WATER POLLUTION CONTROL RESEARCH SERIES 12120 ESW 06/71
i/71
   Magnesium Carbonate,
   A  Recycled Coagulant
    for Water Treatment
    U.S. ENVIRONMENAL PROTECTION AGENCY

-------
          WATER POLLUTION COITROL RESEARCH SERIES
The Water Pollution Control Research Series describes the
results and progress in the control and abatement of pollution
in our Nation's waters.  They provide a central source of
information on the research, development and demonstration
activities in the Environmental Protection Agency? through
inhouse research and grants and contracts with Federal, State,
and local agencies, research institutions, and industrial
organizations.

Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Chief, Publications Branch
(Water), Research Information Division, R&M, Environmental
Protection Agency, Washington, B.C. 20^60.

-------
MAGNESIUM CARBONATE, A RECYCLED COAGULANT
           FOR WATER TREATMENT
                   by
     Department of Public Utilities
           City of Gainesville
   P.  0.  Box 490,  Gainesville, Florida
                 for  the

     ENVIRONMENTAL  PROTECTION AGENCY

    Office  of  Research and Monitoring
          Project # 12120 ESW

              June, 1971

-------
                      EPA Review Notice
   This report  has been reviewed by the Environmental Protection
   Agency and approved for publication.  Approval does not
   signify that the contents  necessarily reflect the views and
   policies  of  the Environmental Protection Agency nor does
   mention of trade names or  commercial products constitute
   endorsement  or recommendation for use,
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $1.00


                              ii

-------
                             ABSTRACT
An entirely new system of water treatment has been developed.   It  is  a
unique  combination of softening and conventional coagulation,  and  may
be used for all types of waters, surface or ground, hard or soft.
Magnesium carbonate is used as the coagulant and lime is added to  pre-
cipitate gelatinous Mg(OH)2, which is as effective as alum for the
removal of both turbidity and organic color.  The floes formed are
larger  and heavier than alum floe, since they are "loaded" with CaCO^.
The sludge, composed of a slurry of CaC03, Mg(OH)2 and clay is  carbonated
with C02 and the Mg(HO)2 selectively and completely dissolved  as the
bicarbonate.  The carbonated slurry is filtered and the filtrate re-
cycled  to the point of addition of chemicals to the raw water  and  re-
precipitated with lime thus recovering both the coagulant and  the  sludge
water.  In small plants, the filter cake of CaC03 can be used  as land
fill.   In larger plants, it is slurried in a flotation cell and the
clay separated and used as land fill.   The purified CaCOo is filtered
and the cake passes to a multiple hearth furnace or Kiln and calcined
to high quality lime.   Chemical treatment costs are substantially  reduced
for most waters and the quality of the treated water is superior to that
treated with alum.

This report was submitted in fulfillment of Water Quality Research
Grant Project 12120 ESW with the Environmental Protection Agency.

Key Words:   Water Purification,  Coagulation, Chemical Precipitation,
            Sludge Treatment,  Industrial Wastes.
                               ill

-------
                               CONTENTS
Section




  I         Conclusions




  II        Recommendations




  III       Introduction




  IV        Theoretical Considerations




  V,        Experimental Materials and Methods




  VI,        Results and Discussion




  VII       Acknowledgements




  VIII      References




  IX        Appendix
 1




 3




 5




11




23




33




77




79



87

-------
                               FIGURES
 1      EFFECT  OF HIGH pH ON PQLIOVIRUS  1  (LSc)  IN
        FLOCCULATED  (500  mg/1 CaCOH)2),  SANDr-
        FILTERED SECONDARY EFFLUENTS  AT  25C                           10

 2      SOLUBILITY DIAGRAM FOR MAGNESIUM IN WATER AT
        ATMOSPHERIC  CONDITIONS,  TOTAL CARBONATE  10~%              13

 3      TEMPERATURE  INFLUENCE ON MAGNESIUM SOLUBILITY                 15

 4      EFFECT  OF pH ON MOBILITY FOR THE INDICATED
        COAGULANT DOSAGES        '                                     17

 5      SOLUBILITY OF MgC03XH20 AS A FUNCTION OF TIME
        FOR THE INDICATED HYDRATE FORMS                                20

 6      SOLUBILITY OF Mg(OH)2 (AS MgC03'3H20) AS A
        FUNCTION OF  pH FOR 23 NATURAL WATERS                           44

 7      PARTIAL TREATMENT COSTS IN $/MG  FOR CaO, C02
        AND MgC03 AS A FUNCTION OF COAGULATION pH                     46

 8      TREATMENT COST IN $/MG AS A FUNCTION OF
        COAGULATION  pH                                                47

 9      TREATMENT COST IN $/MG FOR CaO AND C02 TO RAISE
        THE RAW WATER pH TO 10.5 AND REDUCE THE pH BACK
        TO pH 9.0 FOR STABILIZATION                                   48

10      TREATMENT COST IN $/MG AS A FUNCTION OF THE RAW
        WATER TOTAL  ALKALINITY                                        49

11      TREATMENT COST IN $/MG AS A FUNCTION OF THE
        AMOUNT OF MgC03 PRECIPITATED                                  50

12      SETTLED COLOR AS A FUNCTION OF PARTICLE MOBILITY
        DURING COAGULATION, JACKSON, MISSISSIPPI WATER                56

13      SETTLED TURBIDITY AS A FUNCTION OF COAGULATION
        MOBILITY - LANETT WATER                                       57

14      MAGNESIUM RECOVERY BY CARBONATION                             65

15      FLOW DIAGRAM FOR TURBIDITY REMOVAL PLANT USING
        MgC03 AND LIME RECALCINING                                    69
                               vi

-------
                                                                    Page
                                                                      ^^< 

16      PHOTOGRAPHIC COMPARISON OF Mgco3 AND ALUM FLOC
        DURING RAPID MIXING IN THE REMOVAL 0? ORGANIC COLOR           72

17      PHOTOGRAPHIC COMPARISON OF MgCOo AND ALUM FLOC
        DURING FLOCCULATION IN THE REMOVAL OF ORGANIC COLOR           73

18      PHOTOGRAPHIC COMPARISON OF THE BATE OF SETTLING
        FOR MgCO- AND ALUM FLOGS FORMED IN THE REMOVAL
        OF ORGANIC COLOR                                              74

19      MAGNESIUM CARBON FLOC (MAGNESIUM HYDROXIDE DARK
        WITH ABSORBED COLOR AND CALCIUM CARBONATE CRYSTALS)
        MAGNIFIED 200 TIMES                                           75

20      PHOTOMICROGRAPH OF ALUM FLOC MAGNIFIED 100 TIMES             75

21      PHOTO-MICROGRAPH OF ALUM FLOC MAGNIFIED 200 TIMES             76
                              vii

-------
                                TABLES

NOj                                                                 Page

 1      Chemical Analysts pf Organic Color                            26

 2      Magnesium Carbonate Required for Coagulation of
        Organic Color and Emathlite Turbidity                         33

 3      Coagulation of Puller's Earth Turbidity With
        MgC03 and- With Lime                                           35

 4      Coagulation of a Synthetic-Water Containing
        Both Organic Color and Fuller*s Earth Turbidity
        with MgC03                                                    36

 5      Dosage of MgC03 Required to Coagulate Organic
        Color and Montinorillonite Turbidity                           34

 6      Coagulation of a Highly Colored Synthetic Water
        with MgC03                                                    37

 7      Effect  of Alum as a Flocculant Aid in Color and
        Turbidity Coagulation With Magnesium Carbonate                38

 8      MgCOo and Alum Coagulation of Montgomery, Alabama
        Water                                                         39

 9      MgCO., and Alum Coagulation of Mobile River Water,
        Mobile, -Alabama                                               40

10      Comparison of the Chemical Characteristics of 17
        Raw and Treated Natural Waters                                42

11      Economic Comparison of Treatment Methods for 17'
        Natural Waters                                                54

12      Relationship Between Electrophoretic Mobilities'
        and Settled Color or Turbidity for 12 Natural Waters          55

13      Amount of Magnesium Precipitated as Related to
        Physical and Chemical Characteristics for 17
        Natural Waters                                                59

14      Carbonation of Sludge Produced from the Coagulation
        of 36 Liters of Synthetic Water Containing 200 mg/1
        of Organic Color and 50 mg/1 Turbidity                        61

15      Carbonation-of Sludge Produced from the Coagulation
        of 36 Liters of Synthetic Water Containing 200 mg/1
        Organic Color and 15 mg/1 Turbidity                           62
                                  viii

-------
No;

16      Carbonatton of Sludge Produced from the Coagulation
        of 36 Liters of Synthetic Water Containing 50 mg/1
        Organic Color and 15 mg/1 Turbidity                           63

17      Carbonation of Sludge Produced from the Coagulation
        of 36 Liters of Synthetic Water Containing 15 mg/1
        Organic Color and 15 mg/1 Turbidity                           64

18      Carbonation of Sludge Produced from the Coagulation
        of 36 Liters of Natural Water Containing 200 mg/1
        of Organic Color and 50 mg/1 Added Montmorillonite
        Clay Turbidity                                                64

19      Coagulant Recovery Studies                                    66

20      Magnesium Solubility as a Function of pH for
        Coagulant Recovery Studies                                    60

21      Evaluation of Twice Recycled Magnesium in
        Coagulation of Synthetic Water                                68

22      Coagulation of Atlanta, Georgia Water with MgCO-
        and Alum                                                      88

23      MgCO.j and Alum Coagulation of Baltimore, Maryland
        Water                                                         89

24      Lime and Alum Coagulation of Birmingham, Alabama Water        90

25      MgCOo, Lime, and Alum Coagulation of Chattanooga,
        Tennessee Water                                               91

26      Coagulation of Cleveland, Ohio Water with Lime and Alum       92

27      Coagulation of Detroit, Michigan Water by Precipitation
        of Magnesium Present by Lime Addition                         93

28      Coagulation of Huntsville, Alabama Water With MgCO
        and With Alum                                                 94

29      MgCO^ and Alum Coagulation of Jackson, Mississippi Water       95

30      MgC03 and Alum Coagulation of Lanett,  Alabama Water           96

31      Lime and Alum Coagulation of Louisville,  Kentucky Water       97

32      MgC03, Lime,  and Alum Coagulation of Nashville,
        Tennessee Water                                               98
                               ix

-------
No.                                                                 Page

33      MgCOo and Alum Coagulation of Opelika,  Alabama Water          99

34      Lime, MgCOn, and Alum Coagulation of Philadelphia,
        Pennsylvania Water                                           100

35      Coagulation of Richmond, Virginia Water With
        MgC03 and Alum                                               101

36      MgCOo and Alum Coagulation of Tuscaloosa,  Alabama Water      102

37      MgCOo, Lime, and Alum Coagulation of Washington,  D.  C.
        Water                                                        103

38      Calculated Potential Production of MgCOo'SHLO by
        20 American Cities, 1968                                     104
39      Calculations of Potential Consumption of
        by Water Treatment Plants in the United States                106

-------
                       SECTION I.  CONCLUSIONS
1)  Magnesium carbonate,hydrolyzed with lime, is as effective as alum
    for the removal of both turbidity and organic color from surface
    waters.  The floes formed are larger and heavier than alum floes
    and settle better.

2)  An optimum dosage of magnesium hydroxide is required for acceptable
    treatment of each water.  This requirement is related to the
    physical and chemical characteristics of the water, with organic
    color having the greatest effect.  This relationship, determined
    for 17 natural waters, is:
         Optimum MgC03'3H20 dosage (mg/1)
         = 8.33 + .03 (turbidity)
         + .46 (organic color) - .03 (total alkalinity)
         + .14 (total hardness).
    The coefficients for color and turbidity in the several natural waters
    used were in excellent agreement with the coefficient in the equations
    developed for synthetic waters.

3)  Neither the base exchange capacity of the clay nor the level of
    turbidity present had a significant effect on the optimum coagulant
    dosage.

4)  For each water studied, one specific flocculant was usually found
    superior to all others tested.   In general, however, soft waters
    responded best to 0,5 ppm dosages of alum, whereas either activated
    silica or potato starch produced better floes in hard waters.

5)  The measurement of electrophoretic mobility is an effective analytical
    tool for evaluating coagulation  efficiency, particularly in the
    coagulation of organic color. Charge reversal does not result in
    restabilization of the colloidal particles.

6)  In the study with natural waters, considerably more magnesium was
    found in solution after coagulation than would have been predicted
    from theoretical magnesium hydroxide solubility data.   This may
    have been due to the  fact that true equilibrium was not reached
    in the relatively short settling periods allowed.   A significant
    difference was found  between values for residual magnesium determined
    by EDTA and those determined by  atomic absorption spectrophotometry.

7)  A series of graphs make.s possible the determination of the optimum
    coagulation pH and the chemical  treatment cost for lime, CO,,, and
    MgCOo based on jar test results.   While the cost data are based on
    conservative estimates,  significant savings in treatment costs are
    indicated for most waters without considering the many other benefits
    resulting from the use of this new treatment process.

-------
8)  Carbonation recovered nearly 100% of the magnesium hydroxide from
    the sludge with an insignificant amount of calcium solubilization.
    Release of coagulated color is not a problem where the average color
    in the water is less than 100.  Filterability of the carbonated
    sludge will not be a problem for most waters.  However, in the case
    of very soft waters the sludge will be low in CaCO- and it may be
    necessary to add a synthetic organic polyelectrolyte or an inert
    filter aid to provide satisfactory filterability.

9)  The use of magnesium carbonate in most cases produces a treated water
    with superior chemical characteristics compared to water treated with
    alum.  When magnesium is used, the treated waters have alkalinities
    ranging from 30 to 50 mg/1, giving soft waters sufficient alkalinity
    for calcium carbonate stabilization or softening waters high in
    carbonate hardness.

-------
                     SECTION II.  RECOMMENDATIONS
The problem of neglecting important research areas in new technological
studies is discussed in the Introduction.  While many new and exciting
findings are reported as a result of this work, many more challenging
areas of research have been exposed.  For discussion purposes, these
studies are divided into "applied" and "basic" research studies.

A two year research project should be initiated to implement this new
process, first on pilot plant scale and later in a suitable municipal
water plant.  This study is required before this new process can be
utilized in water plants throughout the country.  The objectives of such
a project would be to study or evaluate the following:

    1)  Operational procedures and problems in the use of this new process
        to treat a soft, colored, highly turbid river water subject to
        wide variations in chemical and physical characteristics.

    2)  Economic considerations of all cost involved in the use of this
        new process including projected capital expenditures for plant
        modifications.

    3)  Instrumentation and control of the carbonation and coagulant
        recycle processes.

    4)  Sludge filterability and the methods available for improvement.
        The effect of increasing hardness on sludge filterability will
        be studied by adjusting the raw water characteristics over a
        fairly wide range.

    5)  The effect of this new process on the future design of floccula-
        tion and settling basins.  Laboratory observations indicate that
        less mixing and settling time is required as compared with alum
        treatment.

    6)  The benefits resulting from the increase in alkalinity and
        hardness resulting from the treatment of very soft water using
        magnesium carbonate.  Comparison will be made of the corrosion
        or deposition potential of waters treated with magnesium
        carbonate and with alum.

    7)  Bacteriological removals in each unit to determine if pre-
        chlorination is required.

    8)  Iron and manganese removal to determine if the high pH is
        sufficient to precipitate the two metals completely.

-------
Magnesium carbonate offers considerable promise in the treatment of
municipal and industrial wastes.  Magnesium forms a very insoluble
complex, magnesium ammonium phosphate, which, when precipitated with
MgCO-j and lime, could provide tertiary sewage treatment at an acceptable
cost.  Simultaneously color, turbidity, and biological and chemical
oxygen demand would be reduced to low and hopefully to satisfactory
levels.  Lime and magnesium values can be recovered and recycled as in
water treatment, thus reducing the chemical costs.  A research project
to investigate this application has been begun at Gainesville, Fla.

Magnesium carbonate treatment of industrial wastes high in color such as
the pulp and paper wastes, tannery wastes, textile dye wastes, etc.
appears very attractive.  More industrial plants have an unlimited supply
of carbon dioxide in the stack gases and many already recover lime.  A
method of coagulant recycling for these highly, colored wastes should be
developed.  This might be accomplished by the addition of a strong oxidant
to the recycle stream after which the magnesium could be recovered by
precipitation as magnesium carbonate tri-hydrate.  The precipitated
magnesium could be lightly calcined, 500-600C, to burn off the organics
and then reused.

The release of absorbed undesirable constituents by carbonation for
coagulant recovery is in need of further study.  For an industrial waste
this might include heavy metals while the release of iron and manganese
would be of concern in water treatment.

Basic research on the chemical interactions between magnesium and com-
plexing ligands should be undertaken.  Complexation, chelation, and ion
pair formation should be studied.  These mechanisms could possibly ex-
plain the lower magnesium values reported by EDTA titration as well as pro-
vide information as to how this problem can be eliminated.  It is possible
that the stability constants for some of these magnesium-ligand complexes
are greater than that for the EDTA-magnesium complex under certain con-
ditions.  Possibly a strong oxidant will destroy these complexes,
allowing an accurate titration of the magnesium by EDTA.

The identification of various magnesium species formed at high pH is an
area for further research.  Possibly polynuclear, strongly hydrated
hydrolysis products form, as has been proposed for aluminum.  Additional
study of the effect of both anions and cations on the mobility of
magnesium hydroxide floes should be conducted.

The removal of tastes and odors is a serious problem faced by practically
all municipal water treatment plants treating soft surface waters.  The
effect of the very high pH values used in coagulation with magnesium
carbonate and lime on the removal of these tastes and odors by activated
carbon and by other methods should be carefully studied.

-------
                      SECTION III,  INTRODUCTION
Iron and particularly aluminum salts have served well in the coagulation
and clarification of water since their common usage in the early 1900's.
Many investigators consider the highly hydrated and gelatinous property
of the hydrolysis products to be a main attribute for their effectiveness.
This property however makes dewatering and disposal of the sludge produced
from the treatment process extremely difficult and costly.  Water plant
wastes are recognized today as an industry-wide pollution problem that
must be solved.  It is estimated that over 1,000,000 tons of alum sludge
are produced each year with less than eight percent receiving treatment
of any kind before disposal.

The characteristics of the waste products from water plants are highly
variable both within and among plants.  A considerable effort has been
made to characterize these wastes with the following ranges in character-
istics being reported:2>3,4
        Total Solids
        Suspended Solids
        Volatile Solids
        BOD,5x  (Ultimate BOD
         considerably higher)
        COD  (Higher value where
         activated carbon present)
1,000 - 17,000 mg/1
75% - 90% of Total Solids
20% - 35% of Total Solids

30 - 150 mg/1

500 - 15,000 mg/1
The reduction of volume and moisture content is of primary concern in
alum sludge disposal.  In a study of two water plants, Neubauer  found
the volume of alum sludge produced  to range from  .12 to  .26% of the total
plant flow.  Methods, which have been employed with varying degrees of
success, to concentrate and dewater alum sludge include:

    1)  Gravity thickening, stirred thickeners, and lamella
        sedimentation  >"'

    2)  Lagoons2'3'5

    3)  Drying beds8
                             Q
    4)  Wedge wire filtration

    5)  Vacuum filtration6

    6)  Pressure filtration

    7)  Centrifugation6

    8)  Freezing6

-------
As a means  of offsetting some of the costs  for treating the sludge,  alum
recovery has been attempted at several water plants,   Jewell10 in  1903
patented sulfuric acid regeneration of alum and in 1923 Mathis11 obtained
a similar patent.  In 1951 Black Laboratories12 suggested utilizing  sulfur
dioxide gas from boiler stacks as a source  of sulfuric acid for alum
recovery at Orlando, Florida.  Roberts and  Roddy13 reported on investiga-
tions for alum recovery at Tampa, Florida which later  was practiced  for
a short while but was discontinued due to operational  problems,  Tampa's
source of raw water varied widely in hardness and organic color content.
Aluminum sulfate was used only during times of high organic color
making recovery an intermittent operation.   Higher dosages of recovered
alum had to be used due to the release from the sludge of organic  color,
reducing the effectiveness of the coagulant.   The Asaka Purification
Plant in Tokyo reports recovery of as much  as 80% of the aluminum  using
the sulfuric acid process.1^ Iron or manganese, which might be present
in the sludge, is also solubilized.  This causes an increase in the  con-
centration  of these elements, making it necessary to waste a portion of
the recovered alum from time to time.1^  Alum recovery is also practiced
at the Daer Works in Scotland** and is being investigated for use at
Minneapolis, Minnesota.^

Lime recovery in softening plants is practiced in several American cities
with profitable operations being reported for two cities.1"'1^  The
excess lime produced is sold to neighboring cities for a profit.   Lime
recovery is presently economically attractive only for softening plants
using 20 to 25 tons of lime a day.

