Water purification

Survival, Water, Medical Field Manuals

Military Manuals

Ellms, Joseph Wilton

Document text

\ 


Gompfimentanj  Oopy 
for  ReviQW 

THE  PftlCE  OF  THI5  QOOK  13 

«     7-^^^ ^ 


WATER  PURIFICATION 


IS 


Eh 


2: 


2; 
3 

•-4 


WATEK 
PURIFICATION 


BY 

JOSEPH  W.  ELLMS 

Member  American  Society  of  Civil  Engineers,  American  Water 

Works  Association  and  American  Chem.ical  Society;  Fellow 

of  The  American  Public  Health  Association  and 

The  American  Association  for  the 

Advancement  of  Science. 


Second  Edition 


McGRAW-HILL  BOOK  COMPANY,  Inc. 
NEW  YORK:  370  SEVENTH  AVENUE 

LONDON:  6  &  8  BOUVERIE  ST.,  E.  C.  4 
1928 


Copyright,  1917,  1928,  by  the 
McGraw-Hill  Book  Company,  Inc. 


PRINTED    IN   THE    UNITED    STATES    OF   AMERICA 


THE   MAPLE    PRESS    COMPANY,   YORK,   PA. 


PREFACE  TO  THE  SECOND  EDITION 

The  author,  in  preparing  the  second  edition  of  this  book,  has 
kept  in  mind  the  advance  in  the  art  of  water  purification  as  it 
has  developed  during  the  past  ten  years.  More  recent  data  have 
been  employed  to  amplify  the  subject  matter  wherever  it  seemed 
desirable.  A  new  chapter  has  been  written  dealing  with  the 
relation  of  the  hydrogen-ion  concentration  of  natural  and  purified 
waters  to  purification  processes. 

The  widespread  use  of  the  book  both  as  a  text  and  a  reference 
has  been  extremely  gratifying  to  the  author,  and  it  is  hoped  that 
in  this  new  edition  readers  may  find  still  more  material  of  an 
interesting  and  instructive  character.  The  author  has  freely 
drawn  on  the  extensive  and  constantly  growing  literature  of  the 
subject  of  water  purification,  and  to  the  many  writers  quoted  he 
desires  to  express  his  acknowledgements  in  preparing  this  second 
edition. 

Joseph  W.  Ellms 

Cleveland,  Ohio, 
Apnl,  1928. 


vu 


PREFACE   TO   THE  FIRST   EDITION 

In  writing  this  book  the  object  has  been  to  provide  the  reader 
with  a  fairly  complete  account  of  the  development  of  the  art  of 
water  purification.  As  a  knowledge  of  the  physical,  chemical 
and  biological  characteristics  of  natural  waters  is  a  prerequisite 
to  the  proper  understanding  of  purification  processes,  a  considera- 
tion of  the  properties  of  various  classes  of  waters  has  been  thought 
advisable.  The  relation  of  polluted  pubhc  water  supplies  to 
water-borne  diseases  has  received  especial  attention  because  of 
its  importance.  The  various  steps  in  purification  processes, 
such  as  plain  sedimentation,  coagulation,  filtration  and  disinfec- 
tion, are  described  in  considerable  detail.  Special  chapters  are 
devoted  to  water  softening,  and  to  the  removal  of  iron  and 
manganese  from  ground-water  supplies. 

The  rapid  progress  made  in  the  art  of  clarifying  and  purifying 
turbid  waters  in  the  United  States  during  the  past  quarter  of  a 
century  has  been  notable.  The  evolution  of  the  rapid  sand 
filter  from  its  crude  beginnings  to  its  present  well-developed 
state  for  purifjdng  waters  of  this  type,  is  distinctly  the  result  of 
research  work  undertaken  during  the  early  part  of  this  period. 
It  was  the  author's  good  fortune  to  have  been  identified  with 
some  of  the  earlier  investigations  of  this  problem,  and  to  have 
been  able  to  follow  its  solution  closely  in  actual  practice  through- 
out the  whole  period. 

The  experiences  of  other  investigators,  as  disclosed  by  their 
published  papers,  have  been  drawn  upon  freely,  and  while  no 
exhaustive  examination  of  the  extensive  literature  of  the  sub- 
ject has  been  attempted,  it  is  believed  that  the  writers  quoted 
and  referred  to  are  sufficiently  representative  to  provide  the 
reader  with  ample  information  upon  the  subjects  discussed. 

The  wTiter  is  indebted  to  a  number  of  his  friends  who  have 
kindly  supplied  him  with  original  information,  illustrations,  or 
assistance  in  writing  technical  descriptions  of  apparatus  for 
some  parts  of  the  book;  and  to  manufacturers  of  special  devices 
used  in  filter-plant  construction,  who  have  loaned  original  draw- 
ings and  photographs  for  reproduction. 

ix 


X  PREFACE  TO  THE  FIRST  EDITION 

The  author  desires  to  acknowledge  especially  the  contribution 
of  the  subject  matter  in  the  appendices  written  by  Mr.  C.  N. 
Miller,  Associate  Member  American  Society  of  Civil  Engineers, 
deaUng  with  the  hydraulics  of  the  flow  of  water  through  filters, 
and  with  the  discharge  of  water  from  waste-water  troughs  in  the 
operation  of  rapid  sand  filters.  The  author  is  also  under  special 
obligation  to  Mr.  S.  J.  Hauser,  Chemist  and  Bacteriologist  of  the 
Cincinnati  Water  Purification  Plant,  for  his  kindly  assistance  in 
reading  the  proof  of  the  book  and  in  preparing  the  index. 

Joseph  W.  Ellms. 

Cincinnati,  Ohio, 
March  1,  1917. 


CONTENTS 

Page 

Preface  to  the  Second  Edition vii 

Preface  to  the  First  Edition ix 

Chapter 

I.  Introduction 1 

II.  Classification  of  Natural  Waters 8 

III.  Transmission  of  Disease  through  Drinking  Water.    .  24 

IV.  The  Effect  of  Improved  Water  Supplies  upon  Health    .  34 
V.  Objects  and  Methods  of  Water  Purification 49 

VI.  Sedimentation 54 

VII.  Types  OF  Settling  Reservoirs  AND  Coagulation  Basins.      66 

VIII.  Practical  Efficiencies  of  Settling  and  Coagulation 

Basins 91 

IX.  Filtration  of  Water 103 

X.  Preliminary    Treatment    of    Water    for    Slow    Sand 
Filters HI 

XI.  System  of  Slow  Sand  Filtration 130 

XII.  System  of  Slow  Sand  Filtration  {Continued) 150 

XIII.  Efficiency  and  Cost  of  Operation  of  Slow  Sand  Filters  168 

XIV.  Rapid  Sand  Filtration 178 

XV.  General  Arrangement  of  Rapid  Sand  Filter  Plants  .  192 

XVI.  Details  of  Rapid  Sand  Filter  Plant  Construction.    .    207 

XVII.  Details    of    Rapid    Sand    Filter    Plant    Construction 

(Continued) 227 

XVIII.  Details    of    Rapid    Sand    Filter    Plant    Construction 

{Continued) 243 

XIX.  Regulating,    Measuring    and    Indicating    Devices   for 

Rapid  Sand  Filter  Plants 254 

xi 


xii  CONTENTS 

Chapter  Page 

XX.  REnrLATiNG,    Measuring   and   Indicating   Devices   for 

Rapid  Sand  Filter  Plants  {Continued) 269 

XXI.  Equipment  for  the  Handling  and  Storing  of  Chemicals 

AND    FOR    THE    PREPARATION    OF    SOLUTIONS 289 

XXII.  Apparatus  and  Methods  for  Applying  Chemicals  and 

THE  Preparation  of  Solutions 313 

XXIII.  Power  Plant,  Pumping  Machinery,  Air  Compressors, 
Air  Tanks,  Wash  Water  Tank  and  Miscellaneous 
Equipment 334 

XXIV.  The  Cost  of  Constructing  Rapid  Sand  Filters.    .    .    .   352 

XXV.  Rates  of  Filtration,  Loss  of  Head  and  Washing  of 

Rapid  Sand  Filters 359 

XXVI.  Efficiency  and  Cost  of  Operation  of  Rapid  Sand  Filters  378 

XXVII.  The  Physical  and  Chemical  Changes  Produced  by  the 
Application  of  Chemical  Coagulants,  and  by  the  Sub- 
sequent FlLTR.\TION   OF  THE   TREATED   WaTER 409 

XXVIII.  The   Effect  of  the   Hydrogen-ion  Concentration   of 

Natural  Waters 426 

XXIX.  Disinfection  of  Water  Supplies 444 

XXX.  Disinfection  of  Water  Supplies  {Continued) 479 

XXXI.  The    Removal    of    Dissolved    Mineral    Matter    from 

Water 499 

XXXII.  The    Removal    of    Dissolved    Mineral    Matter    from 

Water  {Continued) 539 

XXXIII.  The  Control  of  Water  Purification  Processes   .    .    .  559 

Appendix  \ 5(37 

Appe.ndix  B 571 

Tables  1  to  13,  Inclusive 573 

Index 585 


WATER  PURIFICATION 

CHAPTER  I 

INTRODUCTION 

Water  purification  may  be  broadly  defined  as  the  art  of  remov- 
ing foreign  and  polluting  substances  from  solution  and  suspension 
in  water.  The  character  and  amount  of  these  impurities,  and 
the  object  for  which  their  removal  is  undertaken,  naturally 
leads  to  the  treatment  of  this  subject  under  the  following  heads: 

1.  Rendering  impure  water  potable  and  hygienically  safe  for 
drinking  purposes,  and  suitable  for  industrial  uses. 

2.  Removing  putrescible  organic  matter  and  disease-produc- 
ing organisms  from  the  fouled  water  supply  or  sewage  of  a 
community. 

In  the  following  pages  it  is  intended  to  treat  only  of  the  first 
division  of  this  subject,  since  the  second  covers  too  broad  a 
field  to  bring  it  within  the  scope  of  this  volume. 

