn*?*nn.  *" CARBONIC' + VOLATILIZED
                             F ^"*"DW"~V'  ' -««-«'"-"-t«     4
                             r                                 !
regeneration


       CARBONIC' •»• H20 l1s'oo">l8OO'F " CARBON + C °T + H2
                       heat        	-+ c'oKHot
Figure 23. Process steps In the regeneration of granular carbon.

-------
                    spent carbon
                        storage
                              carbon meteringj
                               & dewatering
                                                                      off gas
   effluent
   influent
granular

 carbon

adsorption
                                                       fuel _
                                                        air ^
                                                       steam^
granular

 carbon

regenera
   tion
                     regenerated
                       carbon
                       storage
                                  carbon
                                 cooling
     Figure 24.  GAC regeneration system components.


     The four  basic types  of  GAC  regeneration furnace are as follows:

                              multiple  hearth
                              rotary  kiln
                          c)   fluidized bed
                          d)   electric  moving belt

Each of these  units has  its  individual  advantages and disadvantages, with
nultiple hearth units being  the most  comironly used  1n the United States.

     In 1978,  the PTI site visitation team  inspected the Lurgi, six metric
ton per day, fluidized bed 3AC regeneration  facility at Cusseldorf, Federal
Republic of Germany.  Further information on this installation is provided
in the Dusseldorf plant description of  Appendix  C.  A WABAG designed and
constructed fluidized bed  GAC regeneration  furnace  is 1n operation at the
Wuppertal, Benrath, Federal  Republic  of Germany  water treatment plant.  A
Norlt fluidized bed furnace  is used to  regenerate GAC at the ZUrich, Switzer-
land, Lengg water treatment  plant.  Multiple hearth regeneration furnaces
are known to be in operation at the Alelyckan water treatment plant of
Gfiteborg, Sweden and at the  Church  Milne Treatment  Works, Nottingham, England.

     The prices charged for  granular  activated carbon in Europe generally
are higher (65
-------
     There are three general factors which account for much of the price
difference.  The first is that since SAC is such an energy-intensive product
to manufacture, its price reflects the higher energy costs which prevail in
Europe.  Wholesale prices for petroleum fuel  in Europe were as iruch as twice
those in the United States during 1978, and fuel is estimated roughly to
account for 10 to 20% of the total production cost of GAC in the United
States.

     The second factcr is the higher cost of labor in France, Gerrrany,
Switzerland and The Netherlands, due to high social and welfare taxes,  "^any
of these payments are taxed directly to the employer, often as a percentage
of payroll costs.

     Conversations with utility managers revealed a third factor which may
influence prices, namely a desire to buy from carbon suppliers doing business
primarily in the purchaser's home country.  These national loyalties may
reduce competitive pressures, since each of these countries has very few
activated carton ranjfacturers.

     The prices of carbon which were quoted during the 1978 site visits are
not manufacturers' list prices, but rather the actual costs paid by utilities
and frequently were based upon competitive bidding.  Utility representatives
interviewed agreed that list prices have little bearing on actual cost and
hence are not useful for cost reviews.  They also pointed out that actual
costs paid frequently reflect many influences which relate only to that
single purchase.  For instance, the size of the purchase has an obvious
influence on cost per unit volume.  Also, some bidders may quote prices
below normal profit margins when the initial  purchase is expected to lead to
future, higher margin, sales.  Some contracts also may call for regeneration
services, which is another way for the supplier of activated carbon to
recover from a reduced initial price.  These regeneration contracts may
contain restrictions, however, that require the seller to guarantee high
adsorption efficiency after regeneration, and these limits may boost prices
charged.  Finally, bid prices may be somewhat higher when the seller expects
to be in an advantageous position due to the national loyalties of the
buyer.

     The result of all these factors is that one should recognize that any
or all of these influences may be present in the prices' discussed below.
Hence, some may be above and others below what would be considered an accurate
market price.

     In France, the Soci§t£ Lyonnaise des Eaux et de TEclairage (SLEE)
operates several water treatment plants using GAC.  Several purchases of GAC
had been irade in months immediately prior to June, 1978.  The first of these
was based on a full set of competitive bids, with the terns including a
guarantee of the effectiveness upon reactivation.  The accepted price was
6.9 French francs per kilogram, or approximately 68 cents/pound, based on
the prevailing exchange rate (in mid-1978) of just under 5 francs per dollar.
A separate purchase of additional GAC for the Vigneux-sur-Seine plant was
quoted at 2,800 French francs/cubic Tieter, or about 65<£/lb, based on the
average density of that type of GAC.


                                     82

-------
     In The Netherlands, the City of Rotterdam installed SAC in their new
Kralingen plant in 1977.  The carbon has been leased for two years from the
manufacturer on a basis which includes reactivation services.  Although no
purchase price was set, the utility manager mentioned a likely cost of 1.3CO
Dutch guilders per cubic meter if a flat sales price had been set.  Lsing
the same assumptions as above, this figure converts to 67£(U.S.)/1b, using
the exchange rate prevailing in mid-1978 of 2 Dutch guilders per U.S. dollar.

     Treatment plants at Mulheim and Ddsseldorf, Federal Republic of Gemany
both purchase carbon directly for their own use.  At Ddsseldorf, only a
general cost figure was provided, with the comment that some purchases have
been made below this price of 3,500 Deutsch marks (DM) per ton, or approxi-
mately 80 cents/pound based on the then prevailing exchange rate (mid-1978)
of just under 2 DM/dollar.  The MUlheim plant presented a purchase price
(which did not include reactivation guarantees) of 1,400 DM/cubic meter, or
75
-------
     The fluid bed regeneration furnace operating on-site in DUsseldorf
(Federal Republic of Germany) has been the study of previous research
reports published in the United States (Symons et al_., 1978; Miller e_t al_.,
1978) and in Germany (Poggenburg, 1975).  The furnace has a capacity of
about 6 metric tons/day and operates in a two-stage fashion.  Carbon losses
averaged 8 to 9% per reactivation cycle during the first year of operation,
with about 4 to 5% of this loss taking place in the furnace itself, and the
balance occurring during movement of GAC from the adsorption beds to the
furnace and return.  At DUsseldorf, the spent GAC from each of the three
city water treatment plants is piped by aqueous slurry to the single reactiva-
tion facility installed at the centrally located Holthausen water treatment
plant.

     Although the actual capital and operating costs of this furnace have
not been disclosed as yet, an estimate was made by the utility director that
the total reactivation cost is somewhat less than 30% of the original price
of the virgin GAC.  Based on the figure of 80
-------
     GAC adsorbers in the Federal  Republic of Germany were observed to be of
the pressurized post-adsorber type.   Many of those used in the treatnent
plants of the lower Rhine River are incorporated into a unique, two-level
unit which incorporates filtration in an upper level  and adsorption in a
lower level (Figure 19).   Other facilities incorporate post-adsorber designs
more familiar to United States practice, as illustrated in the plant descrip-
tion of the Dohne (Mtllheim) plant in Appendix E.  The previously referenced
survey of European water treatment practice (Water Research Center, 1977)
describes SAC practice in the Federal Republic of Germany as follows:

     "Activated carbon is extensively used in the treatment of surface
     water, primarily for the removal of organics and taste and odor.
     Activated carbon is used only in the granular form, most'y in pressure
     filters operated at approximately two atmospheres and with a bed depth
     of two meters.  At present, regeneration is carried out by the manufac-
     turers; however, in the immediate future, some treatment plants will
     carry out regeneration on-site using fluidized bed techniques.  Some of
     these regeneration plants are already under construction.

     Karlsruhe University operates a nronitoring service for treatment works
     experiencing difficulties with activated carbon filters.  Their work
     has included identification of organic groups in water and development
     of TCC measurement apparatus and techniques."

     Some GAC installations appear to utilize reinforced concrete, gravity
post-adsorbers.  Due to the high quality of Swiss raw water sources, the
application of GAC appears to be irore of insurance against chemical spills
rather than an everyday need for specific contaminant removal.
                                     85

-------
                                  SECTION 8

                       BASIC PARAMETERS OF BAC SYSTEMS


     The four component steps of BAC processing [(1) chemical oxidation,
oxygenation, aeration, (2) inert media filtration, (3) GAC adsorption and
(4) biological oxidation] will be discussed in this section.  It is important
that each individual step be taken only to the point at which the processing
can be passed along to the next stage, in order to assure maximum cost-
effectiveness.

STEP NO. 1 - CHEMICAL OXIDATION, OXYGENATION AND AERATION

Chemical Oxidation

     Chemical oxidation reactions of organic materials rarely proceed at
constant rates.  If the first oxidation product is more oxidatively stable
than was the starting material, the oxidation reaction rate will tend to
slow down as the starting material becomes depleted.   This is because cherrical
reaction rates usually are concentration-dependent.  In the early stages,
when the readily oxidized starting material  is present at its highest concen-
tration, the oxidation rate will be at its fastest.  However, as the starting
material reacts to produre a more oxidation-resistant material,  its concentra-
tion is lowered, and the overall rate of oxidation will decrease.

     The concentration of chemical oxidant frequently is important' in such
reactions, but less so than is the concentration of the oxidlzable organic
starting material.  Nevertheless, when the initial chemical oxidation rate
slows down,  because of lower concentration of starting material, it is a
natural thought to try to increase the overall oxidation rate by addition of
more oxidant.  In some cases this may be effective, but not if the kinetics
of the pertinent oxidation reactions are slow.  In such cases, the addition
of excessive quantities of oxidant will only increase costs relative to that
for the initial stage of chemical oxidation.

     A primary objective of chemical oxidation, as the first step 1n BAC
processing,  is to convert as much blorefractory dissolved organic material
as possible to biodegradable material using as little oxidant as possible.
However, examination of Figures 13,  14 and 15 (which  show structures of
chemical intermediates formed during enzymatic oxidation of several  types of
organic compounds) indicates a possible secondary role of chemical  oxidation.
The biochemical pathways by which these compounds are oxidized Involve the
presence of many biologically generated enzymes, and  a progression of biologi-
cal oxidation steps to produce smaller compounds.   Even 1f some  of the
starting organic compounds present 1n the water/wastewater to be treated by

                                      86

-------
BAG processing already may be biodegradable, chemical oxidation may result
1n effectively skipping several enzymatic steps in the biological oxidation
process.  It 1s known (R1ce & Miller, 1977) that chemical oxidation of many
organic compounds with ozone or chlorine dioxide produces the same oxidized
organic compounds that are produced upon biological oxidation.

     In considering chemical oxidation as the first step of BAG processing,
one should keep in mind the following key points:

     1)   The relative ease of chemical oxidation of the starting material,

     2)   The relative ease of chemical oxidition of intermediate oxidation
          products,

     3)   The stability of the "end products" to further chemical oxidation,

     4)   The cost-effectiveness of additional  oxidant necessary after the
          Initial oxidation reaction has slowed,

     5)   The cost-effectiveness of increasing  detention time in the presence
          of excess oxidant after the initial oxidation reaction has slowed.

     Point No.  3 is of sufficient interest to warrant more detailed explana-
tion.  Gould &  Weber (1976) have shown that when phenol is ozonized, nany
intermediate organic oxidation products are produced (including catechol, an
intermediate in the 8-ketoad1pate pathway of bacterial oxidation -- Figure
15) and that oxalic acid (HOOC-CCOH) is a relatively stable end product.   In
addition, Gilbert (1977) has shown that the rate of oxidation of water
solutions of oxalic acid with ozone 1s very slow, as measured by the rate of
generation of COj.   On the other hand, when small amounts of hydrogen per-
oxide, H202, were added to the solution, then the rate of conversion of
oxalic acid to  CC>2 upon ozonation increased rapidly.

     Thus it is possible to chemically oxidize  some organic materials to  end
products which  are quite stable, and therefore,  to be unable to convert them
completely to C0£ and water by continued chemical oxidation.  Fortunately,
however, these  stable end oxidation products usually contain more oxygen
than did the original organic materials.  This  means that they are more
likely to be rapidly biodegradable compared with the starting organic mater-
ials before chemical oxidation.

     On the other hand, some biorefractory organics will  require relatively
high dosages of ozone to increase their biodegradabllity.  For example,
Gilbert (1978)  reported on ozone oxidation of dilute aqueous solutions of
aniline, benzenesulfonic acid, 4-chloro-o-cresol and 2-nitro-p-cresol.  He
showed that utilized ozone dosages of 3 to 6 g/g of DOC originally present
were required to increase the biodegradability  of the oxidized products
significantly over those of the starting materials.  In addition, aniline
itself was shown to be more biodegradable than  were its initial oxidation
products.
                                     87

-------
     The significance of the proceeding discussion is that if chemical
oxidation is  incorporated into the BAG process as the initial step, it will
be beneficial to determine what specific organic compounds are present in
the water to  be treated.  This knowledge will allow use of only the minimum
amount of the appropriate chemical oxidant as is necessary to convert the
starting organic material(s) to oxidation products which are now more biodeg-
radable (and  in most cases, less GAC adsorbable) than they were before
oxidation.  Chemical oxidation should not be coupled with BAG processing
when the objective of oxidation is to convert the organic material to CO?
and water in  a single step.  The lower the amount of chemical oxidant which
must be employed to attain this Initial objective, the lower will be the
chemical cost of the overall BAG process.

     Chemical oxidants which have been employed in this initial BAG processing
step include  ozone, potassium permanganate, chlorine and chlorine dioxide
(Rice et. al_., 1979a; 1979b).  It is likely that other strong oxidants, such
as hydrogen peroxide, also can be employed, with the degree of success
dependent upon the relative costs and oxidation powers of each oxidant
involved and  the specific organic materials to be oxidized.

     Kappe (1978) conducted a study of pretreatment methods for the Hagers-
town, Maryland sewage treatment plant.  The Hagerstown plant effluent includes
wastes from 130 industrial firms classified into more than 25 different
product categories.  Some of the industrial wastewater components exert high
immediate and ultimate oxygen demands that either could not be satisfied by
the treatment plant, or were otherwise detrimental to the biological processes
of the treatment system.  Pretreatment methods tested at Hagerstown were
intended to assist the plant to satisfy the oxygen demands by providing
initial oxidation.   The methods used were:   diffused aeration (with and
without the addition of waste activated sludge), addition of sodium nitra-te,
chlorination  (at dosages of 12, 24, 30 and 60 mg/1) and potassium permanganate
(at dosages of 0.53, 1.1, 2.1 and 4.2 mg/1).  Preaeration and prechlorlnation
(at the lowest level of applied chlorine)  proved to be effective pretreatment
methods, increasing the 8005 removal  efficiency of the plant under dry
weather conditions  from less than 702 to above than 90$.   However, KMn04 at
the low levels used did not increase biodegradability of the Hagerstown
sewage.

     In other recently reported studies, Coco et al_.  (1979) evaluated
ozonation as a pretreatment for biorefractory Industrial  petrochemical
wastewaters from halogenated hydrocarbon,  isocyanate, acetylene and styrene
manufacturing industries.  Complete oxidation of the biorefractory components
of these wastewaters with ozone proved to be uneconomical.  However, ozonation
was shown to be effective for pretreatment of wastewaters from the manufacture
of toluene diisocyanate, ethylene glycol,  styrene monomer and ethylene
dichloride.  Batch biooxldation studies showed that ozonation improved the
biotreatability of these industrial wastewaters.

     It should be noted that potassium permanganate has the disadvantage of
producing insoluble manganous hydroxide sludges, which must be removed from
the system.  Chlorine and, in some Instances, chlorine dioxide can produce
chlorinated organic materials, which normally are less biodegradable than


                                     88                             •  '

-------
are non-chlorinated organic materials.  In addition, chlorinated organlcs
may not be removed efficiently by subsequent GAC adsorption.  Even if they
are adsorbed by GAC later 1n the process, the fact that they are not as
rapidly biodegraded means that eventually they can saturate the GAC and
require regeneration of the GAC at that time.  Studies by Rook at the Rotter-
dam (The Netherlands) Water Works (Private Communication, 1978) have shown
that chloroform can be adsorbed by GAC, but is slowly eluted from the column
(before saturation and breakthrough) when chloroform-free water is passed
through the GAC media.  This behavior has been confirmed by McElhaney e_t al_.
(1980) during studies at Philadelphia's Torresdale drinking water treatment
plant.  Thus, chemical oxidation via chlorlnation actually can add to subse-
quent BAG processing costs by cau?Tng early breakthrough of halogenated
organics.

     In general, ozone, hydrogen peroxide, chlorine and chlorine dioxide
will produce only small quantities of insoluble by-products by the process
known as microflocculation (see discussion of Step No.  2, below).   In addi-
tion, neither ozone nor Hj02 are yet known to produce blorefractory organic
oxidation products, and, as of the present state of knowledge,  must be
considered to be the preferred chemical oxldants.  It is also likely that
chlorine dioxide (in the absence of free chlorine) can be a preferred oxidant
for many organic contaminants; however, chlorinated organlcs are produced
under certain conditions (Rice & Miller, 1977).   As will be presented ".ater
in this report, most of the BAC studies reported to date have involved
ozonation or aeration; very little research has been done with  peroxide.

     One major benefit of using ozone as the chemical oxidant is that the
DO content of the aqueous medium becomes elevated concurrently  as  the ozone
is added.   When ozone is generated 1n air, a gaseous mixture containing U
to 2% of ozone In 992 to 98% of air is produced.   When ozone is generated
from oxygen, a gaseous mixture containing  2% to 4X of ozone in  96% to 98% of
oxygen is produced.  Thus, during ozonation,  considerable amounts  of oxygen
are passed through the aqueous solution, thereby elevating the  DO  content
prior to inert media filtration.

     With solutions containing organic materials which are resistant to
ozone oxidation (PCBs, pentachlorophenol,  heptachlorepoxide, etc.), it may
be possible to couple ultraviolet radiation with ozonation to oxidize these
refractory compounds.  Ozone/UV combinations have been shown to increase the
rate of oxidative destruction of such compounds (Prengle & Mauk, 1978;
Arisman et al_., 1979).  Chloroform, which  does not react with ozone to form
soluble cFloride ion In aqueous solutions, has been shown to form chloride
ion when subjected to simultaneous ozone and ultraviolet radiation treatment
(Prengle & Mauk, 1978).

Oxygenation, Aeration

     Chemical oxidation is not always required during step 1.  If the
organic materials present already are biodegradable and are non-biorefrac-
tory, then an oxidant will not be mandatory.  In such cases, the first step
can be merely a preaeration or preoxygenation so as to provide the required
quantity of DO to allow the subsequent biochemical conversions to occur to


                                      89

-------
 the desired degree(s).  Aeration conditions are employed in several drinking
 water  treatment  plants in  the Federal Republic of Germany.  Preaeration of
 the partially treated raw  water is effected before passage through slow sand
 filters and/or through GAC.  On the other hand, wastewaters which contain
 large  quantities of BOD5,  for example, should be pretreated biologically
 before considering BAG treatment for removal of the more biorefractory
 components.  Otherwise large amounts of CO will be required during BAG
 processing, causing large  quantities of sludge to be produced.  In turn,
 these  will require more frequent backwashing of the inert and GAC media.

 STEP NO. 2 - INERT MEDIA  FILTRATION PLUS BIOLOGICAL OXIDATION

     One practical reason  for this intermediate filtration step is simply to
 protect the GAC media from becoming fouled with insoluble materials, which
 would  require more frequent backwashing of the GAC column or bed.   One major
 source of these insoluble materials is the "microflocculation" which occurs
 as a result of chemical  oxidation of the soluble micropollutants.

     The terminology "micropollutant" has evolved from European drinking
 water  treatment practice to describe small amounts of soluble polluting
 components which cannot be removed easily from raw water supplies  by the
 normal processes of flocculation, coagulation, sedimentation and filtration.
 When these micropollutants are chemically oxidized, they become more polar
 in nature, due to the introduction of oxygen into their structures.   In
 turn,   the increased polarity allows them to hydrogen bond to other organic
molecules which contain labile hydrogen atoms, thus increasing the effective
molecular weight of the organic molecules.  If this molecular weight increase
 becomes sufficiently high, water solubility will  decrease.   In addition, if
 polyvalent cations (such as magnesium, calcium, iron, aluminum, etc.)  are
 present, these cations can react with the oxidized micropollutants,  causing
them to coagulate and precipitate.

     The combination of all these secondary effects, which occur after
chemical oxidation, is termed microflocculation,  or flocculation of the
micropollutants.   In many European drinking water treatment plants,  the
ozonation step 1s followed by sand or anthracite filtration in order to
remove the Increased turbidity produced upon ozonation.  An excellent
discussion of the phenomena involved in microflocculation has been presented
by Maier (1979).

     In addition to being water insoluble, many of the materials present in
the microflocculant are biodegradable, because of the introduction of
oxygen into their chemical structures.  As a result, there can be  an increase
 in biological  activity in the solid media through which these materials are
filtered.  Therefore,  if inert media filtration follows chemical oxidation,
 the insoluble flocculated micropollutants will be physically removed from
 solution and a significant portion of the biological activity which  results
 can be caused to occur in the inert media employed.

     In European drinking water treatment plants which are utilizing BAG
 processes, it has been found that when the raw waters contain ammonia, most
of the nitrification (about 80%) occurs in the inert media which precede GAC


                                     90

-------
adsorption  (Rice, 1979a; 1979b).  Therefore, the GAC medium is allowed to
perform its primary tasks, first as an adsorber of soluble organic materials,
and second as a biologically active medium in which at least a portion of
the adsorbed organic material is biologically degraded to COe and water and
in which some adsorption sites are regenerated.

     In sewage treatment, two fluidized sand bed processes have been described
in which biological treatment is conducted.  It is claimed that both processes
save considerable space over classical biological treatment processes.

     Jeris et al_. (1977) describe the Hy-Flo process in which sand beds are
fluidized with water and the medium is maintained aerobic by addition of
pure oxygen.  Three pilot plant studies using the Hy-Flo process were
reported by Jeris et al. (1977) 1n Nassau County, New York.  These ranged in
size from 40,000 gaT/BTy for denitrification to 80,000 gal/day for reduction
in levels of carbonaceous BOD5 and nitrification.  Complete BODs and ammonia
removal was obtained in a fraction of one hour of treatment and in less than
5% of the space required to obtain comparable removals by either the activated
sludge or trickling filter processes.   The empty bed ccntact tine of the
sand beds was 6 minutes.  However, optimum contaminant removals were obtained
when a recycle ratio (recycle flow/primary effluent flow) of 1.5 was employed.
This corresponded to an actual empty bed contact time of 15 minutes.

     A second fluidized sand bed process, the Oxitron system (Anonymous,
1979), also uses pure oxygen and partial recycle.  A pilot plant study was
conducted on wastewater from a corn wet-milling plant in Iowa and 95% 6005
removals were attained.  Estimated total energy required (for oxygen regenera-
tion, feed and recycle stream pumping and biomass separation) for treatment
of a 3.0CO mg/1 BOD5 influent at 1 mgd was 544 hp.   It 1s claimed that a
comparable air-activated sludge system would require 650 hp for powering
aerators and sludge recycle pumps.

     The BASF plant at Ludwigshafen, Federal  Republic of Germany has been
operating an activated sludge plant for treating wastewaters from more than
3CO separate organic chemicals manufacturing processes along with sewage
from several nearby cities.  Effluent from this treatment plant currently 1s
discharged to the Rhine River.  The dally BOD5 load is 150 to 350 tons/day
(average 260 to 270) and the effluent normally contains 10 to 13 tons/day
(95% BODs removal) and 50 tons/day of total organic carbon (TOC).

     The BODs contained 1n the BASF effluent is of no concern to German
authorities, since 1t will  continue to be degraded biologically 1n the Rhine
River.  However, there 1s concern for the COD which is discharged, since it
could contaminate raw drinking water supplies further downstream.

     BASF is conducting pilot plant scale tests which Involve redrculating
the effluent from their large activated sludge treatment plant through GAC
adsorbers (downflow) with continuous aeration or oxygenation.  Sontheimer
(1979b) reported that this pilot plant test system has been operating continu-
ously for 18 months and has been removing over 90% of the TOC (Figure 25)
present in the activated sludge plant effluent.
                                     91

-------
100-

 90-

 80-

 70-

 60-

 50-

 40-
 10-
  9-
  8-
  7
  6-
  5-
  4-
  3-
  2-
  1-
TOC,
mg/L
         GAC influent
    GAC effluent
   r>VY'W'
                                     t   i
                1977
                               1978

   Figure 25. TOC removal upon passage of BASF
            activated sludge effluent through
            biologically enhanced (with 02) GAC.

              (Sontheimer, 1979)
                       92

-------
     Finally, if ozonized wastewater is passed directly into 6AC media
without intermediate filtration, there is the possibility of residual
dissolved ozone being present.  This will be destroyed by reacting with the
GAC, but, consequently, will oxidize some of the carbon,   "he net result
will be a slow loss of GAC.  This effect has been noted at Water Factory 21,
in Orange County, California, where ozonized secondary effluent has been
passed directly into GAC columns.  After a number of months of such operation,
the GAC particles have been found to be much smaller than their initial size
(Argo, 1979, Private Comnunication).

STEP NO. 3 -- REAERATION/REOXYGENATION

     Because the biological activity which occurs in the inert medium
utilizes oxygen (recall that nitrification alone requires 3 to 4.5 equivalents
of DO per equivalent of anmonia-N to produce nitrate), the aqueous solution
which passes through the inert medium can become depleted in 30.  If this
occurs, then inert medium filtration should be followed by a reaeration or
reoxygenation step.  The choice between reaeration or reoxygenation will
depend upon the amount of oxygen which will be required to satisfy the
biological demands of dissolved organic materials during subsequent passage
through GAC.

STEP NO. 4 -- GAC ADSORPTION + BIOLOGICAL OXIDATION

     Following chemical oxidation, inert iredia filtration and reoxygenation,
the aqueous medium now is ready for passage through SAC, during which
adsorbable organic materials present will be adsorbed.  In addition, other
slowly biodegradable organic materials, which were not completely degraded
during passage through the inert filtration medium, will  continue to be
biologically oxidized as the aqueous solution passes through the GAC medium.
This can occur because GAC nedia operating in biologically enhanced modes
contain bacterial colonies on the outer surface layers and in the large
macropores.  The macropores normally represent only about 1% of the total
pore surface area available in 3AC, but these are the only pore openings
large enough to physically house bacteria.  Cn the other hand the micropores,
which nake jp about 99% of the total SAC pore volume, are too small to
accomodate bacteria.  Figure 26 schematically shows the relationship between
micropores, macropores and bacteria.

     GAC macropores are the large openings through which a dissolved organic
molecule must pass on its way into the smaller micropores, where it is
adsorbed.  Thus the dissolved organic molecules are exposed to bacterial
action as they pass by on their way into the micropores.  If and when  the
adsorbed organic molecule is later desorbed from the micropores and diffuses
back out into the main solution stream, it must pass the bacterial colonies
located in the tiacropores once again.

     Bacteria can secrete enzymes which are much smaller in size than  are
the bacteria themselves.  Enzymes are molecular in size, and some may  be
small enough to find their way into the micropores of the 3AC.  In the
presence of sufficient dissolved oxygen, these enzy-ies can cause partial
oxidation of some of the organic molecules adsorbed in the micropores,

                                     93

-------
changing their adsorption characteristics and causing them to desorb.   As
they are desorbed, they pass out of the GAC micropores and past the bacteria
located in the macropores, where they can be further biodegraded.
                      bacterial growth,
                        micropores, 0.05*,
                            average size
                                                  macropore,
                                                      50-100M
                Figure 26.  Schematic representatlon of a GAC particle,

                            showing  pore sizes and  bacterial growths.
     Enzymes are not the only cause of desorption of adsorbed molecules,
however.  Other organic chemicals which are more tightly  bound  by  GAC  than
are the originally adsorbed molecules can be present in the water/wastewater
being treated.  As an example, chlorinated organic molecules are more  strongly
adsorbed (and also are less biodegradable) than  are materials such as  phenols.
Thus GAC can be saturated with phenols and, at the same time, the  biological
activity present can be degrading adsorbed phenols into C02 and water.   If
the concentration of phenol in the water passing through  the GAC,  the  CO,
the size of the GAC column or bed and the detention time  of the aqueous
medium (the empty bed contact time) all  are properly chosen, then  the  system
can attain an equilibrium status.  At this point,  the rate of biochemical
oxidation of adsorbed phenol  will be equal to the rate of adsorption/desorp-
tion of phenol, and the net effect will  be that  the system is in balance.
Under ideal circumstances, all phenol  contained  in the original solution
will be removed, either by adsorption or by biological oxidation,  and  the
effluent will contain no phenol.

     If a chlorinated organic material,  say chloroform, is introduced*.into
this balanced system, the balance will be upset.   Chloroform will  desorb
some of the phenol and itself be  adsorbed.  However,  the  rate of biochemical
oxidation of chloroform is very much slower than is that  of phenol.  The
result will be that the empty bed contact time now will be much too short
for attainment of a chloroform equilibrium such  as was obtained with phenol.
                                     94

-------
 In addition, adsorbed phenol will be desorbed at a rate faster than the
 bacteria can decompose it, and thus non-degraded phenol will be present in
 the BAG effluent.

     As a result, the loading of adsorbed chloroform will increase until one
 of two conditions prevails.  If the GAC becomes saturated with chloroform
 and devoid of phenol, the GAC must be removed from the adsorber and be
 thermally reactivated.  Alternatively, the concentration of adsorbed chloro-
 form can reach a point that even though saturation does not occur, desorption
 (or leakage) of chloroform still begins to occur.   In either event, chloroform
now appears in the GAC effluent in unacceptable quantities, and the GAC must
be reactivated.

     It is clear, therefore, that if a dynamic equilibrium can be estab-
lished whereby the rate of adsorption of organic impurities and biochemical
oxidation of desorbing organic impurities become equal, the biologically
enhanced GAC system will  continue to remove biodegradable organic impurities
from solution without the need to thermally reactivate the 3AC.  It will
only be necessary to maintain a necessary level  of DO and to backwash the
inert media filters and the GAC adsorption medium periodically to continue
the process.  When such an equilibrium has been established, the ability of
the GAC to adsorb sudden surges of non-biodegradable organic compounds
 (which are less polar and, therefore, are more strongly adsorbed by 3AC) is
still available.

     One should not assume from the preceeding discussion that effluents
from GAC columns operating under optimized adsorption and biological  condi-
tions will  be entirely free of dissolved organic carbon.   During early
stages of GAC use, when the activated carbon acts  purely as an adsorber, all
adsorbable dissolved organic carbonaceous material  can be removed from
solution.  However, as biological activity develops over a period of several
weeks, the DOC content of the GAC effluents has been shown to increase.   At
equilibrium, the amount of DOC "apparently removed" from the influent ranges
from 20% to 50%.  We emphasize "apparently removed" because the composition
of the organics in the effluent currently is unknown.  Since the influent
organics have been exposed to biological activity,  at least some of the
organics now present are likely to be products of  biochemical oxidation
and/or decomposition products of the organics originally fed to the GAC
media.  There is also a high probability of by-products of bacterial  metabolic
processes being present.   Finally, high bacterial  colony counts always are
present in GAC effluents, once the biological activity has built up in the
media.  These will be measured as DOC or TOC by normal analytical techniques.

     It is also important to recognize that microorganisms require different
 lengths of time in which to adapt to the particular soluble organic materials
present in the waters being treated.  The term "readily biodegradable"
actually applies only when the microorganisms adapt quickly to their sub-
 strates.  This can take a short time (less than 24 hours), as in the case of
glucose, or significantly longer (several weeks),  as in the case of p-
nitroaniline (Benedek e_t aj_., 1979).
                                     95

-------
     Stephenson et al_. (1980) determined that the effect of ozonatlon on
biodegradation oFbrganic materials has not yet been studied in a thorough
scientific manner and, therefore, undertook an investigation with this
purpose in mind.  In the first report in this series, ozonation of two river
waters and solutions of o- and p-nitroaniline was followed by respirometric
biodegradation studies.  Oxygen uptake and organic uptake during biodegrada-
tion were found to be improved by ozonation, except in the case of p-nitroani-
line, oxidation products of which becane less biodegradable.  These investi-
gators concluded that Intermediates were formed upcn ozonation which either
inhibited biological activity or to which the inoculum was unable to acclimate
over the period of study (3 weeks).

     Stephenson et aj_. (1980) also defined a new parameter which indicates
the fraction of BTssolved organic material made biodegradable by ozonation.
This parameter, f, is defined as:

               f = [ACODQ - (X)ATOC0]/CODQ                  eq. 8

where          f = fraction of COO partially oxidized
             COD » COO concentration after ozonation
           ACOD  = change in COD concentration due to ozonation

           ATOC  = change in TOC concentration due to ozonation

               X = oxygen equivalent required to oxidize organic molecules
                    to COg and water

The value of X was calculated (Stephenson, 1979) to be 2.62 for the two
river waters and 3.33 for nitroaniline solutions.

     Finally, Stephenson et al.  (1980) found that TOC removal  from solution
by a combined ozonation-bTodegradation process ranged from 10% to 70%,
depending upon the specific ozone dosage applied and the water source studied,
and that an ozone dosage of 2 mg/1 was the most cost-effective for the
waters studied.

     In the second part of the above study, Benedek et al. (1980) studied
the effect upon GAC adsorption of river water organics aTter ozonation,
either followed directly by GAC adsorption or followed by biodegradation,
then GAC adsorption.  Prior to biodegradation, a 2 mg/1  ozone dosage increased
the GAC adsorptivity, but at higher ozone dosages GAC adsorptivlty decreased.
However, following ozonation and biodegradation, GAC adsorption of the
remaining organics was Increased.  These authors concluded that pretreatment
(coagulation, flocculation, filtration) can remove about 35% of the TOC in
river water, ozonation removes very little TOC (at a 2 mg/1 dosage), biodegra-
dation removes another 15% and adsorption at a GAC dosage of 160 mg/L removes
an additional 38% of the TOC.
                                     96

-------
                            ___  SECTION 9

                    EUROPEAN BAG APPLICATIONS AND STUDIES


FEDERAL REPUBLIC OF GERMANY

     Introduction of granular activated carbon Into European drinking water
treatment practices occurred before World War II, Its initial applications
being for dechlorination and for taste and odor control (Hopf, 196C).  When
surface waters undergo high levels of prechlorinatlon, considerable amounts
of residual chlorine and chlorinated organlcs are produced in the water
(Sontheimer et,al_., 1978).  European water treatment objectives are to
process surface waters to the same quality as that of natural groundwater
(which does not have to be treated 1n many cases).  Therefore, prechlorinated
waters have to be dechlorinated before they are treated further or distributed
(Sontheimer, 1977a).

pUsseJdprf

     The City of DUsseldorf originally installed ozone for oxidation of iron
and manganese 1n Its sand bank filtered Rhine River raw water 1n the mid-
1960s (Miller et. ajL, 1978; Hopf, 1960).   Today, ozonation is followed by 20
to 30 minutes retention in a holding tank, then by filtration through an
inert medium, then granular activated carbon, then treatment with a small
quantity (up to 0.3 mg/1) of chlorine dioxide for residual.   During storage
of the ozonized raw water at DUsseldorf,  oxidized Iron and manganese ions
are allowed to flocculate and settle.  In addition, excess ozone (and sma'l
amounts of permanganate formed by ozonation of manganese ions) continue to
oxidize dissolved organics.  This holding time is not practiced at all
European plants employing preozonation followed by filtration.

     The Rhine River in the Dflsseldorf area contains considerable amounts of
chlorinated organic materials which are not removed during river sand bank
filtration.  These halogenated organics also are more resistant to oxidation
by ozone than are non-halogenated organics, and thus are less likely to be
converted into readily biodegradable materials.  In addition, halogenated
organics are more tightly adsorbed by the GAC (Ktihn & Fuchs, 1975; Kfllle,
Sontheimer & Steiglltz, 1975).

     Combining the stronger adsorptlvlty of halogenated organics onto
granular activated carbon with their lesser reactivity upon ozonation and
their lower biodegradabllity, simply means that breakthrough of halogenated
organics can occur more rapidly than does breakthrough of non-halogenated^
organic compounds from GAC columns, even though the GAC columns iray contain
optimal biological activity.  Thus, German water works along the Rhine in

                                     97

-------
the Ddsseldorf area monitor their carbon column capacities for Total Organic
Chlorine (TOC1) (by the nethbd of KUhn & Sontheimer, 1973a,b; KUhn, 1974;
KUhn & Sontheimer, 1974), as well as for DOC (by the method of Wfllfel &
Sontheimer, 1974) and UV absorption.  Carbon columns at three Dtisseldorf
plants along the Rhine (Rene, Am Staad, Holthausen) are backwashed every 4
to 6 weeks and are regenerated every 5 to 6 months when the adsorbed TCC1 on
the GAC itself reaches the lower 25% of the carbon beds or if soluble TOC1
appears in the carbon colurm effluent before-adsorbed TOCT reaches the lower
25% of the carbon bed (Miller et aj_., 1978; Poggenburg, 1977, 1978).

     When DUsseldorf activated carbons are regenerated, however, only some
80% of the carbon charge is taken out of the columns.  This leaves a portion
of biologically active carbon in the column so that the level of bioactivity
will not drop significantly when fresh or regenerated carbon is added.  With
fresh carbon columns, about 15 days of operation usually are required for
biological  activity to build up to an effective "steady state", particularly
for ammonia removal (Poggenburg, 1977).

     It is likely that the following processes are occurring simultaneously
in the granular activated carbon beds at CUsseldorf:

1)   Non-adsorbable, biodegradable organic materials are being converted
     biologically into C02 and water as they pass through the bed (as they
     would be degraded biologically in a slow sand filter),

2)   Adsorbable, biodegradable organic materials are being adsorbed and may
     be degraded biologically.  Breakthrough of these materials and those of
     1) above is not observed because the GAC column sizes and flow rates of
     water being processed are designed so that the rate of overall biodegra-
     dation is at least equal to the rate of loading of the GAC columns from
     the oronized waters.

3)   Adsorbable, non-biodegradable compounds (such as those making up Total
     Organic Chlorine, TOC1), are being, adsorbed by the GAC but not being
     biodegraded at any appreciable rate,

4)   Non-adsorbable, non-biodegradable organic materials are passing through
     the GAC unaffected and

5)   Some of the adsorbable, non-biodegradable materials may be desorbed if
     more strongly adsorbed but still non-biodegradable compounds enter the
     GAC bed.  In such cases, those materials so desorbed from the early
     portion of the GAC column can be readsorbed in lower portions of the
     GAC medium.

     Thus, in terms of EPA's THM regulations and proposed SOC regulations,
DUsseldorf plants are using GAC to:

1)   remove THM precursors (by partial oxidation during ozonation, followed
     by filtration through biologically active "inert" media, then adsorption
     onto GAC and/or biological degradation),
                                     98

-------
2)   remove chlorinated synthetic organic chemicals (by adsorption) and

3)   remove those unidentified, adsorbable non-biodegradable SOCs which are
     more strongly adsorbed by the GAC and whose breakthrough would occur
     only after TOC1 breakthrough.

     Over the years which followed Installation of "the DUsseldorf Process,
1t was noted that more dissolved organic carbon was being removed than could
be expected on the basis of the simple summation of the known effects of
ozonatlon and of GAC adsorptive treatment.  When 1t was also discovered that
ammonia levels were much lower after GAC treatment than before, the aerobic
biological activity within the activated carbon columns was recognized as
being beneficial and was examined 1n closer detail.  An extensive, 3-year
pilot plant testing program was conducted on biologically enhanced granular
activated carbon at the Auf-dem-Werder water treatment works in Bremen,
Federal Republic of Germany by Eberhardt, Madsen & Sontheimer (1974), which
will be described in some detail later in this section.

Wiesbaden
     Klotz, Werner & Schweisfurth (1975) reported on a continuing study of
the microbiology in granular activated carbon filters at the Schierstein
water treatment plant in Wiesbaden.   At this plant,  Rhine River water 1s
aerated, settled, chlorinated to the breakpoint,  flocculated, filtered
through sand, then granular activated carbon, then sent to ground infil-
tration.  There is no preozonation of the activated  carbon column.

     Studies of the performances of and bacterial activities in activated
carbon columns at Wiesbaden over a period of 3 years have shown that seasonal
influences are only slight.  There 1s a tendency  for decreased microbial
activity in winter, as Indicated by lower oxygen  consumption and carbon
dioxide production values.  However, changes In raw  water quality also were
shown to cause substantial changes in the colony  numbers (total count of
living and dead bacteria) present 1n the GAC columns.

     Adsorption isotherms were determined for bacteria loaded on the carbon.
At high colony numbers (above lO^O/ml) the system tended to saturation.  At
107 to 108/ml, up to 90% of the bacteria were adsorbed onto the carbon
(Figure 27, upper).  After 20 to 30 hours of operation, adsorption  and
desorption were nearing the steady state (Figure  27, lower) of about 10°
colonies/g of carbon.  Electron scan microscopic  analysis of granular acti-
vated carbons treated differently showed that the bacteria are never present
in greater than a single bacterial layer.  Thus the  total carbon surface
area is only fractionally utilized (about 1%) by  the adsorbed bacteria,
leaving 99% of the total surface (pore) area of the  carbon free for adsorption'
of dissolved organic materials.  This confirms the work of Van Der Kooij
(1975) in The Netherlands.

     Figure 28 shows the numbers of bacterial colonies present in the
Wiesbaden raw water, in the sand filtrate and in  the GAC filtrate for the
nearly 4 year period of early 1972 through 1975.   Bacterial levels 1n the
inlet and after the sand filter remained essentially the same throughout


                                      99

-------
      o

      o
      O)
      in
      o
       c
       o

       o
       o
      N^*

       en
       C

       T>
       m
       o
1012H


1011H

1010H


109H

108H

107-J
           10
             6
                   i    i   i    r    r   i

              106 107 108 109 1010 1011 1012

                 adsorptive  concentration

               (colony numbers/200 ml buffer).
         OB
         c

         •5
         CO
         o
 100


  75-


  50-


  25-


   0
              106 107 108 109 10101011 1012

                 adsorptive concentration

              (colony numbers/200 mL buffer).
Figure 27.  Microbiological loading of GAC--
           dependence on adsorptive capacity.


  (Klotz,Werner & Schweisfurth, 1975)
                        100

-------
       •••••• raw water inlet
       •	-• sand  filtrate
       O.'CB.-^V GAG  adsorber outlet
*
<*-
o
M
V
c
o
o
o
V
+*
o
a
m
f m o m i | o s o n d
       '73
ifmomj | a i o n d
        '74
        m   | o s o n
            1972
     Figure  28. Behavior of microbial  populations on GAC  over 3 years
                  at Wiesbaden, F.R.Germany.
                  (Klotz, Werner  & Schweisfurth, 1975)

-------
this period.  On the other hand, levels of bacterial colonies in the BAG
filtrates were usually less than 100/ml for nearly three years, after which
the levels increased rapidly to approach those of the plant inlet.  This is
an indication that, at least under the Wiesbaden plant operating conditions
and for this raw water, bacterial breakthrough of the BAG occurs after about
3 years of operation.  Bacterial breakthrough was not related to other
effluent parameters, such as TOG, COO, TOC1, etc., and may indicate that
satisfactory BAG operation may only be effective for this length of time (3
years) before regeneration would be required regardless of other parameters
being monitored.

SWITZERLAND

     Activated carbon was installed at ZUrich initially to protect against
oil spillage, later for protection against phenol spills, and for dechlorina-
tion (Schalekamp, 1975).  Zurich's Lengg and Moos plants take raw water from
the Lake of ZQrich, which contains very low concentrations of chlorinated
organics and is otherwise a very clean raw water.  There i's no need for
breakpoint chlorination because of very low ammonia content, but a small
dose of chlorine (T mg/1 maximum) is added at the intake to prevent growth
of mussels.  Granular activated carbon insures dechlorination of this amount
of chlorine.  Before activated carbon filtration, however, the water is
ozonized at dosages of 1 to 1.5 ng/1.

     In plant studies (Schalekamp, 1975), at the end of 7 months both the
top and bottom layers of the carbon beds showed equal loadings of organics,
as measured by UV absorption and COD analyses, thus indicating the need for
regeneration.  However, regeneration of thi's carbon was not_ required because
the continued efficiency of removal of dissolved organic carbon, from the
aqueous Tedium by this carbon remained nearly- the same as that of the new
carbon (Figure 29).  This behavior was attributed by Schalekamp to biological
activity within the carbon bed.

     At the ZUrich Moos plant, the slow sand filter was covered with a1 5 cm
layer of granular activated carbon.  The efficiency of COD1 removal from the
aqueous solution remained essentially constant over the three year period
(about 2.8 mg/1 residual COD in the filtrate); (Figure 30).  This^ performance,
again attributed to bacterial degradation of the adsorbed organics,'was
obtained without reactivation of the carbon, although twice weekly backwashes
were required.

THE NETHERLANDS

     Van Lier et a]_. (1975) describe experiences with granular activated
carbon filtersTn pilot plant studies at Amsterdam.  Three carbon pilot
units were studied side by side, using 2 meter column heights and 8 cu m/sq
m/hr flow rates for 3 months.  The water was treated by iron coagulation,
rapid sand filtration, chlorination, then:

     Process jl_:  Ozonation, rapid sand filtration, granular
                  activated carbon, slow sand filtration,
                                     102

-------
o
CO
                 ,--  RAPID  FILTER

                 ——  ACTIVATED CARBON
             k.O
            3.0
        £E   2.0
1.0
         o
         o
                                                  -x-
                    35TH
                              AOTH
50TH
                                                    WEEK \37't
                  Figure 29.  Efficiency of removal  of  COD  by rapid sand filter and GAC

                                at Lengg plant,  ZUrlch,  Switzerland.
                                    (Schalekamp,  1975)

-------
CD

E
c
0)
3
£    3

o

O

m

c

O
o
o
     1972
                                                         rapid sand filtrate
                                                         slow sand filtrate
                                                      slow GAC  adsorbate
                            1973
1974
1975
      Figure 30. Efficiency of  COD removal  by BAG over  3  years  at  Moos

                  water works, Zurich, Switzerland.
                   (Schalekamp.1975)

-------
     Process #2:  Rapid sand filtration, slow sand filtration,

     Process #3:  Granular activated carbon, slow sand filtration.

     These researchers concluded that:

1)   Ozonation increases bacterial counts considerably after rapid and slow
     sand filtration,

2)   Water treated by Process #1 (with ozone) produces water with better
     color, UV absorption and KMn04 consumption values (a measure of the
     amount of oxtdizable organic matarials present) than waters treated by
     Processes i*2 or #3 without ozone,

3)   Slow sand filtration 1n all three systems reduces bacterial counts,
     color and KMn04 consumption values,

4)   Water qualities by Processes #2 and #3 are about the sanre,  but the
     frequency of backwashing of the slow sand filters is more frequent in
     these processes, which do not include ozonatlon,

5)   Service time of the granular activated carbon column with prior ozone
     treatment of the water was much longer (300 days) than those without
     ozone (175 days) as measured by UV absorption of the filtrates.
     Longer service times were observed for BAC columns which had optimum
     contact times of 20 to 30 minutes (empty bed).

6)   For the same applied DOC, the amount of DOC removed by the  activated
     carbon 1n summer was much greater than in the winter.  This was explained
     on the basis of increased biological activity in the carbon columns at
     the higher summer temperatures.  Oxygen consumption 1n the  winter was
     found to be 0.006 g/hr/kg of carbon and 0.024 to 0.030 g/hr/kg of
     carbon in summer.

CASE HISTORIES

     In this sub-section several case histories of pilot plant and full
scale plant studies of biological activated carbon processes will be discussed
1n some detail.  These are studies conducted at:

(1)  DUsseldorf, Federal Republic of Germany - pilot plant study and operatio-
     nal data
     Bremen, Federal Republic of Germany - pilot plant study
     MUlheim, Federal Republic of Germany -• pilot plant study and opera-
     tional data at the Dohne plant
(4)  Rouen, France - operational data

DUsseldorf. Federal Republic of Germany - Pilot Plant Studies And
     Operational Data

     The City of DUsseldorf began using ozonatlon for taste and odor control
1n 1954 (Hopf, 1970b).  At that time, only 0.7 to 1.0 mg/1 of ozone dosage,


                                     105

-------
along with 7 to 10 mg/1 of penranganate, controlled the problems satisfac-
torily.  Over the years, however, the quality of the Rhine River water
worsened, such that ozone dosages of 1.5 to 2.5 mg/1 of ozone were required,
and sometimes as much as 4 mg/1.  Installation of additional ozone generation
capacity was possible, but this would have been uneconomical since the high
production rate would have been used only infrequently (Hopf, 1970fa).

     As a result, experiments with activated carbon were conducted to
overcome taste and odor problems.  In one section of Dtlsseldorf, only 1.0
to 1.5 mg/1 of activated carbon placed after ozonation was found to be
sufficient to control the taste and odor problems.  This is only l/10th of
the amount of activated carbon that was necessary without the ozone treatment
(estimated at 30 to 40 mg/1).  The average operational lifetime of the
carbon filters for taste and odor removal was 15 months.  This time could be
lengthened if several filters or filter groups could be operated in periodic
stages (Hopf, 1970b).

     Although the river sand bank filtered Rhine River raw water which had
been treated with ozone was, for all  practical purposes, free of bacteria,
the activated carbon filters which follow at DUsseldorf not only are not
sterile, but contain a significant amount of biological activity which was
recognized as being beneficial shortly after activated carbon was installed.

     Ammonia (which is not attacked by ozone) is one of the substances
present in the river sand bank filtrate which is transformed biologically
and is entirely removed during passage through the carbon media.  Even in
freshly filled carbon filters at COsseldorf, biological conversion of
ammonia to nitrate begins after just a few days (Hopf, 1970b).  The dissolved
oxygen content of the effluents from the activated carbon filters 1s 7 to 8
mg/1.

     Water leaving the DUsseldorf activated carbon filters is very low 1n
bacteria content, even though the GAC media are biologically active.
However, the filtering velocity at DUsseldorf is unusally high (35 cu m/sq
m/hr), which means that the total 3 m depth (originally Installed bed depth)
is passed by the water 1n less than 2 minutes (Hopf, 1970b).  This rapid
filtration rate at DUsseldorf is unique, in that all three Ddsseldorf water
treatment plants (Am Staad, Flehe and Holthausen) operate "on demand",
because there is very little reservoir storage capacity available (Miller et
al., 1978).   Thus, all three plants are designed to produce water at peak
demand rates.  As a result, high rates of flow sometimes are employed at
these plants.

     Poggenburg et.aU (1974) reported results of some of the many studies
conducted at one or more of the three water treatment plants in DUsseldorf,
all processing river sand bank filtered River Rhine water.  Studies reported
in this article included pilot plant tests and data obtained from the full
scale plant operations.  A detailed pilot plant study comparing various
types of granular activated carbons was conducted for the removal of various
types of organic chemicals.
                                      1C6

-------
Raw Water Quality at DUsseldorf~
     In the lower Rhine region of the Federal Republic of Germany, the raw
water of the River Rhine contains many organic pollutants, including a
significant fraction of chlorinated organics.  Over the years, the level  of
organic pollution in the Rhine has increased considerably.  For example,
Fuchs & Klihn (1975) reported that in 1962 it was possible to extract samples
of loaded powdered activated carbon (loaded by shaking 40 1  of water 30
minutes with 20 g of powdered activated carbon) with ether and to obtain
about 2 mg of organic extract per liter of ether (Holluta, 1959,  1960).  In
comparison, in 1975 and using dimethylformamide (DMF) as the extractant,
about 8 mg of organic materials was extracted per liter of DMF used.   This
higher amount of organic extract not only is due to an increased  amount of
pollution of the River Rhine, but also to nearly quantitative desorption  by
the DMF as opposed to ether, as well as a higher capacity of the  more modern
activated carbon during the 30 minutes of loading.   Nevertheless, most of
the increased organic material extracted is due to  increased levels of
pollution in the Rhine water (Fuchs & K«hn, 1975).

     Analysis of ether extracts of loaded activated carbons  made  in 1964
showed the presence of a polar, biodegradable fraction along with a non-
polar fraction which comprised about 20% of the total organics extracted
(Holluta, 1964).  Fuchs & KUhn (1975) state:

     "Similar fractions of polar and non-polar materials are found today.
     However, whereas formerly the most important of the non-polar materials
     were mineral oil hydrocarbons, today analysis  shows that the extensive
     group of lipophilic, frequently toxic, organochloro- compounds are much
     more important."

     For example, Fuchs & KUhn (1975) reported some of the then unpublished
work of Stieglitz at the University of Karlsruhe on readily isolated  and
identified volatile compounds from the Rhine River  near Karlsruhe by adsorp-
tion on granular activated carbon, then extraction  first with dioxane (to
remove the non-polar compounds, generally less than 10% of the total  organics
adsorbed), then with DMF.  Some of the compounds obtained are listed  in
Table 15, many of which are halogenated organics.

     Comparison of the quality of River Rhine water at Basel, Switzerland
(the upper Rhine) and at Cuisburg, 30 km north of DUsseldorf, FR3 (the lower
Rhine) in terms of organic sulfur and organochlorine compounds which  are
adsorbed by activated carbon, is given in Table 16.  In Table 17  are  listed
some data obtained at various places along the Rhine and the amounts  of
polar and non-polar organic substances adsorbed by  3AC.

The DUsseldorf Water Treatment Process—
     At all three DUsseldorf City water treatment plants (Am Staad, Flehe,
Holthausen), as well as at almost all water works on the lower Rhine River
in that region of the Federal Republic of Germany,  the general water treatment
procedure consists of five basic unit operations, as shown in Figure 31
(Poggenburg, e_t aj_., 1974).
                                     107

-------
TABLE 15.  THE. MAIN COMPONENTS OF READILY VOLATILE COMPOUNDS  IN  THE  RIVER
            RHINE AT KARLSRUHE
Compound
cyclohexane
trichloroethylene
toluene
tetrachl oroethyl ene
monochlorobenzene
xyl ene
tetrachl oroethane
bromobenzene
cumene
mesitylene
isocumene
B.P., °C
80
87
no
121
132
142
147
156
163
159
159
Compound
dlchlorobenzene
Isobutyl benzene
hexachl oroethane
tetrachl orobutadi ene
pentachl orobutadi ene
trlchlorobenzene
hexachl orobutadi ene
dodecane
tetradecane
pentadecane

B.P., °C
179
170
185 (subl.)
188
—
213
215
215
252
270

Source; Fuchs & KUhn (1975)
TABLE 16.  COMPARISON OF ORGANICALLY BOUND SULFUR AND CHLORINE  IN  GAC  CONTROL
            FILTERS AT DIFFERENT SAMPLING STATIONS
sampling
station
Rhine @
Basel
Rhine 9
Duisburg
adsorbed
organic
substances
mg/cu m
2,900
6,500
sulfur
%
1.5
5.1
chlorine
*
2.3
2.3
organic bound
sulfur
mg/cu m
44
332
organic bound
chlorine
ng/cu m
67
150
Source; Fuchs & KUhn (1975)
                                     103

-------
            Raw River Rhine Water
                       i
                Bank filtration
          Removal of solid materials,
          biological purification,
          manganese absorption
                   Oxidation
          Ozone introduction, KMn04
          manganese oxidation, al-
          teration of the organic
          substances.
             Flocculatiorr filtration
           Precipitation  of manganic
           hydroxide, ozone flocculation,
           flocculation agent and aux-
           iliary agent,  removal of or-
           ganic natter.
                   Adsorption
           Activated  carbon  filter,
           removal  of organic  matter,
           biological  post purifica-
           tion,  nitrification  of  NH4
           Neutralization. and safety chlori

                   NaOH,  C12, C102
Figure 31.  General  procedural  plan for the treatment
            of drinking water in the bank filtration
            works on the lower Rhine  (Dtisseldorf area)
Source:  Poggenburg, ej; aj.., 1974
                    __ 109

-------
TABLE 17.  EFFICIENCY OF GAC CONTROL FILTERS FOR MEASURING DISSOLVED
            ORGANIC SUBSTANCES
sampling
station
Upper Rhine 1? Basel
Lower Rhine @ Duisburg
bank filtrate (DUssel-
dorf-Benrath), Lower
Rhine
bank filtrate
(Hamborn) Lower Rhine
organic
substances
(from DOC)
mg/1
5.6
12.0
5.0
6.0
adsorbed
organic
substances
mg/1
2.9
6.5
4.4
5.3
non-polar
organic
substances
mg/1
0.36
1.19
0.87
0.66
Source: Fuchs & KUhn (1975)
     Step #1 -- River Sand Bank Filtration—This is a technique whereby the
raw Rhine River water is allowed to undergo some biological treatment in the
ground before entering the treatment plant itself.  Wells have been dug into
the sand banks of the Rhine about 50 to 100 meters from the shore.  Water is
pumped from these wells into the plant for further treatment.  About two-
thirds of this well water has passed from the Rhine through the river sand
banks during an average period of three weeks (Miller et al_., 1978).  The
balance is much cleaner groundwater which flows to the~RliTne.

     During the three week time of passage from the Rhine to the wells,
about 60 to 7C% of the organic materials present 1n the river water is
removed from the water biologically.  Those organlcs removed are the easily
biodegraded materials originally present.  The more biorefractory organic
compounds, such as halogenated organlcs, pass through the river sand bank
filters unaffected, and enter the treatment plant.

     In addition to removing the easily biodegraded organic materials
originally present in the River Rhine, the river sand bank filters reduce
some of the nitrate ion present to nitrite and ammonia (Hopf, 197Cb).  Also,
manganese and iron are extracted from the sand banks and are present in the
water entering the treatment plant.

     Step #2 "-Oxidation—Upon entering the plant, the river sand bank
filtered water is subjected to ozonatlon (1 to 3 mg/1 dosage).  Several
benefits are observed during this step.  First, ferrous ions are oxidized to
ferric ions, which hydrolyze and precipitate.  Manganous ions are oxidized
to their manganic and permanganate states.  In the manganic state, the
insoluble manganic hydroxide is formed, which is easily filtered out of
solution.  Manganese in the permanganate state is quite water soluble,
imparting a pink color to the water.
                                      110

-------
     Finally, some oxidation of the dissolved organic materials also occurs.
Some of the organics are partially oxidized completely to COg and water.
This can be deduced from the lower pH after ozonation.  However, most of the
organics which react with ozone are only partially oxygenated.  This increases
their ability to be biodegraded, however.  On the other hand, there are
still other organic materials present which do not react with ozone, and
these (non-readily-biodegradable) materials pass through the oxidation stage
unaffected.

     Step #3 — Flpeculation and filtration—The ozonized water is passed
into a holding tank where it remains about 30 minutes.  During this time
ferric and manganic hydroxides precipitate, some of the oxidized organic
materials flocculate (assisted by the coagulating ferric and manganic hydroxi-
des) and permanganate and residual ozone continue to oxidize more dissolved
organic materials.  As permanganate performs this oxidation function, it is
reduced to the manganic state, at which it precipitates as the insoluble
hydroxide or as manganese dioxide.  As ozone continues to oxidize organics,
it reverts back to oxygen.   Those organic materials which are oxidized
during this holding stage become oxygenated and irore easily biodegraded.

     Following the 30 minute holding time, the treated water is passed Into
a two-stage pressure filter.  The first stage is a 1.5 m depth of granular
carbon which has not been activated.  The original filter material  was 1
meter deep granular activated carbon (0.5 to 2.5 mm grain size), but this
had several undesireable drawbacks (Poggenburg, 1975,).  First, the GAC was
"inadequate as a filter material and for the biological conversion of ammonia
to nitrate".  A considerable fraction of the suspended matter passed through
the upper layer to the second GAC layer, the function of which is to remove
organic materials by adsorption.

     On the other hand, the use of 1.5 m of non-activated carbon for filtra-
tion has produced the following benefits (Poggenburg, 1975):

•    there are now no noticeable amounts of filterable materials in the 3AC
     adsorbers when the combined unit is operated at "reasonable" filtration
     rates,

•    degradation of ammonia occurs in the non-activated carbon bed, rather
     than in the GAC adsorbers,

•    "due to the' good filtration obtained in the upper layer of non-activated
     carbon, the adsorption process is necessarily improved in the GAC lower
     section".

     Most of the filterable substances are retained in the first 0.5 meter
of the non-activated carbon layer, but a bed depth of 1.5 meters is employed
to remove the finer particles and attain a turbidity of 0.1 to 0.2 turbidity
units.

     Step #4 — GAC Adsorption—A mixture of the activated carbons LS Supra
and F-300 are used (Poggenburg, 1975) in the second layer of the pressure
filters.  The granule size of these carbons is 0.5 to 2.5 mm (mean diameter

                        	111

-------
1.4 to 1.6 mm) and the density is about *Q% higher than the activated carbons
used earlier.  The layer heights of GAG in this second section of the pressure
filters were increased from 2 meters to 2.5 meters and the operating velocity
was reduced by about 50% (Poggenburg, 1975).

     The adsorption layers in the DOsseldorf GAC adsorbers are changed twice
each year (Poggenburg, 1975), now that the carbon is used to remove synthetic
organics.  Previously (when used only for taste & odor control), they were
changed once per year.  According to the process control  procedures in use
in 1975 (determination of the total organic components by UV absorption at
250 nm) the adsorption capacity of the GAC for organics (measured by this UV
technique) was only 20% after a throughput of between 400,000 cu m and
6CO,000 cu m of water per adsorber containing 40 cu m of GAC.  The adsorption
capacity of new or freshly reactivated GAC for UV absorbing organics generally
lies between 70 and 80% at this point in the process at DUsseldorf (Poggen-
burg, 1975).  Taking as a basis this degree of loading and a mean operating
velocity of 12 m/hr, the GAC would have to be reactivated or replaced in
each adsorber 4 to 5 times each year if adsorption were the only organics
removal mechanisn occurring in the GAC adsorbers.  The fact that regenera-
tion was not required 4 to 5 tiires per year, as predicted, but rather only
once per year, is attributed to biological removal of UV-abscrbing organic
materials in the GAC adsorbers.

     Step #5 -- pH Adjustment and safety chlorination—Because of the
formation of C02 (from biodegradatTcn or organic materials in the pressure
filters/adsorbers), NaCH is added, fol'owed by a tiaxitr.um dosage of C.3 mg/1
of a mixture of chlorine and chlorine dioxide.  This mixture is produced by
adding exces-s chlorine gas to a solution of sodium chlorite.  The amount of
chlorine added in excess is such as to produce two moles  of free chlorine
(as hypochlorite) for each mole of chlorine dioxide produced (Miller e_t a]_.,
1978).  This low dosage of rrixed disinfectant is sufficient to guarantee
water of acceptable bacteriological quality in the Dtisseldorf distribution
systems.

DUsseldorf Process Performance-
     Figure 32 shows the pertinent water quality parameters measured at
COsseldorf and how they change at each stage of treatment.  Figure 33 shows
the percent removal of organic substances, as measured by DOC, COD and UV
absorption analyses at the various process stages.

     Taking the organics content of the raw Rhine River water as 100%,
almost 60 to 65% is removed during river sand bank filtration.  Ozonation
then lowers CCC and COO values an additional 3 to 5%, but lowers UV absorption
values about 18%.  This fact, coupled with only a slight simultaneous lowering
of DOC and COD, shows that UV-absorb1ng organic materials are changed
structurally, but are not all converted to C02 and water upon ozonation.
Rather, they are converted to other organic materials which still are measure-
able by :CC and COD analyses.

     Finally, GAC adsorption lowers the organics content by an addition 8 to
10%.  Overall, the DUsseldorf treatment process lowers the CCC and COO 70 to
8C% and UV-absorbing organics over 85% (from the Rhine River initial  values).


                                     112

-------
                       treatment
                                         0.02
                                          drinking
                                           water
                                         0.03
   mg/l i_g—
Figure  32.  Changes in Rhine river water quality
           parameters during passage through  the
           Dusseldorf water treatment process.

               (Poggenburg, 1975)
                        113

-------
 However,  starting with river sand bank filtered water entering the plant,
 the combination of ozone and GAC adsorption steps lowers COD and DOC values
 only an additional 10 to 15% and UV-absorbing organics values about 25»
 additionally.
               % remaining
               Rhine
               river

                well
               water

               after
            ozonation


               after
                GAC
                                 DOC      COD      UV
        Figure  33. Changes in organics  parameters In  Ousseldorf
                   water  treatment plants.

                       (Poggenburg, 1975)
Pilot Plant Studies at Dtlsseldorf For Removing Chlorinated Organics
 and Determining Benefits of Biological Activity in GAC Adsorbers""

     An experimental pilot plant installation was constructed at the Flehe
plant which consisted of ten plexiglass columns which could be operated  in
parallel (individually) or in series (one after the other).   Each column was
about 20 meters in diameter and held a charge of granular activated carbon
sufficient to provide a 1 meter bed depth.

     In the first series of tests, eight different types of GAC were compared
with each other, then the four best performing carbons were selected for
more detailed studies.  Many of these studies were conducted with the four
types of GAC operating side by side in the pilot plant test facilities.
Under such conditions, two or three of the test columns were connected in
series.  This allowed measurements to be made at various GAC total  bed
depths.
                                     114

-------
   t    Granular  activated carbons  (labelled Types F, L, N and B) were treated
 with  50,000 cu m of water  ('3.2  million gallons) per cu n of 3AC present in
 the columns, presunably at a flow velocity equivalent to that at which the
 dual  media pressure filters/adsorbers of the large scale plant operated.  At
 various levels of throughput [about 12,000, 25,000 and 50,COO cu i (3.17,
 6.61  and 13.2  million gallons, respectively)/cu m of 3AC passed through
 during the test], GAC samples were renoved from the columns and three values
 were  determined:

 (1)   The Total Organic Substances (TCS) removed from the water 'determined
       by continuous measurement of UV absorption and occasional  parallel
      determination of the concentration of dissolved organic carbcn (:CC)

 (2)  The organic substances which could be extracted from the 3AC with
      dimethylformamide (Extractable Organic Substances  = ECS)  and

 (3)  Adsorbed organochlorine compounds, as chlorine (OrgCl).

      Pertinent data concerning these studies  are plotted on log-log scale in
 Figure 34.   It can be seen that all  three curves have similar shapes  with
 all four granular activated carbons.   It is also clear  that considerably
 more total  organic substances (TOS)  were removed frcm the  water during
 passage through the pilot plant test adsorbers  than were recovered  by extrac-
 tion of 3AC samples with  CMF (EOS).   The DMF  extraction  procedure has  been
 shown (Maier,  1971;  Sontheimer & Maier,  1973)  to extract organic  substances
 adsorbed by granular activated carbon nearly  quantitatively from  the  carbcn.
 Therefore,  the differences between the total  organic substances  removed
 during passage through the GAC adsorbers and  the amount  of organic  substances
 extracted  from the GAC after various  amounts  of water throughput  is that
 amount of  organic materials which were decomposed  biologically  during  passage
 through the biologically  active filters/adsorbers.   This  airount of  organics
 removed biologically was  2 to 3 times  that which was  extracted  by CMF  after
 50,000 cu m of water (13.2 million gallons) had been  passed  through each  cu
 m of GAC,  and  each type of granular  activated carbon  gave  somewhat  different
 results.

      It can also  be seen  from Figure  34 that  after  approximately  25.0CO  cu
 m of water  (6.6 million gallons)  had  been  passed through each cubic meter of
 GAC,  essentially  no additional  increase in  adsorption of chlorinated  organic
 substances  occurred..   It is also apparent that those activated carbons
 which adsorbed the highest amounts of organochloro  compounds (Types L  and ")
 removed less total  organic substances  than did  types  F and  B.   The  maxinun
 airount of  organochloro materials adsorbed  by  GAC type L  was  about 12  kg/cu n
 of GAC after 25,COO cu m  of water (6.6 Trillion  gallons)  had  been  passed
 through 1  cubic meter of  GAC,   At the  same  titre, Type L  GAC  removed abcut 5,
' kg of TCS/cu m of GAC, about 22 kg/cu m of which was  recovered  by CMF  extrac-
 tion, leaving  about 28 kg/cu m of GAC which was destroyed  biologically.

      These pilot plant data obtained at Dusseldorf  are  particularly s1gn1*l-
 cant, since they provide  a practical  limit of organochlcrc natenals  which
 can be adsorbed by a given volume of granular activated  carbon,   "was
 pointed out earlier in thisjection that_^€^se]_dorf  ful1-scale  "lant 'AC


                                       115

-------
                                                            B
         10   20 40 60  10  20 40  60  10   20 40 60  10

             throughput In 103 m3 of water/m3 of GAC
20   40 60
TOS « total org. substances adsorbed and  removed from water
EOS » organic substances extractab/e with OMF
OCI z organic chlorine compounds adsorbed, as Cl
    Figure 34.  Performance of various activated carbons at Flehe test
                plant, Ousseldorf.

                    (Poggenburg je_£ ajk, 1974)
                                   lib"

-------
adsorbers contain a mixture of LS Supra and F-300 carbons.  This 1s presumably
because one has a higher capacity for chlorinated organlcs while the other
has a higher capacity for Total Organic Substances.

     Extensive studies then were conducted on samples of GAC which had been
in use for a period of 6 months using the programmed temperature pyrohydroly-
sis method of KUhn & Sontheimer (1973a).  In this procedure, the sample is
subjected to preliminary drying at 40°C, then is burned at approximately
800°C in an oxygen-water carrier gas which is swept through the furnace.
The temperature is sufficiently high to decompose the organic materials and
produce chloride ion (hydrogen chloride) from the chlorine present, which
then is adsorbed 1n water and this 1s titrated or is determined with an ion-
selective electrode.  This gives a measure of the anount of chlorine origi-
nally present in the organic compound.  Since chloride 1on also 1s adsorbed
from water by the GAC, this must be determined separately.  The loaded
activated carbon is exchanged with 0.1N sodium nitrate and the chloride
removed is determined titrimetrlcally.  The amount of chloride ion originally
adsorbed by the GAC is subtracted from the total chloride ion determined by
the pyrohydrolysis procedure.  This difference gives the amount of chlorine
bound as chlorinated organic materials (Engler-Bunte Institute, 1977).

     In addition, Kfllle (1974) identified several Individual chlorinated
organic compounds by dioxane extraction of the loaded activated carbons,
followed by chromatographic analysis of the extracts.

     Figure 35 summarizes the results which were obtained using the programmed
temperature pyrohydrolysis analyses of the activated carbon adsorbers  connec-
ted in series and filled with various types of activated carbon.   The  total
amount of chlorine found (in g/kg of activated carbon) 1n each type of 3AC
is shown as the upper limit.  This number then is divided according to the
temperature at which the Individual  portions of the total amount either
volatilize or are decomposed.  The hatched portions of Figure 35 indicate
the relative amount of total chlorinated organic material that Kfllle (1974)
was able to identify as Individual chlorinated organic compounds, which
ranged from 7 to 35%.

     Kfllle (1974) was able to identify ten different substances in the
dioxane extracts of carefully dried GAC samples.  These dioxane-extracted
chloroorganics are the non-polar compounds adsorbed.  Pertinent data are
listed in Table 18.  For each individual substance and for each type of GAC,
the amount removed from the water is given as the mg of material  per cu m of
water which must be removed from the water in order to obtain the measured
carbon loading.  From consideration of these data, carbon L appears to be
greatly superior to carbons F and N.

     Poggenburg e_t a_K (1974) also caution that even though small scale
studies with GAC can give consistent data, past experience at DUsseldorf has
shown that large scale tests still must be made in the operating water
treatment plant.  This 1s because of variations 1n the performance of
different batches of granular activated carbon, even though they may be cf
the same type, and variations in wetability, abrasion resistance, amount of
turbidity removed, backwash capacities and regeneration conditions.


                                      117

-------
oo
                                                                           U = upper layer
                                                                           L = lower layer
                           U  U   U   L
                             QAC L
U   U   L
 GAC 8
                    Figure  35  Quantities of individual substances determined by temperature
                               programmed pyrohydrolysis of GAC vs  total org. Cl loading.
                                    (Poggenburg et. gj. . 1974)

-------
TABLE 18.  CHLORINATED ORGANICS REMOVED FROM GAC BY DIOXANE EXTRACTION AT
            FLEHE TEST INSTALLATION, DUSSELDORF*
compound
chloroform
1 ,2-dichloroethane
1 ,2-dichloropropane
trichloroethylene
tetrachl oroethyl ene
bis-(2-chloroisopropyl) ether
o-dichlorobenzene
hexachlorobutadiene
hexachlorocyclohexane
tris-(2-chloroethyl) phosphate
sum of the above chlorinated compds.
stated in mg/cu m of water
total chloroorganics removed
total organic substances
therefore extractable
quantity removed from the water by
GAC of the type:
F
mg/cu m
0.05
0.16
0.36
0.34
0.46
4.52
0.13
0.74
0.11
0.83
7.7
3.9
^26
1,450
853
L
mg/cu m
0.23
0.52
3.C6
0.88
0.82
8.81
0.27
1.79
0.18
1.21
17.8
9.6
MC
1,150
646
H
ig/cu m
0.09
0.21
0.54
0.48
0.46
3.88
0.13
0.75
0.14
1.08
7.8
4.0
^23
1,325
651
* GAC height = 3 m; filtration velocity = 15 m/hr; 78 to 86 cu m water
per kg GAC
Source: Poggenburg et_al_. (1974)
     Comparison of performances of various activated carbons in actual
practice also is nade more difficult by the fact that after a period of 3 to
4 weeks, biological processes occur at significant rates in the large scale
GAC media.  Figure 36 shows some typical organic loadings per kg of GAC
obtained from one of DUsseldorf's plant adsorbers charged with 16,800 kg of
Type F activated carbon and operated over a period of 250 days.  The middle
curve shows the cumulative total organic loading if the carbon had been
acting solely as an adsorber (data obtained by measuring adsorption iso-
therms).  However, the upper curve shows the cumulative amount of organic
materials actually removed from the water.  By adsorption only the cumulative
amount of organic materials removed from water after 250 days of use would
have been about 1,800 kg.  However, the total weight of organics removed by
the GAC was about 2,900 kg.  The difference of 1,100 kg was the amount  of
organic materials removed from the water biologically.
                                      119

-------
    3,000-
  n
    2,000-
M
O
"e
a
a>
t»

°  1,000-
o
a>
c
•a

o
            Poggenburg  et aL, 1974
                                                  organics  actually removed
                                                  by adsorption and
                                                  biological activity
                                                  theoretical organics
                                                   removal by adsorption
                                                          alone    .
                                                    organics actually
                                                    retained in GAC
                                                        adsorber
                      adsorber  content:  16,800 kg of GAC
            V
          'water  throughput, m^/kg of GAC1


               50      100     150    200
                        Operating  time, days
                                                 250
   Figure 36. Organics  removal in  Dusseldorf GAC post-adsorber by
              adsorption and by adsorption + biological activity.
     The lower curve in Figure 36 is  the cumulative amount of adsorbed
organics actually present (determined by DMF extraction).  After 250 days of
use, only 1,100 kg of adsorbed organic materials were shown to be present on
the 16,800 kg of GAC.  The difference between 2,900 kg (total organics
removed from solution)  and 1,100  kg which could be removed from the loaded
GAC by DMF extraction represents  the  total amount of organic materials
destroyed biochemically during the 250 days of use of this GAC.  The differ-
ence between 1,800 kg of GAC capacity acting by adsorption only and the
1,100 kg actual organic loading after 250 days of use represents the adsorp-
tion capacity still available in  this carbon adsorber after 250 days of use.
Therefore, assuming that full saturation of this adsorption capacity can be
attained, it can be concluded that the biological activity in the GAC medium
extends the adsorption  life of the SAC.

     Perhaps more significantly,  the  total amount of organic materials which'
can be removed from the loaded GAC by extraction after 250 days reached a
plateau, as did the adsorption-only curve. Therefore the available adsorp-
tion capacity (the difference between the two lower curves) became constant.
However, the total rate of removal of organics from solution (the upper
curve) still was increasing after 250 days of use.  This is strong evidence
                                     120  _

-------
 to  support  the  conclusion  that some of the organic materials being destroyed
 biochemically are materials already adsorbed on the GAC.   If this were not
 the case, the adsorption capacity of the GAC would have been expected to
 become completely saturated, and the cumulative organic removal curve would
 have reached a  plateau at  the same time as did the other two curves.

     The difference between the 1,800 kg of organic materials removed by
 adsorption only, had the GAC been totally loaded, and the actual 1,1CO
 loading (by CMF extraction) might be viewed as the amount of GAC adsorption
 capacity being reactivated biologically after 250 days of use.  This figure
 (700 kg) is approximately 44% of the total adsorption capacity of the 3AC
 being used.

     Finally, it can be observed from Figure 36 that the contribution of
 biological activity to the removal of organic materials from solution
 became noticeable after 30 to 40 days of use.

 Behavior of DUsseldorf Granular Activated Carbons W.ith Respect to Removal
     of Organochlorine Materials

     Relatively simple analytical methods were desired for controlling the
 performance and reactivation of the GAC adsorbers at CUsseldorf with respect
 to removal of chlorinated organic materials.   To these ends, extraction
methods were developed for these materials, followed by separation techniques
which allow for subsequent gas chromatographlc determinations of the indivi-
dual compounds extracted.  These procedures were developed by the Engler-
Bunte Institute (Park, 1974; Park & Sontheimer,  1973)  and involve initial
extraction of the loaded GAC with dimethylformamide at 50°C.  Mlcrocoulometric
titration of this extract gives the total amount of organochlorine compounds
extracted.  Pyrohydrolysis of the loaded GAC measures  the total  organochlorine
materials present on the loaded GAC.

     Water Is added to the CMF extract and the solution then is extracted
with n-hexane, which separates tre non-polar organochlorine iraterials.   The
 hexane solution then 1s dried and gas chromatographed.  Polar organochlorine
materials remain in the DMF-water solution.

     In Figure 37 are plotted the total concentrations of organochlorine
materials and those of tetrachloroethylene, hexachlorobutadiene and hexachloro-
cyclohexane at GAC bed depths of 0, 100 and 200 cm.   These data were deter-
mined by the above procedures 1n two full scale operating GAC adsorbers at
 the DUsseldorf plant using different activated carbons (types L and F)  being
compared side by side.  The data found were very similar for both carbons at
 the same bed depths.

     About the same data were obtained for total non-polar organochlorine
 compounds when the initial extraction was performed with dioxane rather than
 with DMF, followed by hexane extraction of the DMF extract.  Data shown in
 Figure 38 were obtained using a third full scale DUsseldorf 3AC adsorber.
 From these results, 1t was concluded that only small differences in the data
 were measured using (1) DMF extraction, followed by addition of water and n-
 hexane extraction, (2) Soxhlet extraction of loaded GAC w1th_dioxane and 13)


                                      121

-------
Soxhlet extraction of loaded GAC with  dioxane, addition of water and extrac-
tion with n-hexane.  Therefore,  Poggenburg  e£ al.  (1974) concluded that
extraction of loaded GAC with dioxane  is  a  satisfactory method for determining
the amount of non-polar organochlorine materials adsorbed by GAC.  This
method then was adopted for controlling DUsseldorf's GAC adsorbers, replacing
the then-used UV absorption method.
           o
           <
           o
           a
           JC
           \
           0>
           c

           0
           c
           o
           u
           "c
           <0
           D9
                      •~~-type F  GAC

                      — type L  GAC


          from non-polar chlorohydro-
                carbonsj DMF elution
                         method
             tetrachloroethylene
hexachloro-
   butadiene*1
                     nexachlorocyclohexane
                           GAC bed depth, cm
               10
            Figure  37.  Distribution profiles for organically
                       bound chlorine and some chlorohydro-
                       carbons In Dusseldorf GAC beds

                          (Poggenburg e_t aj... 1974)
     Following adoption of the dioxane/hexane  extraction method for determi-
ning non-polar organochlorine materials  adsorbed,  numerous additional data
were obtained on operating full-scale plant  GAC  adsorbers.  Some of these
data are presented in Figures 39 and 40,

     Figure 39 shows the values obtained for total  extractable organics (Eg)
with DMF, non-polar organics extractable with  dioxane  (Ej), the total amount
of organochlorine compounds adsorbed on  the  activated  carbon  (by pyrohydro-
lysis), the amount of organochlorine compounds in  the  dioxane extracts (Clj)
and the amount of organochlorine compounds in  the  hexane extracts of the
dioxane extracts (Cln).  Data in Figure  39 compare  two different types of
GAC (F and L), one operated at the  DUsseldorf  Holthausen plant, the other
operated at the Dulsburg Wittlaer plant  (both  using  the same  treatment
                                      122

-------
process and sand bank filtered Rhine River water  in the same general location)
and after similar quantities of water had  been  passed through each adsorber.
                             100      200

                          GAC bed depth, cm

             . direct  OMF elution of GAC at 50'C + transfer of
                chiorohydrocarbons from DMF/water to n-hexane

             m> Soxhlet extraction of  GAC with dioxane

             K, Soxhlet extraction of GAC with dioxane + transfer
                of  chiorohydrocarbons  from dloxane/HjO to
                n-hexane
         Figure 38.  Distribution of organically bound
                     chlorine in Ousseldorf GAC beds.