While there are many methods of treating an alum sludge, only in unusual
cases has the treatment been found to be satisfactory  or economically
attractive.  The AWWA conference report on  plant needs1" states

        In  summation, the principal needs are to find  effective
    and economical means, through research, to dispose of water
    treatment plant wastes by direct treatment of sludge, or by
    eliminating undesirable chemicals, such as alum, through changes
    in water treatment methods.

In answer to these needs, an entirely new system of water treatment
chemistry has been developed utilizing magnesium carbonate as the  coagu-
lant.  The  addition of sufficient lime slurry to a water containing
magnesium bicarbonate and/or to which magnesium carbonate has been
added, precipitates both magnesium hydroxide and calcium carbonate.
Carbonation of the sludge solubilizes the magnesium, as magnesium  bi-
carbonate,  which can be recovered by vacuum filtration and the filtrate
recycled and reused.  The filter cake of CaCO^ and clay is easily
handled and disposed of as land fill.  Lime recovery would further
reduce the  volume of sludge to be disposed  of.

The flocculant properties of magnesium hydroxide have  long been evident.
It is these properties, however, which have troubled conventional  water

-------
softening plants,  The gelatinous nature of magnesium hydroxide, in
many ways similar to the hydrolysis products of aluminum, makes the
dewatering of the sludge difficult.  Also, where recovery of lime is
practiced, the magnesium hydroxide must be separated before calcination
because of build-up of insoluble magnesium oxide in the recovered lime.
Various techniques have been developed to separate the magnesium hy-
droxide from the calcium carbonate.  The use of a centrifuge to selec-
tively classify it into the centrate does not provide the degree of
separation needed.  Three-phase selective softening has been used at
Lansing, Michigan for some years during the winter months of low water
demand.1^  In the research activities herein described, the dosage of
lime added in the first phase is just sufficient to precipitate the
calcium bicarbonate and convert the magnesium bicarbonate to the soluble
carbonate.  In the second phase, excess lime is added to precipitate all
magnesium hardness as Mg(OH)2>  This requires a lime dosage well above
the stoichiometric amount.  This second phase effluent with a pH of
about 11.3 passes to the third phase where it is mixed with just enough
raw water to utilize the excess lime in removing calcium hardness.
Phases 1 and 3 sludges are mainly CaCO-, - phase 2 sludge mainly Mg(OH).
Since this process cannot be used to soften waters containing either
color or turbidity and requires extremely careful control, it is seldom
used.

Bureau of Mines Technical Paper No. 684 * describes an industrial process
for separating MgO from its ores, brucite or dolomite, in which the
calcined and finely ground ore is slurried and chilled.  The MgO present
is then dissolved by pure C0 with continuous cooling to neutralize the
high, heat of hydration of the MgO.

                  9 n
Black and Eidsness   were able to selectively dissolve the Mg(OH)2 from
the CaCOo in the lime-soda softening sludge at Dayton, Ohio, thus making
it possible to recalcine the CaCO., and produce high quality quicklime.
This sludge carbonation basin, the only one of its kind in the world,
has been operating successfully since 1958.

While magnesium hydroxide was formerly regarded as a liability it has
been recognized as an effective coagulant.  Flentje,^! in 1927, found
increasing clarification efficiency in the water treatment plant at
Oklahoma City as excess lime was added.  He reasoned this to be due to
precipitation of magnesium as the hydroxide.  Several jar tests were
performed which indicated that magnesium, in the form of magnesium
chloride, is an effective coagulant.  The excess lime treatment was
practiced,in conjunction with ferric sulfate on a full plant scale, to
treat the hard, turbid river water.  The objective was to employ the
magnesium bicarbonate naturally present in the water.  Flentje noted no
decrease in filter runs, less algae in the settling basins, and greater
bacteria removal.  No attempt at coagulant recycle was made and the
excess lime fed was not sufficient to quantitatively precipitate the
magnesium present in the water.

-------
In 19667 Lecompte^^ reported the use of magnesium carbonate as a coagulant
for the reclamation of water within a paper mill,  Lime was reacted with
magnesium carbonate, produced by the reaction of finely ground magnesium
oxide and bicarbonate alkalinity present in the water, to precipitate
magnesium hydroxide.  No attempt was made at magnesium recovery.  The
water to be treated contained 0,5 to 1.0 pounds of suspended solids per
one thousand gallons with fluctuations in organic color.  The chemical
cost of the water produced was estimated at sixty-five dollars per
million gallons with additional disinfection benefits noted due to the
excess causticity.

Although water chemists have long recognized the effectiveness of
magnesium hydroxide as a coagulant, the use of magnesium salts has not
received acceptance for economic reasons.  Both the chloride and the
sulfate cost more per pound than alum and their use, in conjunction with
lime, would increase the non-carbonate hardness of the water being treated
in direct ratio to the make-up dosage required.  The cost of magnesium
carbonate currently quoted at 16
-------
Alum and ferric sulfate have been shown to be effective in the coagulation
of viruses,^?  However, these investigators found that viruses removed
in the floe fraction were not destroyed and active virus can be recovered
from the floe.  Berg e al_   have found that disinfection of polio virus
can be accomplished by high pH,  Figure 1, from this publication, shows
the effect of pH on the survival of polio virus as a function of time.
It can be concluded that to some extent bacteria and virus can be re-
moved by the new treatment process.

In this high pH range of coagulation, complete precipitation of iron and
manganese should occur, possibly eliminating the need for more costly
treatment methods.  This pH environment would be unfavorable for aquatic
growths in settling basins.  The overall effect would be reduction in the
use of chlorine and subsequent cost savings at the same time increasing
the treatment efficiency.

In initial studies of new technology, important areas of research must
be left for future investigations because of time limitations.  The use
and recycle of magnesium carbonate is an entirely new concept in water
treatment chemistry.  Initial research efforts have been planned to
determine if this process is technically and economically feasible.
The scope and objectives of this research were therefore as follows:

    1)  Evaluation of the parameters involved in the use and recycle
        of magnesium carbonate as a coagulant for both organic color
        and turbidity in soft waters.  Studies of both synthetic and
        natural waters are included in this phase of the research.

    2)  Development of a predictive equation to determine the magnesium
        requirements based on the physical and chemical characteristics
        of a water,

    3)  Demonstration of the effectiveness of this new process on a
        broad spectrum of natural waters.  Waters from the largest
        cities in the country were chosen to provide a wide spectrum
        of range in chemical and physical characteristics.

    4)  Estimation of the chemical cost of treatment using this new
        technology and comparison with the chemical costs using alum
        treatment.

    5)  Comparison of the chemical characteristics of the treated waters
        using magnesium carbonate and alum treatment.

-------
   100
    10
                             pH 10.1

                             pH 10.8

                             pH 11.1
                 20
                            40
60         80
   Time-min
                                                            100
                                                                       120
140
Fig. 1.  Effect of High pH on Poliovirus 1 (LSc)  in Flocculated  (500 rag/1 Ca(OH)t),
                       Sand-Filtered Secondary Effluents at 25 C

                       (from  Berg e_t  al. ^)

-------
                SECTION IV,  THEORETICAL CONSIDERATIONS
Magnesium Equilibrium In Water

Magnesium is present to some extent in almost all natural waters.  As a
rule, magnesium content increases with increasing water hardness.  The
ratio of Mg"l~t"/Ca"H" is quite variable but almost always less than 1.

In Pearson's system of classifying metals,   magnesium is classed in
group A, the same as aluminum.  The elements of this group are visualized
as having spherical symmetry with the electron sheaths not readily de-
formed by adjacent charged ions.  Metals in this classification tend to
form insoluble precipitates with OH", CO- , and P07  with simple
electrostatic binding of cation and ligand used to explain complex
stability.

In natural water systems, magnesium can be found in many solid phases.
Considering only a system composed of carbln dioxide, magnesium, and
water, one can calculate which solid phase controls magnesium solubility.
The solid phases of magnesium present are:^-*

                                                  A 30                
                                                  A            -log Kso30
                                                               -
                                                (KCal
    1)  Brucite - Mg(OH)2

        Mg(OH)2(s) * Mg++ + 20H"                  -15.8          11.6


          108             -* + 2pK - 2pH
          log [Mg] = 16.4 - 2pH

    2)   Magnesite - MgCO,,
                        3

        MgC03(s)  * Mg"^ + C03=                      6.7           4.9

          log [Mg^] - -pKSQ - log^ - log a2

          log iMg**] = -4.9 - logc  - log a2
                                 11

-------
                                                               -log
                                              (KCal mol"1)

    3)  Nesquehonite - MgC03'3H20

       MgC03'3H20(s) * Mg44 + C03 +  3H20           +7.4            5.4

         log [Mg*] = -PKSO - log^  -  log 2

         log [Kg"*"1"] = -5.4 - logc  - log  a2

    4)  Hydromagnesite - Mg4(C03)3(OH)2' 3H20

       Mgi(C03)3(OH)2'3H20 * 4Mg+2 + 3C03~2

         + 20H- + 3H20                            +40.2           29.5

         log [Mg^] = -l/4pKsQ + 1/2PK2

           - 3/4 logr  - 3/4 log   -  1/2PH
                   U'p            /

         = -.4 - 3/4 log,,  - 3/4 log a9 - l/2pH
As Nesquehonite is less soluble than magnesite at all  pH values,
magnesite will not be  considered.   In  Figure 2, a pH-stability diagram
is shown for a total carbonate concentration of 10"%.   Brucite is  by
far the least soluble  at pH values  above 9 with hydromagnesite controlling
solubility from pH 9 to approximately  7.5.  Nesquehonite is the least
soluble at pH values below 7.5.  Dolomite, CaMg(C03)2,  is a very  common
stable phase found in  nature but attempts to precipitate a dolomite phase
from supersaturated solutions under atmospheric conditions have been
unsuccessful.

Considerable effort has been expended  in determining  the solubility
product-constant for Mg(OH)p.  The  following table  lists some of  the
values reported in the literature.
            Investigator                    t C          pk,
                                                          SP
       Gallaher32                         25-30         11.28
       Ryzner et_ al.33                       80          11.28
       Krige and Arnold34                    20          10.85
       Travers and Nouvel35                  18          10.60
       Kline36                              25          11.00
       Britton37                          Room          10.64
       Bube38                               25          10.92
       Gjaldbach39                           18          10.92
       Dupre and Bialas40                    18          10.92
                              12

-------
I-'
                            -3
                            -4
                            -S
                                                              9
                                                              pH
                                                                                       13
FIG. 2
                                  SOLUBILITY DIAGRAM FOR MAGNESIUM  IN  WATER AT ATMOSPHERIC
                                  CONDITIONS.   TOTAL  CARBONATE  I0's M
                                   (Stumm, W. and James  J. Morgan.  Aquatic Chemistry,  Fig. 5-10)

-------
             Investigator                   tC         p_kgp

        Herz and Muhs41                      29         10.31
        Kohlrausch and Rose                  18         10.87
        Loven43	10	10.76

Magnesium hydroxide becomes  less  soluble at increased temperature.
Figure 3 taken from Larson,  Lane,  and Neff44 shows this effect.

The solubility products reported were for pure distilled water systems,
extrapolated to zero ionic strength.  Many factors tend to increase
the solubility of magnesium  in natural waters.  Solubility increases
with increase in ionic strength as expressed by the Debye-Huckel relation
ship:

                 pK   = pCKon -  (nZM2 + mZN2)("5
                        r    o
                   so   "   so                ' 'I +

        where:

                 M N  (s) * nM + mN
                  n m

                 Kso -  [M]"[N]' Aj A*

                 u   =  ionic strength

An illustration of this effect on the K   for Mg(OH)  using a pK   =11.0
           u = 0     u  =  .01    u =  .1      u =  .3                S
pK         11.00      10.73     10.20       9.95
  sp

Cotnplexation of the magnesium with both organic and inorganic ligands
increases the solubility.   The formation of ion pairs also tends to
increase the solubility.   Ion pairs  differ from complexes in that the
matal ion and the base  are separated by one or more water molecules
while for a complex the ligand is immediately adjacent  to the metal
cation.    It is reported  that while complex formers present in solution
may often have little or no effect on the solubility of solids, they may
however affect the kinetics of nucleation and of growth and dissolution
of crystals.4-*

Lime is commonly used to precipitate magnesium  from water as magnesium
hydroxide.  The hydroxide  concentration of the water can be increased
to the necessary level  only after converting all of the C02 and
to C03=.  These well known softening reactions  are:

    1)  C02 + Ca(OH)2 * CaCO_3 + H20

    2)  Ca(HC03)2 + Ca(OH)2 * 2CaCO_3 +  2H20
                                  14

-------
.12.0
no
 10.0
 9.0
   30
           50
            Fig. 3
                  70
                                                 150
                                                        170
              110      130
              Temoerature - *F
Temperature Influence on Magnesium  Solubility
                                                                190
                                                                       210
                                                           (from Larson, Lane,
                                                              and Neff44)
The. dashed  curves represent magnesium solubility  (as  parts per
million CaC03);  the solid curves,  pH variation.  The  solubility
curves are  based on the solubility product constants  of  Travers
and Nouvel.35

-------
Magnesium bicarbonate is converted to magnesium carbonate and magnesium
hydroxide on further addition of lime as;

    3)  Mg(HC03)2 + Ca(OH)2 2 MgC03 4- CaC03 + 2H20

    4)  MgC03 + Ca(OH)2 * Mg (OH) 2 + CaCO,

If the magnesium in the water is non^-carbonate hardness, there would be
no net change in total hardness, only an exchange of calcium for magnesi-
as:

    5)  Kg"1"* + S04= + Ca(OH)2 * Mg(OH)2 + Ca^ + S04=

Magnesium carbonate used as a coagulant does not add to the total
dissolved solids as shown in equation (4) .   The lime dosages necessary
for coagulation and softening can be calculated as:

        Reaction                       Lime required, mg/1 of Ca(OH)2

CO, + Ca(OH)9                          C00  x Ca(OH)2   2i  =
  22                                 C02      44

2(HCOo~) + Ca(OH)2                     Alk (as CaCO,) x Ca(OH)2   74  =
                                                   J    CaC03    100

MgCO  '3H,0 + Ca(OH),                   MgCO,,-3H00 x  Ca(OH)2    74  =
    3   l          z                       J   2    MgC03-3H20 100
Mg^ + Ca(OH),                         Mg4^ (CaCO,) x Ca(H)2   _Zi  =
             z                                   J    CaC03     100

In practice 90% pure lime CaO would be slaked and used.  Thus, the total
lime dosage found above should be multiplied  by 100 56  CaO     Or 0.82
times the Ca(OH)  value determined.               90 74 Ca(OH)2

CaC03 suspended in water has been found to be negatively charged while
magnesium hydroxide is positively charged.    While the particles have
been found to coexist, absorption usually takes place and one pre-
dominates, giving the floe either a net positive or negative charge.
For a  water containing both calcium and magnesium, the mobility tends
to become less negative as the pH increases.  Figure 4 demonstrates
this effect found in the coagulation study of water used by Montgomery,
Alabama.  This is due to formation of Mg(OH). which can cause charge
reversal if sufficient magnesium is present.  Flocculation of calcium
carbonate suspensions using coagulant aids is not necessarily accompanii
by a decrease in negative mobility,   These investigators reported thai
mobility in itself is not a reliable indicator of the degree of floccul,
tion.

Salts other than magnesium carbonate could be used as the source of
magnesium.  Once the magnesium is recycled, it would be in the carbonat'
                                16

-------
10.7
                                                       pH VS. HOBIUTt
                                                       O - 50 s/l KgCOj
                                                       Q - 45 g/l MgCO
                .2      .3
                                 Hoblllty(-) U//V/CB
             FIG,  4  EFFECT  OF pH ON MOBILITY   FOR  THE
                     INDICATED  COAGULANT  DOSAGES

-------
or bicarbonate form,  However any make-up magnesium salt, such as MgSO.
or MgCl2, would increase the non-carbonate hardness as shown in reaction
5.

Magnesium Recovery

As discussed in the introduction, the solubilization of magnesium by
carbonation has been practiced by industry for many years.  However,
such processes have generally been carried out with supersaturated
magnesium solutions and pure CC^.  While these processes must be care-
fully controlled, magnesium recovery from water plant sludges is quite
simple, with little control required.  The reactions which take place are:

               Mg(OH)2(s) + C02 * MgC03 + H20

               MgC03 + C02 + H20 2 Mg(HC03)2

Whether the reaction occurs in one or two steps is not known.
                   on
Black and Eidsness,   carbonating a sludge containing Mg(OH)2 and 36 g/1
of CaCO., with 11% C02, found that only 80 mg/1 of CaCOo was dissolved
after 30 minutes' carbonation at a gas flow five times that required to
dissolve all of the Mg(OH)  present.  According to Johnston,   who studied
the solubility of calcium and magnesium carbonates in natural waters, the
equilibrium ratio at 16C is [Mg*f ]/ [Ca"14"] = 14,000 when the partial
pressure of C02 in the atmosphere is great enough to prevent precipitation
of Mg(OH)2>  Another explanation for these phenomena is that a saturated
solution of Ca(HCC>3)2 has a lower pH than a saturated solution of Mg(HC03)2
As carbonation proceeds, the pH is buffered at a pH of approximately 7.5,
due to the Mg(HC03) , allowing little of the calcium to dissolve.

Magnesium hydroxide may also react with the bicarbonate to produce
magnesium carbonate as:

               Mg(OH)2 + Mg(HC03)2 4- 2H20 * 2MgC03'3H20

If complete solution of Mg(OH)2 is desired, obviously precipitation of
MgCO-i'SH^O should be avoided.  In practice, this is avoided by incremental
addition of fresh sludge to the carbonation basin and maintaining a
sludge-water ratio such that a super-saturated solution of Mg(HC03)2 is
not produced.

Production of Magnesium Carbonate

At present magnesium carbonate is produced from four major sources.

    1)  From sea water without evaporation, using sea water and  lime
        as the principal raw materials.

    2)  From bitterns or mother liquors from the solar evaporation of
        sea water for salt.
                                  18

-------
    3)  From denp^well brines.

    A)  From dolomite.

Investigators are not in agreement regarding  the formulae for the several
forms of magnesium carbonate.  For this dissertation it will be assumed
that at least three salts may be prepared by  aerating an aqueous solution
of magnesium bicarbonate:  the penta-hydrate  MgCO~'5H,,0 precipitated be-
low 13.5C; the tri hydrate MgCC^-SJ^O precipitated between 13.5C and
50C and a "basic carbonate" whose composition is most commonly given
as AMgCOyMgCOH^'x FLO precipitated above 50C, most rapid.ly and coir-
pletely by boiling.  Both the penta-hydrate and the tri-hydrate slowly
revert to the basic carbonate, SMgO'ACC^'xH^O, upon exposure to the
atmosphere.  This reversion is accelerated when moisture is present and
at elevated temperatures.  When heated to 100C, dry MgC03'3H20 is quite
stable.    MgCC^'Sl^O exhibits an interesting change in solubility or.
heating to 100C.  Figure 5 demonstrates this increased solubility
effect.  The data for this figure were collected from analytical studies
of the MgCOo'S^O sludge produced in Dayton,  Ohio.  It is assumed that
the conversion to the basic carbonate involves recrystallization of the
aqueous solution.  At 200C, dry material loses water and COp without
the addition of water.  Possibly partial decomposition furnishes some
wa.er which can then assist further conversion. '

Magnesium carbonate, which is used primarily  in the paint, printing,
rubber and pharmaceutical industries, sells from $.16/lb for the tech-
nical grade tc $.22/lb for the USP grade. 50   Most of the product produced
today is the basic carbonate,
As discussed in the introduction, a new source of magnesium carbonate
will be soon available at a greatly reduced cost.  A process has been
developed by A. P. Black and the city of Dayton, Ohio to recover it from
water softening plant sludges. *  Preliminary calculations, included in
the appendix, indicate that as much as 150,000 tons of magnesium carbon-
ate (MgCO^'S^O) can be produced each year by the twenty cities shown.
Substitution of recycled MgCOo'StUO for alum in the more than 4,000
wf.ter treatment plants now using it would require approximately 100,000
tons per year, assuming 85% recovery and re-cycling of the magnesium
and 15% make-up.  These calculations are also included in the appendix.
The cost to produce MgCOs'Sl^O with this new technology has been estimated
to be less than $.02/lb.5l

Colloidal Destabilization

Colloidal destabilization is believed to occur in two steps.  The first,
which is assumed to occur very rapidly, has been referred to as
perikinetic coagulation-5  or coagulation.    In this step, chemical and
physical interaction between the colloid to be removed and the coagulant
takes place.  Two broad theories have been advanced to explain the
mechanism.  The older, chemical theory, assumes stabilization to be due
                                   19

-------
                           H@ofe4 1o I05C for I hr. e
                           200*C for 2 hre. as
                           2OOC tor 2 tir. oe
                           Air drM os HgCOs  5H20
S      10
  TIHE (Uln.)
                                       12
                                               14
FIG. 5  SOLUBILITY  OF  ffgCOs-XHtO  AS  A  FUNCTION  OF
       TIME  FOR THE INDICATED  HYDRATE  FORMS

-------
to chemical interactions, such, as complex formation and proton transfer.
The physical theory emphasizes the concept of the electrical double
layer.  CounterT-ion adsorption and compaction of the diffuse portion of
the double layer are assumed to neutralize the colloidal charge and
bring about coagulation.