Both  the  kind  and  the  amount  of  impurities  in  a  polluted  water 
affect  the  methods  which  can  be  employed  for  its  purification, 
and  this  is  so  peculiarly  true  of  sewage  that  its  treatment  has 
become  an  art  in  itself.  Those  waters  which  are  to  be  purified 
for  drinking  purposes  demand  that  a  degree  of  purity  shall  be 
produced  which  is  not  possible  and  usually  unnecessary  in  the 
case  of  sewage.  The  purification  of  the  latter  is,  as  a  rule,  quite 
incomplete  as  compared  with  that  required  for  a  safe  drinking 
water,  and  obviously  the  object  to  be  attained  is  quite  different. 
In  the  case  of  waters  to  be  treated  for  making  them  suitable  for 
steam-boiler  purposes,  or  for  general  industrial  uses,  a  slightly 
different  object  is  sought.  The  softening  of  hard  waters  or  the 
removal  of  compounds  of  iron  or  manganese  may  be  an  integral 
part  of  the  purification  of  a  public  drinking  water  supply,  in  which 
case  it  is  combined  with  those  processes  utihzed  for  making  a 
water  hygienically  safe.  Of  course,  it  is  quite  possible  to  carry 
on  purification  for  industrial  purposes  independently  of  the  more 
refined  methods  used  for  drinking  water,  and  frequently  this 
course  is  pursued  in  connection  with  manufacturing  plants. 

1 


WATER  PURIFICATION 


HISTORICAL 


Ancient  Systems  of  Water  Supply. — Good  water  for  drinking 
purposes  has  doubtless  been  appreciated  by  the  human  race 
from  time  immemorial,  for  it  does  not  take  a  very  high  degree 
of  intelligence  to  discriminate  between  a  clear,  colorless  and 
odorless  drinking  water,  and  one  that  does  not  possess  these 
outward  attractive  physical  properties.  Among  primitive 
peoples  the  question  of  water  supply  was  never  of  pressing  impor- 
tance, except  in  arid  and  semi-arid  regions.  In  these  latter 
countries  provision  for  securing  and  storing  a  supply  of  water 
was  usually  necessary.  Consequently,  springs  were  sought  for, 
wells  were  dug  and  cisterns  constructed  in  order  that  a  supply 
of  water  might  at  all  times  be  available. 

Wells  were  common  in  ancient  Egypt,  Greece,  Assyria,  Persia 
and  India,  and  from  the  sanitary  point  of  view  they  probably 
furnished  a  safer  drinking  water  than  could  be  obtained  from  the 
surface  waters  in  the  rivers  and  lakes.  Wells  of  great  antiquity 
may  be  found  today  in  Egypt  and  India.  Joseph's  well  at  Cairo 
in  Egypt  is  one  of  the  most  famous  ancient  wells,  and  was 
excavated  in  solid  rock  to  a  depth  of  297  ft.  The  Chinese  were 
familiar  with  the  driving  of  artesian  wells  and  pursued  methods 
similar  to  those  now  in  vogue  in  sinking  them. 

Where  turbid  surface  waters  had  to  be  employed  for  drink- 
ing, domestic  filters  of  unglazed  earthenware  or  of  sandstone 
were  known  to  have  been  used  by  the  ancient  Egyptians  and 
by  the  Japanese.  Clarification  of  muddy  water  by  the  siphoning 
of  the  liquid  from  one  vessel  to  another  by  the  capillary  action 
of  porous  material,  such  as  a  strip  of  cloth,  and  the  consequent 
separation  of  the  water  from  the  suspended  matter,  was  well 
known  to  the  ancients. 

As  population  became  more  dense  and  people  began  to  congre- 
gate in  cities,  the  need  for  larger  volumes  of  water  than  could 
be  suppUed  by  springs  and  wells  became  urgent.  Works  for 
the  collection,  storage  and  conveyance  of  water  were  built  for 
supplying  many  of  the  ancient  cities,  and  the  ruins  of  some  of 
these  works  yet  remain. 

Within  a  year  or  two  archeological  explorations  in  India  have 
shown  that  5,000  years  ago  the  peoples  of  the  Sind  and  Punjab 
were  hving  in  well-built  cities,  and  in  possession  of  a  relatively 
mature  civilization,  with  a  high  standard  of  art  and  craftsman- 


INTRODUCTION  3 

ship,  and  a  developed  system  of  writing.  The  excavations 
brought  to  Hght  houses  and  temples,  massively  built  of  burnt 
brick,  and  provided  with  well-constructed  water  conduits  covered 
with  marble  slabs. 

The  ancient  water  tanks  of  Aden  in  Arabia  for  the  collection 
of  surface  water  from  the  gorges  of  the  volcanic  crater,  at  the 
bottom  of  which  the  old  city  was  located,  afford  an  example  of 
early  impounding  reservoirs  of  an  elementary  type.  These 
tanks  may  have  been  built  by  Persian  engineers  as  early  as  600 
B.C.,  or  possibly  by  the  Romans;  but  they  without  •  doubt 
antedate  the  Christian  era. 

The  infiltration  galleries  for  collecting  ground  water  at  Athens 
in  Greece  w^ere  probably  constructed  over  2,000  years  ago,  and 
the  same  m.ethod  is  still  pursued  to  obtain  a  pure  and  satisfactory'- 
water  supply.  The  rain-water  cisterns  of  ancient  Carthage 
(150  B.C.)  were  divided  into  several  storage  compartments,  two 
of  which  were  apparently  used  for  settling  or  possibly  filtering 
the  water.  Historians  state  that  the  city  of  Laodicea  in  Asia 
Minor  obtained  its  supply  from  the  River  Caprus.  The  latter 
had  its  origin  in  springs.  The  water  w^as  conveyed  to  the  city 
through  a  masonry  aqueduct  over  4  miles  in  length.  A  settling 
tank  with  double  compartments  formed  a  part  of  this  water- 
works system.  In  Jerusalem  underground  cisterns  were  con- 
structed, which  were  supplied  through  masonry  conduits. 

Probably  no  more  elaborate  system  of  public  water  supply 
was  provided  for  any  ancient  city  than  that  for  Rome.  Until 
312  B.C.,  the  Romans  took  their  supply  from  the  River  Tiber 
and  from  springs  and  wells  in  the  vicinity  of  the  city.  The 
increasing  pollution  of  the  river  water,  as  well  as  the  need  for 
a  greater  volume  of  water,  evidently  induced  the  Romans  to 
seek  elsewhere  for  a  supply.  In  consequence,  three  groups  of 
springs  in  the  volcanic  plain  on  the  left  bank  of  the  Tiber  were 
made  use  of,  and  the  water  from  them  conveyed  by  aqueducts 
to  the  city.  A  fourth  group  of  springs  was  also  used  which  were 
located  in  the  mountains  of  limestone  formation  somewhat  more 
distant  than  the  three  other  groups.  This  latter  group  furnished 
the  best  quality  of  water  with  which  the  city  was  supplied. 

E.  H.  D'Avigdor,  in  his  "Water  Works  of  Ancient  Rome"' 
interestingly  describes  the  character  of  the  Roman  water  supply 
as  follows:  "The  Romans  possessed  three  almost  independent 
^  Engineering,  vol.  xxi,  p.  403. 


4  WATER  PURIFICATION 

water  services,  for  which  they  used  water  of  different  degrees 
of  purity.  The  least  clear  and  most  loaded  with  sand,  such  as 
the  Anio  aqueduct  supplied,  was  used  for  public  baths  and  the 
watering  of  streets;  the  clearer  water  from  Tepula  and  Alsietina 
served  for  tanks,  fountains  and  washing  troughs;  while  the  very 
best  (Virgo,  ]\Iarcia  and  Claudia)  was  confined  to  drinking 
purposes,  and  these  springs  were  undefiled  even  after  the  heaviest 
rains." 

Rome  was  supplied  with  water  from  the  above-mentioned 
four  groups  of  springs  through  19  aqueducts,  which  were  built 
between  321  B.C.  and  305  A.D.  The  aggregate  length  of  these 
aqueducts  was  381  miles. 

There  can  be  said  to  have  been  no  real  distribution  system 
for  the  water  entering  Rome  through  the  system  of  aqueducts. 
The  aqueduct  water  flowed  through  small  tanks  in  which  the 
hea^'iest  sand  and  gravel  were  deposited.  To  a  limited  extent 
the  small  distributing  reservoirs  (castella)  near  the  city  served 
a  like  purpose.  From  these  reservoirs  the  water  was  distributed 
to  cisterns,  public  fountains  and  private  residences. 

During  the  ]Middle  Ages,^  when  disease  played  such  havoc  with 
the  people  of  Europe,  polluted  water  undoubtedly  assisted  in 
conveying  infection.  At  the  time  of  the  fall  of  the  Roman 
Empire  many  of  the  aqueducts  built  by  the  Romans  at  Rome  and 
in  the  provinces  were  either  destroyed  or  fell  into  disuse.  The 
Moors  in  Spain  during  the  ninth  century  constructed  some 
important  works,  as  well  as  repairing  in  the  twelfth  century 
some  old  Roman  works. 

Until  1183  A.D.  Paris  obtained  its  entire  supply  of  water  from 
the  River  Seine.  As  late  as  1550,  Paris  used  only  1  qt.  of  water 
per  capita  per  day,  and  at  the  end  of  the  seventeenth  century 
was  using  but  23^  qt.  per  capita  per  day.  With  such  small 
amounts  of  water  being  used  one  can  easily  imagine  what  sanitary 
conditions  must  have  been. 

London  was  first  supplied  in  small  quantities  with  spring  water 
conducted  through  lead  pipes,  and  masonry  conduits.  In  1582 
a  pump  was  erected  on  London  Bridge  to  take  water  from  the 
River  Thames,  and  to  deliver  it  through  lead  pipes  to  the  city. 

By  the  invention  of  the  steam  engine,  pumping  machinery  of 
adequate  capacity  and  power  was  made  possible.  The  growth 
and  development  of  water-works  plants  in  reality  dates  from 

1  TuRXE.^URE  and  Russel:  "Public  Water  Supplies."  ^ 


INTRODUCTION  5 

the  eighteenth  century.  However,  not  until  the  latter  half  of 
the  nineteenth  century  was  very  rapid  progress  made.  The  use 
of  cast-iron  pipe  became  general  in  about  the  year  1800,  and 
gradually  replaced  the  wooden  mains  formerly  used. 