                        (Poggenburg et. aj.., 1974)
     After 90 cii m of water  had  been passed through the adsorbers per kg of
GAC, the total organic materials extracted by DMF from GAC type F was about
60 g/kg and about 40 g/kg  for  type L.  The non-polar organics extracted by
dioxane were 50 and 35 g/kg, respectively, for types F and L, and the total
organochlorine materials found were 2.5 and 5 g/kg, respectively, for types
F and L.  After passage of 210 cu m of water per kg of GAC through the
adsorber, the total extractable organics were about 75 g/kg for the two
types, total dioxane extractables were about 55 g/kg, and total organochlorine
compounds found in the dioxane extracts were 4 and 8 g/kg, respectively, for
types F and L.
                                     123

-------
mr>-

80-
Rfl-

4
O 40-

OD
S 20-
a
c
— 1O-
01 o.
C °
-• ft.
•o «
I 4"
2-
















Eg

^d





Clg
.





fcg

kEd

Cln

*•















Eg
_
Ed





Clg


Old
wt<
E9

Ed



f*l
Clg


Cld
KSKZ














Fn

Ed





PI—
Uy

CI.I
WJ1/M.
Eg

Ed



Clg



cid
W'
tt













               GAG type »  F   L     F    L     F   L

         water throughput «• 90m3/kg  130m3/kg 210m3/kg


              Eg= total  extractable  organics

              Ed=DMF extract (total non-polar extractable organics)

              Clg= total Cl by pyrohydroly sis

              Cld=CI In  dloxane  extract

              Cln = CI in  hexane extract

                         Clha0.22    "Cld = 0.43; ClnsQ.64
           Figure 39. Organics removal with 2 types of GAC in
                      2 Oiisseldorf  treatment plants.

                           (Poggenburg e_taj,1974)

     These data confirmed the earlier pilot plant  test data which  showed
that type F GAC had a slightly higher capacity to  remove  total organic
materials, but that type L could remove double the amount of  organochlorine
materials.

     The cross-hatched areas of Figure 40 show the total  loadings  of non-
polar organochlorine materials on GAC samples  taken from  the  upper, middle
and lower regions of two of the plant operating GAC adsorbers at Dusseldorf,
one using type L GAC and the other using type  F.   As would be expected, the
loadings of total extractable organics and total organochlorine materials
were higher in the upper regions than in the lower.   The  upper layer of Type
F carbon showed total extractable organics levels  of 103  g/kg of GAC and
total organochlorine materials of 1.6 g/kg.  In the upper layer of type L
GAC, total extractable organics levels were  83 g/kg of GAC, but total organo-
chlorine levels were 2.2 g/kg, about 40% higher.
                                     124

-------
                  I    I   total extract, Eg

                         non-polar chlorinated  organics
             o
             o
                 upper mid tower
                   GAC  L
             g
           50

  HHKHBB
upper  mid  lower
  GAC F
       Figure 40.  Change in loading of  2 types of GAC
                    with polar and non-polar organic
                    chlorine compounds  at  large
                    Dusseldorf  treatment plant.


                        (Poggenburg  et  a I... 1974)
     It should be noted also that when the upper layers  of these GAC adsorbers
reached these loadings of organochlorine  compounds, the  lower layers showed
0.25 to 0.3 g/kg loadings of organochlorine materials.   During personal
interviews in 1977 and again in 1978,  Mr. Poggenburg advised that the tech-
nique of measuring the loading of organochlorine compounds as they progress
down the GAC adsorber media by these analytical techniques is used to deter-
mine when the GAC should be regenerated.  Whenever measureable levels of
organochlorine materials are detected  in  the lower 20 to 25% of the Dusseldorf
GAC adsorbers, the upper 80 to 75% of  the GAC is removed and sent to reactiva-
tion.  The remaining 20 to 25% of the  GAC, containing little or no adsorbed
organochlorine materials, is mixed with fresh or reactivated 3AC and acts as
a biological seed so that the GAC adsorbers can reattain biological equili-
brium rapidly.
                                _   12S

-------
      Prior to adopting  these organochlorine analyses, DOsseldorf's GAC
adsorbers were controlled by following organics removed (measured by UV
absorption at 254 nm).  This technique determines the aromatic and unsaturated
organic materials.  Reactivation of Ddsseldorf's GAC was conducted once per
year, using UV absorption as the control analysis.  However, using organo-
chlorine loadings as the controlling analytical procedure, DUsseldorf's GAC
adsorbers now are reactivated every 5 to 6 months.

      Biological activity in the Dttsseldorf GAC adsorbers is considered by
Mr. Poggenburg to be an essential part of the total DUsseldorf treatment
process because the final treated water contains lower dissolved organics
and is ammonia-free.  Furthermore, there is no concern that non-chlorine
containing organics will breakthrough before the organochlorine materials
reach the lower portions of the GAC adsorbers.  Nevertheless, UV absorption
analyses still are conducted routinely at all  three DOsseldorf plants
(Engels, 1978).  In addition, more sophisticated GC and GC/MS analyses also
are conducted on Rhine River water and sand bank filtered treatment plant
influents, so as to note the appearance of organic materials whose fates
when processed by the DUsseldorf treatment techniques are not yet known.

Bremen Pilot Plant Study

     The Auf dem Werder water plant of Bremen, Federal Republic of Germany
has utilized the treatment process illustrated by Figure 41 to treat raw
water from the River Weser.  A comprehensive pilot plant program was carried
out to determine if alternate treatment processes could be found to replace
the slow sand filters, which required a large  proportion of the site area.

     A large pilot plant facility employing granular activated carbon with
preoxidation and preaeration was constructed in 1969, and research was
conducted on biological activated carbon over  a three year period.  The
dissolved organic carbon (DOC)  content of the  raw water was 5 to 10 mg/1,
and the permanganate consumption value numbers varied from 10 to 22 mg/1.
For these tests, the rapid sand filtrate from  the full-scale plant was used
as the pilot plant feed water.   All  pilot plant data obtained were compared
with those obtained after the slow sand filters of the full-scale plant.

Initial  Process Screening Test-
     Many initial screening tests were performed prior to the main pilot
plant operation.  A screening test facility providing a treatment process
similar to that used in the Dusseldorf area to treat lower Rhine River water
was utilized.  The initial  process screening test facility (Figure 42)
consisted of an ozone contact chamber (0.7 m diameter, 3 m high), two holding
tanks (each 2.5 cu m), a sand filter (0.7 m diameter, 3 m height) and an
activated carbon filter (0.8 m diameter, 3 m height).  Provision also was
made to close off the flow of water through the carbon column periodically.
At these times, the water in the carbon columns was recycled back through
the carbon beds (adding oxygen continuously) and analytical parameters were
measured.  These included dissolved oxygen, ~CC, carbon dioxide, ammonia,
nitrate, etc.  In this manner,  material balances were determined at biological
equilibrium.  This was defined as the point at which the rate of consumption
                                     126

-------
raw water-
river Weser
V
chemical .
clarification

1 chemical 1
^ 1 oxidation & 1 " ^
IpH adjustment I
A A
alum Na2SiO3. lime. KMnO4
H2S04
nm IIIIIIMMMHM ri- • 	
1^ f
disinfection 1 ^
V
• I

alow sand 1 ^^m^ «•«
filtration 1 ^
eroundwater
                                           rapid sand
                                            filtration
                                                             ^partial
                                                              recycle
Figure 41. The  Auf dem Werder water treatment  process,
            Bremen, Federal  Republic of Germany.

-------
                           OZONE


                        CONTACTOR
ro
CO
T
                        WATER FROM

                        RAPID FILTER
                                              Hgure 42. Bremen. Germany Water Works Pilot Plant



                                                           (Eberhardt, Madsen & Sonthelmer, 1974)

-------
of dissolved oxygen equaled the rate of formation of C02 (from degradation
of DOC) in the absence of ammonia in the GAC column influent.

     Initially, the depth of the granular activated carbon test bed was 1
meter and the filtration rate was such as to provide an empty bed contact
time of 5 minutes.  Ozone dosages were 1.5 to 2 mg/1.  Under these conditions,
ammonia was removed (biologically by nitrification) as efficiently as in the
full-scale plant, but organics were removed less efficiently in the test
unit.  Increasing preozonation dosages to 4.7 mg/1 and GAC empty bed ccntact
times to 15 minutes (still 1 meter GAC bed depth) gave higher quality water
than that obtained in the full-scale plant, with respect to removal of
dissolved organic carbon (COC), COD, oxygen consumption (by alkaline perman-
ganate), ammonia and 48-hour colony counts.  Pertinent data are listed in
Table 19.

     The testing unit used to gather the data of Table 19  was modified by
incorporating a rapidly driven slow sand filter (RDSSF) after the SAC
adsorber.  This is merely a 1  meter deep slow sand filter  operated at a
filtration rate of 0.5 to 1 m/hr.   This unit was installed to lower the very
high 72 hour colony counts found after the biologically active GAC filter
(see last column of Table 19).

Granular Activated Carbon Screening Test--
     Rapid sand filtrate from the full-scale treatment process was used as
Influent to four granular activated carbon pilot plant units.  Each pilot
plant unit consisted of four columns connected in series as illustrated by
Figure 43.  Since each column contained 1.5 meters of granular activated
carbon, it was possible to evaluate the treatment results  achieved with 1.5,
3.0, 4.5 and 6.0 meters of GAC.

     An initial 11 week GAC screening study was performed  on the four GAC
pilot plant units.  Three units charged with SAC designated as Hydraffins
BD, BR and LS were fed rapid sand filtrate from the full-scale plant.  A
fourth pilot plant unit charged with Hydraffin BD was fed  prechlorinated
rapid sand filtrate (dosage was not stated).

     Review of the data obtained during this 11-week test  period,   summarized
in Table 20, provided the following Information:

1.   Hydraffins BR and LS performed much better in removing dissolved
     organics (as measured by the permanganate oxygen demand values) than
     did Hydraffin BD.  Therefore,, Hydraffin BD was not tested further.

2.   Prechlorination did not improve the ability of Hydraffin BD to remove
     dissolved organics.  If anything, prechlorination was detrimental in
     this regard.  Therefore, prechlorination was not continued in the Ions
     term pilot program.

3.   Colony counts (48-hour incubation) and coliform counts in all activated
     carbon effluents were similar.  However, 72-hour colony counts in the
     prechlorinated Hydraffin BD effluent were 3 to 6 times higher than in
     the filtrates of the other activated carbons tested.


                                     129

-------
    TABLE 19.  TESTS UITH OZONE AND ACTIVATED CARBON FILTERS (JANUARY 8 TO FEBRUARY 2. 1970)
sampling
point


after clari-
fication
rapid sand
filtrate
after
ozonation
after GAC
after RDSSF
after slow
sand filter
(for com-
parison)
PH



7.75

7.63

7.59

7.22
7.18
7.30



DOC
mg/l


4.3

3.7

4.1

2.2
2.4
3.2



COD*
mg/l


12.9

11.4

10.6

9.1
7.9
9.5



UV-abs.
(Corr.)
240 nm,l cm

0.278

0.257

0.163

0.160
0.151
0.225



Permanganate
oxygen, demand
TCK
mg/l
(COD)
21.4

18.4

15.9

13.0
11.7
16.5



Dissolved
02
mg/l

11.7

10.5

16.9

10.1
8.8
6.7




NH4+
mg/l

2.68

2.37

2.16

0.49
0.21
1.41



turbidity
x 10-3
abs.E.

4.b4

0.92

0.99

0.97
0.90
0.89



Colony count/ml
after
48 hrs

6,600

2,700

0

49
9
114



72 hrs

_

_

_

4,600
320
*.



Conditions:
Ozone dosage: 4.7 mg/l.
Flow rate, activated carbon filter: 2 m/hr at 1 m bed height (EBCT - 15 min). Flow rate, RDSSF: 0.5
n/hr at 1 m bed height (EBCT - 1 hr). Slow filter in large scale operation, as normal. DOC, COD and UV
values sometimes from only two specimens. Mean water temperature 1.6°C.
* determined by the bichromate method
** determined by the alkaline permanganate method
' Source: Eberhardt, Madsen & Sontheimer, 1974
co
o

-------
                  GAC
                  column
                    II

n




 GAC
col umn
  #2
                              GAC
                             col umn
                              #3
 T
sampling
 GAC
column
  #4
                                           sampling
                        sampling
                                                                               to next treatment step
Notes:  Each column contains 1.5 meters of GAC bed depth

        Effluent from GAC column #4 can be aerated and recycled to GAC column


  Figure 43.  Schematic of Bremen waterworks pilot plant GAC test units.

                      (Lberhardt, Madsen & Sontheimer, 1974)

-------
         TABLE 20.  COMPARATIVE TESTS WITH DIFFERENT TYPES OF ACTIVATED  CARBON FOR PURELY BIOLOGICAL TESTS.

                    I = MAR.  2A - MAY 5.  1970 = 8.31°C:   II = MAY  8  -  JUNE  16. 1970 = 16..1°C._.
to
ro
San pi ing
point
rapid sand
filtrate
BO filtrate
BR filtrate
LS flltrat"
BO filtrate
chlorinatad
slow sand
filter for
comparison
Dissolved
oxygen
rog/1
I II
11 1 86
9 3 6.8
9.0 5.3
8 6 6.0
9.4 5.9
9.8 7.4
Permangan-
ate 02 con
smtn. mg/1
I II
11 5 11 1
66 80
2.5 3.6
20 28
63 80
9 61 78
A(KMnOi, 02
demand)**,
mg/1
I II

4.8 3 1
9.0 7 5
95 83
51 30
1 90 3 26
pH Value
I II
7 64 7 6'J
7 80 7 65
8.02 7 54
7 73 7 bb
7 38 7 22
7 51 7 56
period- Mar 24 June 16, 1970
rapid sand
filtrate
BD filtrate
BR filtrate
IS filtrate
BD filtrate
chlorinated
slow sand
filter for
comparison
9 7
8.0
6 9
7 2
7.4
8.5
11.3
7 3
3 0
2.3
7 1
8.7

4 0
8.3
8 9
4 1
2 6
7 67
7 73
7 78
7.65
7 30
7 54
Colony counts/nl
after 48 hrs
I II
278 242
364 61
382 70
122 68
246 20
17 89
72 hrs
11
800*
1238
2397
1921
6545

E coli & coll
per 100 ml
I II
1 5
0 1
0 1
0 1
0 0
0 0
overall)
258
149
164
91
70
39
3
<1
<1
<1
0
1
* only i sample ** measured by the alkaline permanganate method
Source: Eberhardt, Madsen 6 Sontheimer, 197*1

-------
4.   Hydraffins BR and LS, during the first six weeks of testing  (Period I),
     reduced the permanganate oxygen demand values from 11.5 mg/1 to 2.5 and
     2.0 mg/1, respectively.  During this period, adsorption appeared to be
     the dominant process removing dissolved organics, and biological activity
     within the GAC media was increasing.  However, during the next five
     weeks (Period II), when significant biological activity appeared to be
     present, permanganate oxygen demand levels in the BR and LS effluents
     increased to 3.6 and 2.8 mg/1, respectively.  In addition, the permanga-
     nate oxygen demands of Hydraffin BR and LS effluents were less than
     half those of the full scale plant.

Long Term Pilot Plant Operation-
     Operation of the previously described Hydraffin BR and LS pilot plant
units was continued for more than three years.  During this time, the other
two 4-column pilot GAC units were used to compare the effectiveness of other
granular activated carbons and to test process parameters over shorter
periods of time.   The pilot plant units were backwashed every two weeks.

     During the next 3 to 4 months, the performance of Hydraffin BR in
removing dissolved organics decreased and reached a level  which remained
essentially constant for the balance of the more than three year test
period.  Figure 44 shows the overall performance of Hydraffin BR in this
regard.  After bioequilibrium had been attained in the GAC unit, about 4 to
6 mg/1  of permanganate oxygen consumption was removed, irrespective of the
value of the influent water, over the last two years of use.   During this
time, rapid sand filtrate from the full-scale plant contained 11.5 to 23
mg/1 of permanganate oxygen demand.  Therefore, the 4 to 6 mg/1 of permanga-
nate oxygen consumption removed during passage through the GAC was equivalent
to 25 to 33% removal  of dissolved organics.   In addition,  permanganate
oxygen consumption values always were higher during summer than in the
winter, probably because of Increased biological activity during periods of
higher water temperatures.

Long Term (3+ Years)  Testing —
     In Figure 45 are plotted the permanganate oxygen consumption values and
the ratios of A(permanganate values)/A(dissolved oxygen consumed) obtained
over the 3+ year period for the Hydraffin BR activated carbon adsorber.
When the ratio was about 1 (which was the case most of the time), then
adsorption and biodegradation of dissolved organics were occurring at about
the same rates.  On the other hand, when this ratio was greater than 1
(during winter),  less dissolved organic material was being removed biochemi-
cally than was being removed from solution by adsorption.   When the ratio
was less than 1,  the authors concluded that biochemical degradation of
adsorbed organic materials was occurring faster than adsorption of organics
from solution.  These results again indicate that adsorptlve sites on the
granular activated carbon either still were available after three years of
use, or were being regenerated biologically.

     Data obtained for the Hydraffin LS GAC adsorber unit over the 38 month
period of use were similar to those obtained with Hydraffin BR, except that
a- slightly longer initial adjustment period was required to reach biological
equilibrium.                           	

                                     133

-------
    25-
 •»  20
O
 c
2
*   15
"W
o
,9   20
                             slow sand filtrate
              GAG effluent
       >nar.
          1970
1971         1972
     years
1973
                            may
   Figure 44. Comparison of chemical oxygen demand (KMnO4>
              removal by slow sand filtration and  BAC at
              Bremen, Federal Republic  of Germany.

                (Eberhardt et aj. ,1974)
   F.gure 45.  Variation of permanganate COD .-•••>-.
              and  a(COD/DO) c«»»r with time of  year.

                Eberhardt et aj,., 1974
                           134

-------
     During the last 9 months of the 3+ year pilot plant test period, the
amount of COg produced during passage through the 3AC columns was measured
and the amount of inorganic carbon produced was calculated.  Pertinent data
are listed in Table 21.  In all cases, the amount of inorganic carbon produced
was greater than the amount of dissolved organics being removed (as measured
by the permanganate oxygen consumption values).  This behavior can be explain-
ed only by the biochemical decomposition of organic iraterials which had been
adsorbed previously by the GAC.                                      —

     Data listed in the lower half of Table 21 show that the biodegradation
processes occur throughout the total 6 meter GAC column depth, and not just
in the first 1.5 meters of GAC bed depth.

     Further indication of biochemical regeneration of adsorption sites was
obtained by extraction studies of loaded GAC samples.  If biological  regenera-
tion occurs at a significant rate, then it would be expected that the loading
of GAC after several years of use would not be as high as that of GAC columns
acting as adsorbers only during the initial several months of use.

     To confirm this hypothesis, samples of Hydraffin BR were removed from
the adsorber columns after 3 years of use and compared with samples of
Hydraffins LS Supra (LSS) and B-12 after only 4 months of use.  All GAC
samples were extracted first with dioxane in a Soxhlet apparatus and  the
solvent evaporated.   The residue corresponds to the amount of non-polar
organic material present on the GAC.  The dioxane-extracted GAC then  was
extracted with dimethylformamide (CMF).   The residue from evaporation of the
DMF corresponds to the amount of polar materials present on the GAC.   The
sum of the two extracts is the total organic loading of the GAC at the time
of sampling.

     It can be seen from the extraction data (shown in Table 22) that the
organic loadings of the first and fourth Hydraffin BR columns after 3 years
of use were 28 to 30 g/kg of GAC.  These loadings compare with 99 and 71
g/kg for the corresponding B-12 columns and 133 and 99 g/kg for the correspon-
ding LSS activated carbon columns.  Such large differences in GAC loadings
were not a result of remarkably higher adsorption capacities of the B-12 and
LSS activated carbons, but to the fact that biodegradation had not yet
reached full swing in the B-12 and LSS carbons after only 4 months of use.

     Table 22 also shows that after 4 months, the fourth GAC columns  in the
B-12 and LSS adsorber units had not yet become fully loaded with adsorbed
organics, since the values measured upon extraction were lower than in the
first columns.  By contrast, the loadings of all 4 columns in the Hydraffin
BR adsorber unit after 3 years of use all were about the same.

     These data indicate that at Bremen's Auf dem Werder plant, removal of
dissolved organic compounds in GAC filters occurs by a combination of
adsorption and biological processes.  Apparently through biological oxidation,
a significant portion of the adsorbed organics are converted to CO? and
water, and adsorption sites become available again to adsorb more dissolved
organic materials.
                              _   _735

-------
          TABLE  21.
AVERAGES OF DATA
ACTIVATED  CARBON
OBTAINED DURING  THE LAST 9  MONTHS  OF THE  38-MONTH  STUDY WITH  BR
AT  BREHEN.  FEDERAL  REPUBLIC OF  GERMANY
Dates
July 25 Sept 12. 1972
Sept 26 Nov 14, 1972
Nov 28 Jan 2J. 19/3
Feb 6 April 3, 1973
April 17 May 22, 1973

July 25 Sep> 12, 1972
Sept 26 Nov 14. 1972
Nov 28 Jan 1. 1973
Feb 6 April 3, 1973
April 17 Hav 22. 1973
geometric average.
raw water
temp
°c
19 9
11 3
5.3
7 0
13 1
X redjctlon
1n oxygen

13 8
12.8
9 7
7.1
12 1
10 8
-n—
25 5
20 7
21 8
23 4
20 1
11 \
oxygen
consumption*
nxj/1
raw
wat»r
14 5
16 4
16 5
15 4
14 0
F 2
12 5
14 5
14 9
13 3
12 3
F 4
10 8
13 0
12 9
11 2
11 1
DOC
eliminated***
mg/l
"n 7
0 39
0 24
0.29
0.39
0 21
0 29
3 4
0 47
0 32
0 21
0 13
0 21
1 4
0 86
0 55
0 bU
0 39
0 34
0 50
UV extinction
at 240 inm X _
1 f\^
water
'20
135
141
131
111
F 2
105
126
130
116
103
F 4
87
144
122
116
98
Inorganic C
formed
ma/l
-frr
1 84
0 b3
0 62
0 76
0 74
0 81
34 14
1 11
0 48
0 bl
0 30
0 65
657
2 95
1 01
1.13
1 06
1 39
1 38
DO
consumption**
F 1 2
4 7
0 7
1 8
1 1
1 4
1 4
7 3
2 4
4 9
2 5
2 9
n>2 Increase
nig/'
F 1 2
6 73
1 96
2 29
2 79
2 70
DO consumption**
mg/l
F 1 2
4 7
0 7
1 8
1 1
1 4
1.62
3 4
2 5
1 6
3 1
1 3
1 4
2 05
1 4
7 3
2 4
4 9
2 5
2 9
3 67
i 4
10 83
3 70
4 13
3 87
5 08
slow sand filtrate
KMnOi,
dlil7iand
10 0
12 3
12 8
12 3
10 6
UV
103
120
119
92
DO consumption
Inorganic C tortned
F 1 2
2 60
1 40
2 94
1 57
1 54
2 00
3 4
2.32
3 52
6 20
4 47
2 28
3 60
1 4
2.49
2 41
4 41
2 39
2 12
2.70
U)
en
           *     measured bv the alkaline permanganate method
           **    higher In winter due to additional decomposition of am onta
           ***   determined by UV extinction and assuming that each 0 038 UV extinction value corresponds  to a DOC of 1 mg/1

           The letter refers to "filtrate" and the following number to the GAC column   Thus, F2 means "In the filtrate from the second GAC column",
           F4 the filtrate from column 4. F 1 2 the filtrate of columns H2. F J 4  the filtrates of columns 1 * 2 f 3 + 4, etc.
                        Source:   Eberhardt,  Madsen  6 Sontheimer,   1974

-------
TABLE 22. EXTRACTION OF ACTIVATED CARBON USED FOR DIFFERENT PERIODS
activated
carbon
f i 1 ter
BR (1-3 mm)
II
II
H
B-12
II
LSS(0. 5-1.0 mm)
M
II
II
filter
stage
1
2
3
4
1
4
1
4
1
4
water
throughout
m3/kij
no
110
110
no
13.6
13.6
11.8
11.8
30.1
30.1
dioxane
extract
gAy
8.0
10.0
6.6
5.9
66.0
33.4
65.3
34.9
53.6
28.2
DMF
extract
g/kg
25.0
25.3
20.4
21.8
33.2
38.0
72.4
62.3
78.2
72.1
total
extract
g/kg
30.0
35.0
27.0
27.7
99.2
71.4
137.7
97.2
131.8
100.3
Source:  Eberhardt, Madsen & Sontheimer,  1974

-------
     Colony Counts—Earlier tests in  which  the  full-scale plant rapid sand
filtrate was oxidized with ozone and  then passed  through GAC adsorbers
showed abnormally high 72-hour colony counts.   Installation of a rapidly
driven slow sand filter (RDSSF) after GAC adsorption  lowered the 72-hour
colony counts to satisfactory levels.   During the 3-year tests with Hydraffin
BR, colony counts shown in Figure 46  were obtained.   In most cases, colony
counts were below 40/ml, but twice (during  consecutive summers) they rose to
very high levels.  After September 1972 a RDSSF unit  was installed after the
BR test unit, and the high levels observed  during the summers of 1970 and
1971 were not repeated.  Bacterial  data listed  in Table 23 were obtained
during the period October, 1972 to May, 1973.
            00
            I*
                                 X
                                 n
£   II

i   1
o   *-
«   -o

O   I
       O
       c
    a.
    a
              100-
               50-
               10-
• •* •o»»»o*»
«
II
"W""'"" 	 L""
                       Jaso"d


                     1970
   filtrate

numbers,
  filtrate
                      1 fmamJ j asondjj fm'mJTi sondlfmamj


                        1971       I   1972      1973
                                                        ?-.,
            Figure 46. Percent decreases In colony numbers and
                       permanganate oxygen consumption (COO)
                       through GAC over 3+ years.

                           (Eberhardt ej, aj., 1974)


     During the 3-year tests  with Hydraffin  LS,  abnormally  high 48-hour
colony counts were observed only once  (May0  1971).  A  RDSSF was installed
after the LS adsorber unit in September,  1970 and,  after an initial 4 month
adjustment period, the 48-hour colony  numbers always were below 20/ml,
except for the May, 1971  data which were  very much  higher (see Figure 47).

     Eberhardt, Madsen & Sontheimer (1974) concluded that the quality of
water following passage through biologically active GAC and a RDSSF was very
high.  At most, River Weser water processed  in this fashion would require
only safety chloHnation (up  to 0.3 mg/1  chlorine dosage) to provide a
stable free chlorine residual for the  Bremen distribution system.
                                     138

-------
       TABU  23.   BACTERIAL  DATA IN RAW WATER. GAC FILTRATE AND RDSFF* FILTRATES




date


Jet. 3, 1972
)ct. 17, 1972
)ct. 31, 1972
tov. 14, 1972
Jov. 28, 1972
tec. 12, 1972
tec. 27, 1972
Jan. 9, 1973
Jan. 23, 1973
:eb. 16, 1973
:eb. 20, 1973
larch 6, 1973
torch 20, 1973
\pril 3, 1973
\pril 17, 1973
lay 8, 1973
lay 22, 1973
aeometric averages
raw water Colony and Coli numbers
i


colony
nos/ml

1240
320
144
96
72
170
220
600
94
200
4?
98
440
850
340
210
300
217.6


E.coli &
col i form
bacteria/100 ml
6
8
2
30 ,
20
28
40
18
no
32
52
8
14
6
8
10
12
15.3
after GAC
column-2
colonies


85
40
14
34
68
39
340
44
41
47
60
32
160
270
84
52
30
59.1
col is


4
4'
?
8
5
10
26
6
56
6
32
0
0
4
2
5
1
3.8
after GAC
column-4
colonies


14
22
2
41
16
4
53
36
8
35
20
0
0
9
9
8
1
6.7
colis


0
0
0
0
3
4
14
4
8
3
16
0
0
0
0
0
2
0.56
t
after RDSSF*
colonies





















10
10
0
3
0
0
2
4
1
0
2
1
0
2
5
g
0
jl.20
colis


0
0
0
0
0
0
4
0
0
0
1
0
0
0
0
0
0
0.14
rRDSSF = Rapidly driven slow sand filter _ ,.. . .. .. . _ ., .. . nn,.,
r J ^n\\YTo- Phorharrlt Marlcon ft "\nnthp i me>r • 1Q7/1 ,
CO
VO

-------
           e
           o
           *«
           a
           E
           3
           
-------
STABLE 24.   COMPARISON OF  VARIOUS  FILTERING  RATES  (ACTIVATED  CARBON is)
,1971
Jan-Feb
March-April
May-June
July-August
Temp.
°C
5.1
8.3
17.4
20.6
pH value
raw
water
7.87
7.91
7.95
8.03
F 1/2
7.71
7.77
-
-
1/4
7.b9
7.62
7.40
7.44
3/4
7.65
7.67
7.52
7.53
DO (mg/1)
raw
niater
11.6
11.0
8.7
7.9
F 1/2
9.9
8.9
-
-
1/4
8.7
7.0
1.9
1.3
3/4
9.7
8.5
4.5
3.4
oxygen consumed*
mg/1
raw
water
12.6
13.1
15.8
16.4
F 1/2
10.2
11.2
-
-
1/4
8.6
9.Q
10.4
11.1
3/4
9.5
10.2
11.8
12.0
Colony
nos/ml
Fl/4
17
32
67
50
3/'
16
29
85
11
Filter 1/2 and 1/4 with vf = 5 m/hr 1/2 = 3 m bed depth
Filter 3/4 with vf =10 m/hr 1/4 = 6 m bed depth
3/4 = 6 m bed depth
 * measured by the alkaline permanganate method
          Source;   Eberhardt, Madsen &  Sontheimer,  1974

-------
      TABLE  25.   COMPARISON  OF  TWO ACTIVATED  CARBON FILTERS OF THE  SAME  BED  DEPTHS  WITH  DIFFERENT
                  FILTERING  RATES
parameter
PH
DO (mg/1)
oxygen consumption (mg/1)*
DOC (mg/1)
COD (mg/1)
UV extinction
no. of colonies
rapid sand filtrate
7.94
9.7
14.4
3.3
12.7
0.139

activated carbon filtrate at
v = 10 m/hr. v = 5 m/hr.
7.59
6.0 ;
10.9 '
2.7
9.5
0.110
26 ,
7.51
3.5
9.7
2.5
8-9
0.109
37
-p*
r
      *  measured by the alkaline permanganate method


               Source:   Eberhardt,  Madsen &  Sontheimer,  1974

-------
 rates must  have been a result of biological decomposition of adsorbed organics,
 since total dissolved organic removal was about the same regardless of the
 filtration  rates.

     Varying GAG Grain Sizes—Two samples of Hydraffin LSS granular activated
 carbon having different average grain sizes (1 m and 1.8 mm) were compared
 over a period of 6 months.  For this study, analyses were made at the start
 of the experiments using fresh GAC samples and data are listed in Table 26.
 It can be seen that although during the first month the larger grain size
 carbon removed more organics (as measured by the differences in KMn04
 consumption values), for the subsequent 5 months, the smaller grain size GAC
 gave better organics removal.  In addition, the differences between the two
 granular activated carbons in removing organics were widening in favor of
 the smaller grain sized GAC as time progressed.

     Recycling Through 3AC to Study Biological Regeneration—After two years
 of operation in a once-through mode, all of the effluent from the fourth
 column of the GAC adsorber unit containing Hydraffin LS was aerated and
 recycled back to the first column.   This recycling was continued for the
 next two months, during which time the effluent from column #4 and the
 influent to column #1 were monitored for KMn04 consumption, CO, CO? formed
 and UV absorption.  Data gathered over this period are listed in Table 27.

     During the initial  recycling period, the dissolved organic content of
 the effluent from column #4 increased, but then decreased after about 1
month to its original value.   After the initial recycling period, the
 amount of C0£ produced decreased to about 5 mg/1.  During this time, however,
 the amount of dissolved organic material removed from solution per pass
 through the GAC (as measured by the permanganate consumption method) remained
 about the same.  Also during this time, the amount of dissolved oxygen
 consumed was about equal  to the amount of C02 produced (about 5 mg/1).  The
 rate of COj production (from biochemical oxidation) was always greater than
 the amount of organics removed from solution;  thus biochemical degradation
 of adsorbed organics was occurring during the 2-month recycling period.

     These data suggest that it may be possible to regenerate granular
 activated carbon columns 1n place, biologically, by employing this recycling
 technique.  However, an estimated 6 month recycling period would have been
 required to attain total  biological regeneration at the Bremen pilot plant,
 and this may not be practical.  Nevertheless, these recycling data indicate
 strongly that biological  regeneration does occur, and that it 1s at least
 technically feasible to consider biological regeneration in place, rather
 than by installing thermal reactivation facilities.  Even partial biological
 regeneration processes could be effective in reducing thermal regeneration
 costs during times when the GAC is fully loaded with organic materials and
 the organic content of the raw water is low.

     Granular Activated Carbons of Different Structures—It had been shown
 by Van der  kooij (1975) and by Klotz, Werner & Schweisfurth (1975) that
 bacteria are too large to fit Into the mlcropores of granular activated
 carbon which comprise about 99% of the available surface area.  Only the
 larger pores near the GAC outer surface are large enough to house the

                                     143

-------
-TABLE 26.   COMPARISON OF  DIFFERENT GAC GRAIN SIZES  (LS ACTIVATED  CAR&QN)
Date
September 1971
October
November
December
January 1972
February 1972
Average, Sept. 1971-
Feb. 1972
Temp.
°C
16.65
12.93
7.16
7.23
2.58
4.06
8.44
3/2 with 0.5 - 1.5 mm range
4/2 with 0.5 - 3.0 mm range
pH value
raw
water
7.83
7.89
7.84
7.77
7.87
7.84
7.84
3/2
/.88
7.78
7.68
7.61
7.71
7.59
7.71
4/2
7.91
7.77
7.66
7.59
7.73
7.61
7.72
DO content
mg/1
raw
water
8.5
9.2
10.1
9.3
11.8
10.3
9.8
3/2
7.1
7.3
8.1
7.3
10.2
7.1
7.8
4/2
7.2
7.3
8.1
7.0
10.5
7.6
7.9
oxygen
consumption*
mg/1
raw
water ,
18.0
22.1
22.1
18.9
18.5
23.2
20.4
3/2
3.4
11M
15.9
14.0
13.9
18.5
11.4
4/2
2.8
11.8
16.6
14.4
14.9
19.7
11.5

* measured by the alkaline permanganate method.



           Source:   Eberhardt,  Madsen  & -Sontheimer,  1974

-------
         TABLE 27.   RESULTS OF  RECYCLING TFSTS AT BREMEN WATER WORKS
1972
•

July 4
July 4
July 5
July 5
July 7
July 7
July 18
July 18
July 19
July 19
July 21
July 21

July 25
July 25
August 1
August 1

August 15
August 15

August 29
August ?9
Average-July 4-Aug. 29
Average-July 4-Aug. 29
difference
PH


7.70
7.43
7.67
7.1b
7.40
6.93
8.05
6.98
7.30
6.80
7.15
6.73

7.30
6.92
7.42
6.95

7.40
7.00

7.27
6.95
7.47
6.95
0.52
oxygen
consumed*
ma/l
11.1
10.4
12.0
13.6
18.0
17.4
14.2
12.3
14.2
13.3
17.1
15.5

13.6
12.0
10.1
9.8

10.8
10.4

10.7
8.5
13.2
12.3
0.9
DO

ma/ 1
6.9
3.5
5.3
0.9
5.8
1.0
8.2
1.4
6.1
1.4
5.6
1.3

b.8
1.3
6.1
1.0

6.0
3.1

6.8
3.2
6.4
1.8
4.6
CO
£
ma /I
5.7
10.1
7.1
16.3
5.5
11.2
3.1
13.2
6.6
11.4
6.6
12.1

7.0
9.7
7.6
11.0

6.6
9.7

7.0
10.1
6.3
11.5
5.2
UV
extinction
240 nm, 1 cm
0.113
0.110
0.148
0.146
0.177
0.175
0.165
0.160
0.172
0.168
0.206
0.195

0.174
0.168
0.155
0.138

0.154 '
0.147

0.140
0.120
0.160
0.152
0.008

Sampling
' Point
before filter 1
after " 4
before " 1
after " 4
before " 1
after " 4
before " . 1
after " 4
before " 1
after " 4
before " 1
after " 4
i
before " 1
after " 4
before " 1
after " 4
]
before " 1
after " 4
l
before " 1
after " 4
before " 1
after " 4

01
         * measured by the alkaline permanganate method
              Source:  Eberhardt, Madsen & Sontheimer, 1974

-------
bacteria.  Therefore, tests were conducted next with Hydraffin B-12, which   -
contains a higher number of large inlet pores than do the other Hydraffins
tested.  Data obtained are listed in Table 28, and these were taken over an
8-ir.onth period.

     During the first 4 months of use, more dissolved organic carbon was
removed from solution (by adsorption) than was inorganic carbon produced.
However, during the fifth month, the amount of inorganic carbon produced
became about the same as the amount of organic iraterials being removed.
After the sixth month, the amount of inorganic carbon produced was greater
than the amount of dissolved organic carbon being removed.  Comparing the
results obtained with Hydraffin B-12 with those from other activated carbons,
it is clear that the B-12 GAC removed a higher percentage of dissolved
organic materials.

     Further testing using fine grained (1 mm) Hydraffin LSS (which has a
higher adsorption capacity for dissolved organics) was conducted for 11
months.  Results from these tests and those shown in Table 24 are plotted in
Figure 48.  The ratios of A(DOC removed/cu m of GAC/hr)  are plotted on the
ordinate against time, and data obtained for Hydraffins  BR, B-12 and LSS
(0.5-1.0 mm grain size)  are presented.  Both B-12 and LSS gave better removals
of dissolved organic materials than did Hydraffin BR. After the initial
period of biology buildup, 32 to 38% of the influent DOC was biochemically
decomposed at a steady rate in the Hydraffin B-12 adsorber.  On the other
hand, 50 to 53% DOC removal occurred in the Hydraffin LSS adsorber, but only
after a somewhat longer initial adjustment period.

     At the end of these 11-month long studies,  samples  of the different
granular activated carbons were extracted with dioxane,  then DMF9 to determine
their total organic loadings.   An average loading of 115 g/kg of GAC was
found for the fine grained LSS, and 85 g/kg for the B-12 carbon.

Summary of Bremen Pilot Plant Test Results --

1)   Dissolved organic substances were first adsorbed by GAC resulting in
     high loadings on the GAC during the first 3 to 5 months of GAC use.
     Thereafter, biological activity reached a steady state, from which
     point a significant fraction of the adsorbed organics was degraded
     biochemically to C02 and water.

2)   After the biological steady state had been attained, 100 to 140 g of
     dissolved organic carbon was bacterially oxidized/cu m of activated
     carbon per day.  Consumption of dissolved oxygen averaged 360 g/cu m of
     granular activated carbon/day and 240 g/cu m/day during winter.

3)   For 30 minute empty bed contact times, about 2 mg/1 of DOC could be
     removed from solution.  This corresponds to about 5 mg/1 of dissolved
     organic materials removed.

4)   Organic materials must be biodegradable.  The amount of biodecomposable
     components present could be increased by preoxidation with ozone.
                                    .146  _

-------
TABLE 28.  TESTS WITH B-12 ACTIVATED CARBON
Date

Sept. 26 Oct. 17, 1972
Oct. 24-Nov. 14, 1972
Nov. 21-Dec. 12, 1972
Dec. 19-Jan. 9, 1973
Jan. 16-Feb. 6, 1973
Feb. 13-Mar. 13, 1973
Mar. 20-Apnl 10, 1973
April 17-May 15, 1973
May 22- June 12, 1973
Raw Water
oxygen
consumed*
mg/1
16.6
16.5
14.1
15.3
19.3
14.8
14.9
14.0
16.5
UV-
extinction
Corr.
0.126
0.150
0.125
0.126
0.173
0.133
0.120
0.110
0.135
% decrease in
oxygen
consumed*
Ir2 1-4
60.3 67.2
39.1 56.2
27.5 48.4
24.6 45.1
18.6 36.7
17.2 36.9
14.9 38.2
22.7 36.0
24.5 36.9
UV
1-4
69.8
49.2
47.8
36.2
31.3
33.3
28.8
32.3
37.0
A02
mg/1
1-4
3.5
3.6
2.6
4.1
4.6
2.4
3.4
3.3
6.9
&co2
mg/1
1-4
1.92
2.98
2.07
3.40
5.85
3.85
4.14
6.73
4.67
A DOC
mg/1
1-4
2.31
1.94
1.56
|.20
I'42
].16
6.90
0.94
1.31
iiinorg. (
mg/1
1-4
O.b2
0.81
O.b6
0.43
1.47
1.05
1.13
1.84
2.64
* measured by the alkaline permanganate method.
Source: Eberhardt, Madsen & Sontheimer, 1974

-------
   specific
   capacity,
  g aDOC/m3 GAC
   5-


   4-


   3-


   2-
   BR-GAC,
     1-3 mm
                       B-12 GAC,
                       formed   £'
                                  "'
                                            LS-supra GAC,
                                           *"  0.5-1.0 mm
               slow filter for comparison
             i ndj
am] j asondj f mamj j asond
        1970
               1971
jfmamjjasond

    1972
j f mam JJ a sondh f
     1973
1974
Figure 48. Variation in specific  capacity for dissolved  organic  carbon
          (DOC) of a GAC filter with operating time for normal (BR)
          and more highly activated carbons.
                 (Eberhardt .el aL.,1974)

-------
     Preozonation also Increased the rate of DOC removal at low temperatures,
     during winter.

5)   Biochemical regeneration of GAC was be followed by measuring the
     amount of C02 generated and relating this to the amount of DOC removed
     from solution.  When the ratio A(DOC removed)/A(inorganic carbon produ-
     ced) was less than 1, the source of the additional CO? must have been
     from organic materials previously adsorbed by the GAC?  When this ratio
     was greater than 1, adsorption of organics was proceeding at a faster
     rate than biodegradation.  When the ratio was about 1, the two processes
     were occurring at approximately the same rate.

6)   During the 3+ years of testing at Bremen, the ratio A(DOC removed) to
     A(inorganic carbon produced) was about 1 most of the time.  During
     periods of low water temperature (winter) the ratio was greater than 1;
     during summer the ratio was less than 1.

7)   The maximum amount of DOC removal  occurred during the initial  period of
     GAC use, when adsorption was the dominating process.   About 80% of the
     DOC was removed during the initial 6 week period.   Thereafter, the COC
     of the GAC effluent rose to a level which remained constant and was
     related to the influent DOC level.  Under the best conditions  found at
     the Bremen pilot plant (large inlet pore GAC, high Initial adsorption
     capacity for the dissolved organic materials present in the Bremen raw
     water, fine grain size), about 50£ of the influent DOC could be removed
     constantly after the biological steady state had been reached.  For
     other granular activated carbons,  the rate of DOC removal at the biologi-
     cal steady state was 20 to 33X.

8)   Satisfactory ^8- and 72-hour colony numbers/ml were obtained after
     passage through biologically active GAC.  However, when the GAC influent
     had been preoxidized with ozone or with chlorine,  the 72-hour  colony
     counts rose to very high levels.  Installation of a rapidly driven slow
     sand filter after the GAC adsorber lowered the 72-hour colony  counts to
     acceptable values.

9)   Coliform counts/100 ml were very low and acceptable in all GAC filtrates.

10)  All GAC experiments conducted at the Bremen pilot plant utilized rapid
     sand filtrate from the full-scale plant, which had been pretreated with
     1  to 3 mg/1 of KMn04-  No experiments were conducted to determine
     whether this permanganate pretreatment increased the biodegradation
     rates.  Treatment of the plant rapid sand filtrate with chlorine (unrepo-
     rted dosage) did not increase the rate of biodegradation as did preozona-
     tion.

11)  Ammonia levels in the River Weser raw waters remained low (below 2
     mg/1) throughout the 3+ years of testing.  Nitrification occurred in
     the rapid sand filters of the full-scale plant, in the biologically
     active GAC adsorbers and in the slow sand filters.  The amount of
     dissolved oxygen required to satisfy the requirements for nitrification


                              _  _  149

-------
     in the GAC and following slow sand filters was sufficiently low that
     additional oxygen was not necessary.

MUlheim, Federal Republic of Germany

     The Rheinlsch-Westfaiische Wasserwerksgesellschaft mbH has taken
advantage of biological activated carbon to radically change the drinking
water treatment process at the 48,000 cu m/day (12.7 mgd) Dohne plant in
MUlheinij, Germany (Sontheimer et al., 1978; Heilker, 1979).  Raw water for
this plant is the River Ruhr,~whTch until mid-April, 1977 was treated by
breakpoint chlorination for ammonia removal, flocculation, sedimentation,
granular activated carbon for dechlorination, then ground filtration.  Over
the years, ammonia concentrations have increased, requiring prechlorination
doses as high as 10 to 50 mg/1.  In turn, these high chlorine doses produced
large amounts of chlorinated organics (Table 29)  which not only were incomple-
tely adsorbed by the carbon columns and passed through the plant into the
distribution system, but also caused frequent regeneration of the activated
carbon columns (every 4 to 8 weeks).

TABLE 29.   ORGANO-CHLOROCOMPOUNDS PRESENT AFTER BREAKPOINT CHLORINATION
Sampling Point
raw water (Ruhr river)
after flocculation + sedimen-
tation
after filtration
after GAC filters
after ground passage + safety
chlorination
DOC1*
PPb
17
—
203
151
92
DOC IN**
ppb
5
—
30
17
18
Sum of halo-
forms, ppb
9
15
23
21
23
CHC13
PPb
<1
6
7
7
9
* DOC1 = dissolved organic chlorine
** DOC1N = dissolved organic chlorine, non-polar
Source: Sontheimer et^a^. (1978).
     During a two year pilot study on the use of preozonation followed by
activated carbon adsorption for removal of ammonia and of organics,  it was
found that breakpoint chlorination could be eliminated completely and the
BAC operation could be relied upon totally for removal of ammonia.   At the
same time the DOC was reduced to the desired levels.

     This process, involving ozonation ahead of sand  filtration and  granular
activated carbon adsorption, was installed and began  operating in mid-April,
1977._At the time, the GAC columns at this plant were 2 meters deep (EBCT =
                                   _ 150

-------
 5.5 minutes) and the GAC, exhausted during use under the old process, was
 used from mid-April until July, 1977.  During late summer, 1977, the carbon
 bed depths were increased to 4 meters (EBCT = 11 minutes).  Fresh GAC was
 charged to the deeper columns and these began operating in November, 1977.
 The process at MUlheim has been operating as predicted by the pilot studies
 and the carbon columns have not yet required regeneration.  The performance
 of the full scale plant process is as effective as was the pilot process at
 the same stage of development (Sontheimer, 1978).

      The newly installed process at Dohne involves preoxidation with about 1
 mg/1 of ozone with addition of poly-aluminum chloride and lime as flocculants.
 Preozonation oxidizes manganese and aids in flocculating the organics.
 After flocculation and sedimentation, 2 mg/1 of ozone is added to oxidize
 dissolved organics.  After a retention time of 15 to 30 minutes, the ozonized
 water initially was again preflocculated using 0.2 mg/1 aluminum chloride
 and 0.1 mg/1  polyelectrolyte, filtered through sand, then passed through
 biological  activated carbon where the bulk of DOC and ammonia are removed.

      During the period April-July, 1977, it was found that the preozonation
 step, coupled with high speed agitiation was so successful in removing
 suspended solids that the second Aids plus polyelectrolyte addition step
 was eliminated when the process was re-started in November, 1977.  However,
 nitrification is promoted by addition of 20 mg/1 of oxygen (obtained from
 liquid oxygen cylinders) before sand filtration (where most of the nitrifica-
 tion occurs).   The ammonia content of the Ruhr River has risen to nearly 6
 mg/1 as of June, 1978.   Addition of gaseous oxygen is controlled by a DO
 residual level  of 7 mg/1 monitored after the BAG columns.

      Table 30 summarizes the the process changes which were made at the
 Dohne plant during the period 1977-1978.  A comparison of the performance of
 the new process versus the old process is given in Table 31.   The COC of
 treated water today is less than half of that treated by the old process.
 Even lower DOC values are being obtained since the depths of the GAC columns
 have been increased.  Finally, no chlorinated organics are being generated
 by prechlorination.

      After activated carbon adsorption,  the treated water is sent to ground
 infiltration (12* to 50 hours retention time), after which it is chlorinated
 (0.2 to 0.3 mg/1) and sent to the Mdlheim distribution system.  Ground
 infiltration at Dohne no longer is considered to be necessary, because of
 the high quality of water which the plant produces.  It will  be maintained,
 however, as a convenient and functional  water reservoir (Heilker, 1978).

      Table 32 shows the bacterial content of waters at the various points in
 the new treatment process.  E. coli counts/100 ml are essentially zero after
"filtration and remain essentially zero after BAC filtration as well.  There
 should be no concern for E. coli passing through the BAC media into the
 product waters because this" strain of bacteria does not survive in the
 presence of the other bacterial strains present in the GAC (Van Der Kooij,
 1978; Schweisfurth, 1978).
                                      151

-------
TABLE  30.   PROCESS  PARAMETERS AT THE  DOHNE WATERWORKS  (MULHEIM)  BEFORE AND
             AFTER CHANGE  OF TREATMENT
Treatment Step
preoxi elation
dosing
power input
flocculation
sedimentation


filtration
with
preflocculation
activated
carbon filter
ground passage
safety
chlorination
Old Treatment
(before 1977)
10-50 mg/1 C12
4-6 mg/1 AT*3
0.1 kW/m3
Ret. time =0.5 min
5-12 mg/1 Ca(OH)2
Ret. time = 1.5 hr


v* - 10.7 m/hr
v* = 22 m/hr
h** = 2 m
Ret. time 12-50 hr
0.4-0.8 mg/1 C12
New Treatment
(April-July 1977)
1 mg/1 03
4-6 mg/1 Al+3
2.5 kW/m3
Ret. time a0.5 min
5-15 mg/1 Ca(OH)2
Ret. time = 1.5 hr
2 ma/1 (]-3
Ret. time = 5 min
v* = 9 m/hr
0.2 mg/1 Al+3
0.1 mg/1 poly-
electrolyte
v* = 18 m/hr
h** = 2 m (4 m)
Ret. time 12-50 hr
0.2-0.3 mg/1 C12
New Treatment
(Nov. 77-June 78)
1 mg/1 03
4-6 mg/1 Al+ 3
2.5 kW/m3
Ret. tine =0.5 Tiin
5-15 mg/1 Ca(OH)2
Ret. time= 1.5 hr
2 mo/1 O1?
Ret. time - 5 min
20 mg/1
liquid oxygen
v* = 28 m/hr
h** = 4 -n
Ret, time 12-50 hr
0.2-0.3 mg/1 Cl,
         *v  = filter velocity
         **h = bed height
Sources:  Sontheimer e_t al_., 1978
          Jekel,  1978 and 1979
                                    152

-------
TABLE 31.  MEAN COC VALUES AFTER THE DIFFERENT TREATMENT STEPS (MULHEIM)
sampling point
raw water (Ruhr river)
after flocculation + sedi-
mentation
after filtration
after GAC filtration
after ground passage
1975
DOC,
mg/1
3.9
3.2
3.2
3.0
1.8
1976
DOC,
mg/1
5.0
4.0
3.8
3.7
2.1
April -July
1977, DOC,
mg/1
3.6
2.9
2.6
2.3*
0.9
Nov. 1977 -
June 1978,
DOC, mg/1
2.4 - 3.7
1.8 - 3.0
1.7 - 3.1
1.0 - 2.6**
—
* filters filled with fully loaded GAC, used in old treatment process
** filters doubled in height and filled with new GAC
Sources: Sontheimer et al., 1978; Jekel, 1978

     Pilot plant data are presented in Table 33 which show the effects of
variation of activated carbons on removal of DOC, inorganic carbon, arraronia
and dissolved oxygen.  In addition, this table also compares the removal  of
these same parameters with carbon column depths of 2.5 m and 5.0 m for two
different activated carbons.

     Conversion of ammonia and dissolved oxygen consumption are fairly
independent of carbon type or column depth.  On the other hand, removals  of
DOC and inorganic carbon produced are affected by the carbon type.  Most
significant, the amount of DOC removed with 5.0 m columns is about 50%
higher than with 2.5 m columns, although the amount of inorganic carbon
measured increases only slightly.

     During pilot studies at the Dohne plant with the BAG process, activated
carbon columns were found to have operational lives of at least one year,
and in some cases two years, without requiring regeneration.  Life of the
full scale carbon colutrns at Dohne now is estimated to be at least two years
(Sontheimer, 1978).  No signs of loss in performance of the activated carbon
have been noted during the first year of operation, and the new carbon
columns have not yet had to be regenerated (Jekel, 1978).

     Additionally, preozonation has extended the running times of the sand
filters from two to seven days before backwashing is required.  Backwashing
of the four GAC filters (each of which contains a different granular activated
carbon being tested on full scale) varies, but was conducted during initial
stages of plant operation every 10 days9 on the average.  However, during
the summer of 1978, "a population explosion of nematodes" was observed in
both the rapid sand filters and GAC adsorbers.  These growths were caused by
the long intervals between backwashes, during which the nematodes developed.
                               	153

-------
 TABLE  32.  GEOMETRIC MEAN VALUES OF BACTERIAL COUNTS FOR THE MULHEIM,
            GERMANY  (DOHNE) PLANT USING OZONE
Sampling place
Raw water (Ruhr)
After flocc'n + sediment1 n
After filtration
After activated carbon
After ground passage
Total Bacterial
Counts/ml
v
14,490
2,340
6,010
3,700
27
**
g
2.0
4.2
4.9
4.0
2.3
E-Col
V
1,620
6.7
«1
«1
«1
i/100 ml
**
g
1.7
3.2
—
—
—
* M = geometric mean
** a = geometric standard deviation
TABLE 33.   PERFORMANCE OF BIOLOGICAL  ACTIVATED  CARBON  FILTERS.  MEAN VALUES
            FOR 6-MONTH OPERATION AFTER  A  3-MCNTH  STARTING  PERIOD  (DOHNE
            PTI n" PIAN~. yniHFTM. RFRMANYI
Activated carbon
type
LSS
LSS
ROW
ROW
NK12
F40D
BKA
bed depth
(m)
2.5
5.0
2.5
5.0
2.5
2.5
2.5
A DOC
mg/1
0.92
1.69
1.09
1.59
0.99
1.26
1.00
A (inorg C)
mg/1
0.83
0.96
0.97
1.05
1.36
1.11
0.97
A NH4+
mg/1
1.31
1.34
1.31
1.34
1.28
1.32
1.28
A02
mg/1
6.32
6.67
6.49
6.71
6.03
6.95
5.99
       Source:   Sontheime£_et_aK,  1978_
                                    154

-------
When the backwashing intervals were reduced to 3 days (just below the time   -
of the reproduction cycle of this genus of microorganisms), the nematodes
disappeared (Heilker, 1979).

     Elimination of breakpoint chlorination at the beginning of the MUlheim
process eliminates formation of chlorinated organics which caused the acti-
vated carbon columns to have to be regenerated every 2 months under the old
process.  The 10 to 50 mg/1 dosage of chlorine previously required for this
step now has been replaced with 3 mg/1 total dosage of ozone, applied in 2
stages.  Additional cost-savings associated with this change include the
labor which was required with breakpoint chlorination.  Formerly, a technician
was required to sample water every two hours and to analyze for chlorine and
for ammonia.  This labor requirement has been eliminated (Sontheimer, 1977a,b).

     Increased algae levels occur in the Ruhr River near MUlheim in the
spring and fall.  These are compensated for at the Donne plant simply by
increasing the preozonation dosage from one to two mg/1  until the algae
bloom has subsided.  The plant operating characteristics remain normal  with
this procedure (Nolte, 1978).

     Because of the success of the BAG process in replacing breakpoint
chlorination at the Dohne plant, the Rheinisch-Westfaiische Wasserwerk-
gesellschaft mbH now is designing BAG into two of its other plants in the
MUlheim area of Germany.  Neither of these plants will employ ground infiltra-
tion of treated water.

     It is important for the reader to realize that Dohne1s raw water (the
Ruhr River) does not contain significant amounts of of synthetic organic
chemicals.  TOG! levels are very low.   In fact, the German government prohi-
bits the discharge of industrial wastes into the Ruhr, because 1t is the
source of drinking water for many cities in the area.  There is considerable
pollution in the Ruhr because of sewage discharges, but the components  of
this generally are biodegradable, or can be made so by oxidative pretreatment.

     As a result, there is only a small chance for non-biodegradable organic
materials to be present after ozonation, and most of the organics entering
the BAG media are expected to be biodegradable.  Therefore, the life of the
BAG columns at the Dohne plant should be at least two years before regenera-
tion is required, based on pilot studies of that length of time.

     At the present time, the criteria to be used to determine when the GAG
is to be regenerated at 3ohne are the breakthrough of DOC and of the organic
materials which absorb at 254 ran in the ultraviolet region.  These criteria
probably will be adjusted on the basis of overall plant performance and the
changes which are noted in the water quality parameters currently being
monitored.  TOC1 analyses are being conducted routinely at Dohne, and it is
possible that this parameter can become a regeneration criterion, if pollu-
tional levels of such halogenated organic materials increase in the future.
                                      155

-------
Rouen-la-Chapelle, France

     At the 50,000 cu m/day (13.2 mgd) plant at la Chapel!e St.  Etlenne de
Rouvray in Seine Maritime (west of Paris near the Atlantic  Ocean), well
waters drawn from near the Seine contain 2 to 3 mg/1 ammonium ion, 0 to 0.2
mg/1 manganese, various micropollutants (detergents, phenols, Substances
Extractable with Chloroform, etc.) and are practically devoid of dissolved
oxygen.  Since 196?, the ammonia content of the raw water has risen from an
average of 0.3 mg/1 to an average of 2.6 mg/1.  This increase required that
the treatment process be improved.  Breakpoint chlorination was  discarded
because it would have required very large contact chambers  (close to 7,000
cu m) and would have produced chlorinated organics which then would have to
be removed.

     After three years of pilot plant testing, the following process (Figure
49) was developed, was installed and began operating in February 1976 (Gomella
& Versanne, 1977; Rice, Cornelia & Miller, 1978):

t    Pre-ozonation (0.7 mg/1) for Mn, organics and adding dissolved oxygen
     to the water

t    Filtration through quartz sand

•    Adsorption in GAC beds

•    Ozonation for disinfection (1.4 mg/1)

»    Post-chlorination (0.4 to 0.5 ng/1)

     This single operation of preozonatlon assures the following:

8    oxygen demands of the materials in water are satisfied,

•    water is oxygenated,

t    complex, biorefractory molecules are broken down and become biode-
     gradable,

•    the content of various micropollutants is lowered,

•    manganese is oxidized and precipitates, to be retained on the sand
     filter so that it does not block adsorption sites on the biological
     activated carbon.

     About 80% of the nitrification occurs in the sand filter beds (100 cm
deep).   Periodic backwashing of these sand filters to remove oxides of
manganese does not upset the action of these bacteria.  Similarly, bacterial
activity on the activated carbon beds (75 cm deep) is not displaced during
backwashing.  The BAC beds are backwashed (by bumping with  air,  then using
water)  once each month, but have not yet had to be regenerated after 2.5
years of operation.  However, one GAC bed was reactivated in early 1979 so
                                     156

-------
    OZOnemnm» I
polluted
well water
T
I

I

I

|
preozonatlon
1
sand
filtration


GAC
adsorption

h

i

I
1

post-ozonatlon
(disinfection)

1

recycle
of
off-gas
ozone
f~j


post -ch lor Inationl
(for residual) I
Figure 49.  The  Rouen-la-Chapelle process,
            Rouen, France.
                    157

-------
as to allow plant personnel to become familiar with the operation (Schulhof,
1979).
     This plant began operating in February, 1976 and showed the perfor-
mances listed in Table 34 for the first two years of operation, respectively.
During the first year the empty bed contact time for the carbon beds was 9
minutes.  During the second year, this was increased to 18 minutes by
decreasing the flow rate of water through the plant.  The percent pollutant
renovals are somewhat better with the longer empty bed contact time.

TABLE 34.  ROUEN-LA-CH"PELLE PLANT OPERATIONAL DATA 0976 AND 1977)
parameter
turbidity
(mastic drops)
ammonia (mg/1
NH4+)
Mn (mg/1)
detergents
(mg/1 DBS)
phenols
(microg/1)
SEC***
(microg/1)
substances
extractable
with cyclo-
hexane
(microg/1)
raw
water
4
1.80
C.15
0.12
6.5
590
1,335
preozo-
nized
-
1.80
0.07
0.09
4.0
470
740
filtered
(sand &
GAC)
-
0.40
0.04
0.06
1.5
250
535
post
ozoni-
zed
2
0.26
0.02
0.03
0
150
410
1
eliml
1976*
—
8656
87*
752
100%
75%
69%
»
nation
1977**
~
95%
90%
75%
100%
75%
70%
av. NH, content of raw water: 0.3 mg/1 in 1968; 2.6 mg/1 in 1975
* EBCT = 9 minutes
** EBCT = 18 minutes
*** SEC = Substances Extractable With Chloroform
Sources-

* Gomel! a & Versanne, 1977
** Versanne, 1978
                                     158

-------
     During a site visitation to the Rouen plant in June, 1977, it was
learned that neither air nor oxygen are added after sand filtration or
before GAC adsorption, and that the dissolved oxygen content of the water
exiting the BAG beds is zero (Versanne, 1978).  This would indicate the
possibility that there may be incomplete nitrification or some denitrification
occurring in the lower part of the 75 cm deep GAC beds.

     Cornelia & Versanne (1977) reported that even though 1 gram of ammoniacal
nitrogen requires 4.57 g of oxygen to be converted to nitrate ion (stoichio-
metry not indicated), only 3.2 g are required at the Rouen plant.  This is
further indication of the occurrence of denitrification processes.  In this
situation, the oxygen source for the nitrifying bacteria would be the nitrate
ion, and as oxygen is removed from the nitrate ions, both nitrite ions and
nitrogen gas can form.

     There should be no dangers from this practice at Rouen, however,
because after BAG treatment, the product water is subjected to ozonation for
disinfection.  The French standard for this process involves first attaining
a dissolved ozone residual  of 0.4 mg/1, then maintaining that 0.4 mg/1
residual ozone concentration for a minimum of 4 minutes.  Under these condi-
tions, Rouen's BAC-treated waters are disinfected, viruses are inactivated,
any nitrite ions formed in the GAC beds are oxidized to nitrate ions and
high dissolved oxygen levels are restored.

     A case history of the Rouen plant has  been published recently by Rice,
Gonrella & Miller (1978).  At least 29 specific organic hydrocarbons and 16
oxygenated organics have been identified in Rouen's raw waters.  Only chloro-
form, carbon tetrachlor1de9 trichloroethylene and trlchlorobenzene were
identified as chlorinated organic impurities.  Rouen does not monitor for
TOC1, but is testing that procedure for possible use as a GAC reactivation
criterion.

     Under these circumstances, it is probable that most of the organic
compounds present in the Rouen raw water are biodegradable after ozonation.
The amounts of identified chlorinated organic compounds are very low.
Therefore, it is to be expected that most of the organic pollutants present
are being removed by the biologically active GAC, and the BAC should have a
long useful life at Rouen.  Nevertheless, the presence of some chlorinated
organic compounds indicates that TOC1 should be monitored frequently, as a
check on the cumulative loading of these materials on the GAC.  When this
practice is instituted at Rouen, perhaps TOC1 will be chosen as the key
regeneration parameter, and the useful life of the BAC at Rouen may turn out
to be shorter than indicated by the parameters listed in Table 34.
                                     159

-------
                                  SECTION 10

 EUROPEAN MICROBIOLOGICAL STUDIES IN ACTIVATED  CARBON  FILTERS  & ADSORBERS

 INTRODUCTION

      Many Investigators  have  observed  the  rapid  growth of bacteria  in
 activated carbon media used as  filters and/or  adsorbers in drinking water
 treatment (McCreary  & Snoeylnk,  1977,  references cited therein; references
 cited throughout this sub-section).  For purposes of differentiation, the
 term  "activated carbon filter" will  be used to describe the use of activated
 carbon  in place of sand  filters, when the  primary function of the activated
 carbon  is as a filtration medium.  On the  other hand, the term "activated
 carbon  adsorber" will be used to describe  the same activated carbon unit
 after sand or dual media filters.

      High levels of  bacteria also have been observed in the effluents from
 6AC filters and adsorbers used for periods of more than a few months.
 Several pertinent observations in this regard were presented at the Conference
 on Activated Carbon  in Water Treatment held in England during April, 1973
 and sponsored by the Water Research Association, Medmenham,  England.  The
 Proceedings of this conference provide valuable information  regarding the
 status of  activated carbon used for treating drinking water  supplies at that
 time.   In  this sub-section we will discuss the status of  European microbiolo-
 gical  studies as it existed in 1973, as reported in five  presentations  made
 at that conference.   Next, more recent European studies will  be discussed.
A portion of these later studies has been presented briefly  1n Section  6.

 EUROPEAN STATUS AS OF  1973

     Melbourne & Miller  (1973) reported details of an extensive study
 conducted over the period 1968 to 1971  at the Colwick (England) plant which
 treats River Trent water.  The plant treatment process consisted of biological
 sedimentation (for ammonia conversion), chemical  addition, sedimentation,
 anthracite/sand filtration,  GAC adsorption and post-chlorination.   During
 this study the behavoir of the plant stream was compared  with that of a
 second stream which was pretreated by chloHnation or caustic soda softening.

     Over  the 3-year period of this study-  it was found that the GAC adsorbers
 removed about 80% of the bacteria present 1n the plant stream entering  the
 carbon beds under "normal conditions" (without prechlorination or presoften-
 inq)   However, when prechlorination or presoftening were practiced, signifi-
 cantly Increased growths of organisms across the GAC adsorbers were observed.
 On the average, waters from the sand filters containing no col1forms/lOO ml,
 zero  E. coli/lOO ml  and no 22'C plate counts/ml showed an average of about
                                     160

-------
 10 coliforms/100 ml, zero to 9 E.  Coli/100 ml,  but 1,000 to 8,500 22°C plate
 counts/ml  in the GAC effluents over an 8-week period (Table 35).

      On the other hand,  even when  these high  levels  of organism growths
 occurred,  final  chlorination successfully eliminated them,  except in  a few
 instances.

   -  Knoppert &  Rook (1973)  had studied the treatment of River  Rhine  water
 at the Rotterdam (The Netherlands)  Waterworks with granular activated
 carbon over the  period 1970  to 1972.   At the  time, this  plant used the
 process consisting of microstraining,  breakpoint chlorination,  iron (III)
 coagulation,  powdered activated  carbon,  flocculation,  sedimentation and
 rapid  sand  filtration.

     Knoppert &  Rook  (1973)  stated,  "The only objection  to  locating GAC beds
 at  the  end  of the purification process  is  the generally-known bacterial
 growth  in the beds, which may give  rise to  high bacterial counts  in the
 delivered water".  Therefore these  investigators conducted  a program  to
 study the possibilities of diminishing this bacterial development by  backwash-
 ing the GAC beds frequently.

     Two equivalent GAC columns were operated in parallel and threshold odor
 numbers were monitored in the effluents.  One column was backwashed daily
 and the other twice monthly.  The column which was backwashed daily ran 52
weeks before taste breakthrough occurred.  During this time, the measured
 22°C bacterial count  levels rose to 10,CCO/ml  after 8 weeks, then diminished
 steadily to about 25/ml by the end of this run (52 weeks).

     The second GAC column, bctckwashed twice monthly, ran only 40 weeks
before taste breakthrough occurred.  Bacterial counts in the effluents also
 rose to 10,000/ml after 8 weeks of operation,  decreased to 100/ml  after 24
weeks and then rose to nearly 1,000/ml at breakthrough (40 weeks).

     The 37°C colony counts of both GAC beds remained at levels  less  than
 10/ml throughout the test period.  The 22°G colony counts of both GAC bed
 effluents were easily controlled by post-chlorination.

     Ford (1973) reported studies conducted at the Foxcote Treatment Works
 of the Bucks Water Board (England), where GAC was introduced into the plant
 process for taste and odor control   in 1960, after an 18-month pilot plant
 study conducted during 1957 and 1958.  The treatment process at this  11,400
 cu m/day (3 mgd) plant involved coagulation (ferric sulfate or alum),  break-
 point chlorination (to 0.5 mg/1 free residual  chlorine), rapid sand filtra-
 tion, GAC, rechlorination and ammoniation.  The GAC adsorber bed was  0.91 m
 high x 2.44 m in diameter (volume 34.2 cu m) and had a 97 second actual
 contact time  (assuming 30* interstitial voids in the carbon), or a superficial
 contact time  of 325 seconds at a filtration rate of 2.8 mm/sec.




                        r a

-------
CT(
ro
1ABLE 15 KtMOVAl Of MCHHIA OUHIW. LA SI 1C SOUA SCFlLnlnG ADO l-KlUllOKItlAT ION AT COIWICK. MIAhL
send filtered""
f 1
toll"
foms/
100.1

500
1.700
1.200
1.400
1.000
3.600
2.600
4,400
340
IkO
820
2.300
2.100
2.300
100

400
220
1,940
960
.600
.240
.500
580
.400
.600
.200
.040
500
000
.620
760
t coll
per 100"
•1

20
35
150
50
455

45
J60
10
4
50
156
90
20
a

a
nil
24
20
130
24
24
240
124
120
42
44
4
168
4
10
22'£ plate
cowit/100
•1
far
1 ,500
1.340
3.900
2.460
2.400
5.400
410
2,500
1,300
BO
60
8511
>5.000
2.800
610
fan

190
920
80
6.000
440
HO
3.200
4,000
400
370
3,900
820
6.400
1.46(1
1.200
G 1
coll
foras/
100 Hi
fc sulf
900
580
490
290
290
550
270
480
40
BO
60
284
550
130
40
1C tulf
360
120
1,080
100
1.390
140
ISO
560
10
110
bb
45
25
60
3.000
80
r coll 22"C plate
per 100 Icount/lOO
•1 1 Hi
te coagu
34
24
8
4
6

2
2
3
1
4
56
34
a
2
te coagu
10
nil
20
5
46
nil
4
30
nil
16
2
nil
nil
nil
38
5
atlon
780
520
730
226
70
M
72
320
21 B
43
55
240
580
370
260
acton

10
850
40
<10
31
89
1.000
360
600
68
380
30
140
>5.000
500
HI (2
coTl
fonts/
100.1

nil
nil
nil
nil
nil
230
nil
1
nil
nil
nil
nil
nil
•11

nil
nil
•11
nil
>500
nil
aa
35
•11
a
nil
nil
38
27

nil
F coll
per TOO
Hi

nil
nil
nil
nil

2
nil
nil
nil
nil
•11
nil
nil
nil

nil
nil
nil
•11
21
nil
7
12
nil
•11
nil
nil
•11
7

•11
TPC. p'ate
count/100
•1

1
5
nil
nil
nil
60
nil
nil
7
2
nil
e
3
2


nil
nil
nil
850
nil
98
800
2
I
nil
Ml
30
6

4
foil
foras/
100 ail

5,000
350
100
680
1.200
1,700
1.160
-lion

700
>5.000
3.000
>5.000
8,600
>2.000
6,000
8.500
>1,200
>3.000
2.400
700
3.200
3.600
3.000
disinfected
« J
coll
fonas/
100 »1

bo
nil
nil
14
1
220
• 11
250
220
nil
1
37
100
9
nil

19
4
nil
nil

nil
3
nil
nil
500
4
at)
3
nil
1
nil
bfTTO
nl

nil
nil
nil
nil
nil

nil
15
a
nil
nil
6
11
3
nil

nil
nil
nil
nil

•II
nil
nil
nil
nil
nil
•I)
nil
nil
nil
nil
22'C plate
count/ 100
al

320
12
nil
250
nil
1.200
nil
nil
>5,000
2
126
156
280
210
150


274
29
nil

nil
>2.000
560
1.600
>1.200
>3.000
nil
630
nil
nil
1

week
no


19
20
20
21
21
22
22
23
23
It
24
25
25
26
27

31
32
32
31
33
34
34
35
35
36
37
37
38
38
39
39
Source Melbourne t Miller. 1973

-------
nated and ammoniated) "over the 5-year period 1968 to 1972.   In 71% of the
cases, the GAC effluents showed higher plate counts than did the  sand fil-
trates, and in 42% of the samples the GAC effluent count exceeded that of
the sand filtrate by a factor of 10 or more.  Rechlorination and  airmoniation
to a level of 0.45 mg/1 of chloramine, did not restore the  bacterial  counts
of the GAC effluents to the levels of the sand filtrates.
          90-
final water-
non-chlorinated
 431 samples
sand filtrate
 495 samples
GAC filtrate
 446 samples
              0- 111- \101-\500+l 0- 111- \101-\500+10-
              10 \100\500\     \10  BOOBOO\     \IO
   11- \101-\500 +
   00 BOO
                      plafe count  colonies/cm3
          Figure 50.  Histogram of agar plate counts,  3 days at 22'C.

                             (Ford. 1973)

     The Water Research Association (England)  examined  samples of  Foxcote
water and tentatively identified the predominant  microorganisms in the  GAC
filtrates as chlorine-forming Flavo bacteria,  along  with some spore-fonr.ing
bacteria.  Their work also demonstrated that three days  incubation is too
short a period of time for the development  of  easily visible colonies of
these bacteria8 and that much higher counts were  obtained after 7  days
incubation.  Using 7 day incubation times,  bacterial  counts obtained with
sand filtrates remained very low.

     The occurrence of biological  activity  in  the GAC adsorbers and their
effluents in spite of approximately 0.5 mg/1 free chlorine residuals in the
influent waters thus was amply demonstrated.

     Ford (1973) also noted that "animal  infestation problems have occurred
when (activated) carbon units have been taken  out of service ?nd,]eft *tan-
ding".   Apart from this aninal infestation  tendency  to  occur in idle filters,
                       	163

-------
 Nais worms were found on  only one occasion  In  the  Foxcote activated  carbon
 effluents.  These were eradicated by addition  of NaOCl  to the  backwash water
 to give a chlorine residual  of 5  mg/1.

      Richard (1973)  reported on treatnent studies  of Seine River water at
 the Vigneux plant upstream of Paris,  France.   At this 1,500 cu m/hr  (36,COO
 cu m/day;  9.5 mgd)  treatment plant comparative tests were conducted  using
 powdered- and granular activated carbons.  The  Vigneux plant process  at the
 time consisted  of microstraining,  clarification (with powdered activated
 carbon),  filtration,  breakpoint chlorination (with 6 to 8 mg/1 of gaseous
 chlorine),  coagulation (alum)  and  flocculatlon (activated silica).   At the
 time of the reported  study,  10 to  20 mg/1 of powdered activated carbon was
 being injected  upstream of the clarifiers.  Chlorine dioxide (0.15 to 0.25
 mg/1) was  injected upstream  of the reservoirs.

      Three  test processes were studied by Richard  (1973):

      1)   clarification with powdered activated carbon,  sand filtration;

      2)   clarification without powdered activated carbon, 3AC filtration;

      3)   clarification without powdered activated carbon, sand filtration,
          GAC adsorption.

      After  250  days of operation, a monitor was installed which controlled
 the  breakpoint  chlorination step to a free residual chlorine level  of 0.25
 to 0.40 mg/1 from that point on.  Under these conditions the pretreated
 plant water gave negative coliforn (24 hours at 37°C)  and E_.  coli  counts.

      Treatment without either  powdered or granular activated carbon produced
waters having very low plate counts.   However,  addition  of powdered activated
 carbon to the clarifier produced filtered water showing  less than  10 colonies/-
ml during cold weather and 10  to 100 counts/ml  during  summer.   Passage
 through granular activated carbon, acting either as a  filter or as  an adsor-
ber,  gave colony counts of about the same order, although slightly higher.
Final sterilization with chlorine dioxide (0.15 to 0.25  mg/1  dosages) produced
waters having zero plate counts/ml.

      Kfllle & Sontheimer (1973) discussed experiences with activated carbon
 in West Germany and specifically reviewed the role of  biology in water
 pretreatment.  They proposed that the action of microorganisms in  granular
activated carbon media is  integrally connected  with the  enrichment  of organic
substances  (by adsorption) which can result in  improvement of substrate
utilization by the bacteria.   As waters containing  organic materials  pass
through granular activated carbon columns or beds,  some  organics are  adsorbed
physically and/or are incorporated into the biomass and/or are oxidized to
carbon dioxide.   In addition,  products and intermediates of  biological
assimilation had been observed (Koppe & Giebler, 1966)  in the effluents from
activated carbon media.

      KBlle & Sontheimer suggested several experimental approaches  to obtain
more  detailed information on the operable physical  and  biochemical  processes


                            .164

-------
occurring 1n activated carbon filters and adsorbers.   These media can be
operated and their efficiencies in eliminating organic materials from solution
can be studied after exhaustion of their theoretical  adsorption capacities.
However, in operational water treatment plants, steady state conditions are
difficult to establish, even over long operating periods, because compositions
of raw waters being treated usually vary from day to day.  Organic conpounds
which are rapidly adsorbed and tightly bound to the activated carbon surface
will displace previously adsorbed organic compounds which are not as tightly
held by the activated carbon surface.

     A second approach is to follow, simultaneously,  both the decrease in
dissolved oxygen consumption and increase in carbon dioxide produced during
passage of water through the activated carbon medium, comparing data treasured
in both the influents and effluents.  The activated carbon filter or adsorber
also can be used as a "bioassay bottle" by recycling effluent to become
influent and following the changes in CO consumption and C02 production.
Studies of this type were being conducted at the Auf-dem-Werder plant in
Bremen, Federal Republic of Germany.  Oxygen consumption upon passage through
activated carbon rose to 7 mg/1 during summer and dropped to 2 mg/1 during
winter, showing that surface water temperature variations affect biological
activities significantly (Eberhardt, Madsen & Sontheimer, 1974).

     Another factor complicating such studies is the kind of pretreatirent
applied before activated carbon treatment.  Kfille & Sontheimer (1973)
state, "There is practically no waterworks without pretreatment -- irostly
using oxidation by ozone".  Some laboratory tests have shown that it is
desireable to add a flocculation step between oxidation  (with ozone) and
activated carbon filters or adsorbers, especially with respect to rerroving
UV-absorbing organic substances (Kdlle & Sontheimer, 1973).

Summary of European Studiesin J973

     The status of published European knowledge of microbiological activity
in activated carbon filters and adsorbers used in drinking water treatment
as of 1973 can be sunirarized as follows:

1)   Bacterial growths in activated carbon were known to occur in plant
     scale media to a significant extent, even when influent waters contained
     as high as 0.5 mg/1 of free residual chlorine (after breakpoint chlori-
     nation).

2)   Biological activity develops rapidly with fresh carbon charges.  GAC
     units used for taste and odor control over a 1-year period developed
     maxima of 10,000 colony counts  (22°C)/ml in the effluents in about 8
     weeks.  During this time the 3AC influent waters contained zero colony
     counts/ml.

3)   Prechlorination or presoftening of  influent waters  had been observed  to
     increase the amount of biological growth in the GAC ledia in some
     cases.
                                     165

-------
4)   Effluents from biologically active GAC media contain a much greater
     number of colonies than do effluents from slow sand filtrates.

5)   Treatment of effluents from biologically active carbon media with
     chlorine or (0.15 to 0.25 mg/1  dosage) chlorine dioxide lowered colony
     counts/ml to acceptable local  biological standards.

6)   Speciation studies of microorganisms in effluents from biologically
     active carbon media had shown  the presence of chlorine-resistant
     flavobacteria (from waters which had been treated by breakpoint chlorina-
     tion) along with some spore-forming bacteria.

7)   Indications were that two and  three day incubation periods were insuffi-
     cient, even at 37°C, for development of easily visible colonies, and
     that 7 days of incubation gave much higher counts.

8)   Animal infestation had been observed in activated carbon units  which
     had been taken out of service  and left standing for two days.
                                                                  4
9)   Bacterial activity was higher  in the summer than in winter.

10)  Products and intermediates from biological assimilation had been
     observed in effluents from activated carbon nedia.

CURRENT EUROPEAN MICROBIOLOGICAL STUDIES

     At present, there are only two European groups known to have published
results of continuing research into the microbiological aspects of activated
carbon systems with respect to treatment of drinking water.  These studies
are being conducted at the KIWA (The Netherlands) by Dr. D. Van der Kooij
and at the University of Saarlands  (Federal Republic of Germany) by Professor
Dr. R. Schweisfurth and his students.  A third research program on this
subject has been established recently at the Engler-Bunte Instltut der
UniversitSt Karlsruhe during 1979.

Research Studies At The KIWA

     Microbiological studies at the KIWA were published by Van der Kooij in
1975 and 1978.  Additional unpublished information was provided by Dr. Van
der Kooij In a personal interview during June 1978.  These two published
articles will be reviewed in this subsection and the discussion will be
supplemented by the private communications supplied during June, 1978.

     In work presented in 1975, Van der Kooij conducted s1de-by-s1de experi-
ments with Norit ROW 0.8 Supra (the 0.8 refers to the particle size in
millimeters) granular activated carbon (GAC), the same granular carbon but
not activated (GNAC) and sand (0.85 to l.CO mm particle size).  Test columns
containing each of these materials  were fed with non-chlorinated tap water
(13 to 17°C) at the rate 3.5 m/hr (3 minutes apparent contact times) over a
period of 10 months.  Colony counts in the columns were deternined at inter-
vals on diluted nutrient agar (0.36 g/1 beef extract, 0.65 g/1 peptone, 10
g/1 agar) after 10 days of incubation at 25°C.  The bacteria were removed

                                     166

-------
from the sand,  GNAC and GAC  by  subjecting the samples to ultrasonic energy.
The numbers of  bacteria found on  the filtering materials were expressed as
numbers/ml of filter volume  and are shown in Figure 51.