After coagulation, orthokinetic coagulation5  or flocculation50 takes
place, normally requiring a longer time period with gentle mixing con-
ditions.  During this step, interparticle bridging of the coagulated
colloids forms larger floe particles.

The first coagulants, alum and iron salts, were chosen for their highly
gelatinous properties.  Later investigators attributed the Shultz-
Hardy-^j" effect as the main attribute of these coagulants.56  However,
more recent investigators found that the hydrolysis products were much
more effective than the trivalent metal cations.52,57,58  Many investigators
have reported the effect of anions on the broadening of the optimum pH
for coagulation.-*" >57,59  ^he displacement of hydroxide ions by highly
coordinating anions has been proposed as the mechanism for these
phenomena.

In 1928, Mattson  demonstrated the relationship between microelectro-
phoretic mobility of colloidal particles and the aluminum salt dosages.
However, this technique remained almost forgotten until 1959 when
Pilopovich et^ al.   studied the effects of pH, alum dosage, zeta
potential, and base exchange capacity of clay particles on coagulation.
Several investigators reported that base exchange capacity was an
important factor in coagulation.52,58  jt was aiso found that, while
good coagulation often occurred near zero mobility, no absolute relation-
ship existed.5

In a series of papers, Packham proposed that coagulation was due almost
entirely to a physical enmeshing or sweeping down of particles by the
highly gelatinous property of the aluminum hydrolysis products.5'  He
found that the type of dispersed phase had relatively little effect on
coagulation conditions.   Packham's work and other recent investigators
seem to support the early contentions that the sticky, gelatinous,
property of a coagulant is possibly most important.  The mechanism of
coagulation seems to be dependent upon the properties of the dispersed
phase and the conditions which are present during coagulation.

Chemical and Physical Properties of Organic Color

While the origin of organic acids found in natural waters is still a
subject of controversy, most investigators report that it is due to
aqueous extraction from soil or decaying vegetation.  Some investigators
propose that color in water is an intermediate step in the transformation
of organic matter from living or decaying woody tissue to the soil
organic complex."2
                               21

-------
The structure of the organic acids has been proposed to be aromatic with
hydroxy, carboxy, methoxy, and carboxylic acid groups.    Black and
Christman, using an electrodialysis  cell with membranes of varying
pore sizes, found that 78% of the organic acid molecules were between
3.5 and 10 my.    However, li-Bljt scatter data for this same organic
color suggested a larger size,  ^  An elemental analysis of organic color
extracted from ten highly  colored natural waters gave the following
range of results,"-^

                Carbon - 44.99  - 54,10%
              Hydrogen -   3.86  -  5.05%
              Nitrogen -   1.46  -  4.23%
                Oxygen - 38.76  - 47,93%

Some investigators have found nitrogen present, but  the carbon/nitrogen
ratio is reported to be higher  for  more highly colored lakes indicating
nitrogen as an impurity.     The molecular weight of  the colored acids
is generally believed to range  from 450 to  10,000.   However Gjessing,
using Dioflo ultra-filtration membranes reported approximately 85 percent
of the organic acids present had a  molecular weight  higher than 20, 000. ^

Kitano observed that organic acids  influenced inorganic solubility
equilibrium and subsequent precipitation products. "^ Shapiro has found
considerably more iron in  solution  in natural waters than would be pre-
dicted by theoretical solubility equilibria.    He proposed peptization
of the iron by humic acids with some chelation as  the possible mechanism.
Oldham and Gloyna^ propose the mechanism  for increased iron solubility
to be the ability of the humic  acids to reduce Fe+^  to its more soluble
form Fe+2 and  the subsequent complexation  of the iron by the humic acids.

Color has been found to vary in intensity  as a function of pH.  Singley
e_t_ aJL   have  developed a  nomograph which  will correct the color  intensity
at any pH to  that at pH  8.3, an arbitrary  standard.

Black e_t _!_   have found  a stoichiometric relationship between the  ferric
sulfate dosage required  for satisfactory  color removal and the raw water
color in a study  of colored waters  from different  regions of the  United
States.
                                  22

-------
           SECTION V,  EXPERIMENTAL MATERIALS AND METHODS
Research efforts should of necessity begin with controlled basic systems
and advance in complexity as information is obtained.  This study begins
with synthetic water, where system variables can be established so it
is possible to study a single variable at a time.  Many methods have
been employed to evaluate coagulation processes'" but the jar test has
been the most widely used and was the primary method chosen for this
study.  An improved version of the jar test apparatus was used.  The
jar test consists of a series of jars, containing the adjusted para-
meters under study, with mixing provided to simulate actual plant con-
ditions.  Normally a settling period follows mixing, where settling can
be evaluated.  Modifications have been made in order to increase the
information obtained.  The parameters measured during this study
included:

    1)  Coagulation pH

    2)  Forms of alkalinity and hardness

    3)  Settled color and turbidity

    4)  Electrophoretic mobility

    5)  Residual magnesium

    6)  Hardness, alkalinity and color of stabilized water

    7)  Visual observation of floe properties and settling rates.

Coagulant recovery was studied in detail for both synthetic and natural
waters.  A volume of water was coagulated to produce from one to two
liters of sludge which was then carbonated, monitoring calcium, magnesium
and organic color released.  The recovered magnesium was reused in order
to evaluate any change in coagulation effectiveness.  Some filtrability
studies of the carbonated sludge were performed.

Montmorillonite and Emathlite Clay Suspensions

The montmorillonite was montmorillonite #23 (Bentonite) obtained from
Ward's Natural Science Establishment, Inc., Rochester, New York.  The
emathlite was obtained from Mid-Florida Mining Company, Lowell, Florida.

Organic Color

Approximately 160 liters of highly colored water was collected from
runoff in the Austin Carey Forest, near Gainesville, Florida.   The water
was very low in turbidity and ionic strength.
                                  23

-------
Magnesium Carbonate

The magnesium carbonate was prepared from water softening plant sludge
at Dayton, Ohio.  The wet magnesium carbonate was shipped to Gainesville
where it was air dried.  The analysis for alkalinity and magnesium of
a solution containing  . 5g of the material allowed calculation of the
hypothetical formula, which was found to be the tri-hydrate of magnesium
carbonate, MgCO^'SH^O.  A chemical analysis performed by the Research
and Analytical Laboratory of the Monsanto Chemical Company, Dayton, Ohio,
of a similar batch of magnesium carbonate produced in April of 1970
indicated the following composition:

            Constituent                         Percent by Weight

            Magnesium Oxide, MgO                      29.44
            Calcium Oxide, CaO                         0.07
            Carbon Dioxide, CO,                       32.50
            Silicon Dioxide, S102                     <0.01
            Aluminum Oxide, A1203                      0.005
            Ferric Oxide, Fe203                       <0.01
            Sulfur Trioxide, 803                      <0,01
            Chloride, Cl"                             <0.001
            Total Insolubles                          <0.01
            Loss on Ignition                          70.67

Flocculant Aids

The flocculant aids studied included an anionic potato starch, Hamaco 196,
Alum, AP^O, and activated silica.

Synthetic Water Constituents

Reagent grade CaCl^, NaHCO^, and ILjSO, were used to prepare stock solution
for adjusting the calcium, alkalinity, and sulfate concentrations of the
synthetic water.

Preparation of Clay Suspensions

Both clays were pulverized by jar milling for twenty-four hours.  Approxi-
mately 20 grams of the pulverized clay and 10 grams of reagent grade
sodium chloride were added to four liters of water.  The slurry was
mixed for twenty-four hours and then dewatered using No. 40 Whatman filter
paper in a Buchner funnel.  The clay was washed with distilled water and
resuspended in four liters of distilled water.  After several hours of
mixing, the slurry was allowed to settle to remove the larger clay
particles.  The supernatant was withdrawn for use as a stock turbidity
solution.
                                                         70
This process was similar to that used by Black and Hannah ? to give a
more uniform suspension, allow greater precision of measurement, and
to promote exchange reactions due to the high zeta potential associated
                                    24

-------
with the sodium form of the clay.  These authors, using the same clays,
found the base exchange capacity of fullers earth to be 26.5 milli-
equivalents per liter and that of the montraorillonite to be 115 milli-
equivalents per liter.  The base exchange capacities of these clays
were determined by the ammonium acetate method used in soil analysis.

Preparation of Organic Color Concentrate

A highly concentrated, pure form of naturally occurring organic color
was needed.  Various methods have been employed to concentrate organic
color such as vacuum distillation, carbon adsorption, freezing, and ion
exchange.  Vacuum distillation was chosen because of its simplicity,
availability of a large vacuum still, and reportedly minor effects on
the chemical nature of organic color.

The water collected was first filtered through Whatman 41 ashless filter
paper.  The color was then concentrated using a Precision Scientific
Flash evaporator.  The operating vacuum was maintained by a vacuum pump
at 4-6 cm of mercury and a temperature of less than 40C.  The capacity
of the vacuum pump limited the evaporation rate to about 2.51/hr.

The evaporation procedure was semi-continuous.  The feed rate was adjusted
to match the evaporation rate.  When the color in the evaporator reached
the desired concentration, the evaporator was emptied and the procedure
repeated.

The color concentrate was filtered through a series of Whatman 40 + 41
paper then through Millipore  . 8y and . 45)j filters.  The concentrated
color was then placed in dialysis tubing and dialized against distilled
water for twenty-four hours.  Chemical analyses of the untreated, con-
centrated, and treated concentrate are shown in Table 1.  All chemical
analyses were run in accordance with the procedures outlined in Standard
Methods with metal ion determined by atomic absorption analysis.  The
two batches of raw water were collected to obtain the desired volume of
color concentrate.  The color concentrate was stored at 4C in tightly
stoppered liter bottles.

Preparation of Synthetic Waters

A synthetic water whose composition was designed to represent, as nearly
as possible, a typical soft surface water of low alkalinity and total
hardness was prepared from the stock solutions listed elsewhere.

The synthetic stock solutions were prepared using deionized distilled
water so that 1 ml = 50 mg of Ca(as CaCO.,) , alkalinity (as CaCO-j) and
SO/~.  Working solutions were prepared by diluting these stock solutions
10 to 1 with deionized distilled water so that 1 ml = 5 mg of the
desired constituents.

The synthetic water had the following composition:
                                 25

-------
                                         TABLE  1
                          CHEMICAL AN'ALYSIS OF ORGANIC  COLOR
pH
Conductivity (
Color (pH 8'. 3)
Acidity" (mg/1 CaCOo)
COD (mg/1)
TS (mg/1)
VSS (mg/1)
NH3N (mg/1)
Organic N (mg/1)
TOC (mg/1)
C/N Ratio
Cu
MN
Fe
Mg
Na
Ca
   0*8/1)
   (mg/1)
   (mg/1)
   (mg/1)
                           Raw #1
                                       Raw #2
4.70
48.5
700
20.1
117
0.190
0.103
0.41
3.14
39.5
12.6
< '.1
0.04
0.03
8.2
1.3
5.1
1 .9
4.45
57.5
690
25.0
108
0.182
0.983




< . 1
0.04
0.04
6/4
1.0
4.3
1.2
                                                   Concentrate
                                                        4.00
                                                      520
                                                      ,150
                                                      320
                                                      ,660
                                                        2,
                                                          76
                                                         2.08
                 Filtered and
             Dlalized Concentrate

                       4.40
                     230
                  13,250
                     200
                   2,010
                       2.00
                       1.66
  1.20
  0.40
  0.70
106
 18
  8.2
  8.0
                                                                              1.20
                                                                              0.20
                                                                              0.45
                                                                             98
                                                                            1 0
                                                                              2.3
                                                                              5.0
Color
COD
Acidity
Conductivity
Volume
                Concentration Factor
                        21.
                        22
                           8
                           8
                        15.9
                        10.3
                        26.7

-------
Milliequiyalents            ppm        Mllliequivalents          ppm
                            10.0       HCOo --^0.500          30.5
                            11,5       SOjZ -^ ---- 0,500          24.0
                            19.5       Cl~  - --- 0.500          17.7
                                                  1,500

Total alkalinity as CaCO-j -r --- <- r- ---- . 25 ppm
Total hardness as CaCOo    <--  .^-w --- 25 ppm
Total dissolved solids    r^_^,r,r,^_ ,_, -, ^^

Alum

The alum stock solution was prepared by adding reagent grade aluminum
sulfate to deionized distilled water so that 1 ml was equal to 10 mg of
aluminum sulfate,  This solution was stored at 4C and used daily to
prepare working solutions by dilution with distilled water so that
1 ml = 1 mg of aluminum sulfate.

MgCQ3;3H20

For dosages of less than 15 mg/1 of MgCOo'SI^O, it was added as a solution
containing exactly 0.5000 g of the material in 1 liter of demineralized
water.  Fresh solutions were prepared weekly.

When dosages greater than 15 mg/1 were to be used, they were accurately
weighed into 15 ml beakers and quantitatively transferred to the stirred
water sample as a slurry.  In the case of synthetic waters, deionized
water was used to prepare the slurry; for natural waters, the water
itself was used.

Flocculants

Alum:

A fresh solution containing 0.1 mg/ml of Al (SO, ) ,. 'ISl^O was prepared
daily from a stock solution stored at 4C.

Starch:

The starch solution was prepared daily by slowly sifting 1.0 gram of
starch into 1 liter of deionized distilled water and rapidly mixed with
a magnetic stirrer.  A working solution was prepared by diluting 10 to 1
with deionized distilled water so that 1 ml = 0.1 mg of starch.

Activated Silica:

The activated silica solution was prepared and  activated in the following
manner :

    1)  10 ml of distilled water was added to a 100 ml graduated cylinder.
                                 27

-------
    2)  5 ml of 348 gram/1 solution of "N" brand sodium silicate added.

    3)  5 ml of 1 N NH.C1 was added with constant stirring to the
        graduated cylinder,

    4)  After 5 minutes1 aging, the solution was made up to 100 ml
        volume with distilled water and mixed thoroughly.

Determination of Turbidity and Color

A Lumetron Model 450 Filter Photometer was used for both color and
turbidity determinations.  Sufficient accuracy was obtained using the
red 650 my filter and 75 mm cell light path for turbidity determinations.
A calibration curve was prepared plotting optical density obtained for
a sample against the turbidity values previously obtained for the sample
using a Jackson Candle Turbidimeter as described in Standard Methods.
The turbidity solutions used for this calibration procedure were prepared
using the emathalite clay stock solution.

Organic color was measured using a 560 my filter and the 75 mm cell light
path.  The calibration curve for color was obtained by plotting optical
density as a function of various dilutions of Platinum-Cobalt Color
Standard.

When color and turbidity were both present in a sample, a procedure
outlined in the Lumetron Operating Manual"^ was followed.  This procedure
takes advantage of the fact that organic color absorbs light more strongly
at shorter wave lengths.  Color would therefore have little interference
in the measurement of turbidity.  The following procedure was used to
determine color in the presence of turbidity:

    1)  For various levels of turbidity the optical densities were
        measured at 650 my.  An average value was determined for the
        ratios between the optical density at 560 my and 650 my.  This
        average value represents a constant for any level of turbidity
        and will be denoted as R.

    2)  The optical density of the sample was determined at 560 my and
        650 my.

    3)  R multiplied by  the optical density at 650 mp represents the
        interference due to turbidity.  Subtracting this product from
        the optical density found at 560 my gives a value which can be
        used to determine the color from the calibration curve pre-
        viously prepared.

Atomic Absorption

Iron, sodium, magnesium, and calcium were determined by use of a Model
1301 Beckman Atomic Absorption Unit in combination with a Beckman DBG
Grating Spectrophotometer and Beckman Potentiometric Recorder with scale
                                    28

-------
expander,  The procedures outlined In Methods for Analyses of Selected
Metals in Watery by Atomic Absorption"^ were' followed.  The standards
were prepared as described with, the exception of iron, for which
reagent grade ferrous ammonium sulfate was used to prepare the standard
solution,  A calibration curve was obtained each time the samples were
run, plotting absorbance versus concentration.

Electrophoretic Mobility

The electrophoretic mobility determinations were made using a Zeta-Meter.
The samples were analyzed immediately after collection using the 8 power
microscope objective and a two-hundred volt potential.  The procedure
outlined in the Zeta-Meter Manual   was followed for all determinations.
Normally 10 particles were tracked for each sample.

Stabilization of Treated Waters

In order to make possible a comparison of both physical and chemical
parameters of waters coagulated with magnesium and with alum, all samples
were stabilized to pH 9.0.  Waters coagulated with MgCOo were stabilized
with CC>2.  Those coagulated with alum were stabilized with freshly
filtered, saturated lime water.

Stabilization of Water Coagulated with Alum:

Approximately 300 ml of settled water, filtered if necessary, was trans-
ferred to a 500 ml beaker, placed on a magnetic stirrer and titrated with
clear saturated lime water to pH 9.0, measured by a glass electrode.  The
sample was then filtered through Whatman No. 40 paper and color, total
alkalinity, total hardness, calcium and magnesium determined.

Stabilization of Water Coagulated with Magnesium Carbonate:

Approximately 300 ml of settled water was transferred to a 500 ml beaker,
placed on a magnetic stirrer, approximately 1 g of reagent grade powdered
CaCOo added and the suspension carbonated by blowing through a pipette
with rapid mixing.  Three to five minutes were required to reduce the
pH to 9.0, measured by a glass electrode, as above.  The suspension
was then filtered through Whatman No. 40 paper and color, total alkalinity,
total hardness, calcium and magnesium determined.

Alkalinity

The alkalinity titrations were performed using 0.02 N ^SO, with phenol-
phthalein and methyl purple indicators as described in Standard Methods.
The sulfuric acid was standardized using standard 0.02 N sodium carbonate,
also as described in Standard Methods.

Hardness

Total and calcium hardness were determined by titrating with carefully
                                   29

-------
standardized EDTA? following exactly the procedures as described in
Standard Methods,  Many determinations were checked by atomic absorption,

pH. Measurement

All pR measurements were made using a Corning Model 7 pH. Meter with a
Corning Combination glass and Ag/AgCl, electrode.  The pH. meter was
calibrated daily using solutions prepared from concentrated standard
buffer solutions purchased from W, H. Curtin and Company,

Jar Test Procedures

The jar test procedures were very similar for the natural and synthetic
waters and the discussion will be applicable to both series.  The
procedure outlined will follow chronological order with the differences
between the series discussed in the order in which they occur.  In
every instance where magnesium is used, it is added in the tri~hydrate
form, MgCOySt^O.  In this text and tables it has been referred to as
magnesium carbonate or
Preliminary Determinations :

Chemical and physical analyses were performed on each natural water prior
to jar testing.  These tests included pH, color, turbidity, alkalinity,
hardness, and magnesium.  A sample of each natural water was filtered
through No. 40  Whatman filter, acidified to approximately pH 2 with
concentrated HC1, and stored in a glass bottle for analysis by atomic
absorption for magnesium and iron.

The coagulant dosages were chosen to give undertreatment of the water at
the lower dosages and overtreatment at the higher dosages.  Based on
previous experience, this range in chemical dosages could usually be
determined from the results of the chemical and physical analyses.  For
synthetic waters, these parameters were, of course, chosen for each jar
test.