Development  of  Modem  Purification  Plants. — The  methods 
employed  in  securing  and  maintaining  pure  water  supplies  have 
been  in  a  large  measure  governed  by  topographical  and  geological 
conditions.  In  different  countries,  where  like  physical  condi- 
tions prevailed,  similar  lines  of  development  have  not  always 
been  followed.  Allen  Hazen,  in  his  book  on  "Filtration  of 
Public  Water  Supplies,"  points  out  that  "it  is  really  marvellous 
how  each  country  has  met  its  problems  of  water  supply  from  its 
own  resources,  and  often  without  much  regard  to  the  methods 
which  had  been  found  most  useful  elsewhere.  England  has 
secured  a  whole  series  of  magnificent  supplies  by  impounding 
the  waters  of  small  streams  in  reservoirs  holding  enough  water 
to  last  through  dry  periods,  while  on  Continental  Europe  such 
supplies  are  hardly  known.  Germany  has  spent  millions  upon 
millions  in  purifying  turbid  and  polluted  river  waters,  while 
France  and  Austria  have  striven  for  mountain-spring  waters 
and  have  built  hundreds  of  miles  of  costly  aqueducts  to  secure 
them.  In  the  United  States  an  abundant  supply  of  some  liquid 
has  too  often  been  the  objective  point,  and  the  efforts  have  been 
most  successful,  the  American  works  being  entirely  unrivalled 
in  the  volumes  of  their  supplies.  I  do  not  wish  to  imply  that 
quality  has  been  entirely  neglected  in  our  country,  for  many 
cities  and  towns  have  seriously  and  successfully  studied  their 
problems,  with  the  result  that  there  are  hundreds  of  water  sup- 
plies in  the  United  States  which  will  compare  favorably  upon 
any  basis  with  supplies  in  any  part  of  the  world ;  but  on  the  other 
hand,  it  is  equally  true  that  there  are  hundreds  of  other  cities, 
including  some  of  the  largest  in  the  country,  which  supply  their 
citizens  w^ith  turbid  and  unhealthy  waters  which  cannot  be 
regarded  as  anything  else  than  a  national  disgrace  and  a  menace 
to  our  prosperity." 

The  above  statement  was  undoubtedly  true  at  the  time  it 
was  written.  In  the  past  30  years,  however,  there  has  been  a 
marked  improvement  in  public  water  supplies.  This  condition 
has  been  chiefly  brought  about  by  an  awakened  public  interest 
in  general  sanitation,  and  especially  in  the  better  quality  of  the 
drinking  water  supplied  for  public  consumption.     Most  of  the 


6  WATER  PURIFICATION 

large  cities  of  the  United  States  are  now  well  protected  against 
water-borne  diseases,  even  though  the  source  of  supply  is  in 
many  cases  badly  contaminated.  The  barriers  set  up,  however, 
must  be  closely  guarded,  and  extreme  \agilance  on  the  part  of 
those  entrusted  with  this  duty  is  the  only  safeguard  for  millions 
of  water  consumers. 

Those  impurities  in  water  which  render  it  turbid,  and  which 
for  the  most  part  are  in  suspension,  make  it  unattractive  as  a 
drinking  water.  It  was  probably  noted  at  an  early  period,  that 
a  muddy  water,  if  allowed  to  remain  quiescent  for  a  short  time, 
lost  more  or  less  of  its  suspended  matter  by  settlement.  In 
storing  muddy  waters  for  irrigation  purposes  this  phenomenon 
could  hardly  have  failed  to  have  been  noticed. 

The  appearance  of  a  muddy  water  was  obviously  improved  by 
settlement,  and  its  suitability  as  a  drinking  water  enhanced. 
If  clear  spring  waters  could  not  be  obtained  for  a  public  supply, 
resorting  to  surface  waters  was  a  necessity.  Hence,  we  find 
evidence  that  in  some  of  the  ancient  water-works  systems  pro- 
vision was  made  for  settlement  of  muddy  waters  in  tanks  or 
settling  basins.  The  "castellee"  and  piscinse  of  the  Roman 
aqueduct  system  evidently  performed  the  function  of  settling 
tanks,  whether  originally  intended  for  that  purpose  or  not.  In 
other  ancient  systems  the  evident  purpose  of  tanks  for  settling 
purposes  is  more  pronounced.  Some  of  these  works  have  been 
cited  in  the  preceding  pages. 

These  early  devices  for  improving  the  quaUty  of  a  water  supply 
were  crude  and  imperfect,  and  can  only  be  recorded  as  the 
beginnings  from  which  the  really  modern  art  of  water  purification 

sprang. 

"\Miile  sedimentation  for  the  clarification  of  water  may  be 
looked  upon  as  the  earliest  step  in  the  art,  nevertheless,  it  can 
only  be  regarded  in  most  cases  as  preliminary  to  more  complete 
methods.  The  straining  action  of  sand  and  gravel  was  doubt- 
less also  noted  by  keen  observers  at  an  early  date,  in  the  same 
way  that  the  clarification  effected  by  settling  tanks  had  been 
observed.  The  application  of  this  principle  to  the  pubhc  water 
supply  of  London  was  made  in  1829.  Filters  of  sand  and  gravel 
were  also  constructed  for  many  of  the  Continental  cities,  more 
especially  in  Germany,  where  the  principles  underlying  the  action 
of  filters  of  this  type  were  carefully  studied. 


INTRODUCTION  7 

Allen  Hazen  has  divided  the  history  of  water  purification  in 
the  United  States  into  three  epochs,  the  first  beginning  with 
James  P.  Kirkwood's  report  on  the  "Filtration  of  River  Waters," 
in  1866,  resulting  from  his  study  of  European  practice;  the 
second  epoch  commencing  with  the  work  of  the  Massachusetts 
State  Board  of  Health  at  its  Lawrence  Experiment  Station,  in 
1887;  and  the  third  with  the  experiments  on  very  turbid  waters 
beginning  at  Louisville,  Ky.,  in  1896,  and  continued  at  Pitts- 
burgh and  Cincinnati  during  the  3  or  4  years  following.  To 
these  three  epochs  the  author  would  add  a  fourth,  which  was 
introduced  in  1908  by  the  experiments  at  Chicago  and  at  Boon- 
ton,  N.  J.,  on  the  disinfection  of  water  with  hypochlorite  of  hme. 
Chlorine  and  its  compounds,  as  well  as  ozone  and  ultra-violet 
light,  have  been  carefully  investigated  as  disinfecting  agents  both 
in  this  country  and  in  Europe.  The  widespread  employment  of 
chlorine  has  amply  demonstrated  its  usefulness  as  a  practical, 
efficient  and  economical  agent  in  water  purification.  This  phase 
of  water  purification  is  well  estabhshed,  and  marks  an  important 
stage  in  the  practical  development  of  the  art. 

References 

1.  L.  F.  Vernon-Harcout:  "Sanitary  Engineering,"  1907. 

2.  Bolton:  "Ancient   Methods  of   Filtration."     Pop.   Sci.   Mo.,   vol.    16, 

p.  495. 

3.  Arthur  S.   Riggs:  "Ancient  Water  Tanks  of  Aden,   Arabia."     Eng. 

News,  vol.  52,  p.  25,  1904. 

4.  "Ancient  Water  Supply  of  Athens."     Eiigineer,  vol.  101,  p.  215,  1906. 

5.  Croes:   "The   Water  Works  of  Carthage."     Eng.  Record,  vol.  25,  p.  8, 

1892. 

6.  Edward   Wegmann:  "The   Water  Works  of  Laodicea,    Asia   Minor." 

Eng.  Record,  vol.  40,  p.  354. 

7.  Edward    Wegmann:  "Ancient    and    Modern    Water    Works."     Eng. 

Record,  vol.  65,  June,  1912. 

8.  E.    H.    D'Avigdor:  "Water  Works  of   Ancient  Rome."     Engineering, 

vol.  21,  p.  403,  1876. 

9.  Geo.  Higgins:  "The  Old  Water  Supply  of  Seville."     Proc.  Inst.  C.  E., 

vol.  38,  334. 

10.  TuRNEAURE  and  Russel:  "Public  Water  Supplies." 

11.  Allen  Hazen:  "Filtration  of  Public  Water  Supplies." 

12.  W.  P.  Mason:  "Water  Supply." 

13.  Engineering  News-Record,  p.  1028,  Dec.  25,  1924. 


CHAPTER  II 
CLASSIFICATION  OF  NATURAL  WATERS 

In  order  to  understand  intelligently  the  methods  employed  in 
water-purification  processes,  it  is  necessary  to  have  a  knowledge 
of  the  composition  of  the  various  classes  of  natural  waters  which 
require  treatment.  Although  all  the  water  which  is  now  flowing 
on  the  surface  of  the  ground,  as  well  as  that  which  is  found  in 
the  ground,  has  a  common  origin,  nevertheless,  this  water  has 
acquired  varied  characteristics,  which  have  been  derived  from 
the  substances  with  which  it  has  come  in  contact. 

RainfalL — From  the  water  surfaces  of  the  earth  the  sun  is 
constantly  evaporating  enormous  volumes  of  water,  which  are 
again  condensed  and  precipitated  upon  the  land  and  water 
surfaces  as  rain,  snow  or  ice.  That  which  falls  upon  the  land 
surfaces  of  the  earth  is  disposed  of  in  different  ways.  Some  of 
the  water  is  evaporated  again;  some  is  absorbed  by  growing 
vegetation;  some  flows  directly  over  the  ground  into  the  rivers 
and  lakes;  and  some  sinks  into  the  soil  to  become  the  subter- 
ranean water  of  the  earth.  The  amount  of  water  precipitated 
on  the  surface  of  the  earth  in  any  particular  region  depends  upon 
a  number  of  factors.  Warm  air  which  is  saturated  with  water 
in  the  form  of  vapor  will  only  precipitate  this  moisture  when 
sufficiently  cooled.  Air  currents  may  be  cooled  by  coming  into 
contact  with  the  cold  surfaces  of  the  earth,  or  by  being  forced 
upward  by  mountain  ranges  into  colder  air  currents,  or  by 
mingling  directlj^  with  colder  air  currents  at  lower  levels.  Pre- 
cipitation of  water  will  take  place  under  these  conditions. 

The  water  thus  precipitated  is  as  near  pure  water  as  may  be 
found  under  natural  conditions.  It  probably  will  contain  minute 
quantities  of  nitrogen  compounds,  such  as  ammonium  salts,  and 
some  organic  matter  consisting  of  plant  spores  and  bacteria. 
It  will  also  contain  in  solution  small  quantities  of  the  gases, 
oxygen  and  carbon  dioxide. 

As  the  descending  rain  or  snow  approaches  the  earth,  it  washes 
out  of  the  atmosphere  suspended  particles  of  dust  and  thus  puri- 
fies the  air.     These  impurities  will,  of  course,  be  found  in  the 

8 


CLASSIFICATION  OF  NATURAL  WATERS  9 

rain  water  if  collected  before  reaching  the  ground.  Rain  water 
caught  in  or  near  our  large  cities  is  much  more  impure  than  that 
collected  in  the  open  country.  The  impurities  swept  from  the 
air  are  relatively  in  small  amounts  as  compared  with  those 
acquired  b}-  the  water  after  reaching  the  earth;  but  they  are  the 
first  of  the  dissolved  and  suspended  constituents  to  be  taken 
up  by  the  water,  and  are  soon  increased  in  amount  and  kind  after 
the  water  reaches  the  ground. 

