Composition and treatment of Lake Michigan water

BIRDSALL

|||

Composition and Treatment

Of Lake Michigan Water

.

,».

v. .

.

.'f'''i'''if,

'«l! ';

. a '«

l|t£:

Chemistry

AM I

"••'»'«!•>HQ *A. iH.

19 09

-

<-V: J' * *'JK " * . ft

fillip*

COMPOSITION AND TREATMENT OF

LAKE MICHIGAN WATER

BY

LEWIS ISAAC BIRDSALL

A. B. Williams College, 1907

THESIS

Submitted in Partial Fulfillment of the Requirements for

Degree of

MASTER OF ARTS

IN CHEMISTRY

IN

THE GRADUATE SCHOOL

OF THE

UNIVERSITY OF ILLINOIS

1909

UNIVERSITY OF ILLINOIS

THE GRADUATE SCHOOL

June 1 i9q9

1 HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPERVISION BY

LEWIS ISAAC BIRDSALL. A....B. ttHJJAMS COLLEGE, 19C7.

ENTITLED G..OI3E.Q.S.If.lCH AM I.EEATME150 CF

LAKE MICHIGAN '.'.'ATER

BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE

DEGREE OF MASTEE Of AB2.S. I.TT C.HEMIS.TEY

In Charge of Major Work

I Head of Department

Recommendation concurred in:

Committee

on

Final Examination

14504-1.

Digitized by the Internet Archive

in 2013

http://archive.org/details/compositiontreatOObird

COMPOSITION AED TREATMENT CF LAKE MICHIGAN WATER

.

IHTRODUCTIOH.

The Great Lakes contain an inexhaustible supply of water

which is available for drinking purposes and for domestic use "by mor^

i

than one-hundred cities in the United States and Canada having

a population exceeding four million.

These lakes also serve as a receptacle for the sewage of

the cities located on their shores. Investigations have sometimes

shown that certain cities which are pumping their sewage into the

lake through one pipe, are at times pumping the same hack into the

city water mains through another pipe.

This pollution of the lake water if continued will

necessitate the purification of the water siipply of every lake

city before the water will be safe for drinking purposes. At the

present time only the very largest cities are able to extend their

intake pipes far enough into the lake to obtain a water that is

safe for drinking.

Dr. J. F. Biehn of the Chicago Board of Health Labora-

tories has shown that with southerly winds the zone* of constant

*Lake Michigan Water Commission Report. Water Supply and the

Condition of the Water of Lake Michigan from the Calumet River to

Howard Avenue, Chicago, by Dr. J. F. Biehn.

pollution of Lake Michigan near Chicago reaches beyond the three

mile line in the Calumet region, then recedes to less than one

mile from shore at Sixty-First street, and then out again between

Twelfth and Twenty-Second streets to a distance of nearly two

***** tor

O

miles. The zone of occasional pollution was found to be fairly

constant and extended out about three miles. The two mile cribs

were constantly surrounded by a zone of water from which over

6C per cent of the samples showed the presence of Colon Bacilli.

The smaller cities which must of necessity have their intakes much

nearer the shore frequently obtain a more or less polluted water.

Even large lake cities such as Cleveland and Chicago

ought to purify their drinking water as the water which they

supply is usually more or less turbid. A turbid water may not

contain typhoid germs but it is not attractive for drinking

purposes and may even be harmful. The truth of this latter state-

ment has been demonstrated at Fort Sheridan, Illinois. There it

has been found that acute attacks of diarrhoeal diseases occur

among the soldiers when the raw lake water which they drink is

turbid.*

*?.eport Lake Michigan V.'ater Commission. Condition of Lake Michigan

Bordering on Illinois ITorth of Chicago, by Dr. Edward Bartow.

It has been proven by experience that highly polluted

waters may be purified and made safe for drinking. Purification

may be accomplished either by slow sand filtration or by the more

rapid method commonly known as mechanical filtration. In the first

method the polluted water is allowed to flow by gravity through a

bed of fine sand, the suspended particles and bacteria in the

water being removed by the sand and the gelatinous film formed on

top of the sand layer. Mechanical filtration differs from the slow

sand method in that certain chemicals such as aluminium sulphate

or ferrcus sulphate are added to the water and a floe formed in

the water "before it is run onto the filter. The floe envelopes

any suspended particles or bacteria in the water and thus ensures

the retention on the filter of very fine particles. The filtration

can thus "be carried on at a much faster rate than is possible in

the slow sand filter. The amount and kind of chemicals required

for purification varies with the suspended matter, the color, the

amount and character of the mineral matter in solution in the water.

The character and amount of the mineral matter is most

important "because of the chemical action between it and the chem-

icals added. We have, therefore, collected all the available data

concerning the mineral content of the great Lakes.

The mineral content of the water from all of the Great

Lakes with the exception of Lake Superior is of a similar nature.

Owing probably to the crystalline character of the rocks in the

drainage area of Lake Superior, this lake has very low mineral

content. The comparative composition of the mineral residue from

the waters of the Great Lakes has been determined by R. B. Dole

and M. G. Roberts of the United States Geological Survey.* Each

* "The waters of the Great Lakes", R. B. Dole. Paper presented

before the American Public Health Association at Winnipeg, Manitoba,

August 19C8.

analysis in the table represents the average analysis of twelve

monthly samples collected during 1906-07. In every case the

samples were collected at the outlet of the lake so as to give a

representative analysis. The results of the analysis are given in

ionic form and the amounts are expressed in parts per million.

Source

Turbidity

Silica SiOs

Iron Fe

Calcium Ca

Magnesium Mg

Sodium ITa

and Potassium K

Carbonate radicle CC 3

Bicarbonate radicle IJCO3

Sulphate radicle SO4

titrate radicle TT0 3

Chlorine CI

Total solids

TABLE I.

ions.

Lake Lake Lake Lake St. LawrenceSuperior Huron Michigan Erie River

2 trace trace 41 45

7.4 12.0 10.0 5.9 6.6

0.06 0.04 0.C4 0.07 0.0J

13.0 24.0 25o0 31.0 31.0

3.1 7.0 8.2 7.6 7.2

KJ —

v

4 .4 4.7 6.5 6.3

0.0 1.8 2.9 3.1 2.9

56.0 1C0.0 112.0 114.

C

116.0

2.1 6.2 7.2 13.0 12.0

0.5 0.4 0.3 0.3 0.3

1.1 2.6 2.7 8.7 7.7

60.0 108.0 118.0 133.0 134.0

~5~

© CM^CMtCnC-HHCO-'vjJo UOII'BD O I> <X> <o en lO O tO tO

g U J-Q& sux^jf) ... i

U ©

CO

ÔH^f) o co H J co cn to o to <r>

jacT sut^JCQ . • • i

© ©

^ noxiITPI * * °i )

H H H C- tO

j> H H COto cn to o to• . • .

en tO H• . . • •

H tO tOH H t>

tO H HCO cn to O to

• . . • .

CO H• . .

tOH H

ttOTT'Bf) W to O ) H t> ^ H co cnC t O W tO i h to tO o to «o

<D 60 H tO tO

M -H03,3 _

co ^5 noTiii:: *!

« °. t H. °. .

g « XB6 s^J co , H g g o to

M HFt

o3 m uoil^j o ca ^ i o to to o co

pq © p; joe! sutbjcs • (

w g w o H tO <0

O o3 !h

C£J W LOtOO lOtOOHOUOITTITI • • * I

§ < -* <= •h s ° a°

Eh M HEH

K — , _ ^ o co c- i cm cn H to ^£ h UOIIÔ o H H H JvOCOO^^

I

g Jed Sa^d . • • •I o M ' *-

#

OH to to

C3

Oo o

H CO No O w «5

52? a o3

<8 n©

© ©© pP •H o5 PJ

o3 ^H oU O Pn-P H H h•H pi o3

£h O CO o

•H •H •H •H-d nd

O o O oCO CO CO CO

1<

oCO

©Po3

^5O)

•H02

©

6c3

CO

oo CO

ta ooC3 CO

© o o-P on3 © ©Fl too 03

©u o pc3 03 CO

O Pi N <do3 o O *Ho rp •H r—

i

CO O•H S ct5

03

CO

CQ

© «ri o HO Pi •H 03

Go H o H •Pc3 03 u H o

O H CO EH

[

I

t

[

I

I

I

I

I

I

(

-6-

When the ions in Table I are calculated into hypothetical

combinations according to the method used in the Laboratories of the

State Water Survey we obtain the mineral content of the water as

shown in Table I fa).

It is readily seen that all of the lake waters are

essentially carbonate waters containing chiefly magnesium and

calcium carbonate. Therefore a treatment found to be successful

in purifying the water from one lake would apply also to the other

lakes with the possible exception of Lake Superior. The amount of

chemicals added in each case would vary with the total amount of

mineral matter present and also with the turbidity of the water at

any given time, but the treatment would be essentially the same.

Because of the increasing pollution of the Great Lakes

and the growing need for purification of the water, we have, in the

Laboratories of the State Water Survey, undertaken a study of the

effect produced by different chemicals when added to Lake Michigan

water, and the smallest amounts of chemicals which may be used to

obtain a satisfactory filtered water. Although cur experiments

have been confined to water from Lake Michigan, we believe that any

results which we may have obtained will be applicable to those other

lakes which contain water similar to that of Lake Michigan.

CCMPOSITICII OF LA1CE MICHIGAU 77ATER.

Before a water can be chemically treated with any degree

of success some idea must be gained concerning the mineral content

of the water and also concerning the extent to which the water has

been polluted. These two factors determine largely the method of

treatment which is to be applied.

-7-

TABLE II.

ions.

Analysis lrzO 5

Samples from Chicago Chicago Chicago {jD 1 cago

Analvsed "bv State77ater

StateV.'ater

C «B ,&Q

.