For both, the quantity of water to total 1 liter after the addition of
all dosages was calculated and added to each jar.

Details of Jar Test Procedure:

For studies where the magnesium carbonate was added as a slurry, at least
two minutes of mixing at 100 RPM was provided after the slurry addition.
The lime slurries were then added to the jars.  The initial addition
required approximately one minute with two additional minutes needed
to rinse the six beakers in order to complete the quantitative transfer
of the lime slurry.  When flocculant aids were used, they were added
approximately three minutes after the lime addition.  Incremental addition
of starch was evaluated using six dosages, one minute apart.  Three were
added during rapid mix and three during the flocculation period,
                                  30

-------
Samples for electrophoretic mobility determinations were taken during
rapid mixing, approximately one minute after flocculant aid addition or
approximately three minutes after lime addition when no flocculant aid
was used.  The mixing speed was then slowed to 10 to 12 RPM and maintained
for 15 minutes.  After visual observations of the floe characteristics,
100 ml samples were collected and filtered through Whatman No, 40 paper
for immediate determinations of alkalinity.

After the flocculation periods, mixing was stopped and the jars allowed
to settle for twenty minutes.  At that time samples of the supernatant
were taken for color and turbidity analysis,  pH determinations were
then made on all jars and the water in selected jars was stabilized,
filtered and analyzed.

The studies using alum were performed in a similar manner.  For several
very low alkalinity waters pre-lime was added first, using a saturated
calcium hydroxide solution to increase the total alkalinity of the water.
Electrophoretic mobilities were not determined on these waters.

Recovery Studies

The recovery of magnesium from the sludges produced in coagulating both
synthetic and natural waters was evaluated.  The synthetic waters were
prepared to give a range in organic color of from 15 to 200 and a mont-
morillonite turbidity range of from 15 to 50.  Coagulation was carried
out in a forty-liter Pyrex jar using a small Lightning mixer with a
rheostat to control the mixing.  The quantity of water to total 36 liters
after the addition of all dosages was added to the jar.  The salt solutions
were then added using the concentrated stock solutions to reduce the
volumes added.

As before, coagulant dosages were estimated from previous jar tests.  The
magnesium carbonate was vigorously slurried and quantitatively added to
the rapidly mixed water.  After approximately three minutes, the lime
slurry was added.  Five minutes after the lime addition, the alum was
added and the rapid mixing continued for two additional minutes.  The
mixing was then slowed and maintained for twenty minutes at a speed which
would deep the floe in suspension.

The floe was allowed to settle for a period between several hours and
overnight in some cases.  The clear supernatant was carefully syphoned
from the sludge layer and a composite sample collected for analyses.

The sludge was then poured into a 2 liter graduated cylinder and measured.
In all but the first two experiments the volume was then made up to 2
liters with distilled water before carbonation.

Sludge Carbonation

Carbonation of the sludge was performed using a cylinder of specially
prepared gas containing 15% C02 and 85% air.  A 2 liter graduated cylinder
                               31

-------
was placed on a large magnetic stirrer for continuous mixing during
carbonation,  The flow of CC^ was regulated using a gas pressure regulator,
so that fine, well dispersed bubbles were produced,  A carborundum stone
diffuser was used to disperse the CCU into the sludge with no attempt
made to measure the gas flow rate,

Fif tyr-milliliter samples were taken at predetermined time intervals and
filtered through Whatman No, 40 filter paper.  pH determinations were
made on the filtrate; 10 ml samples were titrated for alkalinity, and
a dilution of the remainder prepared for color analysis.  After organic
color had been determined, the samples were acidified and stored for
future analysis for magnesium and calcium by atomic absorption.  The
sludge was carbonated in most cases until a pH in the range of 7.5 to
7,0 was reached.

The remaining sludge was filtered through No. 40 Whatman paper using a
vacuum flask and Buchner funnel.  Several f ilterability studies were
made using polymers and calcium carbonate as filter aids.  Two general
methods of evaluation were employed - determination of the time to
dewater 100 ml of the sludge and determination of the total volume that
could be filtered before clogging occurred.
Separan AP   was used in the polymer evaluation.  Incremental addition
of 1 mg/1 of the polymer, followed by determination of the time for
filtration of 100 ml of the sludge provided data used to determine the
effect of the polymer on sludge f ilterability.

The filtrates from several of the recovery studies were stored to be used
as recycled coagulant .

Coagulation Using Recovered Magnesium Bicarbonate

Coagulation, using both the standard jar test apparatus and the 40 liter
Pyrex jar with the variable speed Lightning mixer, was evaluated using
recovered magnesium.  The required volume of recovered magnesium bi-
carbonate to give the desired coagulant dosage was added and the coagu-
lation tests performed as discussed previously.

The coagulation of selected synthetic and natural waters was repeated,
using the solutions of magnesium bicarbonate recovered as described
above, and results identical with those obtained with the original
magnesium carbonate were obtained.  This was done with jar tests and
with "recovered coagulant in the 40 liter vessel.
                                32

-------
                   SECTION VI.  RESULTS AND DISCUSSION
Coagulant Studies of Synthetic Waters

Emathlite clay turbidity was used in the first studies of synthetic
waters, since clays of this type have been employed by others for
coagulation studies.  An experiment was designed to determine the rela~
tionship between the level of turbidity and/or organic color present
and the dosage of magnesium carbonate required for satisfactory treatment
of the water.  The alkalinity and hardness were held constant at 25 mg/1
as CaCO-.  An acceptable treatment would give a settled turbidity less
than 3.5 mg/1 and color less than 15 mg/1.  A minimum of six jars were
required to determine the lowest dosage of magnesium carbonate for each
combination of color and turbidity.  Table 2 summarizes the data used
in the development of this relationship for emathlite turbidity.

                                 TABLE 2

             MAGNESIUM CARBONATE REQUIRED FOR COAGULATION OF
                       ORGANIC COLOR AND EMATHLITE
                                TURBIDITY

                                  Color           Turbidity
	 J
4
20
90
20
90
90
10
50
15
50
200
50
200
200
50
100
60
60
60
20
20
100
100
100
A stepwise, linear regression equation was calculated using a BMD02R
Computer Library program.85  xhe general form of the regression equation
was:

                  Y = A + b^ + b2X2

            where:

            Y = magnesium carbonate dose
            A = constant
           X]_ = the variable, either color or turbidity, which is most
                  significant in reducing the total sums of squares
           X2 = the variable remaining
           b-^ = regression coefficient for X,
           b2 = regression coefficient for X2
                                33

-------
The equation resulting is;

            Y ~ -3,95 + .48 color + .02 turbidity

For the data shown, color and turbidity account for 98.67% of the
variations in the required magnesium dosage with a highly significant
F value of 440 and standard error of the estimate of 5.22.

Lime along with a flocculant, starch, was found to satisfactorily
flocculate the emathlite turbidity as shown in Table 3.  Possibly the
fine particles of turbidity served as a nucleii for calcium carbonate
precipitation which was in turn agglomerated by the starch to a size
which would settle.

Magnesium carbonate*s effectiveness in color removal is demonstrated in
Table 4.  The color present seemed to improve the size of the floe as
well as its settleability.  No attempt was made to measure the magnesiun
in solution after coagulation in these early experiments.  It was found
however, that good floe formation took place at a pH above 11.0; there-
fore the pH of coagulation was maintained from 11.0 to 11.25.

The experiments with montmorillonite clay turbidity were designed in a
similar manner to determine the effect of color and turbidity on the
coagulant dosage.  The data used to develop this relationship are shown
in Table 5 below:

                                 TABLE 5

              DOSAGE OF MgCO., REQUIRED TO COAGULATE ORGANIC
                   COLOR AND MONTMORILLONITE TURBIDITY
MgCO,,
,3
85
90
80
15
15
15
40
Color
200
200
200
50
50
50
100
Turbidity
20
100
60
20
100
60
100
The equation determined is:

               Y = -10.24 + .47 color + .03 turbidity

Color and turbidity account for 99,41% of the variation in required
magnesium dosage with a highly significant F value of 710 and standard
error of the estimate of 3.24.  As with the emathlite turbidity, lime,
aided by a flocculant, was satisfactory in removing montmorillonite clay
turbidity.  In Table 6, the effectiveness of MgCOo in color and
                               34

-------
                                                      TABLE 3
                       COAGULATION OF FULLER'S EARTH TURBIDITY  WITH MgC03 AND WITH LIME
Jar
No.
1
2
3
4
5
6
Dosage in ppip
CO
w> o
r
0
?.
4
6
1C
15
CM
s-^.
n3 ZZ
U 0
95
97
100
102
105
110
v>
n) o\
53 -H
0.4
0.4
0.4
0.4
0.4
"
0.4
PH
10.80
10.90
10.95
10.95
10.95
10.95
u
o
fM
C
O






x
JO
M
Turbid
3.3
2.1
2.0
1.6
1.5
1.0
Mobility

-.95
~.A2
-.54
-.36
-.40
Alkalinity
w
o
46
50
58
52
58
62
<*>
O
o
56
48
44
44
44
40
r<1
O
C_>
32
0
0
0
0
.0
0
o
j->
,0 rn
^ ex
4.J
C/3






M
0
rH
O
CJ






Alkalinity
O
0
U






ro
O
O






Hardness
C






NC






T






Magnesium
as CaC03






_ 	













00
Oi
Characteristics of raw water
Alkalinity as CaC03
Total Hardness as CaC0
                                                                     Comments
                                      25
        Organic Color ..
        Turbidity
        Type Clay
                                       8.30
                             100
                         Eraathli t-Q Cla

-------
                                               TABLE 4
                      COAGULATION OF A SYNTHETIC  WATER  CONTAINING BOTH ORGANIC COLOR
                                  AND FULLER'S  EARTH  TURBIDITY WITH MgCO-j
Jar
No.
1
2
3
4
i
5
6
Dosage in ppm
00 O
2 0
75
100
120
140
160
180
o o
130
144
155
166
177
188
u; -<
0.4
0.4
0.4
0.4
0.4
0.4
-
11.10
11.10
11.05
11.05
11.10
11.10
0
i-H
O
o
>50
>25
14
10
12
16
Turbidity
10.0
2.0
1.5
1.2
1.9
3.8
Mobility






Alkalinity
a
0
60
69
59
62
50
43
0
c_>
118
78
52
40
44
48
CO
o
O
0
0
0
0
JO
0
0
AJ
C3 "H.
c/>



9.0


o
o
o



9


Alkalinity
o



0


0")
0
o:



52


Hardness
C



52


NC



1


T



53


Magnesium
as CaCO-j




















o\
       Characteristics of raw water
       Alkalinity as CaC03 	25
       Total Hardness as CaC03 	25
       pH		8.30
       Organic Color .... 	200
       Turbidity	 	20
Comments
       Type Clay	   Fullers Earth

-------
                                              TABLE 6
                 COAGULATION OF A HIGHLY COLORED SYNTHETIC  WATER WITH MgCO,
Jar
No.
1
2
3
4
5
6
Dosage in ppiv
co
60 O
S 0
40
45
50
60
70
80
CN
<0 &
0 0
V  '
117
120
122
128
134
140
3
i 1
<
.5
.5
.5
.5
.5
.5
PH
11.05
11.05
11.05
11.05
11.05
11.15
p
o
.H
0
o
56
46
45
34
31
23
Turbidity
3.7
6.0
12. C
12.8
2.
M
Mobility
-.81
-.88
-.91
-.80
-.91
-.80
Alkalinity
o
68
60
68
54
60
64
ro
O
O
96
96
80
72
72
72
CO
o
o
ac
0
0
0
0
^0
0
o
j_i
_/~ *^r*I
 G.
JJ
CO





9.0
VJ
o
H
O
u





14
Alkalinity
f>
o
0





4
o
o
o
sc





46
Hardness
C





41
NC





0
T





41
Magnesium
as CaC03





9














U)
         Characteristics of raw water
         Alkalinity as CaCO-j 	25
         Total  Hardness as CaC03 	25
         pH		8.30
         Organic Color .... 	200
         Turbidity	    .	60
Comments
         Type Clay	Montmorillonite

-------
 montmorillonite clay turbidity removal  is  shown.

 The use of starch as a flocculant  for lime coagulation was  studied  for
 both clays.   Fifteen experiments were conducted where all system
 parameters but starch were kept constant.   Starch dosages of  ,2  to  1,6
 mg/1 were found to reduce the final  turbidity in 6 experiments,  to  have
 no effect in 6, and to increase the  final  turbidity in 3.   Three experi-
 ments were performed with all parameters but the method of  starch
 addition kept constant.   The starch  was added as  a single  dose  to  one
 jar and in six increments to the second jar as discussed in a previous
 chapter.   The incremental addition increased the efficiency in one  test,
 had no effect in another, and decreased the efficiency in the third.
 Drew Floe 21,  a cationic  starch, was used  unsuccessfully in one  experiment.

 Study of  Natural Waters

 The first natural waters  studied were obtained from the Talapoosa River,
 a  source  of  water for Montgomery,  Alabama,  Seven sets of jar tests were
 performed on this water with the results from selected jar  tests  shown
 in Table  8.   Color and turbidity removal was comparable to  alum  treatment.
 The hardness and alkalinity  of  the magnesium carbonate treated water,
 44 mg/1 as CaCO^,  would allow pH adjustment for corrosion control.  This
 would not be the case for the alum treatment as the alkalinity and
 hardness  were  13 and 22 mg/1 respectively.  This has led to serious
 corrosion problems because of the  lower buffer capacity.

 Water was also obtained from the Mobile River near Mobile, Alabama and
 evaluated in a similar manner with the selected results for both  treat-
 ment methods shown in Table  9.  It was found that very small dosages of
 alum were very effective  as  a flocculant aid.  In seven sets of jar
 tests with alum addition  the only  variable, a 0.5 mg/1 dosage of  alum,
 gave an average of 17% settled  color reduction and 50% turbidity
 reduction as shown in the Table 7.

                                 TABLE 7

             EFFECT  OF ALUM AS A FLOCCULANT AID IN COLOR AND
             TURBIDITY COAGULATION WITH MAGNESIUM CARBONATE

                                                       Settled    Settled
Montgomery           MgC03         H        Alum       Color    Turbidity

                      50         11.30       0          10.8        1,0
                      50         11,30       0.5         8.7        0.4
                      50         11.30       0.75        9.0        0.5
                      45         11.25       0          13          1.6
                      45         11.25       0.5        12.5        1.3
                      50         11.00       0          15          1.0
                      50         11.00       0.5        12          0.2
                                 38

-------
                                                TABLE
                       MgC03 AND ALUM  COAGULATION OF MONTGOMERY,  ALABAMA WATER
Jar
No.
1
2
Dosage in ppm
60 O
S o
30
35
3 | 45
4
5
6
50
50

CM
C_J O
125
130
125
115
125

s
3
.5
.5
.5
.5
.5
2)
PH
11 . 10
11.20
11.20
11.15
11.20
6.3
o
rH
O
15
10
12
9
7
6
Turbidity
3.2
1.6
2.6
1.6
1.4
0.6
jj
r->
,JD
S
-.83
-.69
I-.44
-.40
-.35
f.19
Alkalinity
0
78
76
88
80
86
0
o
52
56
56
60
60
0
0
o
cc
0
0
0
0
o
6
o
ji X
TO ft,
to



9.0
9.0
9.0
0
o
U



6
5
6
Alkalinity
ro
O
O



4
3
4
o
y



40
41
9
Hardness
C



44
44
13
NC



0
0
9
T



44
44
22
Magnesium
as CaC03













1
;






OJ
           Characteristics  of raw water
           Alkalinity as CaCO-j 	16
           Total Hardness as CaC03 	13
           pH		7.00
           Organic Color .... 	50
           Turbidity		165
           Type Clay		Natural
Comments

-------
                                     TABLE 9
        MgC03  AND ALUM COAGULATION OF  MOBILE RIVER WATER,  MOBILE,  ALABAMA


Jar
No.
1
2
3
i
4
Dosage in ppn1
r- es
COO
%z o
30
35
45
45
i
5
6
40
0
.-^
"3 33
O O
130
130
115
125

115
0

5
i }
.5
'5
.5
.5

.5
23



PH
11.15
11.15
11.20
11.25

11.10
7.2

i-i
o
r-l
O
0
10
10
13
12

10
6
>.
u
l-f
T)
M
^>
M
3
H
2.0
1.6
2.5
2.6

2.6
X
u
H
r |
H
,0
,O
^
-.78
-.69
-.45
-.44


2.0 [-.55
Alkalinity

P"!
O
86
76
84
88

70
0

f"i
0
o
48
56
56
56

50
0

ro
0
y
0
0
0
0

.0
33
o
o p*"!
a o-
to





9.0
9.0
M
o
c-l
0
u





9 
8
Alkalinity

rO
o
u





8
8

r**>
o
o





44
30
Hardness


C





52
38


NC





1
14


T





53
52
s
3 0
-< O

-------

Mobile MgCO-}
40
40
40
40
40
40
40
40
40
40
40

pH.
11,00
11.00
11.00
11,00
11,00
10.95
10.95
11.05
11,05
11.15
11,15

Alum
0
0,25
0.50
1.00
1.5
0
0.5
0
0.5
0
0,5
Settled
Color
14.5
13,0
13,2
10.0
11.2
18
16
16.5
12.5
13.0
10.0
Settled
Turbidity
0.6
0,4
0.3
0
0
1.4
0,6
1.5
0.6
1.0
0.6
Alum addition increased the size and rate of floe growth and made the
electrophoretic mobility less negative, thereby increasing the Van der
Waals attractive forces.  A significant linear relationship between
mobility and the color or turbidity reduction has been found for most
of the waters as will be discussed later.

In the study of waters from cities throughout the country, it was found
that many of the waters contained a considerable amount of magnesium.
For these waters, lime addition precipitated the magnesium present, re-
quiring no additional magnesium carbonate.  In Tables 22 through 37,
selected data from these studies are presented.  In every case magnesium
carbonate gives color and turbidity reductions comparable to alum treat-
ment.  The floe formation with magnesium carbonate occurs at a faster
rate, the floe formed is larger in size, and settling is more rapid due
to the greater floe density.  For the waters of high alkalinity and
hardness, activated silica was found to be the most effective flocculant
aid.

Magnesium present in the natural waters was in most cases in the non-
carbonate form.  Removal of magnesium as Mg(OH)2 by lime addition will
not decrease the total hardness of the stabilized water, merely substitute
calcium hardness for magnesium hardness.  This can be advantageous in
the case of high magnesium waters where the formation of magnesium
silicate scales in hot water heaters is a problem.

The raw water analyses, together with the chemical characteristics of
the waters following both alum and magnesium carbonate treatment are
given for the waters studied in Table 10.  For the 17 waters studied,
treatment with magnesium carbonate gave a stabilized water with alkalini-
ties ranging from 29 to 55 mg/1 as compared with those resulting from
alum treatment, which ranged from 9 to 98 mg/1.  Values for stabilized
hardness ranged from 33 to 83 mg/1 as compared with 15 to 136 mg/1 for
alum treatment.  Six of the waters coagulated with alum are too low in
hardness and alkalinity to use pH adjustment effectively for corrosion
control.  In addition, eight waters would benefit by the reduction in
total hardness resulting from using magnesium carbonate rather than alum.
                                  41

-------
                                            TABLE  1Q

      COMPARISON OF THE CHEMICAL CHARACTERISTICS  OF 17 RAW AND TREATED NATURAL WATERS

| CITY
1
|
Atlanta, Sa.
^Baltimore, Md.
SBirrainghaiE, Ala.
1 (a)
^Cleveland, Ohio
betroit, Mich.
'Huntsvllle, Ala^b)
L'ackson, Miss.
jLanett, Ala. (c)
SLouisville, Ky. .
tMontgoittery, Ala.
ijNashvilic, Term.
Opelika, Ala.
^Philadelphia, Pa.
jXichr.nd, Va.
UTuscaloosa, Ala.
Iwa.ihinr.toiu. U. C.
RAW V/ATER CHARACTERISTICS


Turbidity
104
2
10
15
6
3
13
8
105
106
165
8
14
41
24
1 Total
/"* 1 i
Color iMkaiini}
!
38
4
12
24
5
0
4
27
30
11
50
8
10
14
30
4 26
50 J 15
11
12
74
48
92
80
54
10
17
51
13
71
17
34
27
it
41
Total
;!arr!r>ess
13
40
83
71
127
100'
84
12
17
110
16
86
17
69
43
5
71
Xagncs-
IUTTI as
4
13
25
15
34
10
13
2
6
33
3
16
4
24
7
1
17
MgC03 TREATMENT ALUM TREATKENT j

Total
Mkalinity
38
29'
40
33
36
37
30
39
55
32
44
32
33
34
38
40
40
Total [ Total
Hardness JAlknlinity
3b
53
47
56
68
57
54
39
55
83
44
50
33
66
53
42
68
16
24
71
51
93
86
52
16
18
59
13
73
21
56
29
9
40
Total j
Hardness 1
18 *
54 !
87 i
82
136
112
91
23
23
121
22 5
97 jj
30 f
95 i
55 5
15 I
76 |
(a)  Requested  to be deleted from publication
(b)  Tennessee  "iver used  for source  of rav vater
(c)  Chatt.ihoochee- River used for source of rav water

-------
Solubility of Magnesium Hydroxide

The solubility of magnesium hydroxide has been determined by many
investigators,  In this study it was increasingly evident that the
magnesium remaining in solution after coagulation was many times more
than would be predicted by theory.  There are several reasons for this
apparent increase in solubility.  In the jar tests the time allowed for
equilibrium was usually only one hour.  In plant use, four to six hpurs
are normally allowed for precipitation which should decrease the
magnesium solubility.  It is the magnesium hydroxide which is precipitated
that causes colloidal destabilization and only this portion of the
magnesium dosage can be recovered and reused.