     This figure shows  that  the numbers of bacteria found on the activated
carbon were always about 10  times higher than the numbers on the non-activated
carbon and on sand.  Using more re*ined Tieasurements, Dr. Van der Kooij
(1978a) showed  that the numbers oc bacteria on activated carbon were only 2
to 3 times higher than  on the other two media.
                n
                 5s
                 3 O
                 O >
                 e —
                 o «-
                 "3
                 o
                         108
10'
                         10
                         10"
                           6.
                            02   4  6   8  10  12  14
                                    months  »-
                  ""•"" activated  carbon, Norit ROW  0.8  supra

                  o.-..^ non-activated carbon, Norlt ROW 0.8

                  •«!•».. sand, 0.85-1.00 mm
                 Figure  51.  Bacterial  numbers on GAC,  granular
                             non-activated carbon and sand, each
                             fed with non chlorinated tap water.

                                  (van  der  Koolj, 1975)
     In a second experiment, samples of used  SAC  and 3NAC  (ROW 0.8 Supra)
were taken from adsorbers fed with prefiltered  river water  for a period of
one year.  Wet carbon sanples (150 ml)  were treated with 50 ml of non-
chlorinated tap water and aerated at 25°C.  The numbers of  conforms,
pseudomonads and actinomycetes proliferating  in these  carbons were determined
weekly.  The total numbers of viable bacteria were estimated by colony
counts on diluted nutrient agar and expressed per cu m of  filter volume.
Oxygen consumption was expressed in mg  of 02/1  of filter volume/hr.  Results
are plotted in Figures 52 and 53.  Tab'e 36 shows the  times to attain 50%
reductions in the levels of bacteria present  in the activated and non-
activated carbons (5C% reduction times).

     Van der Kooij (1975) estimated that since  the hypothetical cylindrical
surface area of granular activated carbon is  about 40  sq cm/cu cm of GAC
present, when the colony count is lO^/cu cm,  then a single  bacterium occupies,
                                     167

-------
on the average,  40 sq  microns  of  this available surface area.  Since the
available surface area of GAC  is  much higher than 40 sq cm/cu cm, it was
concluded that there is only a low  density of bacteria occupying the surface
of the activated carbon, even  allowing for the fact that colony count techni-
ques measure only a portion  of the  bacteria present.  Scanning electron
microscope measurements made at the KIWA confirmed this conclusion, as did
similar studies  by Weber, Pirbazari  & Melson (1978) on GAC samples contacted
with chemically coagulated and settled sewage.
                     10
                           10 20  30 40 50 60  70 80
                                days »•
               Figure  52.  Numbers of pseudomonads (•«) and
                           co'iforms UA) or. GAC ROW 0.8 supra (*
                           and on non-activated carbon  ROW 0.8
                           (OA) during  aeration at  25'C.

                                (van der Koolj, 1975)
TABLE 36.  50% REDUCTION TIME (DAYS) OF DIFFERENT  GROUPS OF BACTERIA ON
               ROW 0.8 SUPRA SAC AND ROW 0.8 GNAC
Type of Bacteria
Colony Forming Units
(25°C, 10 days)
Col i forms
Pseudomonads
Actinomycetes
ROW 0.8
Supra
9
3
4.5
>50
ROW 0.8
not act.
8.5
3.5
4.0
>50
Source; Van der Kooij, 1975.
                                     168

-------
     Based on these ireasurements, Van der Kooij (1975) concluded that
adsorption processes operative  in granular activated carbon are not hindered
by the presence of bacterial growths on the activated carbon.

     Figure 51  also shows  that  the number of bacteria on GAC samples decreased
after reaching  a maximum value.  Van der Kooij (1975) concluded that this
decrease is not caused  by  exhaustion of the adsorptive capacity of the SAC,
since the same  relative decrease also was noted with 3NAC and sand.  Van der
Kooij points out that the  decline also might be explained by a shift in the
types of bacterial  flora originally present to types of bacteria which do
not participate 1n colony  counts.
              a
              o
              co
               a
               o
               o
               c
               o
                   io8i
                   10".
               »   103
                                          !„••«««*•
100
                                                  0.1
0.01
                                                 0.001
cs
O
o>
e
c"
o
+•*
a
E
3

-------
10/cu on of filter volume and collforms dropped to 10/cu cm in 23 days from
100 to 1,000.

     On the other hand, the numbers of actinomycetes and the oxygen consump-
tion rates did not show clear decreases during 60 days.   Therefore, from the
data of Figure 52 and of Table 36,  it can be seen that the conforms disappear
faster than do other bacteria.  Since there is hardly any difference in the
50% reduction tiires for the selected bacteria on GAC versus GNAC, Van der
Kooij (1975) concluded that neither coliforms nor pseudomonas nor total
colony-forming bacteria gained any advantage from the organic substances
which were present on the activated carbon which had been in contact with
river water for the one-year period.

     The oxygen consumption rate during passage through either GAC or GNAC
did not decrease with decreasing colony counts, nor did the number of actino-
mycetes (Figure 53).  Van der Kooij (1975) therefore suggested that it is
possible that the oxygen consumption values observed with both carbons were
caused by chemical rather than bacterial processes.  He also suggested that
the relatively constant numbers of actinomycetes bacteria observed over the
60 day period might have been due to their ability to fragment into smaller
organisms (to produce spores).

     Curing June, 1978, Dr. Van der Kooij supplied the following additional
information regarding his studies of biologically active filtration and
adsorber media:

1)   For sampling of filtration of adsorption media, 1 g samples of wet
     medium are taken into sterile, non-chlorinated tap water, then subjected
     to ultrasonics for 2 to 3 minutes.  The supernatant liquid then is
     treated by normal dilution and plate counting.  When water was added to
     the 1 g activated carbon sample treated once by ultrasonics and the
     ultrasonic exposure repeated, more bacteria were found in the second
     supernatant.  The process was repeated 10 times on the same 1 g sample
     of activated carbon.  About 40% of the total bacteria determined were
     found in the first supernatant, 20% in the second, 10% in each of the
     third and fourth extracts, etc.  Even after the tenth ultrasonic treat-
     ment of the original 1 g sample, significant bacterial counts were
     being obtained.  This indicates that bacteria are very tightly held by
     the activated carbon.

2)   Samples from operational water treatment plant granular activated
     carbon filters or adsorbers were taken in 10 to 20 g quantities and at
     various depths.

3)   A bioassay test for biologically assimilable organic carbon in raw
     waters has been developed by the KIWA at one of"the large Dutch water
     treatment plants using river water.  Pseudomonads were cultured in raw
     river water at 15°C and their rate of growth was determined.  Next the
     river water was ozonized, then seeded with the sarre pseudomonad culture
     and their rate of growth again measured.  The growth rate was found to
     be higher after ozonation than in the raw river water.  Such a test
     again shows that ozonation converts some of the dissolved organics

                                     170

-------
     present (which are not easily biodegraded)  into organlcs  which are more
     readily biodegradable.  In addition,  this procedure might be considered
     as a screening test at U.S. water treatment plants to determine whether
     detailed evaluation of an early stage biological processing step would
     be worthwhile.  If no increase in biodegradability is observed after
     ozonation, then preoxidation of any sort will  be ineffective in this regard.

4)   Fresh samples of GAC have shown ireasureable levels of oxygen consump-
     tion before measureable levels of bacterial activity have built up in
     the medium.  Therefore, GAC seems to  cause  some degree of surface
     catalyzed oxidation of dissolved organic substrates.

5)   £. coli bacteria taken into a biologically  active granular activated
     carbon filter or adsorber cannot survive in competition with the other
     species of microorganisms present.   Therefore, in a properly sized and
     operated biologically active filter/adsorber medium, no E. coli bacteria
     are found in the filtrates.  This is  confirmed by long term experiences
     1n Europe with slow sand filters.

     In a paper presented in 1978(b), Van  der Kooij reported continued
studies of the microbiological processes occurring in granular carbon
samples.  In introducing the subject, Van  der Kooij reported that "some
investigators have concluded that an increased contact time between organisms
and adsorbed organic substrates is allowing the  microorganisms to adapt to
the less readily biodegradable organic substances", thus allowing these
harder-to-decompose organic compounds to be degraded during "biological
regeneration" of the activated carbon adsorption sites.

     In this more recent work, 6 cm diameter columns were filled with Norit
ROW 0.8 Supra activated carbon (GAC), the same carbon but not  activated
(GNAC) and 0.8 to 1.0 mm sand.  Again in s1de-by-s1de tests, the columns
were treated with non-chlorinated tap water (14  to 18°C) containing an
average of 3 mg/1 of dissolved organic carbon at 3.5 m/hr filtration rates
(3 minute apparent contact time) for a period of one year.  Numbers of
colony-forming units/ml were estimated by the surface spread technique on 8-
fold diluted Lab-Lemco broth (Oxford CM 15) agar plates.  Samples were
exposed to ultrasonic energy for 3 minutes to detach the microorganisms,
then samples were incubated for 10 days at 25°C.  Removal of organic substan-
ces from solution was followed by measuring the ultraviolet absorption at
275 nm in 5 cm cuvettes.

     Figure 54 shows the colony counts on the filter materials and in their
filtrates measured over the one year period.  Maximum values were attained
in all three media during the first 20 to 30 days.  After these naxima were
reached, the adsorber effluents showed counts of l.OCO to 10,OCO/ml.  Ultra-
violet absorption measurements showed that immediate breakthrough of organic
materials occurred with the GNAC and sand media.  The GAC medium reached 80%
of breakthrough  in 30 days and 90% of breakthrough in 9C days, with respect
to dissolved organic Tiaterials.

     Figure 55 compares colony counts found on  the three filter materials, -.
which  usually were higher on the GAC medium and reached a maximum of 7 x 10

                                     171

-------
E
^
3
•^
U
                                   v
                  .•.in QAC filter
                    ln GNAC  ••
                 MO In sand  "
                         In GAC effluent
                      "»ln GNAC
                     ••••• In drinkinc  weter
                                                               	  X
     10
     10
     10
        0  20  40  6O  80 1OO 12O 140  16O  180  2OO 22O 240 260 280 30O 320340
        Figure 54.  Colony counts In GAC, GNAC and sand filters and In their
                   effluents over 340  days.
                             (van der  Kooij, 1978)

-------
cfu/ml.  Colony counts on GNAC and sand were similar to each other.  Figure
56 shows that the cfu/ml numbers found in all three effluents were about the
same.

     Van der Kooij (1978b) concluded that adsorption of organic materials by
activated carbon therefore is not the cause of the high colony counts usually
observed in the GAC filter/adsorbers.  This was confirned by noting that the
majority of microorganisms isolated from the GAC were able to grow only on
simple, non-adsorbing compounds like acetate, pyruvate and "actate, whereas
adsorbing substances such as aromatic compounds (which are ireasured by
ultraviolet absorption) were not utilized.

          CH3COO"        CH3COCOO"      CH3CH(OH)CCO~

          acetate        pyruvate          1actate

     Effluents from the slow sand filters (which are known to remove bacteria)
at the water treatment plant at the Hague (0.3 m/hr flow rate) were studied.
Bacterial contents in these slow sand filters were 20,000 to 30,COO cfu/ml,
but were less than 100 cfu/ml in the filtrates.  Similar observations had
been made earlier by Schmidt (1963).  Comparing these slow sand filtrate
counts with those of the side-by-side experiments (Figures 55 and 56) sugges-
ted that there was a relationship between flow rates and the nurrber of
micro'organisms in the sand and in the filtrates.  Therefore, colony counts
in the GAC (Norit PKST) present on the slow sand filters of the Hague (probab-
ly a thin layer on top of the sand) were estimated in samples taken from 8
of the operational plant filters (Figure 57).

     Comparison of the data of Figure 55 with those of Figure 57 shows that
colony counts in samples of 3AC taken from the top of the slow sand filters
were 1 to 2 orders of magnitude lower than those observed in the experimental
GAC filters supplied with non-chlorinated tap water.  This confirmed the
conclusion that flow rate through a filter bed strongly affects colony count
levels in the filter materials.

     Since the colony counts in 3AC media usually were greater than those in
GNAC and in sand, even though colony counts in the filtrates did not differ,
Van der Kooij (1979) concluded that this behavior probably was due to the
relatively large surface area/unit volume of 3AC upon which microorganisms
utilizing substrate from the passing water can attach.  Therefore, Van der
Kooij  (1979) concluded that granular activated carbon is a favorable material
for biological filtration processes.

Effects of microorganisms on adsorption of organlcs by GAC—
     Some laboratory experiments next were conducted by Van der Kooij (1979)
to determine the influence of bacterial cells attached to GAC on the adsorp-
tion of 4-nitrophenol  (a readily adsorbed but difficult to biodegrade organic
compound) and 4-hydroxybenzoate anion  (a readily adsorbed and readily biode-
gradable organic compound) by Norit ROW 0.8 Supra granular activated carbon.
                                     173

-------
                                   A_» GAC

                                   a— GNAC

                                   • «. sand
     10
        O   20  40   60  80  100
           cum. %  of  samples
         c

         3
         •fr-
         it-
         41

         "o

         E

         3

         O
              O    20   40   60   80   100
               —*-  cum % of samples
Figure 55.  Colony counts In  GAC, GNAC
            end sand.

              (van der Kooij. 1978)
Figure  56.  Colony counts In  effluents from
           GAC, GNAC and sand filters.

             (van der Kooij, 1978)

-------
          4-n1trophenol
                                                     COO"
                    4-hydroxybenzoate anlon
Adsorption isotherms and  rates  of adsorption of these compounds were deter-
mined with growths of the bacteria Pseudomonas fluorescens (Strain 17)  or
Pseudomonas alcaligenes  (Strain 131) being present on the GAC.
               o
               <
               
-------
     To  cultivate  bacteria on the GAC, ammonium acetate was added to the
 sterilized  solution  to  a final concentration of 10 mg/1 of acetate carbon.
 Bottles  were  inoculated with either Strain 17 or 131 and incubated 3 days at
 25°C in  a rotary shaker (120 revolutions/minute).  The bacteria developed to
 a maximum level of 4 x  107 cfu/ml of medium and the GAC contained about 2 x
 108 cfu/ml  (6 x 108 cfu/g) of GAC.

     Under  these conditions, either 4-nitrophenol or 4-hydroxybenzoate was
 added from  sterilized solutions in final concentrations of 100, 50, 25 and
 10 mg/1.  These compounds also were added to bottles containing sterilized
 GAC without bacteria.  The bottles were placed in a rotary incubator at 25°C
 and the  concentrations of 4-hydroxyphenol and 4-hydroxybenzoate were measured
 by UV absorption at 254 and 269 nm, respectively in 5 ml  of membrane filtered
 samples  after 24, 48 and 144 hours of incubation.  Ultraviolet absorption
 was measured every hour during the first 8 hours after addition in bottles
 containing  an initial adsorbate (phenol or benzoate) concentration of 100
 mg/1.  All  experiments were performed in duplicate.

     Adsorption isotherms of 4-nitrophenol and 4-hydroxybenzoate on the
 activated carbon in the presence and absence of bacteria  were calculated
 from measured concentrations and are presented in Figure  58.   The disappear-
 ance of  these compounds from solutions initially containing 100 mg/1  of
 adsorbate is shown in Figures 59 and 60.   Figures 58, 59  and 60 show that
 the phenol was better adsorbed by the GAC than was  the hydroxybenzoate.
 With both compounds, however, the adsorption equilibrium  was reached  within
 48 to 144 hours.

     Results shown in Figure 58 indicate that the adsorption isotherms  are
 not affected by the presence of bacteria.  Moreover, adsorption of the
 adsorbates  in the 100 ml bottles also was not affected by the presence  of
 bacteria on the activated carbon.  The adsorption rate of 4-hydroxybenzoate
 in the presence of bacteria could not be calculated because both bottles
were infected by adsorbate-consuming microorganisms.  Infection also  occurred
 in some other bottles containing 4-hydroxybenzoate  because this compound is
 so easily biodegraded, compared with 4-nitrophenol.

     Van der Kooij (1979)   concluded the following:

 1)   The number of microorganisms present on a filter/adsorber medium
     depends upon the flow rate of water through the medium,  and is probably
     due to limited transport (diffusion) of the soluble  substrate to the
     microorganisms.

 2)   Adsorption of organic compounds by GAC is not Inhibited by the presence
     of a large number of bacteria on activated carbon.

 3)   Granular activated carbon adsorbers  or filters in operational  drinking
     water  treatment plants sampled by Van der Kooij always contained lower
     colony counts than the levels applied in the laboratory experiments
     described in this paper.  Therefore, in GAC adsorbers used to prepare
     drinking water, hindrance of adsorption of organic compounds by  micro-
     organisms is very unlikely.


                                    176

-------
2 6-
2.4-
2.2-
2.0-
1.8-
1.6-
1.4-
1.2-
« 1.0
o-
2 0.6
0.4-
0.2-










4 nltro phenol „.-««»•—••""""""""""""'
	 J.. 	 "-1Do>
-------
on
so
ng
10
E
0

«
u
"c
9
         100-
          90
          80-
          70-
          60
          50-
          40
          30-
          20
          10-
                                4-hydroxy benzoate
                                         ^ssczry—*	<
                              A GAC  without bacteria
                              A GAC  with bacteria
                         234

                          time, hrs
         Figure 59. Removal of 4-nltrophenol and 4-hydroxy ben-
                    zoate on passage through ROW 0.8 Supra
                    GAC.

                          (van der Koolj.1978)

     On the other  hand, Van der Kooij  (1979)  points out that hindrance of
adsorption  by microorganisms might occur when the influent  water contains
relatively  large amounts of easily blodegraded materials.   In these situations
"extremely  large numbers of microorganisms" develop on the  activated carbon.
Also, adsorption may be affected by the contamination of granular activated
carbon by colloidal and suspended matter.  These last two effects were not
investigated during the 1979    study  of Van der Kooij.
                                    178

-------


• 1 «
olo
i '
H °
olu
9
0
c
a
a
0
a
a
a
a
a
5
o
o
«*
a
0.8
0.7-
0.6
0.5-


0.4-



0.3-



0.2-


0.1 -I
                      \
                               a  GAC without  bacteria

                               •  GAC with  bacteria
                    \      T:
                             \   4-nltrophenol
                          X
    '*V,
                                 Xx
                                  A
                                                         —^
4-hydroxybenzoate
                012345678
                                time, hrs.


         Figure 60. Rate of disappearance of  4-nitrophenol  and
                    4-hydroxybenzoate on passage through ROW
                    0.8 Supra GAC.

                         (van der Koolj, 1978)

Research Studies at the University of  Saarland

     In 1972, a cooperative research program between the City of Wiesbaden
and the University of the Saarlands (Federal Republic of Germany) was  begun
with the objective of studying the microbiology which was present in the GAC
adsorbers at Wiesbaden's Schierstein water treatment plant.  Two doctoral
candidates at the University (P. Werner  and M. Klotz) completed their  Ph.D.
thesis studies in 1979 based on these  studies which were directed by Prof.
Dr. Reinhart Schweisfurth.  These scientists have published three progress
reports of their studies, all of which will be reviewed here.  In addition,
                                   179

-------
all three investigators were interviewed in June,  1978 and provided additional
information on their studies in the form of private communications.

     In work presented in 1975, Klotz,  Werner & Schweisfurth reported
studies at the Schierstein water treatment plant at Wiesbaden,  which had
been conducted since 1972 on plant operating granular activated carbon
columns and pilot plant columns.  Initially, these investigators found that
the determination of colony numbers as  recommended by the DEV (Deutsche
EinheitsVerfahren = German Standard Methods, 1962) (2 days incubation at
27°C) proved to be inadequate, because  only a small portion of  the microflora
present are revealed by this technique.  However,  all nutrient  media tested
showed considerably increased colony numbers after breeding for 7 days, and
this incubation time was used throughout this program and those which these
investigators have reported subsequently.

     The best medium for determination  of microorganisms was found to be P-
Agar which contained few nutrients and  SPC-Agar which was rich  in nutrients
(Standard Methods, 1971).  Both were incubated at 27°C for more than 7 days.
The activated carbon samples were crushed in a simple mixer and the homogeni-
zed material then was processed for the determination of colony numbers in
the sane manner as were the water samples.  The cell numbers (the total of
all living and dead bacteria) were determined by counting microscopically
after collecting on membrane filters and coloring.  Apart from  a few modifica-
tions (not delineated) all other tests  were carried out as recommended by
the DEV whenever possible.

The Schierstein Treatment Process—
     At this plant, Rhine River water,  not river sand bank filtered, is
pretreated directly before ground infiltration to augment groundwater supplies.
The treatment process involves aeration, settling, chlorination, flocculation,
rapid sand filtration, activated carbon adsorption, then ground infiltration.
During the course of this study some of the plant 3AC adsorbers were operated
as long as 3 years without being reactivated.  At the end of this time, 30%
of the influent dissolved organic carbon still was being removed *rom solution
during passage through the GAC media.

     In Figure 61 are shown the mean values of colony numbers determined at
different points in the plant for the period March 1973 to Varch 1974.
There was a decrease in colony numbers  between the Rhine and the entrance to
the activated carbon adsorbers, the largest effect being obtained following
breakpoint chlorination.  However, fresh populations of microorganisms were
formed in the activated carbon media, such that very high colony numbers
were found at the GAC media exits using the 7-day incubation method.  By the
2-day incubation method normally applied for measuring colony counts, as
recommended by the DEV for the control  of drinking water, the water quality
at this point in the process generally was satisfactory.

     Composition of microorganism populations in GAC effluents—Populations
were examined in freshly filled plant GAC adsorbers and in pilot plant test
columns.  The test columns consisted of four successive glass tubes each
with an inside diameter of 4 cm and the total 3AC bed depth was 3.2 m (4 x
                                     180

-------
0.8 m).  Passage of water through each individual  0.8 m deep  column  is
referred to as "a filter step".
            09
            a
u

W
a
S

|

C
O
"3
u
               105J
               104J
               103J
               102J
               101J
t^\

%-" •>  ""  *"
— "5 2
e I ^
K S g
(A
1
G;
2
&C
3
4
II
5
t<
6
"f
T.
9

ollectlng outlet|
                     treatment  steps-
             Figure 61. Mean colony numbers at various points in
                        the  Schierstein plant, March '73-March '74.

                              (Klots et, aj,., 1975)

     The behavior of the microorganism  populations was found  to be the same
in both the pilot plant test adsorbers  and  in the full scale  plant adsorbers.
During the first 20 days the colony numbers  rose linearly and reached a
maximum of 10$ to 10^/ml of waterj,  after which  they  declined  and remained at
a slightly lower level  closer to IG^/ml.   In the early stages of the pilot
plant testing there were large differences  in colony numbers  between the
individual filter steps, but these  disappeared  after about  30 days when the
level stage (with respect to colony numbers) had been reached.  Figure 62
shows the establishment of microorganism populations in the effluents from
                                     181

-------
the pilot plant test adsorbers and Figure 63 shows similar microorganism
establishments in the test  adsorbers themselves.
                                   filter 1
                                   filter 2
                              •mm" filter 3
                               •••• filter 4
                           20     30

                           time, days
40
50
60
          Figure 62.  Development  of  colony counts in pilot
                      plant GAC adsorber effluents  at  the
                      Schlersteln plant.

                            (Klotz  et  Ji.,1975)
     Flow rate studies were conducted in  the  test  columns  using 4, 8 and 20
m/hr water velocities.  The slope of the  microorganism  establishment curves
(Figure 62) during the initial  phase decreased  with  increasing velocity and
with increasing column depth.   The maxima were  more  pronounced as the velocity
decreased.  When the level stage was reached, there  were only slight differen-
ces, however, the lowest colony numbers being observed  at  the lowest flow
rate.

     Microbiological & chemical conditions in a plant operating GAC adsorber—
In Figure 64 are shown the colony numbers (living  bacteria), cell numbers
(living + dead bacteria) and free residual chlorine  contents plotted against
the adsorber depth (after an increasing number  of  filter steps).  The low
levels of colony numbers measured after the first  two filter steps are
                                     182

-------
explained by the low adsorber depth and the lack of sufficient contact  time
of the water passed through the SAC to destroy the  free chlorine  residual.
The largest increase in colony numbers took place between  steps 2 and 3
(after the free chlorine residual  dropped to less than  0.1  mg/1), and there
was no further increase between step 4 and the outlet.
            10
        o
        QL
        in
        o
        a
        ^s
        in
        E
        3
        C
        o
        u
           104-|
           103-|
           io2-]
"""• filter 1
""""*• filter 2
""	 filter 3
•••* filter 4
                      10
                             20
      30
                                            40
                                                    50
                             60
                                time) days
           Figure  63.  Development of  colony counts  in  GAC  of
                       pilot  plant adsorbers at  Schiersteln plant.


                              (Klotz e_t. aL, 1975)


     In the uppermost activated carbon  layers the residual  chlorine content
fell below 0.1 mg/1, where it no longer had any influence  on  the levels  of
living bacteria.  Before the  free chlorine concentration was  lowered by  the
first GAC adsorber, the level of living bacteria  was  low,  although  the total
cell numbers were high.  After the chlorine content was lowered, the number
of living bacteria increased  rapidly, although the total cell numbers remained
relatively constant.

     Changes in dissolved oxygen and carbon dioxide contents  of the water
are indicators of the amount  of microbial  activity, as  shown  in Figure 65.
                                     183

-------
I)
*•*
(B



I

in

v


E
a
c
«t
u
C
a
o
u
    105,
    104-]
    103J
      101-
            ceil number^ ^^.-m.-^
                   
                                                     Figure 65  Oxygen and CO2 contents of GAC

                                                                effluents vs. adsorber depth.
                                                                     (Klotz et. aj,. 1975)

-------
During passage through the GAC media there was  an oxygen  consumption  of
approximately 1.5 mg/1,  most of which occurred  after step 2,  and  a  concurrent
production of approximately 4.5 mg/1 of C02,  which also occurred  nostly
after step 2.
                            TOC.mg/l
              E
              e
              o
              «r
              M
              e
              o
              o
              c
              X
              o
                    0.12-
0.10
                    0.08
O.O6-
                    0.04
                     KMnO4
            ^~-*v
              TOC
11


10 -
   0)
   E
    c
    o
    *rf
    a
    E
    3
    If)
    C
    O
    o
                                                            O
                                                            c
                              £*   JH  *i
                             v  »*  J>
                                 «   «
               Figure 66.  Organics content of  GAC effluents
                           vs.  adsorber  depth.

                                (Klotz £t .ai.. 1975)

     The change in organic content was followed  (Figure 66)  by three parare-
ters.  By UV absorption at 240  nm and TOC analyses,  the content of organlcs
decreased 55% after the fourth  GAC filter step,  and  by about 45% as rreasured
by the KMnO^ consumption.

     Behavior of mlcrobial populations over a three  year period—Figure 67
shows the colony counts/ml at the plant processing inlet (after aeration and
settling), at the 3AC adsorber  inlet (after rapid sand filtration) and  after
passage through GAC adsorbers during the 3-year  period May 1972 through
August 1975.  The numbers 1, 2  and 3 in this figure  indicate changes 1n raw
water qualities which resulted  in distinct changes in colony numbers at the
points indicated.

     It appears noteworthy to the current reviewers  of this  work that there
was an apparent breakthrough of colony counts through the GAC adsorbers
after about three years of use.  This indicates  that there should be a  3AC
regeneration parameter based on colony counts as well as an  organic component
                                     185

-------
CO
en
                   •••••M raw water inlet

                   • ---» sand  filtrate

                   •«•«• GAC  adsorber outlet
               10C
             H. 10*
u>
0)

c
o

"o
u
             4)
             **
             U
             0}

             to
10"




103




102
   10
     1
                          *•••—
          \£\
         ^£X^^.
                 -r^-Mr^
                 r      v /v -i
                                     N.
       m j jason

          J972
                  ^
HV
    M
               I f m a m | j a s o n

                      '73
          I f m a m |  Jason

                '74
I f m a m

   '75
                                                                     I  I
                 Figure  67. Behavior of microbial populations on GAC over 3 years

                            at Wiesbaden, F.R.Germany.
                            (Klotz, Werner & Schweisfurth. 1975)

-------
parameter, at least in plants opting to install  a biological  activated
carbon process involving aeration and chlorination as the pretreatment
steps.  No data are available on microbial breakthrough, however, with
processes using ozonation as the preoxidation step.  This possibility of
bacterial breakthroughs under BAG processing conditions should be evaluated.

     Klotz, Werner & Schweisfurth (1975) also reported that they had determi-
ned large differences in colony numbers, as determined by the 7-day versus
2-day incubation technique, at five other German waterworks,  each of which
uses different treatment rrethods and utilizes different raw water qualities.
They also stated that similar observations have been reported by scientists
in Switzerland and the Netherlands, without giving supporting data or citing
specific references.

     Klotz, Werner & Schweisfurth (1975) also state that there seerred to be
a tendency for decreased microbial activity in the Schierstein plant GAC
adsorbers during winter, as indicated by lower oxygen consumption and C0?
production data (which are not presented).

     Activity of microorganisn populations in the GAC adsorbers—In order to
determine the relative contribution to removal of organic materials frcm
solution by adsorption alone and by adsorption plus microbial activity,
parallel tests were conducted using two pilot plant GAC adsorber test
units.  Each test unit consisted of four glass tubes connected in series
(water passed through the tubes one after the other).  Each glass tube had
an inside diameter of 6 cm.  The GAC bed depth in each glass column was 0.5
m, therefore the total GAC bed depth in each test unit was 2 m (4 x 0.5 m)
and both 4-column pilot plant test units were operated under similar condi-
tions.  One of the test units was kept free of microorganisms by sterile
filtration for nore than two nonths, during which time about 15 cubic meters
of water was passed through this sterile unit before it became bacterially
contaminated.  All other methods of keeping the test units sterile have the
disadvantages of the presence of an additional bactericidal agent and of a
large number of dead organisms.

     In order to increase the amount of biodegradable organic substances,
the inlet water was loaded with about 50 mg/1 of phenol during the second
phase of the test.  This level of phenol is not toxic to the bacteria (Werner,
Klotz & Schweisfurth, 1978).

     Results of these tests are listed in Tables 37 and 38.  When the 3AC
adsorber inlet was unloaded (the sand filtrate), the microorganism contribu-
tion to removal of organics was very small, as far as could be measured by
UV absorption, dissolved organic carbon (DOC) and KMn04 consumption.  On the
other hand, the contribution of microorganisms to the decrease in easily
biodegradable organics concentration was very high.  This is of special
inportance because it rreans that the regrowth of bacteria, which can occur
in distribution systems (feeding upon the easily biodegradable organics),
now can be made to occur in the waterworks itself.
                                      187

-------
TABLE 37.  PERCENTAGE OF BIOLOGICAL REMOVAL OF ORGANIC SUBSTANCES DURING
           PASSAGE THROUGH GRANULAR ACTIVATED CARBON
Raw Water Parameter
DOC
KMn04 consumption
UV Absorption, 240 nm
UV Absorption, 254 nm
BOC?
BODs
BOD20
Sand Filtrate
<1.0%
3.7%
<1.0%
1.2%
69%
46%
17%
Sand Filtrate Using
Phenol -loaded Influent
6,0%
3.1%
2.8%
6.2%
Source: Klotz, Werner & Schweisfurth, 1975
TABLE 38.  MICROBIOLOGICAL SHARE OF OXYGEN CONSUMPTION AND CO? PRODUCTION
             DURING PASSAGE THROUGH GRANULAR ACTIVATED CARBON
Raw Water rarameter
Oxygen cionsunption
Carbon Dioxide Production
Sard Filtrate
58*
61%
Phenol -loaded Sand
Filtrate
68%
64%
Source: Klotz, Werner & Schweisfurth, 1975
     When the GAC adsorber inlet contained 50 mg/1 of phenol, the activity
of the microorganisms was slightly higher in the GAC media.  It must be
remembered, however, that the adsorption capacity of GAC for phenol is very
high.

     Electron-scan microscopic examination of GAC adsorber granules—During
passage of water through SAC adsorbers, the concentration of microorganisms
increases from about 10 colonies/ml to about 100,000/ml.  Since the doubling
period (under optimum conditions in the laboratory) for bacteria is about 20
minutes, this led to the conclusion that the increase in numbers of microor-
ganisms actually measured has to emanate from the bacteria already present
on the activated carbon.  This is because the water remained in contact with
the GAC for a short time only, which was insufficient to allow for the
increases in colony numbers/ml actually measured.
                                                            o
     Tests showed that there were colony numbers of up to 10 /g of wet
material on the activated carbon.  The actual cell number, obtained by
counting cells, could exceed these numbers by almost 100%.

     A variety of electron-scan microscopic tests of activated carbon
samples treated differently showed the sane general distribution of the
microorganisms on 3AC adsorber granules.  As a rule, the microorganisms were
found to be sparsely scattered, and were always in the form of a single
bacterial layer.  The activated carbon surface area available (up to a pore
diameter of 1 micron) was only fractionally utilized (about 1%).  This was
shown to be the case even when the activated carbon had been in contact with
                                      188

-------
a nutrient solution for some  time.  These  findings tend to confim those of
Van der Kooij (1975),  who found  that  only  a  single bacterium was present, on
the average, for each  40 sq microns of  available surface area.

     Adsorption of n-icroorganisms  onto  activated carbon—Tests to determine
the degrees of adsorption of  microorganisms  on  activated carbon were conducted
using starved and washed bacteria  (mixed populations)  in nitrogen-free
environments.  Figure  68 shows an  adsorption isotherm  of bacteria loaded on
activated carbon as a  function of  the adsorptive concentration.  At high
colony numbers (above  lO1^) the  system  tended tcward saturation.  At colony
numbers of 107, 90% of the bacteria were adsorbed.
o

S>
m
o
c

c
o
"o
o
s^
O)
c
•5
(0
o
                         1012-

                         1011H

                         1010H

                         109H

                         108H

                         107H
                             106 107 108 109 10101011 1012

                               adsorptive  concentration
                             (colony numbers/200 ml  buffer).
                         100-
                      t«
                      ^
                       o>
      75-


      50-

      25-


       0
                            106 107 108 109 101°10111012

                               adsorptive  concentration
                            (colony numbers/200 ml buffer).
              Figure 68.  Microbiological loading of GAC--
                         dependence on adsorptive capacity.

                (Klotz,Werner & Schwelsfurth, 1975)
                                      189

-------
     Study of the influence of time on bacterial  adsorption proved  to  be
difficult, since it was impossible to avoid bacterial  increase  followed by
bacterial extinction processes over extended periods  of time (Figure 69).
After an incubation period of 20 to 30 hours,  adsorption and desorption of
bacteria were nearing a steady state equilibrium  condition.   With increasing
ion concentration (phosphate buffer at pH 7.2)  the  bacterial  loading of the
activated carbon Increased as shown 1n Figure  70.   No  dependence of adsorption
on temperature in the range of 5° to 37°C nor  of  pH over the range 5 to 8
was noted.  Dead bacteria were found to be slightly better  adsorbed by
activated carbon than were living bacteria.
               50-,
               40-
            -  30
            a
            c
            •o
            <9
            2.  20-

            u
            4
            O

               10
                                         ° colony  numbers
                                         • cell numbers
           \
10
2*0      30

 time) hrs.
40
                                                        50     <>0
                 Figure  69. Dependence of microorganism loading
                            of  GAC  with time.

                                 (Klotz  et.£j.., 1975)
     In a later report of continuing work at the  Schierstein plant, Klotz,
Werner & Schweisfurth (1976)  pointed out that their  technique for removing
bacteria adhering to activated carbon samples was to mix them with water
using a simple, domestic blendor (Braun  MX 32)  set at exactly the same
rotational speed and for the  saire lengths of time for each sample.  Both low
nutrient P-agar (Wolters & Schwarz,  1956) and high nutrient SPC-agar gave
the optimum numbers of colonies after 7  days of Incubation at 27°C.  Lower
colony numbers were found using meat extract agar incubated at 27° and 37°
as well as using gelatin incubated at 22°C (see German Unit Standards for
Water, Wastewater and Mud Testing, 3rd Edition, 1960).
                                     190

-------
                    25-,
                 o>
                 c
                 TO
                 a
                 o

                 o
                 <
                 o
                         200 ml buffer, pH T.2;
                           KH2PO4/Na2HPO4 , mmoles
              Figure 70. Dependence, of microorganism loading
                         of  GAC  with ion concentration.

                               (Klotz et aj,.,1975)


     Colony numbers were measured  using these  three  nutrient media at the
outlet of the Schierstein plant 3AC adsorbers  for  nearly  one year  (Figure
71).  The colony counts/ml  of the  GAC adsorber influent were about 100/ml,
but the effluents contained 1,000  to 100,000/nl, depending upon the season
of the year (temperature) and upon changes in  raw  water quality.

     When the full-scale pilot plant GAC adsorbers were used for the first
time with new charges of GAC, colony numbers increased both 1n the adsorbers
and in the new 3AC  effluents, even though the  adsorber influent waters
contained, as a rule, 0.5 to 1.0 mg/1  of free  residual chlorine, and sometimes
contained even 4 mg/1 (see  Figure  72).   The maximum  colony numbers (105 to
106/ml of water in  the effluents)  were attained within 10 to 20 days after
placing the GAC in  service.

     Different numbers of colonies were found  at various GAC column depths,
depending upon the  length of time  the particular carbon adsorber had been 1n
service.   This behavior was explained on the basis of free residual chlorine
content.   In the first of four pilot plant test columns connected 1n series,
low colony numbers  and greater than 0.1  mg/1 concentrations of free residual
chlorine  were found.  However a significant decrease in residual chlorine
                                     191

-------
and an increase in bacterial numbers were measured after stage 2, which then
remained constant after stages 3 and 4.   Samples of spent GAC from the plant
adsorbers were found to contain about 108 faacteria/g of wet weight carbon.
V
*•
s



E
^.
in
w
0

1
3
e
            e
            o
            o
            o
                 10SJ
                 104J
                 103J
                 102J
     10'
A
                           V
                                                     SPC-agar
                                      *W»"'X
                                                     P-agar
;

                                 A A
                                    V'i
gelatine
                               100
                              200
                              300  days
                    'jan
                                            nov
             Figure 71.  Fluctuation of colony  numbers in  GAC
                         adsorber effluents at Schierstein plant
                         during 1 year.

                                (Klotz  et  aJL, 1976)


     Increasing the length of the  adsorber columns (by connecting two or
more in series) did not increase the  number of colonies measured in adsorber
effluents.

     Short time reductions in the  number of colonies present in the GAC
adsorber effluents were brought about by backwashing using water from the
adsorber outlets.  The backwash water showed an increase in colony numbers
after being used for backwashing (Figure 73), and these increases could
result only from a lowering of the  counts  in the activated carbon adsorption
media.   This ireans that some of the bacteria were washed out of the activated
carbon columns.  However, the magnitude of the colony count lowering upon
backwashing was not significant, in terms of decreasing the biochemical
degradation efficiency for removing dissolved orgam'cs.  Figure 73 shows
that the numbers of colonies/ml  in  the activated cardon adsorber effluents
at the steady state (before backwashing) was 5 to 8 x 104.  This level  rose
                                     192

-------
above TO5 just before backwashing and dropped to TO3 to" 104 just after
backwashing.  However, the steady state level of colony counts/ml  was  reat-
tained after passage of 1,000 to 1,500 cubic meters (0.25 to 0.4 million
gallons) of water.  This required about 2 days.
                                                outlet
                                    days »-
          Figure  72.  Development of microbiological  activity
                      in full scale adsorber with  new  GAC at
                      Schierstem  plant.

                               (Klotz e_t  a±., 1976)
     In all Investigations conducted by Klotz,  Werner &  Schwelsfurth  (1976)
none of the effluents from any of the biologically operating  activated
carbon adsorbers showed the presence of any fecal  indicators.   Filamentous
fungi and yeasts occurred randomly,  but rarely  in  all  Investigations.  Yeast
numbers were lower than those of filamentous fungi.   No  substantial differen-
ces were found in comparing waters from the Wiesbaden  (Schlerstein) plant
with those of the Frankfurt (Niederrad) Waterworks.   Both of  these plants
are located near each other 1n the southern part of the  Federal Republic of
Germany.  Schierstein draws water directly from the River Rhine and Frankfurt
draws water from the River Main,  It is important  to recognize, however,
that the quality of the River Rhine  is much better than  that  of the River
Main (Werner, 1979).
                                      193

-------
                                                   step 2
                                                         step
4)
J3

3

>

o
S
             10.
backwashing

I  f\   outlet
                  ,•••••«•
                       T
                               8
                     9
10
11
                water throughput, in 10  cu m
             Figure 73. Changes In colony  numbers at each  GAC
                        step and outlet in Schlersteln  plant.

                               (Klotz ej.al.,1976)
     In all test results reported using  the  Schierstein plant GAC adsorbers,
Klotz, Werner & Schweisfurth (1976)  found correlations between microbio-
logical and chemical/physical  data only  with respect to chlorine content and
temperature.

     In pilot plant tests conducted  at the Schlersteln plant, Klotz, Werner
& Schweisfurth (1976)  employed four  identical glass tubes (4 cm diameter,
1.2 m in length) and connected in series, then filled with 80 cm of the same
granular activated carbon used to charge fresh GAC into the full-scale
Schierstein plant.  This amount of GAC corresponded to a wet weight of about
1 kg.  The adsorbers then were filled with chlorinated sand filtrate water
from the main plant.

     Flow velocities of water  through the (fresh) GAC adsorbers (4, 8 and 20
m/hr) were found to influence  the development of microbiological activities.
The higher the velocity the  more slowly  the  maximum values of colony counts/nl
                                     194  _

-------
were reached in the filtrates (Figure 74).   The  maxima were highest for the
slowest flow rate (4 m/hr)  and they were attained  faster at the slow flow
rate.  On the other hand, at the highest flow  rate (20 m/hr), higher colony
counts/g of GAC eventually resulted (Figure  75).   Therefore, more bacteria
were formed per unit time at such flow rates,  which resulted in higher
amounts of microbiological  activity.
       (0
       £


       E
       ^
       ia
       k
       V

       1
       3
       C
        C
        o
       "o
        o
        Figure  74.  Development of colony numbers in effluents
                    of new  GAC pilot  plant adsorbers as a
                    function of flow rate  at  Schierstein plant.

                         (Klotz  et  al.,1976)
     On the other hand, at slow flow rates,  media conditions were created
which had a favorable effect on the reproduction of filamentous fungi.
Yeasts were found infrequently.  Bacfcwasning resulted  in lowering the
colony numbers in the adsorber effluents  for a  short period of time, and
this effect was most pronounced for the adsorber group having the slowest
throughput flow rate.

     Similar behaviors with respect to bacterial growth rates in new batches
of granular activated carbon were observed with LWS, ROW, ROW Supra and
                                     195

-------
F-100 carbons (Figure 76).  LWS Carbon showed about one-half an order of
magnitude lower colony counts/ml in the adsorber outlets  than did the other
three activated carbons tested.
         9
         a
         i
         3
         c

         S=»
         C
         o
         o
         u
                                20     30
                                   days *>
40
50
60
          Figure  75. Development of colony  numbers In  new
                      GAC  charges in Schierstein pilot plant
                      adsorbers  as a function  of  flow rate.

                             (Klotz et. a\., 1976)
     This work of Klotz,  Werner &  Schweisfurth  (1976) confirmed that data
obtained using the pilot  plant test  units could be related directly to data
obtained using the full scale  plant  GAC adsorbers, with respect to colony
numbers and the trends noted  in these  numbers with time.

     Werner, Klotz & Schweisfurth  (1979) presented additional data to
substantiate their earlier conclusions regarding the extent of microbio-
logical activity in the Schierstein  (Wiesbaden) water treatment works
granular activated carbon adsorbers.   Raw water before chlorination generally
contained 2 x 105 colony  counts/ml and these decreased to about 1,000/ml
before entering the activated  carbon adsorbers.  With high level prechlorina-
                                     196

-------
Figure 76. Development of  colony numbers in effluents
           of pilot plant adsorbers  with  different
           types of GAC at  Schiersteln  plant.

                   (Klotz et ai,, 1976)
                          197

-------
tion, almost all the bacteria originally present were kil'ed  and the GAC
adsorber inlet waters contained very few colony counts.

     In the activated carbon adsorbers the number of bacteria again increased
to values around 7 x loVml.  The activated carbon media were shown to
contain about 1,000 times more bacteria per unit volume  than  did the carbon
adsorber effluents.

     The nunber of living cells was determined by enzymatic methods, rather
than by culture methods, and the total number of cells (living plus dead
bacteria) was determined microscopically by counting them after neirbrane
filtration.  Colony count determinations measured up to  2Q% of all  living
bacteria and up to 5% of the total (living + dead) cells.

     Through a population comparison made by numerical taxonomy methods and
comparison of the morphological and biochemical qualities of  the bacterial
strains identified as being present in the activated carbcn -edia (using a
computer program), it was found that:

1)   The ability of the bacteria to adapt in the activated carbon adsorbers
     was less marked than their ability to adapt in raw  water.

2)   There was a higher percentage of pseudomonas bacteria present In the
     carbon adsorbers than in the raw water,

3)   Bacterial populations can be differentiated clearly by the use of
     substrate, and especially by their reactions to toxic substances.

     Based on these findings, Werner, Klotz & Schweisfurth (1979) concluded
that a special microorganism population forms in the activated carbon
adsorbers  (at least those in use at the Schlerstain water treatment plant).

     Identification of microorganism populations—To date, 26 species of 11
genera of microorganisms have been isolated from the Schierstein plant GAC
adsorber effluents by the University of Saarland scientists (Werner, Klotz  &
Schweisfurth, 1979), and these are listed in Table 38A.   Most of the microorga-
nisms present belong to the genera Pseudomonas.  The families Bacillus and
Azomonas also are represented to a considerable extent.   These bacteria
found to date are non-pathogenic in nature, and are generally found in
water.

     Table 39 lists the filamentous fungi and yeasts found in the effluents
from the biologically-loaded granular activated carbon adsorbers of the
Schierstein water treatment plant.  These filamentous fungi and yeasts
occurred rarely and irregularly in the effluents, hence  their role in water
treatment  is considered by Werner, Klotz & Schweisfurth  (1979) to be of
secondary  importance.  The fungi identified to date all  are apathogenic.

     Performance of the bacteria—Activated carbon adsorbs dissolved organic
substances, which can act as substrate for bacteria, but also adsorbs
bacteria.  Figure 77 shows the Freundlich adsorption isotherms for the
adsorption of bacteria onto activated carbon.

                                     198

-------
    10 —
  o
  <
  a
  w
  E
  3
  C

  >>
  C
  O
  o
  u

  a>
  o
 5—
   100

    80 1

c   60 H
•o
(8
O
40

20 -
                              t'o
              I
              5
                          I
                         10
                 absorptive  concentration


            Hog colony number/200 mil
            Q=32 xc°'77   c^— a Q=10KC
                                         °'85
 Figure 77. Bacterial adsorption on GAC.


                 (Werner  et al.,1978)
               __  _ 199

-------
TABLE  33A.  BACTERIAL SPECIES  FOUND  IN EFFLUENTS OF GRANULAR ACTIVATED
  	CARBON ADSORBERS AT THE SCHIERSTEIN  (FRG) PLANT
      Pseudomonas alcall genes            thromobacterium vlolaceum
      Pseudomonas cepacla                Neisseria slcca
      Pseudomonas facills                Acinetobacter calcoaceticum
      Pseudomonas fluorescens            Mlcrococcus luteus
      Pseudomonas lemoignei              Staphylococcus saprophyticus
      Pseudomonas mendocina              Bacillus cereus
      Pseudomonas ruhlandii              Bacillus circulans
      Pseudomonas stutzeri               Bacillus licheniformls
      Pseudomonas spec.                  Bacillus megaterium
      Gluconobacter oxidans              Bacillus pumulis
      Azomonas agilis                    Bacillus thuringensis
      Azomonas insignis                  Corynebacterium spec.
      Azomonas macrocytogenes            Micromonospora spec.

          Source:  Werner, Klotz & Schweisfurth, 1978
TABLE 39.  TYPES OF FUNGI AND YEASTS FOUND IN EFFLUENTS OF GRANULAR
  	ACTIVATED CARBON ADSORBERS AT THE SCHIERSTEIN (FRG) PLANT
FTTarrentous fungi        PhiaTopnora hoffmannii
                         Phialophora mutabilis
                         Taphrina spec.

Yeasts                   Rhodctorula minuta var. texensis
                         Cryptococcus uniguttulatus
                         Candida guillermondii var. guillermondii
                         Hansenula anorcala var. anomala

     Source:  Werner, Klotz & Schweisfurth, 1978
     Because of the great difference in size between bacteria and organic
molecules, the two are separated from each other after adsorption because of
the porous structure of the activated carbon.  The smaller, organic molecules
are adsorbed mostly in the micropores of the activated carbon, which represent
about 992 of the total available surface area.  On the other hand, the much
larger bacteria cannot be adsorbed in the micropores, but only on the surface
and in the macropores, which make up about "56 of the total activated carbon
surface area available for adsorption.   This has a negative secondary effect
on the metabolism of the bacteria.

     In the absence of activated carbon, both bacteria and substrate are
uniformly distributed in the aqueous medium.

     As confirmation of these statements, Figure 78 shows plots of the
metabolic activity of bacteria (measured by oxygen consumption) versus time.
The initial colony counts/ml of bacteria with or without activated carbcn
being present were 1.2 x I08/ml and the water solution also contained 0.1


                                    _ 200

-------
   c
   o
   E
   3
   U)
   C
   O
   u

   c
   V
   O)
   >
   X
   o
J
                                                   •= 1

1
2
bacteria
1.2 x 108/ml

substrate
phenol
0.1 g/l

	 r •"... 	
GAC
0.5g
1 1. 25 mm


               10   20  30   40  50  60   70   80

                        times hrs.


Figure 78.  Effect of GAC on metabolic activity of bacteria.

               (Werner et al., 1978)
                             201

-------
 g/1  (TOO mg/1) of phenol  to act as a substrate for the bacteria.   The upper
 curve (data obtained in the absence of activated carbon)  shows  that the
 phenol  substrate was utilized rapidly by the bacteria  (most  within the first
 24 hours).   On the other  hand,  when 0.5 g/1  (500 mg/1)  of granular activated
 carbon  (1  to 1.25 mm particle size) was added to a second solution containing
 the  sane number of colony counts/ml and the  same amount of phenol,  less than
 half of the substrate was utilized after 80  hours  of contact than  was  utilized
 within  24 hours  when activated  carbon  was absent.

      Activated carbon also provides a  positive effect on  the metabolism of
 bacteria, however.   It  enriches the concentration  of organic substances in
 the  adsorber media  and  increases their  residence  times  in  the adsorber.  In
 addition, there  is  provided a "buffering  action" of the system for organic
 substances  which  are toxic  to the  bacteria present.  These effects are shown
 by the  data  plotted  in  Figure 79,  in which 8  experiments were conducted with
 solutions initially  containing 3.5  x 108  colony counts/ml.  To four sets of
 two  solutions  were added  2.5, 1.0,  0.3 and 0.1 g/1 quantities of phenol.   In
 lower concentrations, phenol can serve as substrate (food) *or the bacteria,
 but  in  the higher concentrations, phenol  is toxic to the same bacteria.

     To one of each  of  the four sets of solutions containing 3.5 x 108
 bacterial colony counts/ml and added phenol  now was added 0.5 g/1  of granular
 activated carbon  (1  to  1.25 mm particle size) and the rates of oxygen consump-
 tion were determined over a period of 200 hours.

     At the highest  phenol concentration  (2.5 g/1 -- 2,500 mg/1) ) the
 solution without activated carbon showed  no metabolic activity (zero oxygen
 consumption over 160 hours) (curve 1), proving that at this concentration
phenol is toxic to the bacteria present.  However, when 0.5 g/1  of granular
activated carbon was added to the duplicate solution containing  2.5 g/1 of
phenol,  this sample  showed the highest rate of oxygen consumption  of all
samples  tested (curve 2).   Therefore, it can be concluded that the activated
carbon adsorbed the  toxic quantity of phenol, rendering the aqueous solution
harmless to the bacteria.   This allowed the bacteria to remain viable. The
adsorbed phenol then was slowly released into solution, in non-toxic concen-
trations (probably by desorption mechanisms), where it either passed through
the activated carbon adsorber to the effluent and/or was captured  and metabo-
lized by the bacteria.  Once adsorbed on the activated carbon,  high initial
concentrations of toxic materials,  such as phenol, can become a  slow releasing
source of dissolved organic carbon  substrate for the bacteria.

     Similarly, the solution containing 1 g/1 of phenol  also  was toxic to
the bacteria (no oxygen uptake noted after 160 hours)  and the duplicate
sample treated with 3.5 mg/1 of activated carbon showed the second highest
oxygen consumption rate over 160 hours.

     Data obtained at the lower phenol concentrations,  however,  seemed to
conflict with  the hypothesis developed 'or the higher phenol  concentrations.
The sample containing 0.3 g/1 of phenol and.no activated carbon  consumed
 oxygen at more than  twice the rate  than did the duplicate sample containing
 activated carbon.  This apparently reversed behavior can be explained,
however, by considering that the adsorption capacity of the weight of


                                     202

-------
           13-
           10-
      c
      o

     «w
      a
      E
      3
      w
      c
      o
      o
      09
      X
      O

1
2
3
4
5
6
7
8
bacteria
3.5 x 108 ml
ohenol substrate
2Sg/l
_, . ":...: 	 i" •-;-..
- i
1
, >
^ < f
-
-
1






1 fl/l



jw-. , :




0.3 g/l




•• - '
; • ^ ;


01 g/l






- • •. ;
GAG, 0 5 g
1 1.25mm



\

: ^

*
Figure 79.  Effect of GAG on metabolic activity of bacteria.


                  (Werner et^ aj... 1 978)
                            203

-------
activated carbon added was sufficient to adsorb the total  amount of phenol
present.  In this case, even low levels of phenol  were not released into
solution, and the net effect was total removal  of biodegradable substrate
from solution.

     This postulate is supported by the fact that the solution containing
0.3 g/1 of phenol was not toxic to the bacteria; the oxygen consumption
measured for this sample was the third highest of the eight samples tested.

     Further confirmation of this hypothesis can be developed by considering
the data obtained with the samples containing 0.1  g/1 of phenol.  Again, the
sample to which activated carbon was added showed no metabolic activity
(zero oxygen consumption after 160 hours) whereas  the sample without added
activated carbon rapidly attained a low level of oxygen consumption.

     The rapid attainment of a constant level of oxygen consumption by the
samples containing 0.1 and 0.3 g/1 of phenol without added activated carbon
shows that the oxygen uptake rates are independent upon the concentration of
phenol (over this concentration range).  With the sample containing 0.3 g/1
of phenol, the fact that this amount of oxygen consumption was not attained
until after about 40 hours can be explained either by there being an insuffi-
cient concentration of bacteria present initially or that this phenol concen-
tration exerts scrre toxicity to the bacteria, which is overcome later either
by adaptation or by the increase in bacterial populations over that period
of time.

Contribution of Bacteria to Water Treatment—-
     The following -merits were found by Werner, Klotz & Schweisfurth (1979)
to hold true for the treatment of Rhine River water with high level chlorina-
tion prior.to filtration and granular activated carbon adsorption.  At the
time this phase of the research program at the Schierstein plant was conducted,
the efficiency of the activated carbon adsorbers in removing dissolved
organic materials was about 80%.  During this study the bacterial contribu-
tions to the removal of organic iraterials (as measured by consumption of
dissolved oxygen and production of C02 compared with the total dissolved
organic carbon removed) were found to be as follows:

•    5% reduction in levels of dissolved organic carbon,

t    about 702 reduction in levels of easily decomposed organic substances
     (BOD2),

•    about M% reduction in levels of difficult-to-decompose organic substan-
     ces (BOD2Q),

•    about 60% of theoretical oxygen consumption,

•    about 60% of theoretical production of carbon dioxide.

     It was noted by Werner, Klotz and Schweisfurth (1979), however, that as
the adsorption efficiency of the activated carbon for dissolved organic
substances decreased, the bacterial contribution to removal of organics

                                      204

-------
 increased  considerably".   At the  Schierstein  plant,  the already low biodegra-
 dability of the  organic  components originally  present in the raw water was
 further lowered  by  the high level  chlorination step.

      Lowering  the amounts of easily biodegraded organic substances in
 biologically operating granular activated carbon adsorbers has a significant
 effect  on the  regrowth of bacteria in water supply distribution systems.
 Through biologically active activated carbon adsorbers these organic materials
 are at  least partially removed in  the plant, rather than in the distribution
 system.   Furthermore, the bacterial activity present causes a continuing
 partial  regeneration of the  activated carbon, and thus prolongs its operating
 life before  reactivation is  required (Werner,, Klotz & Schweisfurth, 1979).

     During  June, 1978, Prof. Dr.  Schweisfurth and his students, Drs.  Klotz
 and Werner were interviewed  regarding some of the unpublished details  of
 their 6 years of microbiological  studies at the Schierstein plant.   The
 following additional information was supplied by these scientists  in the
 form of unpublished information:

 1)   Normal plate counting techniques measure only 5 to 10% of the  living
     bacteria actually present.  However, enzymatically,  20% of the total
     cell numbers can be  counted.   Tctal colony counts determined on culture
     nedia provide information only regarding those types  of bacteria which
     grow on the media.   Not all  types  of bacteria grow on  specific culture
     media.

 2)   E. coli bacteria present in the influents of biologically active
     7iHer/adsorber r.edia are not found in the effluents  because other
     bacterial  strains dominate in the  media and E_.  coli  cannot grow under
     tneje conditions.   They simply die off.

3)   The Schierstain plant in Wiebaden  has an operating rule which  does not
     allow a GAC column to stand  idle and off-line for more than two days.
     Otherwise  plate counts  increase significantly and the  biology  changes.

4)   The dissolved organic content of the Rhine at Wiesbaden is 3 to 4  mg/1,
     which is comprises  mostly of municipal  sewage.   No industries  discharge
     in the area.  Wiesbaden recently has prolonged the residence time  in
     reservoirs ahead of  the Schierstein plant, to provide  3-day residence
     times for  Rhine River waters,  and  which are biologically active.
     Before the reservoir residence time was increased, the 4 mg/1  DOC  level
     of the plant influent water was lowered to 2.5 at the  inlet to the GAC
     media, and to 1 mg/1 exiting the biologically active  GAC adsorbers.
     Since the  increased  reservoir residence times COC levels in the inlet
     to the GAC adsorbers are even lower.

 5)   Although BAC media may seen to operate somewhat like  trickling filters,
     they are not the same.   Trickling  filters irust have a  skin of.slime
     (formed by multiple layers of bacteria) in order to be effective.   BAL
     media remove about 3 times the amount of dissolved organics from solution
     as do trickling filters.  In  trickling filter operation, dissolved
     SgSlS Jri adsorbed only by_the slime coat-Ing, which is of very  low


                                     205

-------
      surface  area,  and  then  1s  degraded.   In  BAC  operation, organlcs  are
      adsorbed on  the  carbon  surface,  in the macropores and also  1n  the
      micropores,  then are  degraded.

 6)    Only  about 1%  of the  available GAC surface area  (the macropores) is
      occupied by  bacteria, which are all greater  than 1 micron in particle
      size.  Therefore,  bacteria occupy only about 1 sq cm/cu m of GAC.

 7)    No clogging  of the BAC media has been observed at the Schierstein plant
      during the past  6  years of use.  On the other hand, the Rhine River at
      Wiesbaden does not contain nutrients which are optimum for good bacterial
    .growths.  If such  optimal nutrients were present, it is possible that
      clogging could occur.

8)    Waters passed  through sterile, virgin GAC with no bacteria present
      contained measureable quantities of C0£ and consumed measureable
      amounts of dissolved oxygen.  This shows that GAC can participate in
      chemical oxidation reactions, and these may continue to occur even
      after bacterial activity has reached a steady state.  In the experiment
      reported earlier (Klotr, Werner & Schweisfurth, 1975) in which a GAC
      column had been kept sterile for two months,  C02 was produced, after
     which time the column became biologically active and the experiment was
     discontinued,  "he amount of C02 produced represented 1.5% of the total
     generated.

9)    In Warburg apparatus studies (measuring iretabolic bacterial  activity by
     oxygen consumption and C02 production), the size of the GAC  granules
     was important.   The smaller the particle size the faster  was the rate
     of oxygen consumption, until  the carbon granule approaches  1 mm.   At
     this point clogging became prevalent and the  apparatus  then  could not
     be aerated or backwashed.   With large GAC particle sizes  the time for
     transfer of adsorbed organlcs to the  bacteria becomes longer.   The
     mechanism probably involves desorption of the organic material  from the
     micropores,  followed by diffusion to  meet the bacteria.   With  larger
     GAC particles,  the process is believed to become diffusion  controlled;
     with smaller 3AC particles the process is desorption controlled.

10)   In considering  biological  mechanisms  occurring  during GAC operation,  it
     is possible  that some bacterially secreted enzymes  can  leave the ,
     bacterial cells,  then may diffuse into the GAC  micropores and  act on
     the adsorbed organics, desorbing them so that they can  diffuse out of
     the micropores  and into the areas occupied by the bacteria.

11)   In another water works (not Schiersteln)  the  GAC was removing  about 30*
     of the dissolved organics present in  the adsorber influents  after 5
     years of operation.

12)   For sampling of an operating pilot plant GAC  adsorber,  the  containers
     are opened (at the top)  and a 0.5 cm diameter sterile pipe  is  inserted
     into the carbon medium to the measured depth.  The open end  of the pipe
     is closed with the thumb of the person sampling and the filled tube is
     removed and emptied.  The 0.5 cm sampling pipe  allows 4 g samples to be


                                     206

-------
     taken.  The Schierstein plant GAC adsorbers are 4 neters high,  3 meters
     in diameter and contain 2 to 3 meters of 3AC depth.   When the full-
     scale plant GAC was charged, 10 to 20 g samples of used 3AC were taken
     at 4 to 5 column depths.  Samples also were taken from the center out
     to the periphery of each bed.  The samples were found to be homogeneous
     (with respect to the bacterial parameters measured)  at the sane bed
     depth of GAC out to the peripheries,  and also were homogeneous  after
     the first 30 cm of depth.

13)  GAC samples were homogenized in sterilized tap water using a Waring
     Blendor type of nixer.   After establishing the optimum dilution to
     provide the best number of colonies for counting, the Blendor should be
     calibrated.  This is done by plotting time of horrogenization versus
     colony counts.  The peak is taken as  the optimum time of homogenization
     for that Blendor (which is always operated at the same speed).

14)  The amount of dissolved organic carbon removed from solution bacterially
     is not precisely equivalent to the arount of C02 produced.  This is
     because some organic carbon is used by the bacteria  for self-synthesis,
     and does not becone liberated as C02.  However, some bacteria die and
     are attacked by living bacteria.  When this occurs,  some of the carbcn
     contained by the dead bacteria is released as COj.  These two effects
     of carbon consumption and carbon release tend to balance each other,
     and it should be possible to follow the biological degradation  of
     organics reasonably quantitatively by ireasuring the  rate of COg formed
     as well as the rate of consumption of dissolved oxygen.

CONCLUSIONS REGARDING MICROBIOLOGICAL ASPECTS OF BAC SYSTEMS

     From the published works and private  communications  described in this
sub-section and elsewhere in this report,  the following major conclusions
can be drawn concerning the nicrobiology present in operational BAC  systeirs
in drinking water treatment plants:

1)   Bacterial activity develops rapidly in fresh charges of granular
     activated carbon (within 5 to 12 days of initiation or flow) and
     reaches a peak of 104 to 106 colony counts per ml of water or 1C6 to
     108 per g of GAC present in 10 to 20  days.  At biological equilibrium,
     counts usually are higher in the GAC  media than in the effluents.

2)   Standard plate count techniques (incubation at 27°C over 2 days) show
     relatively low colony counts in the filter/adsorber media and their
     effluents.  Incubation over 7 days shows much higher colony counts, and
     is the procedure being used by German microbiologists at the University
     of Saarlands in their studies.

3)   These high bacterial activities occur even in the presence of free
     residual chlorine in the 3AC rredia influents.  There is an indication
     that when chlorinated influents (containing as high as 0.5 mg/1 free
     residual chlorine after breakpoint chlorination) are passed through
     biologically operating activated carbon media, chlorine-resistant
     bacteria can be present in the effluents.

                                      207

-------
 4,    High  bacterial  levels  in  the  activated  carbon effluents can  be destroyed
      using low  dosage  levels of  chlorination (0.2 to 0.3 mg/1  in  Germany;  up
      to  0.5 mg/1  1n  France) or of  chlorine dioxide (0.15 to 0.3 mg/1) provided
      that  extraneous chlorine- or  chlorine dioxide-demanding materials are
      absent.

 5)    At  least 20  species of bacteria, 3 species of filamentous fungi and 4
      species of yeasts have been identified  in effluents from operating GAC
      units  at the Schierstein  plant at Wiesbaden.  During the studies
      reported,  this  plant used aeration, breakpoint chlorination and rapid
      sand filtration as pretreatment before  GAC adsorption.  All mlcrobial
      populations identified to date are non-pathogenic soil and water
      organisms.

6)    E_. coli bacteria do not appear to survive in competition with other
      types of bacteria present in biological  activated carbon media used for
      treating drinking water.   Therefore, E_.  cpli baterla are not normally
      found in BAG media effluents.   However,  this conclusion must be based
      on the assumption that the BAG adsorber is properly sized  (provides
      sufficient empty bed contact time)  and  is operated at a sufficiently
      slow throughput velocity.

7)    No publications are known which deal  with the question of  endotoxins
      present in effluents from operational  BAG systems.

8)    Backwashing of operating BAG units lowers bacterial  counts in the
     activated carbon media (and raises  them in the  backwash water),  but not
     sufficiently to lower the degree of water treatment being  sought (organic
     compound removal or nitrification).   Counts return to noriral  within 2
     days after backwashing.

9)   GAC units containing microbiological  growths taken off-line are  not
     allowed to stand idle for more than two  days,  otherwise the microbiology
     changes and colony counts  increase  significantly.

10)  There are no known incidents of biological fouling of operating  BAG
     systems in Europe.  If anything,  the use of ozone  as  the preoxidant
     lengthens the tiire between backwashing of both  the sand filters  and GAC
     adsorbers (at the  Dohne plant, Mill helm,  FRG).

11)  After 5 years of uses  GAC  adsorbers at another  German water works (not
     Schierstein)  were  removing 30% of the dissolved  organic carbon present
     1n the adsorber influents.

12)  On the other hand, results obtained by Van der  Kooij  (1979)   showed
     that colony counts in the filtrates from sand,  granular non-activated
     carbon and granular activated  carbon filters were  about the sane.   This
      indicates that the high colony counts  usually observed in  GAC filter/ad-
      sorbers are Independent of the amount of organic materials adsorbed by
     the GAC.
                                     208

-------
 13)   The maximum removal  of  organic materials present  in BAG adsorber
      influents  (50%  at  biological equilibrium) was found during pilot plant
      studies  at the  Auf-dem-Werder plant at Bremen, FRG, but without the use
      of  preozonation.   In this case, a special 6AC was employed which had a
      high percentage of macropores (to house a higher  level of biological
      activity).   In  all other water treatment plants known to be utilizing
      BAG processes,  the degree of removal of dissolved organic materials at
      the biological  steady state 1s 25% to 35%, even with preozonation.  It
      is  worthy  of noting that 30% removal of dissolved organics is found at
      Wiesbaden  (the  upper Rhine which is not so heavily polluted) using
      aeration and prechlorination without river sand bank filtration.  On
      the other  hand, the same 30% removal of organics is obtained in the
      lower Rhine area at DUsseldorf,  where river sand bank filtration is
      followed by ozonation.   BAG units can be allowed to remain on line at
     water works in  southern Germany for longer times (up to several years)
      before reactivation 1s  required,,  because of the absence of significant
      levels of  halogenated organics 1n the Rhine at this point.   On the
      other hand, in  the Dtlsseldorf area,  reactivation tines of BAG rredia are
     controlled by the presence of halogenated organics progressing through
     the GAG.

14)  At the Schierstein plant in Wiesbaden,  bacterial  breakthroughs were
     noted after 3 years of  use of GAG columns for taste and odor control.
     This suggests that water utilities Installing 3AC  systems  should
     consider a bacterial  monitoring  test to use as a  regeneration control
     parameter, in addition  to whatever organic  parameters  are  also adopted.

15)  In operational  BAG systeirs,  bacteria occupy only  the  outer  surface  and
     nacropores of the activated  carbon granules.   These account  for 1  to 2%
     of the total GAG surface area  in  most activated  carbons  currently
     available.   On  the average,  a  single bacterium occupies  40  square
     microns per cubic meter  of GAG surface  area in operating BAG  systems.
     This amount of  bacteria  present does not  appear  to Interfere  with
     dissolved organic adsorption processes  1n operational  BAG  systems.

16)  High levels of  phenol in the SAG  adsorber influent(which are  toxic  to
     the bacteria present  in  the  absence  of  GAG)  are  lowered  by GAG adsorption
     to levels at which phenol  can  be  utilized as  bacterial  substrate.
     Thus, BAG can be considered  as a  "sink" for adsorbable  toxic  (to bac-
     teria)  organic  materials,,  which then may  be biodegraded  in  the biologi-
     cally active adsorber.

17)  Cne postulated  mechanism of  BAG operation Involves rapid adsorption of
     dissolved organics by the GAG  mlcropores, followed by  slow  desorption
     and diffusion of the  organics  to  areas  where bacteria  are  present.   The
     snail er the GAG partUle size  the greater 1s the rate  of biochemical
     degradation of  organics, possibly because the diffusion  path  is shortened


                                                                        f1-
     problems of clogging and reducing attainable flow rates.


                                     209

-------
18)  Bacterial activities are higher in BAG media during summer than during
     winter, probably because of higher water temperatures.

19)  Samples of GAC known to be bacterlally sterile cause some oxygen
     consumption and CC>2 production 1n waters being passed through  such
     media.  At biological  steady state conditions, however,  this amount of
     C02 produced represents only about 1.5% of the total  produced.
                                     210

-------
                                  SECTION  11

                        ESTIMATED  COSTS  FOR  BAC SYSTEMS


      It must  be  kept  in kept  in mind  that the capital and operating  costs
 for biologically enhanced granular activated carbon systems are determined
 by  a number of factors:

           the type and  amount of  oxidative  pretreatment required,
           the organic composition of  the  raw water (conponents which can and
           cannot  be converted into biodegradable materials),
           the necessity to add oxygen,
           the necessity to maintain pH between 6 and 8 (to maintain optirral
           bacterial activity),
           the empty bed contact time of the GAC colurns or beds,
           the necessity to pretreat the wastewaters to remove components
          which may be toxic to the bacteria.

      In addition, the frequency of regeneration of the GAC will have a major
impact on  the GAC operating costs.  In fact, one of the primary advantages
of operating GAC media in biologically enhanced modes is the saving in
reactivation costs (including attrition losses which occur during nrovement
and  handling of SAC during reactivation) which can be obtained in certain
cases by installing a preoxygenstion or preoxidation step.   If the saving in
GAC  reactivation costs more than offset the  installation of pretreatment to
promote biological activity, then BAC will be cost-effective for the specific
wastewater under consideration.

      In this section, estimated costs for the nrajor components of BAC
systems wi',1 be discussed.  These will include preoxidation with ozone (the
highest cost preoxidation system), GAC adsorption contactors,  Initial loading
of GAC to the contactors and GAC thermal regeneration equipment.

COSTS FOR OZONATION SYSTEMS

      Oxidation with ozone will be the highest cost oxidation step, compared
with  using oxidants such as chlorine, chlorine dioxide, potassium perman-
ganate, hydrogen peroxide, etc., adding pure oxygen or simply the addition
of air.  However, the most widely known biologically enhanced  GAC systems to
date  utilize ozonatlon as the preoxidation step.   Therefore, for purposes of
this  discussion, the costs for ozonation will be summarized.  Those BAC
systems which can utilize lower cost oxidants effectively, or simply oxygen
or air, will be able to lower this portion of their total  costs
proportionately.


                                     211

-------
 Capital  Costs  For Ozonation Systems

     Major  factors  affecting the capital costs of ozonatlon systems Include:

          t   Capacity of ozone generation system
          •   Number of stages of ozone use
          •   Type of ozone contactor off-gas destruction
          «   Type of ozone control system
          •   Space requirements
          •   Standby generation capacity required
          o   Use  of air versus oxygen for ozone generation

     Capital costs  for ozone generation systems range from $850 to $1,150/16
 of ozone generated  per day when up to 100 Ibs/day are required.  This range
 drops to $300  to $500/lb of ozone per day when 1,000 Ibs/day of ozone are
 required (Table 40; G.W. Miller, 1979).  To these equipment costs must be
 added the costs for contacting, building space, control  systems, contactor
 off-gas destruction, etc.  Capital  costs for fully installed ozonation
 systems are given in Table 41  (Gumerman, Gulp & Hansen,  1978).   Costs for
 construction of diffuser contactors (concrete columns,  18 feet  deep with a
 length/width ratio of approximately 2/1) are listed in  Table 42 for various
 sizes of chambers (Gumerman,  Gulp & Hansen,  1978).

     TABLE 40.  COST RANGE OF OZONATION SYSTEMS (FROM AIR)

Ozone Generation Capacity,
Ibs/day of ozone
100
200
300
400
500
1,000
Approx. Capital Cost,
$/lb production cap. /day
850 -
600 -
500 -
450 -
400 -
300 -
1,150
800
700
650
600
500
Source: G.W. Miller (1979)

     Clark and Stevie (1978) assembled available capital  cost data for
ozonation systems and converted 1t into capital  costs per mg/L of ozone
dosage per 1,000 gallons of drinking water disinfected.   These cost data  are
presented in Table 43 for plant sizes ranging from 1  mgd  to 150 mgd.   At  the
lower flow rates (10 mgd and below), the capital costs for generation  of
ozone from air are lower than those for generating ozone  from oxygen.
However, above 10 mgd, the capital costs for generating ozone from oxygen
become less than those for air.  For a 10 mgd treatment plant,  the capital
cost to provide an ozone (generated from air) dosage  of 1  mg/L is estimated
by Clark & Stevie (1978) to be on the order of l.U/1,000 gal.

     One year later, Hansen, Gumerman & Gulp (1979) developed more detailed
capital cost estimates for air-fed ozone generators capable of producing  0.5
to 10 Ibs of ozone/day.   These are given in Table 44, and include costs for
                                     212

-------
TABLE 41.  CONSTRUCTION COST:  OZCNE GENERATION SYSTEMS


Mfgd.
Eqpmt.
Concrete
Steel
Labor
Housing
Subtotal
lisc. &
^ontinq'y
Total
Ozone Generation Capacity, Ibs/day
10

$32,250
--
—
4,840
6,000
43,909

6,460
$49,550
100

143,610
--
--
33,690
8,400
185,700

27.86C
213,560
500

511,960
1,540
1,520
114,110
12,700
642, 70C

96,410
739,110
1,000

685,810
1,540
1,520
143,110
23,400
855,380

128,310
983,690
2,000

1,075,540
2,250
2,210
207,500
35,700
1,318,200

197,730
1,515,930
3,500

1,523,240
2,250
2,210
272,300
41,800
1,841,800

276,270
2,118,070
Source: Gumerman, Culp & Hansen, 1978
TABLE 42.  CONSTRUCTION COST:   CONCRETE OZONE CONTACT CHAMBERS
Concrete Chamber Volume
c\m.
gallons
Excavation &
Sitework
Concrete
Steel
Labor
Subtotal
Misc. &
Contingency
Total
460
3,441
$ 470
850
1,470
2,150
4,940
740
& 5,680
4,600
34,413
1,630
4,950
8,400
12,200
27,180
4,080
31,260
23,000
1/2,063
2,570
8,280
13,570
19,510
43,930
6,590
50,520
46,000
344 J 26
5,150
15,450
23,330
36,120
82,050
12,310
94,360
92,OCO
688,252
10,290
29,810
48,550
69,330
157,980
23.700
181,680
basis: 18 ft deep; length/width ratio = 2/1
Source: Gumerman, Culp & Hansen, 1978
                                    213

-------
TABLE 43.  CAPITAL COSTS FOR DRINKING WATER DISINFECTION WITH OZONE
Capital Costs, i/1,000 gal.
(ozone generated from
air or oxygen)
Capital Cost (from air)
Capital Cost (from oxygen)
Desiqn Caoadtv (mad)
~"l
2.90
4.46
5
1.36
1.50
10
1.11
1.08
100
0.76
0.61
150
0.73
0.58
Source: Clark & Stevle, 1978
TABLE 44.  CONSTRUCTION COSTS FOR SMALL  SCALE OZONE  GENERATION  SYSTEMS*
Cost Category
Manufactured Equipment
Labor
Housing
Subtotal
Miscellaneous & Contingency
Ozone Generation Capacity (Ibs/day)
0.5
$ 11,540
1,860
6,000
19,400
2,910
TOTAL 22,310
Ozone Generating Equipment Costs
($/lb/day of 03 generated) 23,080
Construction Costs
($/lb/day of 03 generated) 44,620
5.0
$ 19,880
3,300
6,000
29,180
4,380
33,560
3,976
6,712
10.0
$ 28,530
4,840
6,000
39,370
5,910
45,280
2,853
4,528
* Assumes ozone is generated from air. Costs include ozone generator,
dissolution equipment, electrical equipment, control instrumentation, but
lot contact chamber. Figures are not stated to include costs for air prepa-
ration equipment (which would add 20% to 25% to capital costs of manufactured
equipment), nor for equipment to destroy ozone in contactor off-gases.
Source: Hansen, Gumerman & Gulp, 1979
                                    214

-------
 the ozone generator,  dissolution  equipment and all required electrical
 equipment and  Instrumentation,  labor and housing, but not costs for the
 ozone  contact  chamber itself, since these are a function of water flow to be
 treated  and  contact time.  These  capital costs are not stated to include air
 preparation  equipment or ozone  destruction apparatus for contactor off-gas
 treatment.   If not Included, these costs must be considered to be somewhat
 low.   For example, ozone generator suppliers normally estimate the capital
 costs  for air  preparation equipment to be about 20% to 25% of the total
 capital  cost of the ozone generation system (Larocque, 1977, Private
 Communication).

     Separate  construction costs vere estimated by Hansen, Sumerman & Cjlp
 (1979) for 18  ft high  (water depth 16 ft) fiberglass reinforced plastic
 ozone contact  chambers to be used 1n conjunction with ozone generation
 systems capable of producing 0.5 to 10 Ibs/day.   Costs for the contacting
 equipment itself are  included with those of the ozone generation system.
 Contactor construction costs are listed in Table 45.

     For larger sized ozone generation systems (10 to 3,500 Ibs/day),
 Gumerman, Culp & Hansen (1979)  estimated capital  costs based upon air feed
 to the generator for up to ICO Ibs/day of ozone generated and from oxygen
 above 100 Ibs/day.  They also assumed that all  oxygen used for ozone genera-
 tion would be generated on-site, and that although the ozonation equipment
would be housed, the oxygen generation equipment would be located outside on
a concrete slab.

     These assumptions made by Gumerman et al.  (1979) are not representative
 ot current ozone generation practice,  eitFer for municipal  water or wastewater
 treatment, for the nrost part.  In drinking water treatment plants,  Miller e_t
 a/L (1978) have shown that of the approximately 1,100 plants which  were
 using ozone during 19779 only 2 generate their ozone  from oxygen.   Only one
 of these plants (Duisburg, Federal Republic of Gertrany)  recycles the oxygen-
 rich ozone contactor off-gases  to the ozone generator.   Even the largest and
 newest water treatment plants using ozone generate more  than 5,000  Ibs/day
 from air feed,,

     On the other hand, most of the newer and larger U.S. wastewater treatment
 plants using ozone also employ the oxygen activated sludge process.   This
means that the oxygen source for ozone generation comes  from an on-s1te,
 oxygen generation plant which generates tons/day quantities of oxygen  at the
 lowest cost, mostly for use in plant biological  reactors.  Therefore,  oxygen
 is available at a lower cost than 1f oxygen generating equipment were  to be
 sized simply to produce the volume necessary to feed  the ozone generators
 alone.

     In addition, 1n oxygen activated sludge plants using ozone, the oxygen-
 rich ozone contactor off-gases are not usually recycled  to the ozone genera-
 tor, but rather are passed into the~bTological  reactor.   This use of ozone
 contactor off-gases is referred to as a "once-through (the ozone generator)
 oxygen"  system.  In this manner, the only loss of oxygen in the system Is
 that which is  dissolved in the ozone-treated wastewater.   Bhargava (1979)
 described an ozone generation system which uses partial  recycle of oxygen

                                     215

-------
TABLE 45.  CONSTRUCTION COST FOR OZONE CONTACT CHAMBERS (SMALL SCALE*)
Cost Category
Manufactured Equipment
Concrete
Steel
Labor
SUBTOTAL
Miscellaneous & Contingency
TOTAL
Contactor Volume (gallons)
850
$ 690
20
10
170
890
130
1,020
2,350
$ 1,270
50
20
280
1,620
240
1,860
5,290
$ 3,160
100
40
470
3,770
570
4,340
8,480
$ 6,960
140
50
610
7,760
1,160
8,920
13,540
$ 8,640
190
70
760
9,660
1,450
11.110
* for use with 0.5 to 10 Ibs/day ozone generation capacity. Contact chambers are
18 ft high, fiberglass reinforced plastic.
Source; Hansen, Gumerman & Culp, 1979

-------
 and which  has  been  installed  1n  the 8 mgd  sewage treatment plant at Murphrees-
 boro,  Tennessee.  This oxygen activated sludge plant using ozone disinfection
 was scheduled  to become operational by mid-1980.  More details on the design
 of  oxygen  recycle systems for generation of ozone can be found in a paper by
 Lestochi et al.. (1979).                                                    *

     Therefore, as with the estimates for  small ozone generators, capital
 costs  estimated by Gumeman et al_. (1979)  are considered to be low (in those
 cases  in which large quantitTes of ozone are generated from air -- for most
 drinking water treatment purposes), but are high by the amount of oxygen
 recycle equipment which normally is not employed.