E.R.C »B .&Q '•

R.H.c.&s.w.

R.R.Survey

.

Survey

Potassium K and5.6 8.3 3.1 3.8 3.1

Sodium Ha

Magnesium Kg 11. V 10 .9 10 .9 10 »y

Calcium Ca 36.2 28 .2 rz A o£ . J.

Oxide of Iron Feg035.0 5.0 .3

and Alumina AI2C3

Hitrate ITO3 2,4 1.0

Chloride CI 4.8 4.2 2 .4f

c .6

Sulphate S0 4 12.0 10 .0 ( .1 . JLA ft

Silica Si0 2 6» 7 1.9 5.3 5.1

-3-

TABLE II.

IOBS

.

Analysis

Sample from

Analysed, by

Milwau-kee, WisM.&St.P.

R.R.

7 8

Btllwau- Milwau-kee,Wis. kee,7.r isM.&St.f. Es&St.P

R.R. R.R.

9 10

Chicago SouthChicago

M.&St.P. B. &R.R. lU&i

Aver-age

Potassium E>

and Sodium Ha

Magnesium Mg

Calcium Ca

Oxide of Iron Fe20 3 +

and Alumina AI2O3

Tlitrate IT0 3

Chloride CI

Sulphate SC4

Silica Si0 2

3.5

10.3

31.8

1.5

6 .6

10.6

31.9

4.8

11.1

34.0

4.5

4.5 £.6 4.5

10.7 9.9 10.8

34.9 35.3 32.9

6.2 3.1

2.9

6.9

7.2

9.1

3.1

8.1

2.7

9.4

4.1

9.5

3.6

1.7

3.6

8.7

4.8

-9-

TTrt-TT-D^ CM tj> tO r-T CO W O COUOTXat) wwwoo o toe-

.iad suft?j{)CO

Too o> ^ o w to H

.iad sijBd to ^ ^ ^ o m ^to co to

„ io co i> in O cr> i oUOlIîJ cvj to W O £> CO j cn

jcad sut^js • ••

CO ^

UOITTlLi to lo to to to o i

* ••••• •I

•

jad sajrej ^ to ^ ^ o to i toto oo i to

jad sut'BJif)

to

g uoxiiik

2 -iad sq.i'BdL

EH

M g UOH133S .iad sut«j:so

2 g uoxitTJleh m J ad s^j^cT

EH

P jcad sui"8j:o

to £> (7> o CO H Oco CO CI c- CO to H• • • •

co 00

o O to o to o• » • • • • • •

o LO LOto CO to

H

00 O to 1 o CO £>

o 1 co o CO• • •

1

1

• • • • •

CO

o J O to• • •

J• • • • •

H to to 1 CO oi—

i

to £> toH

tO o> IG to CO to CO

H CM to CO C2 CO• • • • • # • • •

H LO

to o CO CO IQ o CO• • • • • • » • •

to LQ o to o LO oH to H o

CO

o02 O

o £0 oCO H CO a5 +o O CO o CO

g9 <D P O% P aJ a? wS3 0} Fl p o

0) CD

Pio o5 Ph

CD •d P Fl-p •H o3 H Jh o F3

aJ rfl d a5 oU O ft CO c> U Jh

-P H rH cd M•H d

1B o

4-1^4 O CO pi

•H •rH OCO CO ps

1 CD o •rH

•H •H ri O•d -d «d to to H •HO o o aJ a3 a5

00 CO CO fcr*

F1^ O O

C5

oH

CO

a5 w <xi

p; o th•H «H rHE CO Opl CO

rH 05«aj O rH

•H a5

»d H PC -h o05 00 EH

-10-<D tO ^ CT> O O Cn rH CQbQ no IT'S hcowô>c- CM CM CO

g jocT sutbj:{) • • * * J J '*

> uoixiTI'I co cn cn co to H ^> coo

oH

iioirec I

.19(1 SUTBJf) ' w » ui 1 ^

.lad so.j'Ba i to i H <o co to toH CM CO I CO

rH

UOH^f} tO CO 00 O CO vO I toJ9d SUI'B.If) i w CM O O tO I ^

I

a> CM LQ CO

noflTTH i to to co h w i

j.9d aq.jc'Bji ^ co ^ to e> ^> i **

P2 jcad sui^ô i co m h h cn w i to$25

rwl- •••••• I*

CO CM ^ CO1 I

eh ÔITITKi h to cn ^ cn 10 i

< j.q& 3q.jâ: i • * * • • j i si£3I LQ CO CM tO ^ I ^M, tO 00 1 H

pqM £!M o _ftTT„„ i o> «4* C- 00 LO ! I CO

o uotlwji

iq co co co to i i cnh jad sui^jif) i

|3 m nopiTTN1

EHo

w jed sult3.i£)

cn o o cn rH

CO CM cn CM• • • • • •

Hcn CO to H tO

• • • • •

io tO CM COCO CO

CO 1 cn CM ^« 00

CO 1 to LQ H•

1• • • •

rH LQ

to 1 cn o H H1 • • • •

tO I H to CO CO

H CM CO

tO CO CO o 00 tOCv2 CM o o to• • • • • •

CM

LQ to CO tO H CM• • • • • •

co to tOto CO

o o CM LQ tO

CO lq 1-1 H cn CM• • • • • •

CM

to cn cn LQ• • • • • •

lQ CO CvJ tOto 00

cn c- CO LQ

tO CO CO CO tO j

• • • • •iH I

cn cn CM tO i

• • • •I

__J io tO CM cn I

H CO i

CO to O CM cn

CM CM cn tO O• • • • • •

Hco lQ tO LIJ

• • • • • •

CM cn Hto o

CO

<* o

^ CO

H

O CO

•H •H •H

O o oCO CO CO

•H "H Sw to pQ) 0) <H

£»D m HJtJ o5 cS

a a o

I HI cn

I

I t—

1

I o

1 toI CO

TTrt-r-rr>r\ I CO <J» to <-» w w« iU0TTBD CO CM CM cn O O 1 Î

tOj co cn lO to to i

tLL

- ^ ^ ^ cm cn H i coJiod

I , co c~ i cmH

O co

01 &0 Oo qo a o

co H CO a «5 +o o w O O CO CO

aJ 03 a> 4» O O^^^HJCJCD WWH <3G-P<DH

Q CD rC O 6) ^ Sdj rcj 43 a, ,a £+5 -i-t c5 H Jh od ^ ^ P5 aJ rQ^ o Ph CO O fn

4-> H H ra

•H ,C pi S S o

o CQ

u aJ M rb1—

1

o •H•H •H rH

<r-( CO OO 1H aJ

C2

O <H o H<d •H a3

•H H +>

M •H OO a5 CO EH

I

1

I

t

I

I

I

I

-11-

The mineral content of Lake Michigan water at the south-

ern end of the lake and near enough to shore to he avallahle for use

"by city v/ater works is given in Table II. This tahle shows ten

independent analyses of v/ater from the southern end of Lake Michigan

and the average must represent quite well the composition of the

water

.

If we take the analysis of Lake Michigan water, Table I,

as showing the normal mineral content of the lake and compare with

it Tahle II we may observe an increase in the mineral content in

Table II, the amount of calcium and magnesium especially being

greater. This increase may be due in part to pollution of the water,

but it is probably due for the most part to the stirring up by winds

and currents of the shallow .water in the southern end of the lake

and near the shore.

TURBIDITY OF LAKE MICHIGAN TCATEP

.

The turbidity of water is usually caused by clay, silt,

finely divided organic matter and microscopic organisms, which are

held in suspension in the water.

The turbidity of Lake Michigan water at a distance of

one mile or more from shore is chiefly due to finely divided silica.

The seasonal variation in the turbidity of the Lake is

well shewn in Table III. The determinations of turMdity were made

by the State Tater Survey upon samples regularly collected

from the intake pipes of the city water works at Lake Forest and

Evanston, Illinois. The amount of turbidity is given in terms of

parts per million of silica.

-12-

TABLE III.

June 3, 19C7Jjctivt? r v I c o u

40XJGk UC

June 4, 1907i A V Lk.) U \J XJ.

5

July 1, 1907 20 June 25, 1907 8

July 29, 1907 15 August 26, 19 C 7 20

August 26, 19 C7 5 September 9, 1907 5

October 14, 1907 25 September 30, 1907 40

January 6, 1908 50 October 14, 19C7 30

January 27, 19 08 170 November 11, 190

7

30

March 9, 1908 100 January 13, 1908 180

April 6, 1908 60 March 2, 1908 40

April 27, 1908 80 April 6, 1908 70

May 11, 1908 9C May 4, 1908 30

June 15, 1908 10 May 25, 1908 35

July 6, 1906 5 July 20, 1908 20

August 10, 1908 3 August 17, 1908 20

September 8, 19C8 5 Hovember 30, 1908 20

October 13, 1908 30 December 14, 1908 55

January 4, 1909 2 Februarjr 1909 200

January 18, 1909 3C March 30, 1909 20

February 8, 19 C

9

110

March 8, 19 09 75

April 12, 190 9 60

May 10, 19C9 30

-13- •

As the samples were net collected at "both places the same

day the turbidity at the two places for any one month may vary

considerably. From the table it may be seen however, that in a

general way the turbidity is greatest during the winter months and

least during the summer months because of the change in seasonal and

climatic conditions. This variation in the turbidity of the water

would largely influence the amounts of chemicals added to the lake

water in order to get the best results at minimum cost for purifi-

cation.

ALKALIITITY CF LAKE MICHIGAN WAT&R.