The solubility of magnesium hydroxide under jar test conditions varied
for each of the natural waters studied.  It would have been desirable
to determine this solubility relationship for each water, but because
of a lack of sufficient data, a composite of 70 observations of magnesium
concentrations at varying pH values for all of the waters studied was
used,  A simple linear regression analysis between log magnesium and pH
was used to determine the experimental solubility product, K  .  This
value was found to be 1.66 x lO"-^ with a standard error of the estimate
equal to 1.27,

The solubility relationship found is shown in Figure 6, where the mag-
nesium is plotted as magnesium carbonate tri-hydrate.  This composite
solubility relationship is used in determining recovery efficiencies
and cost estimates for the natural waters studied.

Determination of Conditions for Lowest Treatment Cost

Each water requires a specific amount of magnesium hydroxide for satis-
factory treatment.  The solubility relationship developed allows calcu-
lation of the amount of magnesium carbonate which must be fed to pre-
cipitate this amount at various coagulation pH values.  For the economic
evaluations, three chemical costs will be considered:

    1)  Dosage of 90% quicklime required to provide the optimum pH

    2)  Amount of CCL required to:
        a)  solubilize the Mg(OH)2 in the sludge and
        b)  reduce the high pH of the treated water to the pH of
            stabilization

    3)  Amount of "make-up" MgC03'3H20 to be added.

In addition three alternative conditions are considered:

    Case I.  Lime recovery is practiced, providing CC^ at no cost and
        90% lime at $.006/lb.
                               43

-------
    100-
     10-
      I
      tO 20
Tern PL  29* C
	EMptrlmcnlol K.,-1.66 XIO'10
	Th*or*tleal K^  2.9 X KT"
    Exprlntol Oat*rntl*atlon
                     1060
                                    11.00
                                PH
                                                  11.40
FIG.  6 SOLUBILITY  OF  Mg(OH)2  (AS  MgC05  3H20)  AS  A
       FUNCTION  OF  pH FOR 23 NATURAL  WATERS

-------
    Case II.  Lime is purchased at $,01/lb but CC>2 is available at no
        cost from a source within or near the water plant.

    Case III,  Lime is purchased at $,01/lb and CC^ generated at a cost
        of $.01/lb.

A cost of $.05/lb was assumed for the MgCO^'SHjO,

A series of curves have been developed that will allow a graphical
determination of the pH for coagulation at the least cost.  The lime
required to increase the pH from 10.5 to some desired value is inde-
pendent of the total carbonate present in the water.  In Figure 7, the
cost to raise the pH from 10.5 to the desired pH is shown for the three
cases considered.  Figure 7 also includes the cost of C02 to stabilize
the water back to pH 10,5.  The MgCOo'Sl^O cost curve was developed from
the solubility relationship curve.  It was assumed that MgCO^-Sl^O left
in solution represented a cost, i.e. make-up coagulant.  Summing the
magnesium carbonate cost and the lime and C02 costs for the three cases,
a total cost curve is shown in Figure 8.  The optimum pH values and
costs can then be determined as:

    Case I        pH    =    11,35
                  Cost  =    $9.57

    Case II       pH    =    11.15
                  Cost  =    $10,85

    Case III      pH    =    11.00
                  Cost  =    $12.45

To this cost, the cost to raise the pH to 10.5 and stabilize the water
back to approximately 9.0 must be added.  These costs are related to
the total alkalinity of the water and are shown in Figure 9 assuming
the C02 in the water is negligible.  Summing the costs in Figures 8
and 9 produces the working curve shown in Figure 10.  An additional cost
must be added for the lime required to precipitate magnesium hydroxide
and for CC^ used in the recovery process.  The costs are shown in Figure
11 and are a function of the amount of MgCOo'SI^O precipitated.  Figures
10 and 11 therefore, represent the two basic working curves for calcu-
lating treatment chemical costs.

The total cost found must be reduced by an amount dependent upon the
magnesium present in the raw water, as this will reduce the MgCOo'SI^O
make-up.  If the coagulation pH results in greater than 100% recovery
of MgCO,'3H20 this value will exceed the cost allocated to the MgCOn'SI^O
remaining in solution as shown in Figure 7.   Therefore, before treatment
cost estimates are made for a water, using Figure 6, the pH of saturation
is determined.  When the optimum pH values,  determined in Figure 8, are
higher than the pH for 100% recovery, the treatment costs are calculated
using Figure 8, 9 and 11.  The application of these curves will be illus-
trated using three waters which represent all conditions which might be
encountered.
                               45

-------
  M.OO'
  IZ.OO-
 IOOO-
  aoo-
19
2
  eoo-
  4.00-
  2.00-
    IO.S
           - S0.05/*
o-CoO, COi - |O.OI/*
o-CoO - 80 Ol/*
-CoO-
                  IO.T
                                                                           It.S
         FIG. 7  PARTIAL TREATMENT  COSTS IN $/M.G.  FOR  CoO, C02
                 AND  MgCOs  AS A  FUNCTION  OF  COAGULATION  pH

-------
MOO
aoo
12.00
IIJOO
MuOO
900
tJOO-
                                              mjplmu* coet - g 12.45
                                              optimum pM  - ILOO
                                                      minimum cot - JIO.85
                                                      optimum pH - tt.19
 - C0 * COt  -Ol/*
 - C0-.CM/*
o-C0-.OO/*
                                                                   mlnlmuM  cett - |9.97
                                                                          H - 11.39
   KX90
                    10.70
                                    10.90             11.10

                                             pH
                                                     11.30
                                                                                      11.SO
     FIG.  8 TREATMENT  COST  IN   /M.G.  AS  A  FUNCTION  OF  COAGULATION  pH

-------
                                 12.00-
                                 KJ.OO-
                                 aoo-
-p-
oo
                               o
                               o
                                 4.00-
                                 200-
                                       20     4O     60     80     100     120     140
                                                            ALKALINITY (Mg/1 o*  CoCO,)
                                                                                        160
                                                                                               ISO
                                                                                                      ZOO
                                     FIG.  9  TREATMENT  COST IN  $/M. G. FOR  CoO  AND  C02 TO
                                             RAISE  THE RAW  WATER  pH  TO  10.5  AND  REDUCE
                                             THE pH BACK  TO  pH  9.0 FOR  STABILIZATION

-------
vo
                          20.00-
                           iaoo
                           t&OO
                        O
                        o
                           14.00
                           12.00-
                           IOOO
                                     2O
                                             30
4O     SO     60      70

    ALKALINITY (mfl/l at CoCOj)
80     90
                                                                                               100
                                  FIG. 10  TREATMENT  COST  IN  I/M.G.  AS  A  FUNCTION


                                          OF THE  RAW  WATER  TOTAL  ALKALINITY

-------
                                5OO-
                                ZJOO-
ui
o
                              o
                              Z
                                ix>o-
                                                 10            20

                                                         MgCOj NEEDED (wifl/l)
90
              4O
                                      FIG. 11 TREATMENT  COST IN  $/M.G. AS A  FUNCTION

                                              OF  THE  AMOUNT  OF  MgC05  PRECIPITATED

-------
   Water A ^ Atlanta? Georgia

       Total alkalinity = 11 mg/1
       Magnesium  (as MgCO-j'SH^O) =5,5 mg/1
       Precipitated MgC033H20 required (from jar test data) ^ 32 mg/1

This water had a  small amount of magnesium present in the raw water.
Figure 6 was checked to see that the magnesium recovery was not greater
than 100% for the three cases of treatment.  The pH value which would
represent 100% recovery was greater than 11,6 so the two basic working
curves can be used for all cases for this water,  From Figures 10 and
11 the costs were calculated as:

                                                Case I Case II Case III
Figure 10
   (for a total alkalinity of 11 mg/1)          $10.10 $11.60  $15.38

Figure 11
   (for 32 mg/1 MgCOj                          $ 1.02 $1.60  $2.60
                                                $11.12 $13,20  $18.00

Credit for magnesium in raw water at
   $.05/lb for MgC03-3H20                       $-2.29 $-2.29  $-2.29
                                                $ 8.73 $10.91  $15.69

   Water B - Baltimore, Maryland

       Total alkalinity = 12 mg/1
       Magnesium  as MgCOySK^O = 18 mg/1
       Precipitated magnesium carbonate required = 21 mg/1

Figure 6 shows that the pH for 100% recovery is 11.30, which is less
than 11.35 used in Case I.  Therefore for Case I, an optimum pH of 11.30
will be used.  Cases II and III, pH 11,15 and 11.00, would each provide
less than 100% recovery so Figures 10 and 11 are used.

                                                      Case I
Cost to raise pH  from 10.5 to 11.3 from
   Figure 7                                           $ 2.32

Cost to raise pH  to 10,5 for an alkalinity of
   12 using Figure 9                                  $  .50

Cost to precipitate 21 mg/1 MgCO,'3H20 from
   Figure 11                                          $  .61
                                                      $ 3,43
                                  51

-------
                                                      Case II    Case III
figure 10 (for a total alkalinity of 12 mg/1          $11,65      "$15.45" '

Figure 11 (for 21 mg/1 MgC03)                         $ 1.10      $ 1.70
                                                      $12,75      $17.15

Credit magnesium present in water                     $-7 50      $-7.50
                                                      $ 5,25      $ 9.65

   Water C ^ Washington, D, C.

       Total alkalinity =  41 mg/1
       Magnesium as MgC03*3H20 present =23,5 mg/1
       Precipitated MgC03'3H20 required = 24 mg/1

From Figure 6, the pH for  100% recovery is 10,95.  All cases will use
this as the pH of coagulation as any pH higher will give greater than
100% recovery,

                                                Case I  Case II  Case III
Figure 7  (cost to raise pH to 10,95)            $1.30   $2.20    $4.40

Figure 9  (for a total alkalinity of
   41 mg/1)                                     $  .75   $1.25    $1.98

Figure 11 (for 24 mg/1 MgCOO                   $  .80   $1.30    $2.50
                                                $2.85   $4.75    $8.88

This method of determining costs and optimum pH for coagulation is based
on the following assumptions:

   1)  A  specific amount of precipitated Mg(OH)2 is required for satisfactory
       treatment and  can be determined by jar testing.

   2)  The "pooled" magnesium solubility relationship will estimate the
       actual magnesium  solubility  in practice.  As discussed previously,
       this is a very conservative  estimate and no doubt greater recovery
       efficiencies will be obtained, allowing a lower optimum operating
       pH and subsequent  cost savings.

   3)  All relationships are  based  on the costs shown.  If costs are
       different, the operating  pH  for  the minimum cost will be different.

   4)  A  maximum of  100% recovery of the required  MgC03'3H20 dosage.  It
       is possible  that  a  stable market  for MgC03'3H20 will develop which
       might make it  economically attractive  to recover greater than
       100% of the  required MgC03'3H20  dosage.

   5)  Magnesium present  in  the  raw water is  an asset  and  decreases
       treatment costs.
                                    52

-------
No credits are given for the savings in chlorine,  The costs for the
seventeen natural waters studied are calculated and given in Table 11,
The present treatment costs shown in this table are for use of only
lime and alum and were determined either from laboratory evaluation or
annual reports      supplied by the cities,  A recent survey of treat-
ment costs in Alabama^! was used in cost estimation for the cities in
that state.  Daily water production was taken from the 1962 Public Health
Service Survey'^ where present production was not known,

Electrophoretic Mobility as a Measure of Coagulation Efficiency

The degree to which electrophoretic mobility describes treatment
efficiency determines its value as an analytical tool in the coagulation
process.  A linear regression analysis between particle mobility during
coagulation and settled color or turbidity gave an evaluation of this
relationship for twelve of the natural waters.  Table 12 summarizes the
results.  In every case, mobility showed a highly significant linear
correlation with color reduction,  Five of the twelve waters indicated
a highly significant linear correlation between mobility and settled
turbidity, while two waters were significantly correlated.  Figure 12
illustrates the relationship between coagulation particle mobility and
settled color for Jackson, Mississippi water.  Figure 13 illustrates a
high degree of correlation between coagulation particle mobility and
settled turbidity for Lanett, Alabama water.

The effect of activated silica on the mobility-turbidity relationship
is illustrated for the Cleveland water,  A significant relationship
existed between particle mobility and residual turbidity for seven
coagulation experiments using alum as the flocculant aid.  When the five
experiments using activated silica as the flocculant aid are included,
almost no correlation exists.  This effect was noticeable for all jar
tests using activated silica.  Increased dosages improved the treatment
efficiency without reducing the particle mobility.

Charge reversal did not result in a restabilization of the colloidal
color nor turbidity as would be expected for alum or ferric sulfate
treatment.  In Figure 13, turbidity removal increased as the mobility
became more positive.  In the normal pH range of coagulation with alum,
the predominant hydrolysis species is Al(OH)o which is relatively
uncharged.93  For coagulation with magnesium carbonate, positively
charged magnesium hydroxide will predominate at high coagulant dosages.
A possible "sweeping effect" of the color or turbidity by the excessive
amount of magnesium hydroxide produced could explain the good turbidity
removals at highly positive mobilities.

Prediction of the Required Coagulant Dose

Each of the waters required some minimum amount of magnesium hydroxide
floe for satisfactory treatment.  The amount of MgCO 'SKUO fed as a
coagulant is relatively unimportant since only that portion precipitated
as Mg(OH)2 is effective in coagulation.  Considering the magnesium
                                  53

-------
Ln
-P-
                                                                TABLE   11
                           ECONOMIC  COMPARISON  OF  TREATMENT  METHODS  FOR  17  NATURAL  WATERS
CITY


Atlanta, 3a.
Baltimore, >!d .
Birmingham, Ala.
(a)
Cleveland, Ohio
Detroit, Mich.
Huntsville, Ala.
Jackson, Xiss.
Lanett, Ala.
Louisville, Ky.
Montgomery, Al.1. .
Nashville, Tenn.
Opelika, Ala.
Philadelphia, Pa,
/tichmond, Va.
Tuscaloosa, Ala.
"'ashincton. D. C.
(h)
Avg. Daily
yroci'jction
A<
v- -g/ ' )
32
21
15
22
34
30
22
35
30
33
40
20
10
24
30
10
24
Cost in S/M.G. for M.gCO-, Treatment and Coagulation
pH for Minimum Cost
Cas-- 1 (ci)
C"st
3.73
3.43
2.65
3.50
3.75
3-15
2.80
9. 39
7.59
2.27
9.70
3.80
8.75
2.60
7.37
10.20
2.85
PH
J1.35
11.30
10.55
11.10
10.50
10.50
10.80
11.35
11.35
10.50
11.35
1J .00
11.35
10.95
11.35
11.35
10,50
Case 2 (e)
Cost
10.91
5.25
4.45
5.82
6.30
5.28
4.70
12.12
10.96
3.85
12.23
6.35
10.70
4.35
8.85
12.00
4.75
pH
11.15
11.15
10.55
11.10
10.50
10.50
10.80
11.15
11.15
10.50
11.15
11.00
11.15
10.95
11.15
11.15
10.50
Case 3 (f)
Cost
15.69
9.65
7.12
10.33
9.45
8.30
8.30
17.05
14.64
6.85
18.23
10.30
14.75
C.50
14.43
16.10
8.88
pH
11.00
11.00
10.55
11.00
10.50
10.50
10.80
11.00
10.50
11.00
11.00
11.00
11.00
10.95
11.00
11.00
10.50
(g)
Present
Treatment
Costs
5.94
6.08
12.26
6.25
5.70
6.40
15.00
13.50
5.00
6.12
11.22
5.58
6.30
7.40
11.08
5.70
8.39
                       (a)  Requested  to be deleted from publication
                       (b)  From 1960  PH~ Survey uhero present production not known
                       (c)  Precipitated XgCO,  as Mg(OH)2
                       (d)  Lime recovery, CaO  @ 512.00/ton
                       (e)  C02 source available, CaO @ $20.00/ton
                       (0  CO- & S.Ol/lb., CaO & $20.00/ton
                       (g)  Based on annual report supplied by city or laboratory  evaluation using alum.
                            and alum.
Cost  includes only lime

-------
                    TABLE 12

RELATIONSHIP BETWEEN ELECTROPHORETIC MOBILITIES AND
SETTLED COLOR OR TURBIDITY FOR 12 NATURAL WATERS.

City
Richmond
Montgomery
Atlanta
Jackson
Tuscaloosa
Loriett
Richmond
Montgomery
Clevelanda
Cleveland
Atlanta
Jackson
Birmingham
Philadelphia
Washington, D.C.
Lonett
Huntsville
Opelika
Tuscaloosa
Independent
Variable
Color
Color
Color
Color
Color
Color
Turbidity
Turbidity
Turbidity
Turbidity
Turbidity
Tumidity
Turbidity
Turbidity
Turbid ity
Turbidity
Turbidity
Turbidity
Turbidity
Observa-
tion
8
33
8
11
10
12
8
33
7
12
8
11
12
8
12
12
6
6
10
Correlation
Coefficient
.811**
.811**
.914**
.965**
.929**
.720**
.919**
.757**
.844v.
.299
.458
.822**
.308
. 824*
.498
. 852**
.148
.919**
.628
Standard
Deviation
2.61
5.72
3.57
3.23
2.40
4.70
1.61
4.87
3.18
5.24
7.62
2.36
5.11
1.18
2.19
1.13
3.59
2.67
1.23
aincludes data using activated silica
*denotes significant correlation (a= .05)
**denotes high significant correlation (= .01)
                           55

-------
                                30-
                                25-
                               : 20-
Ln
                              Ul
                              _>
                              Ul
                              in
                                 15-
                                 10-
                                 3-
                                        -.10     -.20    -.30    -.40    -.50    -.60     -.70


                                                            MOBILITY (p/kM. ptr V/cm.)
                                                                                          -.SO
                                                                                                 -.90
                                                                                                       -I.OO
                                   FIG.  12  SETTLED  COLOR AS A  FUNCTION  OF  PARTICLE  MOBILITY

                                           DURING  COAGULATION  , JACKSON  MISSISSIPPI  WATER

-------
I2JO
IO.O
4JO-I
20'
         +.80     +.60
+.40    +.20
  MOBILITY
                                     0      -.20
                                      per V/cm)
                                                   -.40
                                                         -.60    -.80
           FIG. 13 SETTLED  TURBIDITY  AS  A  FUNCTION  OF
                 COAGULATION  MOBILITY - LANETT WATER

-------
present in the raw water, the magnesium added for coagulation, and the
magnesium in solution after coagulation? the magnesium precipitated as
Mg(OR)2 can be calculated.  The lowest amount of Mg(OH)2 to give satis-
factory treatment was determined for each of the seventeen waters.

The required Mg(OH)2 dosage to treat a water must relate to the chemical
or physical properties of the water,  A stepwise linear regression
analysis was made on the data shown in Table 13.  The required Mg(OH)
as MgCO '3H20 was regressed as function of a water*s color, turbidity,
total alkalinity, and total hardness.  Again a BMD02R library computer
program ^ was used for the analysis,  The resulting equation was:

             Minimum magnesium dosage (as mg/1 of MgCO., 3H.?0) =
               8.33 -f ,03 (turbidity) + .46  (organic color) -
               .03 (total alkalinity) + .14  (total hardness).