SURFACE  WATERS 

Flowing  Surface  Waters. — The  character  of  the  land  surfaces 
upon  which  rain  or  snow  is  precipitated  naturally  influences  the 
proportion  of  the  volume  of  water  which  finds  its  way  over  the 
top  of  the  ground  into  the  streams  and  lakes,  and  that  which 
sinks  into  the  ground  to  become  a  part  of  the  underground  water 
of  the  earth.  If  the  land  is  mountainous,  the  rain  rapidly  runs 
off  the  steep  slopes  into  the  valleys,  and  quickly  enters  the 
rivers.  If  the  soil  is  easily  washed  aw'ay,  the  water  will  carry 
with  it  large  volumes  of  sediment,  the  composition  of  which  will 
depend  upon  the  geological  characteristics  of  the  denuded  sur- 
faces. On  the  other  hand,  if  the  precipitation  occurs  in  a  flat 
level  country,  a  much  larger  proportion  of  the  water  will  per- 
colate into  the  soil.  If  vegetation  is  abundant,  some  of  this 
water  will  be  retained  by  the  growing  plants,  and  ui&y  sensibly 
retard  the  discharge  of  surface  water  into  the  natural  water 
courses.  In  water  which  has  been  subjected  to  such  conditions 
there  will  be  found  those  dissolved  salts  with  which  the  water 
has  come  in  contact,  as  well  as  organic  matter  derived  from 
vegetation. 

Flowing  surface  waters  may  be  characterized,  therefore,  as 
those  which  contain  relatively  small  amounts  of  dissolved  mineral 
compounds,  more  or  less  of  suspended  matter,  depending  upon 
topographical  and  geological  conditions,  and  variable  amounts 
of  organic  matter,  likewise  modified  by  the  flora  of  the  region. 

Impounded  Surface  Waters. — Natural  lakes  and  ponds  are 
composed  of  water  which  has  found  its  way  directly  through 
surface  streams  into  the  depressions  in  w^hich  these  bodies  of 
water  lay,  and  of  the  water  which  has  percolated  for  some  distance 
through  the  ground  and  reappears  as  a  surface  water  at  lower 
surface  levels.     In  a  like  manner  the  low-water  flow  of  streams 


10  WATER  PURIFICATION 

is  very  largely  maintained  by  the  ground-water  flow  from  higher 
levels. 

In  consequence,  these  lake  waters  are  virtually  mixtures  of 
surface  and  ground  waters,  and  their  composition  is,  in  a  large 
measure,  intermediate  between  them.  Obviously  local  topog- 
raphy and  the  very  important  factor  of  sedimentation  will 
materially  affect  their  composition.  They  also  have  an  im- 
portant influence  in  rendering  streams  flowing  from  them  more 
uniform  in  their  discharge  on  account  of  the  storage  for  water 
which  they  afford. 

UNDERGROUND  WATERS 

The  water  which  sinks  into  the  ground  does  not  all  imme- 
diately pass  through  the  subsoil  to  underground  channels.  Some 
is  held  by  the  surface  soil,  and  in  hot  countries  may  soon  be 
evaporated  into  the  au';  some  of  the  water  is  absorbed  by  the 
growing  plants;  and  the  remainder  sinks  slowly  or  quickly  through 
the  ground  depending  upon  the  porosity  of  the  latter. 

The  movement  of  this  ground  water  is  governed  exclusively 
by  the  character  of  the  strata  with  which  it  comes  in  contact. 
Impervious  formations  will  block  or  divert  the  flow,  while  per- 
meable strata  will  soon  become  filled  and  form  the  channel  along 
which  the  underground  stream  moves.  The  velocity  of  flow  is, 
of  course,  low,  and  depends  upon  the  porosity  of  the  soil  in  which 
the  water  is  moving. 

All  of  these  factors  have  a  bearing  upon  the  dissolved  constitu- 
ents which  will  be  found  in  an  underground  water.  The  sus- 
pended matter  of  the  surface  water  will  have  been  filtered  out, 
whether  in  a  relativeh'  coarse  form  like  sand  and  silt,  or  in  a 
finer  or  colloidal  condition  like  finely  divided  clay.  On  the  other 
hand,  the  solution  of  mineral  compounds  containing  sodium, 
potassium,  calcium,  magnesium,  ii'on,  aluminum,  silica,  etc., 
will  have  begun,  and  will  continue  as  long  as  the  water  does  not 
become  supersaturated,  and  until  "a  chemical  system  of  balanced 
values"^  is  attained  for  the  various  salts  with  which  the  water  is 
in  contact. 

CONTAMINATION  OF  WATER 

Natural  Impurities. — For  convenience  the  impurities  in  water 
may  be  divided  into  classes  according  to  the  sources  from  which 

^  Chase  Palmer:  "The  Geochemical  Interpretation  of  Water  Analyses." 
U.  S.  Geological  Survey  Bull.  479. 


CLASSIFICATION  OF  NATURAL  WATERS         U 

they  are  derived.  As  has  been  previously  noted  water  in  the 
course  of  its  flow  over  the  surface  or  through  the  ground  dis- 
solves more  or  less  material  as  well  as  carrying  varying  amounts 
in  suspension.  Provided  no  contact  with  the  wastes  of  human 
life  or  activity  has  occurred,  the  acquired  impurities  might  be 
designated  as  natural  impurities,  in  distinction  from  those 
derived  from  sewage  or  manufacturing  wastes.  Water  polluted 
from  the  latter  sources  is  unfit  for  human  consumption  without 
purification  as  will  be  shown  in  a  later  chapter.  Water  which 
has  not  been  subjected  to  such  contamination  may  or  may  not 
be  fit  for  drinking,  depending  entirely  on  the  nature  of  the 
dissolved  and  suspended  impurities  which  it  may  contain. 

Dissolved  Impurities. — The  solvent  action  of  water  upon  the 
seemingly  insoluble  constituents  of  the  earth's  crust  is  enormous, 
aside  from  its  erosive  action.  The  surface  and  subsoil,  as  well 
as  the  deeper  rock  strata,  are  all  subjected  to  the  decomposing 
action  of  the  water  with  which  they  come  in  contact.  Erosion 
assists  the  solution  of  the  mineral  compounds  composing  the 
various  strata  by  reducing  them  to  a  finely  divided  state. 

Disintegration  of  minerals  Kke  orthoclase,  for  example,  is 
able  to  supply  the  bases  sodium  and  potassium;  dolomitic  lime- 
stones may  furnish  calcium  and  magnesium;  and  clay  ironstone 
and  pyrite  may  give  up  iron,  aluminum  and  silica.  The  acids 
with  which  the  above  bases  are  usually  associated  are  carbonic, 
hydrochloric  and  sulphuric  acids.  These  latter  are  derived  from 
the  disintegration  of  the  minerals  in  the  rock  strata  the  same  as 
are  the  bases.  J\Iore  or  less  carbonic  acid  is  obtained  directly 
from  the  air  and  from  the  oxidation  of  carbonaceous  compounds 
in  the  breaking  down  of  organic  matter.  Minute  quantities  of 
nitric  acid  may  be  derived  directly  from  the  air,  and  from  oxida- 
tion of  nitrogenous  compounds.  The  oxidation  of  sulphur  in 
minerals  supplies  sulphuric  acid  in  many  cases.  Sulphates, 
chlorides,  nitrates,  carbonates  and  silicates  exist  as  such  in  the 
earth's  crust  in  enormous  amounts,  and  whenever  they  come  in 
contact  with  water  are  dissolved  directly,  and  in  amounts 
depending  upon  the  supply  available  and  the  solubility  of  the 
salt  being  acted  upon. 

A  classification  of  natural  waters  with  respect  to  their  dissolved 
constituents,  which  is  based  upon  the  chemical  nature  and  the 
proportional  amounts  of  the  radicles  present  in  solution,  has 
been  suggested  by  Chase  Palmer  in  a  paper  entitled  "The  Geo- 


12  WATER  PURIFICATION 

chemical  Interpretation  of  Water  Analyses."'     He  states  that: 

"Nearly  all  terrestrial  waters  have  two  general  properties,  salinity  and 
alkalinity,  on  whose  relative  proportions  their  fundamental  characters 
depend.  Salinity  is  caused  by  salts  that  are  not  hydrolyzed;  alkalinity 
is  attributed  to  free  alkaline  bases  produced  by  the  hydrolytic  action 
of  water  on  solutions  of  bicarbonates  and  on  solutions  of  salts  of  other 
weak  acids." 

According  to  his  classification  waters  may  possess  five  special 
properties  as  follows: 

1.  Primary  salinity;  that  is  sahnity  caused  by  the  sulphates 
and  chlorides  of  the  alkalies  sodium  and  potassium. 

2.  Secondary  salinity;  that  is  sahnity  produced  by  the  sul- 
phates and  chlorides  of  the  alkahne  earths  calcium  and  mag- 
nesium, or  in  other  words  permanent  hardness. 

3.  Tertiary  sahnity;  that  is  in  reahty  an  "  acidity  "  arising  from 
an  excess  of  saline  compounds  over  and  above  that  due  to  primary 
and  secondary  salinity. 

4.  Primary  alkalinity;  that  is  an  alkalinity  produced  by  car- 
bonates and  bicarbonates  of  sodium  and  potassium,  or  in  other 
words  permanent  alkalinity. 

5.  Secondary  alkalinity;  that  is  the  property  caused  by  alkaline 
earth  bicarbonates  such  as  those  of  calcium  and  magnesium,  or 
in  other  words  temporary  hardness. 

By  various  combinations  of  these  properties  five  well-defined 
classes  of  waters  are  possible  and  are  found  in  nature. 

Class  1.  "Waters  characterized  by  properties  1,  4  and  5  as 
stated  above. 

Class  2.  Those  showing  properties  numbered  1  and  5. 

Class  3.  Those  exhibiting  properties  numbered  1,  2  and  5. 

Class  4.  Those  marked  by  properties  numbered  1  and  2. 

Class  5.  Those  possessing  properties  numbered  1,  2  and  3. 

"Surface  waters  appear  to  belong  chiefly  to  the  first  three 
classes,  class  4  is  represented  by  sea  water  and  brines,  class  5 
is  exemphfied  by  mine  (acid)  waters  and  waters  of  volcanic 
origin." 