The alkalinity of a water is usually caused by the car-

bonates and bicarbonates of calcium and magnesium and in some lo-

calities also by the carbonates of sodium and potassium. Unlike a

river water, the alkalinity of which is constantly changing with

every change in the level of the river, the alkalinity of the water

in large lakes remains fairly constant. The alkalinity for Lake

Michigan water covering a period of sixteen months is given in

Table IV. (p. 14)

The alkalinity of a water g°verns the amounts of certain

chemicals that may safely be added to the water without fear of un-

decomposed chemicals remaining in the filtered water. If the

alkalinity were suddenly decreased, as in the case of a river water

during a freshet, some base such as lime or sodium carbonate

would be needed to decompose the sulphates.

The alkalinity of Lake Michigan water is fairly constant

and, therefore, the chemical treatment found to be satisfactory

would not have to be changed for fear of undeconposed chemicals in

TABLE IV

January 1906

-14

Lake Forest116

jLvans ton128

vi mne ujia

126AY ti I cig

123

February19 08 130 130 130

March 1908 124 124 124 124

April 1908 124 124 138 132

May 1908

June 1908

126

112

120118

113

112

119

112

July 1908 122 114 114 116

August 19 C

8

110 116 113

September 19C8 112 109 110

October 1908 118 . 122 120

November 1908 112 120 116

December 1908 124 124 124

January 1909 • 126 130 116 124

February 19C9 122 122

I.Iarch 19 C 9 117 121 119 119

-15-

the filtered water.

POLLUTION OF LAKE MICHIGAN.

The two great problems which every growing city has to

solve for itself are the disposal of its sewage and the securing of

a satisfactory water supply. These two problems are intimately con-

nected in the case of the cities located on Lake Michigan and too

large to obtain their supply of water from underground sources.

These cities naturally turn to the lake for their water supply.

They also pour their sewage into the lake as it offers the easiest

method of sewage disposal. 77hile the cities were still small and

far apart there was little objection to this arrangement; but the

cities have grown and have multiplied in number until now at the

southwestern end of the lake they form an almost continuous front

for miles along the lake shore. The amount of sewage which they

daily ncur into the lake is enormous. Come idea of the amount of

raw sewage daily entering Lake Michigan may be gained from Table V.

TABLE V.Gallons of Sewage Daily. (Estimated)

Fort Sheridan 400,000

Waukegan 1,000,000

Corn Products Refining Co.at VJaukegan 7,500,000

Evanston 5,000,000

Hammond and Whiting, Ind. 1,500,000

Indiana Harbor 1,000,000

Standard Oil Co. Uniting. 18,000,000

-16-

Formerly where the city water works intake and the sewer

outfall were far apart there was little fear of pollution because of

the great dilution of the sewage which entered the lake. How, how-

ever, with the increase in the number of cities and in the amount of

sewage pollution, a lake city is very liable at times to drink jts

own sewage and that of neighboring cities in a more or less diluted

state, the pollution varying with the direction and velocity of the

lake currents

.

There has "been a popular conception that the waters of

Lake Michigan move southward along the western shore, across the

southern end of the lake, and northward along the eastern shore.

This idea of a general lake current has led to the location, of water

intakes and sewer outfalls similar to their relative location on a

flowing stream. That this idea is a wrong one has been shown by

pollution of the city water when the wind blows in certain directions

Milwaukee*, Wisconsin has this arrangement of intake and sewer out-

*Report of the Lake Michigan Water Commission. Report of Milwaukee

City Water Supply, by Major V/. V. Judson.

fall and a t times the city's sewage is carried towards the water

intake

.

Major W. V. Judson*, Corps of Engineers of the War

Report of the Lake Michigan Water Commission. Currents in Lake

Michigan, by Major 7.T

. V. Judson.

Department, has made a careful investigation of the currents in Lake

Michigan and has concluded that the currents found in the lake are

entirely local in their scope and are due to the direct influence of

-17-

winds and atmospheric pressure

.

Barnard and Brewster in their study of the character of

the water in the southern end of Lake Michigan* have corroborated the

* Report of the Lake Michigan Water Commission. The Sanitary Con-

dition of the Southern End of Lake Michigan, ty H. 2. Barnard and

J. H. Brewster.

statement of Major Judson, They found that when the wind was blowing

either from or towards the shore, the water from the body of the lake

tended to move towards the land either below or at the surface while

the polluted shore water .because of counter-currents .washed outward,

Barnard and Brewster also noted by observing the color of the sewage

in the lake and by bacteriological tests that with changing winds,

the sewage from the Standard Oil Works at Whiting, Indiana, visited

at times the water works intakes of Whiting, Hammond and Indiana

Harbor

.

The city of Chicago previous to the opening of the

drainage canal emptied all of the city's sewage into Lake Michigan

and also took its water supply from the lake. The typhoid fever

death rate in Chicago became so great that something had to be dene.

The drainage canal was constructed and the death rate declined. At

the preae-nt time only a very small amount of Chicago's sewage enters

Lake Michigan. Although Chicago's water supply is not liable to

pollution from Chicago's sewage, it is subject to pollution from

other sources.

The shipping in Lake Michigan constitutes a source of

danger which it is impossible to regulate. Dejecta from a typhoid

fever patient may be discharged into the lake near the water intake.

r

-18-

Prcfesscr William T. Sedgewiek* has pointed out that Asiatic

* Lecture delivered at University of Illinois, April 22, 1909.

cholera is liable at any time to appear in Chicago due to infection

of the water supply in this manner.

The dumping of dredgings and garbage into the lake offers

another source of pollution.

The cities both north and south of Chicago are frequently

aiding in the pollution of Chicago water. Barnard and Brewster

have traced the sewage from the Calumet river as far north as the

Hyde Park intake of Chicago's water supply.

Comparative sanitary and bacteriological analyses* of

* Report of the Lake Michigan Water Commission.

water collected from Lake Michigan twelve and one-half miles out

from the mouth of the Chicago river, one mile out, and from the

Chicago river at the foot of Madison Street serve to show the degree

of pollution of the originally pure lake water as one approaches

the Chicago shore. Each analysis in Table IV is the average of five

analyses made in five different laboratories from samples collected

and shipped at the same time. The results are given in terms of

parts per million.

-19-

Turbidity

Color

Odor

Total residue

Chlorine

Oxygen Consumed

I as Free Ammonia

II as Albuminoid Ammonia

II as nitrites

IT as ITitrates

Alkalinity

Hardness

Bacteria per c . o .

Color Bacilli

TABLE VI

.

Chicago river1£ l/2 miles 1 mile foot offrom shore from shore Madison Street

2 3 27

1 4 12

C Musty

153 147 180

3.5 3.6 5.1

1.8 2.5 3.7

.014 .015 .091

.074 .088 .195

.001 .001 .002

.080 .13 .16

114. 115. 118.

121. 124. 127.

32. 792. 32,46C.

Absent A few tests All testspositive positive

-20-

George C. Whipple* speaking of Chicago's water supply

* G. C. Whipple. Typhoid Fever, p. 166.

says :

"Recently typhoid death-rate in Chicago has been lower

than formerly hut from the nature of the sanitary conditions along

the water front and the inevitable contamination of the water result-

ing from shipping and other causes there is no reason to expect that

the city will ever have a permanently low typhoid death-rate until

the v/ater is filtered".

To quote from the summary of the report by Barnard and

Brewster, "the chemical and bacteriological survey of the southern

portion of Lake Michigan adjoining Lake County (Indiana) shows the

water of the lake to be grossly polluted and unfit for use as a

source of water supply for drinking or domestic purposes".

"The laying of intakes further into the lake will not

provide an adequate protection against impure water, since the zone

of pollution extends more than five miles from shore".

TABLE VII

.

TYPHOID FEVER DEATH PATE PEP. 100, CCO

FCP CITIES Oil THE SOUTHER! SHORE OP LAKE MICHIGAN

.

City 1903 1904 1905 1906 1907

Hammond 140. 81. 66. 70. 66.

Whiting 55. 99. 36. 00

East Chicago 55. 13. 13. 160.

Michigan City 75. 78. 42. 48. 42.

Chicago 32.1 20.2 16.5 18.3

-21-

Enough has already "been said to indicate the extent of

the pollution of Lake Michigan and to show the impracticability of

extending the water intakes further into the lake in search of pure

water. The time has come when Lake Michigan water must be purified

"before it is used for drinking purposes.

PURIFICATION OF LAKE MICHI GAIT WATER.

There are various natural agencies at work in Lake

Michigan which tend to self- purification of the water. The first of

these is dilution of the sewage by the large amount of lake water.

This not only lessens the possibility of pathogenic germs reaching

a given water intake but it also aids in destroying the bacteria

and in making decomposable organic matter harmless.

Sedimentation also plays an important part. The heavier

particles in the sewage sink quickly to the bottom of the lake while

the less heavy particles sink more slowly. Where there is storage of

a water for long periods of time most of the suspended matter in the

water settles out. The shallowness of Lake Michigan, however, and

constant agitation of the water by winds and storms keeps the water

near the shore more or less turbid at all times.

Certain organisms in the water and harmless in themselves

may help to destroy the pathogenic bacteria*

The typhoid bacillus does net exist indefinitely in

lake water. Whipple* has shown that only about one per cent exist

*G. C. Whipple. Typhoid Fever, p. 48.

after one month while a few resistant bacilli remain for a much

longer time.

-22-

At the surface of the water sunlight acts as a purifying

agent. Its effect, however, is very slight "beyond a few feet below

the surface

.

These natural agencies might purify the water in time

provided that the amount of pollution was limited. The amount of

pollution, however, is constantly increasing and these purifying

agents are not able to accomplish their purpose. We must, therefore,

resort to artificial purification of the water.

PURIFICATION BY FILTRATION.

The first method of water purification to be developed

and now being used with success in many cities is that known as slow

sand filtration.