     Capital costs for large  scale ozone generation (10 to 3,500 Ibs/day)
 estimated  by Gumerrran et^al.  (1979) are given in Table 46 and include
 equipment  for gas preparation  (assumed to be generation of oxygen), ozone
 generation (from oxygen at more than ICO Ibs/day),  dissolution, off-gas
 oxygen recycling (but may not  include destruction of ozone in the contactor
 off-gases), electrical and instrumentation costs, all  required safety and
monitoring equipment, labor and housing costs.

     The ozone contact chamber for large scale ozone generation was assumed
 by Gumerman et al.  (1979)  to be a covered,  reinforced  concrete structure, 18
feet deep ancTwrth a length/width ratio of approximately 2:1.   Construction
costs for such contact chambers are given in Table  47  (costs  for the  ozone
dissolution equipment Itself are included with the  ozone generation equipment
 in Table 46).

     It should be recognized that once an ozonation system has been installed
for any single purpose, additional  applications for ozone at  the same plant
will involve only the incremental costs for additional  ozone  generation
capacity plus  e second contacting system.   All  of the  one-tire costs  associa-
 ted with installing the Initial ozone generation system (power, housing,
controls, air treatment,  off-gas destruction,  etc.) already will  be present,
and the incremental  cost per pound of ozone generated  now will be less than
those required for initial  installation.

Operating & Maintenance Costs Of Ozonation Systems

     Most of the operating expense to produce ozone is  the cost of electrical
 power; this has been estimated to amount to about 80%  of the  total  ozone
generation costs.   Much information on operating and maintenance costs steirs
from European drinking water treatment plants,  some of  which  have been
employing ozone since 1906.   Miller et. a]_.  (1978) summarized  the costs for
 ozone treatment of drinking water supplies in Europe and Canada.   With
 ozonation doses ranging from 1.5 to 3.0 mg/L and with  ozone generation
 capacities of 1S000 to 3,000 Ibs/day (from air), ozonation costs ranged from
 1.75 to 4.0 cents/1,000 gallons of water treated in 1977.  These figures
 include the amortized capital costs (usually over 20 years at 8% Interest)
 for air preparation equipment, ozone generation, ozone  contacting,  treatment
 of  off-gases from ozone contacting, instrumentation and automation for the
 ozonation  subsystem, and installation and housing for  the ozonation system,
 operation  and maintenance.

                                     217

-------
ro
oo
               TABLE 46.   CONSTRUCTION COSTS FOR OZONE GENERATION SYSTEMS (LARGE SCALE)
Cost Category
Manufactured Equipment
Concrete
Steel
Labor
Housing
SUBTOTAL
Miscellaneous & Contingency
TOTAL
Manufactured Equipment Cost,
$/lb/day of 03 generating
capacity
Construction Costs in
$/lb/day of 03 generating
capaci ty
Ozone Generation Capacity (Ibs/dav)
10
$ 34,210
0
0
5,090
6,430
45,730
6,860
$ 52,590
3,421
5,259
TOO
$152,350
0
0
35,410
9,000
196,760
29,510
226,270
1,524
2,263
500
$543,130
1,630
1,680
120,850
13,600
680,890
102,130
783,020
1,086
1,566
1,000
$ 727,560
1,630
1,680
150,420
25,060
906,350
135,950
1,042,300
728
1,042
2,000
$1,135,720
2,380
2,440
218,100
38,230
1,396,870
209,530
1,606,400
568
803
3,500
$1,615,980
2,380
2,440
286,200
44,770
1,951,770
292,770
2,244,540
462
641
Source: Gumerman, Culp & Hansen, 1979

-------
ro
10
                TABLE 47.  CONSTRUCTION  COSTS  FOR  OZONE CONTACT CHAMBERS  (LARGE SCALE*)
Contact Chamber Volume
cu ft 460
Cost Category gallons 3,441
Excavation & Sitework
Concrete
Steel
Labor
SUBTOTAL
Miscellaneous & Contingency
TOTAL
$ 490
900
1,620
2,260
5,270
790
6,060
4,600
34,413
$ 1,710
5,250
9,270
12,820
29,050
4,360
33,410
23,000
172,063
$ 2,700
8,780
14,980
20,510
46,970
7,050
54,020
46,000
344,126
$ 5,410
16,380
25,750
37,960
85,500
12,820
98,320
92,000
688,252
$ 10,820
31,600
53,580
72,870
168,870
25,330
194,200
* Concrete chambers, 18 ft deep; length to width ratio = 2/1
Source: Gumerman, Culp & Hansen, 1979

-------
     The rather broad range of European ozonation system costs is affected
mainly by the cost of housing for the ozonation system and variations in the
local costs of electrical energy.  If the ozonation system is housed in a
separate bui'ding, as opposed to being retrofitted into an existing building,
then the costs will be about 25% higher (Miller et al_., 1978).

Small Scale Ozone Generation—
     Hansen e_t a]_. (1979) developed operation and maintenance costs for
generation of ozone at the rate of 0.5 to 10 Ibs/day, from air.'  Electrical
energy is required for building, heating, lighting and ventilating, as well
as for ozone generation.  It is not clear from the report by Hansen e£a1_.
(1979) whether costs for air preparation and for destruction of ozone in the
contactor off-gases are included.  Process energy for ozonation is based on
Ib kWh/lb of ozone generated for the smallest system to 11  kWh/lb for the 10
Ibs/day system.  Maintenance costs were estimated to be U of the capital
equipment costs.  Annual operation and maintenance costs for small scale
ozone generators are given in Table 48.

Large Scale Ozone Seneration--
     Gumerman, Gulp & Hansen (1979) developed operation and maintenance
costs for systems generating 10 to 3,500 Ibs/day of ozone.   Below 100
Ibs/day, ozone is assunred by these authors to be generated from air; oxygen
is the feed gas above 100 Ibs/day and is generated on-s1te in a quantity
necessary to feed the ozone generators.   For air feed, power requirements
were based on 11 kWh/lb of ozone generated,  but 7.5 kWh/lb from oxygen.
Annual operation and maintenance costs are estimated in Table 49.

     LePage (1979) has reported the results  of 8 months of operational
experience at the 18 mgd Monroe, Michigan drinking water treatment plant
which began operating with osonation (for taste and odor control) in February,
1979.  This plant is capable of generating 450 Ibs/day of ozone from air and
is designed to apply 3 mg/L of ozone to the  raw water.  After 8 months  of
operation, an average ozone dosage of 1.65 mg/1 had been employed at an
average ozonation cost of 0.634tf/l,000 gal of water treated.

     Knorr (1979)  reported that the new 10 mgd sewage treatment plant being
constructed at El  Paso, Texas will incorporate an average ozone dosage of 5
mg/1 prior to GAC adsorption for an estimated operational  cost of 4
-------
ro
IN)
           TABLE 48.  OPERATION AND MAINTENANCE SUMMARY FOR SMALL SCALE OZONE GENERATION SYSTEMS*
Ozone Generation
Rate (Ibs/day)
0.5
5.0
10.0
Electrical Energy
Building
6,570
6,570
6,570
Process
2,560
21,900
40,150
kHhr/yr)
Totai
9,130
28,470
46,720
Maintenance
Material
($/yr)
$ 120
200
290
Labor
(hrs/yr)
370
550
550
Total
Cost**
($/yr)
$ 4,090
6,550
7,190
* Ozone 1s generated from air
** Calculated using $0.03/kWh and $10.00/hr for labor
Source: Hansen, Gumerman & Gulp, 1979

-------
           TABLE 49.   OPERATION AND MAINTENANCE SUMMARY FOR LARGE SCALE OZONE GENERATION  SYSTEMS*
Ozone Generation
Rate (Ibs/day)
10
100
500
1,000
2,000
3.500
Electrical Energy (kWh/yr)
Building
5,750
9,850
16,420
30,780
71 ,820
123,120
Process
40,150
401,500
1,368,750
2,737,500
5,475,000
9,581,250
lotal
45,900
411,350
1,385,170
2,768,280
5,546,820
9,704,370
Maintenance
Material
($/yr)
$ 1,430
3,060
10,770
14,270
22,120
31,150
Labor
(Hrs/yr)
550
550
910
1,830
2,190
2,920
Total
Cost**
($/yr)
$ 8,310
20,900
61,430
115,620
210,430
351,480
* Assumes ozone is generated from oxygen, except for first line, which 1s ozone generated
from air.
** Calculated using $0.03/kWh and $10.00/hr labor cost.
Source; Gumerman, Culp & Hanssn, 1979
ISJ
ro
IXJ

-------
     On  the basis of information published to date, the empty bed contact
times which have produced the most effective removals of ammonia and dissolved
organics from aqueous solutions by combinations of adsorption and biological
activity have been close to 10 minutes.  However, longer empty bed contact
times of 20 to 30 minutes have been recommended-in some European drinking
water treatment plants (Sontheimer, 1978b).

     Regeneration of GAC normally is done on-site at large water and waste-
water treatment plants, especially when the daily volumes of GAC to be
regenerated or frequencies of reactivation warrant.  On the other hand, when
the GAC does not have to be regenerated frequently, then it usually is cost-
effective for the plant to send its spent GAC back to the manufacturer for
reactivation.  Many of the larger European drinking water treatment plants
using GAC do not have on-s1te GAC reactivation facilities (Rice et al.,
1979).

     Gulp (1980) points out that "good cost data are available from operating
(GAC) installations at: (1) The South Lake Tahoe Public Utility District,
South Lake Tahoe, California (13 years), (2) the Orange County Water District,
Fountain Valley, California (4 years) and (3) the Upper Occoquan Sewage
Authority, Manassas Park, Virginia (capital cost data only ~ plant in
operation only a few months)."  In addition to these three plants, there are
an additional 17 or so operating municipal advanced waste treatment plants
which use GAC in the USA, and these plants also provide detailed cost informa-
tion.  However, Culp (1980) also cautions that "costs taken from wastewater
cost curves which are plots of flow in mgd versus cost (capital  or operation
and maintenance costs) cannot be applied directly to drinking water treatment.
Allowance must be made Tn~tEe~ capital costs for the different reactivation
capacity needed (for the two types of application), and in the operation and
maintenance costs for the actual amount of carbon to be reactivated or
replaced."  These factors are site- and water- (or wastewater)-specific.

Capital Costs For GAC Systems

     Components of capital costs for GAC systems Include the contactor, site
modifications, piping, pumps and valves, GAC fill, contingencies, fees and
regeneration furnaces (assuming on-site reactivation).  The summations of
these capital costs for eirpty bed contact times of 9 and 18 minutes have
been estimated by J.C. Clark (1979) and are presented in Figure 80 for flows
up to 300 mgd.  Of greatest significance is the fact that the SAC capital
costs decrease sharply for plant flows up to 50 mgd, then decrease much more
slowly up to 300 mgd.

Small Scale, Package GAC Columns—
     Gumerman, Culp & Hansen (1979) developed construction costs for factory-
assembled, package GAC columns, which were sized with 7.5 minutes empty bed
contact time to treat volumes of 2,SCO gal/day to 0.5 mgd.  GAC bed depths
were taken to be 5 feet and the surface hydraulic loading rate was taken to
be about 5 gpm/sq ft.  Conceptual design parameters for these package GAC
units are listed in Table 50.
                                     223

-------
ISJ
ro
           TABLE 50.  CONCEPTUAL DESIGN PARAMETFRS FOR PACKAGE GAC COLUMNS  (7.5  m1n  EBCT)*
Plant Flow
gpm gpd
1.7 2,500
17 25,000
70 100,000
175 250,000
350 500,000
Flow Rate
(gpm/sq ft)
5.1
5.4
5.6
5.1
5.5
GAC Coluirns
"TJo.
1
1
1
1
1
Bed Area
(sg ft)
0.34
3.14
12.6
34
64
Diameter
(ft)
0.67
2
4
6.5
9
Housing
Area (sq ft)
60
150
300
375
450
* GAC columns are 5 ft deep; surface loading = 5 gpm/sq ft
Source; Gumerman, Gulp & Hansen, 1979

-------
             a
             a>
            O
            O
            O
                 4001
                 300
200
                 100-
                           100    200    300    400    500

                          av.  daily flow In max. month (mgdl
              Figure 80.  Capital costs for 9- and  18-min.  GAG
                         'empty bed contact  times.

                            (J.C.Clark, 1979)

     Costs for these package GAC columns are  based  on  the use of cylindri-
cal, pressurized, downflow steel contactors designed to operate manually at
50 psi.  Housing costs are included.   Not  included  are supply piping to the
GAC column and handling or conveyance  systems for spent or  regenerated GAC.
Construction costs for this size GAC package  unit (to  handle plant flows of
2,500 gal/day to 0.5 mgd) are  given in Table  51.

Large Scale GAC Contactors--
     For plants treating  larger flows  (1 to 200 mgd),  Gumerman, Culp &
Hansen (1979) have developed cost estimates for gravity flow GAC contactors
(concrete and steel  construction)  and  for  pressure  GAC contactors.

     Concrete construction—Gravity flow GAC  contactors are assumed to be
essentially Identical  to  gravity flow  filtration structures.  Construction
costs were developed for  GAC bed depths of 5  ft and 8.3 ft, which provide
empty bed contact times of 7.5 and 12.5 minutes, respectively, at a water
application rate of 5 gal/m1n/sq ft (Tables 52 and  53, respectively).  Costs
of these facilities Include the contactor  structures cylinder-operated
butterfly valves, liquid  and carbon handling  piping with headers in a pipe
gallery, flow measurement and  other instrumentation, master operations panel
and a housing building.  Not included  are  costs for backwashing pumping, the
initial 3AC charge,  spent or regenerated GAC  handling  outside of the contactor
pipe gallery and SAC regeneration and  preparation facilities.  It was assumed
that all GAC in a single  contactor would be removed and replaced with regene-
rated GAC in a single operation.  In turn, this requires that regeneration
facilities be designed to store both spent and regenerated GAC in quantities
equal to the amount in one contactor.
                                     225

-------
ro
CM
               TABLE 51.  CONSTRUCTION COSTS FOR PACKAGE 6AC COLUMNS (7.5 m1n EBCT)
Cost Category
Excavation &
Site Work
Manufactured
Equipment
Concrete
Labor
Pumps, Valves &
Piping
Electrical &
Instrumentation
Housing
SUBTOTAL
Miscellaneous &
Contlnqencv
TOTAL COST
Source: Han

Plant Flow
1.7 gpm
2,500 qpd
$ 50
740
100
1,100
•*
500
600
5,100
8,190
1,230
9,420
17 gpm
25,000 qpd
$ 50
2,900
250
3,900
1,200
600
6,910
15,810
2,370
18,180
70 gpm
100tOOO qpd
$ 50
7,070
480
6,240
4,300
850
9,180
28,170
4,230
32,400
175 gpm
250^000 qpd
$ 80
14,600
580
9,500
6,400
1,100
10,300
42,560
6,380
48,940
350 gpm
500,000 qpd
$ 80
27,100
700
13,000
8,800
1,300
11,400
62,380
9.360
71,740
sen, Gumerman & Culp, 1979

-------
ro
ro
                                    TABLE 52.  CONSTRUCTION COSTS FOR CONCRETE GRAVITY CARBON CONTACTORS*
Total Contactor Volume (cu ft) and Area (sq ft)
ICU
Cost Category (so
Excavation &
SUework
Manufactured
Equipment
Concrete
Steel
Labor
Pipe & Valves
Electrical &
Instrumentation
Housing
SUBTOTAL
M1sc. & Contingency
TOTAL
Volume/Single Contactor
Cost/Single Contactor
ft) 700
ft) 140
$ 1,660
29.000
12.330
10.630
37.330
33.570
14,730
17,400
156.650
23,500
180.150
3bO ft3
(90.070
3,500
700
$ 3,050
62.660
24,880
18.360
81,410
108.700
42.250
40.480
381,790
57,270
439,060
875 ft3
$109.770
7.000
1.400
$ 4,660
86,130
38,330
27,710
138,800
206,130
42,250
70,590
614,600
92,190
706,790
1.750 ft3
$176.700
35,000
7.000
$ 13,670
335,690
87,850
67,650
327.870
597,380
109,050
291,940
1,831,100
274,670
2,105,770
3,500 ft3
$210.580
/O.OOO
14,000
$ 21,600
582,300
142,410
113,300
468,260
863.970
185.720
514.330
2,891 ,890
433,780
3,325,670
5,000 ft-
$237,550
140,000
28,000
t 36,630
1,080.360
253,520
193.160
920.890
1.463,150
291 .840
968,520
5.208,070
781,210
5,989,280
6,360 ft3
$290.460
* 7.5 mln EBCT; 5 ft GAC bed depth
Source: Gumerman,

Culp & Hansen. 1979

-------
                                 TABLE  53.  CONSTRUCTION COSTS FOR CONCRETE GRAVITY GAC CONTACTORS*
ro
ro
oo
Total Contactq
(cu ft) 1,160
Cost Category (sq ft) 140
Excavation &
Site Work
Manufactured
Equipment
Concrete
Steel
Labor
Pipe & Valves
Electrical &
Instrumentation
Housing
SUBTOTAL
Misc. S Contingency
TOTAL
Volune/Slngle Contactor
Cost/Single Contactor
$ 2,220
29,000
15,010
12,940
45, 4M)
33,570
14.730
17,400
170,320
25,550
195,870
580 ft3
$ 97,940
5,810
700
$ 4.080
62,660
30,300
20,690
99,170
108,700
42.250
40f480
408,330
61,250
469.580
1.450 ft-
$117,400
r Voluire
1 1,620
1.400
$ 6,210
86,130
51.180
38.800
168,990
206,130
42.250
70,b90
670,280
100,540
770,820
2,905 ft3
$192.710
cu ft) and Area (sq ft'
58, 160
7.000
$ 18,240
335.690
111.090
82,360
399.150
597,380
109.050
291,940
1,944,900
291,740
2,236,640
5,810 ft3
$223,660
116,200
14,000
$ 28.800
582.300
1«0,500
137,940
570,050
863,970
185,720
&14.330
3,063,610
459,540
3,523,150
8,300 ft3
$251 ,650

232.400
28,000
$ 48.770
1,080.630
308,640
235.150
1,159.990
1.463.150
291 .840
969,520
5.556,690
833,500
6,390,190
10,560 ft3
$290.460
* 12.5 m1n EBCT, 8.3 ft GAC bed depth
Source. Gumerman, Culp 4 Hansen, 1979

-------
 _    Steel construction-These types of contactors are assumed by Gumerman,
 Gulp & Hansen (1979) to be utilized when more than 30,000 cu  ft of 3AC
 contact volume 1s required.  Costs were developed  for such contactors
 (-Iu   ^c?ec! as °PP°sed to factory-built)  of 20  and 30  ft diameters,  each
 with a GAC bed depth of 20 ft and  an overall  vessel height of  35  ft.  System
 hydraulics were sized for an application rate of 5 gal/m1n/sq  ft,  which
 provides  a 30 minute EBCT.   Other  conceptual  design parameters  are  listed in
 Table 54.

      Steel  vessels are  constructed  of factory- formed  steel  plates,  erected
 at  the job-site.  Units  are  provided with a nozzle-style  underdrain; GAC 1s
 removed as  required  for  regeneration through  multiple carbon drawoff pipes
 in  the underdrain support  plates.   Regenerated GAC is returned  through a
 piping system to the  top of  each contactor.   Costs presented are for a
 complete GAC  contacting  facility, Including vessels, face and interconnecting
 piping, access walkways, cylinder-operated butterfly valves on all hydraulic
 piping  and manually operated ball-  or knife-type valves on the carbon handling
 system, flow  control  and other Instrumentation, master operations control
 panel  and a building  to house the contactors.

      Not included are costs  for GAC supply punplng, surface wash and backwash
 pumping, the  initial GAC charge, spent or regenerated GAC handling 'acilities
 (exclusive of the piping within the contactor building) or GAC regeneration
 or  preparation facilities.  Estimated construction  costs for gravity f'ow
 steel GAC contactors are presented in Tables 55 and £6 for 20  and 30 ft
 diameter units, respectively, both providing 30 minute EBCTs.

     Pressure GAC Contactors--Gumerman,  Culp & Hansen (1979) developed
 construction costs for pressure GAC contactors constructed of  shop-fabri-
 cated steel tankage.  Bed depths of 5, 10 and 20 ft were estimated, providing
enrnty ted contact times o* 7.58 15 and 30 minutes0  respectively, at a hydrau-
 lic loading rate of 5 gal/min/sq ft.  Conceptual design parameters are
 listed in Table 57.  The practical  upper limit plant size  for  this type  of
GAC contactor system is 20 to 25 mgd.

     Costs are based upon downflow operation at a design working pressure of
50 psi using cylindrical ASME code pressure vessels,  which are  either 10 or
12 ft in diameter by 14, 23 or 33 ft in  height, furnished  with  a nozzle-
 style underdrain and designed for rapid  removal of  spent GAC and recharge of
virgin or reactivated carbon.

     Estimated construction costs are presented 1n  Tables  58,  59 and 60  for
 7.5, 15 and 30 minute EBCTs, respectively, and are  for complete GAC contacting
facilities.  These Include vessels, cylinder-operated  butterfly valves,
                                                                      bui -
          .
liquid and GAC handling face piping with headers  in  the  ^c  ""J^

                                                              ?
the GAC columns are totally enclosed
                                     229

-------
         TABLE b4.  CONCEPTUAL DESIGN PARAMETERS FOR STEEL GRAVITY FLOW GAC CONTACTORS; 20 FT GAC DEPTH*
Plant
Flow
(mgd)
10
50
100
200
Total Con
Bed Area
20 ft diam
1,570
7,850
15,700
31 ,400
tactor
(so ft)
30 ft diam
—
7,065
14,130
28,260
No. of Contactors
20 ft diam
5
25
50
100
30 ft dlair
--
10
20
40
Fotal GAC Volume, ft3
20 ft diam
31,400
157,000
314,000
628,000
30 ft diam
—
141,300
282,600
565,200
Plant Area
Requirements (sq ft)
26 ft cliam
6,500
33,000
66,000
126,000
30 ft diam
—
26,000
50,000
95,000
* 30 minute EBCT
Source: Gumerman, Culp & Hansen, 1979
CO
o

-------
         TABLE 55.  CONSTRUCTION  COSTS FOR STEEL GRAVITY FLOW GAC CONTACTORS*
Cost Category
Excavation & Site Work
Manufactured Equipment
Concrete
Steel
Labor
Pipe & Valves
Electrical &' Instrumentation
Housing
SUBTOTAL
Miscellaneous & Contingency
TOTAL
Volume/Single Contactor (ft1)
COST/SINGLE CONTACTOR
Total Contactor Volume (cu ft)
"31 ,4ti<5
$ 2,050
340,970
7,650
3,810
66,220
140,730
50,460
169,000
$ 780,890
117,130
$ 898,020
6,280
$ 179,600
157,000
$ 6,560
1,619,750
27,470
14,040
314,270
675,500
207,800
792,000
3,657,390
548,610
4,206,000
6,280
168,240
314,000
$ 11,600
3,170,980
47,290
24,370
584,390
1,437,110
406,820
1^584,000
7,266,560
1,089,980
8,356,540
6,280
167,130
628,000
$ 21,760
6,137,800
91,580
45,690
1,075,550
2,644,620
787,250
3,024,000
13,828,250
2,074,240
15,902,490
6,280
159,020
* 20 ft diameter GAC tanks; 30 minute EBCT
Source: Gumerman, Culp &

Hansen, 1979
ro
CO

-------
 TABLE 56.  CONSTRUCTION COSTS FOR STEEL GRAVITY FLOW SAC CONTACTORS*
Cost Category
Excavation & Site Work
Manufactured Equipment
Concrete
Steel
Labor
Pipe & Valves
Electrical & Instrumenta
Housing
SUBTOTAL
M1sc. & Contingency
TOTAL
Volume/Single Contactor,
cu ft
COST/SINGLE CONTACTOR
Total Contactor Volume (cu ft)
T4T, 30(T
$ 7,150
1,327,160
29,680
15,230
263,820
565,490
tion 170,640
624,000
$3,003,170
450,480
$3,453,650
14,140
$ 345,370
282,600
$ 13,140
2,595,980
56,180
28,690
488,740
1,092,710
332,910
1,200,000
$5,808,350
871S250
6,679,600
14,140
$ 333,980
565,200
$ 25,020
5,139,970
111,290
55,180
942,800
2,111,020
659,530
2,280,300
$11,324,810
1,698,720
13,023,530
14,140
$ 325,590
* 30 ft diameter GAC tanks; 30 minute EBCT
Source; Gumerman, Gulp & Hansen, 1979
Operating & Maintenance Costs Of GAC Systems

     Major components of operating and maintenance costs include GAC replace-
ment (for losses incurred during backwashing, handling and reactivation),
GAC contactor operation, regeneration furnace labor and materials, fuel  and
miscellaneous expenses.  Of these, costs for GAC replacement are the most
significant (7% estimated losses during reactivation at more than 55£/1b),
followed by fuel (5,000 BTU/lb of GAC reactivated), regeneration furnace
labor and materials, then contactor operation.

Package GAC Columns--
     Hansen, Gumernan & Culp (1979) estimated operation and maintenance
costs based on the conceptual design parameters given in Table 50, which
assume that GAC adsorber units can be preceded  by filtration and that the
GAC would be replaced with virgin or regenerated carbon once per year.

                                     232

-------
               TABLE 57.  CONCEPTUAL  DESIGN PARAMETERS FOR PRESSURE GAC CONTACTORS
Plant
Flow
(mgd)
1
10
50
No. of
Contactors
2
12
60
Contactor
Diam. (ft)
10
12
12
Total Contactor
Area*** (sq ft)
157
1,357
6,786
Total Contactor Vol.*
(cu ft G> detention times }
7.5 mTii
780
6,790
33,930
15 min
1,570
13,570
67,860
30 m1n
3,140
27,140
135,720
Plant Area**
Requirements (sq ft)
1,750
4,800
21 ,000
* Volumes determined at bed depths of 5, 10 and 20 ft.
** Assumes that GAC contactors are totally enclosed.
*** GAC contactors sized for 5 gpm/sq ft application rates
Source: Gumerman, Culp & Hansen, 1979
IN)
OJ
co

-------
Backwashing is assumed once per week, but the facilities otherwise operate
essentially unattended, except for routine maintenance and monitoring the
performance of the 6AC column.  No allowance for administrative or for
laboratory labor (other than for minimal routine quality assurance testing)
is included.  Estimated operation and maintenance costs for the package 3AC
contactors are summarized in Table 61.

     TABLE 58.  CONSTRUCTION COSTS FOR PRESSURE GAC CONTACTORS*
Total Contactor Vol. 'cu ft) & Area (sq ft)
Tcu ft)
Cost Category (sq ft)
Excavation & Site Work
Manufactured Equipment
Concrete
Steel
Labor
Pipe & Valves
Electrical & Instrumentation
Housing
SUBTOTAL
Misc. & Contingency
TOTAL
Volume/Single Contactor, ft3
COST/SINGLE CONTACTOR
780
157
$ 530
49,010
2,190
1,130
8,500
15,250
15,630
32,550
$124,790
18,720
$143,510
390
$ 71,760
6,790
1,357
$ 1 ,440
409,290
5,650
2,830
55,200
135,310
82,910
125,160
817,790
122,670
940,460
565
78,370
33,930
6,786
$ 6,180
1,944,170
24,730
12,360
262,400
679,880
429,660
512,400
3, 871 ,780
580,770
4,452,550
565
74,210
* 5 ft GAC bed depth; 7.5 minute EBCT
Source: Gumerman, Culp & Hansen, 1979
Gravity Flow GAC Contactors - Concrete Construction—
     Gumerman, Culp & Hansen (1979) estimated costs  for this  type  of  SAC
contactor (5 and 8.3 ft depths,  which equate to 7.5  and 12.5  minute EBCTs,
respectively), which are given in Table 62.   Building  energy  costs are  for
heating, ventilating and lighting.   Process  energy is  required  for backwashing
                                     234

-------
 (once/day for 10 m1n at 12 gal/m1n/sq ft) and GAC slurry pumping during
 carbon removal and replacement (assumes regeneration frequency of every 2
 months and a slurry concentration of 3 1b of GAC/gal of water utilized).
 Process energy requirements are essentially Identical for the two different
 GAC depths.

     TABLE 59.  CONSTRUCTION COSTS FOR PRESSURE GAC CONTACTORS*
Total Contactor Vol. (cu ft) & Area (so rt'
(cu «7
Cost Category (sq ft)
Excavation & Site Work
Manufactured Equipment
Concrete
Steel
Labor
Pipe & Valves
Electrical & Instrumentation
Housing
SUBTOTAL
Misc. & Contingency
TOTAL
Volume/Single Contactor, ft
COST/SINGLE CONTACTOR
1,570
157
$ 530
55,460
2,190
1,130
8,990
16,780
15,680
41,850
$142,610
21,390
$164,000
785
$ 82,000
13,570
1,357
$ 1,490
452,720
5,650
2,830
58,570
147,490
82,910
163,000
914,660
137,200
1 ,051 ,860
1,130
87,660
67,860
6,786
$ 6,180
2,161,360
24,730
12,360
280,050
728,540
429,660
700,290
4,343,170
651,480
4,994,650
1,130
83,240
* 10 ft GAC bed depth; 15 minute EBCT
Source: Gumerman, Culp & Hansen, 1979
     Maintenance material costs Include costs for general  supplies,  backwash
pump and GAC transport pump maintenance, instrumentation repair and  other
miscellaneous items.  The cost for replacement of GAC (lost during contactor
operation and GAC regeneration) is not included.   Labor costs Include the
cost of operating the GAC contactorsT"backwashing pumps, GAC slurry  pumps,
instrument and equipment repairs and supervision.
                                      235

-------
Gravity Flow GAC Contactors - Steel Construction-
     Operating and maintenance parameters are the same as for the GAC
concrete contactors.  Backwash frequency was assumed to be once/day for 10
minutes at 12 gal/min/sq ft and the GAC regeneration frequency was assumed
to be every 2 months.  Makeup carbon is not included in the annual operation
and maintenance costs, listed in Table 63~Tor both 20 and 30 ft diameter
columns (30 m1n EBCTs).

     TABLE 60.  CONSTRUCTION COSTS FOR PRESSURE GAC CONTACTORS*
Total Contactor Vol. (cu ft) & Area (sq ft)
leu ft)
Cost Category (sq ft)
Excavation & Site Work
Manufactured Equipment
Concrete
Steel
Labor
Pipe & Valves
Electrical & Instrumentation
Housing
SUBTOTAL
Misc. & Contingency
TOTAL
Volume/Single Contactor, ft
COST/SINGLE CONTACTOR
3,140
157
$ 530
77,300
2,630
1,240
10,340
18,500
16,420
79,050
$ 206,010
30,900
$ 236,910
1,570
$ 171,520
27,140
1,357
$ 1,400
749,560
6,780
3,110
67,370
221,730
87,100
303,420
1,440,460
216,070
1,656,530
2,250
154,140
135,720
6,786
$ 6,180
3,560,370
29,680
13,600
322,060
1,120,350
451,200
1,332,250
6,835,690
1,025,350
7,861,040
2,250
143,960
* 20 ft GAC bed depth; 30 minute EBCT
Source; Gumerman, Culp & Hansen, 1979
Pressure GAC Contactors--
     Electrical energy requirements were developed assuming that this  type
of GAC contactor serves both as a filter and adsorber.   Backwashing  require-
ments were assumed to be once/day for 10 minutes at a  rate  of 12 gal/min/sq

                                      236

-------
               TABLE 61.   OPERATION AND MAINTENANCE SUMMARY FOR PACKAGE GAC COLUMNS
ro
CO
Plant Flow
_2£2L»
1.7
17
70
175
350
JU2L™
2,500
25,000
100,000
250S000
500,000
Enerq
Building
6,140
15,400
30,800
38,500
46,170
v (kWh/yr)
Process
120
1,200
4,840
9,690
24,210
Total
60 260
16C600
35 , 640
48,190
70 , 380
Maintenance
Material ($/yr)
$ 100
275
1,000
2,650
4,880
Labor
(hrs/yr)
100
100
160
210
260
Total
Cost* ($/yr)
$1,290
1,770
3,670
6,200
9,590
* Calculated using $0.03/kWh and $10.00/hr for labor cost
Source: Hansen, Gumerman & Culp, 1979


-------
ro
OJ
00
               TABLE 62.   OPERATION AND MAINTENANCE SUMMARY FOR GRAVITY GAC CONTACTORS
Total Con-
tactor Vol ,
(cu ft)
Electrical Energy (kWh/yr)
Building 1 Process J Total
Maintenance
Material
($/yr)
Labor
(hrs/yr)
Total Cost*
($/yr)
7.5 minute EBCT; 5 ft GAC bed depth
700
3,500
7,000
35,000
70,000
140,000

1,160
5,810
11,620
58,200
116,200
232,400
44,120
151,850
279,070
1,190,160
2,165,890
4,123,490
12.5
44,120
151,850
279,070
1,190,160
2,165,890
4,123,490
690
3,410
6,820
34,080
68,150
136,540
44,810
155,260
285,890
1,224,240
2,234,040
4,260,030
800
2,510
4,020
13,200
21,600
36,700
minute EBCT; 8.3 ft GAC depth
690
3,410
6,820
' 34,080
68,150
136,540
44,810
155,260
285,890
1,224,240
2,234,040
4,260,030
800
2,510
4,020
13,200
21,600
36,700
900
1,500
2,100
4,600
9,000
18,000
11,140
22,170
33,600
95,930
178,620
344,500

900
1 ,500
2,100
4,600
9,000
18,000
11,140
22,170
33,600
95,930
178,620
344,500
* Calculated using $0.03/kWh and $10.00/hr for labor cost
Source : Gumerman, Gulp & Hansen, 1979

-------
               TABLE 63.  OPERATION AND MAINTENANCE SUMMARY FOR STEEL GRAVITY GAC CONTACTORS*
ro
co
ID
Contactor
Diam. (ft)
20
20
20
20
30
30
30
GAC volume
(cu ft)
31 ,400
157,100
314,000
628,000
141,300
282,600
565,200
Electrical Energy (kWh/yr)
"BullcTing
666,900
3,385,800
6,771,600
12,927,600
2,668,000
5,130,000
9,750,000
Process
12,030
60,170
120,340
240,680
54,150
108,300
216,600
Total
678,930
3,445,970
6,891,940
13,168,280
2,722,150
5,238,300
9,966,600
Maintenance
Material
($/yr)
5,350
21,380
37,420
69,490
16,040
26,730
42,760
Labor
(hrs/yr)
3,000
7,000
14,000
27,000
6,800
13,500
26,000
Total Cost**
($/yr)
55,720
194,760
384,180
734,540
165,700
318,880
601,760
* 30 minute EBCT
** Calculated using $0.03/kWh and $10.00/hr for labor cost
Source: Gumerman, Culp & Hansen, 1979

-------
 ft.   Energy requirements  are  for  backwash  pumping,  for pumping of spent GAC
 to  the  on-site  regeneration facilities and return.  GAC was assumed to be
 removed and replaced  every 2  months.  Energy for supply pumping to the GAC
 contactors  1s not  included, nor are costs for makeup GAC or on-site regenera-
 tion.   The  contactors are assumed to be completely  housed.  Other parameters
 are  the same as for the preceeding types of large scale GAC contactors.

     Annual  operation and maintenance requirements  for pressure GAC contactors
 are  listed  in Table 64.

 COSTS FOR REPLACEMENT GAC

     Costs  were developed by  Gumerman, Culp & Hansen (1979) for purchase and
 placement of virgin GAC for use in any of the above-discussed GAC contactors.
 Figure  81 shows a curve for the total costs of purchase,  delivery and replace-
 ment of virgin GAC.  This curve may be used to derive the complete cost of a
 GAC contactor and to determine the cost of makeup GAC lost during contactor
 operation and GAC regeneration.

 COSTS FOR GAC REGENERATION

     Gumerman, Culp & Hansen  (1979) estimated costs for GAC reactivation
 both off-site (when space limitations or volumes of GAC being regenerated do
 not warrant  installation of on-site equipment)  and the several different
 types of on-site GAC regeneration equipment.

 Off-Site Regional  GAC Regeneration - Handling and "ransportation

     In addition to the capital equipment costs for ozonation and GAC
 adsorption discussed above, when GAC is to be reactivated off-site it will
 be necessary for the plant to have available  carbon dewaterlng/storage bins.
Two different design configurations were used by Gumerman,  Culp & Hansen
 (1979)  to develop cost estimates for these facilities.   Storage bins  of
2,000 cu ft and less are elevated, 12 ft diameter,  30 ft  height,  cylindrical
tanks with conical  bottoms, field-fabricated  of braced,  0.25 Inch,  shop-
formed  steel plate protected by a suitable coating  system.   Bins  of 5.COC cu
ft volumes are elevated, 3-hopper, rectangular  tanks.   For larger storage
requirements, multiple units would be used.

     Construction costs for such GAC dewaterlng/storage facilities are
 listed  in Table 65.  Not Included in these costs are paving for the access
area nor for trucks necessary to haul  dewatered GAC to  the regional  regenera-
tion facility.

     A  summary of operation and maintenance costs for off-site GAC reactiva-
tion is presented in Table 66.  The regeneration plant  is  assumed to  be
within  ICO miles of the GAC-using facility.   Annual  fuel  requirements  for
transportation are based on 3.5 miles/gal  diesel fuel  consumption,  and
maintenance materials for the trucks only were  estimated  at a unit cost of
 $0.30/mi1e.  Included in Table 66 are the costs for fuel,  labor and mainten-
ance to load spent GAC from dewatered GAC storage tanks to 30 cu  yd semi-
dump trailers/haul to the regeneration facility, unload,  reload  reactivated


                                     240

-------
TABLE 64.  OPERATION AND MAINTENANCE SUMMARY FOR PRESSURE GAC  CONTACTORS
Total
Surface
Area (ft2)
157
1.357
6,786
Energy (kWh/yr)
"Process
916
7,967
39.746
Building
179,550
492,480
2,154,600
Total
180,470
500,450
2,194,350
Maintenance
Material
($/yr)
1,600
8,020
37,420
Labor
(hrs/yr)
2,000
3,500
7,500
Total Cost*
($/yr)
27,010
58,030
178,250
* Calculated using $0.03/kWh and $10.00/hr for labor cost
Source: Gumertnan, Culp & Hansen, 1979

-------
GAC from bulk storage, return to the treatment plant  and discharge either to
on-site storage tanks or directly to the GAC  contactors.   It was also assumed
that all these operations would be accomplished within an  8-hr day.  The
costs in Table 66 do not include the costs  of regeneration at the regional
facility.  These woulcTEe charged by the owner/operator at a rate probably
based on the reactivation equipment available,  which  are discussed below.
                10'
                10"
                10=
             (A
              I
             «rf
              III
              o
              u
             o
                10'
                                       1 I  I I I It
                   104          105          106

                         GAC  quantity - Ibs
10'
            Figure  81.  Material cost  for GAC, including cost
                        for  purchase,  delivery and  placement.

                (Gumerrnan, Gulp & Hansen, 1979)


On-Site GAC Regeneration

     Gumerman, Gulp & Hansen (1979)  present details of  capital and operation
and maintenance costs for multiple  hearth,  infrared and fluidized bed 3AC
regeneration equipment.   Conceptual  design  parameters,  construction costs
and operation and maintenance costs  are  presented in Tables 67 through 75
for each of the 3 types  of equipment.  All  regeneration furnaces are assumed
                                     242

-------
to be inside buildings, and operation is assumed to be 24 hours/day, 365
days per year.

 TABLE 65.  CONSTRUCTION CCSTS FOR OFF-SITE REGIONAL GAC REGENERATION --
             ON-SITE HANDLING AND TRANSPORTATION FACILITIES
Cost Category
Excavation and Site Work
Manufactured Equipment
Concrete
Steel
Labor
Pipe & Valves
SUBTOTAL
Miscellaneous & Contingency
TOTAL
On-Site Storage Capacity (cu ft)
1,0'tifl
$ 210
3,240
1,170
5,630
12,090
1,380
23,720
3,560
$ 27,280
5,000
$ 370
13,050
1,750
30,900
29,430
3,830
79,330
11.9CO
$ 91,230
20,000
$ 1,470
50,600
6,360
122,500
123,640
14,990
319,560
47,930
$ 367,490
Source: Gunerman» Culp & Hansen, 1979
     Kittredge (1980) sumnarized the economics of GAC regeneration and
described the costs estimated for a fluidized bed 3AC regeneration system
which has been installed at the Manchester Water Works, Manchester, New
Hampshire.  This system has a designed operating capacity of 12.CCO Ibs/day
(500 Ibs/hr) of regenerated GAC and a projected total operating cost of
6.4<£/lb.  The estimated annual operating costs for the Manchester system are
presented in Table 76.  Notice that the cost of makeup GAC represents nearly
54% of the total direct costs and 43% of the total annual operating costs.
The need for makeup GAC is caused largely by losses Incurred during regenera-
tion.  If the frequency of GAC reactivation could be lowered, a significant
savings in GAC operating costs would be obtained.

     Kittridge concluded that GAC can be regenerated by several different
methods and options at costs ranging from 5
-------
ro
-£»
                             TABLE 66.  OPERATION AND MAINTENANCE SUMMARY FOR OFF SITE REGIONAL  GAC  REGENERATION -
                                            HANDLING AND TRANSPORTATION ONLY
GAC Regene
rated (lbs/yr)
30,000
150.000
500.000
1,000,000
3,000.000
Diesel Fuel* (gal/yr)
10 ml*
haul
5.7
28.6
97
194
582
25 ml
haul
14.2
71.4
243
486
1,430
100 ml
haul
57
286
971
1,943
>,829
Maintenance
Material ($/yr)
10 ml
haul
6
30
110
210
650
25 ml
haul
20
90
280
550
1,640
100 mi
haul
60
320
1,090
2.180
6.540
Labor** (hrs/yr)
10 nl
haul
6.8
34
116
232
780
25 ml
haul
11
55
187
374
1.200
100 nr
haul
14
70
238
476
1,428
Total Cost***($/yr)
10 ml
haul
80
380
1.310
2,620
8.710
25-ml
haul
140
670
2.260
4,510
14.280
100 ml
haul
230
1,150
3.910
7.810
23.440
* Based on 3.5 miles/gal for 30 cu yd semi-dump truck
** Labor for loading and unloading GAC and for hauling
*** Calculated using dlesel fuel at $0.45/gal and labor at $10.00/hr
* All distances are one way
Source: Gumennan, Culp & Kansen, 1979

-------
of GAC treatment (18 m1n EBCT) including GAC regeneration, decreases from
13.3
-------
TABLE 68.  CONSTRUCTION COSTS FOR MULTIPLE HEARTH GAC REGENERATION
Cost Category
Manufactured Equipment
Labor
Pipe & Valves
Electrical and
Instrumentation
Housing
SUBTOTAL
M1sc. & Contingency
TOTAL
Furnace Hearth Area (sq
27 37 47
$220.660
117,720
8,330
8,290
109,670
$ 464,670
69.700
$ 534,370
$275,830
147,150
8,330
3,340
109,670
549.320
82.400
631,720
$ 519,830
273,280
8,330
8,340
124,230
934,010
140,100
1,074,110
359
$ 647,140
346,850
14,480
9,190
175.100
1,192,760
178,910
1,371,670
ft)
732
tl, 039, 660
557,060
23,450
14,930
245,790
1,880,890
282,130
2,163,020
1.509
$1,304,880
704,210
48,800
26,980
334.460
2,419,330
362.900
2,782,230
Source: Gumerman, Culp & Hansen, 1979

-------
TABLE 69.  OPERATION AND MAINTENANCE SUMMARY FOR MULTIPLE HEARTH GAC REGENERATION
Effective
Hearth
Area, ft2
27
37
147
359
732
1,509
Regenerated
GAC,
Ibs/day
1,224
1,670
6,624
13,680
32,400
66,960
Flectrical Energ
Building
14,630
14,630
17,550
23,400
35,100
46,800
Process
261,400
326,750
424,770
588,150
849,550
1,307,000
y, kWh/yr
Total
276,030
341,380
442,320
611,550
884,650
1,353,800
Natural Gas
(scf/yr x 106
5.80
7.72
26.2
48.26
108.40
207.75
Maintenance
Material
($/yr)**
$ 2,990
3,740
6,410
8,550
11,760
16,040
Labor
(hrs/yr'
900
950
3,400
6,200
lO.bOO
17,000
Total Cost*
($/yr)
$ 27,810
33,520
87,740
151,630
284,220
496,730
* Calculated using $0.03/kWh, $0.0013/scf and $10.00/hr for labor.
** Makeup GAC costs are not included
Source: Gumerman, Gulp & Hansen,, 1979

-------
ro
.£»
00
               TABLE /I.  CONSTRUCTION  COSTS FOR INFRARED GAC REGENERATION FURNACES
Cost Category
Manufactured Equipment
Labor
Pipe & Valves
Electrical & Instrumentation
Housing
SUBTOTAL
Miscellaneous & Contingency
TOTAL COST
2,400
$160,000
48,000
3,500
21,000
21,000
$ 2b3,500
38,030
$ 291,530
Furnace Ca
16,800
$360,000
100,000
5,500
53,000
60,000
578,500
86,780
paclty (lbs/da\
38,400
$ 620,000
174,000
7,500
81,000
82,000
964,500
144,680
665,280 1,109,180
J
60,000
$ 940,000
235,000
10,000
113,000
149,000
1,447,000
217,050
1,664,050
Source: Gumerman, Culp & Hansen, 1979

-------
          TABLE 72.  OPERATION AND MAINTENANCE  COST SUMMARY FOR INFRARED GAC REGENERATION FURNACE
GAC Regene-
ration Rate
(Ibs/day)
2,400
16,800
38,400
60,000
Enerqy (kWh/yr
Budding
7,540
39,300
61,300
94,300
Process
701,680
4,522,000
10,206,000
15,820,000



Oldl
709,220
4,561,300
10,267,300
15,914,300
Maintenance \
Material ($/vr)\
8,900
21,000
28,000
33,600
Labor \Total Cost*
hrs/yr) \ ($/vr)
2,380
4,900
9,380
13,300
53,980
206,840
429,820
644,030
* Calculated using $0.03/kWh and $10.00/hr for labor
Source; Gumerman, Culp & Hansen, 1979
IS}
.£»
IO

-------
TABLE 73.  CONCEPTUAL DESIGN PARAMETERS FOR GAC REGENERATION -
              FLUIDIZED BED PROCESS
GAC Regeneration
Capacity (Ibs/day)
6,000
12,000
18,000
24,000
Reactor Bed
Area (sq ft)
4
8
12
16
Housing
Requirements (sq ft)
1,400
1,800
2,200
2,600
Source; Gumerman, Gulp & Hansen, 1979
TABLE 76.  SUMMARY OF ESTIMATED OPERATING CCSTS - FLUIDIZED BED SAC
            REGENERATION SYSTEM, MANCHESTER. N.H. WATER WORKS
Item Costs:
Makeup GAC
Labor
Maintenance
Fuel
Power
Steam
Water
Total Direct Cost
Depreciation
Insurance & Taxes
Administration & Overhead
Total Indirect Costs
Total Annual Operating Cost
$/yr
115,500
9,450
13,000
21,000'
7,560
11,740
36,290
$ 214,540
39,250
7,850
7,850
54,950
$ 269,490
4/1 b
2.75
0.22
0.31
0.50
0.18
0.28
0.86
5.11 tf/lb
0.92
0.19
C.19
1.30 
-------
TABLE 74.  CONSTRUCTION COSTS FOR 6AC REGENERATION - FLUIDIZED BED PROCESS
Cost Category
Manufactured Equipment
Labor
Electrical & Instrumentation
Housing
SUBTOTAL
Miscellaneous & Contingency
TOTAL COST
GAC Regeneration Capacity (Ibs/day)
6,000
$ 570,000
180,000
10,000
60,000
$ 820,000
123,000
$ 943,000
12,000
$ 650,000
205,000
11,000
75,000
$ 941,000
141, IbO
$1,082,150
18,000
$ 710,000
225,000
11,000
90,000
$1,036,000
155,400
$1,191,400
24.000
$ 755,000
240,000
12,000
106,000
$1,113,000
166,950
$1,279,950
Source: Gumerman, Culp & Hansen, 1979


-------
           TABLE 75.   OPERATION AND MAINTENANCE COST SUMMARY FOR GAC REGENERATION - FLUIDIZED  BED PROCESS
GAC Regene-
ration Rate
(Ibs/day)
6,000
12,000
18,000
24,000
Process
Energy
(kWh/yr)
131,400
262,800
394,200
525,600
Natural
Gas
(scf/yr)
6,830,700
13,660,000
20,440,000
27,322,860
Maintenance
Material
($/yr)
$ 15,540
17,940
19,400
20,860
Labor
(hrs/yr)
2,400
2,650
3,050
3,330
Total Cost*
($/vr)
$ 52,360
70,080
88,300
105,450
* Calculated using $0.03/kWh, $0.0013/scf for natural gas and $10.00/hr for labor
Source

: Gumerman, Culp & Hanson, 1979
ro

r\>

-------
ro
on
CO
                 TABLE 77.  COSTS OF GAG TREATMENT (18 MIN EBCT),  INCLUDING  REGENERATION
Type of
GAC unit
Pressure
Pressure
Gravity,
steel
Gravity,
steel
Type of GAC
Regeneration
off -site
on-s1te, IR
furnace
on-s1te, mul-
tiple hearth
on-s1te, mul-
tiple hearth
Size of Treatment
Plantj mgd
2
20
75
110
Total Annual
Cost*
$ 117,920
$ 681,180
$2,097,950
$2,919,760
tf/1,000 gal
treated
23.1 t
13.3 t
10.9 t
10.4 t
* Includes amortized capital costs @ 8% (20 yrs), labor @ $10.00/hr, electricity
@ 3tf/kWh and maintenance materials; also natural gas cost @ 0.175 tf/scf and
dlesel fuel @ 45^/gal, when necessary.
Source; Clark, Gulp & Gumerman, 1980

-------
the total costs decrease much more slowly with reactivation periods  from 3
months (lU/1,000 gal) to 8 months (7<£/l,COO gal).   Thus,  pretreatment
concepts which can extend the GAC reactivation period beyond 2 months  will
have a significant cost-savings impact on the GAC treatment process.

 TABLE 78.  ASSUMPTIONS FOR 18 MINUTE EBCT GAC ADSORPTION
Item
Number of Contactors
Diameter of Contactors (ft)
Depth of Contactors (ft)
Volume of GAC/Contactor
(cu ft)
Design Capacity (rrqd)
"T"
3
8
13
653
5
6
12
13
1,469
10
12
12
13
1,469
100
40
20
14
4,396
150
60
20
14
4,396
Source: Clark, Culp & Gumerman, 1980
 TABLE 79.  AMORTIZED CAPITAL AND OPERATING & MAINTENANCE COSTS  FOR
             GAC ADSORPTION (18-MINUTE EBCT). t/1.000 GALLONS
Item
Amortized Capital Cost
Operating & Maintenance
TOTAL COSTS
Desian Capacity (mgd)
~~T~
29.4
22.1
51.5
5
12.7
12.6
25.3
10
9.8
10.8
20.6
100
4.9
7.4
12.3
150
4.6
7.1
11.7
Source; Clark, Culp & Gumerman, 1980
COST SAVINGS WITH BAC SYSTEMS

     As stated earlier, one primary benefit of optimizing biological  activity
in GAC adsorber systems involves extending the operating time of GAC  adsor-
bers, thus effecting savings in costs for regeneration.   In some BAC  systems
operating in European drinking water treatment plants,  operating lives  of
GAC adsorbers are extended 2 to 5 times (Sontheimer,  1979)  by incorporating
chemical preoxidation with ozone before GAC adsorption.   Schalekamp (1979)
noted that in treating Swiss lake waters for potable  purposes,  the  SAC
operating tine is only 6 months if preoxidation with  ozone  is not used.
When ozonation is placed before GAC filtration/adsorption,  the  GAC  can  be
used for about 3 years, an increase in longevity before  thermal  regeneration
is required of a factor of five.
                                     254

-------
                   34-
                   30-
              o>
             o
             o
             o
24-  •
              in
              O
              o

              o
              *«
              u
              3
              •o
              O
16  •
10

 8-
                          1    234567

                           months between reactivation
               Figure 82.  Production cost for  GAC  adsorbers vs
                          period between reactivations.

                    (Clark, Gulp & Qumerman,  1980)

     The question arises,  however, as to the cost-effectiveness  of  Installing
a high capital cost, energy-Intensive ozonation  system ahead of  SAC adsorption
to attain the cost savings which result from extending the  period of GAC
use.  In order to quantify these savings 1n GAC  regeneration,  a  basis for
comparison and several  operating assumptions was developed  by J.C.  Clark
(1979).

     Ttie primary direct cost benefit of the BAG  process  relative to conven-
tional use of GAC with  frequent reactivation is  the reduction 1n the number
of reactivation cycles  required per year,   ""emple,  Barker & Sloane  (1977)
have shown that the majority of GAC operating  expenses can  be attributed to
reactivation.  For a typical  United States water supply  utility, reactivation
costs account for approximately 70 to 75% of the GAC adsorber operating
costs when six reactivation  cycles are scheduled annually (2-month  GAC
life).
                                      255

-------
     Accepting  this cost basis and drawing upon the Information about the
performance of  BAG presented above, a useful operating assumption for making
these calculations is that the maximized biological activity in the carbon
bed enhanced by chemical preoxidation with ozone  (the highest initial cost
chemical oxidant currently in use) will allow a reduction in the number of
reactivation cycles from six to one annually for  the attainment of comparable
water quality (GAC useful life extended from 2 months to 1 year).

     For the activated carbon adsorber operation, this decrease in number of
reactivation cycles per year will cut annual operation costs by about 50%.
This calculation is based on two key assumptions:  first, that each of the
six reactivation cycles adds about 12% (70 to 75% divided by 6) to the
annual operating cost; second, that a utility required to reactivate its GAC
fill only once  per year will be unlikely to install a smaller, therefore
less efficient, reactivation furnace and carbon transfer facility, and
therefore will  face higher costs for a single off-site reactivation cycle
per year than a facility having its own on-site reactivation system and
which reactivates six times per year.  Instead, it 1s assumed that the
utility requiring only one reactivation per year would choose to send its
GAC out for reactivation.

     The use of chemical preoxidation with ozone will  increase the plant's
operating costs, thereby offsetting to some extent the savings achieved by
the reduction in reactivation frequency.   The anount of this increase depends
upon whether the plant treatment process  already included ozone generation
equipment and contact chambers before the introduction of GAC and also on
the level of preozonation dosage required for the particular raw water
quality to be treated prior to GAC adsorption.

A Hypothetical Case Example  (J.C. Clark, 1979)

     To illustrate the cost savings more  accurately and to avoid the generali-
zations in the  introductory section above, an exanple water treatment plant
is used in this section to apply the principles directly.  Since one objective
of this report  is to assess the potentials of transferring this recently
developed European water treatment practice to the United States, a prototype
U.S. water treatment plant is used as an  example with the technical  assump-
tions based upon successful European experiences with BAG.

     The example plant is described fully 1n an EPA report (Temple,  Barker &
Sloane, 1977).  Briefly, the plant supplies a cotrmunity of 265,000 populaticn
with an average daily production of 50 mgd.  The maximum plant capacity is
75 mgd and its  treatment process includes conventional  prechlorlnation,
coagulation, sedimentation, rapid sand filtration and post-chlorination.

     The capital costs for this plant to  add GAC in a post-filtration
adsorber/contactor mode to achieve 9 minutes empty bed contact times have
been estimated  at $10 million, whereas simply replacing the sand with GAC in
existing filter beds would cost $5 million.  These costs include carbon
transfer systems and reactivation furnaces.  The associated annual  operating
costs, assuming 6 reactivation cycles, are estimated to be $1.1  million for
                                     256

-------
GAC  post-contactors and $1.3 million for replacement of sand by GAC, Including
debt service.

      The capital costs of adding an ozone treatment system capable of
applying ozone at dosages up to 2 mg/1 at 50 mgd are shown 1n this same
report  (Temple, Barker & Sloane, 1977) to be $2.4 million.  Operating costs
for  the full ozonatlon system (air treatment, ozone generation, contacting
and  off-gas destruction) are estimated to total $50,000 per year, Including
the  savings likely to result from reduced chlorine dosage.

     Applying the cost savings attributable to the combination of ozonation
followed by GAC (post-contactor mode) to BAG, a new set of cost calculations
can  be made.  The basic design assumptions are that the plant would install
an ozonation system capable of applying up to 2 mg/1  of ozone at 50 irgd and
add  GAC 1n the existing sand filter beds, providing a 9 minutes empty bed
contact time.  Moreover, the operating costs would include one annual reacti-
vation cycle off-site rather than six cycles on-s1te.

     The resulting capital  cost for the ozonation + GAC system would be
about $5.5 million, slightly more than the $5.0 million cost of the conven-
tional GAC system using existing sand filter beds.

     Even though the capital costs are a bit higher,  the savings are substan-
tial  on the operating cost side.  The annual 3AC system operating costs
would decrease by between $500,000 and $600,000, while the ozonation costs
would increase by only some $50,COO.   Thus annual operating savings of
between $450,000 and $550,000 would be achieved under the assumed conditions.

     When both operating costs and the annual debt service expense on the
capital expenditures are combined, the ozone/GAC system would add $1.1
million per year to the total annual  costs of this typical water utility.
This compares to $1.6 million per year for the conventional GAC system
discussed above, showing the combined ozone/GAC system to be about 302 less
expensive.

Energy Considerations (J.C. Clark, 1979)

     Both the reactivation of GAC and the generation of ozone are energy
intensive processes.  Although the amount of energy needed to reactivate a
pound of GAC varies according to furnace type and operating practice, nost
estimates are in the range of 3,500 and 7,000 BTUs per pound, with 5,000
BTUs per pound being a frequently estimated average.   Estimates of electricity
use  for the generation of ozone are generally 1n a narrower range, with 10.5
kwhr per pound of ozone generated (from air) being a currently accepted
figure.  (This Includes the energy required for air preparation, ozone
generation, controls and instrumentation, but not contacting.)  Since 11,000
BTUs are needed to generate one kilowatt-hour 1n a typical mixture of the
various types of electric generating stations (which generate electricity at
a fuel efficiency of about 33%), each pound of ozone generated requires the
use  of 115,500 BTUs of fossil fuel energy (although nuclear generation of
electricity is a lower energy cost substitute).
                                     257

-------
     The conventional use of GAC with 6 reactivation cycles per year and 2.6
million pounds of GAC per cycle would use 78 billion BTUs annually, plus the
energy needed to produce the 1.1 million pounds of makeup GAC, which 1s
estimated to require another 11 billion BTUs.  Thus the total energy use by
the conventional GAC system would be 89 billion BTUs annually.

     By contrast, the example ozone/GAC system fuel use would consist of one
annual reactivation cycle for the activated carbon, rather than six, plus
the energy used by the ozonatlon equipment.  The single GAC reactivation
would account for 13 billion BTUs plus 2 billion BTUs for manufacturing the
makeup GAC.  The ozone generation would require 1.8 million kilowatt-hours
annually, or 20 billion BTUs of primary fuel energy.  The total fuel use for
the example ozone/GAC system thus totals 36 billion BTUs annually.

     By the measure of energy use, the ozone/GAC system clearly 1s more
efficient under these assumptions.  The use of 36 billion BTUs annually 1s
only 402 of the use by the conventional GAC system of 89 billion BTUs.   If
the GAC reactivation can be prolonged to every two years (as is anticipated
at the Dohne plant, MUlhelm, Federal Republic of Germany), the total savings
would be even more substantial.  On the other hand, if the amount of ozone
required for preoxidation should be 5 mg/1 Instead of 2 mg/1 to prolong the
GAC reactivation time to one year, then the energy costs for ozone/GAC would
be about equal to those for GAC alone which would require 6 reactivations
per year.  In addition, if the types of organic materials present In the
specific raw water to be treated already are biodegradable, such that chemical
preoxidation with ozone is not required and a cheaper oxidant can be used to
provide the same extension of GAC operating life before reactivation is
required (one year), additional cost savings will  be effected.

Sensitivity Analysis and Inflation Effects (J.C. Clark, 1979)

     The preceding text has noted several  assumptions about the operation
and costs of a BAC system operated with chemical preoxidation using ozonation,
as well as the conventional GAC system against which 1t is being compared.
Since many of these assumptions may vary somewhat in future circumstances,
the findings shown above need to be discussed 1n terms of their sensitivity
to key assumptions.  In addition, since 1t appears likely that inflation
will continue, Its effects on tradeoffs between capital investment and
operating costs is important to consider 1n maximizing the benefits of
biological activated carbon systems.

     The basic conclusions noted above were that the ozone/BAC system was
30% less expensive on an annual basis than the conventional SAC system, and
that the ozone/GAC design used 612 less fuel.  More specifically, the annual
costs were lower due to smaller operating and maintenance costs which,  nore
often than not, offset the higher capital  costs of the ozone/GAC combination.
This offsetting effect then was annual1 zed to show the combined effect of
the capital, operating and maintenance costs in the first year of the project.

     Although these cost estimates have been considered above as If they
were exact figures, they are more accurately considered as midpoints of a
range of likely future costs.  This range depends largely on specific


                                     258

-------
 design assumptions which, in turn,  depend upon regeneration frequencies,
 local  differences in sites,  existing plants,  raw water quality,  desired
 finished water quality,  etc.  The range of costs can  vary under  a  typical
 set of characteristics  to plus or minus 2C% of the  point  estimates  noted
 above.   This  range is enough to alter the conparison  substantially,  depending
 upon these factors.

     A more likely influence on costs though,  1s the  pace of  inflation in
 coming years.   The basic direction  of inflation  increases will influence the
 comparison between the ozone/GAC  treatment  combination and  conventional GAC
 systems and favors the ozone/3AC  system.   Inflation will  Increase future
 operating  costs, while capital  costs  will remain  as fixed  amounts based on
 investments made in  a particular year.

     More  specifically,  while  the ozone/GAC system 1s 3C%  less expensive in
 the  first  year  (as shown  above),  it would become  40%  less expensive  in the
 10th year  with  an  average 7% annual  inflation rate.   If inflation should
 rise at a  higher average  rate,  the ozone/GAC combination will be favored
 even more  substantially.

 Effect  Of  Operating Variables Cn GAC And BAG Treatment Costs

     Clark  and  Dorsey (1980) reviewed the cost data developed by Gumerman,
 Culp &  Hansen (1978) to examine the influence of changes  in the level of
 operating  variables on the cost of GAC treatment  (18 m1n EBCT).  Losses cf
 GAC per regeneration cycle of 6% were assumed.  Most variables were fcund to
 have a  greater  impact on smaller than on larger treatment plant costs.  The
 only exceptions to this finding were reactivation frequency and capital
 cost.

     Those  variables having the greatest effect on operation and maintenance
 costs are  the system size, reactivation frequency, activated carbon loss  and
 GAC cost.   Power cost has a greater  impact on  smaller systems than  do some
 of the  other variables.   Figure 83 shows a total GAC system cost versus GAC
 reactivation frequency for a 100 mgd post-filter GAC adsorption system at 3
 different 3AC costs.  Assuming 6C<£/lb for the  3AC being reactivated at a
 frequency of 1.1 months, the total system operating cost is about 16<£/1,000
 gal (point  Pi).  If GAC cost Increases to 80i/lb, the reactivation  period
would have  to increase to 1.7 months (point Pa) for total  costs to  remain
 constant.

     Figure 84  shows an analysis made by Clark & Dorsey (1980) in which
 ozone treatrent is followed by GAC adsorption.  Total  production costs in
 (t/l,COO gal are plotted against GAC  reactivation frequency.  These  authors
 concluded  that  if the system initially requires regeneration each month
without ozonation, then the addition of ozone  would have to decrease the
 reactivation frequency to 2.6 months to break  even (pay for the cost ot
 installing  ozonation facilities).  However, if the system w thout ozonation
 has an  initial GAC reactivation frequency of 2 months, Clark & Dorsey cone  ude
 that no  break-even point exists to justify the costs of addition of ozonation.
                                     259

-------
           35-
        OJ
       o
       o
       o
        in
        O
        u
        c
        o
        *«
        o
        3
        •o
        O
                                   1.00/lb GAC

                                   0.80
                                   0.60
                   0.5    1.0   1.5   2.0   2.5   3-0   3.5   4.0

                      reactivation frequency, months
            Figure 83.
Total production cost vs.  reactivation frequency
for SO.60, 0.80 11.00/lb  GAC  for  100 mgd
contactor system.

 (Ciark & Dorsey,  i960)
Practical Examples
Drinking Water Installations —
     Dohne Plant, MUlheim, Federal Republic of Germany—Heilker (1979) has
noted that a BAG process (preozonation, flocculation, sedimentation, ozona-
tion, dual media filtration, GAC, ground passage and safety chlorination)
has replaced the older process (breakpoint chlorination, flocculation,
sedimentation, GAC, ground passage and safety chlorination).   The older
process used breakpoint chlorination for removal of ammonia;  the newer
process removes ammonia biologically.

     Heilker's statement regarding cost comparisons between the two processes
1s as follows:

     "The treatment plants in the Cohne waterworks have been  operating for
     more than 1.5 years using the revised process.  The drinking water
     quality has been significantly improved without increasing treatment
     costs.  The Dohne plant Is less susceptible to disturbance and as a
     result can be operated with 50% of the former staff size.   The activated
     carbon filter runs are 3 to 5 times longer than before."
                                      260

-------
           ID
           CO

           o
           o
           o
           o
           o
           c
           o
           *«
           o
           3
           •o
           o
               20- -
               15- -
10- -
8 mg/L
6 mg/L
4 mg/L
2 mg/L
                                                    GAC without ozone
                                          •4-
                       24      6     8     10    12

                         reactivation frequency, months
                                             14
            Figure 84. Cost of ozone and GAC in combination.

                         (Clark & Dorsey, 1980)

     Dohne's new process required doubling the height of the GAC adsorbers,
adding ozone generation facilities and pure oxygen evaporators (for addition
of oxygen after filtration and before GAC adsorption), all  of which increased
costs.  On the other hand, savings were obtained in the amount of chlorine
used (10 to 15 mg/1 dosages for prechlorination were eliminated), GAC regene-
ration costs (every 4 months with the old process, 1 to 2 years with the  new
process), and labor costs (half the former plant operating  staff).

     Windhoek Process. South Africa—Van Leeuwen (1979) has indicated that
the 1 mgd process used at Windhoek, South West Africa to reclaim sewage for
potable purposes, since 1968, involves breakpoint chlorination 'ollowed by
2-stage GAC adsorption.  Under these circumstances,  the GAC must be regenera-
ted every 90 days.  By contrast, when breakpoint chlorination was replaced
by ozonation (8.5 to 10.5 mg/1 applied dosage at 95% ozone  transfer efficien-
cy) 1n a 1 mgd pilot plant process, the 3AC was operated for at least one
year without requiring reactivation.  This resulted in overall  process cost
savings of 13% at 1 mgd, and are projected to reach 24% at  10 mgd.   These
cost savings are dependent upon the clarification agent used, lime  treatment
being more costly than ferric chloride.  Longer periods of  GAC use  without
reactivation (greater than 1 year) are expected to result 1n even higher
cost savings.
                                      261

-------
      Inclusion of  the ozonatlon unit process added 1.8  to 3.3£/cu m  (6.81  to
 12.49<£/1,000 gal)  In costs  to the 1 mgd treatment line.  This caused a
 decrease  1n GAG operating costs of 2.5 to 5
-------
                                   SECTION 12

               ENGINEERING DESIGN CONSIDERATIONS  FCR BAG  SYSTEMS
 INTRODUCTION
      As  a  first  step,  the  parameters  to  be  considered in designing biological-
 ly enhanced GAC  systems are the saire  as  those which must be taken into
 consideration when designing granular activated carbon beds or columns to
 act purely as adsorbers.   However, there are several additional parameters
 related  to optimizing  the  biological  aspects of specific organic materials
 present  in the system  being treated which must be considered.  These include
 control  of pH and DO levels, pretreatment to remove substances toxic to
 microorganisms, and possibly increased frequency of column backwashing
 because  of increased biological growths.

      Initially, adsorption isotherms  should be measured for the organic
 components present in  the wastewater and which are desired to be rerroved.
This can be accomplished by the procedure described by Rodman, Shunney &
 Perrotti (ly78).   Several  concentrations of the wastewater to be treated are
allowed  to cone to equilibrium with weighed samples of pulverized GAC.   The
equilibrated mixtures are filtered and the 'iltrates are analyzed for the
constituents of concern.   Results are plotted on log-log paper,  with  the
abscissa being in units of Impurities remaining in solution and with  the
ordinate being in units of impurities adsorbed per unit weight of 3AC.

     The phrase "allowed to come to equilibrium" should not be viewed as a
simple matter of several  minutes, several hours, or even several  days.   Peel
& Benedek (1980a) found that the adsorption of phenol  required up to  3  weeks
to attain equilibrium,  and that o-chlorophenol  required up  to 5 weeks to
attain equilbrium.  Sufficient time must be allowed for equilibrium  conditions
to be attained, in order to result in an accurate measure of adsorption
isothems.  The required time to attain equilibrium conditions  can vary,
depending upon the nature of the specific organic materials present.

     Adsorption isotherms provide a good estimate of the effectiveness  of
activated carbon to adsorb the impurities present in the wastewater  to  a
given level.   They also provide an indication of the maximum amount  of
impurities which can be adsorbed by the GAC being tested.   Because of this,
adsorption isotherms are helpful in selecting the most suitable  type  of GAC
to adsorb specific impurities.   However, it must be recognized that  adsorption
 isothems are equilibrium measurements, and will  not indicate how the GAC
will  perform during continuous operation.  Continuous  flow  studies on pilot
plant scale are the best way to obtain data from which to design full scale
plant adsorbers.

                                      263

-------
     It should also be borne in mind that wastewaters which contain high
levels of readily biodegradable organic materials usually can be treated
biologically at much lower cost than by GAC adsorption.  Appropriate biologi-
cal treatment of high strength, highly biodegradable wastewaters also will
lower the amount of GAC subsequently required for adsorption of refractory
contaminants, thereby lowering overall system costs.

     After adsorption isotherms have been determined, empty bed contact
times can be developed for the type of GAC best suited to the adsorption of
the particular pollutants to be removed from solution.  At this point, those
factors which will  affect biological growth and proliferation should be
considered.

PARAMETERS AFFECTING BIOLOGICAL GROWTH IN GAC MEDIA

     The major parameters to be considered in optimizing BAG systems for
treatment of industrial  wastewaters are:

(1)  Presence of materials which are toxic to biological growths (pesti-
     cides,  heavy metals, etc.) — if present, the wastewater must be
     pretreated to remove or destroy them,

(2)  Chemical nature of materials to be removed (organic* and/or ammonia)  —
     necessity for chemical  preoxidation  of organic constituents versus
     preaeration or preoxygenation,

(3)  GAC pore size  distribution — presence of a high proportion of rracropores
     to house bacteria,

(4)  Necessity for  sand  or other "inert"  media filtration before GAC adsorp-
     tion -- for removal  of insoluble nateHals generated by preoxidation
     and as  support media for possible biological  activity,

(5)  Need for supplemental  oxygenation before GAC  adsorption —  to increase
     rate of formation of C02 and maintain aerobic conditions,

(6)  Extension of empty  bed contact tiire  over that determined by adsorption
     isotherms — usually 11  to 18 minutes EBCT will  be required for BAC
     versus  5 to 15 minutes for GAC acting by adsorption only,

(7)  Initial period of operation (several  days) required to  develop and
     establish biological growths.  Nitrifying organisms require several
     weeks to develop,  and may not develop at all  if overpowered by other
     species which  may be present.  Necessity for  biological seeding and/or
     adaptation of biological organisms to organic substrates to be removed
     from solution.   Possible need to add  nutrients (nitrogen and/or phospho-
     rus)  if feed solutions are deficient,

(8)  pH must be maintained between 5 and  9 (preferably between  6 and 8) for
     optimum biological  activity,
                                     264

-------
 (9)   Methods for controlling BAG media — analytical  monitoring for dissolved
      organic carbon,  TOC,  COe (produced biochemically),  dissolved  oxygen,
      ammonia, nitrate,  nitrite,  bacterial  levels  In 3AC  effluents,  etc.,

 (10)  Determination  of 3AC  regeneration parameters  —  breakthrough  of surro-
      gates  or specific  compounds,

 (11)  Backwashing of 3AC and  prefilters ~  need  for air scouring  1n  addition
      to water backwashing,

 (12)  Materials of construction,

 (13)  Safety of GAC  systems.

      Each of these  design parameters will  be discussed in more detail
 below,

 DISCUSSION  OF SPECIFIC  DESIGN PARAMETERS

 Raw Water Con-position

     There  are three  important parameters  of the raw water being considered
 for treatment by  biologically enhanced granular activated carbon systems.
These are:

      (a)  The toxicity of the organic materials present to the biological
          organisms present in the support media,

      (b)  The bicdegradability of the organic materials present (with or
          without a preoxidation treatment step),

      (c)  The pH of the aqueous solution to be treated.

      If materials are present which are toxic to biological  growth, they
must be removed from solution in order that biological growths 1n the inert
media filters and GAC adsorbers rray survive and proliferate.   Common toxicants
include heavy meta1ss cyanides and certain organics such  as  persistent
pesticides  (polyhalogenated materials).

     If the organic materials present are not biodegradable  but are adsor-
bable, they will  be adsorbed by 3AC.   However,  as soon as the GAC medium
becomes saturated with these compounds, they will break through and reactiva-
tion will  be necessary.   On the other hand, 1f chemical  oxidation will
render these refractory organics even slowly biodegradable,  then they can
become nutrients for the bacteria present.  As a result,  the useful adsorption
life of the GAC adsorbers will be extended.

     Preliminary batch testing using ozonation will indicate whether a
chemical preoxidation step will enhance the biological removal of organic
contaminants.  The  blodegradabillty of the wastewater to be  treated should
be measured  (say by oxygen uptake) before ozonation and then again after
ozonation at various dosage levels.  If the biodegradabHUy does not. increase

                                      265

-------
 upon oxidation,  then biological  activated  carbon  will  be  no more  effective
 than treatment with GAC acting as  an  adsorber  alone.

      In conducting  these preliminary  biodegradation batch  tests,  care
 should be  taken  to  use  microorganisms which have  become acclimated to the
 oxidized organic contaminants.

      Biological  growths occur  best at pH 6 to  8 and not at all at high or
 low  pH values.   Therefore, acid  or alkaline wastewaters will have to be
 adjusted to  this pH  range before passage through  GAC.  In addition, if
 pretreatment of  the  wastewater is  necessary, this pH factor must be
 considered.

 Type of GAC

      The GAC utilized for treating the specific wastewater should have a
 high adsorptive  capacity for the particular organic materials present.  In
 addition to  this, the work of Eberhardt, Madsen & Sontheimer (1974) at the
 Bremen,  Federal  Republic of Germany waterworks showed that GAC which contained
 a high  proportion of macropores  removed a higher proportion of DOC from
 solution than did GAC containing a smaller percentage of macropores.   Normal-
 ly,  macropores in GAC constitute about U of the total  surface area.   The
 more  effective GAC used by Eberhardt, Madsen & Sontheimer (1974)  contained
 about 2% macropores.  At biological equilibrium, the 2% iracropore GAC removed
 about 50" of the DOC present, compared with only 25% to 33% DCC removal
 using GAC containing 1% nacropores.

     These data  indicate that research should be conducted on  GAC which  has
 been designed to contain even higher levels of macropores.

 Necessity For Filtration Before GAC Adsorption

      In  European drinking water treatment plants,  it has  been  noted that
 ozone oxidation causes "nlcroflocculation"  of dissolved organic materials.
This is  explained on the basis of oxidative formation  of  carboxyl, alcohol
 and aldehydic groups, which are capable of forming hydrogen bonds with
 similar  groups present in other organic molecules  (Maier,  1979).   Such
 hydrogen bonding effectively increases the  molecular weights  of dissolved
organic materials present.  If these molecular weights  become  sufficiently
high, the materials will come out of solution in the form of  floes.

     Furthermore, polar carboxyl  groups are capable of  linking  with inorganic
cations normally present in waters  e.g., calcium,  magnesium,  iron, aluminum,
etc.   Such linkages with polyvalent inorganic cations also cause  precipitation
of dissolved organic materials.

     Formation of such Insoluble materials  can be  significant,  and,  therefore,
can cause fouling of GAC adsorbers, blinding of the adsorption  sites,  or
 simply cause premature headlosses.   As a minimum,  buildup  of  such Insoluble
materials can require that the GAC adsorbers be backwashed more frequently
 than would be necessary in the absence of these insoluble  materials.   Thus,
 in those cases when chemical  preoxidation is required,  or  when  chemical

                                     266

-------
 flocculants are added In pretreatment,  a  filtration  step  should  be  inserted
 prior to GAC adsorption.

 Need For Supplemental  Oxygenation  Prior To GAC Adsorption

      When one atom  of dissolved organic carbon is converted  into 1 molecule
 of C02,  2 atoms of  oxygen  (1 molecule)  are required.  When 4 atoms of hydrogen
 (the maximum number which  can be associated with the dissolved organic
 carbon atom)  are converted to water, still another molecule  of dissolved
 oxygen is required.  To convert 1 molecule of ammonium ion to nitrate ion,
 hydrogen  ion  and water stoichiometrically requires at least  4.57 molecules
 of dissolved  oxygen.

      For  these  reasons, it is important that sufficient dissolved oxygen be
 present in  solution so that biochemical oxidations can proceed as nearly to
 completion  as feasible.  The solubility of oxygen in water when added by
 aeration  is between 6 and  10 mg/L, depending upon the water  temperature.  On
 the  other hand, when pure  oxygen is added to water,  dissolved oxygen concen-
 trations  as high as 45 to  55 mg/L can be achieved,  again depending upon the
 water tenperature.

     Therefore, it is important for the wastewater treatment chemist to know
 the  biodegradability of his wastewater, so that sufficient oxygen can be
 provided  to maintain the biological organisms in an  aerobic condition.   He
 should also know how much  biodegradation occurs in  the inert media filters,
 so that sufficient oxygen can be supplied prior to  this  treatment stage as
well.

     Dissolved oxygen requirements can  be determined  by  measuring oxygen
 uptake in a Warburg type of apparatus.   When  conducting  such  measurements,
 however,   it is Important to be sure that the  bacteria used have had  tine to
 adapt to  the particular organics present.   Depending  upon  the amount of
 biodegradable materials present in the  wastewater, provision  should  be  made
 to provide sufficient DO in the GAC influent  so as to maintain  a  level  of DO
 in the GAC effluent of at least 3  mg/L  (Sontheimer,  1978).   At  the Cohne
 drinking water plant (Mdlheim,  Federal  Republic of Germany),  for  example,
 the  CO in the GAC column effluents is controlled at  7 to 8 mg/L (Heilker,
 1979b).

     If chemical preoxidation of the wastewater is required,  the  use of
 ozone generated from oxygen will provide a higher DO  level  than will  ozone
 generated from air.

 Empty Bed Contact Time

     Adsorption of dissolved organics by GAC is a rapid  process.   Therefo-e,
 empty bed contact times of GAC columns  which act strictly  as adsorbers
 normally  are on  the order  of 5 to 15 minutes.  On the other hand, biological
 oxidation processes are slower than adsorptive processes,  and longer EBCTs
 are  required.   In the Federal Republic of Germany,  for example, GAC  columns
 2  meters  high were first Installed in drinking water treatment plants in
 Dtisseldorf  and  in Mulheim  for taste and odor control  and for dechlor1 nation,


                                     267

-------
respectively.  When the benefits of biological activity in GAC media were
recognized, however, the depths of the Ddsseldorf and Mulheim GAC columns
were doubled to 4 meters.  Not only did this produce longer EBCTs, but also
provided greater adsorptive capacities for trace organics.  On the other
hand, capital costs for GAC are approximately doubled, as are costs for the
initial charges of GAC.

     On the basis of European drinking water treatment plants using BAG
systems, it appears that EBCTs of 7 to 10 minutes are optimum for maximizing
the benefits of biological activity.

Initial Period Of Biological Adjustment

     From European drinking water experiences, virgin GAC columns require
several days to a week of feeding with water to be treated in order for
significant biological activity to develop which converts COC to C02.
Nitrifying organisms require even longer to develop, on the order of 2 to 3
weeks.  It is probable that some industrial wastewaters will  contain specific
organic components which require even longer for adaptation of microorganisms.
For example, Benedek e_taj_. (1979) have shown that after ozonation, solutions
of p-nitroaniline are not readily biodegradable (as measured  by oxygen
uptake rate) until after about 3 weeks of acclimation of the  microorganisms.
These microorganisms previously had been acclimated to p-nitroaniline.