The  less  there  are  of  dissolved  impurities  in  a  water,  the  better 
it  is  fitted  for  a  pubhc  supply.  Consequently,  water  supplies 
are  commonly  rated  as  "soft,"  "hard,"  or  "alkaline,"  although 
these  classes  may  merge  into  one  another,  depending  upon  the 

lU.  S.  Geological  Survey  Bull.  479,  1911, 


CLASSIFICATION  OF  NATURAL  WATERS         13 

earth  strata  with  which  they  have  come  in  contact.  The  "soft " 
waters  are  ahnost  always  surface  waters,  containing  Httle  mineral 
matter.  This  dissolved  mineral  matter  is  about  equally  divided 
between  sihca,  alkahne  earth  salts,  and  alkali  salts,  although  the 
sihca  is  frequently  greater  than  the  other  two  classes  of  com- 
pounds. The  acid  combining  with  the  bases  is  largely  carbonic 
acid,  and  there  are  usually  about  equal  amounts  of  sulphates 
and  chlorides  present. 

The  "hard"  waters  derive  most  of  their  mineral  matter  from 
limestone.  Sihca  and  the  alkalies  are  usually  present  in  small 
quantities,  and  may  be  in  no  larger  amounts  than  are  found  in 
"soft"  waters.  Sulphates  and  chlorides  are  generally  present 
in  small  amounts,  except  where  the  water  has  come  in  contact 
with  gypsum  deposits  or  with  mine  drainage. 

The  "alkali"  waters  are  quite  variable  in  composition  and  few 
are  fitted  to  be  used  as  public  suppHes.  They  generally  contain 
large  quantities  of  sodium  sulphate  and  chloride;  some  may 
contain  large  amounts  of  calcium  and  magnesium,  and  others 
practically  none.  Some  of  these  waters  contain  little  dissolved 
mineral  matter  other  than  sodium  bicarbonate. 

The  average  hardness  of  water  used  for  public  suppHes  in  over 
300  cities  of  the  United  States  is  graphically  shown  on  the 
accompanying  map.  It  clearly  indicates  the  widely  varying 
character,  as  regards  hardness,  of  the  public  water  supplies  in 
the  various  states.  Waters  from  0  to  55  parts  per  million  of 
hardness  are  regarded  as  quite  "soft,"  while  those  ranging  from 
56  to  100  parts  per  million  are  noticeably  harder,  yet  are  not 
considered  as  being  objectionable.  However,  waters  ranging 
from  101  to  201  parts  per  million  of  hardness  become  less  and 
less  acceptable  as  they  approach  the  upper  limit  of  this  class;  and 
the  waters  in  the  201  to  500  parts  per  million  class  are  decidedly 
objectionable  as  pubhc  supplies  without  treatment.  Some  of 
them  are  too  impregnated  with  mineral  matter  to  attempt  to 
purify  them  by  artificial  means. 

Suspended  Impurities. — Intermediate  between  a  solution,  as 
it  is  commonly  understood,  and  a  suspension  of  finely  divided 
particles,  hke  sand  for  example,  there  may  be  so-called  colloidal 
suspensions  of  certain  substances  in  water  which  possess  peculiar 
properties,  and  which  are  of  considerable  importance  in  connec- 
tion with  water-purification  problems.  For  example,  silica  is 
found  in  this  state  in  many  natural  waters,  especially  those 


14 


WA  TER  P I  'RIFICA  TION 


CLASSIFICATION  OF  NATURAL  WATERS         15 

showing  primary  alkalinity,  i.e.,  waters  containing  sodium  and 
potassium  carbonates.  The  silica  compounds  characteristic 
of  the  clays  show  a  marked  tendency  toward  the  colloidal  state, 
and  render  the  problem  of  the  purification  of  turbid  waters  of 
this  class  almost  a  problem  in  itself. 

The  enormous  amount  of  the  heavier  sediment  carried  by 
many  rivers  consists  largely  of  sand  and  clay.  Any  reduction 
in  the  velocity  of  flow  of  the  water  laden  with  such  sediment 
causes  it  to  be  deposited  on  the  bed  of  the  stream.  Gravel  and 
sand  bars  are  thus  formed  in  river  beds,  which  may  be  shifted 
from  one  point  to  another  by  any  sudden  increase  in  the  velocity 
of  the  current,  such  as  might  be  produced  by  a  flood.  Lighter 
material  like  clay  is  more  slow^ly  deposited  and  more  quickly 
moved  again  by  any  change  in  the  rate  of  flow  of  the  water.  It 
thus  happens  that  almost  all  surface  waters  carry  var3dng 
amounts  of  suspended  material  in  them  which  depend  upon  the 
velocity  of  the  currents  of  water  and  on  the  character  of  the 
bottom  and  shores.  The  action  of  the  wind  on  large  bodies  of 
water  like  the  Great  Lakes,  for  example,  may  stir  up  the  sediment 
on  the  bottom  and  render  the  water,  near  the  shore  especially, 
quite  turbid. 

Compounds  of  iron  and  of  manganese  are  not  infrequently  met 
with  in  the  colloidal  state  in  natural  waters,  and  offer  some  of  the 
most  interesting  phases  of  water-purification  work.  Organic 
matter  found  in  natural  waters  is  probably  always  present  in  this 
form  to  a  greater  or  less  degree.  Vegetable  stain  produced  by 
humic  substances  is  a  marked  characteristic  of  a  very  large  class 
of  natural  surface  waters,  and  as  such  has  received  considerable 
attention  in  the  study  of  purification  problems. 

How  the  removal  of  these  impurities,  whether  in  solution  or  in 
suspension,  is  effected  will  be  discussed  under  the  description 
of  the  various  methods  of  purification  now  employed,  rather  than 
in  this  place.  It  is  only  desired  to  emphasize  the  varied  classes 
of  impurities  which  may  be  found  in  natural  waters,  and  which 
are  virtually  "natural  impurities,"  as  distinguished  from  those 
derived  from  sewage  and  manufacturing  wastes.  This  waste 
material  may  furnish  similar  classes  of  polluting  compounds  to 
those  derived  naturally,  but  their  origin  usually  justifies  their 
consideration  separately. 

Whether  the  dissolved  salts  usually  found  in  natural  waters 
are  objectionable  depends  upon  the  use  to  which  the  water  is 


16  WATER  PURIFICATION 

to  be  put.  A  certain  amount  of  the  chlorides,  sulphates  and 
carbonates  of  sodium,  potassium,  calcium  and  magnesium  are 
by  no  means  deleterious,  and  may  possibly  be  beneficial  in  a 
drinking  water.  On  the  other  hand,  if  too  great  amounts  are 
present  these  dissolved  salts  render  the  water  unfit  for  domestic 
and  industrial  uses,  and  actually  cause  financial  losses  of  no  small 
amount  to  those  obliged  to  use  them.  Compounds  of  iron  or  of 
manganese  are  especially  objectionable,  and  not  infrequently 
have  caused  waters  containing  them  to  be  abandoned  as  sources 
of  supply.  Excessive  amounts  of  the  fixed  alkalies  either  as 
salts  of  the  strong  acids  like  hydrochloric  and  sulphuric,  or  of 
the  weak  acids  like  carbonic,  make  a  water  unsuitable  as  a  public 
supply. 

Microscopic  Plant  and  Animal  Life. — In  many  of  our  natural 
waters  a  luxuriant  growth  of  algae  and  diatoms  is  found  at  certain 
seasons  of  the  year.  The  character  of  the  mineral  and  organic 
constituents  of  the  water,  as  well  as  the  conditions  of  light  and 
temperature  materially  affect  the  extent  of  these  growths.  They 
occur  in  both  still  and  running  water.  Accompanying  the  growth 
and  also  the  decay  of  certain  of  these  organisms  bad  odors  and 
tastes  are  not  infrequently  developed,  and  where  this  occurs 
in  public  water  supphes,  they  become  a  nuisance  entirely  out  of 
proportion  to  their  number  and  size.  Certain  microscopic 
animal  forms  may  also  produce  troubles  of  this  same  character. 

The  odor  of  growth  appears  to  be  due  to  secretions  of  an  oil- 
like character  which  they  produce,  and  is  usually  somewhat 
characteristic  of  the  special  organism  producing  it.  When  the 
organisms  are  in  large  enough  numbers  they  are  capable  of  giving 
an  odor  to  large  volumes  of  water,  and  not  infrequently  spoil 
the  taste  and  odor  of  the  whole  of  a  public  water  supply.  Odors 
of  decomposition  are  usually  very  offensive,  and  notably  so  in 
the  case  of  the  "blue  green  alga?"  or  Cyanophycea. 

It  is  not  probable  that  impurities  of  this  nature  in  natural 
waters  produce  disease  in  human  beings,  when  swallowed  in 
drinking  water.  They  are  very  objectionable  if  they  produce 
a  marked  odor,  and  in  such  cases  are  most  frequently  complained 
of  in  pubUc  water  supplies. 

Bacteria.— Even  lower  in  the  scale  of  plant  life  than  the 
diatoms  and  the  algae  are  found  the  bacteria.  They  are  present 
in  all  natural  waters,  being  the  more  numerous  in  surface  waters, 
and  much  less  so  in  ground  waters.     By  far  the  larger  number  of 


CLASSIFICATION  OF  NATURAL  WATERS         17 

the  various  species  of  bacteria  play  a  beneficient  role  in  the 
economy  of  nature,  and  appear  absolutely  essential  to  many  of 
the  normal  processes  of  development  of  both  plants  and  animals. 
A  few  species,  however,  are  associated  with  disease  in  animals 
and  in  human  beings.  So  far  as  the  contamination  of  water  is 
concerned,  it  is  only  the  organisms  capable  of  producing  patho- 
logic conditions  in  man  that  are  of  interest.  These  virulent 
forms  usually  reach  our  natural  waters  through  the  medium  of 
domestic  sewage  and  manufacturing  wastes.  This  class  of 
impurities  is  considered  in  more  detail  in  the  next  section. 

IMPURITIES  DERIVED  FROM  WASTE  MATERIAL 

Sewage, — The  water  carriage  of  waste  material  of  human  and 
animal  origin  has  become,  in  those  countries  which  pay  any  atten- 
tion to  problems  of  sanitation,  the  most  common  method  for 
its  transfer  to  some  point  of  ultimate  disposal.  Wherever 
public  water  supplies  are  installed,  a  system  of  sewers  will  of 
necessity  follow.  Hence  the  disposal  of  large  volumes  of  fouled 
water  has  become  a  problem  of  great  difficulty,  and  one  that  yet 
awaits  a  completely  satisfactory  solution. 