In brief the slow sand filter is composed of three

essential parts, the top or sand layer, the gravel layer, and the

under drains. The sand layer usually composed of fine sea sand is

from thirty to forty inches in thickness. The sand grains may be *

Allen Hazen*. Filtration of Public Water Supplies, p. 20.

from 0.013 to 0.C4C of an inch in diameter. Particles suspended in

the water to be filtered are caught by the sand grains and held at

the surface of the sand layer. Thus the water first forms at the

surface of the filter a gelatinous layer called by the Germans,

"Schmutzdecke", through which the water passes before reaching the

sand. It is this "Schmutzdecke" which decs the efficient part of

the filtration .llext below the sand is the gravel layer which is

usually two feet in thickness. The layers of gravel are so graded

in size of grain that the sand will not sink down into the coarse

-23-

gravel "below.

The under drains may be made of "brick or of tile. They

are laid at regular intervals "below the gravel layer and serve to

carry off the filtered water.

The depth of water on a slow sand filter may vary in

different localities. It is often from 36 to 52 inches.

In the slow sand filter the water sinks through the sand

due to the force of gravity alone. Thus the rate of filtration is

often slow especially as the filter becomes clogged.

The large area required to afford ample room for slow

sand filters and the accompanying sedimentation basins .provided the

water to be filtered is muddy or turbid, makes this method of puri-

fication often prohibitive because of the great expense involved.

In cold climates unless the filters are covered, the ice

formed on the filters during the winter months is a source of much

trouble and expense. Also unless the filters are covered algae

growths may cause trouble. The cost* of building covers over filters

* Allen Hazen. Filtration of Public T.7ater Supplies, p. 16.

has been estimated at more than 50% of the original cost of the

filters. •

In order to clean a slow sand filter that has become

clogged the filter must be put out of commission while a thin layer

of the topmost sand is being removed. This method of cleaning the

filters requires extra filtering area so that while one filter is

being cleaned another may take its place.

-84-

W5CHAMCAL FILTEATIOH".

The second method of purification is commonly known as

mechanical filtration. This method differs from the slow sand

filtration method in that the water to he filtered is first treated

with some chemical or chemicals whereby a floe or gelatinous precip-

itate is formed. The v/ater is then allowed to stand for a short time

in a sedimentation basin so that the floe may form and settle as much

as possible. In settling, the floe drags down with it any matters

suspended in the water.

The mechanical filter may be circular or square in shape

and is only a few feet in diameter. Frequently the filter area does

not exceed 300 square feet. Such filters nay be arranged in groups

or batteries of any number desired.

In the mechanical filter the filtering material is com-

posed entirely of sand, no gravel being used. The underdrains con-

sist of pipes to which are attached strainers small enough to keep

the sand grains from passing.

Owing to the floe in the water a gelatinous film is soon

formed on top of the sand and this removes particles even smaller

than the bacteria in the water. The rate of filtration by a mechani-

cal filter is much more rapid than slow sand filtration.

The mechanical filter is cleaned much cftener and more

easily. In order to clean the sand in the mechanical filter the

flow of water is reversed and water forced up through the sand from

the strainers while the sand is thoroughly agitated either by rakes,

by the water itself or by compressed air. The wash water is wasted

as is also the water which passes through the filter immediately

-25-

after washing and until the gelatinous surface film has had an

opportunity to form.

The chemicals chiefly used for forming a floe in the

water "before it enters the filter are aluminium sulphate, either alone

or in combination with some "base, and ferrous sulphate in combination

with a "base . Because of its cheapness lime is the "base commonly used,

Experience has proven that the most turbid and most high-

ly polluted waters may "be purified satisfactorily "by mechanical fil-

tration. The number of cities in the United States now using

mechanical filtration is rapidly increasing and already as shown "by

Kazen* the number of cities having mechanical filtration plants is

* Allen Ilazen. Clean Water and How to get it. p. 80.

nearly three times the number of those which are using slow sand

filtration. We have, therefore, obtained the following experimental

data with a view to determining the best chenicals or combination of

chemicals for use in the mechanical purification of Lake Michigan

water, and also the smallest amounts of coagulants which might be

used to produce a satisfactory floe.

-26-

CHEMICAL TREATMENT CF LAKE MICHIGAN WATIR.

The water used in all of the experiments was obtained

from the Fourteenth Street Pumping Station of the Chicago city water

works

.

The lake water was first treated with a saturated solu-

tion of lime water containing in one cubic centimeter 1.11 milligrams

of calcium oxide, CaO. To 1000 cubic centimeters of the water

contained in clear glass stoppered bottles was added varying amounts

of lime as shown in Table VIII.

The effect produced by the lime is represented diagramat-

ically in Plate I. The hardness was determined by the soap method

and shows approximately the calcium and magnesium content of the

filtrates . The lime added first combines with the free carbon dioxide

in the water to form the soluble bicarbonate of lime as shown by the

rise in the hardness and the alkalinity to methyl orange. With the

addition of mere lime there is a decrease in the hardness and

alkalinity shewing that calcium and magnesium have been precipitated.

The minimum is reached when 4 l/E grains per gallon of lime are added.

Further addition of lime causes an increase in the hardness and

alkalinity in proportion to the amount of lime added.

The maximum treatment is indicated when the alkalinity to

phenolphthalein is equal to one-half the alkalinity to methyl orange.

The chemical reactions taking place may be shown as follows

;

Ca(0H) 2 2C0 2 = CaC0 3 H2C0 3

Ca(0H) 2 + CaC0 3 E 2C0 3 = 2CaC0 3 + 2H2

2Ca(0H) 2 MgCOa H»C03 = 2CaC0 3 * Mg(0H) 2 + 2H 2

The reaction with lime is almost immediate and the pre-

-27-

Pu o pO CD Ph O«H S -H

P 02 H•H +3 HMO fn-H

co o > d fiCO O Phcd

p«dP3W

op05

P-PrH•H

Pi

CD pPh O

•HCO H-P HPi -H«J sPh

LO •O to tO

to to CM O GO <X)

H H H H

LO CO CVJ tO Oto to O 00 tO

H rH H H

tO

LO

totO

lO

tOtO

tO

C\2

tOtO

•

M BMM o> PI

i aM<EH

Q-P

P)o cd

<H BO

P P O>>aJ o >P P O•H OP•H HH05

H CD

CP

p05

PPrH•Hft*

op

Po

HCD • HH (h O

>s «J O >P rP O•H PP•H OhH HCo O>H pH CO

Ph

uCD pPh O

•HCO HP HP *H05 SPh

hCD PPh O

•HCO rHP H.P -HaJ BPh

CD Pi

Ph O•H

a hP HP -H=5 BPi

CD PiP CD PSa5 Ph OPi •HP CO HH P H•H Pi -H

05 SPh

oH H a>

H

tO £>

rH

o

LOCO

otO LO

oLO

HLO

(HLO LO

toCO LO <vt< LO

o eoH H C\2 to

O COH

toH

LOCO

toto

CO pt) p o

CD CD -H p H aB *d 05 CD rH Cj

•H-d P Ph 05 Ohi 4 tt) M

coLO

tO LO to w o 00 LO to

H to C7> LO o to w 00• • • • • • • • •

H rH C*2 to LO LO

CD I

p B Pi »d o -h co

p -H CD CD «H -P pp! H -P "dp P CD

O a5 td P <D PgtH S 5) O O O<=} o B

o orH

oCM

oto

o oLO

oto

o oCO

o ocr> O

H

Pip a i

p -P rp o -h CO

p 05 <D *H -P pO «i CO p CD

B p pj op<3 Vl O O CD

o e

ooooooH

cvl

OooH

ooo

to ^

ooo

LO

ooorH

tO

OOo

oooH

CO

OooH

ooo

oH

OOOH

HH

r

i

-29-

cipitate settles quickly. Lime serves as a good reagent for softening

the lake water but the precipitate formed is too crystalline to serve

as a good coagulant.

Experiments were next tried using alum alone. Varying

amounts of alum were added as shown in Table IX. The alum used is a

commercial brand containing approximately 18% alumina. The solution

of alum used in all of the experiments contained 4.28 grams per liter

of aluminium sulphate so that one cubic centimeter of the solution

added to a liter of water was equivalent to one quarter of a grain per

gallon. As shown in Plate II, the alkalinity is decreased steadily by

the alum added but there is still enough alkalinity remaining to pre-

vent the presence of undecomposed alum in the filtered water. The

hardness remains practically the same in amount although the bicar-

bonates of clacium and magnesium are partly changed to sulphates by

the alum added. The permanent hardness of the water is thus increased

It is important in mechanical filtration to have the floe

form quickly and settle out readily as this means a saving in the

size of the precipitation and sedimentation basins. The length of

time required for the formation of a floe with alum alone is shewn

in the table as is also the relative length of time required for

sedimentation. Owing to the small cross section of the glass

cylinders used and the friction at the sides, the data for sedimen-

tation is only relative. A good floe was obtained in less than one

hour when 2 1/2 grains of alum were used.