     Once the bacteria have adapted to the particular industrial  wastewater,
however, continuous operation of GAC systems should be easily maintainable.

Analytical Monitoring

     Measurements for specific contaminants of the industrial  wastewater
being treated,  COD, TOC, ammonia, dissolved oxygen, etc.,  in  the influent
and/or effluent of the GAC adsorbers are apparent.   In addition to these
determinations, however, it is also helpful to analyze effluents from
biologically operating filtration or GAC adsorption media for carbon dioxide
content (which  can be followed in some cases by accurate determination of
pH).   By comparing the amount of inorganic carbon produced (the amount of
elemental carbon contained in the C02 measured) with the amount of dissolved
oxygen being consumed during the same period of time, one can determine that
his treatment system is or is not in biological equilibrium.   At biological
equilibrium, the ratio of DO consumed to inorganic carbon produced will te
close to 1.

     Knowing the ratio of Inorganic carbon produced to the amount of orcanic
carbon removed from solution also is useful.  If this ratio is greater than
1, more CO? is being produced than organic carbon is being removed from
solution.  This situation indicates that biological regeneration °f some of
the active adsorption sites of the GAC is occurring   Converse Y. when the
ratio is less than 1, this indicates that more dissolved organic carbon :s
being removed by adsorption than by biochemical decomposition.

     As indicated in preceding sections, it is Important to  provide suffi-
cient DO to the solutions being treated to maintain the bacteria in an

                                      268

-------
 aerobic state.   Since industrial  wastewaters  fed  to  GAC  adsorbers  are
 expected to contain considerably  more  DOC  than  drinking  waters,  it likely
 will  be necessary to add  pure  oxygen before sand  or  other  inert  media  filtra-
 tion.   This can  be accomplished during ozonation,  in  the event that chemical
 preoxidation with ozone is  chosen as a pretreatment  step,  by generating  the
 required ozone from pure  oxygen.

     Monitoring  the sand  or inert medium filter influents  and effluents  for
 dissolved oxygen  also will  be  necessary, so as to  determine the  amount of 20
 consumed biologically during filtration, as well as  to determine the amount
 of additional oxygen  necessary to be added prior to  passage o* the  solution
 through  biologically  enhanced  GAC.

 GAC Regeneration  Parameters

     GAC regeneration parameters  should be developed on  the basis of the
 specific organic  materials which  must be removed from the wastewater and for
which_GAC was selected as the  treatment process to deal with them.  If
chlorinated  organics are the contaminants to be removed,  for example,  then
GAC reactivation  frequency should be based upon their breakthrough.  Once
the specific polluting contaminants to be removed have been identified, then
specific or surrogate chemical  analytical procedures can  be developed  for
lower cost monitoring.  The use of TOC as a surrogate analysis probably will
be a useful  procedure in this regard.

     It  should be recognized that the objective of GAC treatment 1s to
remove those contaninants  which can be removed by adsorption;  the objective
of promoting aerobic biological activity In the GAC is to remove as much of
the biodegradable fraction as possible.  Under Ideal  circumstances,, break-
through of the GAC system will  be determined only by those chemicals which
are adsorbed but which cannot be made biodegradable.   If  breakthrough  of
biodegradable organics occurs first, this means that the  system has not yet
been designed optimally.

Backwashing of Filtration  Media and GAC Adsorbers

     Air scouring is used  initially during backwashing 1n all  European
drinking water treatment plants known to be using  biologically enhanced SAC
adsorption processes.  Air scouring is  required to loosen the  biological
growths adhering to the filtration or GAC media.   In actuality,  air scouring
is more akin to "bumping", and  as soon  as the  agglomerated filtration  or GAC
medium has been broken up  by such bumping,  then backwashing is continued
with water.

     Biologically operating European sand filters  normally are backwashed
first by air scouring, then with a mixture of  air  plus water,  then with
water alone.  In backwashing 3AC media, however, air scour then water
backwashing  is practiced.   Apparently the density of GAC  is too low to allow
use of the  (air + water) treatment without incurring significant losses of
GAC.
                                     269

-------
     Air  scour  or  air  backwash  is a concept which has found considerable
 application  recently in wastewater treatment, but little or no application
 in United States drinking water facilities.  In water treatment practice,
 the combined effects of heavy prechlorination dosages as well as so-called
 surface washing have been sufficient to enable satisfactory filter cleaning
 during the backwashing cycle.

     A report of the American Water Works Association Subcommittee on
 Backwashing of Granular Filters (Cleasby et al., 1977) recommends provisions
 for air scour or for surface wash when comTitTbns exist sjch as those which
 occur with biological activity promoted in the beds.  While surface wash
 units are common in modern U.S. water treatment plants,  air scour systems
 are not.  Cleasby et aj_.  (1977) categorized backwashing  methods as follows:

 (1)  High rate backwash - full bed fluidization and substantial bed expansion
     (20 to 50%) normally is proceeded and followed by a Icwer rate backwash.
     This backwash system can be used for single or multi-media filters.

 (2)  Low rate backwash with little or no bed fluidization or ted expansion -
     auxiliary scour is essential  to low rate backwash.   This  backwash
     system can be used for single nedium filters only.

 (3)  Water backwash with surface water wash only.

 (4)  Water backwash with air auxiliary:

     (a)  Air scour followed by low rate water backwash  - for  use with
          single medium filters only.

     (b)  Air scour followed by high rate water backwash -  for use with
          single and multi-media filters.

     (c)  Simultaneous  air scour and low rate water backwash,  followed by
          high rate water backwash alone - for use with  single medium
          filters alone.

     (d)  Simultaneous  air scour and low rate water backwash,  followed by
          high rate water backwash alone - for use with  single and multi-
          media filters.

     It appears that filter backwash techniques nust be  reviewed if chlorina-
tion ahead of filtration 1s discontinued.   Based on the  AWWA Filtration
Committee's findings and European  practice, provisions of either air scour
or surface wash appear to be necessary.  Provision of surface  wash without
prechlorination would appear to result in shorter filtration runs with
associated Increases 1n operating  costs.   For example, a midwest water
treatment utility with  filters equipped with surface wash estimated a re<;uc:
tion in backwash cycles from 60 hours  to 30 hours.   Conversion of a filtration
system from water backwash to water backwash with air auxiliary would require
complete reconstruction of existing filter bottoms to accomodate the air
distribution systems.
                                     270

-------
      In those European water treatment plants  in  which  biologically  enhanced
 GAC has been installed to  replace  breakpoint chlorination  followed by  GAC
 later in the process  for dechlorination (for example, at the 3ohne plant,
 Mulheim, Federal  Republic  of Germany),  both the sand filters and the 3AC
 adsorbers have been found  to be able  to operate nearly  twice as long without
 breakpoint chlorination and  with ozone  preoxidation before backwashing
 becomes necessary.  This is  clear  indication that the anticipated biological
 fouling of filters when prechlorination is eliminated may not occur, or, if
 it  does, that it  may  be controlled by  the low  level preozonation required to
 promote aerobic bacterial  activity in  the filter and adsorber rredia.

     The frequency with which backwashing of biologically enhanced filters
 or  GAC  adsorption media will be required usually is determined by headlosses.
 However,  at  the Cohne  drinking water treatment plant (Mulheim, Federal
 Republic  of Germany),  filter runs as long as 1  week could be achieved before
 headlosses built  up sufficiently to require backwashing.  On the other hand,
 during  the summer of 1978,  a "population explosion" of nematodes was observed
 to  occur  in the rapid  sand filters and SAC adsorbers.   Since the reproduction
 cycle of  these organisms is just over 3 days,  it was concluded that nematode
 development was a result of the more than one week backwashing intervals.
 When backwashing  intervals  were lowered to 3 days, the nematodes disappeared
 completely (Heilker, 1979).

     Valencia and Cleasby  (1979) have recently published an excellent
 discussion of ve^dty gradients in granular filter backwashing, which
 provides  the engineer with  a valuable tool  for analyzing shear forces in
 filter  beds and their effects upon the media cleaning  process.

 Materials of Construction

     Granular activated carbon systems will  cause  corrosion problems  1f they
 are not anticipated and provided for in the original  design.   It 1s  not
 expected, however, that there will  be any significant  increase in  corrosion
 behavior of GAC media as a  result of operating  with maxinized  biological
 activities.  Perhaps the best source of information as  to antl-corrosion
measures  to be taken is the carbon supplier himself, as  well as  operational
 plants where GAC has been in use.

     At the new Kra'ingen drinking water treatment plant in Rotterdam,  The
 Netherlands, the granular activated carbon adsorbers  are constructed  of
 carbon  steel, but are coated on the inside with a  proprietary  coating.   GAC
 is  supported on a coated steel plate with plastic  nozzles inset  in  the  plate
 to  provide flow distribution (Rice et_al_.,  1979).

     Heilker  (1979) has described construction and corrosion inhibition
 aspects of the GAC adsorbers recently enlarged and rebuilt at  the  Dohne
 drinking water  treatment plant in MUlheim,  Federal Republic of jermany.
 Steel GAC adsorbers were selected initially, because they were better
 suited  for overall process design and were easier  to work with in modifying
 existing  water  treatment plants.  However,  corrosion problems  caused high
 repair  costs.


                                     271

-------
     After  10 years of testing different adsorbers with additional cathodic
corrosion protection, a method has been adopted at Mill helm which offers
"nearly complete protection against defects".  The 12 double-staged adsorbers
are being modified as follows:  all interior welded seams are being straigh-
tened and ground.  Rust is being removed by sand blasting.  Condensate is
prevented by aeration during coating.  Solvent-free epoxy resin, resistant
to abrasion and to ozone, 1s used, and a 450 to 500 ym layer is applied.
The pore!ess coating is tested with 1.5 to 2 kV.

     Further experimentation has shown that the interior adsorber surfaces
can be "completely protected" if a combined cathodic protection with sacrifi-
cial anodes is used and the electric power supply is transmitted over plati-
num-covered titanium anodes.

     Costs for all  corrosion protective devices being applied to GAC adsorbers
at Mill helm amount to approximately 10% of the total  cost of the treatment
plant modifications required to convert the old (breakpoint cnlorination)
process to the BAG process recently installed (Heilker,  1979).

Mechanical Considerations
     Culp (1980) reviewed GAC use in advanced wastewater treatment (AWT)
systems and pointed out that the experience has varied from excellent to
very poor.  Culp also states that the failures of 3AC systems in AWT applica-
tions have not stemmed from deficiencies in the basic GAC processes  of
adsorption oTbrganics, but, rather, from mechanical  probleirs,  which have
included:

e    failure of GAC column linings due to improper Installation or faulty
     material,

•    corrosion of GAC storage tanks,

t    corrosion of surface wash equipment and other metal parts  in GAC
     contact tanks,

t    numerous failures of GAC column underdrains, where the wrong type of
     sand filter type underdrains had been installed, or where  the Installa-
     tion was defective,

•    failure to make provisions for maintaining aerobic conditions at all
     times in the GAC column influent water,

•    failure to provide means for excluding air or venting air  from GAC
     column backwash lines prior to initiation of backwashing,

ft    lack of adequate means to dewater and feed measured amounts of spent
     GAC to the reactivation furnace,

•    corrosion of furnace parts.
                                     272

-------
 •     installation  of  sub-standard or defective refractories 1n reactivation
      furnaces,

 •     improper  installation of furnace refractories,

 •     inadequate pretreatment of wastewater prior to application of GAC,

 •     lack of adequate means for GAC movement and conveyance, apparently due
      to the lack of knowledge concerning the physical handling properties of
      GAC.

      Culp (1980) concludes that any and all of these rather serious problems
 could have been avoided quite easily by proper application of the best
 current engineering design knowledge and practices for GAC systems.

Taking GAC Filter/Adsorbers Out Of Service

     Plant personnel at European drinking water treatment plants  which have
been using granular activated carbon for some years advise that GAC columns
or beds that have been in plant operation should  not be allowed to stand
idle or off-line for more than one day.   If this  caution is not heeded, the
biomass in the activated carbon medium changes and when the unit  is placed
back in-line, contamination of the product drinking water occurs.   When
operating units are taken out of service, the activated carbon is  sent for
reactivation.  If the units are to be taken off-line only for  a few days,
water is passed through the units continuously until  they are  placed back 1n
line  (Engels, 1978, Private Communication).

Safety Considerations oF GAC Adsorbers

     Strudgeon e_t_ al_.  (1979) have discussed several serious accidents which
have occurred with GAC systems and which have resulted in at least two
fatalities.   These accidents occurred in adsorbers which contained wet GAC,
which is an  excellent adsorber of atmospheric oxygen.   In two  instances,
workers apparently entered enclosed GAC adsorbers without positive pressure
masks to provide oxygen and were suffocated because of the lack of sufficient
oxygen.

     Other potential hazards associated with GAC  systems include  the
following:

 (1)  GAC in dry form contains significant quantities of fines  and dust.  As
     a result, all GAC handling systems pose the  possibility for  creating a
     dust explosion hazards unless precautions are taken.

 (2)   If biological growths within 3AC adsorbers are allowed to become
     anaerobic, and if sulfates are present, quantities of HeS can be
     generated.  This toxic gas also can pose a hazard to workers entering
      the adsorbers, although its presence usually can be readily detected
      nasally in small quantities.
                                     273

-------
(3)  When dry and confined, GAC provides a relatively solid footing.
     However, after having been fluidized with the proper amounts of water,
     it has the characteristics of a mixture with virtually no support, thus
     allowing a heavier weight to sink within the mixture.  The term "quick-
     sand effect" has been applied to this meretricious behavior of GAC and
     similar systems.

(4)  GAC also will "bridge", as soil does, and personnel entering a GAC
     vessel must insure that the bed is well  drained and that bridging has
     not developed, which possibly could collapse and bury the individual.

Alternative GAC Adsorber Designs

     Recently, Carnes and Burstein (1980) have addressed all  of the engineer-
ing factors involved with designing GAC adsorption systems.   This elegant
treatise should be consulted by those wishing more detailed  information on
this subject.  In addition, Bernardin (1980)  has discussed the subject of
problems and solutions of granular activated  carbon operations.
                                     274

-------
                                 REFERENCES
American Water Works Association,  1971,  Water Treatment Plant  Design, Am.
     Water Works Assoc., Inc.,  Denver,  Colorado.

Anonymous, 1979, "Space-saving  Wastewater Treatment",  Chem.  Engrg.,  July  2;
     p. 47-8.

Argo, D.G., Culp, R.L. & Hansen,  S.P.,  1978,  "Ozone  and Biological Activated
     Carbon", Presented at Am.  Chem.  Soc., 176th  Nat!.  Meeting,  Miami Beach,
     Florida, Sept. 11-15.

Arisman, R.K., Crase, T.C., Zeff,  J.D.  & Musick,  R.C.,  1979,  "Destruction of
     PCBs in Industrial and Sanitary  Waste Effluents by the  Ultrox  (UV-
     Ozone) Process", presented at AIChE 86th National  Meeting,  Houston,  TX,
     April 1-5.  Am. Inst. Chem.  Engrs., New  York,  N.Y.

Baker, M.N., 1949, The Quest For Pure Water,  Am.  Water Works Assoc., Denver,
     Colorado.

Batchelor, J.S., 1977, "Activated Carbons for Effluent and Water Treatment",
     Effluent & Water Treatment J., April, 1977.

Benedek, A., 1977, "The Effect of Ozone on Activated Carbon  Adsorption  — A
     Mechanistic Analysis of Water Treatment  Data",  Presented  at Intl.  Symp.
     on Advanced Ozone Technology, Toronto, Ontario, Canada,  Nov.,  1977.
     Intl. Ozone Assoc., Vienna, VA.

Benedek, A., Bancsi, J.J., Malaiyandi,  M. & Lancaster,  E.A.,  1979,  "The  Effect
     of Ozone on the Biological Degradation and Activated Carbon Adsorption
     of Natural and Synthetic Organics in Water.   Part II.   Adsorption",
     Ozone Sci. & Engrg. 1(4)-.347-356.   See also Stephenson,  P., Benedek, A.,
     Malaiyandi, M. & Lancaster, E.A.,  "Part  I.  Ozonation and Biodegradation",
     Ozone Sci. & Engrg. 1(3):263-280.

Benedek, A. & Richard, Y., 1980, "Advances in Biological Treatment", presented
     at the  IWSA Congress, Paris, Sept.

Bernardin, F.E., Jr., 1980, "Granular Activated Carbon Operation:  Problems
     and Solutions", in Proc. 8th. Ann..  WWEMA  Indl.  Poll. Conf.,  Houston,  TX,
     June 4-6,  p. 93-101.

Bhargava, P., 1979, "Ozone Disinfection of Pure Oxygen Activated Sludge
     Using Unique Partial Recycle System", presented at 52nd Ann. Conf.,
     Water Poll. Control Fed., Houston, TX, Oct. 9.

                                     275

-------
Blanck, C.A.,  undated,  ca.  1976,  "Taste  and Odor Control Utilizing Granular
     Activated Carbon in the Plains  Region of The American Water Works
     System",  Am.  Water Works Service  Co., Inc., Western Region.

Blue, P.T., 1979,  Infilco Degremont  Inc., Richmond,  VA., Private Communication.

Brewer, W.S. & Carmichael,  W.W.,  1979, "Microbiological Characterization of
     Granular Activated Carbon Filter  Systems", J. Am. Water Works Assoc.
     71:738-740.

Cairo, P.R., McElhaney, J.  & Suffet, I.H., 1979, "Pilot Plant Testing of
     Activated Carbon Adsorption  Systems", J. Am. Water Works Assoc. 71:660-
     672.

Calgon Corp.,  undated,  Basic Concepts  of_ Adsorption  on Activated Carbon,
     Filtrasorb Dept.,  Water Mgmt. Div., Calgon Corp., Pittsburgh, PA.

Carnes, B.A. & Burstein, C., 1980, "Alternative GAC  Adsorber Design", presented
     at 8th Ann.  Indl.  Poll. Conf.,  Houston, TX, June.

Carswell,  J.K., Or., 1977,  U.S. Environmental Protection Agency, Cincinnati,
     Ohio.  Private Communication.

Characklis, W.G.,  1973, "Attached Yicrobial Growths.  I. Attachment and
     Growth",  Water Research 2:1103.

Chidal, J., 1976,  Cie.  GSnirale des  Eauxs Paris, France.  Completed question-
     naire for Public Technology Inc.  survey describing Choisy-le-Roi plant.

Clark, J.W., Viessman, W, & Hammer„  M.J., 1971, Water Supply &  Pollution
     Control,  2nd Ed., The Intl.  Textbook Co.,  Chapter 11,  Biological
     Treatment Processes".

Clark, J.C., Jr.,  1979, "Economic Aspects of Controlling Synthetic Organics",
     presented at Seminars on Control  of Organic Chemical Contaminants  in
     Drinking Water, Jan. 16/17,  Feb.  13/14 & March  13/14.  U.S. EPA, Office
     of Drinking Water, Washington,  D.C.

Clark, R.M. & Stevie, R.G., 1978, "Meeting  the  >1nking Water Standards:
     The Price of Regulation".  Presented at the Nat'l. Conf. on Drinking
     Water Policy Problems.  Resources For The  Future, Washington, D.C.,
     March 6-8, 1978,

Clark, R.M. & Dorsey, P., 1980, "Influence of Operating Variables on the
     Cost of Granular Activated Carbon Adsorption Treatment", U.S. EPA,
     Municipal Environmental Research Lab., Cincinnati, Ohio.

Clark, R.M., Culp, R.L. & Gumerman,  R.C., 1980,  "Cost of Granular Activated
     Carbon Adsorption in the USA",  U.S. EPA, Municipal Environmental Research
     Lab., Concinnati, Ohio.
                                     276

-------
Cleasby, J.L., Arbolea, J., Burns,  C.E.,  Prendiville,  P.W.  &  Savage,  E.S.,
     1977, "Backwashing of Granular Filters",  J.  Am. Water  Works  Assoc.
     69(2):115-126.

Coco, J.H., Klein, E., Howland,  D., Mayes,  J.H.,  Myers,  W.A.,  Pratz,  E.,
     Romero, C.J. & Yocum, F.H., 1979,  "Development of Treatment  and  Control
     Technology for Refractory Petrochemical Wastes",  EPA Report  No.  EPA-
     600/2-79-080.  U.S. EPA,  Robt. S.  Kerr Env.  Res.  Lab., Ada,  OK.

Coin, L., Hannoun, C.  & Cornelia, C., 1964,  "Inactivation of Poliomyelitis
     Virus by Ozone in the Presence of Water", la Presse Medicale 72(37):2153-
     2156.

Coin, L., Gomella, C., Hannoun,  C.  & Trimoreau, J.-C., 1967,  "Ozone  Inactiva-
     tion of Poliomyelitis Virus in Water", la Presse  Mfidlcale 75(38):1883-
     1884.

Culp, R.L., 1980, "GAC Water Treatment Systems",  Public  Works, Feb.,  pp.
     83-87.

Directo, L.S., Chen, C.-L. & Kugelmar,  I.J.,  1977, "Pilot Plant Study of
     Physical-Chemical Treatment",  J. Water Poll. Control Fed. 49(10):2081-
     2098.

Dostal, K.A., Pierson, R.C., Hager, D.3.  & Robeck, G.G., 1965, "Carbon Bed
     Design Criteria at Nitro, West Virginia", J. Am.  Water Works Assoc.
     57:663.

Eberhardt, M., 1975, "Experience with the use  of Biologically Effective
     Activated Carbon", in Translation of Reports on Special  Problens of_
     Water Technology, Vol. 9 - Adsorption, H. Sontheimer,  editor.EPA
     Report No. E'PAi50079^76-030l Dec.  1976,  pp.  331-347.

Eberhardt, M., Madsen, S. & Sontheiner, H., 1974, "Untersuchungen zur
     Verwendung Biologisch Arbeitender Aktivkohlefilter  bei der Trinkwasser-
     aufbereitung", Heft 7, Vertiffentlichungen des Bereichs und des  Lehrstuhls
     ftir Wasserchemie Leitung: Prof. Dr.  H. Sontheimer,  Univ.  Karlsruhe,
     Germany.  See also Wasser/Abwasser 116(6):245-247 (1975).

Edwards, R.W., 1979, 'Control of Synthetic Organic Chemicals by Granular
     Activated Carbon - Theory, Application and Regeneration", presented  at
     EPA seminar on Control of Organic Chemical Contaminants  in Drinking
     Water, Philadelphia, Pennsylvania, January, 1979.  Public Technology,
     Inc., Washington, D.C.

Engels, C., 1978, DUsseldorf Waterworks,  Federal  Republic of  Germany,
     Private Communication.

Engler-Bunte Institut der UniversitSt Karlsruhe, 1977, "Determination of  the
     Organic Substances Dissolved  in Water by Adsorption on Activated
     Carbon".  From Practical Control Methods 1n the Treatment of Drinking
                                     277

-------
     Water,  published by EBI  and DVGW  Research  Bureau, Richard Willstatter
     Allee 5, 7500 Karlsruhe, Federal  Republic  of Germany.

Fiessinger,  F.M.  & Monteil, A.,  1980,  "Case  Studies  for  the Abandonment of
     Prechlorination", presented at Ann.  Conf.  of Am. Water Works Assoc.,
     Atlanta, 3A, June 18.

Flentje, M.  & Hager, D.G.,  1964, "Reevaluation  of Granular Carbon Filters
     for Taste and Cdor Control", J. Air.  Water  Works Assoc. 56(2):191-197.

Ford, D.B.,  1973, "The Use  of Granular Carbon Filtration for Taste  & Odor
     Control", Proc. Mater  Research Assoc. Conf., Reading, England, pp. 263-
     278.

Froelich, E.M., 1978, "Control of Synthetic  Organic  Chemicals by Granular
     Activated Carbon — Theory, Application and Reactivation Alternatives",
     presented at EPA seminar on Control  of  Organic  Chemical Contaminants  in
     Drinking Water, Cincinnati, Ohio, October, 1978.  Public Technology,
     Inc., Washington, D.C.

Fry, J.C. & Staples, D.G.,  1976, "Distribution  of Bdenovibrio Bacteriovirus
     in Sewage Works River  Water and Sediments", Appl. & Envmntl. Microbiology
     31:469-474.

Fuchs, F. & KUhn, W., 1975,  "The Use of Activated Carbon to Analyze Natural
     Waters With Regard to  Their Behavior in Waterworks  Filters", in Transla-
     tion of_ Reports on_ Special  Problems of_ Water Technology, Vol.  9 -
     Adsorption,  H. Sontheimer8  editor.   EPA Report  No.  EPA 600/9-76-030,
     Dec., 1976,  pp. 182-207.

Gauntlett, R.B. & Packham,  R.F., '9/3a,  "The Use of  Activated Carbon in
     Water Treatment", Proc.  Water Research  Assoc. Conf., Reading,  England,
     pp. 31-51.

Gauntlett, R.B. & Packharr.,  R.F., 1973b,  "The Removal of  Organic  Compounds  in
     the Production of Potable Water", Chem. &  Ind., Sept., 1973, pp. 812-
     817.

German Unit Standards for Water "egting, Wastewater  Testing and  Mud Testing,
     T960, 3rd Edition, Chemie Pub!.,  Weinheim, Federal  Republic  of Germany.

Gerval, R., 1978, "Ozonation  of Raw Water Before Treatment in Drinking Water
     Plants", presented at Ozone Technology  Symposium, Los Angeles, Calif.,
     May, 1978.  Intl. Ozone  Assoc., Vienna, VA.

Serval, R., 1979, "The Different Locations of Ozone  in Treatment  Lines.
     Study of the Cperating Mechanisms According  to  These Locations", presen-
     ted at 4th World Dzone Congress,  Houston,  TX5 Nov.  26-29.   Intl. Ozone
     Assoc., Vienna, VA.
                                     278

-------
Glaze, W.H., 1978, "A Comparison of Ozone and  Ozone/Ultraviolet  for  the
     Destruction of Refractory Organic  Compounds  in  Water".   Presented at
     Ozone Technology Symposiums, Los Angeles,  Calif.,  vay,  1978.   Intl.
     Ozone Assoc., Cleveland,  Ohio.

Gilberts E., 1977, "Chemische  Vorgflnge  bei  der Ozonanwendung", in  WASSER
     BERLIN '77 - Ozon und Wasser - Wasserchem'e,  p.  277-293.  AMK Berlin  or
     Intl. Ozone Assoc., Vienna, VA.

Gilbert, E., 1978, "Investigations on the Changes  of Biological  Cegradability
     of Single Substances Induced by Ozonolysis",  presented at Ozone Technol.
     Symp., Los Angeles, May 23-25.  Intl Ozone Assoc.,  Vienna,  VA.

Gomella, C. & Versanne, D., 1977, "Le Role de  TOzone dans  la Nitrification
     Bact§rienne de 1'Azote Amironiacal  -- Cas  de  I'Usine de la Chapelle
     Banlieue Sud de Rouen (Seine Maritime), France".   Presented at  Third
     Intl. Symp. on Czone Technol.„ Paris,  France, May,  1977.  Intl. Czore
     Assoc., Vienna, VA.

Gould, C.P. & Weber, W.J., Jr., 1976, "Oxidation  of  Phenols by Ozone", J.
     Water Poll. Control Fed.  48(1):47-60.

Grombach, P., 1975, "Reaktivierung von  Aktlvkohle  im Wasserwerk",  3az-Eaux-
     Eaux Us§es 55(9):549-553.

Guirguis, W.A., Cooper, T., Harris, J.  & Ungar, A.,  1976a,  "Inproved Perfcr-
     mance of Activated Carbon by Pre-Ozonation',  Presented at 49th  Annual
     Conf. of Water Pollution  Control Fed., Minneapolis, Minn.8  Oct., 1976.

Guirguis, W.A., Hanna, Y.A,, Prober, R., Meister,  T.  & Srivastava, P.K.,
    "l976b, "Reaction of Organic* Nonsorbable  by  Activated  Carbon  with
     Ozone", in Ozone/Chlorine Dioxide Oxidation  Products of Organic
     Materials, R.G. Rice & J.A. Cotruvo, editorsTIntl. Ozone  Assoc.,
     Vienna, VA (1978), pp. 291-301.

Guirguis, W.A., Jain, J.S., Hanna, V.A. & Srivastava, P.K., 1976c, "Ozcne
     Application for Disinfection in the Westerly Advanced  Wastewater "reat-
     ment Facility", in Forum on Ozone Disinfection. E.G. Fochtman,  R.G.
     Rice & M.E. Brownings editors.  Intl. Ozone  Assoc., Vienna, VA, pp.
     363-381.

Guirguis, W.A., Melnyk, P.B. & Harris,  J.P.,  1976d,  "The Negative  Impact  of
     Industrial Waste on Physical-Chemical Treatment", presented at  31st
     Purdue Indl. Waste Conf.s Lafayette, Indiana,  May,  1976.

Guirguis, W.A., Cooper, T., Harris, J.  & Ungar, A.,  1978, "Improved  Perfor-
     mance of Activated Carbon by Pre-Ozonation",  J. Water Pollution Control
     Fed. 50(2):308-320.

Gurrerman, R.C., Gulp, R.L. & Hansen, S.P., '978,  "Estimating Costs For Water
     Treatment as a Function of Size and Treatment Efficiency",  EPA  Report
                                     279

-------
     No. EPA-600/2-78-182.   U.S.  EPA,  Municipal  Environmental  Research  Lab.,
     Cincinnati, Chio.

Gumerman, R.C., Gulp, R.L.  & Hansen,  S.P.,  1979,  "Estimating Water  Treatment
     Costs", EPA Report No. EPA-600/2-79-162a (Vol.  1  -  Summary)  and  -162b
     (Vol. 2 - Cost Curves  Applicable to 1  to 200 mgd  Treatment Plants).
     U.S. EPA, Municipal Environmental Research  Lab.,  Cincinnati, Ohio.

Haaar, O.G. & Flentje,  M.,  1965,  "Removal  of Organic Contaminants by  Granular
     Carbon Filtration", J. Am.  Water Works Assoc. 57(11):144C (1965).

Hagar, D.G. & Fulker, R.D., 1968, "Adsorption and Filtration with Granular
     Activated Carbon", J.  Soc.  Water Trtmt. & Examn., 17.

Handbook of Chemistry & Physics,  1975-1976, 56th Edition.   CRC Press,
     Cleveland, Ohio, pp. D-141-143.

Hanna, Y., Slough, J.,  Jr.  & Guirguls, W.A., 1977, "Ozone  as a Pretreatment
     Step for Physical  Chemical  Treatment - Part II",  Presented at  Symposium
     on Advanced Ozone Technology, Toronto, Ontario, Canada, Nov.,  1977.
     Intl. Ozone Assoc., Vienna,  VA.

Hansen, S.P., Gumerman, R.C. & Culp,  R.L.,  1979,  "Estimating Water  Treatment
     Costs. Vol. 3.  Cost Curves Applicable to 2,500 gpd to 1  mgd Treatment
     Plants".  EPA Report EPA-60C/2-79-162c.  J.S. EPA,  Municipal Environmen-
     tal Research Lab., Cincinnati, Chio.

Hasler, J.W., 1974, Purification With Activated  Carbon,  Chemical  Publishing
     Co., Inc., New York, N.Y.

Heilker, E., 1978, Rheinisch-Westfa'llschen Wasserwerksgesellschaft  mbH,
     MUlheim, Federal Republic of Germany.   Private Communication.

Heilker, E., 1979a, "Design of Water Treatment Adsorption  Systems", presented
    • at NATO/CCMS Conf. on Adsorption Techniques in Drinking Water  Treatment,
     Reston, VA., Apr.  30-May 2.   U.S. EPA, Office of  Drinking Water, Washing-
     ton, D.C.

Heilker, E., 1979b, "The Mdlheim Process for Treating  Ruhr River Water",  J.
     Am. Water Works Assoc. 71(ll):623-627.

Holluta, J., 1959, "Zwei Jahre Seruchsnessungen  am Miederrheln",  Kommunal-
     wirtschaft 6:223-229.

Holluta, J., 1960, "Geruchs- und 3eschmacksbeeintr3chtigung des Irinkwassers.
     Ursache und Bekanrpfung".  Wasser-Abwasser 1C!:1018-1C23 & 1070-1078.

Hopf, Vi., 1960, "Versuche mit Aktivkohlen zur Aufbereitung des DUsseldorfer
     Trinkwassers"9 Wasser-Abwasser 101(14):33C-336.
                                     280

-------
Hopf, W., 1970, "Water Treatment With Ozone and Activated  Carbon  (The
     DUsseldorf Process), Part II.   Notes on the Use  of Czone", Wasser-
     Abwasser *n(3):156-164.

Huisman, I.L., 1978, "Developments  in Village-Scale Slow Sand  Filtration",
     Progress in Water Techno!.  11(1/2):159-165.

Hutchinson, M. & Ridgway, J.W.,  1977, "Microbiological  Aspects of  Drinking
     Water Supplies", in Aquatic Microbiology,  Soc. for Appl.  Bacteriology,
     Symposium Series No. o,  F.A. Skinner & J.M.  Shewan, editors,  Acadeiric
     Press, London & New York, pp.  179-218.

Cekel, M., 1977, "Biological  Treatment of Surface Waters in  Activated
     Carbon Filters", presented at Meeting of KIWA, Water  Research Center &
     Engler-Bunte Inst., Jniv. Karlsruhe, Federal  Republic of  Germany, Cct.
     27-28, 1977.

Jekel, M., 1978, Engler-Bunte Institut, Univ. Karlsruhe, Federal  Republic of
     Germany.  Private Communications.

Jekel, M., 1979B "Experience  With Biological Activated  Carbon", presented at
     Intl. Conf. on Oxidation Techniques in Drinking  Water Treatment,
     Karlsruhe, Federal Republic of Germany, Sept. 11-13.  Engler-Bunte
     Inst. der Univ. Karlsruhe.

Jeris, J.S., Owens, R.W., Hickey, R. & Flood, F., 1977, "Biological  Fljidized-
     Bed Treatment for BOD and Nitrogen Removal", J.  Water Poll.  Control
     Fed. 49(5):816-831.

OOntgen, H., 1975, "Manufacture and Properties  of Activated  Carbon",  1n
     Trans'atior p_f_ Reports on Special Problem^ of_ Water Technology,  Vol.
     9 - Adsorption, op. cit., pp.  16-28,

Kittredge, D., 1980, "The Economics of Carbon Regeneration;  State of the
     Art", 0. New England Water Works Assoc. 94(1):1-23.

Klotz, M., Werner, P. & Schweisfurth, R., 1975, "Investigations Concerning
     the Microbiology of Activated Carbon Filters",  in  Translation of Reports
     on Special Problems of Water Technology. Vol. 9  -  AdsorptiohT op_.  cit.,
     pp. 312-330.

Kappe, D.S., 1978V "Pretreatnent of the Combined Industrial-Domestic Waste-
     waters of Hagerstown, Maryland.  Vol. I".   U.S.  EPA Report No.  EPA-
     600/2-78-043a, Robert S. Kerr Environnental Research  Lab., U.S.  EPA,
     Ada, OK  74820=

Klotz, M., Werner, P. & Schweisfurth, R., 1976, "Microbiolologlcal Studies
     of Activated Carbon Filtering for the Treatment  of Drinking  Water:
     Research and Development in Water Works Practical  Applications', Sci.
     Reports on Study and Planning of the City Works, Wiesbaden,  A.G.  3:75-82.
                                     281

-------
Knoppert, P.L. & Rook,  J.J.,  1973,  "Treatment of River Rhine  Water  With
     Activated Carbon", Proc.  Water Research Assoc.  Conf.,  Reading,  England,
     pp.  109-125.

Knorr, D.B., 1979, "Direct Recharge For El  Paso, Texas",  in Proc. Water
     Reuse Symp., Washington,  D.C., March 25-30.  Am.  Water Works Assoc.
     Research Foundation, Denver CO., pp. 212-238.

KBlle, W. & Sontheiner, H., 1973, "Experiences With  Activated Carbon in  West
     Germany", Proc. Water Research Assoc.  Conf., Reading,  England,  pp.  347-
     367.

Kfille, W., 1974, "Suitability Testing of Water Works Activated Carbons on
     the Basis of Their Adsorption  Characteristics for Crganic Chlorine
     Compounds", Report of the Nuclear Research Center, Karlsruhe  (KFK 1969
     UF)  of April, 1974, dealing with Water & Wastewater Chemistry  Investiga-
     tions, 1972/73.

Kfllle, W., Sontheimer,  H. & Steiglitz, L.,  1975, "Eignungsprdfung von
     Wasserwerks-Aktivkohlenanlage  Ihrer Adsorptionseigenschaften ftlr
     Organische Chlorverbindungen", Vom Wasser 44:203-217.

Koppe, P. & Giebler, G., 1966, "Untersuchungen Uber  die Ozonzehrung von
     wassern", Wasser-Abwasser 107(8):196-200.

Kornegay, B.H., 1978, "Control of Synthetic Crganic  Chemicals by Activated
     Carbon — Theory,  Application  and Regeneration  Alternatives",  presented
     at EPA seminar on  Control of Organic Chemical Contaminants in  Drinking
     Water, Los Angeles, Calif., Nov. 14-15, 1978.  U.S.  EPA, Office of
     Drinking Water, Washington, D.C.

Kdhn, W., 1974, "Untersuchungen zur Bestimmung von Organischen Chlorverbin-
     dungen auf Aktivkohle",  Dissertation,  Fak. f. Chemie-Ing. Wesen, Univ.
     of Karlsruhe, Karlsruhe,  Federal Republic of Germany.

Kdhn, W.  & Fuchs, F., 1975, "Lntersuchungen zur Bedeutung der Organischen
     Chlorverbindungen  und Ihrer Adsorbierbarkeit",  Vom Wasser 45:217-232.

KUhn, W.  & Sontheimer,  H., 1973a, "Einige Untersuchungen zur Bestlnmung  von
     Crganischen Chlorverbindungen  auf Aktivkohle1,  Vom Wasser 41:65-79.

Ktlhn, W.  & Sontheimer,  H., 1973b, "Einfluss Chemischer Umsetzungen  auf die
     Lage der Adsorptlonsgleichgewichte an Aktivkohlen",  Vom Wasser 40:115-123.

Klihn, W. & Sontheimer,  H., 1974, "Zur Analytischen Erfassung Organischer
     Chlorverbindungen  mit der Temperaturprogrammierten Pyrohydrolyse",  Vom
     Wasser 43:327-341.

KUhn, W., Sontheimer, H. & Kurz, R., 1978,  "Use of Ozone and Chlorine in
     Water Works in the Federal Republic of Germany",  1n Ozone/Chlorine
                                     282

-------
     Dioxide Oxidation Products of Organic Materials,  R.3.  Rice & J.A.
     Cotruvo, editors.  Intl. Ozone Assoc., Vienna,  VA,  pp.  426-441.

Kussmaul, H., 1978, "Efficiency of Ground Passage in Crfnking Water Treatment",
     presented at Intl. Conf. on Oxidation Processes in  Drinking Water
     Treatment, Karlsruhe, Federal Republic of Germany,  Sept. 10-13,  1978.
     Engler-Bunte Inst. der Univ. Karlsruhe.

Latoszek, A. & Benedek, A., 1979, "Soire Aspects of the Micrcbiology of
     Activated Carbon Columns "raating Domestic Wastewater", Erviron.  Sci.  &
     Techno!. 13(10):1285-1287.

Lawrence, C.H.S 1968, "California Plant Uses Ciatcmite and  Carbon Filters",
     Water & Wastes Engrg.9 Jan., 1968.

Larocque, R., CANOZONE, Montreal, Canadas 1977.  Private Comnunication.

Lepage, W.L., 1979, "A Plant Operator's View of Ozoration",  in Proc.  Seminar
     On_ 'The Design and Operation of_ Drinking Water  Facilities Using Ozone
     or CfiTonne Dioxide, R.3. Rice, editor.  New England  Water Works
     ATsoc.s Cedham, MA.S pp. 225-24C.

Le Pauloue", J., 1978, TRAILIGAZ, 3arges-l§s-Gonesse, France.  Private
     Communication.

Lestochi, L.J., Byrne, C.I, & Horg, S.N., 1979, "Generation of Czone From
     Oxygen for Drinking Water Treatment", in Pr_oc_.  Seminar On 'The Design
     and Operation p_f Or inkiing Water Faci 11 ties Us iing  Ozone" or_ Chlorine
     D|ox|QeY, R.G. Rice, editor, New England Water  Works Assoc.V"Dedham,
     MA, pp. 121-130.

Lurgi, 1976, Hydraffin, Aquaffin. Powdered and Granular Activated Carbon For
     Water Pur ifi cat yon, Manufacturing", AppTTcation. Testing, Publi en.  fib.
     T-T188/3.76, Lurgf Apparate-Technik"GrnbH',~CTiemotechn1k, Frankfurt/vain,
     Federal Republic of Germany.

Maier, D., 1971, "Untersjchungen zur Chemischen Regeneration von Erschfip'ten
     Wasserreinigungskohlen", Dissertation, Fak. f.  Chemie-Ing. wesen,  Univ.
     Karlsruhe, Federal Republic of Germany.

Maier, D.p 1979, "Microflocculation By Ozone", in Oxidation Techniques  in
     Drinking Water Treatment, W. KUhn & H, Sontheimer,  editors, U.S. EPA"
     Report No7~ETA"-570/9-79-020c pp. 394-417.  U.S. EPA, Office of Drinking
     Water, Washington, D.C.

Maqsood, A. & Benedek, A.,  1977, "The Effect of Low 'emperature on Organic
     Removal and Denitrification in Activated Carbon Colunns"-, J. Water
     Poll. Control Fed. 49:2107.

McCreary, J.J. & Snoeyink, V.L., 1977, "Granular Activated Carbon in Water
     Treatment", J. Am. Water Works Assoc. 69(8):437-444.
                                     283

-------
McElhaney, J. & McKeon, W.R., 1978, "Enumeration and Identification of
     Bacteria In Granular Activated Carbon Columns", presented at Am.  Water
     Works Assoc. Technical Mtg., Louisville, KY, Dec.   Am. Water Works
     Assoc., Denver, CO.

McElhaney, J., Coyle, J.T., Santo, J. & Burllngame, G., 1979, "The Effect of
     Bacterial Activity on TOC Removal Using Granular Activated Carbon",
     presented at Am. Water Works Assoc. Technology Conf., Philadelphia, PA,
     Dec.  Am. Water Works Assoc., Denver, CO.

Melbourne, J.D. & Miller, D.G., 1973, "The Treatment of River Trent Water
     Using Granular Activated Carbon Beds", Proc. Water Research Assoc.
     Conf.. Reading, England, pp. 73-108.

Metcalf & Eddy, Inc., 1972, Wastewater Engineering; Collection, Treatment,
     Disposal, McGraw Hill Book Co., New York, N.Y.

MUler, G.W., Rice, R.G., Robson, C.M., Scullln, R.L., KOhn, W. & Wolf, H.,
     1978, "An Assessment of Ozone & Chlorine Dioxide Technologies for the
     Treatment of Municipal Water Supplies".  U.S. EPA Report No. 600/2-78-
     147.  U.S. Environmental Protection Agency, Municipal Environmental
     Research Laboratory, Cincinnati, Ohio  45268.

Miller, G.W., 1979, "Costs of Treating Drinking Water With Ozone", in Proc.
     Seminar On 'The Design and Operation of Drinking Water Facilities Using
     Ozone or~Chlorine Dioxide', R.G. Rice, editor.  New England Water Works
     Assoc., Dedham, MA, pp. 189-204.

•Milliner, R., Bowles, D.A.. & Brett, R.W., 1972, "Biological Pretreatnent at
     Tewkesbury", Water Trtmt. & Examn. 21:318-326.

Moat,  A.G., 1979, Xicrobial Physiology. John Wiley & Sons, New York, N.Y.,
     Chapters 3 & T.

Moser, R.H., 1978, Letter, dated July 25, 1978, to Fred E. Schlaegal,
     included as Attachment D of Comments and Recommendations Concerning
     EPA's Proposed National Interim Primary Drinking Water Regulations,
     Coalition For Safe Drinking Water, Indianapolis, Indiana, August, 1978.

Nolte, H., 1978, Rhein1sch-WestfSil1schen Wasserwerksgesellschaft mbH,
     Mulheim, Federal Republic of Germany.  Private Communication.

Osborne, D., 1979, "Experience at Wilne Treatment Works", presented at
     NATO/CCMS Conf, on Adsorption Techniques in Drinking Water Treatment,
     30 April - 2 May 1979, Reston, VA.  U.S. EPA, Office of Drinking Water,
     Washington, D.C.

Park,  Y.K. & Sontheiner, H., 1973, "The Transformation and Concentration of
     Pesticides From Dioxane Extracts Into n~Hexane"s Vom Wasser 41:81-91.
                                     284

-------
Park, Y.S.,  1974,  "The Investigation  of Polar Organic  Chlorine  Compounds  in
     Waters  and on Activated Carbons",  Dissertat1on9 Univ.  Karlsruhe,
     Federal  Republic of Germany.

Peel, R.  & Benedek, A.,  19770 "The Modelling  of Activated  Carbon Adsorbers
     in the  Presence of Biooxidation",  Water-1976.  AIChE  Symp. Series
     166:25.   Am.  Inst,  Chem. Engrs.s New York,  N.Y.

Peel, R.  & Benedek, A.,  1980, "Dual Rate Kinetic Model  for Activated Carbon
     Adsorption, Parts I and II",  accepted for publication in J. Am. Soc.
     Civil Engrs.   Am. Soc.  Civil  Engrs., New York, NY.

Piet, G., 1979, "Behavior of Micropollutants  in  River  Water during Ground
     Passage", presented at  Intl.  Conf.  on Oxidation Processes  in drinking
     Water Treatment, Karlsruhe,  Federal Republic of Germany, Sept. 10-13,
     1978.  Engler-Bunte Inst.  der Univ. Karlsruhe.

Piet, G.J. & Zoeteman, B.C.J.,  1980,  "Organic Water Quality Changes Curing
     Sand Bank And Cune Filtration of Surface Waters In  The Netherlands", J.
     Am.  Water Works Assoc.  72(7):400-404.

Poggenburg,  W., Engels,  C.,  Weissenhorn, F.J., Fuchs,  F. & Sontheimer,  H.,
     1974, "Investigations on the Application of Activated Carbon for  the
     Treatment of Filtrate Fron the Banks of  the Rhine1, Special Printing
     from the CVGW, Series,  Water, No.  7, varch 27-29,  1974, Dulsburg.
     Deutscher Vereln des Gas-  und Wasserfaches e.V.,  6236 Eschborn, Federal
     Republic of Germany.

Poggenburg,  W., 1975, 'Activated Carbon Filters in Water Treatment Plants.
     Processing Techniques - Engineering - Operation",  in  Translation  of
     Reports on_ Special  Problens or_ Water Technology,  Vol. 9 -  Adsorption,
     op.  cit.. pp. 74-97.

Poggenburg,  W., 1977, Wasserwerk DUsseldorf,  DUsseldorf, Federal Republic of
     Germany.  Private Communication.

Poggenburg,  W., 1978, Wasserwerk CUsseldorf,  DUsseldorf, Federal Republic of
     Germany.   Private Communication.

Prengle,  H.W., Jr. & Mauk. C.E., 1978,  ""Ozone/UV Oxidation of  Pesticides in
     Aqueous Solution", in Ozone/Chlorine Dioxide Oxidation Products of
     Organic Materials, R.G. Rice & J.A. Cotruvo, editors.  Intl. Ozone
     Assoc.7 Vienna, VA, p.  302-320.

Presecan, N.L., 1978, "Effects of Biological  Growth  on 3ranular Activated
     Carbon Adsorption Systems", presented at WWEMA  Annual Indl. Poll.
     Control Conf., St. Louis,  MO, 13 April.   Water  &  Wastewater Equipment
     Mfgrs.  Assoc., McLean,  VA.

Rafelson, M.E., Brink'ey, S.8.  & Hayashl, J.A., 1971,  Basic Biochemistry.
     3rd Ed., McMillen Co.,  New York, NY.


                                     285

-------
Rice, R.G. & Miller, G.W.,  1977,  "Reaction Products  of Organic  Materials
     With Ozone and Chlorine Dioxide in Water",  presented  at  Symp.  on  Advanced
     Ozone Technology, Toronto,  Ontario, Canada.   Intl.  Ozone Assoc.,  Vienna,
     VA.

Rice, R.G., Gomella, C. & Miller, G.W., 1978,  "Rouen,  France  Water  Treatment
     Plant:  Good Organics and Airmonia Removal with  no Need to  Regenerate
     Carbon Beds", Civil Engineering, May, 1978,  pp.  76-82.

Rice, R.G., Miller, G.W., Robson, C.M. & KUhn, W.,  1978,  "A Review  of  the
     Status of Pre-Ozonation of Granular Activated  Carbon  For Removal  of
     Dissolved Organics and Ammonia in Water and Wastewater Treatment",  in
     Carbon Adsorption, P.N. Cheremisinoff & F.  Ellerbusch, editors.   Ann
     Arbor Science Publishers, Inc., Ann Arbor,  Michigan,  p.  485-538.

Rice, R.G., Miller, G.W., Robson, C.M. & Hill, A.G.,  1979a, "Ozone  Utiliza-
     tion in Europe", presented at AIChE 86th  Natl.  Meeting,  Houston,  TX,
     April.  In Water — 1979. G.F. Bennett, editor,  p.  117-134 (1980).   Am.
     Inst. Chem. Engrs., New York, N.Y.

Rice, R.G., Robson, C.M. & Miller, G.W., 1979b,  "Ozone For Drinking Water
     Treatment — Current State Of The Art", in  Proc.  Seminar On 'The  Design
     and Operation of Drinking Water Facilities  Using flzore or  Chlorine
     ERoxide', R.G. Rice, editor^  Rew England  /later Works Assoc.,  Dedham,
     MA, p. 1-48.

Richard, Y., 1973, "Experience With Activated Carbon in France", in Proc.
     Water Research Assoc. Conf., Reading, England,  pp.  313-346.

Richard, Y. & Fiessinger, F., 1977, "Emploi ComplSrentaire des  Traitments
     Ozone et Charbon Actif", presented at 3rd  Intl.  Symp. on Ozone Techno!.,
     Paris, France, May, 1977.  Intl. Ozone Assoc.,  Cleveland,  Ohio.

Richard, Y., 1978, Soc. DegrSmont, Rueil Malmaison,  France.   Private Corr-
     munication.

Richard, Y., 1979, Soc. DegrSmont, Rueil Malmaison,  France.   Private Com-
     munication.

Rodman, C.A., Shunney, E.L. & Perrotti, A.E.,  1978,  "Biological Regeneration
     of Activated Carbon", in Carbon Adsorption  Handbook,  P.C.  Cheremisinoff
     & F.E. Ellerbusch, editor?]!   A~n~n Arbor Science  Publishers, Inc.,  Ann
     Arbor, MI, p. 449-483.

Romero, J.C., 1970, "The Movement of Bacteria Through Porous  Media",  Ground
     Water 8:37-48.

Rook, J.J., 1972, "Production of Potable Water From A Highly  Polluted  River",
     Water Treatment & Examn. 21:259.

Rook, J.J., 1974, "Formation of Haloforms During Chlorination of Natural
     Waters", Water Treatment & Examn. 23(2):234-243.


                                     286

-------
Rook, J.J., 1976, "Developments in Europe",  J.  Am.  Water Works  Assoc.
     68(6):279-282.

Rook, J.J., 1978, Wasserwerk Kra1ingens  Rotterdam,  The Netherlands.   Private
     Conrnunication.

Sala-Trepot, J.M., Murray, K.  & Williams,  P.A., 1972,  "The Nletabolic Diver-
     gence in the Meta Cleavage of Catechols by Pseudo^onas Putida", Eur.  J.
     Biochem.  28:347.

Schalekamp, M., 1975, "Use of Activated  Carbon  in the  Treatment of Lake
     Water", in Translation of Re^jrts on  Special Problems of Water Technology,
     Vol. 9 - Adsorption.   Q£. cjjb.,  pp. 128-159.

Schalekamp, M., 1979, "Swiss Experiences With Granular Activated Carbon
     (GAC) in the Treatment of Drinking  Water", presented at NATO/CCMS Conf.
     on Adsorption Techniques in Drinking  Water Treatment, Reston, VA, Apr.
     30-May 2.  U.S. EPA,  Office of Drinking Water, Washington, D.C.

Scheidtmann, W., 1975, "Investigations of the Optimization of Pretreatment
     When Using Ozone", in Translation of Reports on Special Problems of_
     Water Technology, Vol. 9 - Adsorption,   op. cTt., pp. 98-TTI.

Schmidt, K.H., 1963, "Die Abbauleistungen  der Bakterienflora bei der Langsam-
     sandfiltration und ihre Beeinflussung durch die RohwasserqualltSt und
     andere UmwelteinflQsse.  Biologische  Studien zur kunstlichen Grundwasser-
     anreicherung", Verflffentl. d. Hydrol. Forschungabt. der Cortmunder
     Stadtwerke AG, No. 5.

Schmii.it, K.H.j, 1978, "Experience With Removal of Micropollutants in Slow
     Sand Filters'% presented at intl. Conf. on Oxidation Techniques in
     Drinking Water Treatment, Karlsruhe,)  Federal Republic of Germany.
     Engler-Bunte Inst. der Univ. Karlsruhe.

Schulhof, P., 1979, "French Experiences  in the Use of Activated Carbon For
     Water Treatment", presented at NATO/CON'S Conf. on Adsorption "echniques
     in Drinking Water Treatment, Reston,  VA, 30 AprH-2 May.  U.S. EPA,
     Office of Drinking Water, Washington, DC.

Schulhof, P., 1979, Cie. GeriSrale des Eauxs  Par1sB France.  Private Communica-
     tion.

Schulhof, P., 198C, "Water Supply in the Paris Suburbs:  Changing Treatment
     for Changing Demands", J. Am. Water Works. Assoc. 72(8):428-434.

Schweisfurth, R., 1978, Jniv. of Saarlands,  Homburg, Federal Republic of
     Germany.  Private Commur.ication.

Scott, C.D., Hancher, C.W., Holladay8 D.W. & Dinsmore, 3.B., 197E, "A Tapered
     Fluidized Bed Bioreactor for Treatment of Aqueous Effluents Froti Coal
     Conversion Processes", presented at Symp.  on Environmental Aspects of
     Fuel Conversion-II, Hollywood, FL,  15-18 Cec.

                                     287

-------
Sontheimer, H. & Maier, D. ,  1973,  "Untersuchungen zur Verbesserung  der
     Trinkwasseraufbereltungstechnologie am Niederrhein  (1.  Berlcht)",
     Wasser-Abwasser 113:187-193.

Sontheiner, H., 1975, "Considerations on the Optimization  of Activated
     Carbon Use 1n Waterworks", in Translation of Reports  on Special
     Problems of Water Technology, Vol.  9 - Absorption,  op.  ci t . ,  pp. 208-
Sontheimer, H., 1977a, Engler-Bunte Institut,  tniy.  Karlsruhe,  Karlsruhe,
     Federal Republic of Germany.   Private Communication.

Sontheimer, H., 1977b, "Biological  Treatment of Surface Waters  in  Activated
     Carbon Filters", presented at Seminar on Current Status  of Wastewater
     Treatment & Disinfection With Ozone, Cleveland, Ohio,  Sept.  15,  1977.
     Intl. Ozone Assoc., Vienna, VA.

Sontheimer, H., Heilker, E., Jekel, M., Nolte, H.  &  Vollrrer,  F.-H.,  1978,
     "The MUlheim Process", J. Am.  Water Works Assoc. 70(7):393-396.

Sontheiner, H., 1978, Engler-Bunte Institut, Univ. Karlsruhe, Karlsruhe,
     Federal Republic of Germany.   Private Communication.

Sontheimer, H. , 1979, "Development, Problems,  Aims and Significance  of the
     Oxidation Process in the Treatment of Drinking  Water", in  Oxidation
     Techniques in Drinking Water Treatment, W. Ktlhn & H.  Sontheimer,  editors,
     U.S. EPA Report No. EPA-57C/9-79-020.  U.S. EPA, Office  of Drinking
     Water, Washington, DC.

Sontheimer, H. , 1980, "Experience With Rlverbank Filtration Along  the  Rhine
     River", J. Am. Water Works A^soc. 72(7):386-390.

Standard Methods for Examination of Water & Wastewater, 1971, 13th Edition,
     S . 651, Am. PubT. Health Assoc . , New York, N.Y.

Stephenson, P., Benedek, A., Malaiyandi, M. & Lancaster, E.A.,  1979,  "The
     Effect of Ozone on the Biological Degradation and Activated Carbon
     Adsorption of Natural and Synthetic Organics in Water.  Part I.   Ozona-
     tion and Biodegradation", Ozone Sci. & Engrg. 1 (3):263-280.

Strudgeon, G.E., Lewis, B.J., Albury, W.W. & Clinger, R.C., 1980,  "Safety
     Considerations in Handling Activated Carbon", J. Water Poll.  Control
     Fed. 52(10) .-2516-2522.

Symons, J.H., editor, 1978, Interim Treatment Guide  For Controlling  Organic
     Contaminants in Drink ing'llater Using Granular Activated  Carbon,  U.S.
     EPA, Water SupplylTesearch Divlsionr Cincinnati, fthio, Jan.  1978.

Temple, Barker & Sloane, Inc., 1977, "Economic Impact Analysis  o" a  Trihalo-
     methane Regulation for Drinking Water", U.S. EPA, Office of Drinking
     Water, Washington, D.C., August, 1977.
                                     288

-------
Temple, Barker & Sloane, Inc., 1978,  "Revised Economic Impact  Analysis  of
     Proposed Regulations on Organic  Contaminants  in Drinking  Water", U.S.
     EPA, Office of Drinking Water, Washington,  3.C.,  July,  1978.

Tsezos, M. & Benedek, A., 1980, "Removal  of Substances by Biologically
     Activated Carbon in a Fluidized-bed  Reactor",  J.  Water  Poll.  Control
     Fed. 52(3):578-586.

U.S. Environmental Protection Agency, 1973, "Process Design  Manual  For
     Carbon Adsorption", Report No. EPA 625/l-71-002a.  U.S. EPA,  Washington,
     DC.

U.S. Environmental Protection Agency, 1978, "Proposed  Regulations  for
     Control of Crganic Chemical Contaminants in Drinking Water',  Federal
     Register, Feb. 9, 1978, 5756.

Valencia, J.A. & Cleasby, J.L., 1979, "Velocity  Gradients 1n Granular Filter
     Backwashing", J. Am. Water Works Assoc. 71(12):732-738.

Van Der Kooij, D., 1975, "Some Investigations Into  the Presence and Behavior
     of Bacteria in Activated Carbon  Filters", in  "ranslation  of Reports  on
     Special Problems of_ Water Technology, Vol.  9_ - Adsorption,  op.  cit.,
     pp. 348-35T

Van Der Kooij, D., 1978, Keuringsinstituut voor  Waterleidingartikelen  KIWA
     NV, Rijswijk (ZH), The Netherlands,   Private  Communication.

Van Leeuwen, C., 1979, "The Design and Application  of Packed Column Adsorber
     for Ozonation in Water Reclamation", presented at 4th World Ozone
     Congress, Houston, ~X9 Nov. 26-29.  Intl. Ozone Assoc., Vienna,  VA.

Van Lier, W.C., Graveland, A., Rooks  J.J. & Schultink, L.J., 1975,  "Experien-
     ces With Pilot Plant Activated Carbon Filters  in Dutch  Waterworks",  in
     Translation o£ Reports oil Special Problems  of Water Technology,  Vol.
     9 - Adsorption.   £p_. cit., pp. 160-181.

Versanne, D., 1978, Cie. Gen. des Eaux, Rouen, France.  Private Communication.

Wachs, A., Narkis, N. & Schneider, M., 1977, "Organic Matter Removal  From
     Effluents by Lime Treatments Ozonation, and Biologically Extended
     Activated Carbon Treatment", presented at Third Intl. Symp. on Czone
     Technology, Paris, France, May,  1977.  Intl.  Ozone Assoc., Vienna, VA.

Water Pollution Control Federation, 1977, Wastewater Treatment Plant
     Design, A Manual Q-F Practice, WPCF,  Walhlngton, TjTCT"

Water Research Center, 1977, Water Purification In The EEC,  A State Of
     The Art Review, Pergamon Press,  Elmsford, New York.

Weber, W.J., Jr., 1972, Physico-Chemico Processes for Water Quality
     Control, Wiley  Interscience, New York, N.Y."


                                     289

-------
Weber, W.J., Jr., 1974, "Adsorption Processes",  Pure &  Appl.  Chem.  37(3).
     Intl. Union of Pure & Appl.  Chem.,  Butterworth & Co.  Ltd.,  London,
     England.

Weber, W.J., Jr. & Ying, W.C., 1977, "Integrated Biological  and  Physico-
     chemical Treatment for Relamation of Wastewater",  Proc.  Intl.  Cpnf. on
     Advanced Treatment & Reclamation of Wastewater, IAWPR,  Johannesburg,
     So. Africa, June, 1977.

Weber, W.J., Pirbazari, M. & Melson, G.L., 1978, "Biological  Growth on
     Activated Carbon:  An Investigation by Scanning Electron Microscopy",
     Environmental Sci. & Techno!. 12(7):817-819.

Weber, W.J., Jr., 1978b, Discussion of "Improved Performance  of  Activated
     Carbon by Preozonation", by W. Guirguis,  T. Cooper,  J.  Harris  & A.
     Ungar, J. Water Poll. Control Fed.  50(12):2781-2785.

Weber, W.J., Jr. & Van Vliet, B., 1978,  "Fundamental Concepts for Appli-
     cation of Activated Carbon in Water and Wastewater Treatment", presented
     at Symp. on Activated Carbon Adsorption of Organics  From the Aqueous
     Phase, Miami, FL, Sept. 1978.  Am.  Chem.  Soc., Washington,  DC.

Weissenhorn, F.J., 1977, "The Behavior of Ozone in the  System and  its
     Transformation", Wasser Berlin-1977, p. 140-147.  AMK-Berlin,  Federal
     Republic of Germany; Intl. Ozone Assoc.-  Cleveland,  Ohio.

Werner, P., Klotz, M. & Schweisfurth, R., 1979,  "Microbiological Studies  of
     Activated Carbon Filtration", presented at Intl. Conf.  on Oxidation
     Techniques in Drinking Water Treatment, Karlsruhe.. Federal  Republic  of
     Germany, Sept. 11-13, 1978.   Engler-Bunte Inst. der Univ. Karlsruhe.

Werner, P., 1979, Univ. of Karlsruhe, Federal  Republic  of Germany.   Private
     Communication.

White, G.C., 1972,  Handbook of Chiorination,  Van Nostrand Reinhold Co.,  New
     York, N.Y.

Wolf, H.W., Camp, B.C. & Hawkins, S.J., 1979,  "Pyrogenic Activity  of Carbon
     Filtered Waters", U.S. EPA Report No. EPA-600/1-79-009,  U.S.  EPA,
     Health Effects Research Lab., Cincinnati, Chio.

Wfllfel, P. & Sontheimer, H.f 1974, "Ein Neues Verfahren zur Bestimmung  von
     Organisch Gebundenem Kohlenstoff im Wasser Durch Photochemische Oxida-
     tion", Vom Wasser 43:315-325.

Wolters, N. & Schwarz, W., 1956,  "Untersuchungen Uber Vorkommen  und Verhal-
     ten von Mikroorganismen", Arch. Hydrobiol.  51:500-541.

Woodward, R.L., Dostal, K.A. & Robeck, G.G., 1964, "Granular Activated
     Carbon Beds For Odor Removal", J. Am. Water Works  Assoc. 56(3):287-295.
                                     290

-------
                                BIBLIOGRAPHY


Abram, J.C., 1973, "The Characteristics of Activated Carbon",  Proc.  Water
     Research Assoc.  Conf., Readings,  England,, pp.  1-30.

Ackerman, R.A. & Cosentino» I.A., 1979,, "Recreational  Area Water Reuse
     System", in Proc.  Water Reuse Symp., Washington,  DC,  25-30 March.  Am.
     Water Works Assoc., Denver, CO,  pp. 1755-1761.

Ahlberg, N. & Rupke,  J.W.G., 1974, "Modes of Activated Carbon  Utilization in
     Wastewater Treatment", Proc. Physical-Chenical  Treatment, Activated
     Carbon Adsorption in Pollution Control, Seminar held  in Ottawa, Ontario,
     Canada, Oct. 24, 1374"Environmental Protection  Service, Environment
     Canada.

Aldridge, J.L., 1973, "Design of Systems for Applying  Powdered Activated
     Carbon", Proc. Water Research Assoc. Conf., Reading,  England,  pp. 145-
     158.

Anonynous,  1974, "Ozone-Carbon Dye Waste Treatment", "Textile Ind. 138(10):-
     43, 45.

Argo, D.3., 1978S "Control  of Organic Chenical Ccritaminants in Drinking
     Water", Presented at EPA Seminar on Control cf Organic Contaminants in
     Drinking Water,  Los Angeles, Calif., Nov.  L'.S. Environmental  Protection
     Agency, Office of Drinking Water, Washington, D.C.   20460.

Argo, D.3., 1979, "Control  of Organic Chemical Contaminants in Drinking
     Water", Presented at EPA Seminar on Control of Organic Chemical Contami-
     nants  in Drinking Water, Dallas, Texas, March.   U.S.  Environmental
     Protection Agency, Office of Drinking Water,  Washington,  D.C.   20460.

Argo, D.G., 1980, "Cost of Water Reclamation by Advanced Wastewater Treatment",
     0. Water Poll. Control Fed., 52(4):750-759.

Bailey, D.A., Bayley, R.W.  & Waggott, A., 1973, "Treatment of Wastewaters
     With The Aid of Activated Carbon", Proc. Water Research AI$
-------
Baldauf, 3., 1977, "Competitive Adsorption Data  in  Bisolute  Systems",
     presented at the Activated Carbon Meeting of KIWA,  Engler-Bunte  Institute
     and Water Research Center, Karlsruhe, Fedederal  Republic  of  Germany,
     Oct. 30-31, 1977.

Baldauf, G., Prick, B. & Sontheimer,  H.,  1977,  "A Model  for  the Description
     of Competitive Adsorption of Two-Solute Mixtures",  presented at  the
     Activated Carbon Meeting of KIWA, Engler-Bunte Institute  and Water
     Research Center, Karlsruhe, Federal  Republic of Genrany,  Oct.  30-31,
     1977.

Baldauf, G., 1979, "Prediction of Organic Breakthrough Patterns From  Labora-
     tory Tests", Presented at NATO/CCMS  Conf. on Adsorption Techniques  1n
     the Treatment of Drinking Water, Reston, Va.,  Apr.  30-May 2.

Becker, D.L. & WHson, S.C., 1978, "The Use of Activated Carbon for the
     Treatment of Pesticides and Pesticidal Wastes", in  Carbon Adsorption
     Handbook, P.N. Cheremisinoff & F,C.  Ellerbusch, editors,  Ann Arbor
     Science Publishers, Inc., Ann Arbor, Michigan, pp.  167-214.

Beebe, R.L., 1973, "Activated Carbon Treatment of Raw Sewage in Solids
     Contact Chambers", EPA Report No. EPA-R2-73-183, 98 pp.  U.S.  Environmen-
     tal Protection Agency, Washington, D.C.  20460.

Benedek, A., 1974, "Carbon Evaluation and Process Design",  Proc.  Physical -
     Chemical Treatment, Activated Carbon in Pollution Control, Seminar  held
     in Ottawa, Ontario, Canada, Oct. 24, T9~74.   Environmental Protection
     Service, Environment Canada, Ottawa.

Benedek, A. & Najak, A., 1975, "The Biological  Regeneration  of Activated
     Carbon", Presented at 43ch Annual Conf. of Water Pollution Control
     Fed., Miami Beach, Fla. (Oct.).   Water Poll. Control Fed., Washington,
     D.C.

Benedek, A., 1978, "Simultaneous Biodegradation  and Activated  Carbon  Adsorp-
     tion - A Mechanistic Look", presented at the Symposium  on Activated
     Carbon Adsorption of Organics From the Aqueous Phase, Miami, Florida,
     Sept., 1978.  Am. Chem. Soc., Washington,  D.C.

Benedek, A., 1979, "The Effect of Chemical Cxldatlon on Activated Carbon
     Adsorption and Biodegradation", Presented at NATO/CCMS  Conf. on  Adsorp-
     tion Techniques in the Treatment of  Drinking Water, Reston,  Va.s Apr.
     30-May 2.

Benedek, A., Stephenson, P., f^alaiyandi,  M. & Lancaster, E., 1979,  "The
     Effect of Czone on the Biological Degradation and Activated  Carbon
     Adsorption of Natural and Synthetic  Organics in Water", Presented  at
     4th World Ozone Congress, Houston, Texas,  Nov. 26-29.   Intl.  Ozone
     Assoc., Vienna, VA.
                                     292

-------
Bernardin, F.E.,  19739  "Cyanide Detoxification  Using Adsorption and Catalytic
     Oxidation on Granular Activated  Carbon", J.  Water  Poll. Control  Fed.
     45(2):221-231.

Berndt, C.L. & Meidl,  J.A., 1979,  "Startup  and  Operation  of  a  6.5 mgd  PACT
     Process With Wet  Air Regeneration",  Presented  at 53rd Annual Meeting of
     the Ohio Water Pollution Control  Conf,9 Toledo, Ohio.,  June 20-22.

Bernhardt, H., 1970, "Entkeimung von  AktivkohlefHtern  durch Erwarmung",
     Schriftenreihe der Vereins Wasser0  Eoden und Lufthygiene,  Berlin,
     31:165-169.

Bertrandy, M.C.,  1974,  "Filtrable Residual  Organics From  Biooxidation  and
     Activated Carbon  Adsorption of Domestic Wastewater", Master's Thesis,
     Wastewater Research Group,, McMaster Univ., Hamilton, Ontario, Canada.

Besik, F., 19730  "High  Rate Adsorption - Biooxidation of  Domestic Sewage",
     Water & Sewage Works, June 1973,  pp. 68-71.

Besik, F., 1975,  'Simultaneous Removal of Nitrogen  and  Organics in New
     Activated Sludge  Process", Special  Conf. on Nitrogen As A  Water  Pollutant,
     Copenhagen,  Denmark, Aug. 18-20,  1975.   Intl.  Assoc. of Water Poll.
     Research.

Bettens, L., 1979, "Powdered Activated Carbon in an Activated  Sludge  Jnit'1,
     Effluent & Water  Treatment J.  19(3)-.129-135.

Birkmeier, J.L.,  Sr.,  LaRocca, S.A. & Haulsee,  R.E.,  1978,  "Activated  Carbon
     Removes Pesticides From Wastewater At  Research Facility",  Indl.  Wastes
     Sept./Oct.„  p.  20-24.

Bishop, D.F., et a]_. „  1967S "Studies  on Activated Carbon  Treatment",  J.
     Water PolTT Control Fed. 30:188.

Blanck, C.A., 1978, "End Taste and Odor Complaints  With 3ranular Activated
     Carbon", The American City & County, Review Issue, pp.  9-10.

Blanck, C.A., 1979a, "Taste and Odor Control  Utilizing  Granular Activated
     Carbon in the Plains Region of The American Water  Works System",
     Presented at Am.  Soc. Civil Engrs.  Mtg., Central  Illinois  Section,
     March 22.  Am. Soc. Civil Engrs., New  York, N.Y.

Blanck, C.A.s 1979b, "Plant Operations as Related to Trihalomethanes",
     presented at 70th Annual Meeting, Illinois Section,  Am. Water Works
     Assoc., Oak Brook, 111., March 21.

Blanck, C.A., 1979c, "Trihalomethane Reduction  in Operating  Water Treatment
     Plants", J.  Am. Water Works Assoc.  71(9):525-528.

Blanck, C.A., (date unknown),  "In-Plant Control of  Taste  and Odors  in the
     Plains Region of the American Water Works  System'.  Am. Water Works
     Service Co., Inc., Fairview Heights, I1U

                                     293

-------
Blanck, C., 1980,  "GAC Value Exceeds  Expectation", Water  &  Sewage Works,
     Jan., p.  40-42.

Boenlng, P.M., Beckmann D.D. & Snoeyink,  V.L.,  1980,  "Activated Carbon
     Versus Resin  Adsorption of Humic Substances", J. Am. Water Works
     Assoc. 72(l):54-59.

Bradley, K.C., 1974,  "Design of an Activated  Carbon Wastewater Treatment
     System for a  M.u Hi-Product Organic Chemical  Plant",  Proc. Physical-
     Chemical  Treatment,  Activated Carbon 1n  Pollution  Control, Seminar held
     1n Ottawa, Ontario,  Canada, Oct. 24, WfT.Environmental Protection
     Service,  Environment Canada, Ottawa.

Brodtmann, N., DeMarco, J. & Greenberg, D.,  1978, "Critical  Study of Large
     Scale Granular Activated Carbon  Filter  Units for the Removal of Organic
     Substances from Drinking Water", presented at the  Symposium on Activated
     Carbon Adsorption of Organlcs From the  Aqueous Phase,  Miami, Florida,
     Sept., 1978.   Am. Chem. Soc., Washington,  D.C.

Bull, R.J., Pereira,  M.A. & Blackburn, K,U,  1979, "Bioassay Techniques for
     Evaluating the Possible Carcinogenicity of Adsorber  Effluents", Presented
     at NATO/CCMS  Conf. on Adsorption Techniques in Drinking Water Treatment,
     Reston, Va.,  Apr. 30-May 2.

Burns, D.E. & Shell,  G.L., 1973, Physical-Chemical Treatment of a Municipal
     Wastewater Using Powdered Activated Carbon", EPA Report No. EPA-R2-73-
     264, 230 pp.   U.S. Environmental Protection Agency,  Washington, D.C.
     20460.

Burns, D.E. & Shell,  3.L., 1974, "Carbon Treatment of a Municipal Waste-
     water", J. Water Poll. Control Fed.  46:148-164.

Burley, M.J. & Short, C.S., 1973, "Economics  of Activated Carbon Treatment",
     Proc. Water Research Assoc. Conf., Reading, England, pp.  203-240.

Cairo, P.R., McElhaney J. & Suffet, I.H., 1979a, "Design  and Operational
     Experiences With Activated Carbon Adsorbers: Treatment of Delaware
     River Water", presented at NATO/CCMS Conf.  on Adsorption Techniques  in
     Drinking Water Treatment, Reston, Va.,  Apr. 30-May 2.

California State State Water Resources Control  Board, 1978,  "The Effects  of
     Advanced Wastewater Treatment on Trace  Organic Compounds", Calif. State
     State Water Resources Control Bd.s Publicn. No.  61,  July, 147 pp.

Carnahan, R.P. & Keinath, T.M., 1973, "Competitive Adsorption From Hetero-
     geneous Aqueous Solutions Onto Activated Carbon",  Proc.  Am. Chem. Soc.
     Natl. Meeting.,  April 8-13s General  Papers, p. 101"!  A"m.Them. Soc.,
     Washington, D.C.

ChSdal, J. & Schulhof, P., 1979, "Reduction  of the Content  of Chlorine
     Compounds by a Treatment Combining Physlco-Chemical  and Biological
     Processes", in Cxidation Techniques in  Drinking  Water  Treatnent, W.


                                     294

-------
     KUhn & H.  Sonthelmer, editors,  U.S.  EPA Report No.  EPA-57C/9-79-020.
     U.S. EPA,  Office of Drinking Water,  Washington,  DC,  p.  176-194.

Cheremisinoff,  P.M. & Morresi,  A.C., 1978,  'Carbon Adsorption  Applications",
     in Carbon  Adsorption Handbook,  P.N.  Cheremisinoff & F.  Ellerbusch,
     editors, Ann Arbor Science Publishers,  Inc.,  Ann Arbor, Mich.,  pp.  1-
     54.

Chian, E.S.K. & DeWalle, F.,  1975, "Sequential  Substrate Re-noval  in  Activa-
     ted Sludge Systems Fed With Naturally  Occurring  Wastewater",  Prog.
     Water Techno!. 7:235-241.

Chian, E.S.K.,  Chang, Y., DeWalle, F.B.  & Rose, W.B., 1975,  "Combined
     Treatment of an Organic Chemical Water by Activated Sludge Followed by
     Activated  Carbon, Proc.  30th Purdue Univ.  Indl.  Waste Conf.,  May  6-8,
     1975.  Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan, pp.
     966-972.

Clark, R.M., Dorsey, P. & Adams, J.Q., 1979, "Influence of Operating Varia-
     bles on the Cost of Treatment by Adsorption", Presented at NATO/CCMS
     Conf. on Adsorption Techniques  in the  Treatment of Drinking Water,
     Reston, Va.9 Apr. 30-May 2.

Cohen, J., 1974, "Introduction to Physical-Chemical Treatment", Proc.
     Physical-Chemical Treatment, Activated Carbon in Pollution Control,
     Seminar held in Ottawa,  Ontario, Canada, Oct. 2T, 1974.Environmental
     Protection Service, Environment Canada, Ottawa.

Consumer Product Safety Comnissiont  1979, "Horre-Use Carbon-Only Water
     Fillers",  Federal Register ^4(180):53559-53560.

Cookson, J.T.,  Jr., Ishizaki, C. & Jones, C.R., 19728 "Functional  Surface
     Groups on Activated Carbon Responsible for Adsorption of  Crganics From
     Water", Water — 1971, L.K. Cecil,  Ed., AIChE Symp.  Series 68(124):157-
     168.

Cookson, J.T.,  Jr., 1975, "The Adsorption Chemistry of Crganics From Waters
     on Activated Carbon", Water —  1975, G.F. Bennett, Ed., AIChE Symp.
     Series 71(151):376-386.

Cookson, J.T.,  Jr., 1978, "Adsorption Mechanisms:  The Chemistry of Organic
     Adsorption on Activated Carbon", in Carbon Adsorption Handbook, P.N.
     Cheremisinoff & F. Ellerbusch8  editors., Ann Arbor Science Publishers,
     Inc., Ann Arbor, Michigan, pp.  241-280.

Cooper, J.C. & Hagar, 3.S«, 1967, "Water Reclamation With Granular Activated
     Carbon", in Water Reuse, L.K. Cecil, editor.   Chetn. Engrg. Progress
     Symp. Series~53T78TTT85"-192.  Am. Inst. Chem. fcngrs., New York, NY.

Culp, 3.L. & Suhr,  L.G.,  1973,  "State of the Art of Activated  Carbon Treat-
     ment of Wastewater", Presented at the Physical/Chemical  Treatment -
     Activated Carbon in Water  Pollution Control Seminar, Environment

                                     295

-------
     Canada, Environmental  Protection Service,  Technology  Transfer  Seminar,
     Vancouver, Canada.

Culp, R.L. & Hansen, S.P.,  1980,  "Carbon Adsorption  Enhancement  With  Ozone",
     J.  Water Poll. Control Fed.  52(2):270-281.

Culp/Wesner/Culp, 1978,  "Estimating Costs for Granular Activated Carbon
     Systems in Water Purification — Based on Experience  in  Wastewater
     Treatment", presented  at meeting of Natl.  Drinking Water Advisory
     Council, Washington, D.C., Aug. 22-23, 1978.  U.S. EPA,  Office of
     Drinking Water, Washington,  D.C.

Cummins, B.B. & Nash, H.D., 1978, "Microbiological  Implications  of  Alterna-
     tive Treatment", Presented at AWWA Technol.  Conf., Dec.  4., Louisville,
     Ky.  Am. Water Works Assoc., Devner, Colo.

Daigger, G.T. & Grady, C.P.L., Jr., 1977, "A Model for the Bio-Oxidation
     Process Based on Product Formation Concepts", Water Research 11:1049-
     1057.

Daniels, S.L., 1972, "The Adsorption of Microorganisms Cnto Solid Surfaces",
     in Developments in Industrial Microbiology 13:211-253.  Soc. Indl.
     Microbiology, WasTington, D.C.

Davis, W.R., 1972, "Control of Stream Pollution at  the Bethlehem Plant",
     Iron & Steel Engr.  Year Book, p. 315.

DeJohn, P.B., 1974, "Factors to Consider When Selecting Granular Activated
     Carbon for Wastewater Treatments presented at  29th Annual Indl.  Waste
     Conf., Purdue Univ., W. Lafayette, Indiana,  May 7-9,  1974.

DeJohn, P.B. & Adams, A.D., 1975, "Treatnent of Oil  Refinery Wastewaters
     With Granular and Powdered Activated Carbon",  Proc. 30th Purdue  Univ.
     Indl. Waste Conf., May 6-8, 1975.  Ann Arbor ScTenceTuFl 1shers  TncT,
     Ann Arbor, Michigan, pp. 216-232.

DeJohn, P.B., 1976, "Factors to Consider When Treating Dye Wastes With
     Granular Activated Carbon", Proc. 31st Purdue  Univ. Indl. Waste
     Conf., May 4-6, 1976.   Ann Arbor Science Publishers,  Inc.,  Ann Arbor,
     RTcTfigan, pp. 375-384.

DeMarco, J. & P.R. Wood, 1978, "Design Data For Organlcs Removal by Carbon
     Beds"t presented at Envmntl. Engrg. Div., Natl. Conf. on Engrg.  Design,
     ASCE, Kansas City, MO, 10-12 July.

DeMarco, J., 1979, "Practical Application of Adsorption Techniques  in
     Drinking Water", Presented at NATO/CCMS Conf.  on Adsorption "echniques
     in Drinking Water Treatment, Reston, Va., Apr.  30-May 2.