It  is  obvious  that  some  method,  even  though  it  is  not  entirely 
satisfactory,  must  be  used  to  get  rid  of  this  polluted  water  or 
sewage,  and  the  easiest  way  has  been  to  turn  it  into  the  natural 
water  courses.  In  this  manner  much  of  the  surface  water  on 
thickly  settled  land  areas  has  become  polluted  with  material 
dangerous  to  the  health  of  human  beings,  who  unwittingly  or  of 
necessity  drink  the  water  thus  contaminated.  Since  disease  has 
been  found  to  originate  so  largely  from  specific  plant  and  animal 
forms,  microscopic  in  size,  which,  having  produced  the  disease, 
are  discharged  from  the  body  chiefly  in  the  excreta  and  the  urine, 
the  conveyance  of  disease  through  sewage  to  water  has  been 
pretty  definitely  proven. 

Ground  waters  as  well  as  surface  waters  may  become  polluted 
by  sewage.  The  discharge  of  sewage  on  the  surface  of  the  ground 
or  into  cesspools,  or  the  leakage  or  overflow  of  vaults,  may  furnish 
the  dangerous  pollution  to  well  and  spring  waters  by  direct 
percolation  through  fissures  in  the  rock  strata,  or  by  more 
indirect  routes  through  the  soil  itself. 

Manufacturing  Wastes. — In  many  industries  there  remains 
after  the  mamifactured  product  has  been  completed,  a  great 
deal  of  waste  material,  which  for  economic  reasons  it  is  not  worth 


18  WATER  PURIFICATION 

while  to  work  over.  Much  of  this  material  is  in  suspension  and 
solution  in  relatively  large  volumes  of  water.  Its  disposal  by 
the  easiest  method  is  to  dump  it  into  the  nearest  body  of  water. 
Water  fouled  with  such  material  is  totally  unfit  for  human  con- 
sumption. Frequently  the  material  renders  even  the  best 
methods  for  the  purification  of  domestic  sewage  inadequate,  and 
its  proper  disposal  becomes  a  special  problem  in  almost  every 
case. 

The  waste  liquids  from  textile  works,  dye  works,  straw-board 
factories,  paper  mills,  abattoirs,  meat  packing  establishments, 
dairies,  etc.,  furnish  material  which  is  obviously  difficult  to 
dispose  of,  and  which  must  pollute  in  the  foulest  manner  any 
natural  water  into  which  they  may  be  turned. 

NATURAL  METHODS  OF  PURIFICATION 

Sedimentation. — Some  of  the  impurities  which  a  natural  water 
acquires  in  the  course  of  its  flow  may  be  lost  under  certain  favor- 
able conditions.  For  example,  a  water  laden  with  suspended  clay 
or  fine  sand  will  deposit  this  material  as  soon  as  the  velocity  of 
the  water  is  sufficiently  retarded  as  was  previously  explained. 
This  process  of  sedimentation  is  one  of  the  most  important  of 
the  natural  methods  of  purification,  and  plays  an  important  part 
in  our  artificial  methods  as  well.  In  flowing  streams  the  deposi- 
tion of  sediment  is  intermittent,  being  active  during  low  stages 
of  the  stream  when  the  rate  of  flow  is  relatively  low,  and  much 
diminished  or  practically  nil  in  flood  periods.  At  such  times  the 
scouring  action  of  the  current  causes  much  that  has  been 
deposited  to  be  again  placed  in  suspension,  and  thus  carried 
further  toward  its  ultimate  disposal  in  the  sea.  In  this  way  the 
immense  deltas  at  the  mouths  of  rivers  like  the  Mississippi  and 
the  Nile  are  formed. 

Effect  of  Sunlight. — The  purifying  action  of  sunlight  on  certain 
vegetable  compounds  in  colloidal  suspension,  such  as  the  brown 
coloring  matter  in  many  of  the  streams  and  lakes  in  the  north 
central  and  northeastern  parts  of  the  United  States,  is  worth 
mentioning  in  this  connection.  A  certain  amount  of  bleaching 
out  of  this  coloring  matter  is  apparently  effected  when  this  class 
of  waters  are  impounded  in  natural  lakes  or  artificial  reservoirs. 
Oxidation  of  the  carbonaceous  matter  probably  occurs,  and 
sedimentation  in  the  quiet  water  undoubtedly  assists  in  the 
clarification. 


CLASSIFICATION  OF  NATURAL  WATERS         19 

Precipitation  of  Compounds  from  Solution. — Dissolved  salts 
are  not  usually  readily  removed  once  they  have  gone  into  solu- 
tion. The  chlorides,  sulphates  and  nitrates  of  either  sodium, 
potassium,  ammonium,  calcium  or  magnesium  will  be  retained 
on  account  of  their  great  solubility.  Carbonates  and  bicar- 
bonates  of  the  fixed  alkalies,  as  well  as  of  ammonia,  are  also 
very  soluble.  On  the  other  hand,  the  bicarbonates  of  the 
alkahne  earths  have  a  rather  limited  solubihty  and  may  be 
deposited  from  solution  if  the  excess  of  carbon  dioxide,  which  is 
necessary  for  their  retention  in  solution,  is  in  any  way  removed. 
Ground  waters  in  particular  may  become  heavily  charged  wdth 
bicarbonates  and,  on  being  brought  to  the  surface  where  the 
pressure  is  diminished,  wall  lose  some  of  their  free  carbon  dioxide 
and  deposit  their  monocarbonates,  which  are  much  less  soluble. 
This  is  particularly  true  of  calcium  carbonate.  Magnesium 
carbonate,  however,  is  considerably  more  soluble. 

Oxidizable  salts  hke  ferrous  sulphate,  ferrous  carbonate,  and 
corresponding  salts  of  manganese  occurring  in  ground  waters 
may  be  deposited  from  solution  on  exposure  to  the  air.  Such 
purification  can  be  hastened  by  aeration  and  thus  render  some 
unsuitable  deep  well  waters  entirely  acceptable  as  a  source  of 

water  supply. 

"Acid  mine  waters"  are  usually  contaminated  with  dissolved 
iron  compounds,  which  not  infrequently  find  their  way  into  sur- 
face streams.  As  much  of  this  iron  may  be  in  an  unoxidized 
state,  the  exposure  to  the  oxygen  of  the  air,  and  to  that  dissolved 
in  the  surface  water,  soon  converts  the  iron  to  the  form  of  the 
insoluble  ferric  oxide.  Organic  matter  in  colloidal  suspension 
may  retard  the  precipitation  of  the  iron,  and  cause  the  latter  to 
assume  a  colloidal  state  itself. 

In  waters  in  which  the  alkahnity  is  due  to  sodium  and  potas- 
sium carbonates,  colloidal  solutions  of  silica  and  alumina  are 
sometimes  found,  and  on  account  of  their  slight  solubihty  may 
be  deposited,  should  this  "primary  alkalinity"  be  diminished. 
Such  a  diminution  can  be  effected  if  waters  of  this  class  come 
into  contact  with  chlorides  and  sulphates  of  hme  and  magnesia. 
These  latter  salts  will  react  with  the  fixed  alkaline  carbonates, 
forming  carbonates  of  lime  and  magnesia  and  the  chlorides  and 
sulphates  of  sodium  and  potassium.  The  latter  salts  are  without 
power  to  assist  in  holding  the  sihca  in  solution. 


20  WATER  PURIFICATION 

Filtration. — The  natural  filtration  of  water  through  the  soil 
effects  a  high  degree  of  purification  provided  the  ground  is  of  the 
right  character.  Sand  and  gravel,  when  not  too  coarse,  afford 
an  excellent  purifying  medium.  Suspended  impurities,  organic 
matter  and  oxidizable  salts  are  removed  as  a  result  of  the  strain- 
ing action,  and  the  chemical  and  biological  changes  induced 
during  filtration.  The  action  of  both  sedimentation  and  filtra- 
tion in  purifjang  natural  waters  is  perfectly  normal,  and  one 
which  is  constantly  going  on.  To  these  agencies  we  owe  the 
potability  of  most  of  our  ground  waters,  and  by  a  study  of  the 
principles  underlying  these  natural  processes  we  have  been  able 
to  design  and  operate  our  modern  water-purification  plants. 

PURIFICATION    BY    MEANS    OF    MINUTE    PLANT    AND    ANIMAL 

ORGANISMS 

Thus  far  in  considering  natural  methods  of  purification  only 
those  agencies  have  been  especially  noted  which  are  effective 
without  the  intervention  of  organized  plant  and  animal  life. 
In  the  cycle  through  which  inert  mineral  matter  passes  into 
organized  matter,  and  then  back  again  into  inorganic  compounds, 
life  in  some  of  its  most  marvellous  forms  plays  an  important  and 
essential  part.  These  forms  belong  both  to  the  animal  and 
vegetable  kingdom,  and  for  the  most  part  are  microscopic  in 
size.  The  borderland  between  plant  and  animal,  in  these  almost 
invisible  organisms,  is  extremely  ill  defined;  but  no  matter 
how  classified  their  importance  in  the  economy  of  nature  is 
fundamental. 

All  living  organisms  which  float  about  in  water  between  the 
surface  and  the  bottom  are  designated  by  biologists  as  "plank- 
ton." They  are  moved  about  by  the  currents  and  the  wind 
chiefly,  although  they  have  shght  powers  of  locomotion.  They 
also  possess  the  peculiar  ability  to  remain  suspended  in  the  water 
with  little  effort  on  their  part.  The  plankton  can  be  divided 
into  two  general  classes: 

"the  food  producers  or  plants,  which  assimilate  inorganic  matter  and 
build  up  organic  compounds  by  means  of  their  chromophyll  coloring 
matter;  and  the  food  consumers  or  animals,  such  as  the  microscopic 
protozoa,  rotifera,  etc.,  together  with  the  larger  ones  up  to  the  fishes." 


CLASSIFICATION  OF  NATURAL  WATERS         21 

In  the  lecture  from  which  the  above  quotation  was  cited,  Dr. 
Marsson^  concisely  epitomizes  the  relations  of  these  two  groups 
by  stating  that  the 

"vegetable  component  of  the  plankton  is  the  fundamental  food  supply 
or  condition  of  existence  for  all  aquatic  life.  It  comes  from  the  prod- 
ucts of  the  decomposition  of  the  albumen  which  finds  its  way  into  the 
water  from  decaying  animals  and  plants,  as  well  as  from  sewage.  The 
self-purifjdng  power  of  natural  waters  is  merely  the  maintenance 
of  the  proper  equilibrium  between  retrogressive  and  progressive 
metamorphosis." 