The State Board of Health of Ohio* has carried out ex-

*~.7ater ana Sewage Purification in Ohio. State Board of Health

Report 19C8 . p 309

.

rd©

© fn

S -H fn

•H pi oEh cr"*-

(17

H

I

•H

CQ

) Pi

Pi O© -HS+>

c5

S•HEH

rd©h o S•H ^ O O ^£ O H -P O

h

-30- Pi•Ha

otO

W* CO CQ

H fH h

o CO COH

•

Pi•HSinH

• • •

CO CQ CQ

fn H fn

vO CM

•

Pi PI Fl

•H •H •H •H •Hgw

lO Q LOCM to CM CM H

CQ CQ CQ CQ CQ CQ •

H fn fn fn fn N Hr-< fl fl

CM CM H

Pi p; FJ f3

•H •H •H •H •H •H

PiaPI

aM c£h

O lo LO LO LO LOLO to to to H H

CQ

U

cm

• CO •

fn H ».p U I PiCQPtOHPtOHO© O J> ft S H

fn

Wi © CO

•H Ctf fn ©HOch p, aJ Ph-h -h

#» A H

iO OtO COH H

LO OtO <MH H

1 <DI

CM! H

I LOI CMI H

CM

1 LOI COI H

XI

r—i

pq

EH

OH

br-H

<3

o ©•P bO •

53 H>s aS fn

•P fn O•H O OJ3•H H31?PH ©

H »

O H<tH O

F*

I ©H P•H C3

«H fnP

CQ

-P fH •

^ CO I fl

o3 ftH oft -H -H

S H

CQ

P H I Pi

fH © rH O0} ft'H »Hft S H

COHrH

COrHH

HHOH

vOOH

OH

OOH

COC7>

LO

cn

CO

COCO

LOCO

00CO

LO£>

OPJ•H •

© fH

H fH

>s 03 OP ^ O•H -PPJrS•H ftrH H05 oA4 P3H ©< A

Ph

CO •

fH • P fn I PJ

O rH fn © rH O9H O 03 ft'H «H

> ft B H

I © CQ •

rH -P +» fn I PJ

•H 3j fn © rH OCh fn CtJ ft'H -HP ft S rH

o o o o o O O o o o o• * • • • •

CM CM CM

o o o o O O o o o o o• • • • • *

CM CM CM

E CQ

PH O'Hs

CO

f3—

fH O ^©HOcrj cq a} ftH co

o

HC>D

OS

LO o LO O LO a LO O LO• • • • • • • • • *

H H CM CM to LO LO

oS2S

CM to LO CO

I

(

I

I

I

I

I

I

1

periments with alum which tend tc shew the large amounts of alum

necessary to form a floe in clear waters. It was also found that

the addition of clay to a clear water causes a floe to form much

more readily and with smaller amounts of alum.

The use of lime with alum suggested itself as a means of

decreasing the amount of alum required to form a floe. nIe first

tried varying amounts of alum with equivalent amounts of lime.

Table X shows the amounts of lime and alum added, and the

results obtained are shown diagramatically in Plate III. By compar-

ing Table X with Table IX and Plate III with Plate II it is seen

that practically no advantage is gained by adding an equivalent

amount of lime with the alum. The only effect seems tc be an

increase in the hardness and somewhat less variation in the methyl

orange alkalinity. The amount of floe formed when the lime is added

is practically the same as when alum alone is used and there is

little difference in the time required for reaction and sedimentation

Alum with an excess of lime was next tried. The results

are given in Table XI.

The floe formed in a few minutes and settled rapidly.

The softening effect of the lime is well shown in Plate IV. The

amount of lime used is excessive and would probably result in the

depositing of lime on the sand grains of the filter. Such an

effect has been observed at Lorain, Chio*, when an excess of lime

*77ater and Sewage Purification in Ohio. State Board of Health

Report. 19C8. p 164.

was used with iron sulphate.

We next tried to find an amount of lime which when added

X

H

EH

MH

e

*dCD

CD t-i

e -h•H pi

Pi -HO T3

I PJ

C OCD -H

EH CfCH © S Po co ccjCD

Pi

d©

S -rl Vi O O h•H pi O HP OEH OhcH "tH ^

CD

Pi

• CO •

to fn Pi »p u I Piaôrih ohoA) 0<H O O ft-H -Hso > ft a

H

»dPi J <D co

03 H-P-P ^ •

W «H 05 Jh <D I £Ch Pi Q ftH O

+3 ft -H -r-l

E H

O CD

-P JsO •

PI P* fH

<tf h O-P 5m O•H O OC•H H IH^ H^ P «HH CD

to•4s h i fl

iH Pi CD H Oo g5 ft >h -h

t> ft EHCD COPP fH

c3 Pi CD I PJ

Pi a5 ftH OP ft iH -H

H

P!O -HPHaiP &

•H -PPJ ^•H ftH iHGj O

H ©<$,d

P-i

• CO •

Pi Pi • P Pi ! Pi

Pi O H Pi © H Oo a! ft«H ti

O > ft S H

I © CO •

H P P Pi I PJH aj Pi © H OCh fH 25 ft -h -HP ft EH

©• S -dP -H (D

rH CO

^ s

O

CO Pi

PI Pi o•H © H O25 ftH fig

Pi a5 O

PI P! pi Pi PJ

•H •H •H •H •Hm33- a a >H g

o O o O OCvJ CM to to

CO CO CO CO 03 CO CO CO 03 •

Pi Pi Pi Pi Pi Pi Pi

r^H & & A\J * rn tO LO to w H

H H

Pi Pi Pi Pi Pi Pi Pi

•H •H •H •H •H •H

S3S3S3

fl>-* a>—

i

* g

o O o o LO LO oLO to H H H

•

CO

•

CO • •

Pi Pi Pi

LO H H

CO CO I 1 cr» 1 o ! J

CVJ CvJ 1 to 1 CO I 1 1

__1Pi I 111 r-l 1 iH 1 1—

1

] rH 1

GO CO 1 LO 1 CO 1 CO J O 1

CvJ

Hcvi 1

1

to 1 to 1 to 1 1

H H 1 H I H 1 H 1

CO 00 O cn to W Cvi CVJ

H H H rH H O o O O O o.1

1 t—JII _

1

I 1,—1 f—\ r-| H H H

CO <o 00 o CO <o CVJ O O OH H H H H o o O O o OH H H H H H H H H H H

cvj

CvJ CV3

1

1

I

W lO co ,—

(

O to «o

H CXI CO LO l> cn o H CVJ

C5

S CO Pi'—• pi rd pi Pi O ^P rH CD «H CD H O

CtJ CO 03 ftH COd Pi cd—

tiD £i0 c-i

O H

O 1HPh

LO o LO LO o LO o LO o• • • • • • • •

H H CVJ CVJ to to LO

cvj tO LO tO CO oH

-34-

-35-•

Pi Pi

•H •H<d S daCP •

OCD Pi i i R o•H 5h 'H pj o H iH

•H ^ O O *HEH OiCH CP S P

CP 05 03 03•

h -P Pi

CM H

p; p; Pi Pi

•H •H •H •H

S B s S

o O o OH H

Pi

Pi PI P! Pi•H •H •H •Hfl s b s

O o o ost4

M

h!PQ

EH

OCOCO

Wo

«=1

H5

EHIH

wE-i

-=1

WEH

EH<

5

o

«*i

Hi

-H°

CD Pi O S9 -h Pi O O Pi

•Hp* O H -P OEH tftH'H *H

cu

Pi

• CO

^ Pi • PCO Pi O HEl O'H OCD OPIrduas

m

t> Ph

Pi I Pi

Pi CD H Ocrj ft-H «H

S H

I cp CO

H P P•H Gj H<H Pi 05 PVHP Ph e

P!

o•H

O CP

-P fc£ • CO

R M k '-P >h S>s (3 Pi O H Pi CD I OP !n O^H O c3 pi|H «H•H O O > Ph *H HPI B•H H I CP CO

H >3 H P P Pi I PI

cdrCj •Hrf^cPHO<H Pi c3 PVH h

H CP P Q^ S H

Pi

O «H • CO

-P CP Pi Pi • P Pi •

H Pi O H Pi CP |

ha! o aJ PhH op ,d o > Ph -h <h

S rH•HPi•HrHa)

MH<3

P,3

iHOPiOAPh

I CP CO •

rH P P Pi I Pi•h aJ Pi CP H O

a} pvh «hPiP Ph

ordCP

CO

Pi

S H

Pi oo ca

•H h H•P H CQ aJ CD H WS Cv5 pi U 03 r-J«4 hD ttu <J

pi•HSo

H

o

COto

to

LP

Pi•Hfi

to

Pi•HSCXI

C7>

to

Pi•HB

pi Pi Pi Pi pi Pi

•H •H •H •H •rl •Ht

s s B B B Pi

IP 10 t> 10

CO »CO 1 CO 1 LP

IP 11 COH

IP 1 1 to!

10UO 1

«£> 1 torH

vO to COH/ rH

to to to IP IP COHH

•

O•

O•

O• • • • *

O•

O•

C\J CVJ CM CXI CX2

O O O O O O O• • • • • • • • • •

CVJ CM <X2 CvJ

O IP O LP LCj IP1• • • • • • • • •

H H to to LP

<HO

cp idB o

PI

o oM rH 05P -H 03 3! OHO

H p! Pi P< aJ

b0 bO

o

LP rH to CX> LP H toO CM LP O O to 00

• • • • • • • • • •

to to H^ LP LP LO LP

H CVJ to £> CO cr> O 1H=£9=

-37-

with the alum would cause a slight softening effect and yield a

satisfactory floe which would settle quickly. Table ZII shows the

results when a constant amount of alum is used with varying amounts

of lime. A satisfactory floe was obtained with 1.5 grains of alum

and 1.99 grains of lime. The floe formed in one-half hour and

settled readily. The amount of floe obtained was practically

equivalent to that obtained with 2 l/2 grains of alum when alum alone

was used. Owing to the cheapness of lime in comparison with alum,

there would be quite a saving in the cost for chemicals used. If

bulk lime at a cost of $4.50 per ton is used and alum at a cost of

$25. CO per ton, we can compare the cost of the two treatments as

follows

:

TABLE XIII.Cost Cost Total

Al 2 fS04) 3 CaO eostGrains per per perper Grains million million million

gallon per gallons gallons gallonsAl«(S0*)a gallon of water of water of water

used CaO. treated. treated treated

2,0 $3.57 03.57

2.5 $4.46 $4.46

1.5 1.99 $2.68 $.64 $3.32sodium

With alum and an equivalent amount of,. hydroxide the hardnessA

and alkalinity remain practically the same. The floe is formed at

about the same rate as when alum and an equivalent amount of lime are

used.