The independent variables explained 67.6 percent of variations of the
dependent variable, the minimum magnesium dosage, with a standard error
of the estimate of 7.1.  The variables were added in the order of greatest
reduction of the regression sums of squares, i.e., turbidity explained
more of the variation of the dependent variable than did hardness.  This
equation has regression coefficients almost identical to those found for
the synthetic waters.  While turbidity was significant in explaining
relationships, organic color determines the required coagulant dosage.
The average color for the seventeen waters studied was only 18.  Had it
been higher, possibly more statistical significance would have resulted,
as was the case for the synthetic waters.

It is interesting to note that neither the base exchange capacity nor the
level of turbidity present influenced the coagulant dose to any great
extent.  Montmorillonite clay, emathlite clay, and natural turbidity
each gave similar predictive equations.  One hundred JCU of turbidity
would require only 2 to 3 mg/1 of precipitated MgCOo'SH^O for coagulation.

Charge reversal did not occur in coagulating emathlite  or montmorillonite
turbidity.  In several synthetic water experiments using kaolinite clay
turbidity, a very low, base-exchange capacity  clay, charge reversal was
common.  Possibly charge reversal in natural water coagulation is re-
lated to the base exchange capacity of the turbidity present, which could
then be a measure of the type of clay present.

Coagulant Recovery

As previously noted, magnesium is separated  from  softening plant sludge
at Dayton by carbonation.  The raw water at Dayton is a very  clear,
hard, ground water with no turbidity or organic color present.  Coagu-
lation of turbid waters containing organic color  would  be a different
situation.  Turbidity and organic color become highly concentrated in
the sludge and would be expected to influence  magnesium recovery.  Ten
experiments, four with natural and six with  synthetic waters, evaluated
these effects.  The results  of five of  these experiments are  shown in
                                58

-------
                                TABLE :13

       AMOUNT OF MAGNESIUM PRECIPITATED AS RELATED TO PHYSICAL AND
              CHEMICAL CHARACTERISTICS FOR 17 NATURAL WATERS


Mg Required             Turbidity    Color    Alkalinity   Hardness
(mg/1  as MgC03,3H.20)    (j, c.  )    (:Pt.Co.)    (mg/1)       (mg/1)
40.
30
10
15
22
10
25
24
30
35
35
25
24
32
21
20
22
165
105
4
10
13
14
106
50
24
7.5
6
2.5
41
104
2
7.5
15
50
30
26
12
4
10
11
15
30
27
5
0
14
38
4
8
24
13
17
4
74
54
17
51
41
27
10
92
80
34
11
12
71
48
16
17
5
83
84
17
110
71
43
12
127
100
69
13
40
86
71
                              59

-------
Tables 14 through 18,  Figure 14, a plot of the values shown in Table
16, graphically demonstrates the 'relationship between Kg"*"*", 3++  and
color as carbonation proceeds.

Table 19 summarizes the results  of the recovery experiments.  These
experiments show that nearly 100% recovery of magnesium from the sludge
was possible.  Several experiments indicated greater than 100% recovery
of magnesium from the sludge.  A combination of analytical error and
release of magnesium by the montmorillonite clay on carbonation no doubt
accounted for this error.  The release of color on carbonation would
not be a problem in coagulating  waters with an average color of 100 or
less.  More highly colored waters would require either llwasting" of a
portion of the recovered coagulant or a more elaborate recovery process.

In every case the carbonated sludge filtrate was free from turbidity,
indicating that release of turbidity would not be a problem.  The amount
of calcium, as calcium carbonate, released by the sludge on carbonation
ranged from 35 mg/1 to a high of 200 mg/1.  Assuming that the sludge
volume was equal to 1% of the total flow, the high value would represent
an increase of only 2 mg/1 total hardness.  The rate of carbonation and
the amount of turbidity in the raw water did not appear to effect the
solubilization of Mg(OH)_ or the release of color and calcium.  Color
release was closely related to magnesium solubilization, generally
reaching a maximum when about 90% of the magnesium had been recovered.

An interesting aspect of the sludge recovery experiments was the fact
that considerably less magnesium was found in solution after coagulation
in the 36 liter vessel than in the 1 liter jar tests.  The magnesium
values found for these studies,  at the indicated pH values, are shown
in Table 20.

                               TABLE 20

               MAGNESIUM SOLUBILITY AS A FUNCTION OF pH
                   FOR COAGULANT RECOVERY STUDIES
            1.63                      9.4                  11.25
            0.50                      3.3                  11.30
            0.60                      3.4                  11.25
            0.30                      1.7                  11.35
            0.80                      4.5                  11.20
            1.32                      7.5                  11.45
            0.32                      1.8                  11.30
            0.11                      0.6                  11,30
            0.65                      3.8                  11.30
                                    60

-------
                          TABLE  14

     CARBONATION OF SLUDGE  PRODUCED FROM  THE COAGULATION OF
       36 LITERS OF SYNTHETIC WATER CONTAINING. 200 mg/1
             OF ORGANIC COLOR AiND 50 mg/1 TURBIDITY
Time (Min.)
1.5
3.0
5.0
10.0
15,0
20.0
30.0
45.0
60.0
Mg^dng/l)
12
16
14
36
63
105
190
260
270
Ca++ (rag/1)
18
23
30
28
27
34
40
50
40
Color
76
138
140
410
440
720
920
1040
1020
1460 ml of sludge made up to 2 liters with distilled water.
                         61

-------
                         TABLE 15

        CARBONATION OF SLUDGE PRODUCED FROM THE COAGULATION
      OF 36 LITERS OF SYNTHETIC WATER CONTAINING 2QO mg/1
              ORGANIC COLOR AND  15 mg/1 TURBIDITY


Time (Min.)     Ms"(mg/1)     Casing/I)
5
10
15
20
30
45
60
75
90
100
110
18
25
52
86
130
184
212
218
248
255
265
24
30
18
28
54
30
34
37
48
48

85
100
205
370
480
760
820
1080
1180
1180
1020
 1360 ml of sludge made up to  2  liters with distilled water.
                            62

-------
                         TABLE 16

 CARBONATION OF SLUDGE PRODUCED FROM TOE COAGULATION OF 36
       LITERS OF SYNTHETIC WATER CONTAINING 50 mg/1
             ORGANIC COLOR AND 15 mg/1 TURBIDITY


Time (min.)    pH   Mg  (mg/1)   Ca^Cmg/l)   Organic Color
                                              (Rt.Co, units)
5
10
15
20
25
30
45

8.35
7.75
7.35
7.00
6.85
6.70
17
26
32
40
45
48
52
22
25
25
30
40
50
65
100
138
165
146
214
130
146
 1300 ml"sludge made up to 2 1 with distilled water before
carbohation.
                          63,

-------
                      TABLE  17

 CARBONATION OF SLUDGE PRODUCED FROM THF COAGUIATION
   OF 36 LITERS OF SYNTHETIC WATER CONTAINING  15 nig/1
          ORGANIC COLOR AND  15 mg/1 TURBIDITY

                                                 Color
Time (Min,
5
10
15
20
25
40
1375 ml
) PH
9.20
7.20
6.85
6.60
6.60
6.50
of sludg
Mg (mg/
6
16
26
25
27
30
e made u
1) Ca (mgy
25
29
54
66
94
108
p to 2 liter
'!) (Pt.C.o, units)
12
15
32
50
30
32
volume.
                       TABLE  18

  CARBONATION OF SLUDGE PRODUCED FROM THE COAGULATION
       OF 36 LITERS OF NATURAL WATER CONTAINING
     200 mg/'l OF ORGANIC COLOR AND 50 mg/1 ADDED
           MONTMORILLONITE CLAY TURBIDITY
                     ,  ,
Time (Min.)   .pH   Mg   (mg/1)
                                                Color
                                     (rag/1)   (pt.Co. units)
5
10
15
20
30
40
--
--
9.35
8.90
7.20
6.85
13
35
54
76
112
119
33
28
32
31
46
48
40
165
240
360
410
545
 Total sludge volume 2820 ml.
experiment.
                                1410 ml made up to 2 1 for this
                          64

-------
60-
30 
40-
30-
20-
IO-
                                     240
                                     200
                                      40
o-Mo.** by Atomic Absorption
o-Co** by Atomic Abvorptlon
A-Color (uncorrtctcd for pH)
                                                                                    60
                                                                                   50
                                                                                   40
                                                                                   30
                                                                                   20
                                                                                   10
                  \O
                                  20              30
                                       TIME  fmln.)
                                                                  40
                  FIG. 14 MAGNESIUM  RECOVERY  BY CARBONATION

-------
        TABLE .19



COAGULANT RECOVERY STUDIES
Water
Montgomery
Montgomery
Atlanta
Austin Cary Forest
Synthetic
Synthetic
Synthetic
Synthetic
Synthetic
Synthetic
Color
(Pt.Co.
units)
50
50
38
200
15
50
200
200
50
15
Tur-
bidity
(J.Q. )
165
165
104
50
50
50
50
15
15
15
MgCOo.
3H20
Adaed
(Grams)
0.15
0.45
1.26
2.80
0.53
1..40
2.88
2.88
0.72
0.36
MgCOo.
3H20J
in Sludge
(Grams)
0.07
0.25
1.15
2.70
0.38
1.41
3.00
3.00
0.59
0.34
Percent
Color
Released
11.0
4,5
4.2
25.0
6.1
18.0
28.0
16.0
9.0
25.0
% MgC03
Recovered
From
Sludge
45
54
92
96
73
100
106
103
83
95

-------
This could have been due to a mass action effect increasing the MgCOH),,
precipitation rate,  However the longer time allowed fox settling^ from
3 to 12 hbutSj, is a more likely explanation,  Both effects would be
present in actual plant conditions tending to make the economic evaluationf
based on higher magnesium solubility, more conservative,

Limited studies of sludge filterability showed that the carbonated sludge
was considerably easier to filter than the uncarbonated sludge,  Treat-
ment of the carbonated sludge with 1,5 mg/1 of anionic polymer increased
the filterability while higher dosages tended to have a negative effect^
Approximately 1 g/1 of CaCOo greatly improved filterability producing a
dry, friable cake,  Indicati-ons are that only sludges produced in treat-
ing very soft waters will present any problem in dewatering.  Addition
of an inert material, such as CaCO,,, or the use of polymers may be
necessary for these waters.  The CaCO-j precipitated in stabilization
may be adequate for this purpose,

Coagulation With Recovered Magnesium

Table 21 demonstrates the effectiveness of recovered magnesium.  Three
levels of magnesium carbonate were used to determine the effectiveness
of twice-recovered magnesium bicarbonate in solution.  This magnesium
had been used twice in previous experiments.  The resulting settled
color, turbidity, stabilized chemical characteristics of water so
treated were almost identical with those of water treated with fresh
material.  The reused magnesium bicarbonate was that recovered from
treating a highly colored water, thus representing the most unfavorable
conditions.

Application of the Process

The use of this new coagulation process in large water treatment plants
allows the recovery and reuse of both the lime and the magnesium carbonate
with carbon dioxide supplied from the recalcination of the lime.

The majority of the more than 4000 water plants in the United States
however, will not find it economically attractive to recover lime.  Since
the COo produced in recalcination greatly reduces the treatment cost, it
would be advantageous to locate an inexpensive source near the plant.
In many cases such a source exists in the form of diesel or natural gas
engine exhaust, stack gases from incinerators or power plants, or some
other industrial source.  As noted in Table 11, the difference in treat-
ment costs between Cases II and III is due entirely to the purchase of CO,.,.

A flow diagram of the unit operations involved is shown in Figure 15.
The sludge from the settling basin, containing MgCOH)-, CaCO,, turbidity,
color, etc., is pumped to the carbonation tank.  Carbon dioxide is
introduced with rapid mixing for a detention time of approximately one
hour.  The magnesium is completely solubilized as MgCHCO^)- or MgCO-,
while the CaCOo, turbidity, etc. remain as solids.  Vacuum filtration
separates the solids so that the liquid phase, containing the magnesium,
                                       67

-------
CTv
00
                                                       TABLE 21
                   EVALUATION OF TWICE  RECYCLED MAGNESIUM IN -COAGULATION OF SYNTHETIC  WATER


Jar
No.
*1
*2
*3
4
5
6
Dosage in ppir
CO
00 O
"2: O
40
50
60
40
50
60
CM
 rc
0 O
^^
130
142
153
112
117
122
g
3
5!
.5
.5
.5
.5
.5
5


PH
11.15
11.15
11.10
11.15
11.00
11.15

u
o
r-l
O
C_>
36
34
33
34
41
30
>,
.u
H
T3
r^
,0
)-l
D
...H J
4
8
5
6
8
12
>>
u
H
tH
H
,0
O
zz
-.62.
-.61
-.71
-.62
-.70
-.67
Alkalinity

SB
O
.82
60
50
68
46
70

CO
O
CJ
96
92
84
100
100
76

f>
o
u
C
0
0
0
0
0
0
o
u
^ Si
<$ a*
U
in


9.0


8.7
Vj
o
r-H
O
O


20


18
Alkalinity

CO
o
o


6


0

m
o
u


44


49
Hardness

C


41


47

NC


0


0

T


41


47
B
3 o
H O
o> w
C O
oo
w w
S CO


4
>

6
















j

             Characteristics  of  raw water
             Alkalinity as CaC03 	     25
             Total Hardness as CaC03 	2_5_
             ?H		
             Organic Color .... 	110_
             Turbidity		60_
        Comments
*Twice recovered solution of Mg(HC03)2 used as
 coagulant
            Type Clay	Montmorlllonite

-------
ON
\o
                                                 FILTER BACK WASH
                                 RAW WATER
                                        R.M.
                                            TUX-
                                                   SETTLING BASIS
:HEMICAi.
 F"o  n
L ' 
                                                   SLUDGE
<        I
-[

  . -  I
                                                       ._.co2	i
                                                     'I V.F. II PLOT.
                                                             TUMIDITY
                                FIG.  15 FLOW  DIAGRAM  FOR  TURBIDITY  REMOVAL

                                        PLANT  USING  MgC03 AND  LIME  RECALCINING

-------
can be recycled for reuse.  If lime is not recovered, this filter cake
represents a plant waste which could be handled easily and disposed of
as land fill or it might find application in agricultural use as a pH
stabilizer for soil.  Carbon dioxide, from either a source available
near the plant or by generation, is supplied to the carbonation and
stabilization units.

The dashed  lines in Figure 15 show the additional unit operations
necessary for lime recovery,  The filter cake is slurried and the clay
turbidity, color, etc. floated off in a small flotation cell.  This
process, which has been shown to be successful at a very minimal cost,
purifies the calcium carbonate before recovery.  The small volume of
float, which can easily be dewatered and  disposed of as land fill,
represents the only waste product from the entire process.  The relatively
pure calcium carbonate is then dewatered by centrifugation, and burned
in a lime kiln producing calcium oxide and carbon dioxide.  The calcium
oxide is then slaked and reused while the carbon dioxide is used for
stabilization of the water and carbonation of the sludge as shown in the
flow diagram.

Sludge thickening before carbonation could prove to be desirable.  For
most waters, the sludge before carbonation will rarely exceed 1% solids,
suspended in water containing 70-80 mg/1 of hydroxide alkalinity.
Thickening the sludge to 3% solids would allow recycling of the clear,
highly alkaline supernatant which would have a volume equal to 2/3 of
the total sludge flow.  This would reduce the required lime dosage and
also reduce the amount of carbon dioxide required for sludge carbonation
as well as the size of the carbonation basin.  These advantages must be
weighed against the capital cost and operational problems resulting from
the additional unit required.

Plants reusing lime must carbonate sludge on a continuous basis.  For a
one hour carbonation detention time, a fifty million gallon per day
plant would require only an 11,000 gallon carbonation basin, assuming
the sludge volume is 0.5% of the total flow.  A five million gallon per
day plant, recovering magnesium on a batch basis, would require a
25,000 gallon basin for the same conditions.  In addition, at least a
25,000 gallon storage tank for the magnesium carbonate solution would
be necessary.  Sludge thickening prior to carbonation would reduce the
size of both tanks.

Where lime recovery is not practiced, plants treating a moderately turbid
water could vacuum filter the carbonated sludge on an intermittent basis.
All of the carbonated sludge could be recycled until the solids reached
an undesirable level.  Removal of the solids by vacuum filtration would
allow the cycle to continue.  The increased solids concentration could
possibly increase the coagulation and settling efficiency, particularly
for waters very low in turbidity.
                                 70

-------
Photographic Comparison, of the Formation of Floes Produced With
Magnesium Carbonate and With Alum

In all of the coagulation studies, the magnesium carbonate floes formed
more rapidly and were generally larger in size and more dense than the
alum floes.  Photographs taken during coagulation of organic color with
alum in one jar and with MgCO^ in the other allow comparison of both the
rate of floe formation and the physical characteristics of the two floes
formed.  The raw water which was coagulated contained 250 mg/1 of organic
color with an adjusted alkalinity- and hardness of 30 mg/1.  The dosage
of alum required for good coagulation was established as 85 mg/1.  The
optimum dosage of MgC03 was 140 mg/lf hydrolyzed with 160 mg/1 of hydrated
lime,  A dosage of 0,5 mg/1 of potato starch was used as a flocculant
with the magnesium carbonate treatment.  Five minutes of rapid mixing,
fifteen minutes of slow mixing, and twenty minutes' settling was provided
as in previous jar tests.

Figure 16 shows the water betore addition ot coagulants and live minutes
after the coagulants were added.  Figure 17 shows the floe formed seven
minutes after coagulant addition.
                                   71

-------

                   Before coagulant addition
             Five minutes after coagulant addition

FIG. 16 PHOTOGRAPHIX COMPARISON OF Mg003 AMD ALUM FLOG DURING
        RAPID MIXING IN THE REMDVAL OF ORGANIC OOLOR
                               72

-------
FTC. 1?  PHDTOGRAPHIC ODMPARISON OF Mg003 AND ALUM FLOG DURING FLOCCULATION IN THE
         REMDVAL OF ORGANIC  OOLOR

-------
FIG. 18  PHOTOGRAPHIC COMPARISON OP THE RATE OF SETTLING FOR Mg003 AND ALUM FLOGS FORMED
         IN THE REMDVAL OF ORGANIC  COLOR

-------
  A
v  y  
        NT
                           .
 FIG. 19 MAGNESIUM CARBON FLOG (MAGNESIUM HYDROXIDE DARK
      WITH ABSORBED COLOR AND CALCIUM CARBONATE
      CRYSTALS) MAGNIFIED 200 TIMES
                i
FIG. 20 PHOTO-MICROGRAPH OF ALUM FLOG MAGNIFIED 100 TIMES
                   75

-------
a-
               PIG. 21  PHD TO-MICROGRAPH OP ALUM FLOG MAGNIFIED 200 TIMES

-------
                    SECTION VII,  ACKNOWLEDGEMENTS
Acknowledgement is extended to the following personnel of the City of
Gainesville, Florida whose advice and assistance was most helpful in
carrying out this study;

       John R, Kelly, Director of Public Utilities
       Charles H, Oakley, Director of Finance
       H, E, Wbrsham, Controller
       R, P Vogh, Sup^t, Water and Waste Water
       Linda Warrington, Account Clerk

Acknowledgement is also extended to the Department of Environmental
Engineering, University of Florida, for making available laboratory
space and all needed facilities for carrying out this study, and for
the helpful advice and assistance of the following:

       Dr. E, E. Pyatt, Department Chairman
       Dr. Wt H. Morgan, Assistant Chairman
       Dr, J. E, Singley, Professor of Water Chemistry
       Roger Yorton, Graduate Student
       Roy Burke, Graduate Student
       Jeanne Dorsey, Secretary,

Acknowledgement is extended to Mr. Edmond P. Lomasney, Project Officer,
for his cooperative direction and many helpful suggestions.
                                  77

-------
            SECTION VIII.  REFERENCES
 1,  Hudson, H. E., Jr.,  "How Serious is the Problem?"
       Proc. 10th San. Eng. Con-f. of Illinois, Urbana,
       Illinois (1968).

 2.  Neubauer, W.  R.   "Waste Alum Sludge Treatment,"  J.
       AWWA, 60:819'(i968).

 3,  O'Brien, E. F, and Gere, B. A.  "Waste Alum Sludge
       Characteristics and Treatment,"  N.Y. State Dept. of
       Health, Research Report No. 15 (1966).