The  groups  of  microscopic  plant  forms  known  as  the  algse, 
diatoms,  fungi  and  bacteria  exist  in  enormous  numbers  in  all 
natural  surface  waters,  and  to  some  extent  in  ground  waters. 
The  algse  and  diatoms  through  their  pecuhar  cellular  structure 
are  hving  laboratories  in  which  light  is  the  energy  which  tears 
the  carbon  from  carbonic  acid,  and  the  nitrogen  from  its  simpler 
compounds  and  convert  them  into  starch,  sugar  and  albumen. 
Thus  oxygen  is  liberated  and  becomes  available  for  oxidizing 
organic  matter  and  preventing  putrefactive  changes. 

The  fungi  and  bacteria  find  their  nutriment  in  dead  organic 
matter,  and  are  the  primary  agents  for  its  decomposition  into 
simpler  compounds.  Bacterial  activity  is  associated  with  the 
using  up  of  large  amounts  of  oxygen,  where  the  latter  is  available; 
and  in  such  cases  non-putrefactive  disposal  of  contaminating 
impurities  in  water  is  in  process  in  distinction  to  putrefactive 
changes  where  the  oxygen  is  not  present. 

The  bacteria  are  the  natural  food  for  many  of  the  microscopic 
animal  organisms.  The  latter  include  the  protozoa,  infusoria 
and  metazoa,  and  where  these  organisms  are  found  in  abundance, 
bacteria  and  food  for  bacteria  will  also  be  present.  They  thus 
become  indexes  of  pollution  in  water,  quite  as  indicative  as  the 
bacteria  themselves.  When  the  food  supply  is  gone  they  must 
die  also. 

Since  the  smaller  plant  and  animal  organisms  are  the  source 
of  food  for  the  fishes,  and  they  in  turn  for  human  beings,  the 
cycle  of  matter  from  man  through  human  wastes  to  mineralized 

1  Max  Marsson:  "The  Significance  of  Flora  and  Fauna  in  Maintaining 
the  Purity  of  Natural  Waters,  and  How  They  are  Affected  by  Domestic 
Sewage  and  Industrial  Wastes."  Eng.  News,  Aug.  31,  1911.  Trans,  by 
Emil  Kuichlixg. 


22  WATER  PURIFICATION 

compounds  and  back  again  to  man  is  complete.  The  natural 
methods  of  self-purification  of  water  are  going  on  ceaselessly  and 
effectively,  but  the  agencies  ordained  for  this  purpose  must  have 
time  and  opportunity  to  do  their  work. 

References 

1.  United  States  Geological  Survey: 

(a)  Water  Supply  Papers: 

"Conservation  of  Water  Resources."     No.  234,  1909. 

Herman  Stabler:  "Some  Stream   Waters  of  the   Western   United 

States."     No.  274,  1911. 

R.  B.  Dole:  "The  Quality  of  Surface  Waters  in  the  United  States." 

No.  236. 

Herman  Stabler  and  Gilbert  H.  Pratt:  "The  Purification  of  Some 

Textile  and  Other  Factory  Wastes."     No.  235,  1909. 

(b)  Water  Supply  and  Irrigation  Papers: 

D.  D.  Jacksox:  "The  Normal  Distribution  of  Chlorine  in  the  Natural 

Waters  of  New  York  and  New  England."     No.  144,  1905. 

Herman   St.\bler:  "Prevention   of   Stream   Pollution   by    Distiller}- 

Refuse."     No.  179,  1906. 

Herman  St.\bler:  "Stream  Pollution  by  Acid  Iron  Wastes."     No. 

186,  1906. 

(c)  Bulletins: 

Chase  Palmer:  "The  Geochemical  Interpretation  of  Water  Analysis." 
Bull.  479,  1911. 

2.  United  States  Department  of  Agriculture: 

Office  of  Public  Roads: 

A.  S.  Cushman:  "A  Study  of  Rock  Decomposition  under  the  Action 

of  Water."     Cir.  38. 

Bulletins: 

A.  S.  Cushman:  "The  Effect  of  Waters  on  Rock  Powders."     Bull. 

92,  1905. 

Bureau  of  Chemistry: 

J.  K.  Haywood  and  B.  H.  Smith:  "Mineral  Waters  of  the  United 

States."     Bull.  91,  1907. 

3.  Mass.  State  Board  of  Health  Report  for  1892: 

T.  M.  Drown:  "On  the  Mineral  Constituents  of  Some  Natural  Waters 
in  Massachusetts." 

4.  Engineering  News: 

Dr.  M.\x  Marsson:  "The  Significance  of  Flora  and  Fauna  in  Main- 
taining the  Purity  of  Natural  Waters,  and  How  Thej'  Are  Affected  by 
Domestic  Sewage  and  Industrial  Wastes."  Aug.  31,  1911.  Trans,  by 
Emil  Kuichling.  J.  D.  Watson:  "Pollution  of  the  River  Taine." 
Feb.  8,  1912. 

5.  Journal  New  England  Water  Works  Association: 

G.  C.  Whipple  and  D.  D.  Jackson:  " Asterionella:  Its  Biology,  Its 
Chemistry  and  Its  Effect  on  Water  Supplies."     Vol.  14,  No.  1. 


CLASSIFICATION  OF  NATURAL  WATERS         23 

F.  S.  HoLLis  and  H.  N.  Parker:  " Chlamydomonas  in  Spot  Pond 
(Mass.)."     Vol.  14,  No.  1. 

R.  S.  Weston  :  ' '  The  Occurrence  of  Cristatella  in  the  StorageReservoirs 
at  Henderson,  N.  C."     Vol.  13,  No.  1. 

T.  M.  Drown:  "Odor  and  Color  of  Surface  Waters."     Vol.  2,  No.  3, 
1888. 
G.   Transactions  American  Microscopical  Society: 

D.  D.  Jackson:  "A  New  Species  of  Crenthroix  (C.  Manganifera)." 
Vol.  23,  May,  1902. 

G.  C.  Whipple  and  H.  N.  Parker:  "On  the  Amount  of  Dissolved 
Oxygen  and  Carbonic  Acid  Dissolved  in  Natural  Waters,  and  the 
Effect  of  These  Gases  upon  the  Occurrence  of  Microscopic  Organ- 
isms."    Vol.  24,  May,  1902. 

H.  N.  Parker:  "Notes  on  the  Growth  of  Synura  in  Lake  Cochituate, 
Mass."     Vol.  30,  No.  2,  April,  1911. 

7.  American  Naturalist: 

D.  D.  Jackson:  "Movements  of  Diatoms  and  Other  Microscopic 
Plants."     Vol.  23,  No.  461,  1905. 

8.  Technology  Quarterly: 

D.  D.  Jackson  and  J.  W.  Ellms:  "Odors  and  Tastes  of  Surface  Waters 
with  Especial  Reference  to  Anabaena."  Vol.  10,  No.  4,  December, 
1897. 

9.  Geo.  C.  Whipple:  "Microscopy  of  Drinking  Water." 

10.  Surveyor: 

"Algae  and  Water  Supplies."     Aug.  25,  1911. 

J.\mesScott:  "TheChara:  A  Water-purifying  Plant."     Aug.  25,  1911. 

11.  Proceedings  Engineers'  Society  of  Western  Pennsylvania: 

T.  P.  Roberts:  "Acids  in  the  Monongahela  River."     November,  1911. 

12.  E7ig.  Cont.: 

Thorndyke  Saville:  "The   Nature  of  Color  in  Water."     January 

10,  1917. 

13.  W.  D.  Collins:  Water  Supply  Paper  No.  496,  U.  S.  Geological  Survey, 

1923. 

14.  "A  Study  of  the  Pollution  and  Natural  Purification  of  the  Ohio  River." 

Public  Health  Bull.  143,  U.  S.  PubUc  Health  Service. 


CHAPTER  III 

TRANSMISSION  OF   DISEASE  THROUGH  DRINKING 

WATER 

The  discharge  of  sewage  and  waste  material  of  all  kinds  into 
the  streams  and  lakes  obviously  affords  ample  opportunit}'  for 
disease-producing  organisms  to  enter  the  sources  of  most  of  our 
public  water  supplies.  Those  diseases  peculiar  to  the  intestinal 
tract  of  the  human  body  are  the  ones  most  likely  to  be  dis- 
seminated in  consequence  of  this  common  practice;  and  hence, 
typhoid  fever,  cholera,  dysentery  and  gastro-intestinal  disturb- 
ances have  come  to  be  regarded  as  derived  in  a  large  measure 
from  polluted  drinking  water,  wherever  these  diseases  are  endemic 
or  even  epidemic.  Anyone  or  all  of  these  diseases  may  be  trans- 
mitted in  other  ways,  but  where  the}'  are  widespread,  some  com- 
mon carrier  of  infection  is  generally  found  to  be  the  source,  and 
a  common  drinking-water  supply  usually  offers  the  most  favorable 
opportunity,  for  transmitting  the  disease. 

Probably  most  of  the  diseases  transmitted  by  water  are  of 
bacterial  origin.  The  "spirillum  cholera "  of  Asiatic  cholera, 
the  ''bacillus  typhosus"  of  typhoid  fever,  and  the  "bacillus 
dysenteria3"  of  dysentery  have  all  been  found  in  contaminated 
drinking  water.  Pathogenic  protozoa  ma}'  also  produce  certain 
diseases,  and  in  the  case  of  one  form  of  dysentery,  an  amoeba 
is  known  to  be  the  cause.  If  the  theory  advanced  by  Sedgwick 
and  IVIacNutt,  that  inflammatory  diseases  of  the  respiratory 
organs  may  be  also  to  some  extent  water-borne,  is  accepted,  then 
polluted  water  supplies  are  chargeable  with  another  group  of 
diseases  particularly  prevalent  among  all  classes  of  people. 

The  virility  of  disease-producing  organisms  upon  their  entrance 
to  a  water  is  of  importance  with  respect  to  the  real  danger  which 
they  possess.  This  ability  to  live  and  retain  their  vitality  in  a 
medium  foreign  to  their  natural  habitat  is  also  of  consequence, 
for  if  the  power  to  reproduce  the  disease  is  soon  diminished  and 
eventually  destroyed,  then  the  length  of  time  before  their  vitality 
is  lost  is  of  the  utmost  importance.  Much  experimental  work 
has  been  done  to  determine  the  period  elapsing  before  certain 
pathogenic  organisms  die,  when  placed  in  water  under  varying 

24 


TRANSMISSION  OF  DISEASES  25 

conditions.  Laboratory  experiments  throw  some  light  on  this 
problem,  but  are  not  usually  conclusive,  because  of  the  artificial 
conditions  imposed.  In  almost  all  the  experimental  work  the 
number  of  organisms  diminish  in  time,  and  usually  very  rapidly. 
This  may  be  the  result  of  a  decreasing  food  supply,  or  to  toxic 
compounds  eliminated  in  the  course  of  growth  or  decay,  which 
kill  off  rapidly  the  less  resistant  organisms. 