Alum and an equivalent amount of sodium carbonate give

practically the same results as do alum and caustic soda. This may

be seen by comparing Tables XIV and XV and Plates VI and VII. The

only advantage gained by using caustic soda or soda ash in place of

38

M

hh

oCOEh

EO

ftM>H

I

Eh FnOCOEH

o

EH

<EHCOr—

i

Oo

0)

e•HEH

CD

Pi•rl

1

•HO

o;

-hCh © g -P©Pi

CQ c3+3

©

•HEH

u o BH H O O ^p! O H -P OcriCH «H "VH

091

Pi

Pi

03 U OCQ O <H© OPirdPi

aJ

• CO I•

^H-P U rH $3

^ C'H Oa} ft g *Hft

I 05 CQ

rH P P•H o3

<M Pi-p

I

f-l rHlH © «Hcrj ft S

Pi

o•HrH

O O+3 £i0 • CQ I

*

rnp, ^ -PhHC

-PPlOÔCSjftB'H•H O O t> ft HPi

«H HH >> I © CQ I

^ 4^ H d (1) H CH 0) =h Pi a} ft 6 O<4 £ ft *h

-p irl

©

43 05

•H$3 P•H ,13

H ftaj o^ S3H a

Ph

CQ I•

Pi O rH Pi 0) .|H Oo tH o cd ft E -h

O 5> ft rH

I O CQ I•

H P P U rH J3

•H 03 U © «H O«H Pi 0) ft £ «HP ft H

9 °

? OH (S

ft <a oU3

CQ

HPi

C5

C5

PI—OMHO

© rH CQft 05^

t)0 wrH

ft

CQ CQ CQ CQ CQ CQ CQ CQ

rH Pi Pi Pi Pi Pif ,M M

.a ^3 ^3

CM A -

)

rH rH _ tr 1

to

to tO to to to to CM CM

• *

Pi Pi•H •Hso LO

.irn

CQ CQ CQ CQ CQ CQ

Pi Pi Pi Pi Pi

rC rS

to to CV CM CM CM

LO cO H H to H CO OCM OJ CM CM o O t> CD

r—

|

H rH H H H

LO LO O O H CO tO £>

cm CM CM CM O a> £> LO

rH H H H

LO H W CM O O COo rH H H o C7>

H rH H rH

LO O H O CO CO to to

O rH H H CO E-

H H H H

O O O co iO HrH

O o O cO o CO OH

1 CO O CT» CO CM

l H to E- H LO o> CO

1• • • • • • •

lH H H to

LO LO lO LO LO LO U. J

• • • • • • • •

rH H rH H H rH H H

H CM CO LO o 6- CO

I

I

i

-39-

i

!

WPMX

• o> PhM

wp 1m HP

Eh OCO

p

-40-

©© PB-rj•H £

©P

H -P05

O -PCO

oT) 4»

© P ©a «h p a£ p O H B£4 cn«H U

© op <H

CQ

h h •p hCQ P O rH P 0)

W O 'H o d ft<D O |> PhPi

P<A

WI © CQ

H 05 P ©

43 ft

I•

H Pr-l OB "HH

I•

H P•H OS -H

O ©•P ftf •

P P P>s3 p oP OSH O•H O O t>

P

CO I P•P hri OH p © «H «rt

c6 ft S HPh

•H HH !>»

X =H ©

I O 02 I PH+'-P P rH O•H G3 P © •H "H.

<H P ttj ft B rH+3 ft

O p4° *H •

O P>»«H P+5 C5 O

P +f•H rPH ftaJ oM PiH a«3rP

CO i p^ • +3 P H OO H P © -H -H«H o ai ft S H

I 0) CO 1 pH -P P P rH O•H 3 p O tH t-i

«H P d Ph S H43 ft

CO PP o trj

H P H Oa5 oh cj

P ft a) ^5£»0.

co p -*p o «•

•H P rH OCO <D H 01p ft d^-C5 tD N

rH

O

r%4

PdrH

•

P•

p•

P•

P1 •H •H vri •H

aS3 a a aIP

to to to

111

CO CO CO CO CQ Id CQ CO CQ

p P u P u P HrP rP rP rP rP rP rP

CO to to

• • • • •

P P p P p1

•H •H •H •H •H

J

a a>—

<

a a a

1ipH to CVJ

2 • • • •

CO CO CO CO • •

p p p p P PrP rP rP rP rP rP

CO IP to H H

1 J 1Cv2 J 1 1 to 1

CM 1 1 1 CM. 1 1 1 CVJ !

rH r-j J ) J H | I 1 H 1

O O 1 I I O 1 1 1 O 1

•41 CvJ 1 1 1 CM 1 1 l CVJ 1

rH H 1 » J rH 1 ) 1 rH 1

CO to to IP lO IP

H H rH H H H rH H H H rHrH —i

1 1 r-l rH H H rH rH H H

CO cvj CvJ N CO Cvj O O OH H rH H H H H rH H rH rH

H H H H H rH H rH rH rH H

cvj C\J CVJ CM CO CVJ CVJ CVJ CvJ CVJ C\J

c\j CVJ C\J CM CVJ CVJ cvj CVJ CVJ CvJ CVJ

1 CO lO cvj O CO CvJ OJ rH tO IP o> CVJ tO CO

1

1rH rH H H H

1

1

IP O tp IP O IP O IP O• • • • • • • • • •

1 rH rH CM to to LP

CvJt0îPOJ>-C0CT>OrH

I IIII IIII III

1

I

[

i

I

[

I

I

I

-41-

-42-

>MmPQ<

mEH

OPn

<O

SMPiocn

p;

CD

Pi

•Hpi

!

O1

i PPi o

rd (D'riCD fi-P

-P

CD

•H

EH cr'CH <H

Pi

o SPt o o uO H -P O

• CO

^ ^ h -P Uu o o u ®

>CO

CO O^HCD Op:

ua5

a5

a.ft

iH O•H >H

I CD CO

H P -P f-»

•H 03 Pi ©«h a} ftP ft

H O

op

4^•H

•HrHC3

4*1

H<3

5*0 • * CO•

c; ^ hHP ^ i p!

O'H > 5) ft-H -H

c o ft 0H

-pCD

I CD CO

H P P•H 03 p och p g3 ft'H -HP ft 0H

+3 CD

Hp ^•H 4*

£,3•H ftH H03 O^ PIH CD

• • CO

Jh ^ rH P ^ ISh O O !n CD H OO <H t> C3 ft-H -H

O ft 6 H

I CD CO •

H P P P I Pi

•ri a3 P CD H O

43 ft SH

OH

oH

OHH

OHH

O

CO

P

CO

CO

p

VO

p•H

oto

CO

p

lO

CO

p

CO 02 CO CO CO

p p U u P&V.O CM CM CM

CO 1 11

CM 1 1CM

H 1 ) rH

cm 1 1 1 CM

CM 1 1 1 CM

H ] 1 1 H

O H H CM CM

H rH rH H HH H H H H

O O O O OH H H rH HH rH H H H

O O O O O

CO

p

CO

CM

toH

OHrH

CO

P

HHH

COO

O

O O O O

CO

p

rHrHH

CO

5

orH

CO

P,S

CM

to CM CM rH

• • • •

Pi P Pi

•H •H •H •H

s S S S

o LO o Oto CM CM CM

J CM I

I rH I

JLO 1

I CM I

1 H

OrH

CO tO VOO O OH H H

CO Jh

P on•h^hoC) OrlOP ft a5 wC5 LQ aS

CMO to o O to H

H ID CO H to• • • • • *

H H rH rH CM CM

co-

co ££ o ^•H H i-i O03 CD rH COU ft 03--Cj5 tO N

OJ25

PIc3

HPH

lO o•

LO•

O•

LO•

O•

LO•

O•

LO«

o•

H rH CM CM to to LO

CM to LO VO CO10

-44-

line is that the hardness is not increased by the addition of the

"base. This advantage is. offset, however, by the greater cost of

caustic soda and soda ash as compared with lime, lime being the

cheapest "base obtainable.

Ferrous sulphate was substituted in place of aluminium

sulphate and experiments carried out in the same way. The iron

sulphate solution used contained in 1 cubic centimeter 4.28 mil-

ligrams of FeS04 so that 1 cubic centimeter of the solution added to

a liter of the water was equivalent to one quarter of a grain per

gallon of FeSCi .

Ferrous sulphate alone docs not yield a floe when added

to Lake Michigan water . Some base is needed to decompose the sul-

phate and yield the flocculent ferrous hydroxide which upon oxida-

tion becomes ferric hydroxide and settles out of the solution.

Because of its cheapness lime is the base commonly used.

Table XVI shews the results obtained when varying

amounts of iron sulphate and an equivalent amount of lime were added

to the water. The alkalinity is gradually decreased while the hard-

ness at first increases and then decreases with the alkalinity. The

time required for the floe to form ana for sedimentation is much

greater than is required with alum and an equivalent amount of lime.

At the end of 24 hours cylinder So. 1 in Table XVI had not yet shown

any floe. The water was filtered and the color determined by the

platinum cobalt method*, was found to be 10 parts per million.

* American Public Health Association. Standard Methods of Water

Analysis, p 21.