DeWalle, F. & Chian, E.S.K., 1974, "Removal of Organic Matter by Activated
     Carbon Columns", Environ. Engr. Div., Am. Soc.  Civil  Engrs. ICO:1085-
     1104.


                                    296

-------
DeWalle, F.B.0 Chian, E.S.K. & Small,  E.M.,  1976,  "Organic  Matter Removal
     by Powdered Activated Carbon Added to Activated Sludge Jnits",  J.  Water
     Poll. Control Fed., 1n Press.

DeWalle, F.B. & Chian, E.S.K., 1977,  "Biological  Regeneration  of Powdered
     Activated Carbon Added to Activated Sludge Units",  Water  Research
     11:439-446.

DiGiano, F.A. & Webers W.J., Jr., 1972, "Sorption  Kinetics  in  Finite Bath
     Systems", J. San. Engr. Div., Proc. Am.  Soc.  Civil  Engrs.  SA 6:1021.

DiGiano, F.A. & Weber, W.J., Jr., 1973, "Sorption  Kinetics  in  Infinite  Bath
     Experiments", C. Water Poll. Control  Fed.  45(4)-.713-725.

DIGiano, F.A., 19793, "General Considerations in Assessing  the Beneficial
     Aspects of Microbial Activity on  GAC",  Presented at NATO/CCMS Conf. on
     Adsorption Techniques in Drinking Water Treatment,  Reston, Va., Apr.
     30-May 2.

DiGiano, F.A., 1979b, "Influence of Microbial Activity on the  Performance  of
     GAC", presented at AWWA Pre-Conference  Seminar on Organics, San Francisco,
     CA, 24 June.  Am. Water Works Assoc., Denver, CC.

Directo, L.S. & Chen, C.L., 1974, "Pilot Plant Study of Physical-Chemical
     Treatment", Presented at 47th Ann. Conf. of Water Poll.  Control Fed.,
     Denver, Colo.

Directo, L.S., Chen, C. & Miele, R.P., 1978,  "Two-Stage Sranular Activated
     Carbon Treatment", U.S. EPA Report No.  EPA-6CO/2-78-170.   U.S.  Environ-
     mental Protection Agency, Cincinnati, Ohio  45268.

Dobbs, R.A., Middendorfs R.J. & Cohen, J.M.,  1978, "Carbon  Adsorption
     Isotherms For Toxic Organics", Municipal Environmental Research Labora-
     tory, Office of Research & Development,  U.S.  EPA, Cincinnati, Ohio
     45268.  Unnumbered Report, dated  May, 1978.

Dobias, J. & Starz, E., 1977, "Ozonelnsatz In Kleineren Trinkwasserversor-
     gungsanlagen", Intl. Synp. on Ozone & Water,  Wasser-Berlln - 1977, pp.
     253-260.  AMK-Berlin, Federal Republic of Germany;  Intl.  Ozone Assoc.,
     Vienna, VA.

Dressman, R.C., McFarren, E.F. & Symons, J.M., 1977, "An Evaluation of  the
     Determination of Total Organic Chlorine (TOC1) in Water by Adsorption
     Onto Granular Activated Carbon, Pyrohydrolysis and Chloride :on Measure-
     ment", presented at Water Quality Technology Conf., Kansas City, Mo.,
     Dec. 5-6, 1977.  Am. Water Works  Assoc., Denver, CC.

Eberle, S.H., Sttiber, H. & Donnert, D., 1975, "Study on the Adsorption
     Properties of Aluminum Oxide and  Its Application for the Purification
     of Groundwater Containing Humic Substances", in Translation of Reports
     on Special Problems of_ Water Technology, Vol. 9 - Adsorption, H. Sont-
                                     297

-------
     heimer,  editor.   EPA  Report  No.  EPA  600/9-76-030, Dec. 1976, pp. 405-
     413.

Elmaleh, S.,  LaBaquSre,  H.  &  Ben  Aim,  R.,  1978,  "Biological Filtration
     Through  A Packed Column",  Water  Research  12:41-46.

English, J.N., Masse, A.N., Carry,  C.W.,  Pltkin,  J.B. & Hasklns, J.E.,  1971,
     "Removal of Organics  From  Wastewater by Activated Carbon", Water —
     1970, Chem. Engrg.  Prog. Symp. Series 67(107):147-153.

English, J.N., Carry, C.W., Masse,  A.M.,  Pitkin,  J.B. & Dryden, F.D., 1974,
     "Denltrification in Granular Carbon  and Sand Columns", J. Water Poll.
     Control  Fed. 46(1):28-42.

Evans, F.L.,  III & Ryckman, D.W., 1963,  "Ozonated Treatment of Wastes
     Containing ABS", Proc. 18th  Indl. Waste Conf.,  Purdue Univ., Lafayette,
     Indiana, pp. 141-157.

Evans, T.M.,  SchilUnger,  J.E.  &  Stuart,  O.G.,  1978,  "Rapid Determination of
     Bacteriological  Water Quality by Using Limulus  lysate", Appl.  Env.
     Microbiol. 35:376.

Famularo, J., Pannu,  A.S.  & Mueller,  J.A., 1978»  "Prediction of Carbon
     Column Performance From  Pure Solute  Data",  Presented at 51st Natl.
     Conf. of the Water Poll. Control  Fed., Anaheim,  Calif., Oct. 1978.
     WPCF, Washington, D.C.

Fenco Consultants, 1977, "The Removal  of  Organics From Potable Water Sup-
     plies By Ozonation and Activated Carbon;   Biodegradation and Adsorption",
     proposal submitted to Dept.  of Health and Welfare, Canada.  Fenco
     Consultants Ltd., Toronto, Canada.

Ferguson, J.F., Keay, G.F.P., Merrill, M.S. &  Benedict, A.H., 1979,  "Pow-
     dered Activated Carbon In  Contact Stabilization  Activated Sludge", J.
     Water Poll. Control Fed.  51(9):2314-2323.

Ferrara, A.P., undated (but after 1975),  "Replacement of Filter Media With
     Granular Activated Carbon".   ICI United States,  Inc.

Ferrara, A.P., undated (but after 1977),  "Controlling Bed Losses of Granular
     Activated Carbon Through Proper  Filter Operation".  ICI United States
     Inc.

Ferrara, A.P., 1980,  "Controlling Bed Losses of Granular Activated  Carbon
     Through Proper Filter Operation", J.  Am.  Water  Works Assoc. 72(1):60-
     63.

Fiessinger, F., 1978, "Large  Scale Applications  of Granular Activated
     Carbon for Potable Water Production  in France",  presented at Symp. on
     Activated Carbon Adsorption  of Organics From the Aqueous Phase, Miami,
     Fla., Sept., 1978.  Am.  Chem.  Soc.,  Washington,  D.C.
                                     298

-------
Fiessingers F., 1979a, "The Cost of Activated Carbon  "reatnent in  France",
     Presented at NA70/CCMS Conf. on Adsorption  Techniques  in  Drinking  Water
     Treatments Reston9 Va.9 Apr. 30-May 2.

Fiessingers F., 1979b, "Biological  Activated Carbon....A Myth  or A Promising
     Reality?", Aqua 2.

Fiore, J.V. & Babineau, R.A., 1977, "Effect of an Activated Carbon Filter
     on the Microbial Quality of Water", Applied & Environmental Microbiology
     34:541-546.

Fisher, J.L., 1980, "How to Conduct a GAC Organics Removal  Pilot Study",
     Public Works, June, p. 1CO-102.

Flotard, R.D., Ford, D.L. & Harrison, W., 1979,  "Use  of  Granular Activated
     Carbon to Remove Trace Organics From the Effluent of a Class-B Refinery:
     A Case Study", Natl. Tech.  Info. Service, Springfield, VA,  Report  Mo.
     CONF-7906120-1.

Flynn, B.P. & Barry, L.T., 1976, 'Finding A Home For  Carbon: Aerator (Pow-
     dered) Or Column (Granular)",  Proc. 3]_st Purdue  Univ.  Indl. Waste
     Conf., May 4-6, 1976.  Ann  Arbor Science Publishers Inc., Ann Arbor,
     Michigan, pp. 649-660.

Flynn, B.P., Robertaccio, F.L. & Barry, L.T., 1976, "Truth  or  Consequences:
     Biological Fouling and Other Considerations in the  Powdered Activated
     Carbon - Activated Sludge System", Proc. Jljt, Indl. Waste Conf.,
     Purdue Univ., pp. 855-862.   Ann Arbor Science Publishers,, Inc., Ann
     Arbor, RTcTfigan.

Flynn, B.P. & Stadnik, J.G., 1977,  "Startup of a 38 mgd  Powdered Activated
     Carbon - Activated Sludge (PAC~) Treatment Systeir at OuPont's Chambers
     Works", Presented at the 5Cth Annual Water Poll. Control  Fed. Conf.,
     Philadelphia, Pa., Oct. 3.

Fokken, B. & Kurz, R., 1979, "Removal of Purgeable Organic  Chlorine Cotr-
     pounds by Activated Carbon Adsorption", Presented at NATC/CCMS Conf. on
     Adsorption Techniques in Drinking Water Treatment,  Reston, Va., Apr.
     30-May 2.

Ford, D.L. & Buercklin, M.A., 1972, "The Interrelationship  of  Biological-
     Carbon Adsorption Systems for the Treatment of Refinery and Petro-
     chemical Wastewaters", Presented at the 6th Intl. Assoc.  Water Poll.
     Research Conf., Jerusalem,  Israel, June.

Ford, D.L., 1976, "Current State of the Art of Activated Carbon Treatirent",
     in Proc. Open Forum on_ Management of Petroleum Refinery Wastewaters.
     F.S. Manning, editor.  U.S. EPA, RoFt. S. Kerr Environmental  Research
     Labs., Ada, OK, p. 295-344.

Ford, D.L., 1977, "Putting Activated Carbon in Perspective to  1983 Guide-
     lines", Indl. Water Engrg.  14(3):20-27.


                                     299

-------
Ford, D.L. & Manning, F.S.,  1978,  "Treatment of Petroleum Refinery  Waste-
     water", In Carbon Adsorption  Handbook,  P.M.  Cherem1s1noff  &  F.  Ellerbusch,
     editors, Ann Arbor Science Publishers,  Inc., Ann  Arbor,  Michigan,  pp.
     687-732.

Frick, B., 1977, "Use of Adsorption Data For Characterization of  Organlcs  1n
     Natural Water Systems", presented at Activated Carbon Meeting,  WRC-
     KIWA-EBI, Karlsruhe, Federal  Republic of Germany, 27-28  Oct.   Engler-
     Bunte Inst., Univ. Karlsruhe, FRG.

Frick, B., Bartz, R., Sontheimer,  H.  & DiGiano, F., 1977, "Problems  of
     Designing Filter Plants", presented at Activated  Carbon  Meeting,  KIWA,
     Engler-Bunte Inst. & Water Research Center,  Karlsruhe, Federal  Republic
     of Germany, Oct. 30-31, 1977.  Engler-Bunte Inst. der Univ.  Karlsruhe.

Frick, B., 1979, "Prediction of Multlcomponent Adsorption Behaviour 1n
     Activated Carbon Adsorbers -  Equilibrium Aspects", Presented at NATO/CCMS
     Conf. on Adsorption Techniques in Drinking Water  Treatment,  Reston,
     Va., Apr. 30-May 2.

Fritz, W. & Schluender, E.-U., 1974, "Simultaneous Adsorption Equilibria  of
     Organic Solutes on Activated  Carbon", Chem,  Engrg. Sci.  29:1279-1282.

Fulton, G.P., 1980, "New York City's Pilot Plant Filter Studies,  Part  One",
     Public Works, April, p. 62-64.  "Part Two", Public Works,  May,  p.  85-
     88.

Gardiner, E.R., 1973, "Experience  With Powdered Activated Carbon  For Taste
     And Odor Control", Proc. Water Research Assoc. Conf., Reading,  England,
     pp. 241-262.                           '

Gardner, F.H., Jr. & Williamson, A.R., 1976, "Naval Stores Wastewater
     Purification and Reuse by Activated Carbon Treatment", EPA Report No.
     EPA-600/2-76-227, Oct., 1976.  Industrial Environmental  Research
     Laboratory, U.S. EPA, Cincinnati, Ohio  45268.

Garland, C.F. & Beebe, R.L., 1970, "Advanced Wastewater Treatment Using
     Powdered Activated Carbon in  Recirculating Slurry Contactor  Clariflers",
     U.S. Fed!. Water Qua!.  Admin. Contract No. 14-12-400.  U.S.  Environ-
     mental Protection Agency, Washington, D.C.  20460.

Garrison, W.E., Gratteau, J.C., Hansen, B.E. & Luthy,  R.F., Jr.,  1978,
     "Gravity Carbon Filtration to Meet Reuse Requirements",  J. Environ.
     Engrg. Div.f Am. Soc, Civil Engrs., Dec. 1978, pp. 1165-1174.

Gasser, J.A., Chen, C.-l. & Miele, R.P., 1978, "Fixed  Growth  Nitrification
     of Secondary Effluent", J. Environ. Engrg. Div.,  Am. Soc.  Civil Engrs.
     104(1):77-92.

Glushchenko, V. Yu., Levagina, T.G. & Pershko, A.A., 1975, "Effect  of the
     Nature of a Carbon Surface on the Adsorption of Nitre Compounds From
     Aqueous Salt Solutions", Kol!o1dny1 Zhurnal 37(1): 134-136.

                                     300

-------
Goodall, J.B. & Hyde, R.A.,  1979,  "Current United Kingdom  Practice  In  the
     Use of Granular Activated Carbon 1n Drinking Water  Treatment",  Presented
     at NATO/CCMS Conf.  on Adsorption Techniques  1n  Drinking  Water  Treatment,
     Reston, Va., Apr. 30-May 2.

Grandjacques, B.L. & Waller, G.,  1973, "Design of Carbon Beds",  Proc.  Water
     Research Assoc. Conf.,  Reading,  England,  pp. 159-176.

Guarino, C.F., Radziul9  C.V., Cairo,  P.R., Blair, ~.D.,  Pence, M.M., Apto-
     wicz, B.S. & Welntraub, N.,  1980, "Design and Economic Considerations
     of GAC Systems", Water & Sewage  Works, Sept., p.  58-59,  63-64,  66-67.

Gulf South Research Institute, 1979,  "Results  Fron Phase 1 Report of GSRI on
     Home Water Treatment Units",  U.S. EPA8 Office of  Drinking Water,  Washing-
     ton, DC.

Guy, M.D., Mclver, J.D.  & Lewis,  M.J., 1977, "The Removal  of  Virus  by  a
     Pilot Treatment Plant", Water Research 11:421-428.

Halaby, D., Helm, J.M. & Zolin, P., 1978, "Economic  Evaluation of Full Scale
     Carbon Regeneration Systems", presented at Symp.  on Activated  Carbon
     Adsorption of Organics From the  Aqueous Phase,  Vianri, Fla., Sept.   Am.
     Chen. Soc., Washington, D.C.

Hals, 0. & Benedek9 A.0  1973, "Simultaneous Biological Treatment and Activa-
     ted Carbon Adsorption", presented at 46th Annual  Conf.,  Water  Poll.
     Control Fed., Cleveland, Ohio, Get.

Hals, 0, 1974, "Biological Treatment  and Activated Carbon  Adsorption", M.
     Engr. Thesis, ^ater Research Group, Mcfaster Univ.s Hamilton,  Ontario,
     Canada.

Hanna, Y.A., Meister, T. & Slough, J., Jr., 1977, "Ozone as a Pretreatment
     Step for Physical-Chemical Treatment Process.  Part I",  Presented at
     Seminar on Current Status of Wastewater 'reatnent and Disinfection  With
     Ozone, Cincinnati,  Ohios Sept.  Intl. Ozone Assoc., Vienna, VA.

Hansen, K.H., Jr., 1979, "Control  of  Synthetic Crganlc Chemicals by Granular
     Activated Carbon -- Theory,  Application and Reactivation Alternatives",
     presented at Seminar on Control  of Organic Chemical Contaminants  in
     Drinking Water, Dallas, Texas, March 13-14.   U.S.  EPA, Office  of  Drinking
     Water, Washington,  D.C.

Hattori, T., 1973, Microbial Life in  the Soil; An Introduction.  M.  Dekker,
     Publishers, New York, N.Y.

Hay, W.C., 1975, "Pilot Plant Studies of the Use of  Radiation-Induced
     Oxidation for Treatment of Industrial and Municipal WastewaterJ,  Proc.
     WWEMA 3rd Annual Indl. SoIns.  Ann Arbor Science Publishers,  Inc.,  Ann
     Arbor,~MTchigan, pp. 507^527.
                                     301

-------
Heath, H.W., Jr., 1977, "Combined Powdered Activated Carbon  -  Biological
     ("PACT") Treatment of 40 mgd Industrial  Waste", presented at  Symp.  on
     Indl. Waste Poll.  Control, ACS Natl.  Meeting,  New Orleans,  LA,  Var.  24.
     Am. Chem. Soc., Washington, D.C.

Helfgott, T., Hunter, J.V. & Rlckert,  D.,  1970,  "Analytical  and  Process
     Classification of Effluents", J.  Sanit.  Engr.  Div.,  Am.  Soc.  Civil
     Engrs. 96(SA3):779-803.

Heukelekian, H. & Heller, A., 1940, "Relation Between Food Concentration and
     Surface for Bacterial Growth", J. Bacteriology 40:547-558.

Heymann, E., 1975, "Practical Experiences  in  the Use of Flocculation and
     Filtration Connected in Series to Granular  Activated Carbon Filters",
     in Translation of Reports o_n_ Special  Problems  of_ Water  Technology,  Vol.
     9_ - Adsorption,Tr. Sontheimer, editor, op.  eft., pp. 112-127.

Hiser, L.L. 8 Tarazi, D.S., 1971, "Application of Carbon  Techniques  to
     Biological Systems", Water « 1970, Chem. Engrg. Prog.  Symp.  Series
     67(107):267-278.

Hodges, W.E., 1978, "Filtration Pilot Studies for the City of Los  Angeles",
     presented at Forum of Assoc. of Environ. Engrg, Profs.,  Atlantic City,
     N.J., June 28, 1978.

Hoign6, J. & Bader, H., 1978, "Ozone Initiated Oxidations of Solutes in
     Wastewater: A Reaction Kinetic Approach", Prog. Water Techno!.  10(5/6):-
     657-671.

Holladay, D.W., Hancher, C.W., Chilcote, D.D. & Scott, C.D.,  1976, "Biodegra-
     dation of Phenolic Waste Liquors in Stirred Tank, Columnar and  Fluidized
     Bed Bioreactors", presented at 69th Ann. Meeting, AIChE,  Chicago, 111.,
     Nov. 28-Dec. 2.  Am. Inst. Chem.  Engrs., New vork, N.Y.

Holluta, J., et al_., 1960, "Untersuchungen Uber Geruchsbildene Organische
     Stoffe infWasser und Deren Isolierung und Identiflzierung", Monats-
     heftbull. SVGW 40:105-112.

Holluta, J., 1964, "Organische Extraktstoffe  in Oberfla'schenwa'ssern  und
     Deren Bedeutung ftlr die TMnkwasserversorgung", Forschung und Fortsch-
     ritte 38(6):164-166.

Hfllzel, G., 1977, "Control of Virgin and Reactivated Carbon for Drinking
     Water Treatment", presented at Activated Carbon Meeting of the  KIWA,
     Engler-Bunte Inst. and Water Research Center,  Karlsruhe,  Federal Republic
     of Germany, Oct. 30-31, 1977.  Engler-Bunte Inst. der Univ. Karlsruhe.

Hfllzel, G., 1979, "Laboratory Activated Carbon Test Methods  For Water
     Utilities", presented at NATO/CCMS Conf. on Adsorption ~echniques in
     Drinking Water Treatment, Reston, Va., Apr. 30-May 2.
                                     302

-------
Hopkins, C.B., Weber, W.J.,  Jr.  & Blooms,  R-»  1968,  "A Comparison  of Expanded
     Bed and Packed Bed Adsorption Systems",  U.S.  EPA Report No.  TWRC-2,  74
     pp., U.S. Environmental Protection Agency,  Cincinnati,  Ohio.

Hopkins, C.B., Weber, W.J.,  Jr.  & Bloom,  R.,  1970,  "Granular Carbon Treat-
     ment of Raw Sewage", U.S. EPA Report Mo.  ORO-10750 DAL  05/70.   U.S.
     Environmental Protection Agency, Municipal  Environmental  Research  Lab.,
     Cincinnati's Ohio.

Horvath, R.S., 1972, "Microbial  Co-Metabolism and  the Degradation of Organic
     Compounds in Nature", Bact. Rev. 36:146-155.

Houston, C.W., 1978, "Bio-Regeneration of Hydrocarbon-Exhausted Activated
     Carbon", Tech. Completion Report, U.S.  Cept.  of Interior,  Office of
     Water Resources Research Project No. A-C63-RI.   Natl. Tech.  Info.
     Service Report No. PB 283,651.

Huang, J-C. & Steffens, C.T., 1976, "Competitive Adsorption  of Organic
     Materials by Activated  Carbon", Proc. 31st  Purdue Univ. Indl.  Waste
     Conf., May 4-6, 1976.  Ann Arbor Science Publishers,  Inc., Ann Arbor,
     MTcFigan, pp. 107-121.

Hutton, D.G., 1978, "Coirbined Powdered Activated Carbon -  Biological ~reat-
     nent", in Carbon Adsorption Handbook, P.N.  Cheremisiroff & F.  Ellerbusch,
     editors, Ann Arbor Science Publishers,  Inc.9  Arn Arbor, Michigan,  pp.
     389-448.

Hutton, D.G., 1979, "Removal of Priority  Pollutants  by the DuPont 'PACT'
     Process", Presented at 7th Ann. Indl, Poll. Control  Conf., WWEMA,
     Philadelphia, Pa.s June 5-7.  Water  & Wastewater Equipmt.  vfgrs.
     Assoc., McLean, Va.

Hutton, D.3. & Temple, S., 19798 "Priority Pollutant Removal:   Comparison
     of DuPont  'PACT1 Process and Activated Sludge", Presented at 52nd  Ann.
     Conf. Water Poll. Cortrol Fed., Houston, Texas, Cct.  10.   Water Poll.
     Control Fed., Washington, D.C.

Hutton, D.3., 1980, "Removal of Priority  Pollutants", Irdl.  Wastes 26(2):22,
     24, 26.

Jain, J.S. & Snoeyink, V.L., 1973, "Adsorption From Bisolute Systems on
     Active Carbon", J. Water Poll. Control  Fed. 45(12):2463-2479.

Jain, K.K., 1978, "A Carbon Adsorption Process for Highly Concentrated
     Municipal Waste and Low Water Temperatures",  in Carbon  Adsorption
     Handbook, P.N. Cheremisinoff & F, Ellerbusch,  editors,  Ann Arbor
     Science Publishers, Inc., Ann Arbor, Michigan,  pp. 627-660,

Jain, K.K. & Bryce, A.J., 1978, "Feasibility of Munitions  Wastewater Treat-
     ment by Adsorption-Oxidation:, in Carbon Adsorption Handbook, 1oc_.
     cit., pp. 661-686.
                                     303

-------
Janata, W., 1975, "Investigations Into the Control  of Activated  Carbon
     Filters at Waterworks", in Translation of Reports cm  Special  Problems
     of Water Technology, Vol.  9 - Adsorption, H.  SontheTmer,  editor, op.
     cTt., pp. 230-237.

Jeris, J.S., et a!., 1974, "High Rate Biological  Denitrification Using  a
     GranularTlZTTdlzed Bed", J. Water Poll. Control  Fed.  46:2118.

Johnson, R.L. & Baumann, E.R.,  1971, "Advanced Organlcs Removal  by Pulsed
     Adsorption Beds", J. Water Poll. Control  Fed.  43:8.

Johnson, W.R. & Jenkins, C.R.,  1978, "A Study for Improving the  Aerobic
     Media Trickling Filter for Wastewater Treatment", Final  Report of
     Contract 31-14-31-0001-6051, Office of Rsch.  & Technol.,  Dept.  of
     Interior, Washington, D.C.  Natl. Tech. Info.  Service Report No. PB-
     288,775.

Jones, J.L., Kuo, M.C.T., Kyle, P.E., Radding, S.B.,  Semrau,  K.T.  & Somogyi,
     L.P., 1979, "Overview of the Environmental Control Measures and Problems
     in the Food Processing Industries", EPA Report No. EPA-600/2-79-C09,
     U.S. EPA, Cincinnati, Ohio.

Joschek, H.-I., Dehler, J., Koch, W., Engelhardt,  H.  & Geiger, W.,  1975,
     "Formal Kinetic Model For The Biological  Decomposition Of Industrial
     Wastewater", Chemie-Ing.-Techn. 47:422-428.

Joyce, R.S. & Sukenik, V.A., 1964, "Feasibility of Granular Activated Carbon
     Adsorption For Wastewater Renovation'1, Report No. AWTR-10,  Env. Health
     Series, Water Supply & Poll. Control, U.S. Dept. of Health, Education  &
     Welfare, Public Health Service^ Cincinnati  Ohio.

Joyce, R.S., Allen, J.B. & Sukenik, V.A., 1966, J.  Water Poll. Control.
     Fed. 38(5):813.

Juhola, A.J., 1973, "Regeneration of Activated Carbon", Proc.  Water Research
     Assoc. Conf., Reading, England, pp. 177-202.

JUntgen, H., 1975b, "Phenomena of Activated Carbon Regeneration", in Trans-
     lation of Reports on Special Problems of_ Water Technology,  Vol. 9_ -
     Adsorption, H.SontFeliner, editor, op. cit.,  pp. 269-283.

Kalinski, A.A., 1972, "Enhancement of Biological  Oxidation of Organic
     Wastes Using Activated Carbon in Microbial Suspensions",  Water &
     Sewage Works, p. 115.

Kawazaki et,al_., 1965, Water & Wastewater (Japan)  6:632-648,  778-780.

Keinath, T.M. & Weber, W.J., Jr., 1968, "A Predictive Model for  the Design
     of Fluid-Bed Adsorbers", J. Water Poll. Control  Fed.  40(5):741-765.
                                     304

-------
King, D.L.  & Verma,  R.D.,  1968,  "The Role  of Paniculate  Substrates  in
     Biotlc Degradation",  Proc.  23rd Indl.  Waste  Conf., Purdue  Univ.,
     Lafayette,  Indiana.

Klein, J.,  1975, "Comparative Assessment of Adsorbents",  in  Translation  of
     Reports on  Special  Problems of Water  Technology,  Vol. 9 -  Adsorp-tion,
     H. SontheTmer,  editor,  gg_.  cTt., pp.  215-229.

Klemetson,  S.L.  & Scharbow,  M.D., 1979B  "Filtration  of Phenolic Compounds  in
     Coal  Gasification Wastewater", J.  Water Poll. Control Fed, 51(11):2752-
     2763.

Knopp, P.V., Gitchel, W.B.,  Meidl,  J.A.  &  Berndt,  C.L., 1978,  "Wet Oxidation
     Regeneration",  in Carbon Adsorption Handbook, P.N. Chernnisinoff &  F.
     Ellerbusch, editors,  0£. cit.j, pp.  539-626.

Knoppert,  P.L.,  Oskan, G.  &  Vreedenburgh,  E.G.H.,  1980, "An  Overview of
     European Water  Treatment Practice1', J.  An. Water  Works  Assoc. 72(11):592-
     599.

Kfllle, W.s  1975, "Use of Macroporous Ion Exchangers  for Drinking Water
     Purification" in "ranslation of Reports on Special Problens of^  Water
     Technology, H.  Sontheimer,  edTFor,  op.  cTF.,  pp.  405-4T3.

Koppe, P.,  Sebesta,  G. & Herkelmann, H., 1974a9  "VorlSufige  Mittei^ung Uber
     die Biochemische Oxidation  Einer Schwer Abbaubaren Substanz in  Gegenwart
     von Aktivkohle", Gesundsheits-Ing.  95(2):33-35.

Koppe, P.,  Sebesta,  G. 4 Herkelmann, H.9 1974b,  'The Biochemical Oxidation
     of a Slowly Deqradable  Substance in the Presence  of  Activated Carbon:
     Biocarbon Unit", Gesundsheits Ingenieur 95:247-248.

Koppe, P.,  Imhoff, K.R., Herkelmann, H.  &  Sebesta, 3., 1979, "Some Practical
     Aspects of the  'Biocarbon Process1",  Water Research  13:865-871.

Kreissl, J.F. &  Westrick,  J.J.,  19720 "Municipal  Waste >eatment by  Physi-
     cal-Chemical Methods",  in Application of New Concepts of Physical-
     Chemical Wastewater Treatmint, W.W. EcEenfeTder,  Jr. &~T.K. Ceci1,
     eds., Pergamon  Press, New York, N.Y.

KUhn, W., 1978,  "Fornation and Behavior of Polar  Organochloro Compounds",
     presented at Intl. Symp, on Oxidation Techniques  in  Drinking Water
     Treatment,  Karlsruhec Federal Republic of 3ermanys Sept.  11-13, 1978.
     Engler-Bunte Inst. der Univ. Karlsruhe.

Kussmaul, H., 1979,  "Purifying Action of the Ground  in the Treatment of
     Drinking Water", in Oxidation Techniques in  Drinking Water Treatment,
     W. KUhn S H. Sontheimer, editors, U.S. EPA Report No.  EPA-570/9-79-020.
     U.S. EPA, Office of Drinking Water, Washington, DC.
                                     305

-------
Lang, E.W., Timpe, W.G. & filler, R.L.,  1975,  "Activated  Carbon  Treatment  of
     Unbleached Kraft Effluent For Reuse",  U.S.  EPA  Report  No. EPA-660/2-75-
     004, April, 1975.  Industrial Environmental  Research Laboratory,  U.S.
     EPA, Cincinnati, Ohio  45268.

Lang, E.W., Stephens, J.W. & Miller, R.L.,  1977,  "Activated Carbon  Treatment
     of Kraft Bleaching Effluents", U.S. EPA Report  No.  EPA-600/2-77-119,
     June, 1977.  Industrial Environmental  Research  Laboratory,  U.S.  EPA,
     Cincinnati, Ohio  45268.

Latoszek, A. & Benedek, A., 1974, "Some Aspects  of the Microbiology of
     Activated Carbon Columns Treating Domestic  Wastewater", Water  Research
     Group Report, McVaster 'Jniv., Hamilton, Ontario,  Canada.

Lawson, C.T. & Fisher, O.A., 1973, "Limitations  of Activated Carbon For
     Upgrading Petrochemical Effluents", in Water -  1973, G.F. Bennett,
     editor.  AIChE Symp. Series 70(136):577^817

Lawson, C.T., 1976, "Cautions and Limitations on the Applications of Activa-
     ted Carbon Adsorption to Organic Chemical Wastewaters", in  Proc.  Open
     Forum on Management of Petroleum Refinery Wastewaters, F.S. Manning,
     editor.  U.S. EPA, Robt. S. Kerr Environ. Research Labs., Ada, OK,  p.
     345-358.

LeClair, B.P. & Wei, N., 1974, "The Use of Powdered  Activated Carbon and
     Chemicals in the Treatment of Raw Sewage",  Proc.  Physical-Chemical
     Treatment, Activated Carbon i_n_ Pollution Control, Seminar held in
     Ottawa, Ontario, Canada, Oct. 24,  1974.  Environmental Protection
     Service, Environment Canada, Ottawa.

Lee, D.D. & Scott, C.D., 1977, " A Tapered Fluldized-Bed Bioreactor for
     Treatment of Aqueous Effluents From Coal  Conversion Processes", presented
     at 70th Ann. Meeting of AIChE, New York,  Nov. 13-17.  Am.  Inst. Chem.
     Engrs., New York, N.Y.

Leuser, R.M., 1978, "Pulsed Bed Decolorizing System at the  Pepsico  Sugar
     Refinery - A Case History", in Carbon Adsorption Handbook,  P.N. Cheremi-
     sinoff & F. Ellerbusch, editors, op cit., pp. 779^8T1T

Li, A.Y.L. & DiGiano, F.A., 1980, "The Availability of Sorbed Substrate  For
     Micrcbial Degradation on Granular Activated Carbon", presented at 53rd
     Annual WPCF Meeting, Research Symp.,  Las Vegas, Nev.,  Sept. Water
     Poll. Control Fed., Washington, DC.

Lindstrom, F.T., Haque, R, & Coshow, W.R.,  1970,  "Adsorption From Solution.
     III. A New Model for the Kinetics of Adsorption-Desorption  Processes",
     J. Phys. Chem. 74(3):495-502.

Lipscombe, F., 18628 British Patent No.  2,887.
                                     306

-------
Lombana, L.A. & Halaby,  D.,  19788  "Carbon Regeneration  Systems",  in  Carbon
     Adsorption Handbook^, P.M.  Cheremisinoff & F.  Ellerbusch,  editors,  op_.
     cit., pp. 905-922.

Love, O.T.j Jr., Robeck, G.G.,  Symonss  J.M.  & Buelow, R.W.,  1973,  "Experience
     With Activated Carbon  in the  USA", Proc.  Water Research Assoc.  Conf.,
     Reading,, England, pp.  279-312.

Love, O.T., Jr., Carswell,  J.K., Stevens,, A.A. & Symons,  J.M.,  1975,  'Evalua-
     tion of Activated Carbon", U.S.  EPA, Water Supply  Research Lab.,  Cincin-
     nati, Ohio, 3 Feb.

Love, O.T. & Synons, C.M.,  1978s, "Operational  Aspects of  Granular Activated
     Carbon Adsorption Treatment"» Craft Paper0 U.S.  EPA, Water Supply
     Research Laboratory, Cincinnati, Ohio.

Lowry, J.D. & Burkhead,  C.E., 1978S  "The Role of Adsorption  in  Biologically
     Extended Activated Carbon  Columns", presented at Symp.  on  Activated
     Carbon Adsorption of Organics From the  Aqueous Phase, Miani,  Fla.,
     Sept. 1978.  Am. Chem.  Soc.s  Washington,,  D.C.  Also0 J. Water Poll.
     Control Fed. 52(2)-.389=398 (1980).

Lurgi Express Information,  1977, "Reactivation of Activated  Carbon.
     Thermal Reactivation of Granular Activated Carbon  in the  Two-Stage
     Fluid Bed", Publicn. Mo. T-1191/2.77, Lurgi Apparate-Technik 3mbH  -
     Chemotachnik division,  3-6, Frankfurt/Main, Federal  Republic of Germany.

Lyman, W.J., 1978, "Applicability  of Carbon  Adsorption  to the  Treatment of
     Hazardous Industrial Wastes", in Carbon Adsorption Handbook.  P.N.
     Cheremisincff & F.  Ell^rbuscn,  edTtorsT op. cit., "pp7nTPT66.

Lynch, W.O. & Potter, L.R.,  1979,  "Start-up  and Initial Operation of Physi-
     cal-Chemical Wastewater Treatment Facilities at Cortland,  New York",
     presented at 52nd Annual WPCF Conf., Houston, TX,  10 Cct.   Water  Poll.
     Control Fed., Washington,  DC.

Maier, D., Fuchs, F. & Sontheimer, H.,  1976, "Bestimmung  von Crganischen
    •SMuren in Wassern und auf Aktivkohle",  Wasser/Abwasser  117(2):70-74.

Maqsood, R. & Benedek, A.,  1973,  "The Feasibility of the  Physicochemical
     Treatment of Sewage at Low Temperatures-', Proc. Intl.  Symp., Research
     S Trtmt. of; Wastewaters in Cold Climates, Univ. STsFatchewan at Saskatoon,
     Canada, Aug.

Maran, E. & Sulik, D.J., 1977,  "Advanced Waste Treatment  Design Protects
     Lake Alton", Water & Sewage Works, April,, pp. 68-69.

Martin, R.J. &  Iwugo, K.O., 19785  "Studies on Residual  Organics in Biologi-
     cal Plant  Effluents and Their Treatment by the Activated  Carbon Adsorp-
     tion Process", presented at Syir.p.  on Activated Carbon Adsorption  of
     Organics From the Aqueous Phase, Miami, F1a.s Sept., 1978.  Am. Chetn.
     Soc., Washington, D.C.


                                     307

-------
Masschelein, W.J.. 1979, "Belgian Experiences in  Practical  Applications  of
     Adsorption Techniques in Treating Drinking Water",  presented  at  NATO/CCMS
     Conf. on Adsorption Techniques 1n Drinking Water Treatment, Reston,
     VA., Apr. 30-May 2.  U.S. EPA, Office of Drinking Water,  Washington,
     DC.

Matsuoka, H., 1973, "Ozone Treatment of Industrial  Wastewater",  PPM 4(10):-
     57-69.

Matthews, J.E., 1978, "Treatment of Petroleum Refinery,  Petrochemical  and
     Combined Industrial-Municipal Wastewaters With Activated  Carbon  —
     Literature Review", U.S. EPA Report No.  EPA-600/2-78-200, Robt.  S.  Kerr
     Environmental Research Lab., U.S. EPA, Ada,  Oklahoma   74820.

Mattson, J.S., Mark, H.B., Jr. & Weber, W.J., Jr.,  1969,  "Identification of
     Surface Functional Groups on Active Carbon by  Infrared Internal  Reflec-
     tion Spectrophotometry", Anal. Chem. 41:355-358.

Mattson, J.S., Mark, H.B., Jr., Malbin, M.D., Weber, W.J.,  Jr. & Crlttenden,
     J.C., 1969, "Surface Chemistry of Active Carbon: Specific Adsorption  of
     Phenols", J. Colloid & Interface Science 31(1):250-264.

McCarty, P.L., Reinhard, M. & Argo, D.3., 1977, "Organics  Removal  by  Advanced
     Wastewater Treatment", presented at the  Annual Conf.,  Am. Water  Works
     Assoc. Anaheim, Calif., May, 1977.  AWWA, Denver, Colo.

McCarty, P.L., Argo, D. & Reinhard, M.» 1979, "Operational  Experiences With
     Activated Carbon Adsorbers at Water Factory  21", J.  Am. Water Works
     Assoc. 71(ll):683-689.

McCrodden, B.A., 1974, "Operation of a Refinery Wastewater Treatment  Plant
     Using Filtration of Carbon Adsorption",  in Prpc. Physical-Chemical
     Treatment - Activated Carbon Adsorption  in Pollution  Control, Seminar
     held in Ottawa, Ontario, Canada, Oct. 247^974.  Environmental Protection
     Service, Environment Canada, Ottawa.

McCrodden, B.A., 1979, "Treatment of Refinery Wastewater Using A Filtration-
     Activated Carbon System", U.S. EPA Report EPA-600/2-79-C66, Robt. A.
     Kerr Environmental Research Lab., U.S. EPA,  Ada, Oklahoma  74820.

McGuire, M.J., 1978, "Feasibility Analysis and Implementation  of Synthetic
     Organic Chemical Control Strategies", presented at  Symp.  on Activated
     Carbon Adsorption of Organics From the Aqueous Phase,  Miami,  Fla.,
     Sept., 1978.  Am. Chem. Soc., Washington, D.C.

McGuire, M.J. & Suffet, I.H., 1979, "Adsorption of  Organics From Domestic
     Water Supplies", J. Am. Water Works Assoc. 70(9):621-636.

Meijers, A.P., 1979, "The Use of Carbon Filtration  in The  Netherlands",
     presented at NATO/CCMS Conf. on Adsorption Techniques  in  Drinking Water
     Treatment, Reston, Va., Apr. 30-May 2.  U.S. EPA, Office  of Drinking
     Water, Washington, D.C.


                                     308

-------
Meljers, A.P., Rook, J.J.,  Schultinkj,  B.8  Smeenk,  J.3.M.M.,  van  der  Laan,  J.
     & Poels, C.L.M., 1979, "Objectives and Procedures  for SAC Treatment",
     J. Am. Water Works Assoc.  71(11):628-637.

Merk, W.0 1979, "Prediction of  Multicomponent Adsorption  Behavior  in Activa-
     ted Carbon Adsorbers -- Kinetic Aspects",  presented  at  NATO/CCMS Conf.
     on Adsorption Techniques in Drinking  Water Treatment, Reston9 VA.,  Apr.
     30-May 2.  U.S. EPA, Office of Drinking Water,  Washington,  D.C.

Miller, G.W. & Rice, R.G.,  19789 "European Water Treatment Practices —  The
     Promise of Biological  Activated Carbon", Civil  Engrg.0  Feb.,  81-83.

Miller, 3.W., Rice R.G., & Robsonc  C.M., 1978,  "Large Scale  Applications  of
     Activated Carbon With Ozone Pretreatment", presented at Symp. on
     Activated Carbon Adsorption of Organics From the Aqueous Phase, Miami,
     Fla., Sept., 1978,  Am. Chem.  Soc., Washington, D.C.

Miller, R., 1979, "Treatment of Ohio River Water", presented at  NATO/CCMS
     Conf. on Adsorption Techniques in Drinking Water Treatment, Reston,  VA,
     Apr. 30-May 2.  U.S. EPA,  Office  of Drinking Water,  Washington, CC.

Miller, S., 1980, "Adsorption on Carbon:  Solvent Effects on Adsorption",
     Environ. Sci. & Techno!. 14(9):1037-1049.

Miltner, R.J., 1979, "Results for Chio River Valley Water Sanitation Corrmis-
     sion (ORSANCO) Studies", presented at NATO/CCMS Conf. on Adsorption
     Techniques in Drinking Water Treatment, Reston9 VA,  Apr. 30-May 2.
     U.S. EPA, Office of Drinking Water, Washington, DC.

Mizumoto, K. & Horie, M.9 1974, "Dyeing Wastewater Treatment by  Combination
     of Ozone and Activated Carbon", Japan Textile News 89:238.

Morris, C.C. & Weber, W.J., Jr., 1966, "Adsorption of Biochemically  Resis-
     tant Materials From Solution.   Adv. Waste Trtmt. Report-16.  U.S.
     Environmental Protection Agency,  Municipal Environmental Research Lab.,
     Cincinnati, Ohio.

Morrison, T.J. & Edwards, L.L., 1978,  "Ozonation of Water for Salmonid Fish
     Rearing Facilities.  Pilot Plant  Results", presented at Intl. Ozone
     Assoc. Workshop on Aquatic Applicns.  of OzoneB  Orlando, Florida, Nov.,
     1978.  Intl. Ozone Assoc., Vienna. VA.

Moss, W.H., Schade, R.E., Sebesta9  S.J., Scheutzowf  K.A., Beck P.V.  & Gerson,
     D.B., 19779 "Full-Scale Use of Physical/Chemical Treatment  of :omestic
     Wastewater at Rocky River, Chio", J.  Water Poll, Control Fed. 49(11 ):2249-
     2254.

Mueller, J.A. & Mancini, J.L.,   1975S "Anaerobic Filter - Kinetics  and
     Application",  in Proc. 30th Purdue UrrN. Indl.  Waste Conf., ^ay 6-8,
     1975.  Ann Arbor Science Publishers, Inc., Ann Arbor,"RTchigan, pp.
     423-447.
                                     309

-------
MUller, G. & Bernhardt, H.,  1976,  "Comparative Bacteriological  Investiga-
     tions on an Experimental  Activated Carbon Filter Used  For  Chlorinated
     and Ozonized Groundwater  and  Reservoir Water",  Forum,  Umwelt,  Hygiene
     2(27):393-396.

Murin, C.J. & Snoeyink, V.L.,  1979,  "Competitive Adsorption of  2,4-Dichloro-
     phenol and 2,4,6-Trichlorophenol  in the Nanomolar to Micromolar Concentra-
     tion Range", Environ. Sci.  Techno!. 13(3):305-311.

Myers, L.H., 1976, "Pilot Plant Activated Carbon Treatment  of Petroleum
     Refinery Wastewater", in  Proc.  Open Forum on Management of Petroleum
     Refinery Wastewaters, F.S.  Manning, editor, 26-29 Jan.  U.S.  EPA,  Robt.
     S. Kerr Environinental Research  Labs., Ada,  OK.

National Research Council, 1979, "An Evaluation  of Activated Carbon For
     Drinking Water Treatment",  NRC, Washington, DC.   Natl. Tech.  Info.
     Service, Springfield, VA, Report No. PB-299.968/8WP.

O'Brien, J.E. & Alsentzer, H.A., 1977, "Demonstration Plant for the Treat-
     ment of Wastewaters From  Tank Truck Cleanings",  AIChE  Symp.  Series
     73(167): 189-192, Water-1976.  Am. Inst. Chem.  Engrs., New vork, N.Y.

Oehler, K.E., 1977, "Development of  a Method for Chemical Cxidation During
     Processing of Surface Waters",  presented at Activated  Carbon  Conference,
     Oct., 1977, Karlsruhe,  Federal  Republic of Germany.  Engler-Bunte  Inst.
     der Univ. Karlsruhe.

O'Farrell, T.P. & Menke, R.A., 1978, "Operational Results for the  Piscataway
     Model 5 MGD AWT Plant". U.S.  EPA Report No. EPA-600/2-78-172,  Sept.,
     U.S. EPA, Cincinnati, Ohio, 91  pp.

Osborne, D.J., 1979a, "Activated Carbon Treatment",  printed in  On-Stream,
     No. 21, Winter Issue.  Humphries & Glasgow, Ltd., 22 Carlisle Place,
     London SW1P UA, England.

Osborne, D.J., 1979b, "Experience  With Multi-Hearth  Furnace at  Wilne Treat-
     ment Works, Nottingham, U.K.",  presented at NATO/CCMS  Conf.  on Adsorption
     Techniques in Drinking Water Treatment, Reston,  VA, Apr. 30-May 2.
     U.S. EPA, Office of Drinking  Water, Washington,  DC.

Otte, G. & Rosenthal, H., 1978,  "Water Quality During a One Year Operation
     of a Closed, Intensive Fish Culture System", presented at  Intl. Ozone
     Assoc. Workshop on Aquatic Applications of Ozone, Orlando, Florida,
     Nov. 1978.  Intl. Ozone Assoc., Vienna, VA.

Oulman, C.S., 1980, "The Logistic  Curve as a Model ror Carbon Bed  Design",
     J. Am. Water Works Assoc. 72(l):50-53.

Pahren, H.R. & Melton, R.G., 1979, "Mutagenic Activity and  Trace Organics  in
     Concentrates From Advanced Wastewater Treatment Plant  Effluents",  in
     Proc. Water Reuse Symp.,  Washington, DC, 25-30  March.   Am. Water Works
     Assoc. Research Foundation, Denver, CO, p.  2170-2186.


                                     310

-------
Parkhurst, J.D., Dryden,  F.3.,  McDermott,  G.N.  & English,  J.,  1967,  "Poirona
     Activated Carbon Pilot Plant",  J.  Water Poll.  Control  Fed.  35:R70.

Paulson, E.3., 1969, "Adsorption as  a Treatment of  Refinery Erfluent",
     Calgon Corp. Report.  No.  C-623  for CCRW Subcommittee  of Chemical
     Wastes, Am. Petroleum Inst., Washington, D.C.

Peel, R.G. & Benedek, A.,  1980a, "Attainment of Equilibrium in Activated
     Carbon Isotherm Studies",  Env.  Sci.  & Techno!. 14(1):66-71.

Peel, R.G. & Benedek, A.,  1980b, "Dual  Rate Kinetic Model  for Activated
     Carbon Adsorption, Part I.  Prediction of Column Performance Using  the
     Surface Diffusion Model",  accepted for publication in J.  An. Soc. Civil
     Engrs.  Am. Soc. Civil Engrs.,  New York, N.Y.

Peel, R.G. & Benedek, A.,  1980c, "Dual  Rate Kinetic Model  for Activated
     Carbon Adsorption, Part II.  A  Simplified -rining Force Model",  accepted
     for publication in J. Am.  Soc.  Civil  Engrs. Am. Soc.  Civil  Engrs.,  New
     York, N.Y.

Reel, R.3. and Benedek, A., 1980d, "Dual  Rate Kinetic Model of Fixed  Bed
     Adsorber", J. Env. Engrg.  DiVo, Arr.  Soc. Civil Engrs.  EE4:797-813.

Pendygraft, G.W., Schlegel, F.E. & Huston, M.J., 1979, "The EPA-Proposed
     Granular Activated Carbon Treatment Requirement:  Panacea or Pandora's
     Box?"  J. Am. Water Works Assoc. 71(2):52-60.

Perrotti, A.E. & Rodman,  C.A.,  19745 "Factors Involved With Biological
     Regeneration of Activated Carbon", Water -- 1974, 3.F. Bennett,  Ed.,
     AIChE Symp. Series 7C(144):317-325,   Am. Inst. Chem.  Engrs., New York,
     N.Y.

Perry, D.L., Smith, O.K.  & Lynch9 S.C., 1980,, "Development of Basic  Data  and
     Knowledge Regarding Organic Reiroval  Capabilities of Commercially Avail-
     able Home Water Treatment Units Utilizing Activated Carbon;  Final
     Report, Phase 2", U.S. EPA, Office of Water Supply, Washington,  DC.

Peyton, G.R., Burleson, C., Huang, F., Lin, S. & Glaze, W., 1979, "Treatment
     of Potential Drinking Water Supply Pollutants  With Czone and Ozone/UV",
     Presented at 4th World Ozone Congress, Houston, Texas, Nov.  26-29.
     Intl. Ozone Assoc.t Vienna, VA.

Piet, G. & Morra, C.F., 1979,  "Behavior of Hicropollutants in River  Water
     During Bank Filtration",  in Oxidation Techniques in Drinking Water
     Treatment, W. KUhn & H. Sontheimer, editors.  J.S. EPA Report No.  EPA-
     570/9-79-020.  U.S. EPA, Office of Drinking Water, Washington,  DC,  p.
     608-619.

Pitkat, C.A. & Berndt, C.L., 198C, "Textile Waste Treatment at a Municipal
     PACT Facility", presented at 35th Annual Purdue Indl. Waste Conf.,  12-
     14 May.  Purdue Univ., School of Civil Engrg., Lafayette, IN.
                                     311

-------
Poels, C.L.M., undated,  but after 1977,  "Toxicologlcal  Quality of  Rhine
     Water and Implications for Water Treatment  Processes",  KIWA,  Rijswijk,
     The Netherlands.

Poon, C.P.C. & Shunney,  E.L.,  1973,  "Have  A  Space  Problem?", Water &  Wastes
     Engrg. 10(3):B-2-B-6.

Popper, K., Camirand,  W.M., Williams, G.S. & Mecchi,  E.P.,  1978,  "Regenera-
     tion of Spent Activated Carbon  With Formaldehyde", U.S. Patent,  4,073,747,
     Feb. 14, 1978.

Prengle, H.W., Jr., Mauk C.E.  & Payne, J.E., 1977,  "Ozone/UV Oxidation of
     Chlorinated Coirpounds  in  Water", in Forum on_ Ozone Disinfection, E.G.
     Fochtman, R.G. Rice &  M.E. Browning,  editors.Intl. Ozone Assoc.,
     Vienna, VA, pp. 286-295.

Prober, R., Hanna, Y.A.  & Guirguis,  W.A.,  1977,  "Toward a Model for Activated
     Carbon Treatment in the Presence of Significant  Bacterial Srowth".
     Presented at WWEMA  Meeting on Wastewater Treatment, Atlanta,  3eorgia,
     April, 1977.

Prosche, M.A., 1976, "Activated Carbon Treatment of Combined Storm and
     Process Waters",  in Proc. Open  Forum  on Management of  Petroleum  Refinery
     Wastewaters, F.S. Manning,"editor,  26^79 JanT  07S~.TPA, Robt. S. Kerr
     Environmental Research Labs., Ada,  OK,  p. 399-410.

Randtke, S.J., 1978, "Discussion of  Guirguis, Cooper, Harris & Ungar,
     1978", 0. Water Poll.  Control Fed.  50(11)-.2602-2603.

Reimers, R.S., Englande, A.O.  & Miles, H.B., 1976,  "A Quick Method for
     Evaluating the Suitability of Activated Carbon Adsorption for Waste-
     waters", Proc. 31st Indl. Waste Conf.,  Purdue Univ., May 4-6, 1976.
     Ann Arbor Science Publishers Inc.,  Ann  Arbor,  Michigan, pp.  395-408.

Reynolds, T.D. & Hawkins, S.J., 1980, "Water Treatment Project:   Observations
     on Use of GAC in Practice", U.S. EPA  Report No.  EPA-6CO/1-80-023, U.S.
     EPA, Health Effects Research Lab.,  Cincinnati, Ohio.

Rice, R.3., Miller, 3.W., Robson, C.M. & KUhn, W.,  1977,  "Biological  Activa-
     ted Carbon", presented at Intl. Symp. on Advanced Ozone Technology,
     Toronto, Ontario, Canada, Nov., 1977.   Intl.  Ozone Assoc., Vienna, VA.

Rice, R.G., Miller, G.W. &  Robson, C.M., 1978a,  "Potentials of Biological
     Activated Carbon for the Treatment  of Industrial Wastewaters", presented
     at WWEMA Annual Indl.  Poll. Control Symp.,  St. Louis,  Mo., April, 1978.
     Water & Wastewater  Equipment Mfgrs. Assoc., McLean, Va.

Rice, R.G., Miller G.W.  & Robson, C.M.,  1978b, "European Drinking  Water
     Treatment Practices",  presented at  the  Annual  Meeting  of the  Natl.
     Assoc. of Water Cos.,  Baton Rouge,  La., Oct.,  1978.  Natl. Assoc. of
     Water Cos., Washington, D.C.
                                     312

-------
Rice, R.G., 1979a, "Biological  Activated  Carbon"9  presented  at  EPA  Seminar
     on Control  of Organic Chemical  Contaminants  in  Drinking Water,  Dallas,
     Texas, March, 1979.   Public 7echnology8  Inc., Washington,  D,C.

Rice, R.G., 1979b, "Biological  Activated  Carbon  -  A  Status Report",  presen-
     ted at 4th  World Ozone Congress,  Poustons Texas,  Nov. 26-29.   Intl.
     Ozone Assoc. s, Vienna, VA.

Rice, R.3., 1980,  "Ozone  Gives  Boost To Activated  Carbon", Water &  Sewage
     Works, April, pp. 40-82.

Rice, R.G., Robson, C.M.  & Miller, G.W.,  1981, "Biological Activated Carbon
     and its Potentials for Treating Industrial  Wastewaters", Final  Report
     of Grant No.  R-804385-01  to Public Technology,  Inc.  U.S.  EPA,  Office
     of Environmental Engrg, &  Techno!.,  Washington, DC.

Richard, Y., 1972, "Experiment  on the  Industrial Treatment of Drinking  Water
     by Activated  Carbon", Degr€mont S.G.E.A., Rueil Malmaison, France,  5
     Dec.

Richard, Y. & Fiessinger0 F.9  undated9 but after 1972,  "Le Traitement
     Industrie! le  Ces Eaux Potables  par le Charbon Actif", CegrSmont,
     S.G.E.A., Rueil Malmaison, France.

Richard, Y. & Blue, P., 1978,,  "Ozone Pretreatment  of Drinking Water",
     presented at  101 Ozone Technology Symposium,  Los  Angeles,  Calif.,  23-25
     May.  Intl. Ozone Assoc.,  Vienna0 VA.

Richard, Y., Brener, L. & Leblanc, C.9 1979,,  "Optimization of Potable
     Water Treatment Lines With A View to Halogenous Compound Reduction",
     Presented at  4th World Ozone Congress, Houston, Texas,  Nov. 26-29.
     Intl. Ozone Assoc.,  Vienna, VA.

Rizzo, J.A., 1976a, "Activated  Carbon  Clears  Effluent",  Oil  & Gas J. 74(22):-
     52-56.

Rizzo, J.A., 1976b, "Case History:   Use of Powdered  Activated Carbon in an
     Activated Sludge System",  in Proc. Open  Forum on  Management of Petroleum
     Refinery Wastewaters, F.S. Manning,  editor, 26^?9 Jan.   U.S.TPA,  Robt.
     S. Kerr Environmental Research  Labs., Ada,  OKD  p.  359-374.

Robeck, 3.G., 1978, "Progress Reports  July 1  to  Sept.  30,  1978,  EPA Drinking
     Water Research Activities, p.  1.   U.S. EPA, Water Supply Research  Lab,
     Cincinnati, Ohio 45268.

Robertaccio, F.L., Mutton, O.G., Grulich, 5.  & Goltzer,  H.L., 1972, "Treat-
     ment of Organic Chemical Plant Wastewater with  the DuPont  PACT Process",
     presented at AIChE Nat!. Meeting, Dallas, AIChE Symp.  Series  No. 125,
     Vol. 65.
                                     313

-------
Roberts, P., Gujer, W.  & Eugster,  J.,  1977,  "Re1n1gung  von  Kommunalem
     Abwasser Mittels Aktivkohle Nach  Schwach  Belasteter Biologischer
     Reinigung und Filtration,  Vom Wasser 48:47-70.

Rodman, C.A. & Shunney, E.L., 1970, "A New Concept For  The  Biological
     Treatment of Textile Finishing Waters", Water — 1970,  Chem.  Engrg.
     Prog. Symp. 67(107):451-457.

Rodman, C.A., 1971, "Bio-Regenerated Activated Carbon Treatment of Textile
     Dye Wastewater", EPA Report No. 12090 DWM.   U.S. Environmental  Protection
     Agency, Indl. Environ.  Research Lab., Cincinnati,  Ohio  45268.

Rodman, C.A. & Shunney, E.L., 1971, "New Concepts For "Treating  Coloured
     Organic Contaminated Wastewater", Presented  at Symp. on Environ.
     Engrg. Aspects of  Pollution Control, London, England,  June 22-23, Soc.
     Environmental Engrs., London.

Rom, D., Wachs, A.M. &  Rotel, M.,  1980, "Pilot Plant Studies of Water  Renova-
     tion in a System Combining Ozonation With Activated Carbon Treatment",
     presented at 53rd  Annual WPCF Conf., Las  Vegas, Nev.,  1 Get.   Water
     Poll. Control Fed., Washington, DC.

Rosenthal, H. fit Sander, E.,  1975,  "An  Improved Aeration Method  Combined
     Waste Foam Renoval in a Seawater  Recycling System", Intl.  Council for
     the Exploration of the Sea, Mariculture Committee, E:14, 1-16 (Fisheries
     Improvement Committee),

Rosenthal, H., von Westernhagen, H. &  Otte,  G.,  1978, "Maintaining Water
     Quality in Laboratory Scale Sea Water Recycling Systems",  Intl.  Council
     for the Exploration of the Sea, Mariculture Committee  Report, C.M.
     1978/F:10.

Rosenthal, H. & Otte, G., 1979, "Ozonation in  an Intensive  Fish Culture
     Recycling System", Ozone Sci. & Engrg.  1(4):319-327.

Sander, R., 1977, "Formation and Removal of Chlorinated By-Products in a
     Pilot Plant Unit", presented at Seminar on Activated Carbon,  Karlsruhe,
     Federal Republic of Germany,  Oct. 30-31,  1977.  Engler-Bunte  Inst.  der
     Univ. Karlsruhe.

Sander, R.: "979, "Effect of Prechlorination on Activated Carbon Adsorp-
     tion", presented at NATO/CCMS Conf. on Adsorption  Techniques  in  Drinking
     Water Treatment, Reston, VA,  Apr. 30-May  2.   U.S.  EPA, Office of
     Drinking Water, Washington, DC.

Savage, P.R., 1979, "Waste Disposal With An Energy Bonus",  Chem. Engrg.,  vay
     21, p. 116-117.

Scaramelli, A.B. & DiGiano, F.A.,  1970, "Upgrading the  Activated Sludge
     System by Addition of Powdered Carbon", Water & Sewage Works  120(9):90-94.
                                     314

-------
Schalekamp, M.  & Bakker,  S.P.,  1978,  "Use and  Thermal  Regeneration  of
     Activated  Carbon in  Switzerland1,  Effluent &  Water  Trtmt.  J. 18(1):28-32.

Schalekamp, M., 1979, "The Use  of GAC Filtration to  Ensure  Quality  in
     Drinking Water From  Surface Sources", C.  Am.  Water  Works Assoc. 71(11):-
     638-647.

Scherm, M.  & Lawson, C.T., 1977, "Pilot Demonstration  of Renovation and
     Reuse of Wastewaters From  Organic  Chemical  Manufacturing',  Indl.  Water
     Engrg., Oct./Nov. issue, p. 16-22.

Schulhof, P., 1979, "An Evolutionary  Approach  to Activated  Carbon Treatment",
     J. Am. Water Works Assoc.  71(ll):648-659.

Schuliger, W.G., 1974, "Equipment Design Considerations",  in  Proc.  Physical-
     Chemical Treatment Activated Carbon Adsorption  in Pollution Control,
     Seminar held in Ottawa,  Ontario, Canada,  Oct.  24T 1974.Environmental
     Protection Service,  Environment  Canada, Ottawa.

Schuliger, W.G., 1978, "Purification  of Industrial  Liquids  With Sranular
     Activated  Carbon: Techniques For Obtaining and  Interpreting "ata  and
     Selecting  the Type of Commercial System",  in  Carbon Adsorption Handbook,
     P.N. Chereirisinoff & F.  Ellerbusch, editors,  op.  cit., pp.  55-84.

Semmens, M.J. & Goodrich, R.R., Jr.,  1977, "Biological Regeneration of
     Ainnoniurn-Saturated Clinoptilolite. I. Initial  Observations", Environ.
     Scu & Techno!. 11(3):255-265.

Semmens, M.J.,  1977V "The Feasibility of Using Nitrifying Bacteria  To
     Assist The Regeneration  of Cl InoptilolUe", in  Proc.  32nd  Purdue  Univ.
     Indl. Waste Conf., May 10-12, 1977.  Ann  Arbor Science Publishers,
     Inc., Ann  Arbor, Michigan, pp.  733-744.

Semrens, M.J.,  Wang, J.T. & Booth,, A.C., 1977,  "Nitrogen Removal by Ion
     Exchange:  Biological Regeneration of Clinoptilolite",  J. Water Poll.
     Control Fed. 49(12)=2431-2444.

Semmens, M.J. & Porter, P.S., 1979,  "Ammonium  Removal  by Ion  Exchange:
     Using Biologically Restored Regenerant",  J. Water Poll.  Control Fed.
     51(12):2928-2940.

Shelby, S.E., Koon, J.H., Marks, D.R. & Scott, H.A., Jr., 1980, "Adsorption
     of Chlorinated and Non-Chlorinated Organics From a Pesticide Manufactur-
     ing Plant Waste Stream", presented at 53rd Nat!.  Conf. WPCF, Las  Vegas,
     Nev., 1 Oct.  Water Poll.  Control  Fed., Washington, DC.

Shanra, B. & Ahler, R.C., 1977, "Nitrification and Nitrogen Removal",
     Water Research 11:897-925.

Shuckrow, A.J., Bonner, W.F., Presecan, N.L.  & Kazmierczak, E.J.,  1972,  "A
     Pilot Study of the Physical-Chemical Treatment of Raw Wastewater  at the
     Westerly Plant in Cleveland, Ohio", Water Research 6:619.


                                     315

-------
Shunney, A.L., Perotti, A.E.  & Rodman,  C.A.,  1971,  "Decolorization  of
     Carpet Yarn Dye Wastewaters",  Am.  Dyestuff Reporter  60(6):32,  34-36,
     38, 40.

Siemak, R.C., Trussell, R.R., Trussell, A.R.  & Umphres, M.D.,  1979,  "How  to
     Reduce Trihalomethanes in Drinking Water", Civil  Engrg.,  Feb.  1979,  pp.
     49-51.

Skovronek, H.S., Dick, M. & des Rosiers, P.E., 1977,  "Selected Uses  of
     Activated Carbon for Industrial  Wastewater Pollution Control",  Indl.
     Water Engrg., 14(3):6-13.

Skovronek, H.S. & Becker, D.L., 1977, "Pollution Control  by Adsorption",
     presented at 7Cth Annual Meeting of the  Am. Inst.  Chenu  Engrs., New
     York, N.Y., Nov. 13-17,  1977.   AIChE, New York,  N.Y.

Skovronek, H.S., 1978, "Industrial  Case Histories",  in  Carbon  Adsorption
     Handbook, P.N. Cheremisinoff & F.  Ellerbusch,  editors, op.  cit.,~w>.
     85-130.

Slade, J.S., 1978, "Enteroviruses in Slow Sand Filtered Water",  J.  Inst.
     Water Engrs. & Scientists 32(6).-530-536.

Smith, O.K., Lynch, S.C., Gebhart,  J.E. & Monteith,  C.S., 1979,  "Development
     of Basic Data and Knowledge Regarding Organic  Removal  Capabilities  of
     Commercially Available Home Water Units  Utilizing  Activated Carbon  -
     Preliminary Report - Phase I", U.S. Environmental  Protection Agency,
     Office of Water Supply,  Washington, DC.

Smith, S.A., Chapman, R.L. & Butterfield, O.R., 1979,  "Tahoe-Truckee Water
     Reclamation Plant First Year in Review",  Proc.  Water Reuse Symp.,
     Washington, D.C., Mar. 25-30,  pp.  1435-1445.  Am.  Water Works  Assoc.
     Research Foundation, Denver, Colo.

Smith, S.B., 1974, "Techniques of Activated Carbon  Regeneration", in Proc.
     Physical-Chemical Treatment Activated Carbon Adsorption in Pollution
     Control, Seminar held in Ottawa, Ontario, Canada,  Oct. 237 1974.
     Environmental Protection Service, Environment  Canada,  Ottawa.

Smithson, G.R., 1978, "Regeneration of Activated Carbon:  Thermal, Chemical,
     Solvent, Vacuum and Miscellaneous Regeneration Techniques", in Carbon
     Adsorption Handbook, P.N. Cheremisinoff & F. Ellerbusch,  editors,  op.
     cTtT8 pp. 379-904.

Snoeyink, V.L., McCreary, J,J. & Murin, C.J.,  1977,  "Activated Carbon
     Adsorption of Trace Organic Compounds",  U.S. EPA Report No. 6CO/2-77-
     223, U.S. EPA, Indl. Environ.  Research Lab., Cincinnati,  Ohio   45268.

Snyder, A.J. & Alspaugh, T.A., 1974, 'Catalyzed Bio-Cxidation  and Tertiary
     Treatment of Integrated Textile Wastewaters",  U.S.  EPA Report  No.
     660/2-74-C39, June, 1974.  U.S. EPA, Indl. Environ.  Research Lab.,
     Cincinnati, Ohio  45268.


                                     316

-------
Sontheimer, H., 1975a, "The Importance of Adsorption Processes  1n  drinking
     Water Treatment", 1n Translation pjf Reports on Special  Problens  of_
     Water Technology, Vol. 9 - Adsorption,  H.  Sontheimer,  editor,  op_.  ctt.,
     pp. 1-T5-;

Sontheimer, H., 1975b, "Basic Principles  of Adsorption  Process Techniques",
     1 n Translation of. Reports on_ Special  Proble-ns  of_ Water Techology,  Vol.
     9_, op_. cit., p.  29-66.

Sontheimer, H., 1975d, "Realistic Laboratory Test Methods  for the  Evaluation
     of Activated Carbon", in Translation of Reports on  Special  Problems  of_
     Water Techno!ogy, Vol. 9 - AdsorptTon7~H~SontheTm'erSi  editor,  op.  cit.,
     ppTT50-268.

Sontheimers H.9 1975e, "Theory and Practice  in  the  Use of  Adsorption  Proces-
     ses", in Translation of Reports  pj^ Special  Problems of Water  Technology,
     Vol. £- Adsorption, H7 Sontheiner. editor,~op cit.,  pp. 414-417.

Sontheimer, H., Wtilfel, P. & Safert,  F., 1977,  "Verbesserung der Biologis-
     chen Abbaubarkeit der Organischen Stoffe in Biologlsch Gereinigten
     Abwflssern Durch Fine Ozonbehandlung", presented at  3rd Intl.  Symp. or
     Ozone Technology, Paris, France, May.  Intl. Czone  Assoc., Vienna, VA.

Sontheirer, H., 1978a, 'Biological Treatment of Surface  Waters  in  Activated
     Carbon Filters", OZONews, July,  1978, Part 2 - Technical Paper Section.
     Intl. Ozone Assoc.,  Cleveland,, Ohio.

Sontheiner, H., 1970a, "Process Engineering  Aspects in the Combination  of
     Chemical and Biological Oxidation", in  Oxidation Techniques in Drinking
     Water Treatment, W.  KOhn a H. Sontheimer,  editoFs.  U.S. EPATeport  Nol
     EPA-570/9-79-020.  U.S. EPA, Office of Drinking Water, Washington, pp.
     702-714.

Sontheimer, H., 1979b, "Biologisch-Adsorptive Trinkwasseraufbereitung in
     Aktivkohlefiltern -- Das VUlhelmer Verfahren", Rhe1n1sch-Westf3l1sche
     Wasserwerksgesellschaft mbH (MUiheirr a.d.  Ruhr) & DVGW-Forschungsstelle
     am Engler-Bunte Institut der University Karlsruhe, Feb.

Sontheimer, H., 1979c, "German Experience in Activated Carbon Treatment",
     presented at NA70/CCMS Conf. on  Adsorption Techniques in Drinking  Water
     Treatment, Reston, VA, Apr. 3C-May 2.  U.S. EPA, Office of Drinking
     Water, Washington, DC.

Sontheimer, H., 1979d, "Applying Oxidation and Adsorption  Techniques:  A
     Summary of Progress", J. Am. Water Works Assoc. 71(11):612-617.

Sontheimer, H., 1979e, "Design Criteria and Process Schemes for 3AC Filters",
     J. Am. Water Works Assoc. 71 (11).-618-622.

Spahn, H., Brauch, V., Schlunder, E.U. & Sontheimer, H., 1974,  "Auslegung
     von Aktivkohlefiltern zur Wasserreinigung.   Tell I:  Untersuchung  der
     Adsorption am Eizelkorn", Verfahrenstechnik 8(8), 8 pp.


                                     317

-------
Spahn, H., Brauch, V., SchlUnder,  E.U.  & Sonthelmer,  H.,  1975,  "Auslegung
     von Aktivkohlefiltern zur Wasserreinlgung.   Tell  II.  Theoretische und
     Experimented Bestimmung der Beladungsfelder In Aktlvkohlefestbetten",
     Verfahrenstechnlk 9(1),  5 pp.

Stephenson, P., 1979, "The Effect of Ozone on the BiodegradabHity  of
     Refractory Organics In Water", M.  Engr.  Thesis,  McMaster Univ., Hamilton,
     Ontario, Canada.

Stevens, A.A., Seeger, D.R.,  DeMarco, J. & Moore, L., 1979,  "Removal of
     Higher Molecular Weight Organic Compounds by the Granular  Activated
     Carbon Adsorption Unit Process", presented at NATO/CCMS Conf.  on
     Adsorption Techniques 1n Drinking  Water  Treatment,  Reston,  VA, Apr.  30-
     May 2.  U.S. EPA, Office of Drinking Water, Washington, DC.

Stewart, D.R. & Sierka, R.A., 1979, "Effects  of Preozonatlon of Domestic
     Wastewater on Activated Carbon Adsorption", presented at 52nd  Annual
     WPCF Conf., Houston, TX.  Water Poll. Control  Fed.,  Washington CC.

Strack, B., 1975, "Operation, Problems  and Economy of Activated Carbon
     Regeneration", in Translation of Reports on Special  Problems of Water
     Technology, Vol. 9 - Adsorption, 0£_. at., pp. 284-311.

Suffett, I.H., Brenner, L., Coyle, O.T. & Cairo, P.R., 1978, "Evaluation  of
     the Capability of Granular Activated Carbon and XAD-2 Resin to Remove
     Trace Organlcs From Treated Drinking Water", Env. Sci.  & Techno!.
     12(12):

Suffett, I,H.S McGuire, M.J., Josephson, 0. & Ember,  L.R., 1978,  "Cleanup:
     That Cld Black Magic Works Again!11, Environ. Sc1. & Technol. 12(10):-
     1138-1149.

Suffet, I.H., 1980, "An Evaluation of Activated Carbon for Drinking Water
     Treatment:  A Report From the National Academy of Science",  J. Am.
     Water Works Assoc. 72(1):41-50.

Suidan, M.T., Cross, W.H. & Khan, K., 1979, "Treatment of Phenolic  Waste-
     water with Activated Carbon Filters.  Technical  Completion Report,
     Sept. 1977 - Sept. 1979.  Natl. Tech. Info. Service, Springfield, VA,
     Report No. FE-2756-T1.

Suzuki, J., 1976, "Study on Ozone Treatment of Water-Soluble Polymers. I.
     Ozone Degradation of Polyethylene  Glycol in Water",  J.  Appl. Polytrer
     Sci. 20:93-103.

Suzuki, J., Nakagawa H. & Ito, H., 1976, "Study on Ozone Treatment  of
     Water-Soluble Polyners. II. Utilization  of Ozonized Polyethylene
     Glycol by Bacteria" J. Appl. Polymer Sci. 20:2791-2797.
                                     318

-------
Suzuki, M.,  Tada,  Y.  & Kawazoe,  K.,  1977,  "Comparison of Treatment Processes
     for Strong Wastewater From  Chemi-Mechanical  Pulp Mill",  Pacific Chem.
     Engrg.  Congr. (PAChEC '77)  2:1322-1329.  Am.  Inst. Chem.  Engrs., New
     York, N.Y.

Symons, J.M.,  1976,  "Summary of  Granular Activated Carbon  Practice 3ata",
     Report dated  Feb. 17, 1976.   U.S.  EPA, Water Supply Research Laboratory,
     Concinnati, Ohio  45268.

Symons, J.M.,  1979,  "Practical Applications of  Adsorption  Techniques 1n
     Drinking  Water Treatment",  presented  at  NATO/CCMS Conf.  on Adsorption
     Techniques in Drinking Water Treatment,  Reston, VA, Apr.  30-May 2.
     U.S. EPA, Office of Drinking Water, Washington, DC.

Symons, J.M.,  Carswell, J.K.,  DeMarco,  J.  & Love,  O.T., Jr.,  1979, "Removal
     of Organic Contaminants From Drinking Water  Using Techniques Cther Than
     Granular  Activated Carbon Alone -  A Progress Report",  Drinking Water
     Research  Div.,  Municipal  Environmental Research Lab.,  Cffice of Research
     & Developments,  U.S. Environmental  Protection Agency,  Cincinnati, Ch1o
     45268.

Taylor, R.H.,  Allen,  M.J. & Geldrelch,  E.E.,  1979, "Testing of Home Use
     Carbon Filters", J. An. Water Works Assoc. 71(10):577-579.

Thibault, 3.T., Steelman, B.L. & Tracy, K.D., 1978, "Enhancement of the
     Refinery  Activated Sludge Process  With Powdered Activated Carbon',
     Presented at  6th Ann. Indl.  Poll.  Control  Conf., WWEMA,  St. Louis, Mo.,
     April 13.  Water & Wastewater Equipment  Mfgrs. Assoc., McLean, VA.

Tien, C., 1978, "Bacterial Growth and Adsorption  1n Granular  Activated
     Carbon Columns", presented  at Symp. on Activated Carbon  Adsorption of
     Organics  From The Aqueous Phase, Miami,  Florida, Sept.  Am. Chen. Soc.,
     Washington, D.C.

Tifft, E.G., Jr. & Tamays, T.A.,  1979,  "Drinking  Water Treatment With
     Ozone/Granular Activated Carbon",  Presented  at 4th World Ozone Congress,
     Houston,  Texas, Nov. 26-29.   Intl. Ozone Assoc., Vienna,  VA.

Timpe, W.G. &  Lang, E.W., 1974,  "Activated Carbon Treatment of Kraft Mill
     Effluent  For  Reuse", Water  — 1973, G.F. Bennett, ed., AIChE Symp.
     Series 70(136):585-591.  Am. Inst. Chem. Engrs., New  York, N.Y.

Tomlinson, T.G. & Snaddon, D.H., 1966,  "Biological Oxidation  of Sewage by
     Films of Microorganisms", Air & Water Poll.  Inst. J.  4:865.

Van Der Kooij, D., 1978, "Processes During Biological Oxidation 1n Filters",
     presented at Intl. Conf. on Oxidation Techniques  in  Drinking Water
     Treatment, Karlsruhe, Federal Republic of Germany, Sept.  11-13,  1978.
     Engler-Bunte Inst. der Univ. Karlsruhe.

Van Leeuwen, J. &  Prinsloo, J.,  1980, "Ozonation  at the Stander Water Reclama-
     tion Plant",  Water SA 6(2):96-102.

                                     319

-------
Van Vuuren, L.R.J., Clayton, A.C.  & Van Der Post,  D.C.,  1978,  "Current
     Status of Water Reclamation at Windhoek",  Presented at  51st Ann. Conf.
     of the Water Poll. Control  Fed.,  Anaheim,  Calif.,  1-6 Oct.   WPCF,
     Washington, D.C.

Von Dreusche, C., Jr., 1978, "Process  Aspects of Regeneration  in a  Multiple
     Hearth Furnace", in Carbon  Adsorption Handbook,  P.H. Cheremisinoff & F.
     Ellerbusch, editors, 0£. cit., pp. 923-954.

Waggott, A. & Bayley, R.W., 1972,  "The Use of Activated Carbon for  Improving
     the Quality of Polished Sewage Effluent",  Water  Pollution Control, p.
     417.

Wallace, R.N. & Burns, 3.E., 1976, "Factors Affecting Powdered Carbon
     Treatment of a Municipal Wastewater", 0. Water Poll. Control.  Fed.
     48(3):511-519.

Waller, 6., 1976S "Granular Activated  Carbon and its  Re-Use  in Water and
     Waste Treatment", Prog. Water Technology 8(2/3):39-45.

Wallis, C., Stagg, C.H. & Melnick, J.L., 1974,  "The Hazards  of Incorporating
     Charcoal Filters Into Domestic Water Systems", Water Research  8:111-113.

Warner, H.P. & English, J.N., 1978, "Wastewater Treatment For  Reuse and its
     Contribution to Water Supplies",  U.S. EPA  Report No. EPA-600/2-78-027,
     March, 40 pp.

Wasserlauf, M., 1974, "Toxicity Removal From Kraft Mill  Effluents by Activa-
     ted Carbon", in Proc. Physical-Chemical Treatment Activated Carbon
     Adsorption jn_ PoTTuTiorf Control,  Seminar held in Ottawa,  Ontario.
     Canada, Oct. 24, 1974.  Environmental Protection Service, Environment
     Canada, Ottawa.

Weber, W.J., Jr. & Morris, J.C., 1964, "Equilibrium and Capacities  for
     Adsorption on Carbon", J. Sanit.  Engr. Div., Am. Soc.  Civil Engrg.
     90:SA3, 79.

Weber, W.J., Jr., 1964, "Competitive Interactions in  Adsorption From Dilute
     Aqueous Bi-Solute Solutions", J.  Appl. Chem. 14:565-572.

Weber, W.J., Jr. & Gould, J.P.,  1966,  "Sorption of Organic  Pesticides From
     Aqueous Solution", in Organic Pesticides lp_ the  Environment, Advances
     in Chemistry Series, No. 60C A,.  Chem. Soc.,~WTshington,  D.C., ppe 280-
     304.

Weber, W.J., Jr. & Keinath, T.M., 1967, "Mass Transfer of Perdurable Pollu-
     tants From Dilute Aqueous Solution in Fluidized  Adsorbers", Chem.
     Engrg. Progress Symp. Series 63(74):79-89.

Weber, W.J., Jr., Hopkins, C.B.  & Bloom, R., Jr., 1970, "Physlco-Chemical
     Treatment of Wastewater", J.  Water Pollution Control Fed. 42:83-99.
                                     320

-------
Weber, W.J. Jr., 1970, "Discussion of High Quality Reuse  Water By  Chemical-
     Physical Wastewater Treatment, J. Water Poll. Control  Fed.  42(3):456-
     463.

Weber, W.J., Jr., Hopkins, C.B.  & Bloom,  R., Jr.,  1971,  "Expanded  Bed
     Adsorption Systems For Treatment of  Sewage Effluents", Water  —  1970,
     AIChE Symp. Series 67(107):541-553.   Am.  Inst.  Chem.  Engrs.,  New  York,
     N.Y.

Weber, W.J., Jr., 1973, "The Prediction of the Performance  of Activated
     Carbon For Water Treatment", Proc. Water Research Assoc.  Conf.,  Reading,
     England, pp. 53-72.

Weber, W.J., Jr., Friedman, L.D.  & Bloom, R., 1973,,  "Biologically  Extended
     Physicochenrical Treatment",  in Proc. 6th Intl.  Conf. £n_ Advances  in
     Water Poll. Research, Jerusalem. June~5^23, 1972, S.H. Jenkins,  eoTtor.
     Pergamon Press, Oxford, England and  New York, N.Y.,  pp. 641-649.

Weber, W.J., Jr. & Crittenden, J.C., 1975, "MADAM I  - A  Numeric Method For
     Design of Adsorption Systems", J. Water Poll. Control  Fed.  47(5):924-
     940.

Weber, W.J., Jr., 1977, "Integrated Biological & Physical-Chemical Treatment
     for Reclamation of Wastewater", Indl. Water Engrg.  14:20-27.

Weber, W.J., Jr., Pirbazari, M.  & Herbert, M.D., 1978, "Removal  of Halogena-
     ted Organic and THM Precursor Compounds From Water  by  Activated  Carbon",
     Environmental & Water Resources Engrg. Dept.8 Univ.  of Michigan,  Ann
     Arbor, Michigan.

Weber, W.J., Jr., 1978, Discussion of "Improved Performance of Activated
     Carbon by Preozonation", by W. Guirguis, T. Cooper,  J. Harris &  A.
     Ungar, J. Water Poll. Control Fed. 50(12):2781-2785.