 4.  Russelman, H. B,.  "Characteristics of Water Treatment
       Plant Wastes,-"  Proc. 10th San. Eng. Conf. Illinois.
       Urbana, Illinois (1968).

 5.  McWain, J. D.  "Dewatering by Lagoons and Drying Beds,"
       Proc. 10th San. Eng. Conf. of Illinois,  University of
       Illinois, Urbana, Illinois (1968).
                              *
 6.  Doe, P. W.  "A Report on the Disposal of Sludge from
       Water Treatment Plants,"  British Water Works Assoc..
       Jubilee Traveling Scholarship (1966-67).

 7;  Young, E. F.   "Water Treatment Plant Sludge Disposal
       Practices in United Kingdom,"  J. AWWA,  60:717 (1968).

 8,  proudfit, D.  P.   "Selection of Disposal Methods for
       Water Treatment Plant Wastes,"  Proc. 10th San. Eng.
       Conf. , University of Illinois, Urbana,  Illinois ("1968)

 9.  Sankey, K. A.  "The Problem of Sludge Disposal at the
       Arnfield Treatment Plant,"  J. Inst..of Water
       Engineers,  21:367 (1967).

10.  Jewell, W. M., U. S.  Patent 718,465 (1903).

11,  Mathis, W. R., U. S.  Patent 1,441,731  (1923).

12,  Black Laboratories.   "The Disposition of Water Plant
       Sludge," Unpublished laboratories report to the
       Orlando Utilities Commission,  Orlando,  Florida (1951).

13.  Roberts, J. Morgan and C.  P. Roddy.  "Recovery and Reuse
       of Alum Sludge at Tampa, ".  J.  -AWWA ,  52:857 (1960).
                            79

-------
14.  Fujita, H.  "Tokyo's Asaka Purification Plant,"
       Water and Sewage Works, 114:73 (March 1967),

15.  Fulton, G. P.  "Disposal of Wastewater from Water Fil-
       tration Plants,"  J. AWWA, 61:322 (July 1969).

16.  American Water Works Association Research Foundation.
       "Disposal of Wastes from Water Treatment Plants,"
       August 1969 (Water Pollution Control Research Series).

17.  Black, A. P., C.  F. Wertz and C. R. Henry.  "Recalcin-
       ing Softening Sludge at Miami, Florida,"  Water Works
       Eng; (March 1951).

18.-  Manual for Water Works Operators.   Texas Water and
       Sewage Works Assoc. , 199-200  (1959).

19.  Doermer, A., F. W. Holbrook and 0.  W. Fortner.  "The
       Bicarbonate Process for the Production of Magnesium
       Oxide," Technical Paper 684, U.  S.  Department of the
       Interior, U, /S. Gov. Printing Office, 1946.

20.  Eidsness, F, A. and A. P. Black.  "Carbonation of
       Water Softening Plant Sludge,"  J.  AWWA, 49:1343
       (1967).

21.  Flentje, M. E,  "Calcium and Magnesium Hydrates,"
       J. AWWA, 17:253 (1927),-

22.  Lecompte, A. R.  "Water Reclamation by Excess  Lime
       Treatment of Effluent,"  TAPPI,  49:121 (December
       1966).                   '

23.  Black, A, P. and  W. T. Eiffert.  "Recovery of  Calcium
       and Magnesium Values from Lime-Soda Softening Sludge
       at Dayton, Ohio,"  Ohio - Michigan  Pollution Control
       Conf. (May 1970).

24.  Houston, A. C.  8th Research Rpt.  of  Metropolitan
       Water Board, London, England.

25.  Hoover, C. P.   "Water Softening as  an Adjunct  to Water
       Purification,"   J. AWWA,  17:751  (June 1927).

26.  Riehl, M. L.,  H.  H. Weiser and R.  T.  Rheins.   "Effect
       of Lime Treated Water upon Survival of Bacteria/1
       J. AWWA, 44:466 (May 1952).

27.  Chaudhure, Maley  and R.  S.  Engelbrecht.  "Removal of
       Viruses from Water by Chemical Coagulation an-d
                       J. AWWA,  62:563  (Sept. 1970).
                          SO

-------
 28.   Berg,  Gerald,  R.  B.  Dean,  and D. 'R.  Dahling.   "Removal
        of Poliovirus  1  From Secondary  Effluents by Lime
        Flocculation and Rapid  Sand Filtration," J.  AWWA
        60:193  (February 1968).                        ~

 29.   Pearson,  R,  G.   "Hard  and  Soft Acids and Bases," j
        American Chem.  Soc.,  85,  3533 (1963).

 30.   Laitinen, H, A. Chemical Analysis.   McGraw-Hill  Book
        Company, Inc., New York  (1960).

 31.   Graj,  D.  L.  and Goldsmith,  J.  R.   "Some  Hydrothermal
        Syntheses  of Dolomite and Protodolomite," J. Geol..
        64,  173 (1956).

 32.   Gallaher, W. W. and Buswell, A. M.   "Investigations  of
        Chemical Reactions Involved  in Water Purification,"
        Illinois Water Survey Bulletin 22  (1920-25).

 33.   Ryznor, J. W., Green, J., and  Winterstein, M.  G.   "Deter-
        mination of  the  pH of Saturation of Magnesium  Hydro-
        xide,"  Industrial and Engineering Chemistry 38:  1059
        (1946).

 34.   Krige, G. J., and Arnold,  R.   "The Aging  to Aqueous
        Suspensions of Magnesium Hydroxide," J.  So.  African
        Chem. Inst.,  1,  62 (1948).

 35.   Travers,  A.  and Nouvel, C.  "On the Solubility of Magne-
        sium Hydroxide at High Temperatures,"   Compt.  Rend.
        (Fr.),  188:  499 (1929).

 36.   Kline, W. D.   "The Solubility  of Magnesium Carbonate
        (Neoquehonite) in Water at 25C + Pressure of  Carbon
        Dioxide up to One Atmosphere," J.A.C.S., 51, 2096
        (1929).

 37.   Britton,  S.  C.   "Electrometric Studies of the Precipita-
        tion of Hydroxides.  Part I.  Precipitation of Magne-
        sium, Manganous, Perrous, Cobalt, Nickel, and  Thorium
       Hydroxide  by.the Use of  the Hydrogen Electrode,"
       J. Chem. Soc; London, 127, 2115  (1925).

 18.  Bube, K.   "Uher Magnesiumammoniumphosphat," Z. Analyt.
       Chem. 49,  553 (1910).

39.  Gjaldbach, J. K.,  "Untersuchungen  uhes die Loslichkeit
       des Magnesiumhydroxyds,"  Z.  Anorg.  Ally. Chem., 144,
       150 (1925).

40.  Dupre,  Jun. and Bialas,  J.  "Zur Bestimmung der Los-
       lichkeit Von  Magnesia und Zinkoxyd  in  Wasser auf
       Grund das Elektrischen Lutvermagen," Z. Angew Chem.,
       54-55 (1903).	
                          -81

-------
41,  Herz, W. and Muhs, G.   "Uber das Gleichgewicht Mg(OH)?
       + 2NH/C1 MgCl9 + 2NH/01-I," Z. Anorg. Chem. , 38,  138-
       41  (1904).   Z      *

42.  Kohlrausch, F. and Rose, F.  "Die Loshlichkeit einiger
       schwer loslichen Korper im Wasser, heurteilt aus der
       elektrischen Leitungsfahigkeit der Losungen,"   Z^
       Physlk. Chem., 234-43 (1893).

43.  Loyen, J. M.  "Chemishes Gleich guvicht in Ammonia
       kalischen Magnesium salz loserigen,"  Z. Anorg.  Chem.,
       11: 4-415 (1896).

44.    Larson, T. E., Lane, R. W. and Neff, C. H.  "Stabiliza-
       tion of  Magnesium Hydroxide in the Solids-Contact
       Process," J.AWWA, 51: 1551 (Dec. 1957).

45.  Stumm, We'rner and Morgan, James J.  Aquatic Chemistry.
       Interscience,'New"York, N.Y. (197077

46.  Black, A. P. and Christman, R. F.  "Sludge from Lime-
       Soda Softening," J. AWWA, 53: 6, 737 (June 1961).

47.  Johnston, J.  "The Solubility Product Constant of Calcium
       and Magnesium Carbonates,"  J.  Am. Chem. Soc.,  37:
       2001 (1915).                '"

48.  Monis, R.  The Chemical Process Industries.  McGraw-
       Hill Book Company,  Inc., Second Edition, New York,
       N.Y. (1956).

49.  Arnov.,  R.  I.  "Conversion of Magnesium Carbonate
       Trihydrate into Basic Carbonate,"  UKR. Khim Zh,
       32(5), 521-3 (1966) (Russ.).

50.  Oil, Paint Drug Reptr. 198: 10, New York",'N.Y. (1970).

51.  Black, A. P.  Unpublished Engineering Report to the
       City of Dayton, Ohio  (October, 1970).

52.  Langelier,  W. F.  and  Ludwig, H. F.  "Mechanism, of
       Flocculation in the Clarification of Turbid Waters,"
       J. AWWA,  41: 163 (1949).

53.  LaMer, V. K. and Heutz, T. W.  J. Phy. Chem., 67:
       2417 (1963).                       	

54.  Markle,  S.   "Mechanism of Coagulation in Water Treatment,"
       J. ASCE,  Sanitary Division, 3:  3117 (May, 1962).

55.  Fair,  G. M., Geyer, J. C., Morris, J. C.   Water Supply
       and Waste Disposal,  John Wiley & Sons,  New York
                         82

-------
56.   Bartow, E. and Peterson,  B.   "Effect of SaTts on the
       Rate of Coagulation and the Optimum Precipitation of
       Alum Floe."  Ind.  Eng.  Chem.  20:51 (1928)..

57.   Packham,  R. F.  "Some Studies of the Coagulation of
       Dispersed Clays with Hydrolyzing Salts."  j_. Colloid
       Science. 20: 81-92 (1963).

58.   Pillepovich, J.  B,,  Black,  A. P.,  Eidsness, F. A., and
       Stearns, I. W.   "Electrophoretic Studies of Water
       Coagulation,"  J.  AWWA,  50:  1467  (1958).

59.   Black, A. P. and Rice, Owen.   "Formation of Floe by
       Aluminum Sulfate."  Ind.  Eng.  Che/n. 25: 811 (1933).

60.   Stumm, W. and Morgan, J.  J.  "Chemical Aspects of Coagu-
       lation," J. AWWAt  54: 971  (1962).   Discussion, p. 992,
       Black,  A. P.

61.   Mattson,  S.  "Cataphoresis  and  the Electrical Neutrali-
       zation  of Colloidal Material," J.  Phy. Chem.,  32:
       1532 (1928).

62.   Christman, R. T.  and Ghassemi,  M.   "Chemical Nature of
       Organic Color  in Water,"  J. AWWA,  55: 1347166 (1966).

63.   Black, A. P. and Christman,  R.  T,   "Chemical Characteris-
       tics of Colored Surface' Waters,"  J.  AWWA, 55: 753
       (1963).

64.   Birge, E. A. and Juday, C.   "Particulate and Dissolved
       Organic Matter in Inland  Lakes," Ecol. Monographs,
       4: 4^0  (1934).

65.   Gjessing, E. T.   "Ultrafiltration  of Aquatic Humus."
       Communication  in Environmental Science and Technology,
       37"? (May 19TDT

66.   Kitano, Y., Kznomori, N., and Tokuyama, A.  "Influence
       of Organic Matter on Inorganic Precipitation."
       Organic Matter in Natural Waters,  U.  of Alaska
       Symposium.

67.   Shapiro,  J.  "Effect of Yellow  Organic Acids on iron and
       Other Metals in Water," J.  AWWA, 56,  1062-81 (1969).

68.   Gloyna, F. E. and Oldham, W.  R.   "Effect of Colored
       Organics on Iron Removal,"  J.  AWWA, 61: 11 (1969).

69.   Singley,  J. E.,  Harris, R.  H.,  Maulding, J. S.  "Correction
       of Color Measurements to  Standard  Conditions,"
       J. AWWA. 58: 4  (April  1966).
                          83

-------
70,  Black, A, PM Singley, J. E,, Whittle,  G, P,,  and Maulding,  J,  S,
       uStoichiometry of the Coagulation of.  Color Causing Organic
       Compounds with Ferric Sulfate," J, AWWA. 53; 10 (Oct, 1963),

78.  Te Kippe, J, E, and R, K, Ham,  '^Techniques of Coagulation Testing,
       Part I,"  J, AWWA, 62 (September 1970),

79,  Black, A. P, and S, A, Hannah,  "Electrophoretic Studies of
       Turbidity Removal by Coagulation with Aluminum Sulfate,"  J.  AWWA,
       53:  4, 438 (April 1950),

80,  Replaceable Bases in Soils Devoid of Carbonates.  In Official
       Methods of Analysis of the Association of Official Agricultural
       Chemistry,  George Banta Publishing Co,, Menasha,  Wisconsin
       C8th ed, 1955), pp, 39142.

81.  Standard Methods for the Examination of Water and Waste Water,
       APHA, AWWA, and WPCF, New York (12 ed., 1965).

82.  Lumetron Photoelectric Colorimeter Model 450 for Nessler Tubes,
       Photovolt Corp,, 95 Madison Avenue, New York 16, New York.

83.  Methods for Analyses of Selected Metals in Water by Atomic
       Absorption,  Fishman, M. J. and Downs, S. C. Geological Survey
       Water Supply Paper 1540-c (1966).

84.  Zeta-Meter Manual,  Zeta Meter, Inc., Second Edition, New York,
       New York (1968).

85.  Dixon, W, J.  BMP Biomedical Computer Program.  University of
       California Press, Berkeley and Los Agenles (1968).

86.  Steel, R. G. and Tonie, J. H.  Principles and Procedures of
       Statistics.  McGraw-Hill Book Company, Inc., New York, N.  Y.
       (1960),

87.  117th Annual Operating Report of the Detroit Metro Water
       Department (1969),

88,  Annual Operating Report, Nashville,  Tennessee (1969-70).

89.  Philadelphia Water Department's Report, January 1968 - June 1969.

90.  Operating Record for the Year 1970.,  City of Jackson, Mississippi.

91.  Treatment Cost Survey of Surface Water  Plants in Alabama, Alabama
       State Health Department (1970).
                               84

-------
92,  Municipal Hater Facllltles->--Commtfnltles of 25 ?000^ Population and
       Qyer? United States and Possessions as of January 1? 1962.  U.S.
       Public Health Service Pub, 661,

93,  Sullivan, J, H, and Singley, J,  E,  "Reactions of Metal Ions in
       Dilute Aqueous Solution;  Hydrolysis of Aluminum,"  J. AWWA,
       60:  11, 1280-1287 (Nov. 1968),
                                  85

-------
SECTION IX.  APPENDIX
         87

-------
                                             TABLE 22
                    COAGULATION OF ATLANTA, GEORGIA WATER WITH MgC03  AND ALUM
Jar
No.
1
2
3
4
5
6
Dosage in ppm
ro
W> O
s o
20
30
40
40


CX!
/  \
03 S
C_> O
91
97
103
118


E
D
f 1
<
.5
.5
.5
.5
10
13
PH
11.15
11.15
11.15
11.25
7.50
7.48
1-4
O
^
0
o
24
20
15
2
9
8
furbidity
23.0
Mobility
-.45
14.oL.34
4.6
1.0
6.C
3.C
0
+ .41


Alkalinity
:r
o
64
76
68
92 '
0
0
oo
O
0
60
64
76
60
0
0
CO
0
0
0
0
0
0
10
8
0
u
J3 X
S) C^
u
to


9.0
9.0

9.0
t-i
o
r-i
O
u


11



Alkalinity
fl
O
0


0
0

0
ro
O
0


45
38

16
Hardness
C


45
38

16
NC


0
0

2
T


45
38

18
Magnesium
as CaCO-j


16
10
















oo
00
          Characteristics of  raw water
          Alkalinity as CaC03 	
          Total Hardness as CaCO-j 	
          PH		
          Organic Color .... 	
          Turbidity 	 	
          Type Clay		
         Magnesium as CaCOj
   11
   13
                               Comments
Raw water mobility -1.24
    7.65
   38
  104
Natural

-------
                                           TABLE  23
                 MgC03  AND ALUM COAGULATION OF BALTIMORE, MARYLAND WATER
Jar
No.
1
2
3
4
5
'*
Dosage in ppm
CO
bOO
S O

15
20
.25


cv
x-\
^s
95
105
105
107


s
3
r-t
<
2
1.5
.5
.5
8
3D
pH
11.15
11.15
11.15
11.15
6.00
6.00
v<
o
r-{
O
u






Turbidity
2.1
3.5
2.5
2.1
1.3
0.6
Mobility
-.57
-.47
-.33
0


Alkalinity
M
O
74
70
64
56
P

CO
O
O
48
60
56
60
0

f>
o
o
PC
0
0
0
0
10

0
u
JD 32
nj a.
4J
CO
9.0


9.0
9.0

J
o
H
O
o






Alkalinity
CO
o
o
0


0
-

n
o
CJ
x.
54


40
24

Hardness
C
54


40
24

NC
32


29
30

T
86


69
54

Magnesium
as CaC03
10


16
















00
vO
        Characteristics of raw water
        	;	 
        Alkalinity as -CaC03 	
        Total Hardness as CaC03 	
        pH.		
        Organic  Color  ....  	
        Turbidity  	  	
        Type Clay		
         Magnesium as .CaCO-
                              Comments
   12
   40
    6.00
Natural
   13

-------
                                             TABLE 24
                   LIME  AND ALUM COAGULATION OF BIRMINGHAM, ALABAMA WATER
Jar
No.
1
2
3
4
5
6
Dosage in ppir
tN!
x~-\
 re
o o
s '
90
130
140
150


e
3
i 1
<
.5
.5


15
20
vO
cfl CT>
33 rH



.5


pH
10.70
11.15
11.18
11.25
7.50
7.45
M
O
H
O
O
8
7
8
8
8
6
Turbidity
15
4.5
5.5
4.6
1.0
0
Mobility
-.55
-.58
-.58
0


Alkalinity
5C
O
45
72
80
81
0
0
f>
O
U
70
48
48
62
0
0
ro
O
c_>
T"
0
0
0
0
$2
60
o
u
,0 W
CO B.
4-1
to

9.0

9.0

8.8
M
0
rH
O
CJ>

3

3

7
Alkalinity
n
O
0

2

0

8
ro
O
O
32

37

52

63
Hardness
C

39

52

71
NC

5

9

16
T

44

61

.87
 rt
H O
to c_>
0) CO
C U
oo
cd w
2 to

14

18

25














VD
O
Characteristics of raw water
Alkalinity as CaC03 	74
Total Hardness as CaCC^ 	83
pH.  .		7.60
                                                                   Comments
                                                                Raw water mobility -.77
          Organic Color ....
          Turbidity 	
          Type Clay	
          Magnesium as CaCO-
                            12
                            10
                         Natural
                            25

-------
                                  TABLE  25
       MgC03, LIME,  AND ALUM COAGULATION  OF CHATTANOOGA,  TENNESSEE  WATER
Jar
No.
1
2
3
4
5
6
Dosage in ppm
CO
00 O
"Z, 0

15
20
25


CM
.*~s
H> 33
CJ O
v^
120
116
125
128


e
< i
_.SL_
.5
.5
.5
.5
15
20
pH
11.15
11.10
11,15
11.15
7.30
7.25
M
o
1-1
o
u
19
15
15
11
11
8
Turbidity
15
10
10
7
3.0
1.8
Mobility
-.81
-.42
0
K38


Alkalinity
*-r<
0
86
64
86
82
0
0
(O
O
U
56
60
68
72
0
0
CO
O
u
0
0
0
0
40
38
o
4J
.0 ad
CO O.
iJ
w
9.0
9.0

9.0

9.0
^
o
r~\
O
o
6
6

4

7.0
Alkalinity
CO
o
o
0
0

0

0
CO
8
x.
42
38

44

51
Hardness
C
42
38

44

51
NC
20
18

22

31
T
62
56

66

82
Magnesium
as CaC03
11
13

16

15














Characteristics of raw water
Alkalinity as CaC03 	
     48
Total Hardness as CaCO-j
     71
                               Comments
Mobility  of raw water -1.07
pH	
Organic Color . . .
Turbidity  	
Type Clay  .  . . . .
Magnesium as CaCOg
      7.85
     24
     15
Natural
     15