!To color was visible in the ether filtrates. Thus if the water

-45-

9

EH

Pii-h

En<m

CO

ElOIK

Eh OPi

0)

O U0-H•H d

03

m i

•H U vid O *dcj<*h © -P

to

I Pi

Pi o03 iH

33

+3

O Pi

EH cr*^ rH -P O©

Pi Pi *

W ^ O HCO O ^ OCD OPI

«dPicvj

w

co i piP fnn OU © -H *H05 ft H

> ft

I (1) DQ

rH -P -Prl tj p«H Pi flj

-P Pi

PH O

ft H

o ©•P ID

B>> <a

P Pi

•H O•H H3£^ -PH 0)

• M I Pi

^ U • P Pi r—I OPi O rH Pi © -H -r-i

O'HO cs3 ft rHO f> ft

I CD CO I Pi

H P P Pi rH O•H Hi Pi 0) *H *H«H Pi 03 ft HP ft

o £P -H • CO

© Pi Pi »-P Pi

>jrH Pi O rH Pi ©-p d O <H O G} ft'•H rCj O t> PP1-P

I BH O•H «ri

H•H ^rH ftco* oM PI

rH ©

ft

I O CO

rH -P P Pi i

•H Cfl P <D '

f+H ^ k5 ftp ft

I PIH OH «HH

©• S fd co

4» «H ©" PI

Pi O OH <d "H <D rH gJ

rd ftH O03

o

PI• oP Pd-H G

tD bD

Pio o

Ph i—i CO<} CO U3 On ©

«h d p, ft a fii

O £i0 to

CO

Pi

ocm

•

CO CO BQ CQf ,H H M MAcm CM CJ CV2

H H H HH O oH

©P<

,3

Pi•H

LO O LO OCQ

LQ

CQ

CO

Pi•H

H

O

©

Pi•H

CO# • • o o

w w CQ •

P< Pi Ph Pi Pi

1 o ^ AO C\2

CMH CM

<h d.

1 •H H H H1

no 0> CO CVJ CM rH

o 00 CO C\J O LO CM CMCO CO CO CM

1 H f i r—

1

1 1 i i rH rH H r—

)

CO w o j> CO CO o<# <# CO CO CM CM CM

rH H H H H rH rH H H rH

WJ CM —ji i CM CO H H CJi

H H H H O O tf> a* CO E>

rH H H H H rH

CD CM O CO O ' O CO CO

H H rH o o o CO oH H H H H H

O O O O rH rH o o o o• • • • • • • »

CM CM CM H

O o O o o O o o o• • • • • • •

cm CM <o ca rH

1 o O o o o o O o o1

l

*o CO o CM CO

LQ

CO

5

PJ

•H

O

CCH

CO

o

wPiO 0)

PO

oorH

HOow u

oHoo

o

§rH

CM tO LO CD OH

I

(

[

1

I

J

I

I

I

I

I

I

I

I

I

i

I

1

-47-

treated with lime and ircn is filtered before the reaction has had

time to take place, the filtered water is liable to be more or less

colored. For drinking purposes a colorless water is much to be

preferred. If the reaction between the lime and iron has not pro-

ceeded far enough before the water is filtered there is danger of a

floe forming in the filtered water. Such a floe might be injurious

in certain industrial processes. As for example, texile fabrics

washed in the water would be apt to be stained by the ircn and thus

rendered unfit for dyeing.* It is, therefore, important to have the

reaction complete before the water is filtered.

* Harold Collet. Water Softening, p 116,

Iron sulphate with an excess of lime was next trded to see

if the reaction could be hastened. The results are given in Table

XVII. The softening effect of the excess of lime is shown in Plate

IX by the sudden drop in the alkalinity and hardness after which they

remain fairly constant. The speed of reaction increases greatly

with the increase in the amount of lime and the precipitate settles

readily

.

It has already been shown above (page 32) that at Lorain,

Ohio, a large excess of lime tended to cause incrustations on the

sand grains of the filter. Such an excess of Ijme as is shewn in

Table XVII would therefore be impracticable.

Table XVII a. shows the results when 2.5 grains per gallon

of iron sulphate and varying amounts of lime were added to the lake

water. The alkalinity to methyl orange and the hardness decrease

quite regularly with the increase in the amount of lime as shown in

Plate IX a. The alkalinity to phenolphthalein is fairly constant

-40- •Ha

o

Pi•Ha

o

# • • •

Pi Pi Pi Pi Pi

•H •H •H •H •H •H •H

idcln a

>•* a a

O o W m o o oCV2

cS•HEH

Pi I I PJ•H Pi -H Od o >d a -ha1 "Vh © a -p

© 03 C3

p. -p

rd<d

opi o Sg -H P4 O O Pi

•H pi OHP O

CD

Pi

wPi

PI•Ha

o

Pi

CO

Pi

CV2

Pi•Ha

oH

Pi

CQ

Pi

Pi

Pi

PI•Hao

H,3

Pi Pi

• • • • • •

PJ Pi PI Pi Pi Pi

•H •H •H •H •H •H

a a a a a a

o in o oC\2 H H rH

• 02

03 Pt Pi • -P Pi I PJ

WPiOHPtCDrHOO O O C3 ft'H -Hpi o t> Ph anrdPi I CD

W -h tj -P p j p!

Pt Pi CD rH O^> a pvh -h

ft an

tO

tO

tO

tO

tO

tOto

tO

tO

O

tOCO

tO to

<o to to

PQ

EH

OCOCOHOS*1

O CO

>» a3P u•H Op:•H hH^ PH CD

• 03

^ U * P Pi 1 Pi

Pi O rH Pi Q> H OO'H O Cj PVH 'H

O > ft 0HI CD

rH -P 03

•H 33 +3 Pi I Pi

Ch Pi Pi CD H O+5 C5 PVH -H

Ph aH

cr. GO

m

to

H

to

rH H H

tO

to

o

PiO -HP CD

H>s cd

p ^•H +a

c ^•H PmrH Hc3 O^ PSH CD

<JP̂H

03

Pi H -P Pi l Pi

Pi O rH Pi CD H Oo «H O 03 P4-H -H

o > ft an

J CD 03 •

rH P P Pi ! Pi

•H 0} Pi CD H O'H Pi 03 ft'H -HP ft B H

o CD 03 pi

a Pi Pi o• «H CD "H CD H OPH 10 3} ftrH 05

a h v, o3 o3 t^D tQ

Pi

O 03 Pi <*

43 Jh-tJ fi P OOa v-i CD *H CD rH CO

«C 02 a3 ftrH CD

^ d Pi o5 pt,

oH

tOrH

m

H

tO

rH

COH

COH

rH

CO oCV2

OCv2

C2

o02

o m O in O in o• • • • • • •

rH rH C\2 CV2 to to

to02

toto CO o

• •tO•

CO•

o• •

to

*

to

•

to

• m in

oin

02

CD

rHft

s03

HCrj

SHO

Pi

orHOO

oft

I.O ^ m to CO

-50-

TT•H

""FT"•H •ri •H

PS"•H

B PicjM aM

lO LO O O oto to tO

•H8oto

MM>MHHimE-t

HHi

oCOEH

et-i

>H

5

E-iM£§:

OHHHO03EH^

—

O

EH

-4EH

Oo

Pi I •*

•h h -H Api o o oori«iH B 'H

03 -PPi c3

02

a a a

o8•HEH

Pi o S•ri in O O Upi O H -P O

CD

U

03 Pi

03 Uo oPi OPi

a$

03 I•

Pi H U -H OO O Cj ft -rt

Ch f> ft H

I 03 I

rH P P Pi H Pi

•H ij h ffl t) OPi (0 ftB -h-P ft rH

O•P

>»P•HPI•HHcs5

rî

5b •

p! Jh p, H P £i Hct3 Pi O O P. -H

Pi o «H i> «3 ft MO O ft

PI 0) 03 I

H P P H rH

*H 5h cf ft BP ft

PJ

O•ri

rH

Pi

O•ri

H

P!

O -HP

rH .

• • 03

Pi Pi H PPi O O

'Pi

Pi

05 ftp ri o•H Ppj ri•ri ftrH rHCA O^ P!H ©<J ri

Ph

Ph

l

H P 03

'ri 35 -P<ri Ph PiP aJ ft

ft

I•

H PJ

•H O

I•

H Pi

H OS -H

rH

03

PJ•Ha5

Pi

03

C•ri

0$

rj5

PIOMHOrH 03

ft orj oto

Pi

oPi H

rHft 0}

OCO

O

03

hri

CM

LO GO O« M WH H H

LO LO £>CM CM HH H H

lOHH

lOHH

O

o

3rH

£>rHrH

O

OrH

LO

CM

COO*

o

oCM

OCM

LO

LO

CM

CM

Pi

ri

E-

•Hs

O

Pi Pi Pi Pi Pi

•H •ri ri •H •H

s s s 6 a

o LO LO rHCM CM CM rH

CO £>O COH

o COH

LO

tO

CMCO

CO

LOLO

toLO

00

to

oLO

Ht-

LO

H

lO

CMlO

O

to

LO

LO

COto

LOto

o O O O o O o• • • • • • •

CM CM LO CD oH H H CM

o o o O O O oCM H £>

H H H H

O LO o LO O LO o• • • • • • •

CM CM to to LO

LO LO LO LO LO LO LO• • • • • • •

CM CM CM CM CM CM CM

to LO <o CO

-52-

until 5 grains per gallon of lime are used after wnieh the curve

steadily rises. As shown by the two alkalinity curves the maximum

amount of lime which could safely be used without overtreating is

4.5 grains per gallon. With 2.5 grains per gallon of iron sulphate

and 2.0 grains per gallon of lime there was obtained in one hour a

floe which required one hour for sedimentation. With 2.5 grains per

gallon of lime three quarters of an hour was required. If iron

sulphate at $9.00 per ton and lime at $4.50 per ton were used for

the treatment, the cost would be as follows:

TABLE XVIII.

Grainsper

gallonFeS04

GrainspergallonCaO

CostFeS04per

milliongallonsof watertreated

CostCaOper

milliongallons

of watertreated

Totalcostper

milliongallonsof watertreated

2.5 2.0 $1.61 $.64 $2.25

2.5 2.5 $1.61 $.80 $2.41

By comparing the above table with Table XIII, it is readily seen

that the use of iron sulphate in place of alum decreases very much

the cost of treatment.