Weissenhorn, F.J., 1975, "Testing of Activated Carbon Filters in Waterworks",
     in Translation of Reports on Special Problems of_ Water Technology,  Vol.
     9 - Adsorption, H. Sontheimer, editor, op. cit., pp.  238-249,

Werner, P., Klotz, M. & Schweisfurth, R., 1978, "Microbiological Studies  of
     Activated Carbon Filtration", in Oxidation Techniques  in Drinking Water
     Treatment, W. KUhn & H. Sontheimer,  editors, U.S. EPA Report No.  EP"A~^
     570/9-79-020, p. 678-688.

Werner, P., Klotz, M. & Schweisfurth, R., 1979, "Investigations Concerning
     the Microbiology of SAC Filtration for Drinking Water Treatment",
     presented at NATO/CCMS Conf. on Adsorption Techniques  in Drinking Water
     Treatment, Reston, VA, Apr. 30-May 2.  U.S. EPA, Office of Drinking
     Water, Washington, DC.

Wilson, S.C. &  Krantz, G.E., 1978, "Source Listing for the Application of
     Activated Carbon Technology", in Carbon Adsorption  Handbook,  P.N.
     Cheremisinoff &  F. Ellerbusch, editors, op. cit., "pp.  955-1022.


                                      321

-------
Wfllfel, P. & Sontheimer, H.,  1977,  "Amelioration  de  la  Degradation  Blologique
     des Eaux Reslduaires par un Traltment a  VOzone",  presented  at 3rd
     Intl. Symp. on Ozone Techno!., Paris, France, May.   Intl.  Ozone Assoc.,
     Vienna, VA.

Wood, P.R. & DeMarco, J., 1978,  "Effectiveness  of Various Adsorbents In
     Removing Organic Compounds  From Water.  Part I  - Removing  Purgeable
     Halogenated Organics", presented at Symp.  on Activated  Carbon  Adsorption
     of Organics From the Aqueous Phase", 176th ACS  Mtg., I*!1am1 Beach, Fla.,
     10-15 Sept.  Am. Chem. Soc., Washington, DC.

Wood, P.R. & DeMarco, J., 1978,  "Effectiveness  of Various Adsorbents 1n
     Removing Organic Compounds  From Water.  Part II -  Removing Total Organic
     Carbon and Trlhalomethane Precursor Substances", presented at  Synp.  on
     Activated Carbon Adsorption of Organics  From the Aqueous  Phase, 176th
     ACS Mtg., Miami Beach, Fla., 10-15 Sept.   Am. Chem.  Soc.,  Washington,
     DC.

Wood, P.R., 1979, "Treatment of  Contaminated  Ground  Water",  presented at
     NATO/CCMS Conf. on Adsorption Techniques in  Drinking Water Treatment,
     Reston, VA, Apr. 30-May 2.   U.S. EPA, Office of Drinking  Water, Washing-
     ton, DC.

Wood, P.R., Jackson, D.F., Gervers, J.A., Waddell, D.H.  & Kaplan, L., 1980,
     "Removing Potential Organic Carcinogens  and  Precursors  From  Drinking
     Water", U.S. EPA Report No. EPA-6CO/2-80-130a.   U.S. EPA,  Municipal
     Environmental Research Lab., Cincinnati, Ohio.

World Health Organization, 1975, "Health Effects  Relating to Direct and
     Indirect Reuse of Wastewater for Human Consumption", Report  of an  Intl.
     Working Meeting, held at Amsterdam, The  Netherlands, Jan.  13-16, 1975.
     Tech. Paper No. 7.

Ying, W.C. & Weber, W.J., Jr., 1978, "Bio-Physicochemlcal Adsorption Systems
     for Wastewater Treatment: Predictive Modeling for  Design  and Operation",
     in Proc. 33rd Annual Purdue Univ. Indl.  Waste Conf., 9-11  May. Ann
     Arbor Science Publishers, Inc., Ann Arbor, MI,  p.  128-141.

Ying, W.C. & Weber, W.J., Jr., 1979, "Bio-Physicochemical Adsorption Model
     Systens For Wastewater Treatment", J. Water  Poll.  Control  Fed. 51(11):-
     2661-2677.

Zaidi, S.A. & Tollefson, E.L., 1976, "The Physical-Chemical  Treatment of
     Sour 3as Plant Process Wastewaters", J.  Can. J. Petrol. Technol.,
     April-June, 1976, pp. 39-47.

Zanitsch, R.H. & Stenzel„ M.N.,  1978, "Economics  of  Granular Activated
     Carbon Water and Wastewater Treatment Systems", in Carbon  Adsorption
     Handbook, P.N. Cheremislnoff & F. Ellerbusch, editors,  op. cit., pp.
     215-240.
                                     322

-------
Zeff, J.D., Leitis, E.  & Crosby,  D.G.,  1979,  "Study of  the  Chemistry  of  the
     UV-Ozonation of Refractory Organic Compounds  in Water",  Presented at
     4th World Ozone Congress,  Houston, Texas,  Nov.  26-29.   Intl.  Ozone
     Assoc., Vienna, VA.

Zobell, C.E., 1937, "The Influence of Solid Surfaces Upon  the Physiological
     Activities of Bacteria in  Sea Water",  J.  Bact.  33:86.

Zogorski, J.S. & Faust, S.D.,  1978, "Operational  Parameters for Cptimun
     Removal of Phenolic Compounds From Polluted  Waters by  Columns of
     Activated Carbon", in Carbon Adsorption Handbook,  P.N. Cheremisinoff  &
     F. Ellerbusch, editors, op_.  cit.,  pp.  753-778.

Zuckerman, M.M. & Molof, A.M.,  1970,  "High  Quality Reuse Water by  Chemical -
     Physical Wastewater Treatment".  C. Water Poll.  Control Fed. 4Z(3):437-456.
                                     323

-------
                                 APPENDIX A

               AUBERGENVILLE PLANT (SUBURBS OF PARIS, FRANCE)
BACKGROUND
     The Aubergenville water treatment plant of the SociSte" Lyonnaise des
Eaux et de TEclairage (SLEE) produces approximately 100,000 cubic meters of
water per day (26.4 mgd) from 23 wells which are situated along the south
side (left bank) of the Seine River downstream of Paris, between Les Mureaux
and EpOne.

1961 CONSTRUCTION

     The plant was designed to treat groundwater drawn from an area between
the Seine River and the Alluets Forest by means of wells driven In the
Senonian limestone strata.  The groundwater initially contained only a small
amount of ammonia and iron with iron bacteria.   The original  treatment plant
which was placed in service in 1961 Included the following process steps:
     !al
     (b)
          cascade aeration
          biological nitrification
          filtration
          post-disinfection with chlorine dioxide.

     A few months of operation caused a drawdown of the groundwater level of
the originally high quality water source.  This resulted in an intrusion of
groundwater from other sources, particularly from the Seine River.   The
quantity of flow from the Seine was accentuated by increases in the river
level due to the construction of reservoirs and locks to allow passage of
larger barges in the Seine.  As a result, water from certain wells  developed
unpleasant tastes and exhibited increased levels of organic matter, detergents
and bacteria.  In particular, there was a large increase in the number of
filamentous iron bacteria.

1969 ADDITIONS

     Major additions to the treatment plant were'made operational  in 1969.
These modifications Included chemical clarification ahead of nitrification.
Chemical addition ahead of clarification included aluminum sulfate  for
coagulation, activated silica for flocculation and powdered activated carbon
for elimination of detergents, organics and bad tastes.   Facilities were
provided to enable ozonation of the water after the filtration step.  Ozone
addition is for taste and odor elimination, virus inactivation, removal  of
micropollutants and detergents.

                                     324

-------
     Since 1969, the treatment process has been as follows:  chemical
clarification, cascade aeration,, biological nitrification, ozonation and
post-disinfection with chlorine.  In Figure 85 the treatment process schema-
tics of this plant are compared from 1961 to 1969, then from 1969 until
1978.

DETAILED PLANT DESCRIPTION

(a)  Twenty-three, 30-meters deep wells,, each producing 30 to 40 cubic
     meters per hour of water.

(b)  Chemical Addition:

          Aluninum sulfate         -    10 mg/1
          Activated silica         -     2 mg/1
          Powdered activated carbon-    10 mg/1

(c)  Four, 1.5CO cubic meters/hour Pulsator flocculator-clarlfier units.
     Each unit is 22.3 x 21.6 x 5.1  meters water depth.

(d)  Cascade aeration.

(e)  Twenty-one biological nitrification units  (described  in detail  below).

(f)  Gravity sand filtration - 5 cubic meters/sq m/hr.

(g)  One ozonation systems, consisting of the following  components:

     Two, 440 cu m/hr variable speed, positive  displacement blowers.

     One, water-cooled, heat exchanger type,  after-cooler.

     One, Freon refrigerant-cooled drier to reduce air  temperature  to 5°C.

     One, 2-cell, activated alumina  desiccator  drier  to reduce  air  dew point
     to minus 60° C.

     One, 550 tube,  horizontal tube,  Welsbach water-cooled  ozone generator
     operating at 50 Hertz, with a production capacity  of  22 kg of  ozone  per
     hour.

     One, 2-compartment,  countercurrent flow (ozone/water), 5 meters  water
     depth, porous tube diffuser,  ozone contactor.  Each compartment  provides
     6  minute ozone contact tiires, with an overall ozone contact time of  15
     minutes.  Total  ozone dosage -  0.5 mg/1.

     One, 200°C contactor off-gas ozone destructor.

(h)  Post-disinfection - Chlorine is added at a booster pump station  189
     meters from the plant.
                                     325

-------
1961  to 1969
    Cascade
  Aeration
 Biological
   Nitrification
  Filtration
 Post-Disinfection
     (C102)
                              aluminum  sulfate
                              activated silica
                     powdered activated carbon
                        -Phosphorus
 1969 to present
                                                  Coagulation
                                                  Flocculation
                                                  Clarification
   Cascade
     Aeration
 Biological
   Nitrification
                                                   Filtration
                                                    Ozonation
                                                         J/
Post-Disinfection
     (Chlorine)
  Figure  85.  Aubergenville water treatment plant.  Process Diagrams,
                                326

-------
Biological Nitrification System

     The details of biological  nitrification at Aubergenville are as  follows:

     There are 21 nitrification beds at the AubergenvWe plant,  each with
     dimensions of 4.9 m (16.1  ft)  by 5.9 m (19.3 ft)  for a surface area  of
     28.9 m (311 ft).   They have the following  characteristics:

          Bed depth 1s 2.5  m (8.2 ft).

          Filtration rate at design flow 1s 11.65 m/hr (4.76 gpm/sq ft).

          Upflow type.

          Intermittent aerations,  on for 2 minutes and  off for 6.   This 6
          minute aeration time  is less  than the time required for water to
          pass across  the filters.   This mode of aeration decreases the
          bicarbonate  equilibrium by decreasing the quantity of CCg eliminated
          by  stripping  and  does  not hinder nitrification.

          The filter material used  is pozzolanic (volcanic  stone)  with a
          grain size of 0.5 to 0.15 cm.

          Aeration  is  accomplished  by blowing air through the bed.

          The quantity  of phosphorus  added is between  0.1  and 0.2  mg/1,
          expressed as  P20s.  A  phosphorus residual is not  detected at the
          outlet.

          Nitrification beds  are  backwashed with water, once per week.

          Once every six months,  the  media are  dosed with hypochlorite to
          eliminate filamentous  bacteria.   This dosing is carried  out for
          24  hours.  Reseeding of the filter then requires  one week.

          Once every two years the  media are removed from the beds, washed
          in  media  scour and  put  back in place.   Reseeding  then is  immediate.

          Efficiency:   An influent  ammonia concentration  of 3 mg/1  is
          reduced further to  only trace quantities of  ammonia after sand
          filtration.
                                     327

-------
                                  APPENDIX  B

         CHOISY-le-ROI  (EDMOND  PEPIN  PLANT), SUBURBS OF  PARIS,  FRANCE


 BACKGROUND

     This plant currently  is designed  to treat 800,000  cubic meters per day
 (211 mgd) of River Seine water and has an  average dally flow of 470,000 cu
 m/day  (124 mgd).  In 1961  the  first  300,000 cu m/day leg of the plant was
 installed, and by 1969 the balance of  the  800,000 cu m/day treatment capacity
 was operational.  Ozonation was installed  in 1967 for color removal, taste
 and odor control, removal of phenols, detergents, polyaromatics, micropollu-
 tants  (dissolved organic materials)  and for bacterial  disinfection and viral
 inactivation.  The plant is owned by the Syndicat des Communes de la Banlieue
 de Paris pour les Eaux, a public agency, but is operated under contract by
 the Compagnie G§n€ra1e des Eaux.

 ORIGINAL PROCESS (Chfidal. 1976)

     Seine River water (which commonly contains 6 to 10 mg/1  of total
 organic carbon), is chemically treated with flocculant (hydrolyzed aluminum
 chloride), powdered activated carbons, then by breakpoint chlorination  (for
 ammonia removal), with sodium hydroxide for pH correction and chlorine
 dioxide for destruction of organic manganese complexes.   Following addition
 of these chemicals, the water is treated by sedimentation and rapid sand
 filtration,  dechlorinated by addition of sodium bisulfite,  then ozonized (2
 to 5 mg/1 applied dosage, average dosage 4 mg/1)  and treated  with chlorine
 to provide a residual disinfectant for the distribution system.

     In 1977, the post-disinfectant was changed to chlorine dioxide to
 provide a more stable distribution system residual.

     This process has the disadvantage of producing chlorinated organic
materials which are not readily removed during subsequent processing nor by
 the dechlorination step.

 BIOLOGICAL REMOVAL OF AMMONIA

     Gerval  (1978) has described experiments which have led to replacement
of the breakpoint chlorination step by a biological  process for the conversion
 of ammonia to nitrate.  This is done by preozonizing the raw  Seine River
water as it enters the plant raw water reservoirs.  After preozonation,  the
water is retained about two days in the raw water reservoirs  before being
 treated further by addition of the same chemicals as used in  the old process,
with the exception of chlorine and sodium bisulfite.

                                     328

-------
      The preozonation step itself involves addition of low levels of ozone
 (up to 1.25 ng/1)  over a short contact time (2 minutes).   In Table 81  are
 listed values  obtained for detergents,  organic carbon,  COD and  ammonium ion
 concentrations contained by the raw water,  by the  water after 2-day storage
 with no preozonation,  and by the water after  preozonation  (1.25 mg/1 ozone
 dosage)  plus 2-day storage.   The contents  of  detergents and  CCD are nearly
 halved,  the organic carbon  content  is  lowered from 8  to 7  mg/1  and the
 ammonium content is lowered  from 4  to  3.2 mg/1  by  preozonation  followed  by
 2-day retention.   Nitrification  now also can  occur in the  sand  filters.

      Several advantages  have  been realized  from the preozonation  treatment
 which produce  savings  in  both chemicals and processing  costs, as  well as
 producing a higher  quality finished water.  First, the  post-ozonation dosage,
 required for viral  inactivation, can be lowered by at least 2Q%.  According
 to French public health standards,, whenever ozone  1s used for disinfection
 purposes, it must be applied under the following conditions:  after satisfying
 the  Initial ozone demand and attaining a 0.4 mg/1  of residual ozone in the
water, this 0.4 mg/1 residual then must be maintained for a mlnimutr of 4
minutes.  In plant practice, this residual  normally is iraintained 6 to 12
minutes, to be certain of meeting the 4 minute requirement.  This treatment
standard for viral  inactivation is based on the pioneering work  of Coin et
al_.  (1964; 1967) and is further described by Miller et.al_., (1978).     ~~~

     Gerval (1978)  states that by the old treatment process, without preozona-
tion, the average ozone dosage necessary to provide viral  inactivation (to
attain and maintain 0.4 mg/1 of dissolved ozone) was 4.2 mg/1.  During the
period of time the pilot plant studies were conducted employing  1.25 mg/1 of
preozonation, the average amount of ozone required for post-ozonation was
halved, to 2.1  Tig/1.  Thus the total amount of ozone dosage required was
1.25 mg/1 preozcnation plus 2.1 mg/1 post-czonation, or 3.35 mg/1, a savings
of about 20%.

     Preozonation also was found to lower the amount of process  chemicals
normally added, not only by eliminating the need for chlorine (used in the
breakpoint step) and sodium bisulfite, but also because of the flocculation
effect caused by ozone oxidation which lowers the  amount of flocculant,
powdered activated carbon and chlorine dioxide required.  In addition,  the
tirre between backwashings of the sand filters was  doubled.   Finally, when
the amrronium ion concentration 1n the inlet water  was not greater than 1
mg/1, the ammonia level in the sand filter outlet  did not exceed 0.1 mg/1.

     As a result of these process improvements, it is now possible for the
Choisy-le-Roi plant to employ very low levels of chlorine for post-disinfec-
 tion and formation of residual for the distribution system.  Using the
modified process which Includes the preozonation technique, organic halogen
compounds (probably trihalomethanes) were less than 10 m1crog/l  In the pilot
 plant work reported by Gerval (1978).

      Preozonation of raw water followed by retention over a period of time
 to allow biological activity to lower the contents of ammonia and organic
 carbon, as well as  to allow reductions in levels of chemicals added and
 reduction in the amount of ozone required for viral Inactivation  has been


                                     329

-------
practiced for several years at the Moscow, Russia water treatment plant
(Schulhof, 1979).  Based on the successful use of the process 1n Moscow,
preozonation is being installed to replace breakpoint chlorination at the
Choisy-le-Roi plant in the Paris suburbs.  The process is scheduled for
full-scale operation in 1980 (Schulhof, 1979).  In addition, pilot plant
studies are being conducted at other water treatment plants in the Paris
suburbs and at most of the large scale water treatment plants operated by
Compagnie G£n£rale des Eaux (Le Pauloue1, 1978).

     Figure 86 shows the comparative schematic diagrams of the treatment
process used at Choisy-le-Roi  in 1978 and the modified process which includes
preozonatlon.

     TABLE 81.  EFFECT OF PREOZONATION (1.25 mg/1) BEFORE STORA3E

Raw water
2-day storage
(no preozonation)
2-day storage
with
preozonation
Detergents
(irq/1)
0.16
0.13
0.08
Organic C
(mg/1)
8.3
8
7
COD
(mg/1)
19
n
6
[NH4+1
(mg/1 )
6.2
4
3.2
Source: Gerval , 1978
     A recent article by Schulhof (1980) gives more details of the benefits
of preozonation at Choisy-le-Roi and two other suburban Paris plants and the
improvements being Incorporated at these three plants to maximize biological
removal of pollutants in (1) preoxidized reservoirs ahead of the treatment
processes (2) in biologically operating sand filters and (3) in biologically
operating GAC adsorbers.
                                     330

-------
  1967 to  1980
                                                     starting 1n 1983


Seine River raw
water

^ 	
chlorine (breakpoirt)


Seine River raw
water
1
preozonation
(up to '.25 ng/1)
^


^.. _ ,. 	 nU aHi ictr,£nt /NafHl 	 , Z

Sedlneitatlon
i
rapid sard
filtration
4.
dechloriratlon
(NaHSO,)
1
Czonatlon
(4 ng/1)
I
CIO,

^
/
Sedirentafon
i
rapid sand f'ltra
tion (nitrification
\
r
Post ozonation
(2.1 Tig/')
^
ozone
}-. Off-
' gas
, recycle
1
1
1
A
1
1
1
) 1
1
I
I
1
1
i
— i
t
:io2

       T
to distribut-on                                       to distribution




     Figure 86.    Choisy "e  Roi water  treatmeit plant   Process Diagrans.
                                 331

-------
                                 APPENDIX C

                   DUSSELDORF, FEDERAL REPUBLIC OF GERMANY
 BACKGROUND
     Three water treatment plants are operated by the City of DUsseldorf
currently, and include the Flehe (88,000 cu m/day), Am Staad (119,200 cu
m/day) and Holthausen (192,000 cu m/day) installations.  These three plants
use the identical process to treat river sand bank filtered water from the
River Rhine, and which is described below.  The Am Staad plant has been
using the process of ozonation followed by GAC since 1961, and is the longest
operating water treatment plant known to use this sequential  combination of
water treatment steps.

     However, there are at least two other water treatment plants in the
vicinity of 3Usseldorf which also draw water from the same region of the
Rhine and use closely similar treatment processes.   These are the Duisburg
Wittlaer III Wasserwerk and the Wuppertal Benrath plants.  Plant data
presented in this section of the report for unnamed water treatment plants
utilizing river sand bank filtration was obtained from one or more of the
plants noted above.

     A detailed discussion of the CUsseldorf water treatment  process was
presented in Section 9 of this report, along with pertinent data showing the
effectiveness of the various process steps.   These  will  not be duplicated
here.   Instead, additional plant performance data will  be presented and
discussed.

RIVER SAND BANK FILTRATION

     The five plants noted above make excellent use of the natural aquifers
adjacent to the River Rhine.  Figures 32 and 33 illustrate that the majority
of water treatment is achieved in the so-called bank filtration (German:
(Jferfiltration) step.  On the other hand, a disadvantage of using this
method of pretreatnent of organics  is the necessity for subsequent ozone
treatment for removal of iron and manganese.

     Water is drawn  from 10 m (32.8 ft)  to 30 m (98.4 ft) deep  wells situated
50 m (164 ft) to 250 m (820 ft)  from the banks of the River Rhine.  The
aquifers in which the wells are located  consist of  sand and gravel deposits.
Removal  of turbidity and associated pollutants in the sand bank treatment
results  in a high quality sand bank filtered water.   Additional  treatment
removes  iron, manganese, taste,  odor and dissolved  organics.


                                      332

-------
 OZONATION

      Ozone Is  generated  by means  of horizontal  tube, water  cooled,  ozone
 generators manufactured  by Gebruder Herrmann of Koln (Cologne),  Federal
 Republic  of Germany.   Each of  the ozone generators contains 432  glass  tubes
 and  generates  ozone from dried air  by  the application of power at 50 Hertz
 and  voltages ranging  from  9,000 to  16,000 volts.  These tube-type generators
 were  installed 1n the  early 1970s and  replaced  the water-cooled, Otto  plate
 units originally Installed  1n  1961.  Cooling of  the ozone generators is by
 means of  a closed loop cooling water system using chemically treated boiler
 feed water.  The cooling water is cooled, in turn, by passing it through a
 heat exchanger located in  the  raw water supply line.

     Ozonized  air 1s drawn from the ozone generator by means of the negative
 pressure  induced by passing treated water (approximately 1% of the total
 plant flow) through a ventuH  nozzle situated near the top of a vertical
 pipe having 10 meter submergence.   The ozonized water mixes with the sand
 bank filtrate which passes down the vertical pipe and discharges Into the
 contact charrber.   The 10 meter submergence and overall  contact chamber size
 provides a retention time of five minutes.  The ozonized water from the
 contact chamber proceeds to a holding tank which provides  an additional 30
minutes retention time.

     Over  the  life of the Am Staad plant 'since 1961),  the ozone dosage
 necessary  to produce the desired quality of finished  water has had to be
 increased from an average of 1  mg/1  to an  average of  3  mg/1.   This ozone
dosage is set manually to maintain approximately 0.1  mg/1  of ozone in
 the off-gases from the 30-m1nute detention tanks.  This level  of ozone 1s
determined both by the potassium Iodide wet chemistry procedure and  by plant
personnel   "sniff-testing" the off-gases,,   Residual  ozone 1n the off-gases
from the contact  chambers and the  holding  tanks is  destroyed by passage
through wet granular activated  carbon.   However, catalytic destruction of
the ozone-containing off-gases  has been tested at DUsseldorf plants  and will
be installed to replace the wet GAC procedure (Welssenhorn,  1977).

FILTRATION AND GAC ADSORPTION

     Water is punped at pressures  of 6 bars (87 ps1)  to 7  bars (101  psi)
 from the ozonation holding tanks to the bilevel filtration/adsorption
 units.  These units are constructed of steel, 8 rreters  high and 5 meters in
 diameter.  Water flows downward through an upper 1.5  meter deep filter layer
 and then downward through a 2.5 meter depth of granular activated carbon.
 Both the filter layer and the adsorption layer rest on  layers  of support
 gravel, each based on a steel support plate containing  plastic nozzles
 (approximately 31 per sq m).  Each layer can be backwashed separately.

     The hydraulic loading rate in the filtration/adsorber units has been 20
 cu m/sq m/ hr  (m/hr), or 8.8 gpm/sq ft, but recent plant expansions  have
 reduced the loading rate to 12 m/hr (4.8 gpm/sq ft).   The  older coated steel
 units have been on-Hne since 1961 with satisfactory  service from the coating,
 with the exception of physical  wear in the vicinity of the media removal and
 changing fittings.


                                     333

-------
     The  filtration medium  is a so-called  "preactivated" carbon having the
following characteristics:

               Granule size        -    0.9 to 2.5 mm
               Mean granule diameter -  1.7 mm
               Bulk weight         -    580 kg/cu m

     Biological nitrification, as well as  high degrees of removal of mangan-
ese, iron and turbidity, is achieved in the filtration stage with provision
of 1.5 meters of filter media.

     The  adsorption medium consists of a mixture of Lurgi LS Supra and
Chemviron (Calgon) F-300 granular activated carbons having the following
characteristics:

                    Granule size        -    0.5 to 2.5 mm
                    Mean granule size   -    1.4 to 1.6 mm

     Filtration layer runs of approximately 24 to 48 hours are observed.
Backwash  cycles for the adsorption stage are 4 to 6 weeks, with periods
between regeneration ranging from 5 to 6 months.   Initial air scouring,
followed  by a 5-minute water backwash, is the procedure used for both the
upper and lower layers.

FINAL TREATMENT STEPS

     The  final treatment steps include the addition of sodium hydroxide 'or
pH control (to neutralize C02 produced biochemically) and 0.1 to 0.3 mg/1 of
chlorine dioxide to disinfect and to maintain a residual  within the distribu-
tion system.

     The  three City of Dusseldorf water treatment plants  are unique in that
there is very little reservoir storage for finished water.   As a result,  all
three plants are designed to operate on demand.

GRANULAR ACTIVATED CARBON REGENERATION

     Over the years since the ozone/GAC process was installed in Tusseldorf,
the increased GAC levels of halogenated organics  (TOC1) in the Rhine (which
are not removed during river sand bank filtration)  has shortened the GAC
regeneration cycles to such a point (every 5 to 6 months) that it became
economical to install  reactivation facilities  at  the Dusseldorf plants.
Spent GAC now is transported from the Am Staad and  Flehe  facilities to the
fluidized bed regeneration furnace situated at the  Holthausen plant.   This
furnace, which was in  shakedown during a site  visit in May,  1977,  was fully
operational  in June 1978,

     The GAC regeneration furnace was supplied by Lurgi and  has a GAC
regeneration capacity  of 6 metric tons per day.   It is to be the central
regeneration facility  for Dusseldorf, and space 1s  available for two additio-
nal furnaces to be Installed if and when needed.   Spent GAC  is flushed from
a storage hopper to a  dewatering screw which regulates the carbon to the


                                     334

-------
 upper or drying stage of the fluidized bed furnace which is maintained at a
 terrperature of 200° to 300°C by means of natural gas heating.  The dried GAC
 passes to the lower or reactivation stage where steam is injected into the
 chamber which is maintained at a temperature of 600°C to 800°C, again using
 natural gas heating.  The SAC enters a quench tank from which it is flushed
 to a storage hopper.  The furnace off-gas is treated by means of a cyclone
 after-burner and heat exchanger.

     Another fluidized bed granular activated carbon regeneration furnace is
 installed at the Wuppertal Benrath water treatment plant.  While this plant
was not inspected by the 1978 site visitation team, it is known that the
unit was designed and constructed by WABAG, a design/construction engineering
firm headquartered in Kulmbach, Federal  Republic of Germany.   The Wuppertal
regeneration facility was installed in January, 1978 but was  still  in'shake-
down in June, 1978.  The maximum capacity of the unit is 240  kg of GAC/hr,
but the recommended rate of application  is 100 to 150 kg of GAC/hr.

     A third fluidized bed GAC regeneration furnace is in operation  at the
ZUrich (Switzerland) Lengg water treatment plant (Grombach, 1975).

     It has also been reported (Water Research Center, 1977)  that there is a
multiple hearth GAC regeneration furnace  at the Alelyckan water treatment
plant at Goteberg, Sweden.   This same reference notes that "one granular
activated carbon regeneration plant is currently in use  at Church Wilne,
England."  This English regeneration unit has been discussed  by Osborne
(1979).

PLANT OPERATIONAL DATA

     Operational data from the individual  plants 1n and  around CUsseldorf
are not readily available either from the plants or from the  Engler-Bunte
Institute of the University of Karlsruhe,  which conducts research for many
German waterworks.  The Institute places  great emphasis  on the confidentiality
of the raw and/or unpublished analytical  data developed  by and for  their
client waterworks.  However,  information  was  provided for unnamed plants  on
the lower Rhine which utilize the same treatment process as does Dtlsseldorf.
Similar data for other water plants treating  river waters in  the Federal
Republic of Germany were provided as well,  identifying only the river source
and the treatment process.

     Table 82 contains actual plant operational  data showing  the degree of
purification obtained at the various stages of treatment of River Rhine
waters in the Dlisseldorf area.   Table 83  lists DOC and halogenated  organlcs
data obtained during a single day of operation at one of the  five plants  in
the region.

     Table 84 shows data obtained at a lower Rhine water treatment  plant
which uses a process consisting of river  sand bank filtration, permanganate
 (oxidation)/polymer sedimentation, filtration and GAC treatment.  These data
were determined from the three operating  trains in the plant,  and are averages
of data obtained throughout 1977.
                                     335

-------
TABLE 82.  LOWER RHINE RIVER WATER TREATMENT DATA FOR PLANTS UTILIZING
           RIVER BANK FILTRATION/OZONATION/FILTRATION/GAC ADSORPTION.
           AVERAGES FOR 1977
Parameter
PLANT #3
DOC (mg/1)
COD (mg/1)
UV
PLANT #4
DOC (mg/1)
COD (mg/1)
UV
PLANT #5
DOC (mg/1)
COD (mg/1)
UV
Raw Rhine
River Water
4.55
12.7
11.62
4.46
13.11
11.07
4.03
11.13
11.20
After River
Bank Filtration
1.8
4.88
3.95
2.16
5.54
5.13
2.17
6.07
5.10
After Ozonation
& Filtration
1.70
2.83
2.03
••**
2.41
1.90
3.30
After
GAC
1.15
3.33
1.23
1.45
3.47
1.47
1.43
3.19
1.87
tote: DCC = dissolved organic carbon
COD s chemical oxygen demand
UV = relative absorbance at 250 nm
     Table 85 shows data obtained at a plant utilizing river sand bank
filtration, followed by GAC adsorption directly.  Data presented are the
averages obtained during 1977.

     Table 86 shows average data for 1977 at a River Danube plant (without
river sand bank filtration) using ozonatlon, filtration and GAC adsorption.

     The following conclusions may be drawn from the data of Tables 82
through 86:

1)   River sand bank filtration is an effective method of removing waterborne
     pollutants, as measured by DOC, COD and UV adsorption.  The degree of
     pollutant removal is variable, however, as 1s seen by comparing data
     from Plants No. 3, 4 and 5 with those from Plants 2 and 6.

2)   Data from Plants No. 3, 4 and 5 are in general agreement as to the
     levels of treatment attained after river sand bank filtration.   Percen-
     tage removals of COC, COD and UV absorbing materials by the treatment
     process consisting of ozonation/filtration/GAC adsorption is as follows:
                                      336

-------
     TABIE 83.  PLANT J4.  CHLORO-ORGANIC MATERIAL  DATA FROM A SINGLE DAY'S SAMPLING IN 1977
Water Treatment
Process
Rhine River
raw water
river banK
filtration
ozonation
filtration + GAC
adsorption
0.2 mg/1 chlori-
nation
Parameter
DOC CHC13 CHBrCl2 CHBr2Cl CHBr3 CC14
jag/l^g/i
4.7 3o9

2.2 1.4
2.0 0.9

1.6 1.1
1.3 1.3
C2H4C12 JC2HC13 C2C14 C2H2C14
.

—

—
—

—
0.1
—

—
--

—
0.2
; -- 2.4

— 0.03
— 0.02

— 0.04
0.8 0.04
—

—
--

—
__
0.8 1.1

0.9 0.7
0.2 0.3

0.1 0.05
0.2 0.06
—

—
—

—
—
OJ
OJ
—I

-------
                              % Reduction 1n Levels of

          Plant No. 3
          Plant No. 4
          Plant No. 5

TABLE 84.  LOWER RIVER RHINE WATER TREATMENT DATA FOR PLANTS  USING RIVER BANK
           FILTRATION/KMnOi/POLYMER SEDIMENTATION/FILTRATION/GAC.   AVERAGES
           FOR 1977
DOC
36
33
34
COD
32
37
47
UV
69
71
63
Parameter
PLANT #6
TRAIN #1
DOC (mg/1)
COD (mg/1)
UV
TRAIN #2
DOC (mg/1)
COD (mg/1)
UV
TRAIN #3
DOC (mg/1)
COD (mg/1)
UV
Raw Rhine
River Water
4.64
14.10
11.07
4.64
14.10
11.07
4.64
14.10
11.07
After River
Bank Filtration
2.46
7.17
6.37
2.46
7.17
6.37
2,46
7.17
6.37
After KMnOd
& Sedimentation
2.07
3.75
2.05
3.78
2.24
3.81
After
3AC
0.96
0.95
1.26
1.63
0.99
2.50
0.81
TABLE 85.   LOWER RHINE RIVER WATER TREATMENT DATA  FOR  PLANT  UTILIZING  RIVER
           BANK FILTRATION/GAC ADSORPTION.   AVERAGE  FOR  1977
Parameter
PLANT #2
DOC (mg/1)
COD (mg/1)
UV
Raw Rhine
River Water
4.25
11.23
9.13
After River
Bank Filtration
0.88
2.77
1.47
After
GAC
0.74
2.47
1.16
                                     338

-------
TABLE 86.  UPPER DANUBE RIVER WATER TREATMENT DATA FOR PLANT UTILIZING
           SEDIMENTATION/OZONATICN/FILTRATION/GAC.  AVERAGE FOR 1977
Parameter
PLANT #1
DOC (mg/1)
COD (mg/1)
UV
Kaw Danube
River Water
2.26
5.51
4.98
After
Treatment
1.61
1.49
COSTS

     Only limited cost data were made available during the site visit to the
Dusseldorf waterworks.  This brief Information 1s presented below.

     Total revenues for the waterworks in 1976 were about 73 million Ceutsch
marks (DM), up from 55 million DM in 1974.  The average rate charged for
water was 0.9 DM per cubic meter 1n 1978, with the flat residential  customer
rate set at 1.2 CM/cu m, approximately equivalent to $2.20/1,000 gallons.

     Treatment costs (exclusive of distribution and administration)  were
reported (Poggenburg, 1978) to comprise somewhat less than 40% of the total
annual Dusseldorf water utility costs.  The figure of 0.28 DM/cu m was
mentioned as the total water cost in a recent year, excluding returns on
capital.  Costs at CUcseldorf should not be compared with those of nearby
MUlheim, for example, because the Dusseldorf Wate>" Works pays more for
street usage.

     The waterworks provides capital for new projects from its own funds set
aside in previous years and from new bond issues.   Bonds were sold recently
(prior to 1978) at an interest cost of about 6.5%.  The waterworks is a
private stock corrpany, with the City of Dusseldorf owning 100% of the stock.
Due to its private status, Its revenues are drawn entirely from rates
charged for water and its financing 1s separate from other city capital
projects.

     The new fluidlzea bed furnace for reactivating spent GAC from all three
DUsseldorf water treatment plants has been operating for over a year.  The
two-stage furnace operates at 200° to 300°C at the first stage for drying
and at 600° to 800°C at the second stage for reactivation.  An afterburner
(which attains 600°C outlet temperatures) prevents air pollution and prevents
even a visible steam plume.  Absence of a visible steam plume is required by
local air pollution control regulations.  A cyclone removes particulates to
a level of 75 parts per billion.

     Energy requirements are 55 to 65 cu m/hr of natural gas (AHgas ^8,40)
kcal).  At present, the furnace is operating with 4 to 5% GAC losses in the
furnace, with a total loss of 8 to 9% (the difference 1s lost 1n transporting
GAC from the adsorber to the furnace and returning).  However, the 4 to 5%
                                      339

-------
furnace losses depend to some degree on the quality of the GAC and the
temperature and precision of the furnace operations.  The regeneration
capacity is 6 tons/day (250 kg/hour).

     Although no cost figures were made available on either the capital
invested tn the furnace or its operating costs, the total cost of regene-
ration was said to be a bit less than 30% of the cost of virgin GAC.  With
European GAC prices 1n the range of 70 to 80<£/lb, this percentage suggests a
reactivation cost on the order of 20£/1b.

     Capital costs for a single GAC column at DUsseldorf were said to be
120,000 DM, without outside piping.  The installed total  capital  cost for
water treatment was 3,000 DM/cu m of water treated per hour.  For each 1,000
cu m/hr of installed treatment capacity, the capital cost was 1,OCO x 3,000
= 3,000,000 DM.

     The City of DUsseldorf treats 400,000 cu m/day of water and pumps
500,000 cu m/day from the wells.  The additional 100,000  cu m/day of well
water is sent to industries and nearby cities for local treatment.  The
staff at Dusseldorf totals 375 people, but only 2 people  operate the plants
around the clock.
                                      340

-------
                                 APPENDIX D

                    MORSANG-SUR-SEINE (SUBURBS OF PARIS)


GENERAL CHARACTERISTICS

     The Morsang plant is located on the Seine River, 35 km upstream of
Paris near the new town of Evry and supplies water to Evry as well as to the
Paris suburbs.  In concept and when completed about the year 2000, Morsang
is expected to produce about 1,000,000 cu m/day (264.2 mgd) of drinking
water from the Seine River.

     Morsang is being constructed in successive,  star-shaped plant stages,
with each star being capable of producing 225,COO cu m/day (59.4 mgd).   Each
star at Morsang will be constructed around a central station with each
treatment stage occupying one point of an In-aginary three pointed star.   At
the center of each star will be the central  station, which will  house chemi-
cals, activated carbon and ozonation equipment.   Laboratories, pumping
station and administration are housed in a separate building.   The general
plant layout-is shown in Figure 87.

     The first stage of the first 3-pointed  star  began operating in 1970.
In 1975 the second stage of the first star was completed, and  construction
of the third stage of the first star will depend  upon results  obtained  from
a detailed pilot plant testing program,  which was to be completed during
1979.

     The Morsang plant is owned and operated by the Soci§t£ Lyonnaise des
Eaux et de TEclairage, one of two large, integrated water making companies
in France.

     Raw Seine River water at Morsang is fairly clean:  TOC is about 3  mg/1,
humic acids about 7 mg/1, ammonia about 0.5  mg/1  and taste threshold about
12.

PLANT OPERATION

     Raw Seine River water is coarse screened (1.5 mm mesh), then is subjected
to breakpoint chlorination (for removal  of ammonia, usually less than 0.5
mq/1) with 3 to 6 mg/1 of chlorine added as  the gas.  Prechlorinated water
then is pumped to the center of the first star where it is split into two
streams.  One stream of 50,000 cu m/day  (13.2 mgd) is sent to the first
stage of the star, and the second stream of 75,000 cu m/day (19.8 mgd)  is
sent to the second stage of the star.

                                      341

-------
                                                     Control
                                                      Room
                                                                    Future treat-rent
                                                                    building
                                                          Certral Co-trol SJuilding
                                                          for f ret Treat-iert &. d ng
      Figure  87.   General  layout of the Morsang-sur-Seine treatment works.
First Stage of the  Star   (Maximum flow:  3,300  cu m/hr --  871,860 gal/hr)

     Prechlorinated water is  settled  and  clarified  using  a  665 sq m s'udge
blanket Pulsator clarifier (Degremont) operated  at  an hydraulic loading rate
of 3.15 cu m/sq m/hr  (m/hr),  then filtered  through  six type T Aquazur V
(DegrSmont) sand filters.   Coagulation 1s accomplished with addition of 60
to 80 g/cu m of aluminum  sulfate  and  is monitored by  flocculation tests and
zeta potential measurements.  Activated silica (1 to  2.5  g/cu m)  is made by
adding sulfuric acid  to alkaline  sodium silicate, and its dosage  depends
upon the measured sludge  cohesion coefficient.   Powdered  activated carbon
normally is added ahead of the clarifier  at 15 to 25  g/cu m.

     Water is passed  through  the  sand *11ters, each having  an area of 60
sq m, at an hydraulic rate  of 6 cu m/sq m/hr (m/hr).   After sand  filtration,
the water is ozonized in  two  baffled contact chambers,  each having a volume
of 215 cu m (10.5 m long  and  4.5  m high)  and designed to  achieve  a contact
time of 8 minutes at  peak flow.   About 67%  of the total ozone generated is
added to the first contact  chamber in which the  initial ozone demand is
satisfied and a residual  ozone concentration of  0.4 mg/1  is attained.   The
remaining 33% of the  total  ozone  dosed is added  in the  second chamber,  where
the Tevel of residual dissolved ozone is maintained at  0.3  to 0.4 mg/1.   The
contact time 1n each chamber  1s 4 minutes at maximum  flow (8 minutes total
ozonation at peak flow).   Water flow 1n each chamber  1s countercurrent  to
the direction of the upward flow  of ozone-containing  air  (see Miller et_ al..
1978, p. 120).  The residual  dissolved ozone is  monitored at the  outlet of
both contact chambers at  0.3  to 0.4 g/cu m.
                                      342

-------
     This  treatment sequence will be referred to subsequently as Process #1.

 Second Stage of the Star;   (Maximum flow: 3,300 cu m/hr -- 871,860 gal/hr)

     Settling and clarification of the prechlorinated water are conducted
 with a 208 sq m Superpulsator clarifier (DegrSmont).  Filtration is through
 four Aquazur sand filters, each having a unit surface area of 63 sq m, at an
 average hydraulic loading of 8.75 m/hr and a peak hydraulic loading of 13
 m/hr.

     Ozone is produced (for the first two stages of the star) by two Degremont
 generators, each having a production capacity of 6.6 kg/hr.  Contact times
 (8 minutes at maximum flow) and other ozonation conditions are identical
 with those of the first stage.

     Following ozonation is 6AC adsorption through 1 m deep beds of Chemviron
 F-400 carbon, using four Aquazur filters each having a surface area of 63
 sq m.  These are operated at a  filtration velocity of 13.1 m/hr at maximum
 flow.  These Aquazur filters are especially designed for activated carbon
and are called Mediazur filters.

     After sand filtration and  before ozonation,,  the water is divided  into
two equal  portions.   The first  fraction is ozonized before GAC adsorption
and the second is  passed through the GAC adsorbers  then ozonized.   These  two
sequences  of treating water in  the second stage  of the star (termed Process  #2
and Process 13,  respectively) have been monitored full-time since  October,
1975 and the water qualities compared with those  of Process #1  conducted
over the same period of time.   Initial  results  of this comparative study
 (obtained  over the period October, 1975 through  December,  1976)  were  reported
by Richard & Fitssinger (1977)  and are discussed  below.   The three processes
are shown  schematically in Figure 88,  along with  a  fourth pilot plant  process
 (Process #4)  which will be discussed later.

     Parameters studied during  the 15-month program reported by Richard &
Fiessinger (1977)  for Processes #1,  #2 and #3 included:

 (a)  Organic matter — expressed as mg/1 of oxygen consumed as measured by
    . KMnO- oxidation in an acid medium,

 (b)  Taste threshold — measured at 30°C by the  dilution method,

 (c)  Organo-halogen derivatives -- detected by gas chromatography  combined
     with an electron capture detector.  These were extracted with Uvasol
     pentane (Merck).  Results  given by Richard  & Fiessinger (1977) were
     limited to chloroform and  dichloromethane only.

 RESULTS OF 15 MONTH STUDY COMPARING PROCESSES #1, #2 & #3

 Organic Matter

     Figure 89 shows the percent reduction in concentration of organic
 matter obtained by each of the  three processes.

                                      343

-------
                            RAW SEINE RIVER WATER
        PRECHLORINATED RAW WATER
CLARIFICATION WITH
 POWDERED ACTIVATED
  CARBON (PAC)
        \
  SAND FILTRATION
     OZONATION
CLARIFICATION
 WITHOUT PA':
 SAND FILTRATION
CLARIFICATION
 WITHOUT PAC
 SAND FILTRATION
1
i
OZONATION
1
\
GAC
ADSORPTION

i
Treatnu
Line t
(full-see

r
»nt
tz
lie)



i
G,
ADSO
OZO
i
Tr
L
(ful


r
\C
OPTION
'
NATION
r
satment
ine #3
1-scale)
i
OZONA
1
G
ADSO

1
Trea
Lin
(pilot

TION
r
AC
RPTION

F
tment
e #4
plant)
     Treatment
      Line #1
   (full-scale)

Note:  all treated water leaving the plant is  chlorinated  for  residual
 Tigure 88.  Morsang-sur-Seine water treatment  plant—processes  operating  in
              1977.
                                     344

-------
100 .

90 -

80 -
* 70 -
4)
o
co 7: 6O -
•£» c
tn ™
o>
o
c 5O.
c
o
1 40-
•o
** 30 .
treatment line 1 10° -
A clarification with PAC
90 -
* send filtration
* ozonatlon
so-
il /"•/ A* y\ 7°
1 \ r^Jr%\
JaWx^ JSr/4" 6° "
*7 1 7 v v ?
*"* 1 / i 50 -
1 /
1*
* 40-


G


1
J
1
1
F
j<
""f
I
1

treatment line 2 10° -
clarification w/o PAC
90 -
* sand filtration
| & ozonatlon
i 80 -
1 • GAC adsorption
J .^-A
JSM^-^-Vi 7° "
1 / V-A ...
it X t,^^ \ ^«"'*
fcj j ^ f »^ ja 1
*•*» f O W-* 
-------
 Process  #1 —
      Starting with  a  raw water  level  of  3  mg/1,  some  30  to  60%  of  the
 organic  material  is eliminated  by  clarification  with  10  to  20 mg/1 of
 powdered activated  carbon.   During the summer of 1976, this dose was increased
 to  25 to 30 mg/1.   Sand  filtration removed all carbon fines and increased
 overall  organics  removal to  55  to  65%.

      The dosage rate  of  ozonation  was found  to affect the organics removal
 greatly.  During  the  first three months, the ozonation dosage was  too  low
 and the  organic content was  the same as  or higher than that obtained after
 sand  filtration.  Increasing the ozone dosage to 1.2  mg/1 raised the total
 percent  removal of  organic matter  to 60  to 70%,  an additional 8% over  sand
 filtration.

 Processes #2 and #3—•
      Clarification  at the rate  of  8.75 m/hr without powdered activated
 carbon removed 40 to  50% of  the organic material and  sand filtration removed
 an additional 5%  (total organic removal by powdered activated carbon and
 sand  filtration: 45 to 55%.

      Process j2--During the first  three months of the study the ozone
 dosage again was too  low, but when this was increased to 1.2 mg/1, the total
 removal of organic material increased to 55% to  70%.  However, water treated
with  powdered activated carbon  by  Process #1 was of higher quality, with
 regard to organic matter content, after ozonation (1.4 mg/1  of ozone)  than
that  obtained from Process #2 (1.6 mg/1 ozone).

      During the first three tr.onths, GAC adsorption (after sand filtration
 and ozonation) increased the overall percent removal of organic materials to
 80 to 85%.  This figure fell to 70% when the water was cold and rose to 75%
when  the water was warm.  The increased organics removal  observed during
warm weather could have been a  result of bacterial activity in the GAC
adsorbers.  The 80% to 85% levels of organics removal  were obtained after
the first three months period and  Indicates that the additional  5% to  10%
removal also may have been a result of bacterial activity in the GAC beds.

      Process #3--Sand filtered water was passed  through granular activated
carbon.  This increased the total organics removal to 80 to 85%, but this
 level  fell to 70% when the water was cold.

     A drought occurred during  this test program, and the level  of organics
removal fell to 67% by Process  #3 but was 75% by Process  #2.  This higher
rate of organics removal by Process #2 was attributed to "the intense oxygena-
tion of the ozonized water and  the biological oxidation which it causes in
 the water".

     When the drought ended, the quality of water which passed through the
GAC adsorbers in Process #3 improved, but was still  slightly lower (about
 50%)  than that obtained by Process #2.
                                      346

-------
Taste Threshold

     During the 15 month study the raw water taste threshold rose from 10 up
to a level of 20 to 30 units during the drought.

Process #!--
     Clarification with powdered activated carbon followed by sand filtration
lowered taste thresholds to two units in most cases.   Upon ozonation, however,
"new, unpleasant tastes" were formed, and the threshold level rose to three
units.  During the drought the taste threshold of clarified water rose to 4
to 10 units (even with powdered activated carbon dosages being increased to
20 to 25 mg/1) and the level 1n sand filtered water rose from 3 units to 6
units.  During this period, the level in ozonized water remained at 2 to 3
taste units.  This level was considered to be "not perfect, but acceptable
to customers".

Process #2—
     After ozonation and GAC adsorption, the taste threshold remained
consistently at 1  unit over the 15 month period.

Process #3—
     Sand filtration followed by GAC adsorption without ozonation produced
taste threshold values of 1 for the first 10 months,  then the carbon "comple-
tely lost its capacity for removal of taste" and the  threshold value rose to
2 units.

     The difference in activated carbon performances  (with ozonation in
Process #2 and without ozonation in Process #3) was attributed to "the dual
role of. ozone1':

     a)   Ozone oxidized non-polar, or taste-causing, organic molecules Into
          polar molecules which have a less powerful  taste; and

     b)   Ozonation aerates the water and stimulates  biological  activity in
          the GAC adsorbers.

Formation and Elimination of Halogenated Organic Derivatives

     The chloroform content of the Seine raw water at the Morsang plant 1s
well below 1 microg/1.  Chlor1nat1on at the clarlfier (with 3 to 6 mg/1 of
chlorine) produces chlorinated organics, and chloroform levels of 5 to 15
microg/1 are found routinely at this point.  Figure 90 shows results of
chloroform analyses obtained during the 15 month study.

Process #1 —
     Powdered activated carbon treatment lowered the  chloroform content 40
to 50X, leaving 0.1 mg/1 of total chlorine in the clarified water.  Additional
chlorine (0.2 mg/1) had to be added to the filter Inlet to protect it from
algae and zooplankton, but this additional chlorine dose produced "no more
than 1 microg/1 of additional chloroform".  After clarification and sand
filtration, chloroform levels found were 5 to 7.5 microg/1.
                                      347

-------
15-
           10-
CO
 o>
 4.
 n
O
i
o
                     line 1
                                        15-
                                        10
                              raw water  5
                                water
                 o clarification  with PAC
                  A sand  filtration
                   » ozonatlon
                  son
                     month
                                A    '!   \
                                             line 2
                                                                15-
                                                                10-
       * raw water
      » raw water +
clarification w/o PAC   5
4 sand filtration
o ozonatlon
• GAC adsorption
                                                                       line 3
                                                                        •raw water
                                                                        *raw water  + CI2
                                                                 o clarification w/o PAC
                                                                 A sand filtration
                                                                  • GAC adsorption
                                                                            6    ri    d
                                                                               month
                                                                                      j    f
                      Figure  90.   Evolution of  CHCI3 in Morsang waters.

                                        (Richard & Fiessinger, 1977)

-------
     Ozonation lowered the chloroform level to 3.5 to 6.5 m1crog/l, but by
air stripping, not by chemical oxidation.

Processes #2 and #3—
     Prechlorinated water was clarified without addition of powdered activated
carbon.  The amount of chloroform produced tended to increase with longer
retention times in the clarifier, and vice-versa.  After sand filtration,
the average chloroform concentration was 10 microg/1.

Process #2--After ozonation followed by GAC adsorption, the chlorofom
concentration varied between 7 and 13 microg/1.  (It is apparent that the 1
meter deep GAC beds had been saturated with chloroform and that breakthrough
had been attained early in the 15 month study).

Process 13--The same chloroform range (7 to 13 microg/1) was obtained after
sand filtration followed by GAC adsorption (no ozonation).   Thus, ozonation
followed by GAC adsorption should have no capability to remove chloroform
after the adsorption capacity of the GAC for chloroform has been exceeded.

Conclusions From the 15-Month Study of Processes  #1,  #2 and #3

     Richard & Fiessinger (1977)  drew the following conclusions from this
15-month comparative study:

1)   Formation of chloroform can  be reduced by optimizing the coagulation
     step (before chlorination),  which lowers  the hunvic content to a minimum.

2}   After prechlorinat^on,  powdered activated carbon (PAC)  removes residual
     chlorine from the water, adsorbs chloroform  precursors and reduces  the
     level  of chloroform by  about bO%.

3)   Over the 15-month period, sand filtration followed by  GAC adsorption at
     10 m/hr (no ozonation)  removed 40% of the chloroform.

4)   Ozonation before or after GAC adsorption  removed some  chloroform,  but
     by air stripping, not by chemical  oxidation.

5)   Chloroform concentrations at Morsang have never exceeded 16 microg/1.

6)   Process #1 (clarification with powdered activated carbon, final  treatment
     with ozone) produces "less than perfect"  organoleptic  qualities, which
     will be affected directly by any change in the amount  of raw water
     pollution.

7)   Process #3 [clarification (without powdered  activated  carbon), sand
     filtration, GAC adsorption,  post-ozonation]  produces water with "very
     good organoleptic qualities".  With sudden surges of raw water pollution,
     however, it is better to use powdered activated carbon so as to prolong
     the useful life of the GAC.

8)   Process #2 [clarification (without powdered activated carbon), sard
     filtration, ozonation, GAC adsorption] gave "very appreciably improved


                                      349

-------
     results".  Ozone aerates the water and oxidizes non-polar compounds
     which cause unpleasant tastes.  This "Increases the activity of the
     carbon and prolongs Its useful life considerably".

     (No data are reported by Richard & Fiessinger, 1977, regarding the
     criteria measured to determine the useful life of the GAC adsorbers.
     However, these are presumed to be organoleptlc parameters, such as
     taste threshold, 1n order to maintain low taste thresholds.)

9)   Biological phenomena 1n the GAC appear to be stimulated by ozonatlon of
     the water before passage through the GAC adsorbers.

     The next phase of the research program at ^lorsang-sur-Seine was a pilot
plant study of the treatment of raw Seine water by a process 1n which break-
point chlorlnatlon was eliminated (Process #4).  This pilot plant study
began 1n early 1977 and was to have been completed during 1979, after which
a decision was to be made as to the treatment process to be Installed on
full-scale 1n the third stage of the first star.   Very recent data (Fiessinger
& Montiel, 1980) presented at the 1980 Annual  American Water Works Association
Conference Indicates that the decision will  be to install Process #4.

PILOT PLANT STUDIES WITHCUT PRECHLORINATION

     In Process #4, the prechlorination step has  been eliminated.  Raw Seine
River water is clarified with alum and lime  (no powdered activated carbon),
decanted, sand filtered, ozonized, passed through GAC adsorbers and post-
chlorinated.   Chemviron F-400 granular activated  carbon is  used in Process
#4 and the bed depth is 1.4 m.   Dissolved oxygen  levels in  the raw Seine
River water at Morsang are 3 to 5 mg/1.  Prior to the GAC adsorption step in
Process #4 the DO is 10.4 mg/1  and drops to  8.8 mg/1 after  GAC adsorption.
Ammonia levels 1n the Seine River raw water  are 0.1 to 0.2, but 0.1 mg/1 in
the product water.   Therefore,  nitrification 1s Insignificant in Process #4.

     For all  four processes, data on organic materials content (by the KMnC4
and UV absorbance methods), percent reduction  of  organic materials content,
TOC, taste threshold, fluorescence, humlc acids content and chloroform
content at various  points in the different treatment processes are presented
in Figures 91  through 94 (for Processes #1,  #2, #3 and #4,  respectively).
Figure 95 1s  an overview diagram of the four treatment processes showing the
points where samples for the analyses given  In Figures 91 through 94 were
taken.   Data presented in Figures 91 through 94 are averages of 5 samples
taken from side-by-side process operation.

     For comparison, data for all four processes  are summarized in Table 87.
The quality of water produced by Process #4  appears to be the highest  of all
processes, but comparison 1s difficult to make since the GAC used in Process
#4 was placed in use about 2 years after that  used in Processes #2 and #3.

     Therefore, the advantages  of not prechlorinating at Morsang remain
debatable.  It is still too early to predict the  final  process that will be
installed in the third stage of the first star (Richard,  1979).
                                      350

-------
  4-
  3-
  2-
  1-
  0
      org. mat*Is. -KMnO4
100
 50-J
org. mat'ls. -
    %  removal
  5-| fluorescence
  4-
                   10-f numic  acids
                    8-
     org. mat'ls. by UV
0.5-
                                          11
                       haloform comp'ds.
                           2
                                     -CHCI3
                                          11
 1-raw  waters 2-after  prechlorinatloni  3-after clarifi
 cation; 6~after sand filtrationj 11-after  ozonation

 Figure 91. Process 1 performance parameters at
            Morsang plant.

               (Rlchard,1978)
                        351

-------
  5-
org. mat'ls.-KMn04   5—  JOG
     org. mat'ls. -
      % removal
                         taste number
                         th=threshold
                                              12
  5-
  4
  3-
  2-
  1
  0
     fluorescence
                    10-
                     8-
                     6-
                     4-
               9  12  2-
                         humic acids
1.0-«
0.5.
           org. mat'ls.   50-
                                haloform
                                 comp'ds.- CHCIg
                                       9-,12
7-after sand filtravion; 9-after ozonation;
   12-after GAC adsorption

Figure 92.  Process 2  performance parameters at
            Morsang  plant.

                 (Richard, 1978)
                         352

-------
      org.mat'ls.- KMnO^   5*.  TOG
      org. mat'ls. -
         % removal
taste number
   = threshold
0.5
                                    halof orm
                                      comp'ds.- CHCI3
                                         1.3  14
   13-after GAG  adsorption,- 14-after ozonation


   Figure 93. Process 3  performance  parameters
               at  Morsang plant.

                   (Richard,  1978)
                        353

-------
5-
4-
3-
2-
1-
100-
50-
5-
4-
3-
2-
1-
1.0-
0.5-
0-
org. m
1
or
fl
1
org
nniaHIHB
1
BKH

5
at is.- r\mnu4 -
03 4-
G 3-
10 ^ 24

g. mat'ls..
removal 1O

5
uor<
. m<
5
•BH
rr ^ 10"

iscence 1Q-
8-
6-
4.
8 1C 15 2.
it Is. by UV 5Q_
40-
30-
	 20-
•BaaMABBBMBM 0-
1
TOC
G
A
10 C
8 \* 1S
taste number
th=threshold
1

5 8 10 15
humic acids
1

5 TV0 15
haloform comp'ds.
CHCI3
1 8 10 IS
  5-after clarification! 8-after sand filtration;
  10-after  ozonation; 15- after GAC adsorption
Figure 94.  Process 4  performance parameters
           at Morsang  plant.

                 (Richard, 1978)
                      354

-------
    to
  PARIS
NOTE.
            indicates sampling point
     Figure 95.
Analytics" sarpling points for the 4 water treatnent processes
  at Horsang-sjr Seire p'ant, Frarce
                                    355

-------
CO
en
CTl
TAB
IF 87 COMPARISON OF PROCESSES 1, 2, 3 AND 4 AT MOR'ANG SUR SEINE DATA AVERAGED OVER 5 MONTHS OF SIDE BY SIDE OPERATION.
Parameter
Organic
Materials
(by KMnOj)
TOC
Raw
Water
3.8
mg/1
3.0
mg/1
Taste Threshold 12
Fluorescenc
Organic
Materials
(by UV)
Humlc
Acids
CHCla
i 3.9
0.9
7
mg/1
1
m1crog/1
Finished Water Parameter by Process
Process 11
1.6 mg/1
1.2 mg/1
1
0.8
0.2
1.8 mg/1
26 mlcrog/1
Process 12
1.4 mg/1
1.2 mg/1
none perceptable
0.4
0.2
1.0 mg/1
30 m1c rog/1
Process #3
1.3 mg/1
1.1 mg/1
2
0.5
0.2
1.2
26 ralcrog/1
Process 14
0.8 mg/1
0.9 mo/1
none perceptable
0.2
0.05
0.5 mg/1
1 mlcrog/1
               Note:  Post-chlorln-
               atlon (0.5 0.6 mg/1
               dosage) follows all
               treatment processes
               for distribution
               systems
process  II.
 chlorlnatlon,
 coagulation,
 flocculatlon,
 clarification  (pow
 dered activated
 carbon),
 sand filtration,
 ozonatlon
                                                                Process  »2:
chlorlnatlon,
coagulation,
flocculatlon.
clar1f1cat1on(w1thout
powdered activated
carbon),
sand filtration,
ozonatlon,
GAC filtration
                      Process 13:
chlorlnatlon,
coagulation.
flocculatlon,
clarification (without
powdered activated
carbon),
sand filtration,
GAC filtration,
ozonatlon
Process 14:
coagulation,
flocculatlon,
clarification (without
powdered activated
carbon,
sand filtration,
ozonatlon,
GAC filtration

-------
 FUTURE  PLANS AT MORSANG

     The  pilot plant study of the process without prechlor1nation was to
 have been continued through 1979, then a decision was to have been made as
 to the  process to be installed in the third stage of the first star.  A
 recent  presentation (Fiessinger & Montiel, 1980) implies that the decision
 will be to install Process #4.

     Granular activated carbon adsorption has been proven to be an efficient
 process step and is being retrofitted into the first stage of the star
 (Process  #1) followlr, ozonation.

     When 3AC adsorption in Process #1 becomes operational in 1979, powdered
 activated carbon was to be eliminated as a permanent treatment step, but
 will be maintained in the ready state for discontinuous addition in the
 event of  incidents of pollutlonal surges.

     After the first stage at Morsang has been retrofitted with GAC adsorp-
 tion, there will  be sufficient GAC installed in the South Paris region
 plants of SocietS Lyonnaise des Eaux et de TEclairage (in the Viry, Vigneux
and Morsang plants) to warrant a central  GAC regeneration facility.  Plans
for constructing such a facility were being considered during 1978.

     The Degr§mont water treatment plant at Crly (in Paris), which currently
uses ozone treatment,  also is being retrofitted with GAC adsorption to
follow after ozonation.

BIOLOGICAL REMCVAL CF ORGANIC MATERIALS BY GRANULAR ACTIVATED CARBON -

     RESULTS OF A MODELING STUDY AT MORSANG-SUR-SEINE

     In 1977, Benedek reported a study conducted at the Morsang-sur-Seine
 treatment plant using data obtained from Processes #ls  #2 and #3 of the
full-scale plant.  The objective of this  study was to try to establish the
mechanism(s)  of organics removal  in the biologically active activated carbon
adsorbers and, thereby,  to attempt to develop application and design princi-
ples for the ozonation/granular activated carbon treatment combination.
This study will  be reviewed during this discussion because it presents
further insights  into the 15-month study  previously conducted by Richard &
Fiessinger (1977).

     A mathematical model for describing  activated carbon particles with
 biological activity has been developed by Maqsood and Benedek (1977) and by
 Peel and Benedek (1977).  The basis of this model Involves mass transfer 1n
a differential SAC column segirent containing percolating water, a bacterial
 "film" and activated carbon.  The concept of a uniform bacterial  "film"  was
 presented to allow simplification of the mathematical expressions derived.

     The major required Inputs for the model were based on batch tests for
 (a) adsorbable Isotherms for the carbon surface equilibrium reaction (with
 adsorbable solutes) and (b) batch adsorption kinetics of solutes.  In addir
 tion, an  estimate was -nade of the bacterial degradation rates of solutes,


                                      357

-------
and these are expressed as zero order, carbon surface-based reactions in
either batch or continuous column tests.  AdsorbablHty and blodegradability
were assumed to be operative on the same organic fractions.  Finally, the
presence of refractory organics, possibly produced by the bacteria, was
considered to be negligible.

     Respiration measurements were made in a Hach BOD apparatus on pulverized
activated carbon samples, to determine the amounts and rates of blodegradation
of organic materials.

     Isotherms were determined on ozonated and non-ozonated water and
pulverized activated carbon 1n terms of TOC, COD (determined by !
-------




o
o
emoved/g
y
o
H-
0)
0.12-
.
0.10-

-
0.08-
.
0.06-
«
0.04-
0.02-1
J
a
Q
m
n a
a °
a n
a n
S Q
o
B 0
1 °
on
t
• a a with 03
o
D without 03
13
&
      0.02  0.06  0.10  0.14   0.18   0.22   0.26   0.30
                 g TOC applied/g GAC

Figure 96. Effect of  ozone on GAC adsorption at
            IViorsang  plant.

              (Eenedek. 1M77)
                       359

-------
     Bacterial respiration rates also were similar, although more activity
was observed at the top of the adsorber receiving ozonized water.  Respiration
rates corresponded to the degradation rates of the organic materials.

     In these modeling studies, Benedek (1977) used the isotherms and
kinetic constants determined at Morsang to predict breakthrough curves, as
well as cumulative adsorption curves.  Figure 97 is a plot of C/C0 versus
the throughput parameter, I, which represents the ratio of substrate fed to
the adsorptive capacity of the GAC.

     The non-adsorbable component of the feed immediately broke through
(Figure 97).  Thereafter, the concentration of solute in the effluent
remained unchanged until a value of Z = 0.7 was reached (27 days for the
ozonized feedwater).  Subsequently, complete breakthrough occurred in an
additional 27 days.

     On a cumulative plot (Figure 98) using a GAC adsorber containing no
biological activity, the adsorption curves rise rapidly and thereafter level
off as no more adsorption occurs.  The lower capacity of the activated
carbon for TOC in the ozonized water may be caused by ozonation rendering
the TOC more polar and less readily adsorbed.

     With biological activity present in the GAC adsorber, cumulative plots
showed a non-zero slope after adsorptive exhaustion.   Figure 99 shows that
TOC removals are very sensitive to the assumed values of Rf, the biological
reaction rate.  Moving from the observed value of Rf =  2.6 x 10-11  g/sq
cm/sec to 5.0 x 1Q-11  would result in virtually complete elimination of
adsorbable and biodegradable organic materials in the feed waters.

     The cumulative removals expected on the basis of Benedek1s modeling
studies for the Morsang GAC adsorbers used with preozonized versus non-
ozonized water are shown in Figure 100.  Up to an applied TOC loading of 0.23
g/g of GAC, the GAC adsorber which followed non-ozonized water would have
adsorbed more organic materials (both biodegradable and non-biodegradable).
After 0.23 g/g of GAC applied TOC loading, however, the GAC adsorber which
followed ozonized water would be superior in all respects (because of biodeg-
radation occurring 1n the GAC adsorber), although under the conditions of
Process #3 at Morsang, only slightly more so.

     Ozonized water data (Figure 96) and prediction (Figure 100) are compared
in Figure 101.  The apparent agreement is acceptable, although the data
should match reasonably well  since Rf was obtained from the data of Figure
96.  This agreement shows that the model is capable of matching operational
data.
                                      360

-------

     t.oo-
     0.90-
     0.80-
     0.70-
     0.60-
     0.50-
     0.40-
     0.30-
     C.2O-
     0.10-
       0
         0  O.2 0.4 0.6 0.8  1.0  1.2 1.4 t.6  1.8 2.0 2.2 2.4
                throughput parameter, Z

Figure 97.  Breakthrough curve for the ozonated GAG
           adsorber without biological activity  (DQ=2.8 x10  cm2/sec).
               (Benedek,1977)

-------
co
a\
ro

O
0
CO
T3
93
remoi
O
O

O.14-
0.12-
0.10-
0.08
0.06
O.04
0.02






without O3
S^ with 03
'
                            0-020.04°-060.080-10 0.12 °'14 0.16 °-180.20°-220.24



                                   g  TOC applied/g  GAG




               Figure 98.  Cumulative plots* GAG adsorption vs. loading 5  with


                           and without ozone (DQ= 2.8 x 108cm2/sec,- Rp-0.0)





                                      (Benedek, 1977)

-------
CO
cr>
CO
                O
OB
^

TJ

V
»
O

E

4)
^


O

O

!-



Oi
0.20-



0.18-



0.16



0.14-



0.12-



0.10-



0.08-



0.06



0.04



0.02-



   0
                                                                       -11
                                                          = 0.0
                           ° °-020.04a06O.Oa010 0.12 °'14 0.16 °-180.20°-220.24


                                    g  TOC  appiied/g GAC



                 Figure 99. Effect of bloiogica! activity on the cumulative TOC


                            removal of a GAC adsorber.
                                       (Benedek. 1977)

-------
     u
     <
     o

     a
      O

      E
o
o
H

co
0.20-



0.18-



0.16-



0.14-



0.12-



0.10



0.08



0.06



0.04



O.O2
                           without O3
                                        03
                0 0.020-040.060-080.100-120.14°.160.180-200.220-24

                         g TOO applied/g GAC




Figure 100. Comparison of ozonated* and non-ozonated" GAC


            adsorbers ('Rf=2.1 x10"11i "Rf = 1.8 x 10'11).



                         (Benedek,1977)
                              364

-------
      0.12-
O
<
a
•o
V
>
o
E
o
o
H-

O>
computer prediction
          0 o.o#-04ao6Q08aioa12o.i4°-16o.i eW


                   g TOG applied/g GAG



Figure 1O1.  Comparison of  actual vs. computer-predicted data


            for ozonized water/GAC adsorber.
                       (Benedek,
                             365

-------
                                  APPENDIX E

            THE DOHNE PLANT AT MULHEIM, FEDERAL REPUBLIC OF GERMANY


 BACKGROUND

      From 1962 until April, 1977, the Donne plant at Mdlheim processed Ruhr
 River water directly (without sand bank filtration) by the following sequence:
 breakpoint chlorination  (for the removal of ammonia), chemical coagulation,
 f"occulation, clarificaticn, filtration, dechlorination (by passage through
- GAC columns), ground passage and chlorination for residual before introduction
 into the three MUlheim distribution systems.  Prior to 1962, the Ruhr at
 Mulheim was treated simply by slow sand filtration followed by chlorination.

      German law prohibits discharges of industrial wastes into the Ruhr
 because the Ruhr is used as a drinking water supply by a number of cities,
 but allows the discharge of sewage (which is biodegradable).  Local  industrial
 wastes are transported 60 to 80 km north of MUlheim to the Emscher River for
 discharge.  Treatment of the Emscher for potable water applications  is
 prohibited by German law because of the allowed industrial discharges.

      Ammonia levels frequently rise to as high as 4 to 6 ng/1  because of the
 prevalence of sewage in the Ruhr.  This results in prechlorination dosages
 as high as 50 mg/1  for conversion of ammonia-nitrogen to elemental  nitrogen.
 German law allows a maximum level of 0.6 mg/1  of ammonia in finished drinking
 water.   Higher levels will require addition of more than 0.6 mg/1  of chlorine
 to produce a stable chlorine residual.  A maximum dosage of 0.3 mg/1 of
 chlorine is allowed by German drinking water laws, but this can be  raised to
 0.6 mg/1 dosage in  exceptional  circumstances,  with prior approval  of health
 authorities.

 THE DOHNE TREATMENT PROCESS, 1970 to 1977

      High prechlorination dosages were employed (10 to 50 mg/1) to reirove
 ammonia.  These chlorination levels produced high levels of chlorinated
 organic materials (above 200 m1crog/l) which were not easily removed by SAC
 "dechlorination"  or by subsequent ground passage.  Table 88 shows  the old
 Dohne water treatment process and the median levels of Dissolved Organic
 Chlorine (DOC1),  Dissolved Non-polar Organic Chlorine (OOC1N),  total haloforms
 and amounts of chloroform produced over the last two years that the  old
 process was employed (1975 and 1976)  (Sontheimer e_t al.,  1978). Maximum
 DOC! values as high as 0.5 mg/1  (500 microg/1) have been observed  at Dohne
 under extreme conditions.
                                      366

-------
     During  the  period 1969 through 1972, detailed pilot plant  studies  had
been conducted at  the  Bremen,  Gernany water works on the biological  activated
carbon process (Eberhardt,  Madsen & Sontheimer, 1974; Eberhardt,  1975).   The
BAG studies  at Bremen  showed that amnonia can be "removed" biologically
(converted to nitrate)  rather  than by employing breakpoint chlorination  with
its attendant production  of chlorinated organics.  German drinking water
regulations  allow  a maximum of 50 mg/1  of nitrate in finished water.
      TABLE 88. OLD TREATMENT PROCESS AT 'HE COHHE PLANT, MULHEIM, GERMANY, AND MEDIAN VALLEr DF
                  CHLORINATED ORGANICS PRODICEO "HJRISG 1976 AMD 1977


Process Stage
taw water (10
mg/1 SS,
mostly
organics
Chemical
addition


Hocculatlon S
sedimentation
Sand filtration
iranular
activated
carbon
'.round passage
Safety chlor1rat1on


Treatnent
—



10 50 mg/1 Clz
4 6 Ttg/1 Al+3
0.1 '
-------
      The pilot plant testing  program was  continued until mid-1978,  to  check
 the full  scale plant operations.  At all  times, the performances of the full
 scale plant  have  paralleled those of the  pilot plant.  Through 1979, the GAC
 columns  have not  had to be regenerated and, based on the length of  time that
 the pilot plant GAC  columns functioned without regeneration, it is  predicted
 (Sontheimer,  1978) that the full scale plant GAC columns should operate up
 to  at least  two years before  regeneration will be required.

      The  Dohne plant can process 48,COO cu m/day (12.7 mgd) of Ruhr River
 water (without sand  bank filtration), and is located in the center of a
 residential  neighborhood; in  fact, 4 residential back yards butt against the
 plant property.   Breakpoint chlorination had been employed as a pretreatment
 step  since 1962,  and because  of the residential location, special  attention
 had been  paid  to  guard against chlorine leaks or spills,.

 THE DOHNE PILOT PLANT PROGRAM (Jekel, 1977)

      This pilot plant facility began operating at the Dohne plant in Septem-
 ber,  1976, under  the direction of Prof. Dr. H. Sontheimer.   Dr.  Martin Jekel
 of the Engler-Bunte  Institut der UniversitSt Karlsruhe carried out the pilot
 plant investigations at the Dohne water treatment plant.  The BAC  process
 currently being used on full-scale by the Dohne plant is based upon the work
 performed by Dr. Cekel in this pilot plant.

     A schenatic diagram of the Cohne pilot plant is  shown  in Figure 102.
 Ruhr  River water  (without being treated by river sand bank  filtration)  was
 saturated by air using an injector,  then treated with po1y(aluminum chloride),
 clarified in a Pulsator, and the Pulsator effluent ozonized in a  bubble
 column using ozone generated in air.

     The major portion of the ozonized water was filtered directly through  a
 two-stage sand filter then passed through GAC column  adsorbers.  The pilot
 plant filters consisted of two groups of two sand filters each and the
 adsorbers consisted of three single  GAC contactors,  each filled with 2.5 m
 of GAC.

     A minor portion of the ozonized water was mixed  with a small  amount of
 aluminum  sulfate  (2.5 mg/1 = 0.2 mg/1  Al+3), filtered through a  double  layer
 sand  filter, then passed through two GAC adsorbers.

     A parallel pilot plant process  was operated in which full-scale plant
water which had been treated by breakpoint chlorination and flocculation
 then was treated by ozonation, sand  filtration and GAC adsorption.

     All GAC adsorber/contactors were operated at flow rates of  10 cu  n/sq
rn/hr, which was equivalent to an empty bed contact time of  15 minutes.   Five
different commercial  grades of GAC were tested in the Dohne pilot  plant.

      In addition to the two parallel pilot plants described above,  two
 fluidi/ed bed reactors were set up for studying the  biological  oxidation of
 ammonia to nitrate.  One of these reactors was supplied with air-saturated
 Ruhr River water,  the other with ozonized effluent from the flocculator.


                                     368

-------

air
Inject 01
~>






J r^ J
(.> ^^^^
£






dec
1f*n
ve

k.
o
4-t
u
s
t
f




O> Ruhr
U3 River
(raw water)

tossing and
•ssure
sel
3











"1r — s.
Injector '
Ruhr — ££V-
a
V
flocculating ^
agent '









r





























itxlng
tth aj









chamber
1 tat ion
Y


































u
lA
Q.
I-
3
3
U
U
o
I*.



i


River v
(raw water)

Figure U2




t-
u
a
t
X
V

ozone
gassing and
reaction
(Col unns/









.




















-
•*





i
1
ozone













ozone
jjenrtr












_l




^
\J





dual granular
iredia activated carbon filters—. 	
filter / / 7 / ^^ 	 --^


1
!












floccu^ting
agent — /
Scrematlc diagram of Donne pilot plant process (MUlhelir)


W/ft
ifm
I
'////.
M
W
^J
P

f-





O
c




r







J

x








^^




















.










"^
T








*"*













~\



















p

P






J




""

•#



i
I
V
—
_

















-
f


Source   Jekel,  1977.

-------
The purpose of these studies was to obtain information concerning collapses
or poisonings of the blomass which had been noted during the proceeding
year.

Comparison of Treatment With & Without Breakpoint Chior1nation

     Table 89 compares the performance of the two pilot plants with and
without breakpoint chlorlnatlon.  These plants were operated side by side
under the same processing conditions.  Data shown in Table 89 are the mean
values of DOC (Dissolved Organic Carbon) obtained during October and November,
1976.  Prechlorination led to almost no reduction In the initial 5.3 mg/1
DOC content after flocculation, whereas without prechlorlnation, a 1.4 mg/1
reduction in DOC level was obtained after flocculation.  In later treatment
steps (ozonation, filtration, GAC adsorption), the pilot plant without
prechlorination produced better DOC removals than did the pilot plant with
prechlorination.

TABLE 89.  COMPARISON OF TREATMENT WITH & WITHOUT BREAKPOINT CHLORINATION
     Treatment Step
Ruhr River water

After flocculation

After 03 & sand filtration

After GAC adsorption
                                        DOC, mg/1
WithU2  without Cl2
  5.3

  5.3

  4.7

  3.4
5.3

3.9

3.7

2.4
     Mean values of pilot plant data obtained during Oct. & Nov., 1977

          Source:  Jekel, 1977
     The higher DOC concentration in the prechlorinated water caused an
early breakthrough of the GAC adsorber, as shown in Figure 103.   This
figure shows the percent reduction in concentration of organic materials (as
measured by UV absorption at 254 nm) versus the amount of water passed
through the GAC for the two process waters.  The upper curve (with biological
treatment in the GAC) was extended at least up to a flow of 40 cu m/1 of
GAC, without any significant decrease in DOC removal  efficiency.   This
amount of flow corresponded to a GAC column operating time of more than 1
year.

GAC Adsorption Versus Biological Activity

     Fable 90 compares the mean DOC concentrations in GAC column  effluents
obtained during the first three months of use (during which time  COC removal
was primarily by adsorption) with the COC concentrations obtained during the
next 8 months of operation (during which time DOC removal  was primarily by

                                      370

-------
biodegradation).   Significant  differences were observed for the five  activated
carbons during the initial  adsorptive period.  On the other hand,  during  the
biological period which followed, few differences in :OC removals  were
observed.  Only the F-400 GAC, which has a smaller grain diameter,  performed
better than the other four  carbons.  This indicates that biological activity
does not depend upon adsorptive capacity, but may depend upon  grain size  of
the GAC.          10Q.
3  =
             c c
            •o «-
            4, x
            t- «>
                                   without  breakpoint
                                          chlorination
                          with  breakpoint
                                chlorination
                         2   4   6   8  10  12  14  16  18 20

                          water/liter of GAC,
           Figure 103. Influence of prechlorination on the
                        effectiveness of a process consisting
                        of flocculation, sedimentation, ozona-
                        tion, filtration  and GAC treatment.

                                 (Sontheimer et,a±., 1978)

     Figure 104 shows  the  removal perfornances of the pilot plant  GAC adsor-
bers in g of DOC/cu  m  of 3AC/day and the total  organic  loadings of the
various types of GAC tested during the first 18 months  of operation  (Jekel,
1979).   These data were interpreted by Cekel (1979)  as  indicating  that the
breakthrough behavior  of biologically active GAC adsorbers is  influenced,
among other factors, by the adsorbed biologically resistant organic  substances.

     Figure 104 also shows that an average of about  75% of the total organic
materials removed is biologically oxidized, while the remainder, primarily
biologically-resistant materials, is removed by adsorption.  Due to  the
enrichment of these  compounds on the GAC surfaces, recognizable by the
increased GAC loading  values in the lower section of Figure 104, the total
removal performance  of the GAC adsorbers clearly falls  off, particularly
during the third half-year period of operation.

     Ouring this third half-year period, a clear improvement in the  quality
of the raw water occurred  in the Ruhr river, the influent COC  decreasing
from about 2.5 to 1.8  mg/1.  This intensified the observed decrease  in 3AC
performance.   Nevertheless, Cekel (1979) concluded that the operating time
of biologically active GAC adsorbers for the treatment  of Ruhr river water
                                     371

-------
175---T-  r-i
«  100--
1 w
5 C
E a

o M 50--
t, -
                                 a
(Adsorption


,Biological

 Oxidation
«JL
^ eo-
g 60-
yj 40-
= 20-
s
• •
0 0
C
»






—

^^

—





^™ '



^"™

—


""""











—


^™*





        1st half-jr  2nd half-vr 3rd half-y




  Figure 104.  Performance & loading  of

                BAG  adsorbers.