-------
                                   TABLE  26
           COAGULATION OF CLEVELAND, OHIO WATER WITH  LIME AND  ALUM
Jar
No.
1
2
3
4
5
6
Dosage in ppm
CM
^s
rt 33
CJ> O
^-^
90
90
160
180


M
rl tsl
W O.
1!
2.0
4.0




a
3
iH
<



.5
8
10
PH
10.35
10.35
11.20
11.35
7.90
7.80
vj
o
tH
O
O
2
2
2
2
2
2
Turbidity
4.8
3.3
6.0
2.7
4.0
2.6
Mobility
-1.10
-i.o;
0
+ .57


Alkalinity
32
O
14
14
76
88
0
0
n
O
o
44
50
28
28
0
0
f>
0
o
i-r*
0
0
0
0
as
86
0
U
& re
03 CU
ij
in
9.0
9.0
9.0
9.0

9.0
M
O
H
0
O
1
1_




Alkalinity
<*
0
o
0
0
0
0

0
c->
O
O
X
39
47
38
36

98
Hardness
C
39
47
38
36

98
NC
31
32
32
32

38
T
70
79
70
70

136
e
3 <^
H O
w o

-------
                      TABLE  27
     COAGULATION  OF DETROIT,  MICHIGAN WATER BY
PRECIPITATION OF  MAGNESIUM PRESENT  BY LIME ADDITION
Jar
No.
1
2
3
4
i
i '
5
6
Dosage in ppm
c^3
CO Dll
O O
75
40
120
120
150

z
rH




.5
15
ii
H CM
in o
II
4.0
1.0
3.0



pH
10.70
10.50
11.10
11.10
11.20
7.60
V-i
o
rH
O
o






Turbidity
3.5
5.5
9-0
6.0
5.0
2.0
Mobility


-.34
-.53
-.46

Alkalinity
X
o
14
2
50

74
0
o
u
38
56
28

24
0
ro
0
o
0
0
0

0
74
o
.u
,0 J
CO O*
ij
an

9.0

9.0
9.0
9.0
u
o
o






Alkalinity
o

0

0
0
0
CO
o
o

42

37
35
86
Hardness
C

42

37
35
86
NC

31

20
17
26
T

73

57
52
112
Magnesium
as CaC03



20
9
30














                 100
Characteristics of raw water
Alkalinity  as CaC03 	0_
Total Hardness as CaCO-j 	
pH		
Organic Color ....  	
Turbidity 	  	
Type Clay		
Magnesium as CaCOo
                                             Comments
                   7.90
                   2.5
                Natural
                  30

-------
VD
                                                      TABLE 28
                        COAGULATION  OF HUNTSVILLE, ALABAMA WATER WITH  MgC03 AND WITH
                                                                                 ALUM
Jar
No.
1
2
3
*4
Dosage in ppn|
ro
000
S f-5
0
15
20
20
5 J20
6 j 30
Cs
rt pi
o o
0
120

r-t
^
7
.5
125 I .5
f
125
110
115
.5
.5
.5
pH
7.5
11.15
11.15
14
O
o
o
5.0
3.0
1.0
11.155 3.0
10.95J 1.0
11.05
1.0
Turbidity
3.0
3.8
2.5
6.4
JD.O
10.0
Mobility
-.48
-.37
0
-.56
K48
0
Alkalinity
a
o
0
88
92
86
54
62
o
0
48
60
64
44
. 44
o
o
w
50
0
0
0
.0
0
o
.JO X
CO P
u
9.0

9.0



V)
o
-i
o
o






Alkalinity
cT
6

2



0
re
52

28


Hardness
c |NC
58

30
33

24
1
1
|

T
91

54



Magnesium
as CaCOj
20

14
Calcium
as CaC03
71

40




| j i
31 1 1
24 I
!
Characteristics of raw v;ater
Alkalinity as CaC03	54_
Total Hardness as CaCO-j         84
                                                                     Cosssnts
                                                              *Hal95 at  .5 mg/1
                                           7.5
Organic  Color .
Turbidity  .  . .
Type Clay  .  . .
                                          13
                                      Natural

-------
                                              TABLE 29
            MgC03 AND ALUM COAGULATION OF JACKSON, MISSISSIPPI WATER
Jar
No.
1
2
3
4
5
. 6
Dosage in ppm
CO
bOO
 0
20
30
40
50


r-i
s~\
nj js
O 0
V^
90
95
113
119


e
rH
^
t-1
0
0
0
0
.9
8
o
jj
.0 SB
to o.
ij
V)



9.0

9.0
M
0
iH
O
e>



5

8
Alkalinity
CO
o
o



2

0
o
o
u
te



38

16
Hardness
C



39

16
NC



0

7
T



39

23
Magnesium
as CaC03



11
















VO-.
           Characteristics of raw water
           Alkalinity as CaC03 	10
                                                         Comments
                                     12
                                      7.55
                                     27
Total Hardness as CaC03
pK	  	
Organic Color ....  	
Turbidity-		7.4
Typ Clay .......     Natural
Magnesium as CaCO,          2

-------
                                    TABLE 30
            MgC03  AND ALUM COAGULATION OF LANETT,  ALABAMA WATER
Jar
!No.
1
2
3
4
5
6
Dosage in ppm
CO
OOO
s o
30
35
40
35


CN
s~^
 w
o o
**^t
105
108
110
125


e
3
i-H
<
.5
.5
.5
.5
10
15
PH
11.15
11.15
11.15
11.25
7.10
7.00
K
0
iH
O
0
30
13
11
13
13
11
Turbidity
5.0
3.0
2.0
2.4
3.6
3.6
Mobility
-.32
K.37
K47
+ .34


Alkalinity
K
0


94
87
0
0
n
O
c_>


60
54
0
0
en
O
O
22


0
0
.7
5
o
4J
.C K
tO O.
JJ
V3



9.0
9.0
9.0
j-i
o
r-l
0
o



4
5.0
5.0
Alkalinity
n
O
o



0
2
2
ro
0
o
X



53
16
12
Hardness
C



53
18
14
NC



8
5
9
T



61
23
25
Magnesium
as CaCO-j


14
13
















Characteristics^ of raw water
Alkalinity as CaC03 	  17
Total Hardness as CaCO^ 	17
PH	  	7.55
Organic  Color ....  	30
Turbidity		105 ___
Type Clay		Natural
Comments

-------
                                             TABLE 3Z
                  LIME AND ALUM COAGULATION  OF LOUISVILLE,  KENTUCKY WATER
Jar
No.
1
2
3
4
5
6
Dosage in ppm
0>l
td /*~s
0 35
O
*^r
60
120
100
140


e
3
t-i
<


.5
.5
10
15
ii
vH e>j
CO O
II
2.0
2.0




PH
10.75
11.20
10.90
11.25
7.35
7.30
i*
o
r-i
o
o
3
3
10
4
6
4
Turbidity
14.5
8.0
7.8
4.5
1.8
1.4
Mobility


K38
4-. 73


Alkalinity
35
0
34
92
42
80
0
0
ro
O
U
60
60
60
56
0
0
CO
0
o
32
0
0
0
0
50
48
0
u
.0 35
(0 (X
u
w

1
9.0
9.0
9.0

>j
0
rH
0
O


5
4
5

Alkalinity
ro
O
<_>


0
0
0

CO
O
U


32
28
59

Hardness
C


32
28
59

NC


61
55
62

T


83
81
121

Magnesium
as CaCO-j


17
12
















\o
-J
          Characteristics of raw wat_er
          Alkalinity as CaC03 	51
          Total Hardness as CaCp3 	110
          PH		7.50
          Organic Color .... 	11
          Turbidity		106
          Type Clay		Natural
          Magnesium as CaCOj             33
          Comments
Mobility of raw water -.98

-------
                                             TABLE 32
                MgC03, LIME, AND  ALUM COAGULATION OF NASHVILLE,  TENNESSEE WATER
Jar
No.
1
2
3
4
5
6
Dosage in ppm
P")
00 O
H o


10
15


CM
x-N
(0 22
0 0
v^s
110
150
116
120


E
iH
<
.5


.5
LO
12
pH
10.95
11.20
11.00
11.00
7.55
7.50
vj
o
t-{
o
(_>
8
7
8
8
8
7
turbidity
5.4
2.8
6.0
4.6
2.6
1.6
Mobility
-.63
-.60
-.67
-.57


. Alkalinity
pr!
O
43
98
48
56
0
0
(*>
O
O
54
58
44
44
0
0
CO
o
O
SB
0
0
0
0
6,6
65
o
4_>
.
.0 E
RJ a
u
CO
9.0
9.0

9.0

9.0
i-i
o
i-i
o
u
8
4

6

8
Alkalinity
CO
O
O
2
0

4

12
rn
o
o
40
38

33

61
Hardness
C
42
38

37

73
NC
16
20

15

24
T
58
58

52

97
Magnesium
as CaCOo
15
12

19


Stabalized
Turbidity
0.3


0









.so
CD
          Characteristics of raw water
          Alkalinity as CaCO-j 	71
          Total Hardness as CaCO-j 	86
          pH.		7.60
          Organic Color ....	8
          Turbidity		7 .5
          Type Clay		Natural
          Magnesium as CaCO^             17
Comments

-------
                                        TABLE 33'
                 MgC03 AND ALUM COAGULATION OF OPELIKA, ALABAMA WATER
Jar
No.
1
2
3.
4
5
6
Dosage in pptr
PO
60 O
S "
5
10
20
25


C-J
X~N
 s
u o
NX
90
92
106
109


s
fH
<
.5
.5
0
.5
7
LO
pH
10.95
11.10
11.10
11.15
6.90
6.80
t-i
0
H
O
0
7
1
2
3
7
4
[Turbidity
13
16
12. f
3.4
1.0
0
Mobility
-.81
-.68
-.49
+ .43


Alkalinity
*i"
o

66

72

0
CO
0
0

32

32

0
CO
0
y

0

0

14
o
tj
n) "E.
u
C/5



9.0

9.0
u
o
^i
0
o






Alkalinity
fo
o
o





3
CO
O
CJ





18
Hardness
C





21
NC





9
T





30
Magnesium
as CaCO-j

10


















Characteristics of raw water
Alkalinity as CaC03 	17
Total Hardness as CaCO^ 	17
pH		6.95
Organic Color ....  	10
Turbidity		14
Type Clay		Natural
Comments

-------
                                                   TABLE  34
              LIME, MgC03,  AND ALUM COAGULATION  OF PHILADELPHIA,  PENNSYLVANIA WATER
Jar
No.
1
2
3
*4
5
6
Dosage in ppm
CO
oo o
 0


10



CN]
X-N

-------
                                        TABLE ..35
             COAGULATION OF  RICHMOND,  VIRGINIA WATER WITH MgC03 AND ALUM
Jar
No.
1
2
3
4
5
6
Dosage in ppw
f>
60 O
s o
20
25
30
40


CNJ
x-x
-i
o
T-H
O
O


5
5
12
8
Alkalinity
CO
o
o


0
0
0
0
ro
0
U
Z


40
38
29
31
Hardness
C


40
38
29
31
NC


15
15
24
24
T


55
53
53
55
3 CO
H O
tn o
a) n)
C 0
60
r; w
S tf


15
17
















Characteristics of raw water
Alkalinity  as CaC03 _
Total Hardness as
pH
                               Comments
   27
   43
Organic Color  ....
Turbidity ......
Type Clay ..... .
Magnesium as CaCOo
   30
   24
Natural

-------
10
                                                    TABLE  36
                          MgC03  AND ALUM COAGULATION OF TUSCALOOSA,  ALABAMA WATER
Jar
No.
1
2
3
4
5
6
Dosage in ppm
CO
w> o
S o
15
20
20
25


CN'
x-x
to a
o o
V '
83
86
97
100
1
2
B
D
H

O
*



0
.3

0
u
jz a
CO C.
4->
W



9.0
9.0

S-l
o
r-1
0
CJ>






Alkalinity
CO
O
c_>



2
3

ro
O
O
a:



38
6

Hardness
C



40
9

NC



2
6

T



42
15

Magnesium
as CaC03



12
















                                        4.0
Characteristics of raw water
Alkalinity as CaC03 	
Total Hardness as CaC03 	5.0
pH		6.3
Organic  Color .... 	26
Turbidity		    4
Type Clay	'. 	Natural
                                                                    Comments

-------
                                                   TABLE 37-
                       MgC03,  LIME,  AND ALUM COAGULATION OF WASHINGTON, D.C.  WATER
Jar
No.
*1
*2
3
4
5
6
Dosage in ppm
CO
00 O
Z. o

10
15
20


OJ
s~\
rt re
u o
^^x
120
127
130
117


(=
3
,-!
-i
o
i-H
O
<_>
12
8
7
11
10
6
Turbidity
3.0
2.2
2.0
2.6
3.0
1.5
Mobility
0
K49
K41
+ .72


Alkalinity
a
o
76
78
72
56 '
0
0
n
0
c_>
48
48
54
64
0
0
CO
O
CJ
-r-
0
0
0
0
37
33
o
u
JD a:
CO CL
4-1
in



9.0

8.4
u
o
t-H
0
u



5

7
Alkalinity
f>
0
o



4

0
CO
O
u
3=



44

40
Hardness
C



48

40
NC



27

36
T



75

76
Magnesium
as CaC03



18

17














o
          Characteristics of  raw water
          Alkalinity as CaCO-j 	
          Total Hardness as
          PH
          Organic Color ..
          Turbidity
          Type Clay
          Magnesium as CaCOo
                                Comments
    41
    71
     7.50
    15
Mobility of raw water -.96

*.5 mg/I Hal96
    50
Natural
    17

-------
                   TABLE 38

CALCULATED POTENTIAL PRODUCTION OF MfeCOo.SHoO BY
               20 AMERICAN CITIES, 1968   *
City
Indianapolis
Des Moines
Kansas City, Mo.
Kansas City, Kan.
Flint, Mich.
Lansing, Mich.
Minneapolis
St. Paul
St. Louis
Omaha
Cincinnati
Columbus
Oklahoma City
Fort Wayne , Ind .
Dayton, Ohio
San Diego
St. Louis Co. W.C.
Austin, Tex.
New Orleans
Wichita, Kan.

M.g. Water
treated
daily
80
30
98
32
v36
X23
70
46
190
68
112
38
*15
X25
x70
X50
60
X35
X120
35

MB++ in
untreated
water, ppm
22
33
18
21
25
33
13
14
14
23
10
.28
X31
20
33
29
17
19
11
14

Amount
removed
ppm
18
29.
14
17
21
23
9
10
10
19
6
.24
"27
16
24
25
13
15
7
10

Annual
production
of MgCOo.
3H20 tons
12,500
7,200
11,800
4,500
6,400
4,500
5,000
3,800
16,400
11,000
5,800
7,900
3,300
3,300
14,500
10,400
6,800
4,500
7,200
3,000
1-49 , 800
Annual gross
revenue at
2c per Ib.
$ 500,000
288,000
472,000
180,000
256,000
180,000
200,000
152,000
656,000
440,000
232,000
316,000
132,000
132,000
680,000
416,000
136,000
180,000
288,000
120,000
$5,956,000

-------
                              TABLE 3'gr (Continued)



                                    Comments

196711968

 Morse plant 54 mgd Mg+tl5 - 4 ~ 11   11 x"54 = 594
 Dublin plant 38 mgd MgT"^28 - 4 = 24   38 x 24 = 912
 Both plants 92 mgd Mg  20 -.3 =12   92x12=1104
 Preferred method would probably recover at Dublin only, and calculated on
   this basis.
 'Oklahoma City has three treatment plants
 Lake Overholzer - 6 mgd Mg,, = 29
 Lake Hefner     -15 mgd MgjT = 31
 Lake Draper     -18 mgd Mg   =  4
     So calculations based on recovery at Hefner plant only.

 St. Louis County Water Co. has four plants, 3 deriving water from the Missouri
   River, one from the Minimac River.
 Calculation based on largest plant only.

xExtrapolated from 1962 data.

-------
                     TABLE 39
              CALCULATIONS OF POTENTIAL
CONSUMPTION OF MgC03,3^0 BY WATER TREATMENT PLANTS
              IN THE UNITED STATES-
460 Plants  Serving  55,757.215 People  Use  Alum
 70 Plants  Serving  11,282.820 People  Use  Iron Salts
530 Plants  Serving  67,040.035 People  Use  Al or Fe


(67 x 106) people x (55 x 103) gal, yr. = 3.7 x 1012 gal/yr.

     3.700,000,000,000 gal. = 3,700,000 mg/yr,

25 ppm MgC03 . 3H20 = 200 Ibs/mg.

     15% make- up = 30 Ibs/mg,

3,700,000 x 30 = 111,000,000 Ibs.

     Ill ,000,000 Ibs.
                      = 55,000 tons/yr.
         2000
3400 filtration plants below 4 mgd (25,000 people)

     25%  Treat  3 mgd  = 2600 nigd
     2570  Treat  2 mgd = 1700 mgd
     50%  Treat  1 mgd = 1700 mgd
              Total     670W mgd

6,000 rogd x 20 Ibs/mg x 365 days = 43,800,000 Ibs.

     43,800,000 Ibs.
                    = 22,000 tons/yr.
          2000
Total of all plants 77,000 tons/yr,  1963 basis

Corrected for 1971 production - 102,000 tons/yr.


*From 1964 USPHS summary of municipal water facilities  in
Communities of 25,000 or more.
                            106

-------
1
Access/on Number
w
5

r. Subject Field & Group
05D
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Organization
Department of Public Utilities, City of Gainsville, P.O. Box 490,
Gainsville, Florida
     r;<;e
     Magnesium Carbonate, A Recycled Coagulant  for Water Treatment
] Q Authors)
Dr. A. P. Black
Dr. C. G. Thompson
16

21
Project Designation
EPA Grant Project 12120 ESW
6/71
Note
 22
     Citation
 23
   Descriptors (Starred First)

   Water Purification, Coagulation,  Chemical Precipitation, Sludge Treatment,
   Industrial Wastes.
 25
     identifiers (Starred First)
 27
   Abstract
   An entirely new system of water  treatment has been developed.  It is a unique
   combination of softening and  conventional coagulation, and may be used for all
   types of waters, surface or ground,  hard or soft. Magnesium carbonate is used as
   the coagulant and lime is added  to precipitate gelatinous MgCOH)-, which is as
   effective as alum for the removal of both turbidity and organic color.  The floes
   formed are larger and heavier than alum floe, since they are "loaded" with CaCO^.
   The sludge, composed of a slurry of  CaCOo, Mg(OH)2 and clay is carbonated with
   C02 and the Mg(HO)2 selectively  and  completely dissolved as the bicarbonate.  The
   carbonated .slurry is filtered and the filtrate recycled to the point of addition
   of chemicals to the raw water and reprecipitated with lime thus recovering both
   the coagulant and the sludge  water.   In small plants, the filter cake of CaGO,
   can be used as land fill.  In larger plants, it is slurried in a flotation cell
   and the clay separated and used  as land fill.  The purified CaCO~ is filtered and
   the cake passes to a multiple hearth furnace or Kiln and calcined to high quality
   lime.  Chemical treatment costs  are  substantially reduced for most waters and the
   quality of the treated water  is  superior to that treated with alum.

   This report was submitted in  fulfillment of Water Quality Research Grant Project
   12120 ESW with the Environmental Protection Agency.
Abstractor
	C.  G.  Thompson
                             Institution
                                Department of Public Utilities
                             SENO WITH COPY OP DOCUMENT. TO: WATER RESOURCES SCIENTIFIC 1NFORMAT tOWCENT ER
                                                       U.S. DEPARTMENT OF THE INTERIOR
                                                       WASHINGTON. D. C. 20240

                                                                               * CPO: 1070-369*930
WR:IO2 (REV  JULY 1969)
WRSIC

-------
ENVIRONMENTAL PROTECTION AGENCY
   Publications Distribution  Section
   Route 8,  Box 116, Hwy. 70, West
   Raleigh, North  Carolina  27607
     POSTAGE AND FEES PAID
ENVIRONMENTAL  PROTECTION AGENCY
        Official Business
                                        If your address is incorrect, please change on the above label;
                                        tear off; and return to the above address.
                                        If you do not desire to continue receiving this technical report
                                        series, CHECK HERE LJ ; tear off label, and return it to the
                                        above address.

-------