The results obtained when iron sulphate and an equivalent

amount of sodium hydroxide are added to the water are shown in Table

XIX. As shown in Plate X, the methyl orange alkalinity remains

fairly constant while the hardness decreases. A more gelatinous

floe is obtained with sodium hydroxide than with lime, but the much

greater cost for the caustic soda offsets this advantage. The time

for reaction and for sedimentation are practically the same for both

treatments

.

-53-

rdco •

U U I I PCD «H O *H P OS pi «H d © -H•HO4 CD S 4sEH CD CO oj

H +»

d*h o s•H U O O Vc

0) d p rH 4* oS 0«H

©u

u u

bo09

CO

ft

© Hft -rl

PO•HH

(1)

P

Ua

i ©

•H CO

«H «H4»

rH

ft

P I

© i—

I

ft'H6

HrH

03

5

COrHrH

H«Hl

CQ

2

tOHH

CQ

CM

3

03

«H u PrP

CM ca cm

H H Hm

CQ 03 CO

P prP rP

^ w o>rH H OH H H

OH

OHH

oH

CO

CQ CQ CQ

P H P,P ^3

CM

Hto

p P P•H •H •Hs . so a otQ to to

to o ocr> o> a*

o tO tOCO CO

O ©P> t»0 •

P P>S 53 H

P O•H O OP•H rH

3£H CD

IH•H

03P h 1 fl

Is OH Oa5 ft *h *h

ft e hCQ

-P P I P^ OH Oa} ft-ri-nft 0rH

CO H «* CM C?> t> CO CO

H CM CM CM H H rH H rH H HH H rH H H H rH rH rH H rH

CO O CM O to <tf ^ O a» o CO

H cm CM CM H H rH H o O HH H H H H rH rH rH H H H

pO -H•P CD

H>a c5P rfl

•H -P

p.HOPCD

rH PP O

03• -P» P

rHS3

1 CD

H P«H aJ

«H P4^

CQ

rH

aft

rH I

CD HPh-h

s

rH 1

CD rHft'H

o•I

J.

Pio•H

o o o O O O O O O o O00 CM

HtoH

<orH

tX)

HCOH

HC\2

HCM

HCM

toCM

SI1

o o o O O o O O O o OCO

H<orH

tOH

oH

COrH

oCM

oCM

O Q-i

CM

o o oCM H

Po

pCQ

09

o

CD

O 03

CO d

PO CD

rH CQ

•H H

03

P•HCJ

rH

So

Or Wft aJ O

la;

03

P rH •

•H CD rH OcrJ ft aJ corH CD

o

cr.

CJ

in

CO O to to CM H o oCO rH o CO tO

• • • • • • • • •

rH rH rH CM CM CM CM

o O LO O in O m OrH H CM CM to to m

CM tO o CO a* oH

I

I rH CJ 03

CD

rH rPO PH O

1 OJ OI

P

I

I

I

I

I

I

I

I

I

-55-

The results of treatment with iron sulphate and sodium

carbonate are shown in Table XX. The floe produced is gelatinous

and similar to that produced by sodium hydroxide. The greater

length of time required for the reaction makes the sodium carbonate

less desirable than the sodium hydroxide. Because of its greater

cost this treatment is net so practical as the treatment with lime

and iron.

CCrGLUSIClI.

A careful consideration of the preceeding data shows

that,

1. Lake Michigan water is seriously polluted and should

be purified before it is used for drinking.

2. Llechanical filtration is a most satisfactory and

economical method of purifying the lake water.

3. Treatment of the lake water with lime alone softens

the water but does net yield a satisfactory floe for purification.

4. Treatment with alum alone recpuires 2 l/2 grains per

gallon of alum in order that a good floe may be formed in less than

one hour

.

5. Vfhen alum and lime are used for treatment a good floe

is obtained in one-half hour with 1.5 grains per gallon of alum and

1.99 grains per gallon of lime. As compared with the cost of treat-

ment when alum alone is used this method shows a saving of $1*84 for

each million gallons of water treated. By this treatment the hard-

ness of the water is reduced from 125 to 75 parts per million.

6. The use of alum with sodium hydroxide or sodium

carbonate increases largely the cost of treatment and does not yield

I

-56-

pq

EH

CO

-4PhoCO

3

EH

Ph

00

OWi-i

(D

•HEH

J Pi

PJ O

CD

fH I

•H fH-Hfi O -d O 'Hcr"CH <l> S -PCD CO 03

fH P

CD

fn O e•H fn O O fnOHP OCT'HH <h mhCD

fn u •

CO ft O HGO O HH QCD O f>

P5THfHc3

w

CO

4* fH I

fn CD H03 Ph'Hft £

Pio

i CD CO

HP-P•H 05 fn

fn

fH I PJ

O CD

-P 60PJ

>2•H OPI•H H33^ PH CD

• CQ •

fn fn *P fn I PJ

fn O H fn CD rH OO 'H O a} Ph «rl «H

O > ft S H| CD CO •

H -P -P fH I Pi*H c3 fn OH OHH fn 03 Ph'H tHP Ph S H

PJ

O *HP CD

HJô3-P A•H -PPJ^•H ftrH iHo3 OM PiH CD

Ph

• CO •

fH fH «-P fH I PJ

fn O H fn CD rH OO 'H O 03 Ph'H iHO > Ph CH

1 CD CO •

rH P P fn I PJ

•H 03 fn <D H Och ?h csS Ph *h «hP Ph SH

«hO CO

o3 mCD

CO PJ WPi o o

_ tH fn H OCS CO S) CO H N

B re) pi fn Ph o3 03

60 60P3OCO

• o+> fH-d

*H CD

PJ +o o

fn rH CO•a!) CO o3 b rH <U

sh pi fn Ph o3 PhO 60 60

o

CO CO

fn

.3

CQ

CO

fH

^3

CQ

CO

fH

CQ

CO

fH

CQ

CO

fH

cq

PJo

co

CD fnP ^J•H CD

-P P H OO PhCQPJ Ho pJ

CO «H

Oo3

CD

fH

HrH

CO

fH

CQ

toCQH

CQ

HH

<X>

HH

CO

fn

rCj

CQ

CO

fn

rCj

CD OrH H

CO CQ CO CO CO CO

fH fH fn fn fH fH

CQ CQ CQ CQ

H H H Hcc CQ CQ rH H rH

1CO

I ! 1

1 J

1|

CQ1 J !

to1

1 1 IH

1 1 1H

1 I

|

1

lO1 I 1

CO1

1

CQ) 1

iCQ

1

!

i

l 1rH

1 ] JrH

0> CQ <£> CO rH toH CQ CQ CQ CQ to to to

H H rH H rH H rH H

00 O H CQ ^ t-O CO CT>

H CQ CQ CQ CQ CQ CQ CQ

H H rH H H H H rH

O O O o O O O o o o

O O O o o O o o o o

o>H

COto

O lO tocn

CQH

Hto

oiTi *o

COCO

•

H H rH H H

O lO o LTi O lO O LO O•

rH H CQ CQ to to LO

H CQ to CO orH

o o

rH CO

-*s SH

<D

^3Po

fn fHO OrH rHo oo o

1

-58-

any practical advantage over the treatment with lime and alum.

7. 7/hen iron sulphate and lime are used for treatment a

satisfactory floe is formed in one hour with 2.5 grains per gallon

of iron sulphate and 2.0 grains per gallon of lime at a cost of

§2.25 per million gallons of water treated.

8. Sodium hydroxide with iron sulphate yields a more

gelatinous floe than does the lime, "but the cost of treatment is

much greater.

9. Sodium carbonate with iron sulphate forms a gelati-

ous floe, "but the reaction is much slower than when sodium hydroxide

is used. The cost of treatment is also much greater than the treat-

ment with lime and iron.

10. T/hen iron sulphate is used as the coagulant , there is

danger of the iron appearing in the filtered water unless the reactioi.

is completed "before the water goes on the filter. The use of a

slight excess of lime hastens the reaction and decreases the danger

of iron appearing in the filtered water.

11. A comparison of the lime and alum with the lime and

iron sulphate method shows that by using the iron method there

results a saving of $1.07 for each million gallons of v.ater treated.

The reaction with lime and alum, however, is twice as rapid as betweer,

the iron and lime so that there is a saving in the size of the sedi-

mentation "basins. The Y/ater treated with alum and lime has a clear

crystalline appearance which is not true of the water treated with

iron sulphate. Should the iron appear in the filtered water it would

"be injurious in certain industrial processes, while the floe from

alum "because it is colorless, would be entirely harmless. In spite

of its greater cost, we "believe the alum and lime treatment to be

-59-

best adapted for use with Lake Michigan water.

To Dr. Edward Bartow, Professor of Sanitary Chemistry,

under whose direction this work has been done, I wish to extend my

thanks for many valuable suggestions and for assistance in carrying

on the work. To all those who so kindly furnished analyses of the

Great Lakes water, I also extend my thanks.

BIBLIOGRAPHY

Clean Water and How to get it. Allen Kazen.

Data of Geochemistry . F. W. Clarke. United States Geological

Survey Bulletin.

Filtration of Public Water Supplies. Allen Ilazen.

First P.epcrt of the Lake MioMgan Water Commission.

The Waters of the Great Lakes. E . B. Dole.

Typhoid Fever. G. C. Whipple,

Water and Sewage Purification in Ohio. Chic State Board of Health

P.eport. 19C8.

Water Softening. Harold Collet.

Water Purification at Louisville. G. W. Fuller.

Standard Methods of the American Public Health Association.

Documents

Composition and treatment of Lake Michigan water