                (Jekel.1979)
                     372

-------
at the Cohne pilot plant was prolonged by a factor of approximately 4 1n
comparison with GAC columns operating purely by adsorption.
TABLE 90.  COMPARISON OF DOC REMOVED DURING FIRST 3 AND SECOND 8 MONTHS OF
           USE OF GRANULAR ACTIVATED CARBON FOR 5 ACTIVATED CARBONS
Treatment Step
After sand filtration
After GAC contacting
LSS 2.5 m
LSS 5.0 m
ROW 2.5 m
ROW 5.0 m
NK-12 2.5 m
F-400 2.5 m
BKA 2.5 m
DOC, mq/1
10-12, 1976
(adsorption)
3.5

2.2
1.2
1.9
1.0
2.4
1.6
1.9
1-8, 1977
(biological)
2.6

1.7
1.1
1.8
1.3
1.8
1.5
1.8







Source; Jekel , 1977
     Ozonized clarifier effluent treated with 2.5 mg/1  of aluminum sulfate
then passed through LSS GAC during the adsorption period (October to  December,
1977) resulted in further reduction in DOC levels, due  to precipitation of
insoluble alum saHs of (probably) organic acids, formed by oxidation during
ozone treatment.   Pertinent data are compared in  Table  91,  which  also show
that a deeper GAC column removed nore DOC.

Chemical Balancing of Biological Activity

     It was concluded that biological activity in an  aerobic,  filter  -
adsorber system leads to:

     1)   biological degradation of dissolved organic compounds  (DOC)
     2)   the formation of C02 (from which the amount of inorganic carbon
          formed was calculated) from the DOC,
     3)   oxidation of ammonia to nitrate and
     4)   consumption of dissolved oxygen by both processes.

     Measurement of these three parameters yielded information as to  the
rate of biological activity occurring during passage  through GAC  adsorbers.
Each mg/1 of DOC removed biologically consumed 1  mg/1 of dissolved oxygen
and produced 1 mg/1 of inorganic carbon (calculated from the amount of
                                     373

-------
measured).  Each mg/1 of ammonia oxidized used 3:6 mg/1 of DO (based upon
the equation Nfy* + ZOg—*-N03' + 2H+ + H£0) and produced 3.4 mg/1 of
nitrate ion.

TABLE 91.  REMOVAL OF DOC AFTER TREATING OZONIZED WATER WITH Al+3
     Treatment Step
After flocculation + 03 +
     alum + sand filtration

After GAC adsorption

     LSS       2.5 m

     LSS       5.0 m
DOC, mg/1
10-12, 1977
(adsorption)
     3.2



     1.8

     0.9
Mean values of data obtained during the period of study

          Source:  Jekel, 1977
     Table 92 shows the mean values for reduction in levels of DOC,  NH4+ and
DO and mean values for increases in inorganic carbon found over the  period
vlanuary to March, 1977.  During this time the nrean water temperature was
6.8°C.   Nearly all DOC removed in the first 2.5 m of GAC columns was found
as inorganic carbon in the effluent.  The ratio, -ADOC/+A(inorganic  carbon)
was unity or slightly higher for all GACs tested, except for the NK-12,  for
which the ratio was 0.83.  A ratio of less than 1 indicates that nearly  all
of the DOC being reiroved was being degraded biologically.   A ratio of about
1 indicates that the BAC system was in biological equilibrium.

     However, in the next 2.5 m of activated carbon (see data for 5  m SAC
columns), the ratio -ADOC/+A(inorganic carbon) was significantly greater
than 1.  Therefore, more DOC was being removed from solution than was being
degraded biologically.  Since the data for 2.5 m GAC column heights  show
that biological equilibrium had been attained (rate of COC removal was equal
to the rate of inorganic carbon production), the additional DOC being removed
by the 5 m high columns must have been a result of adsorption in the lower
half of the columns, in which biological equilibrium had not yet been attained.

     Data in Table 92 also show that the ammonia levels were reduced to  near
zero in the first 2.5 m of GAC depth without showing any significant perfor-
mance differences between 3AC types.  Increasing the GAC depth to 5m did
not increase the amount of ammonia converted to ftfa* J1?;!'}"^-*
means that nearly all of the ammonium ion was oxidized in the first  2.5  m
depth of GAC.
                                      374

-------
TABLE 92.  BIOLOGICAL ACTIVITY IN PILOT PLANT GAC COLUMNS, JAN.-MAR., 1977
Treatment Step
After GAC adsorpffo
LSS 2.5 m
ROW 2.5 m
NK-12 2.5 m
F-400 2.5 m
BKA 2.5
LSS 5.0 IT
ROW 5.0 m
ADOC
mg/1
i
1.1
1.2
1.0
1.3
1.2
1.6
1.7
A (inorganic
carbon)
mg/1
1.0
1.1
1.2
1.2
0.9
1.2
1.3
iboc/A(in-
org C)
l.T
1.09
0.83
1.08
1.33
1.33
1.31
ANH4*
IP9/1
1.43
1.41
1.45
1.43
1.44
1.46
1.47
A02
mg/1
7.2
7.2
7.1
7.1
7.1
7.7
7.7
Sand filtrate contained 2.6 mg/1 DOC and 1.53 mg/1 NH4+
Mean water temperature: 6.8°C
Mean values of data obtained over the 3 month period
Source: Jekel , 1977

     Table 93 shows similar data obtained during the summer period of June
to August, 1977 when the mean water temperature was 19.5°C.  The feed to the
GAC adsorbers was ozonized water from the full-scale plant after floccula-
tion and sand filtration.  As a result, the ammonia concentration was zero
and the pilot plant GAC columns thus showed no nitrification activity.
Comparison of the amount of DOC removed with the amount of inorganic carbon
produced at 19.8° (Table 93) or at 6.8°C (Table 92) showed no significant
differences because of water temperature.

     However, it j_s_ significant that during the June to August, 1977 period,
the ratio of ADOC/A(inorganic carbon) was slightly less than 1  for the
activated carbons NK-12, F-400 and BKA (at 2.5 m bed depths) and for LSS and
ROW activated carbons at 5.0 m depths.  This Indicates that more DOC was
being converted to Inorganic carbon than was being removed from the inf'uent
solution, which implies that biological regeneration of the GAC was occurring.
The biological reactivation of GAC is believed to have been demonstrated
unequivocally later in the full-scale plant (see Figure 108, later in this
Appendix).
                                      375

-------
.TABLE  93.   BIOLOGICAL ACTIVITY  IN  PILOT 6AC COLUMNS, JUNE-AUG., 1977
Treatment Step
After GAC Adsorpti
LSS 2.5 m
ROW 2.5 m
NK-12 2.5 m
F-400 2.5 m
BKA 2.5 m
LSS 5.0 m
ROW 5.0 m
ADOC
mg/1
on
0.9
0.9
1.0
1.2
1.0
1.3
1.3
A (inorganic
carbon)
mg/1
0.8
0.8
1.1
1.3
1.2
1.4
1.4
ACOC/A
(inorg C)
1.13
T.13
0.91
0.92
0.83
0.93
0.93
mg/1
0
0
0
0
0
0
0
AO?
mg/1
2.8
2.7
2.4
3.0
2.0
3.5
3.5
Sand filtrate contained 2.8 mg/1 COC and ^ess than 0.05 mg/1 NH^"1"
Mean water temperature: 19.5°C
Mean values of data obtained over the 3 month period
Source:

Jekel, 1977
     Table 94 presents data obtained over the 6-month period (January to
June, 1977) after a 3-month starting period,  "he ratios of £.DOC/A(inorganic
carbon) were always greater than 1, except for NK-12 activated carbon (at
2.5 m bed depth).  These data, gathered over double the "length of time than
those of either Table 92 or Table 93, indicate that slightly more DOC was
being removed than was being converted into inorganic carbon.  Therefore,
slightly more adsorption of DOC was occurring than was biological degradation.

Bacterial Counts in the GAC Adsorber Effluents

     Table 95 shows total bacteria counts in the effluents of a 2.5 and 5.0
m high GAC column with pilot plant process water involving no prechlorination.
The data indicate that as biodegradable DOC was removed in the 3AC columns,
the number of bacteria in the effluents decreased.  This shows that it is
possible to convert DOC into biodegradable organic materials which then can
be biodegraded in the water treatment plant (in the GAC contactors);   The
greater the amount of biodegradable DOC that can be removed in the treatment
plant, the less will be the amount of biological aftergrowth which can occur
after safety chlorination and after the chlorinated water is sent to  the
distribution system.  Such treatment practice also can be expected to lower
the amount of chlorine required to provide a measureable chlorine residual
in the distribution system.
                                      376

-------
TABLE 94.  BIOLOGICAL ACTIVITY IN PILOT GAC COLUMNS.  MEAN VALUES FOR A
            6-MONTH PERIOD (JAN-JUNE. 1977) AFTER A 3-MONTH STARTING PERIOD
<\fter
SAC
LSS
LSS
ROW
ROW
NK-12
F-400
BKA
Bed
Depth
2.5 m
5.0 m
2.5 m
5.0 m
2.5 m
2.5 m
2.5 m
A(in-
ADOC org C)
mg/1 mg/1
0.92 0.83
1.69 0.96
. 1.09 0.97
1.59 1.05
0.99 1.36
1.26 1.11
1.00 0,97
Source: Sonthelmer et al.
-


ADOC/A(in-
org C)
1.11
1.76
1.12
1.51
0.73
1.14
1.03
ANH4*
ng/1
1.31
1.34
1.31
1.34
1.28
1.32
1.28
, 1978

A02
mq/1
6.32
6.67
6.49
6.71
6.03
6.95
5.99







!
|
" . " - -i
TABLE 95.  TOTAL BACTERIAL COUNTS IN BIOLOGICALLY ACTIVE GAC ADSORBERS
     Treatment Step
<\fter sand filtration

After GAC Adsorption0 2.5 m

After GAC Adsorption, 5.0 m

     M  = geometric mean value
MG
          On
6,040

  747

  253
        4.9

        8.1

        4.8
        = geometric standard deviation

          Source;  Sontheimer et a1.s 1978
Fluidized Bed Reactors For Ainnonia Oxidation

     During the suraner of 1976S pilot plant studies were conducted at the
Donne plant using air-saturated Ruhr River water passed upflow through
fluidized sand beds for biological removal of amrronia rather than breakpoint
chlorination.  Aeration provided sufficient dissolved oxygen to allow
nitrification of all aononia originally present at the time.  This technique
has been developed at the Water Research Center at Medmenham, England.  The
biological activity was very effective during the first months but then two
                                      377

-------
 collapses  of the  biomass  occurred.   High concentrations of heavy metals were
 found  1n the fluidlzed  sol Ids, adsorbed or  incorporated from the river
 water,  and these  were suspected of having poisoned the nitrifying biomass.

     A  new fluidized bed  pilot plant reactor was set up and was fed with
 ozonized effluent from  the full-scale plant (no breakpoint chlorinatlon).
 Heavy metals were removed effectively during the flocculation step.  The
 fluidized  bed reactor was operated during the June to August, 1977 period at
 the rate of  10 m/hr and the mean water temperature was 19.5°C.

     Mean  data obtained during this period are shown in Table 96, which also
 shows data obtained from aerated Ruhr River water without flocculation and
 ozonation.   The level of ammonia decreased with no collapses or poisoning of
 the biomass using the flocculated and ozonized water over a period of 1
year.  In addition,  about 1 mg/1  of DOC was removed biologically in each
 reactor.  The amount of dissolved oxygen utilized corresponded to that
 expected from the biological  oxidation reactions observed.

 TABLE 96.   BIOLOGICAL OXIDATION IN FLUIDIZED BED REACTORS
Treatment Step
Ruhr River water + aeration
After amnonia reactor
Decreases in levels
After clarification + 03
After ammonia reactor
Decreases in levels
NH4+
mg/1
0.91
0.09
0.82
0.39
0.05
0.34
DOC
mg/1
4.5
3.5
1.0
3.1
2.2
0.9
DO
mg/1
8.5
2.6
5.9
9.1
4.6
4.5
Flow velocity: 10 m/hr
Mean water temperature: 19.5°C
Mean data values obtained over the 3 month period
Source: Jekel, 1977
                                     378

-------
 THE BAG  PROCESS AS  INSTALLED, APRIL  1977

      Based  on  the successful early pilot plant results with the new ozonatlon
 process,  the breakpoint chlorination step at the full-scale Cohne plant was
 eliminated  in  April, 1977 and ozonation of the clarifier effluent was begun.
 Dohne's  treatment process as 1t was  instituted in mid-April, 1977 is shown
 in  Figure 105  (Sontheimer et al., 1978).  Ruhr River water, which contains
 about  10 ng/1  of suspended soTTds (mostly organics) is pumped into a flash
 mixing tank {1.8 x  1.8 x 5.2 m) which contains a high speed aspirating
 turbine which  uses  2.5 kw of power per cu m of tank volume.  Water is dosed
 with 4 to 6 mg/1 of poly(aluminum chloride) and an average of 1 mg/1 of
 ozone.  Most of the ozone used in this preozonation step is obtained frorr
 the off-gases  from  the second (major) ozonation step.   In fact, all  of the
 off-gases from the  second ozonation step are drawn into the flash mixing
 tank by the suction action of the aspirating turbine.   Additional  quantities
 of fresh ozone, as  required, are added to these off-gases.   Additional  ozone
 is required whenever the manganese content o* the raw water is  high  (see
 below) or when algae levels rise in the river (Spring  and Fall).

     Residence time in the flash mixing chamber is about 0.5 minute.
 During this time,  nearly all  of the ozone added in this chamber is utilized
 in performing various oxidation functions.   Only 3% of the  ozone  originally
 generated is present in the off-gases from the flash mixing chamber.   The
 combination of ozone plus  high  speed  mixing results in inproved flocculation
of suspended solids  in the presence of the  hydrolyzed  aluminum  chloride.

     After flash mixing,  the water is pumped  to a Degremont Pulsator for
clarification where  5 to 15 mg/1  of Ca(OH)2 could be added, if  needed.   The
 Pulsator consists  of 2 basins,  each 27.7 x  13.2 x 4.2  meters in dinension.
Distribution of water in the Pulsator is achieved by means  of perforated
asbestos cement pipes,  which allows  for good  final  flocculation within  the
floe blanket.   Clarification tine  is  1.5 hrs.   By this treatment,  turbidity
of the water is lowered to 0.4  to  0.8 FTU.

     The ozonation system is supplied by Trailigaz, and consists  of  2  ozone
generators, each containing, 220 water-cooled  dielectric tubes and  each
 capable of generating 4 kg of ozone/hr at a concentration of 20 to 25  g of
ozone/cu m of air (8 kg/hr total  ozone generating capacity).

     The ozone dosage at this point in the  process averages 2 mg/1,  the
 contact time is 5  minutes, and  this step is controlled by nonitoring  0.3
mg/1 of dissolved ozone residual  at the exit  of the contact chamber.   Ozone
contacting efficiency in this single  dlffuser chamber  is 90%.   A  further
 reduction of turbidity level  usually  is achieved at this point  by  the  forma-
 tion of larger floes, caused by agglomeration of residual  solids.   In  addi-
 tion, organics are oxidized to  more polar compounds which are capable  of
 being flocculated by aluminum cations and also become  more biodegradable.

     Off-gases from this primary ozonation step (containing 10% of the
 originally generated ozone) are drawn back into the flash mixing  tank  by the
 high speed aspirating turbine and are injected into the raw water.  If the
 manganese content of the water  after the Pulsator clarification step is

                                      379

-------
     poly(a1um1ni.m chloride)
          Ca(OH)7  	
                           **• distribution
Mgure 105.  Process schers for surface water  treatnent  at  trp  ;orne water
             atnity 1n Millheiir after  introduction  of  coirtrneci  chemical  and
             biological  oxidation.
                    (Sonthener et. al_., 1978)
                                   380

-------
greater  than 0.05 mg/1, then additional fresh ozone 1s added to the flash
mix chamber.  Manganese content is monitored at this point in the process.
Additional ozone also  is required when the algae content of the raw water
rises  (in the Spring and Fall).  During such periods, simply doubling the
preozonation dosage from 1 mg/1 to 2 mg/1 overcomes any tendencies of the
algae  to clog the filters; the treatment process is otherwise unaffected
(Nolte,  1978).

     The use of off-gases from the primary ozonation step in the flash
mixing chamber allows  97% of the generated ozone to be utilized in performing
useful work.  The flash mixer off-gases contain only 3% of the ozone genera-
ted, and this is destroyed by passage through a Degussa catalytic destruction
unit (palladium on alumina) at 60°C before the gases are'discharged to the
atmosphere.

     Ozone is monitored (1) in solution after the primary contact chamber,
(2) in the primary contactor off-gases (by a Hartman & Braun monitor,  Metz,
France) and (3) in the ozoneur room atmosphere (by a Wallace & Tiernan KI
monitor).  This last monitor will  cause an alarm to be signalled at the
control panel If the ozone concentration in the room atmosphere exceeds 0.05
ppm.  In addition, the plant provides DrSger tubes for plant personnel to
determine the ozone concentration nanually at various places in the plant,
if desired.

     After ozonation, the water is pumped to 4 horizontal  pressure sand
filters (1.6 m high — see Figure 106), for filtration at the rate of  '.1  cu
m/sq m/hr (11 m/hr).  It had been planned to dose 0.2 ng/1  of poly(aluminum
chloride) and 0.1  ng/1  of polyelectrolyte after installation of new dual
iredia filters in December,  1978,  if necessary for further turbidity removal.

     After sand filtration, the water is passed through 4 GAC adsorbers,
each containing media which is 2  m high and 5 m in diameter,  at the rate  of
22.5 cu m/sq m/hr.  During the period April to July, 1977,  these carbon
contactors contained the same GAC as had been used for dechlorination  in  the
old process.   Because of this, the GAC was "fully loaded" with chlorinated
organics.  Empty bed contact times of these 2 m high GAC contactors is 5.5
minutes.

     Following passage through the GAC columns, the water then passes  into
15 injection wells for storage in the ground where it remains 12 to 50
hours.  When removed for distribution (from wells within 50 m of the infiltra-
tion points), 0.4 to 0.8 mg/1 of chlorine is added.  This low dosage provides
a stable, measureable 0.1  mg/1 chlorine residual  in the Maiheim distribution
system.  The measureable O.T mg/1  chlorine residual is required by German
drinking water regulations.  However, even the 0.4 to 0.8 mg/1  safety  chlori-
nation dosage is above that normally authorized under German drinking  water
regulations (0.3 mg/1, maximum),  and an exception had to be applied for and
granted to allow this high a dosage to be used at the Cohne plant.
                                      381

-------
     30 meters long, 3.9 meters diameter
     Sand media: 1.6 meters deep, 1.2 to 1.8 mm particle size
     Support Gravel:  Three layers    2 to 3 mm particle size
                                      3 to 7 mm
                                      7 to 15 mm
                            3.9 m
plastic nozzle
                                                       1.6m quartz sand
                                                       support gravel
                                                   ;oncrete
     Filter hydraulic loading rate:
            backwashing rate      :
                                    18 cu m/sq m/hr
                                 :  20 cu m/sq m/hr
    Chemical filter aid:  0.2 mg/1 Al+3
                          0.1 mg/1 polyelectrolyte
    In Decembers 1978, convert media to:
         bottom layer 0.6 to 1.2 mm particle size hydroanthracite
         top layer    2 to 3 mm particle size low density GAC

Figure 106.   Pressure filters at the  Dohne  (MUlheim)  plant.
                                  382

-------
 PROCESS  PERFORMANCE, APRIL-JULY, 1977

      In  Table 97 are listed the Dissolved Organic Carbon (COC), ultraviolet
 absorbances and ratios of UV/DOC measured at various process points for 1975
 (old  process), 1976 (old process) and April to July, 1977 (new process using
 old GAC).  DOC is determined after passing the sample through a 0.45 micron
 filter by oxidizing all dissolved carbonaceous materials with ultraviolet
 radiation and measuring the amount of CO? formed (Wfllfel & Sontheimer,
 1974).

     Ultraviolet absorbance of process waters is measured at 254 ran in 1 or
 5 cm cells, but then is calculated for cell lengths of 1 meter.  This
 measurement relates to the amount of carbon-carbon unsaturatlon contained in
 the molecular structures of the dissolved organic materials.   Unsaturatlon
 can consist of individual carbon-carbon double bonds, aromatic nuclei (multi-
 ple and conjugated C=C bonds)  or carbon-oxygen (C=0, carbonyl) functions.
Ozonation destroys many isolated carbon-carbon bonds and most aromatic
moieties, converting them to COg or to carbonyl  groups,  thus  lowering (but
not totally eliminating)  the total  UV absorbance value.

     The UV/DOC ratio by itself is  not as meaningful as  is  the change of
this ratio 1n coirparison with  changes 1n both components.  If DOC values
remain constant while the UV absorbance decreases,  this  indicates that
unsaturated organics are being converted to saturated, dissolved, oxidized
organic products (which may contain more carbonyl  functions), but not to C02
and water.  If both UV absorption and DOC decrease,  this indicates that
unsaturated organics are  being destroyed and  that  DOC 1s being converted to
C02 and water.

     Table 98 compares the mean DOC values obtained  by the  old process
during 1975 with the mean values obtained during May to  August, 1977.
During this later period, the  GAC adsorbers were operated with the exhausted
carbon used in the old treatment process.  It is apparent that ozoration
enhanced the blodegradation of DOC  during ground passage.  In addition,  it
 is also clear that chlorinated organics were  not being synthesized since the
prechlorination step had  been  eliminated.  Ammonia  levels were reduced from
an average of 1.07 mg/1 in the Influent to zero  after ground  passage.   Some
ammonia (0.25 mg/1) was removed biologically  during  flocculatlon and most  of
 it (0.60 mg/1)  during sand filtration.

 PROCESS MODIFICATIONS, NOVEMBER, 1977

     After the initial testing period of April  to  July,  1977, the following
 plant modifications were made  during August to October,  1977, the plant was
 restarted in November, 1977 and has been operating continuously since:

 1)   The granular activated carbon  depths were doubled from 2 meters to 4
     meters (see Figure 107).   This increased the  empty  bed contact time in
     each filter from 5.5 to 11 minutes.  At  the same time,  each of the four
     GAC columns was charged with a different fresh  granular  activated
     carbon (NK-12, F-300, LSS and  AG-1).  Three of  these four activated
     carbons had been studied 1n the pilot plant.   The intent of the Dohne

                                      383

-------
    TABLE 97.  MEAN DOC AND UV-EXTINCTION VALUES FOR THE DIFFERENT TREATMENT STEPS AT THE DOHNE PLANT
Sampling Point
Raw water (Ruhr)
After floccula-
tion & sedimen-
tation
After filtration
After GAC
adsorption
After ground
passage
1975
T5UCT
mg/1
3.9
3.2
3.2
3.0
1.8
UV
254 nm
m-1
6.8
4.5
4.4
4.0
3.1
Dv/DOC
ratio
1.8
1.4
1.4
1.3
1.8
1976
"DUC~
rig/1
5.0
4.0
3.8
3.7
2.1
UV
254 nm
m-1
9.1
5.5
5.6
5.3
4.0
OY/TJOTT
ratio
1.8
1.4
1.4
1.4
1.9
Aoril-Julv 1977*
DOC
mg/1
3.6
2.9
2.6
2.3
0.9
UV
254 nm
m-1
6.1
3.2
1.8
1.6
1.4
UV/DOC
ratio
1.7
1.0
1.0
0.7
1.6
Nov. 1977-
June, 1978**
DfiC, mg/1
2.4 - 3.7
1.8 - 3.0
1.7 - 3.1
1.0 - 2.6
—
* GAC adsorbers filled with fully loaded GAC, used during old process
** Source: Jekel, 1978, otherwise Sontheimer et al_. , 1978
co
00

-------
                                  ventilation
            4  it
      Sranular
     Activated
      Carbon
3 layer support    »
 (gravel)
   3AC outlet pi
                                                                 .5 -\
                                                             \- raw water
backwash
 water
                                 treated water
                                               Inlet for
                                              backwash air
     3AC backwash rate:   27  to  30  n/nr;  air  ther water;
                         backwash  each  iQ  to "4 days
  Figure  "07.  3AC adsorbers at ;ohne plant, MUlheim,  Feteral  Republic of
               3enany.
                                     385

-------
     plant management 1s to test all four activated carbons in the full
     scale plant, side by side, to select the best performing type at some
     future date.  One-fourth of the sand filtered water is passed through
     each GAC contactor.  Thus the water is in contact with only one of the
     four activated carbons being tested.  After GAC contacting, the effluents
     from the four GAC contactors are blended in the slow sand filters, in
     the injection wells and in the underground reservoir.

TABLE 98.  COMPARISON OF OLD & NEW PLANT PROCESSES FOR REMOVAL OF DOC &
Treatnent Step
Ruhr River water
After flocculation
After 03 + sand filtration
After GAC contacting
After ground passage
old process
1975
DOC
mg/1
3.9
3.2
—
3.0
1.8
new process
Mav-Auqust, 1977
DOC
mg/1
4.0
3.1
2.8
--
0.9
NH4*
mg/1
1.07
0.82
0.22
—
0





Source; Jekel , 1977
2)   Pure oxygen is added after ozonation, at the pumps which send'ozonized
     water to the sand filters.  Some 80% of the nitrification has,been
     found to occur in the sand filters and the remainder occurs  iii the GAC
     contactors.  During the summer,  all  of the nitrification occurs in the
     sand filters.

     Since each mg of ammonia-nitrogen requires 3.6 mg of dissolved oxygen
     for conversion to nitrate (assuming  the equation:  NH4* + 202     NOj-
     + 2H+ + H20), it is apparent that a  raw water ammonia level  of 5 to 6
     mg/1 will require 18 to 22 mg/1  for  nitrification alone.  In addition,
     biological conversion of dissolved organic carbon to C02 and water also
     requires dissolved oxygen (2 mg  of DO for each mg of DOC converted to
     0)2).  Such levels of dissolved  oxygen cannot be supplied solely rrom
     air, which saturates the water during ozonation.   Aeration produces
     maximum DO levels of only 8 to 11  mg/1, depending upon the water tempera-
     ture.  Addition of pure oxygen,  however, can provide DO levels  up to 45
     mg/1.

     Liquid oxygen (LOX) is stored in a 3 cu m tank in the room which used
     to house the chlorine cylinders.  This is vaporized (2 LOX evaporators
     are installed, one for backup) and the gaseous oxygen is introduced to
     the ozonized water at the pumps  which send the water to the  sand
                                      386

-------
     filters.  This pump provides considerable mixing action for water and
     gas.  No special corrosion problems have been encountered during the
     first seven months of operation and no changes in the materials of
     construction of the pumps have been required.

     The DO is monitored at the outlet of the GAC contactors.   Addition of
     gaseous oxygen is controlled by maintaining a level  of 7 mg/1  at the
     GAC column outlets.  Average oxygen dosage to attain this 7 ng/1 residual
     DO level usually is around 10 mg/1.  As a rule, excess dissolved
     oxygen can cause outgassing, which can cause mechanical problems
     during filtration.  However, since the oxygenated water at the Dohne
     plant is under 1.6 bar pressure, no outgassing problems have been
     observed during the first 7 months of operation.

3)   The second poly(a1uminum chloride) * polyelectrolyte addition  step
     (after the primary ozonation step) has not yet been  found to be neces-
     sary.  This is because the flash mixing step utilizing poly(aluminum
     chloride), 1 mg/1 of ozone and high speed agitation  has been so effective
     in lowering turbidity (to 0.1 FTU after sand filtration).   Sigrist
     turbidity meters have been installed after sand filtration to  monitor
     turbidities at this point (see later discussion on head loss vs turbidity
     breakthrough of the sand filters).

     As of June, 1978, the Dohne plant had a total of 41  monitoring points
in the plant.

PERFORMANCE OF MODIFIED BAC PLANT PROCESS

     Plant performance data obtained over the period November 1977  through
May, 1978 are listed on the far right hand side of Table  97.  It is c'ear
that the DOC of water exiting the GAC contactors was lower than that obtained
at the same point during the April to July, 1977 study.   This difference may
be due to one or more of several  factors:

     1)   the Increased EBCT of the 4 m SAC columns (11 minutes vs  5.5)

     2)   the use of virgin GAC in November 1977 vs "fully loaded"  3AC used
          earlier

     3)   higher levels of dissolved oxygen (pure oxygen  was not added
          during the April to July, 1977 tests).  This effect would be
          expected to be minor compared with the first two.

     When the four new GAC columns were started up In November 1977, DOC
values in their effluents were very lows on the order of  1.2 to 1.6 mg/1.
However, these effluent levels increased over the next few months,  reaching
2.1 to 2.6 mg/1 during the May to June, 1978 period.  These increases were
attributed to the Spring algae bloom.  Pertinent data in  this regard are
listed in Table 99.
                                      387

-------
TABLE 99.  DOHNE PLANT PERFORMANCE DATA SINCE NOV.  1977.   DOC DATA, mg/1
Sample Point
Raw water
After 03 + floccln
After sand filtrn
After GAC
-Lurgl NK-12
-Calgon F-300
-Lurgi LSS
-Essen Rsch AG-1
11 -12/77
3.0
2.5
2.5
1.3
1.2
1.6
1.4
1-2/78
3.0
2.7
2.6
1.7
1.0
1.9
1.5
3-4/78
2.4
1.8
1.7
1.5
1.2
1.6
1.5
i
5-6/78
3.7*
3.0
3.1
2.5
2.1
2.6
2.6







* higher COC levels during this period are attributed to the Spring algae
bloom
     Data of Table 99 are Interpreted by Dr.  M.  Jekel  of the Engler-Bunte
Institute of the University of Karlsruhe as  indicating adsorptive  removal  of
DOC during the initial 6 month period of carbon  use.   During this  time,
biological activity was building up to equilibrium very slowly  (because  of
the low water temperatures) and the ratio -ADCC/+A(Inorganic carbon)  was
greater than 1.  The May to June 1978 data indicate that biological  activity
had increased.  The ratio of -ADOC/+A(inorganic  carbon) passed  through  unity
and became slightly less than 1.

     The values for ADOC and A(inorgan1c carbon) are plotted in Figure  108
for the period November, 1977 through July,  1978.   During July, the  rate of
formation of inorganic carbon increased significantly, and the  ratio
-ADOC/+A(inorganic carbon) decreased to 0.55.

     This behavior confirms the data obtained during the Bremen pilot plant
studies.  During the initial 3 to 4 months,  the  removal of dissolved  organics
by virgin GAC took place primarily by adsorption.   The maximum  removal  of
organics by adsorption took place for about 6 weeks,  after which the  degree
of removal of organics decreased to 25 to 33% over the next 3 to 4 months.
As the biological activity in the GAC media increased, however, the  amount
of dissolved organic carbon removed from solution by adsorption plus  biologi-
cal activity increased (at the Donne plant)  to just over 50% of the  amount
removed by virgin GAC.  This 50+% removal remained constant over the  last
two months for which data are reported (the 6th  and 7th months  after  virgin
GAC startup).
                                      388

-------
co
00
              6
                1.0-
                0.6-
              U  O 2
              O
                                                                 .,]
cur IM»II
                                                       ADO<:
                                                 (Jekel, 1979)
                        Dec Jan  Fek Alar  Apr  May June July

                        '77         1978
              Figure 108. Initial  performance of  BAG  adsorbers
                        in the Dohne water works

-------
      Based upon the nore  than  two year long  pilot  plant  program which  he
 conducted  at the Dohne  plant,  Dr. Jekel  anticipates  that this  5Q+% dissolved
 organics removal  will continue for  up  to 2 years before  reactivation of the
 GAC may be required.  Dohne's  plant management has budgeted for regeneration
 on this basis (see  later  section in this Appendix  on Costs).

      Dr. Jekel  also believes that had  the new GAC  columns at Dohne been
 started up during the summer period of higher water  temperatures, high
 levels of  biological activity  would have been established in a much shorter
 time, say  2 to  3  months.

      Substantiation of  the increasing  biological activity in the GAC columns
 is shown in Table 100,  which lists  the  increase in inorganic carbon concentra-
 tions obtained  during passage  through  the GAC contactors for the periods
 January to  April, 1978, and May to  June,  1978.  During the May to June
 period, production  of C02 more than doubled in all  4 GAC columns.   These
 data  should be  coupled  with those of Table 99, which show that the COC in
 the GAC column  effluents was about  67% higher during the same period than
 during March  to April,  1978.

 TABLE ICO.   OOHNE PLANT PERFORMANCE DATA, JAN.-JUNE,  1978.   BIOLOGICAL
                PRODUCTION OF INORGANIC CARBCN IN SAC CONTACTORS
Inorganic Carbon Produced
(C02 as C), mg/1)
SAC Contactor Contg*
NK-12
F-300
LSS
AG-1
Jan-April, 1978
0.2
0.2
0.3
0.3
May- June, 1978
0.5
0.6
0.6
0.8
k During these studies there was no ammonia in the GAC influent.
Therefore, no nitrification was occurring.
     When the water temperature rose in the Spring of 1978,  the biological
activity in the activated carbon columns increased significantly,  and the
amount of C02 produced more than doubled.   The amount of CO? produced
during the May to June, 1978 period with the carbons LSS and AG-1  actually
was greater than the amount of DOC being removed from the GAC influent
(Table 101).  This indicates that some biological  regeneration of  the SAC
must have been occurring.  Since the only sources  of dissolved organic
carbon to produce C02 were the influent water and  what had been adsorbed by
the GAC during the winter period, the excess C02 produced over that present
in the influent water must have arisen from the biodegradation of  adsorbed  DOC.
                                      390

-------
TABLE 101.  DOHNE PLANT PERFORMANCE DATA.  COMPARISON OF DOC REMOVED VERSUS
            INORGANIC CARBON PRODUCED IN BIOLOGICALLY ACTIVE GAC CONTACTORS
                    DURING MAY - JUNE, 1978
GAC Type
LSS
AG-1
BUT
NK-12
F-300
Source:

DOC removed, mg/1
0.5
0.5

0.6
1.0
1norg C produced, mg/1
0.6
0.8

0.5
0.6
Jekel, 1978. Private Communication
     A significant benefit of the new BAC process to the Rheinisch WestfSlis-
chen Wasserwerksgesellschaft mbH, is that the amount of chlorine dosage
required to produce the 0.1 mg/1 of chlorine residual has been lowered fron
0.4-0.8 mg/1 to 0.2-0.4 mg/1, halving the amount of chlorine formerly used
for this purpose and bringing Dohne's chlorine dosage down to levels used
with less polluted and treated surface waters, such as lakes, reservoirs and
some groundwaters.

     During the pilot plant testing program at Dohne, pronounced slowing
down of biological activity was noticed in the GAC adsorbers whenever the
DOC in the influent to the adsorber dropped below 2 mg/1 and approached 1
mg/1.  This slowing down of biological  activity is believed to have been
caused by a shortage of biodegradable substances over a longer period.

RELATIONSHIP OF EPA'S PROPOSED ORGANICS REGULATIONS REGARDING GAC TO DCHNE
PLANT PERFORMANCE

     EPA's proposed regulations for the use of GAC to remove synthetic
organic chenicals (SOCs) from U.S. waters would require that the GAC be
regenerated when the TCC in the GAC effluent becomes 0.5 mg/1 higher than
that obtained with virgin GAC.  The proposed regulations also would require
that the TOC removed by virgin GAC be at least 50% of that in the water
influent to the GAC column.

     It can be seen from the data of Table 99 that if EPA's proposed regula-
tions were to have been applied to the modified Dohne treatment process (4 m
GAC columns; 11 minute EBCT) that 50% TOC removal was obtained by only one
of the four activated carbons being used.  Only the F-300 carbon removed 50%
or more of the influent (to the GAC column) TOC,and continued to provide
this amount of TOC removal through April, 1978  (5 months).  However, It Is
probable that longer empty bed contact times with the other three activated
carbons would have produced the desired initial 50% removal of TOC.
                                      391

-------
      On the other hand, continued removal  of TOC by all  four activated
 carbons at Dohne would have met the proposed EPA requirement of no  more than
 a 0.5 mg/1 maximum increase in effluent TOC  concentration  through April,
 1978 (5 months).

      On the basis of these  considerations  alone,  one would conclude that
 only the F-300  3AC would meet  the proposed EPA  requirements  at  Dohne,  with
 an 11  minute EBCT,  and that this  carbon would have  to be regenerated after  5
 months of use.

      However, when biological  equilibrium  had been  obtained  in  April,  1978
 (after 5 months of use), the F-300  activated carbon now allowed 0.9 mg/1 of
 DOC more in the GAC column  effluent than did the  virgin activated carbon.
 At this  time, only about 33% of the DOC  in the  GAC  influent was being  renoved.
 After  an additional  2  months of use, however, the degree of COC removal had
 risen  to about 50%  (see Figure  1C8).  This performance was expected to
 remain essentially  constant  (33 to  50% DOC removal) for at least the next 18
 months  (through 1979).  At this 11 minute enpty bed contact time, F-300 3AC
 would  not,  however, meet the proposed EPA regulation which would Unit the
 effluent TOC level  to  only 0.5 mg/1 above that obtained with virgin GAC.

     On  the other hand, performances of all four of the biologically active
 GAC  colunns at Dohne during the period Kay to June, 1978 are considered to
 be  satisfactory by  the RWVI, and no plans are being made to reactivate the
 GAC  until after two years of use.  Even though only 33 to 50% of the influent
 DOC  is being removed,  the GAC is not saturated with the very strongly
 adsorbed, more refractory, halogenated organic materials which are present
 in  the Ruhr River only in very low concentrations.  GAC column performances
 at Dohne are being followed closely, not only by DOC, UV and UV/DOC ratios,
 but also by TOC1 and TOC1N analyses.  As long as the rate of COC removal
 remains  constant at 33 to 50% and there is no indication of TOC1, TOC1N or
 bacterial breakthroughs, Dohne plant management does not plan to regenerate
 the GAC, at least during the first two years  of use.

 SAND FILTER OPERATIONS

     Donne's 4 sand filters (before the GAC contactors)  are 3.9 m diameter,
 30 m long cylinders, with sand 1.6 m deep and fed with ozonized then oxygena-
 ted water at the rate of 11 m/hr (9 to 10 minutes EBCT).   Before ozonation
was incorporated into  the process, the running time of these filters was 2
 days, then backwashing was required because of increased head loss.   Since
 ozonation has been installed, however, the running times have increased to 7
 days.  Furthermore, the indicator of backwashing need during 1978 no longer
was head loss, but rather breakthrough of turbidity (0.4 to 0.8 FTU before
 sand filtration -- 0.1 FTU after).  When the  turbidity of the filtered  water
 rises above 0.2 FTU, the filters are backwashed.  Breakthrough of turbidity
 rather than head loss  is the reason that ~ohne has installed turbidity
monitors after sand filtration (Jekel, I978b).

     In  December, 1978, however, Dohne plant  management planned to change
 the sand filters, because it wants to control this sand filtration step by
 head loss, not by turbidity breakthrough.  The new filters  will  be composed

                                      392

-------
of dual carbon media.  The lower layer will be hydroanthracite and the upper
layer will be a light density granular activated carbon.  This light density
GAC  is a filtration medium, and has very poor adsorption qualities.

     During the summer of 1978, "a population explosion of nematodes" was
observed to have developed both in the rapid sand filters and in the following
GAC  adsorbers at the Cohne plant.  These growths were shown to have been
caused by the extended intervals between backwashes, during which the nema-
todes developed.  When the backwashing cycles were reduced to 3 days (just
below the time of reproduction of these microorganisms), the nematode problem
disappeared (Heilker, 1979).

GAC  COLUMN BACKWASHING

     Dohne's GAC contactor columns are backwashed first with air to loosen
the  biomass which tends to cause GAC particles to stick together at times.
After loosening the mass with air, water alone is pumped through the columns
at the rate of 27 to 30 m/hr.  The time between backwashings is controlled
by head loss through the columns.  Because of the fact that there are four
different granular activated  carbons being tested in the Dohne plant, backwash-
ing  frequencies of the four GAC contactors are not the same.   Each GAC has a
different grain size, and head loss buildup occurs at a different rate for
each of the four contactors.   On the average,  however, backwashing was being
performed every 10 to 14 days during early 1978.   However,  because of the
development of nematodes (see above), the backwashing time  subsequently has
been lowered to 3 days (Heilker, 1979).

     Turbidity of the water after the GAC contactors ranges from 0.1  to 0.2
FTU, and averages 0.14 FTU.

     The four activated carbons being studied  in  the Dohne  plant have
different adsorption capacities, which were determined during the pilot
plant testing program.  The full-scale plant 4 meter virgin GAC columns were
started up during the winter  of 1977, when water temperatures were below 8°C
for  the first 3 to 4 months of operation.  As  a result, bacterial  growth
buildup was very slow in the  GAC media.  This  was confirmed by the data of
Table 100, which show low levels of C02 being  produced during this period.

     During backwashing, only small amounts of carbon fines are removed.
These had not been sufficient to require the addition of any makeup SAC
during the first 7 months of  plant operation.

     No buildup of slimes has been observed at the Dohne plant in either the
sand filters or in the GAC contactors during the first 7 months of operation.
Neither were slimes observed  during the 2 year pilot plant  study.   Therefore,
it can be concluded that the  pretreatment steps,  including  ozonation, prevent
the  buildup of such detrimental materials.

BACTERIAL TESTING

     Table 102 shows bacterial count data obtained at various points in the
Dohne plant process.  It 1s clear from these data that the  SAC effluents

                                      393

-------
 from the old GAC columns contained considerable biological activity.  After
 5 months of use, effluents from virgin GAC columns showed much lower bacterial
 counts.  However, it is equally clear that E_._ coli are not present in the
 effluents from columns containing old or new GAC.  Two groups of European
 microbiologists (KIWA — The Netherlands and Univ. of Saarlands — Federal
 Republic of Germany) agree that IE. coli cannot survive in the GAC columns in
 the presence of the other strains of water and soil bacteria which are
 present.

 TABLE 102.  BACTERIAL COUNT DATA AT DOHNE PLANT WITH BAG TREATMENT PROCESS
sampling Point
tew water
\fter flocln +
sediment.
Hfter sand filt.
Ifter GAC
^fter ground
passage
With Old GAC*
Total Counts'/ml
M
0
14,490
2,340
6,010
3,700
27
°g
2.0
4.2
4.9
4.0
2.3
E. COl
"g
1,620
6.7
« 1
« 1
« 1
7100 ml
ag
1.7
3.2
—
—
—
April, 1978 (new SAC)**
counts/ml
20,000
20,000
90
57
22
E. coli/ICO ml
—
—
0
0
2.2***
M = geometric mean a = geometric standard deviation
k Source; Sontheimer et_ aj_. , 1978
** Source; Heilker, 1978, Private Communication
*** About 20% of the water present after ground passage comes from
infiltration from the Ruhr River. This infiltrated water is the
source of the E. colls found.



     The Dohne plant waters are analyzed routinely for total  bacterial
counts, £. coli and algae in the raw water, filtered water, after ozonation,
after GAC and after ground passage.  Raw water from the Ruhr  River is
analyzed once each day and groundwater stations are analyzed  once each  week.

COSTS

     Substantial cost data were obtained during visitation to the Cohne
plant in June, 1978.  These will be presented in three sections:   (a) Costs
For Enlarging GAC Contactors And For GAC, (b) Costs For Ozonation Equipment
and (c) General Cost Data, which include Costs For Plant Operations  and
rates charged to water customers.
                                      394

-------
 Costs  for Enlarging  GAG  Contactors and  for GAG

     The  originally  installed GAC contactors  (which provided a 2 m depth of
 GAC) were enlarged in  late  1977.  Plant managenent had planned for this
 eventuality when  the original GAC contactors were installed, and had designed
 the building and  associated appurtenances to be able to accomodate the
 enlarged  GAC depths.   Currently, the four GAC contactors are 5 m in diameter
 and 8.5 m high.   The current depth of GAC contained in each contactor is 4
 meters.

     Enlargement  of  the  four GAC contactors was completed at a cost of
 300,000 Deutsch marks  (DM)  ($150,000 at an exchange rate of 2 CM/$), including
 necessary modifications  to the building.  On the other hand, four new GAC
 contactors of the current size at the Donne plant would have cost an estimated
 600,000 DM ($300,000).  A single new GAC contactor would have cost approxi-
 mately 250,000 DM ($125,000) on a turnkey basis.

     In addition, 450,000 DM ($225,000) were expended for the purchase of
 320 cubic meters  (about 100 netric tons) of GAC to charge these contactcrs
 with virgin GAC in November, 1977.  The GAC cost represents an average of
 about 75
-------
 only  on a weekly basis.  These extended backwashing cycles are attributed
 largely to the  new ozonation process.  It had been hoped that extended
 backwashing  times would have provided substantial cost savings both in
 pumping costs and 1n wastewater treatment charges.  However, the nematode
 problem arose,  which necessitated lowering the backwashing times to every 3
 days, in order  to avoid their development.

 Ozonation System Costs

     The capital cost of the total ozonation system for the Dohne plant was
 2,200,000 DM in 1976.  This figure includes air compressors and driers,
 ozone generators, associated piping and control instrumentation and construc-
 tion of the ozone contact chambers.   The ozone generation capacity is 8
 kg/hr (422 Ibs/day), which provides for ozone dosages of 3.5 to 5 mg/1,
depending upon the rate of water flow through the plant.

     The energy demand of the ozone generation system at Dohne is 20 kwhr/kg
of ozone generated,  or 9.1  kwhr/lb.   It was not clear whether this figure
 includes electricity used by the air preparation unit.  If not,  an additional
21% (1.9 kwhr/lb)* should be added.   This  would result in a total energy
demand of 11  kwhr/lb for air preparation and ozone generation.   An additional
1.8 kwhr/lb* of energy is required by the  ozone contact system (diffusers
plus off-gas recycling to the preozonation step).   This makes a  grand total
of 12.8 kwhrs/lb of  ozone generated  and applied.   [The figures  of 1.9 and
 1.8 kwhr/lb required for air preparation and ozone contacting,  respectively,
were obtained from a questionnaire completed by The RWW during  1977 (see
Miller et al_.,  1978)].

     Although the addition  of ozonation to the Cohne water treatment process
involved a capital cost of  2,200,000 DM ($1,100,000)  and  an Increase in
purchased electricity,  5,086 kwhrs/day (523 to 755 DM; $262 to  $378 at 0.09
to 0.13 DM/kwhr local  power cost), these increased costs  have been offset by
several  operating cost savings:

1)   Much less  chlorine is  used in the new process than in the  old process.
     A minimum  of 96% of the chlorine formerly required (1,059  to 5,295
     Ibs/day) has been eliminated.

2)   A reduction in  labor force by 7 or 8  persons  for an  estimated total
     annual  savings  of 300,000 DM ($150,COO).

3)   The GAC reactivation cycle has  been extended from a  conservatively
     estimated  hypothetical 6 months to a  projected minimum of two years —
     an annual1zed savings  of just over 600,000 DM ($300,000)/year in GAC
     reactivation costs alone.                  • *

     Although individual Items of cost savings can be deduced by comparing
the old and new treatment processes, other modifications  of the  process
required increases in costs.  Heilker (1979)  summarized the cost comparisons
for water treatment  at the  Dohne plant by  the old and by  the new processes
and made the following ststement:
                                      396

-------
      "The treatment plants 1n the Donne waterworks have been operating for
      more than 1.5 years using the revised process.  The drinking water
      quality has been significantly improved without increasing treatment
      costs.  The Dohne plant is less susceptible to disturbance and as a
      result can be operated with 50% of the former staff size.  The activated
      carbon filter runs are 3 to 5 times longer than before."

      In addition to the Dohne plant, the Rheinisch-Westfaiischen Wasser-
werksgesellschaft mbH also owns and operates several other plants near
MUlheim.  Based upon the performance and cost benefits obtained using the
ozone/GAC process at the Dohne plant,  two other plants have been redesigned
and are being modified to use the ozone/GAC process for both plants.  Comple-
tion  and startup of these plants is expected during 1979.

      Bids were received in early 1978  for the ozonatlon equipment for both
new plants.   The larger of the two plants requires a 36 kg/hr (1,900 ibs/day)
ozone generation system which cost 2.9 million DM ($1.45 MM).  This price
included all related hardware, such as the air preparation equipment, ozone
generators,  turbines (for first stage  ozonatlon), contacting (for second
stage ozonation),  associated instrumentation and controls and contactor off-
gas destruction equipment.   On the other hand, this price does not include
the second stage ozone contact chambers or buildings.

     The ozonation system at the smaller plant will generate 14 kg/hr (739
Ibs/day) and has been purchased for 1.6 million DM ($800,OCO).   This price
included and excluded the same items as the larger plant,  but was  described
as a more complicated installation requiring more piping at several  points
in the water treatment system.

      In Table 103 cost figures and ozone generation capacities  are compared
for the three RWW ozone installations  discussed above.   It is evident that
economies of scale result in lower costs per unit of ozone generation at the
larger installation.   However, site-specific factors (such as the  need for
relatively more piping at the smaller  plant) also affect the cost/lb of
ozone generation capacity/day.

General Cost Data

     Residential customers  of the RWW  in Mill helm were charged 0.89 CM/cu 11
of water supplied during 1978.   This is somewhat lower than the rate charged
by other major German municipalities.   At the exchange rate prevailing
during June, 1978 (slightly under 2 DM/$), this rate converts to about
$1.65/1,000 gallons.

     Costs of treatment plant operation at Dohne were described by RWW
management as falling into  three catagorles — treatment,  distribution and
pumping (for both treatment and distribution).  These categories account
for,  respectively, 21%, 462 and 33% of total operating costs.  A breakdown
of the costs within each of these categories is included 1n Table  104.
Several interesting conclusions can be drawn from these data in comparing
German treatment practice at the Dohne plant of MUlheim with typical United
States practices.  Treatment costs of  2U are a relatively low percentage of

                                      397

-------
                    TABLE 103.  COSTS FOR OZONATION SYSTEMS* IN MULHEIM, FEDERAL REPUBLIC OF GERMANY
GO
£
Plant
Dohne
"A"
"B"
Ozone Generation
Capacity
kg/hr
8
14
36
kg/day
192
336
854
Ibs/day
423
741
1,905
Capital
Cost
MM of DM
2.2 (1976)***
1.6 (1978)
2.9 (1978)
Cap Cost of
Ozone Generation
DM/kq/dav
11,458
4,762
3,356
$/lb/day*
2,600
1,080
761
* includes cost of air preparation, ozone generation, turbine contacting, contactor
off-gas treatment, controls, but no contact chambers.
** at an exchange rate of 2 DM/$
*** includes building costs

-------
total costs.  Labor also represents a snraller part of the total.  On the
other hand, costs related to capital (interest and depreciation) are high.
Taken together, these three observations show that a irore capital-intensive
system is used at Dohne, which relies less on labor than is typical in
United States plants.

TABLE 104.  BREAKDOWN OF COSTS AT DOHNE PLANT. MULHEIM. FRG
Category of
Cost
Labor
Energy
Materials
Taxes
Depreciation (c)
Interest (d)
Notes;
Treatment
(21% of Total)
10%
15%
17%
2%
17%
39%
TOOT"
Distribution
(46% of Total )
10%
1%
2%
7%(a)
29%
35%
TCOT"
Pumping
(33% of Total)
32%
36%
1%
1%
17%
13%
TOOT"

(a) Property taxes to local communities
(b) Taxes to authorities providing raw water supply
(c) Includes capital set aside for future building
(d) Interest costs estimated to be 5 to 6%
     Cost of water to MUlheim's residential  customers averaged 89 pfennigs/-
cu m (1.68^/1,300 gallons) in 1978.  Power costs are 9 to 13 pf/kwhr (4.5 to
6.5i/kwhr) at Dohne, but are lower at other MUlheim plants where power is
generated on-site.  The RWW did not raise the price of water to their custo-
mers when the BAG process went on-stream in 1977.
     Labor costs at Dohne are 2,800 DM ($1,400)/month.
employs 10 machine operators (2/shift) + 1  plumber.

FUTURE PLANS AT MULHEIM
Dohne currently
     Based upon the successful incorporation of the BAC process into the
Dohne p'ant, the RWW management is incorporating the process into two
additional plants which also draw Ruhr River water.  These plants will
process 72.0CO and 25,000 cu m/day of drinking water.
                                      399

-------
     Both processes will be nearly "identical to the current Donne process,
except that ground passage of treated water will be eliminated.  Ground
passage could be eliminated at Dohne now, from a treated water quality point
of view, but will be retained to act as a storage reservoir.  In addition, a
Superpulsator will be used at the smaller plant instead of the Pulsator,
because the newer equipment can be used at much higher upflow rates (12
m/hr).

     Additionally, the sand filters at Dohne were scheduled to be replaced
in December, 1978 with dual media filters.  These were to be hydroanthracite
(0.6 to 1.2 mm particle size) covered with a layer of light density (2 to 3
mm particle size) GAG (which has very little adsorption capability.  The
reason for this change is to allow backwashing of the filters to be controlled
by headless rather than by turbidity, which is the present control  mechanism.
                                      400

-------
                                  APPENDIX F

        ROTTERDAM,  THE NETHERLANDS  —  KRALINGEN WATER TREATMENT  PLANT
Background

     The  new Kralingen plant became  operational  in March, 1977.   It wai
constructed  to  replace the older Honingerdtjk plant which dated  back to
1874.  Raw water for the Kralingen plant  is  supplied by the River Maas
through two  Biesbosch storage reservoirs.  An alternate water source is the
Berenplaat storage reservoir and, under emergency conditions, River Rhine
water may be drawn through the Nieuwe Maas emergency Intake.  Figure 109
shows the Kralingen plant water sources.
          RHINE RIVER
             i
 ME;SE RIVER
 (HAAS)
                                                             RhBIE RIVER
          OVDE MAAS

        DGR3JMCY
                          HISJWE MAAS

                            areY INTAKE
          BEREMPLAAT

        3TCRA3E
  BIES BOSCH

STORASH R2S
                                 KRALBGEN RESERVCIR
            Figure 139.  Kra'lngen (Rotterdai) plant water sources.
     The Kralingen plant is of  interest to students of European  water
treatnent practices for a number of reasons, including the following:

1)   Kralingen is a totally new plant  which incorporates the latest  knowledge
     of  water treatnent processes,
                                      401

-------
2)   Being recently constructed, Kralingen provides contemporary cost
     information,

3)   Kralingen incorporates a high level of process monitoring for organic
     chemicals, especially trihalomethanes, due to the interest in these
     problems by Dr. Johannes J. Rook, Chief Chemist.  Dr. Rook is a pioneer
     in studying and identifying the mechanisms of formation of trihalo-
     methanes in drinking water.  Dr. Rook's basic objective is to minimize
     the amount of chlorine used in treating Rotterdam's surface water
     supplies, while producing high quality finished drinking water.

The Biesbosch Reservoirs

     These two reservoirs store River Maas water, which flows sequentially
through the first reservoir, then the second.  Each reservoir is 20 meters
deep and the two provide a combined retention time of three months.  This
storage time equalizes wide variations in River Maas flows, which are especial-
ly affected by the annual Spring thawing of snows.

     The second reservoir has been fitted with aeration devices, prinarily
for the control of algae, to prevent stratification, to promote biological
decomposition of dissolved organic materials and to promote conversion of
ammonia to nitrate ion by nitrification.  Airmonia levels in the reservoir
influent from the Maas are as high as 4.5 mg/1 during winter months, but
this usually drops to less than 1 mg/1 during the 90-day reservoir storage
in winter and to zero during summer.  Coliform levels drop to about 10% of
their influent levels during the first 30 days of storage in these reservoirs.

     Promotion of aerobic biological activity in the Biesbosch reservoirs
incorporates biological pretreatment into the Kralingen plant treatment
scheme.

     The cost for installing these two reservoirs was equal to the cost for
installing the Kralingen treatment plant itself (see cost section).

     Water flows from the Biesbosch reservoirs to the Kralingen plant
through a 20 kilometer long transmission line which requires 15 hours
residence time.  The Kralingen plant is designed to operate without prechlori-
nation, however unacceptable head losses occur in the transmission line when
water temperatures exceed 10°C during summer months.  This is due to buildup
of biological growths in the main.  Therefore, up to 4 mg/1 of chlorine is
added at the reservoir to control biogrowths in the transmission line during
summer periods.  Prechlorination dosage is controlled by monitoring head
loss in the transmission main.  Nominal prechlorination dosage is set at 1
mg/1, but when head losses increase, more chlorine is added.  Prechlorination
at the reservoir for protecting the transmission line is referred to as
"transport chlorination".

     Prechlorination is not required during winter months when the water
temperatures are below 10°C.  Therefore, trihalomethanes are produced at
Kralingen only during summer periods when transport chlorination must be
practiced.


                                     402

-------
The Kralingen Treatment Plant

     The site of the Kralingen plant is adjacent to that of the old Honinger-
dijk plant, as this is the focal  point of the Rotterdam water supply system.
Transmission irains and the Nieuwe Maas intake previously used by the Honinger-
dijk plant can be used by Kralingen.

     Process selection was based  on a lengthy (about 2 years) semi-industrial
scale pilot plant study.   Figure  110 shows details of the individual process
steps and Table 105 summarizes the number of operational units and their
capacities for each process step.

     Low lift pumps draw water from the raw water storage reservoir at the
end of the 20 km transmission line from the Biesbosch reservoirs.   Iron
sulfate coagulant is added in static mixers prior to discharge to floccula-
tion, which is achieved in a 4-compartment unit utilizing horizontal shaft
paddle mixers.  The flocculated water is clarified in 55° lamella settlers
(Parkson process).

     Clarified water flows to a 5-minute retention time, aspirating turbine,
ozone contacting chamber, and subsequently to an 11-minute retention time
tank.  A design dosage of 3 mg/1  of ozone is added to produce a residual
dissolved ozone level of approximately 0.4 mg/1 in the water emerging from
the contactor.  However, the ozonation step is controlled by monitoring the
level of ozone in the contactor off-gases at the level of 1 g of ozone/cu
meter.  This is done because Kralingen plant personnel have lore confidence
in being able to monitor ozone in the gaseous phase than in solution.  Ozone
is analyzed in both phases spectrophotometrically (with a Sigrist unit), but
in solution a *i1m of oxidized (by ozone) micropolljtants gradually builds
up which interferes with light transmission and results in inaccurate readings.

     Ozone in the contactor off-gases is destroyed in heated catalytic
units.

     Ozonized water flews to multi-media filters (20 cm of anthracite on
sand, supported on gravel) at a rate of 15 to 20 cu m/sq m/hr (m/hr).
Filter backwashing is provided by air, then water backwash, when the headless
reaches a maximum of 1.5 meter.

     Filtered water is pumped from the dual media compartment filter clearwell
below the filters by medium lift pumps to the pressurized GAC columns.
Filter and GAC column backwash water also is drawn from the filter clearwell.
A depth of 4 neters of GAC is provided in the twelve, 8-meter high, 6-meter
diameter, cylindical steel GAC contactors.  These GAC contactors are construc-
ted of carbon steel with a special proprietary coating on the inside surfaces.
The GAC is supported by a coated steel plate with plastic nozzles inset in
the plate for flow distribution.   A 10-minute empty bed contact time is
provided in the 3AC contactors.

     Backwashing of 3AC columns is conducted about once/week by using the
filter backwash pumps.  The process is controlled by head loss of 5 11 WC.
                                      403

-------
          River Haas
                        1
          Biesbosch ReservoTs
          Raw Water Storage Reservoir
          Low Head Punping Station
          F'occulators
          Clar-*-ers
          Czore Coitacto-s
                        I
          3AC Contactors
          Fin-shed Water Reservoirs
                                                      aerat-on
Dual Media Filters
>
f
F-lter Clear Well
,
f
Medium -ift Pjrpirg Station
                                                  ' transport chlorTation
                                                     'summer only)
                                                  chenlcals
che-rca's
          H-gh Lift Purpirg Stafon	

                         i                ,
          Water Oistr-butioi Systen       I
                                                 post  chlor-nation
                                                   (C  5 ng/' dosaqe
Figjre "1C.   Process  flow d^agran of the Kralingen water treatment

              plant,  Rotterdam, "he Netherlands.
                           404

-------
TABLE 135.  DESCRIPTION OF TREATMENT PROCESS UNITS AT KRALINGEN PLANT
description of unit
raw water storage basin:
low lift pumping station:
constant speed
variable speed
coagulation: static mixer
flocculation: four compartment,
horizontal shaft, paddle
flocculators
clarifiers: lamella (Parkson
process) separators
ozonation: single stage, 5-minute
retention time, aspirating turbine
holding tank
filtration: dual media (0.8 m sand,
0.8 m anthracite). Unit filter
area -- 36 m^
medium lift pumping station:
medium lift pumps
filter & GAC backwash pumps
GAC contactors: 10 minute EBCT,
4 m GAC depth; unit filter area =
28.1 m2
finished water storage reservoirs
high lift pumping station: pumps
number of
capacity operational
of unit units
87,984 m3

1.5 m3/sec
1.5 m /sec
1.94 m /sec

3
0.4 m /sec
3
0.4 m /sec
3
0.4 m /sec
0.4 m /sec




1 m /sec
1 m /sec

0
0.194 m /sec
30,000 m3
0.83 m3/sec
1

1
1
1


5

5

5
5


10

1
1


10
2
6
number of
backup
units
—

—
2
1


1

1

1
1


2

3
3


2
—
2
                                   405

-------
     The GAC 1n the columns is manufactured by Norit (currently Super,
Normal and mixed Super/Normal) and leased from Norit.  Spent GAC is returned
to the Norit plant, some 20 miles distant, for reactivation.  Any losses of
the leased GAC during backwashing or transportation to Norit are charged to
the Kralingen plant at the rate of 1,300 Dutch gilders/cu m.

     During the plant visitation (June, 1978), only 8 of the GAC columns
were being used.  Of the eight, seven were scheduled to be on-line at any
one time, with the GAC from the eighth column being regenerated at the Norit
plant.  Currently, GAC regeneration is expected to be required every one to
two years, based on chemical laboratory analyses being developed by Dr. Rook
(see later section).

     Water flows from the GAC columns to above-ground, enclosed finished
water reservoirs having a capacity of 30,000 cu m.  High lift pumps draw
from the reservoirs and discharge to the Rotterdam distribution system.
Prior to storage, the processed water is treated with chlorine to attain a
free residual of 0.1 to 0.2 mg/1, which normally requires chlorine dosages
(total of 0.3 to 0.6 mg/1).  During summer, a peak chlorine dosage of 1 mg/1
is required to attain these chlorine residuals.  At the extremities of the
distribution system, total chlorine residual is zero to 0.1 mg/1.

Water Quality

     Table 106 lists typical water analyses obtained during January, 1978.
Biesbosch raw water analyses after entry into the plant but prior to treatment
are shown as "raw water".  Finished water analyses are made after residual
chlorination.  It should be noted that influent CCD and TOC levels of 8 and
4 trg/1 were lowered to 4 and 2.8 mg/1, respectively, by the Kralingen treat-
ment process.  Dr. Rook believes COD to be a better parameter for monitoring
effectiveness of GAC adsorption efficiency than is TOC.

Trihalomethane Production & Removal by GAC

     Several well-known papers by Dr. Rook detail his pioneering studies on
the formation of trihalomethanes in Rotterdam drinking water.  These papers
are cited in the bibliography section of this report.  In addition, Dr. Rook
provided the unpublished data of Table 107, which show that removal of THM
precursors by the Kralingen water treatment process (ozonation followed by
GAC adsorption) is efficient only with virgin (or freshly regenerated)
activated carbon.  From June through November, 1977 prechlorination of
Biesbosch reservoir water was practiced at levels of 4 mg/1.  Total THM
levels over this period averaged 108 microg/1 in the plant influent, 72
microg/1 before ozonation and 71 microg/1 after ozonation.

     Without GAC adsorption and following 0.5 mg/1 post-chlorination dosage,
TTHM levels in ozonized water were 100 microg/1.  With virgin GAC and post-
chlorination, TTHM levels in ozonized water dropped to 17.5 microg/1, but
with 4-month old GAC, TTHM levels in ozonized and post-chlorinated waters
rose to 61 microg/1.
                                     406

-------
TABLE 106.  ANALYSIS OF RAM AND FINISHED  MATERS AT  KRALINGEN. JANUARY 1978
parameter
color, mg Pt/1
temperature, °C
turbidity, JTU,
KMnO. demand, mg/1
nitrite, mg/1
nitrate, mg/1
sulfate, mg/1
NH. saline, mg/1
NH. albuminoid, mg/1
Fe, mg/1
Mn, v-g/1
DO, mg/1
phenol, yg/1
detergents, yg/1
COD, mg/1
TOC, mg/1
Br, yg/1
hexachlorobenzene, i.g/1
- HcH, yg/1
- HcH, yg/1
- HcH, yg/1
cholinesterase inhibitors, in
parathion equivalents, yg/1
raw water*
mih.
9
5.5
0.60
10
0.019
18.5
78
0.20
0.17
0.02
6
11.2
1
45
8
3.5
120
—
<0.01
<0.31
0.01
0.04
avg.
12
6.0
1.6
11
0.050
18.8
83
0.26
0.20
0.03
7
12.2
1
50
8
3.9
125
<0.01
<0.01
<0.01
0.01
0.07
max.
14
7.5
3.9
12
0.070
19.0
90
0.33
0.22
0.05
9
12.6
2
55
9
4.1
135
--
<0.01
<0.31
3.01
0.08
finished water
mm.
<1
5.5
0.05
3
--
19.5
113
0.34
0.09
0.01
2
11.2
<1
10
2
2.2
60
<0.01
--
0.01
3.31
0.02
avq.
<1
6.0
0.10
4
0.000
19.9
116
O.C6
0.11
0.02
3
12.4
<1
11
4
2.8
65
<0.01
—
0.01
0.01
0.03
max.
1
7.5
0.25
5
—
20.0
119
C.ll
0.13
0.04
3
13.0
<1
15
5
3.7
80
<0.31
--
C.01
0.01
3.34
* raw water = water influent to the plant
                                     407

-------
              TABLE 107.  HALOFORM PRODUCTION IN KRALIN6EN  PLANT.  JUNE  1977  -  MARCH  1978
sampling point
plant Influent
after coagulation,
lamella settling, sand
filtration & evap'n.
after 5 min. ozonation
no GAC +0.5 mg/1 post-
chlorf nation
after virgin GAC +0.5
mg/1 post-chlori nation
after 4 month old GAC +
0.5 mg/1 post-chlori-
nation
data taken
average data,
June-Nov. 1977
average data,
June-Nov. 1977
average data,
June-Nov. 1977



prechlorlnatlon
at reservoir
4 mg/1
4 mg/1
4 mg/1
0
0
0
haloforms produced (mtarog/1!)
CHC13
55
35
34
51
3.3
22
CHCl2Br
34
23
24
34
3.6
15
CHBr2Cl
17
12
11.3
14
4.6
13
CHBr3
2
1.6
1.8
1
6.0
11
total
THMs
108
72
71
100
17.5
61
o
00

-------
     Thus Dr. Rook concludes that GAC 1s effective in removing THMs from his
waters only for short periods of time (less than 4 months).   He also concludes
that once his waters have been prechlorinated,  ozonation is  not effective
either in reducing THM concentration levels or  in preventing formation of
more THMs upon subsequent post-chlorination.

     From July to December 1977, the chloroform levels in Kralingen plant
influent water and in one GAC column effluent followed the behavior pattern
shown in Figure 111.  Prior to July, 1977 the plant influent water was not
chlorinated, therefore chloroform levels in the plant influent were zero.
Starting in July, 1977, the amount of chloroform formed immediately increased
to levels of 30 to 60 m1crog/l.  Most of this was adsorbed quickly by the
GAC, but the SAC effluent quickly began to show the presence of chloroform.
During September/October, 1977 the chloroform level in the GAC effluent
peaked at about 15 to 20 mg/1 then began to fall.  When prechlorination was
ceased in December, 1977, however, chloroform still was measured in the GAC
effluent for several months thereafter, falling to levels near zero by
March/April, 1978.  A second GAC column showed  the same behavior over the
same period, except that the chloroform concentration in the effluent levelled
off more slowly after prechlorination was ceased.

     Dr. Rook concludes from these and other data that GAC adsorbs chloroforn,
but that desorption of chloroform begins almost immediately.  After cessation
of prechlorination, all chloroform measured in  the 3AC ^iltrates is present
because of desorption.

Biologically Active 3AC

     Dr. Rook has noticed biological activity in the Kralingen GAC adsorbers,
but its buildup was slow during the first year  of operation.  Its current
contribution to the overall efficiency of the treatment process is relatively
small.  At water temperatures above 10°C (summer) biological activity is
operative in both the double layer filters and  in the GAC adsorbers.  Ammonia
levels of 0.3 mg/1 (winter) drop to 0.10 to 0.15 mg/1 by nitrification in
the dual media filters.  There Is no ammonia 1n the Kralincen plant influent
during summer because it 1s all nitrified in the Biesbosch reservoirs.

     The new Kralingen plant began operation in March, 1977  with virgin 3AC.
Reactivation of GAC did not begin until March,  1978.  With virgin GAC,
effluents from the adsorbers contained 0.9 mg/1 of TOC and 1 to 2 irg/1 of
COD (from Influent values of 4 and 8 mg/1, respectively).  After the second
month, the GAC effluent contained 2 mg/1 TOC.  By January, 1978, the TOC of
the GAC effluent had risen to 2.8 mg/1 (622 of that 1n the plant influent'.
At the same time, the average COD level in the GAC effluent was 50% of that
in the plant Influent.  Therefore, 1n terms of EPA's proposed GAC regeneration
criteria (remove 50% of the influent TOC), the  effective life of the Kralingen
activated carbon was less than 1 year.  However, this GAC performance was
considered to be satisfactory at Rotterdam.

     Detergent levels in the Kralingen water averaged 50 microg/1 during
January, 1978 and were lowered to 11 microg/1 during treatment.  Dr. Rook
noted that new GAC removed nearly all of the detergents originally present.


                                      409

-------
30-60

  nncrog/1
 A

 in

1C
t—


     0}
     O

     O
     O
               Influent-
           July,  1977
                                         THMs due to transport chlorinatlon
                             10,000  12,000
                              bed  volumesi
                                                                        ceased prechlorination
                                                                       elution of CHCU adsorbed;
                                                                       ^hows desorptlon is operative
                                    Sept    Oct
                                                                 Dec
           Figure  111.   Desorption of trihalomethanes  from GAC at Kralingen plant.

-------
However, after some 6.0CO bed volumes had passed through the GAC,  hardly any
detergent removal  occurred.   However, after continued passage of an additional
6,000 bed volumes  of water through the same GAC, renoval of detergents
increased up to a  level  of about 50% (Figure 112).   Dr.  Rook believes  that
desorption of detergents may have been occurring during  the period when no
removal was apparent and that the biological activity had not yet reached a
point at which biodegradation of detergents was significant.  Upon continued
use. however, biological degradation (and/or desorption) became responsible
for the removal of about 50% of the influent detergent concentrations.

     Biological activity in the Kralingen plant dual  media filters and  GAC
adsorbers is present, but does not appear to be removing as nuch ~OC and
ammonia as at the  MUlheim, Germany Donne plant.  This can be explained  on
the basis of the biological  decomposition reactions which occur 1n the
Biesbosch reservoirs during 90-day storage with constant aeration.  During
this storage period, most of the amnonla is nitrified and the readily
degraded carbonaceous organic materials are decomposed 1n the reservoir.
The balance of the TOC and COD which is removed in  the Kralingen plant
proper is a result of flocculatlon, clarification,  ozonation, filtration,
GAC adsorption and biological degradation 1n the filtration media and  GAC
adsorbers.

GAC Regeneration Parameters

     Dr. Rook began sending the Kralingen GAC out for regeneration in
March, 1978 (one or two columns at a time).  However, he believes that  even
though the GAC may be passing relatively high levels of TOC through it  in
the form of dissolved organic materials which are not strongly adsorbed by
GAC (or are easily desorbed), the less polar, more  strongly adsorbed halogen-
ated organic materials of concern still are being adsorbed.  This belief is
based upon work which had been conducted on Rhine River waters by the  Engler-
Bunte Institute of the University of Karlsruhe, Germany.

     As a result,  Dr. Rook has been developing analytical procedures for the
measurement of chlorinated organic materials.  TOC1 and DOC1 analyses  are
conducted following the Engler-Bunte Institute procedures.  In addition, an
"Ether-extractable Organic Chlorine" (EOC1) analysis has been developed by
Dr. Rook.  This extraction procedure employs petroleum ether (30°C boiling
point) for extracting GAC effluent.  As soon as the level of EOC1  rises
significantly, Dr. Rook plans to regenerate the particular GAC column(s)
involved.

     The March, 1978 GAC regeneration was based on  TOC rise in the 3AC
effluent.  However, as of June, 1978, regeneration  now will be based on a
maximum level of TOC1 and/or EOC1.  These parameters for GAC regeneration
may be changed in the future by Dr. Rook as more data are gathered regarding
the Kralingen plant performance.

Colony Counts

     Table 1C8 lists representative colony counts/ml obtained during March,
1978 and May, 1978.  There is little question that  counts are low after the

                                      411

-------
         100%
-**
ro
      o
      03
      01
      S-
      O)
12,000 bed volumes
          Y
                                                                time
                     Figure M2.  Removal of detergents by  Kralingen plant GAC.

-------
TABLE 108.  REPRESENTATIVE COLONY COUNT DATA AT KRALINGEN PLANT
sampling point
reservoir - before
chlori nation
reservoir - after
chlorinatlon
plant Influent
after lamella settling
after ozonatlon
after GAC
plant Influent
after lamella settling
after ozonatlon
after sand filtration
after GAC
data obtained
during
March 1978

March 1978
March 1978
March 1978
March 1978
March 1978
May 1978
May 1978
May 1978
May 1978
May 1978
* incubated at 22°C; grown on agar
no. of
samples
123

122









colony
counts/ml*


—
19-20
7
3
10
150
2,000-
5,000
70
10,900
5,800
no. of samples having counts of
oTmT
109

64









1-10/ml
12

32









10-100/ml
2

24









100-1,000/ml


2 (av. 500)










-------
ozonation step but high after GAC adsorption.   However,  the post-chlorination
step (0.5 iig/1 chlorine dosage) lowers plate counts 1n the Kralingen product
water to levels which meet the required public health standards.

Capital Costs And Financing

     The Kralingen plant was completed in 1977 at a total  capital  cost of
170 million Dutch gilders, stated in April,  1978 values.   Since the Rotterdam
accounting procedure includes the statement  of all  assets  in estimated
current market values, figures were quoted to the site visitation  team and
are discussed below in April, 1978 values.  The market exchange rate during
June, 1978 was approximately 2.1  Dutch gilders per U.S.  dollar.  Hence,  the
Kralingen plant cost $81 million.

     Table 109 breaks down the capital costs of the components  of  the
plant; the figures are in millions of Dutch  gilders and  1n U.S.  dollars.
          TABLE 109.  CAPITAL COSTS OF KRALINGEN PLANT
Cost Categories
General items (land purchase, engi-
neering, financing, overhead, archj
tects, city planning fee, etc.)
Service building (offices, labora-
tory, mechanical workshop, pilot
plant)
Streets, landscaping
Storage reservoir for raw water,
pumping station
Emergency inlet pump station
Treatment plant (see separate
breakdown of components)
Carbon facilities (excluding GAC)
Finished water storage tanks
High lift pumps
Shelter facilities for war
emergencies
Items unaccounted for
TOTAL
April 1978 values
(•nil! Ions of Dutch gilders)
40
24
12
5
2
46
7
9
6
1
18
170
millions
of $ I'S
19.05
11.43
5.71
2.38
0.95
21.90
3.33
4.29
2.86
0.48
8.57
80.95
                                     414

-------
     Additional information was provided on the costs of each component of
the treatment plant itself.  The elements making up the total of 46 MM
gilders are shown in Table 110.

          TABLE 110.  TREATMENT COMPONENTS AT THE KRALINGEN PLANT
Plant Component
Construction - concrete portions
Piping
Heating, cooling, ventilation
Electrical
Chemicals - initial stock
Electrical installations for
chemical dosing and metering
Coagulation
Sedimentation
Sludge treatment & drying
Ozonation equipment*
Filtration
Medium lift pump station
Control system - automation
TOTAL
April 1978 values
(millions of Dutch gilders)
13.5
1.8
1.9
2.2
3.7
2.1
1.2
3.0
0.1
4.3
3.4
2.2
7.0
46.4
millions of
$ US
6.43
0.86
0.90
1.05
1.76
1.00
0.57
1.43
0.05
2.05
1.62
1.05
3.33
22.10
* 'includes air preparation, ozone generation, turbine contactors and off-
gas destruction equipment
     These costs were financed by the authority through funds drawn from
general revenues rather than through Issuance of new debt.  Since it is a
stock company owned by the city of Rotterdam, the water authority incorporates
its capital needs with those of the city and acquires funds in this way.  In
return, the water company pays for the cost of this capital in the form of
"interest" and depreciation payments to the city in future years.  For the
Kralingen plant, that annual payment is 7 MM gilders ($3.33 MM) annua'ly,
for a plant presently worth 170 MM gilders ($81 MM), about 4% per year for
the current worth of the plant and a higher percentage of its original cost.
                                      415

-------
     The total gross assets for the water company were 540 fM gilders ($257
MM) in April, 1978 values.  Including the new plant less accumulated deprecia-
tion of 140 MM gilders ($66.67 MM), net assets totalled 4CO MM gilders
($190.5 MM).  Hence the new plant was a large portion — about 75% — of the
net assets of the coirpany existing before its construction.  Such a large
capital program would strain the capabilities of utilities lacking the
resources and city ownership which Rotterdam enjoys.

Annual Operating Costs

In this section are discussed first the operating costs of the Kralingen
plant and then a breakdown of the overall system operating costs.  The plant
costs are based only on general categories of expenses, while the system
costs are based on the generalized steps in water supply from raw water
procurement to treatment and distribution.

     The operating costs of the plant total about 7 MM Dutch gilders ($3.33 YM)
annually.  Table 111 presents an approximate breakdown of these costs.

     TABLE 111.  ANNUAL OPERATING COSTS OF KRALINGEN PLANT
Cost Category
Wages - management & technical staff
Wages - operators
Maintenance
Utilities
Chemicals
Miscellaneous
TOTAL
millions of
Dutch gilders
0.9
1.0
2.5
1.2
0.5
0.9
7.0
millions of
$ US
C.43
0.48
1.19
0.57
0.24
0.23
3.34
     In addition to the 7 MM gilders for operating the Kralingen plant, the
water company pays 7 MM gilders to the City of Rotterdam to cover the cost
of the capital funds provided for the construction of the plant.  Hence, a
total of 14 MM gilders ($6.67 MM) is required annually to cover the full
plant costs.  On a unit basis, assuming average annual production from the
plant of about 120,000 cu m/day (32 mgd), these total treatment costs amount
to 1.2C gilders/1,COO gallons, or 57(t(U.$.)/!,000 gallons.

     Water treatment, however, is only one part of the total  process of
potable water supply.  For the entire Rotterdam system, plant treatment
costs about 30 MM gilders ($14.29 MM), or only about 26% of the total costs
of potable water supply.   The Kralingen plant costs of 14 MM gilders ($6.67 VM)
represent 47% of this total, while the plant provides about 38% of the
                                      416

-------
system production.  Table 112 presents a breakdown of the Rotterdam system
annual costs, excluding costs associated with a distilled water plant which
offsets its own operating expenses with separate revenues.

          TABLE 112.  ANNUAL KRALINGEN WATER SYSTEM COSTS
cost category
Raw water - procurement &
reservoirs
Treatment/production
Distribution
Administration & laboratory
City use of water
Other
TOTALS
costs in millions
of Dutch gilders
41
30
32
5
6
4-
118
% of total I millions
5 of $ US
35%
26%
27%
4%
5%
3%
100%
19.52
14.29
15.24
2.38
2.86
1.90
56.19
Company Revenues And Rates

     The Rotterdam City Council establishes the rates to be charged for
water.  The stated intent of this rate-setting authority is to cover ful'.y
all of the costs of providing water services.

     The total revenues in 1977 were 112 MM gilders ($53.33 MM),  which is
slightly less than the operating expenses shown above due to the  unbilled
provision of water to the city itself.  The total  amount of water delivered
during the year was 118 MM cu m (31,175.6 million  gallons), at a  calculated
average rate of C.95 gilder/cu m ($1.20/1,000 gal).  Table 113 shows these
totals and a breakdown by class of service.

     Since the water system serves 1.1 million people on a direct retail  and
a wholesale basis, the usage and revenues per person served can be calculated
from these figures.  The usage is 59 cu m/person/year, which converts to  42
gallons/capita/day.  At the assumed rate of 1.16 gilders/cu m ($1.46/1,000
gal), this amounts to about 70 gilders ($33.33)/person/year, or about 23
gilders ($10.95)/month for a family of four.
                                      417

-------
TABLE 113.  ROTTERDAM UATER SYSTEM REVENUES,  WATER DELIVERIES  AND  RATES
use category
Residential
Industrial
Wholesale
City use &
unaccounted for
TOTAL
revenues - millions
of Dutch gilders
49.7
45.0
14.6
2.4
111.7
deliveries -
millions of
cu m
42.8
51.0
22.4
1.4
117.6
calculated
rates, gilders
per cu m
1.16
0.88
0.65
	
0.95
US $ per
l.COO
gal
1.46
1.11
0.82
	
$1.195
                                    418

-------