The  presence  of  pathogenic  forms  in  a  drinking-water  supply 
denotes,  of  course,  all  the  possibilities  of  dangerous  infection. 
Nevertheless,  sanitarians  have  of  late  regarded  the  number  of 
such  organisms,  and  the  length  of  time  which  they  may  have 
been  in  the  supply,  as  factors  of  much  importance  in  the  epi- 
demiology of  disease.  This  quantitative  feature  is  of  consider- 
able significance  in  connection  with  the  qualit}^  of  a  water 
obtained  by  the  methods  commonly  employed  in  the  purification 
of  polluted  waters. 

Spiiillum  Cholerae  and  Bacillus  Typhosus.^ — The  cholera 
spirillum  and  the  typhoid  bacillus  are  the  pathogenic  organisms 
which  have  been  most  studied  in  water-borne  diseases.  It  is 
very  doubtful  whether  either  of  these  organisms  will  multiply 
outside  the  bod}',  or  in  impure  water.  The  cholera  spirillum 
is  not  very  resistant  to  adverse  conditions  outside  the  human 
bodJ^  It  is  killed  in  10  min.  by  a  temperature  of  60°C.,  easily 
destroyed  by  chemical  disinfectants,  and  does  not  long  retain 
its  vitality  in  association  with  the  ordinary  saprophj-lic  bacteria 
in  the  water. 

The  typhoid  bacillus  is  probably  more  resistant  than  the 
cholera  organism  to  outside  influences.  Laboratory  experiments 
have  demonstrated  that  the  typhoid  bacillus  will  live  in  sterile 
water  in  glass  vessels  for  3  months,  and  in  unsterilized  ground  and 
surface  waters  for  several  weeks.  Jordan'  showed  by  his  experi- 
ments with  typhoid  cultures  placed  in  sacks  of  collodion  and 
parchment  and  suspended  in  flowing  water,  that  they  would 
retain  their  vitality  under  natural  conditions  for  at  least  4  or 
5  day's.  Geo.  A.  Johnson's  experiments  at  Columbus  Ohio,- 
in  which  he  modified  Jordan's  technique,  showed  that  the 
ability  of  the  bacteria  to  pass  through  the  walls  of  the  parch- 
ment sacks,  might  indicate  that  conclusions  drawn  from  the 
disappearance    of   the  bacteria  in  Jordan's  experiments,   were 

^Jordan,  Russel  and  Zeit:  Jour.  Infect.  Diseases,  1,  p.  641,  1904. 
-Eng.  Record,  vol.  52.  Sept.  23,  1905. 


26  WATER  PURIFICATION 

somewhat  misleading.  If  the  organisms  actually  escaped  from 
the  sacks,  failure  of  samples  withdrawn  from  the  latter  to 
develop  typical  cultures,  did  not  necessarily  mean  that  the 
typhoid  bacilli  had  died.     Jordan  concludes/  that: 

"It  is  possible  that  water  may  continue  to  be  the  vehicle  of  infection 
during  a  much  longer  period  (than  4  or  5  days),  but  the  available  data 
point  to  a  comparatively  short  duration  of  life  of  the  specific  germ  in 
the  water  of  flowing  streams." 

Houston-  has  shown  that  samples  of  Thames  River  water 
inoculated  with  a  typhoid  emulsion  and  stored  at  temperatures 
ranging  from  32°F.  to  98.6°F.,  developed  negative  tests  for 
typhoid  in  9  weeks  at  the  low  temperature,  and  in  2  weeks  at 
the  highest  temperature.  Intermediate  temperatures  gave  nega- 
tive results  in  conformity  with  the  results  stated  above,  viz., 
the  higher  the  temperature  of  the  water,  the  shorter  the  period 
of  life  of  the  organism.  The  history  of  typhoid  epidemics  tends 
to  confirm  in  a  measure  the  data  obtained  in  these  experiments. 
It  emphasizes  the  protective  value  of  ample  periods  of  sedimenta- 
tion and  storage  of  polluted  waters  used  as  public  supplies. 

Isolation  of  Cholera  and  Typhoid  Organisms  from  Water. — 
The  actual  isolation  of  these  two  organisms  from  polluted  water 
has  been  accomplished  only  in  a  comparatively  few  well-authenti- 
cated cases.  In  the  case  of  cholera  the  organism  is  discharged 
from  the  intestines  in  enormous  numbers,  but  not  in  the  urine. 
Its  appearance  in  sewage  and  polluted  water  would,  therefore, 
be  expected,  and  has  been  demonstrated.  In  1892  Dunbar 
isolated  the  spirillum  of  cholera  from  the  polluted  water  of  the 
Elbe,  during  the  epidemic  in  Hamburg.  Koch^  also  reports  its 
isolation  from  the  water  of  two  Altona  reservoirs  supplied  also 
from  the  Elbe. 

The  isolation  of  the  typhoid  bacillus  from  natural  waters  also 
offers  a  great  deal  of  difficulty,  although  it  probably  is  more 
virile  and  capable  of  living  longer  in  natural  water  than  the 
cholera  organism.  The  bacilli  of  typhoid  fever  are  discharged 
from  the  human  body  both  in  the  urine  and  the  feces.  From 
9  to  14  days  after  infection  has  taken  place  are  required  before 
the  disease  fully  develops.  This  characteristic  feature  of  typhoid 
fever  makes  the  tracing  of  infection  through  natural  waters  much 

^  E.  O.  Jordan:  "General  Bacteriology." 

2  Eng.  Record,  vol.  65,  June  1,  1912,  p.  G08. 

^  Zeit.  fur  hygiene  und  Inject.  Krank.,  14. 


TRANSMISSION  OF  DISEASES  27 

more  difficult,  for  although  the  bacilli  may  have  been  present 
and  caused  the  disease,  they  will  have  probably  disappeared 
before  suspicion  is  thoroughly  aroused  as  to  their  possible  presence 
in  the  water.  Comparatively  few  cases  have  been  recorded, 
therefore,  in  which  the  bacillus  has  been  isolated  and  shown  to 
have  been  the  probable  cause  of  a  case  of  typhoid  fever. 

Other  Water-borne  Diseases. — Intestinal  diseases  and  some 
gastric  troubles  may  be  and  probably  frequently  are  caused  by 
organisms  found  in  water.  Among  infants  this  perhaps  is  truer 
than  with  adults.  Epidemics  of  diarrhea  and  dysentery  are  not 
uncommon  and  have  been  traced  to  impure  drinking  water. 
The  possible  infection  of  a  water  supply  by  anthrax  (B.  anthracis) 
derived  from  animals  sick  with  the  disease  produced  by  this 
organism  is  rather  remote,  but  not  impossible.  It  is  of  more 
theoretical  interest  than  practical  that  the  pathogenic  organisms 
B.  anthracis  and  B.  tetani^  have  both  been  isolated  from  river 
water  by  Zeit  and  Fiitterer;  but  it  goes  to  show  the  possibihties 
of  water-borne  infection.  Sewage-polluted  waters  may  contain 
all  known  pathogens  as  well  as  saprophytes,  and  what  role  the 
latter  forms  may  play  in  disease  is  by  no  means  a  settled  question. 

The  relation  between  pneumonia,  bronchitis  and  other  inflam- 
matory diseases  affecting  the  respiratory  organs  and  polluted 
drinking  water  has  been  noted  above.  The  ascertainable  facts 
relating  to  this  phase  of  water-borne  diseases  are  few  and  difficult 
to  satisfactorily  classify.  The  data  already  collected  by  Dr. 
W.  T.  Sedgwick  and  his  associates  are  extremely  valuable,  and 
further  confirmation  of  their  deductions  is  hoped  for. 

EPIDEMICS  OF  WATER-BORNE  DISEASES 

Cholera 

Probably  no  disease  is  more  truly  characterized  as  a  "filth 
disease"  than  is  cholera.  In  certain  parts  of  India  it  may  be 
said  to  be  endemic.  Explosive  outbreaks  are  not  uncommon,  and 
the  spreading  of  the  disease  by  contact  is  probably  constantly 
going  on.  The  insanitary  nature  of  the  personal  habits  of  the 
lower  classes  of  natives  affords  ample  opportunity  for  transmitting 
infection,  and  the  streams  and  lakes  frequently  serve  as  carriers. 

1  "Report  of  the  Sanitary  Investigation  of  the  lUinois  River  and  Its 
Tributaries."     IlUnois  State  Board  Health,  p.  85,  1900. 


28  WATER  PURIFICATION 

In  1817  a  violent  epidemic  of  cholera  broke  out  in  Jessore  in 
Bengal,  which  rapidly  spread  over  a  larger  part  of  British  India. 
It  continued  unabated  for  3  years,  and  then  began  to  spread  into 
China  and  Persia.  In  1823  the  disease  had  reached  Asia  Minor 
and  Russia.  For  the  next  7  years  it  did  not  advance  westward 
any  further,  but  a  fresh  outbreak  in  1830  in  Russia  caused  the 
disease  to  spread  all  ov^er  the  latter  country  and  into  northern 
Europe  and  the  British  Isles.  During  the  next  5  years  it  spread 
southward,  invading  northern  Africa. 

Another  epidemic  started  in  India  and  China  in  1841,  reaching 
Europe  in  1847;  another  began  in  1850  and  entered  Europe  in 
1853,  and  was  carried  across  the  Atlantic  to  North  and  South 
America,  where  it  was  particularly  severe.  The  epidemic  of 
1865-66  was  less  extensive  than  its  predecessors.  Since  1832 
eight  epidemics  of  cholera  have  occurred  in  the  United  States, 
the  last  being  in  1873. 

With  a  better  idea  of  the  true  cause  for  this  disease  in  par- 
ticular, and  with  improved  methods  for  combating  infection  and 
contagious  diseases  in  general,  cholera  has  not  been  widely 
prevalent  in  Europe  or  the  United  States  for  a  great  many  years. 
Constant  vigilance  is  required  for  its  suppression,  however,  and 
only  by  prompt  action,  where  sporadic  cases  are  discovered,  have 
the  health  authorities  prevented  epidemics.  How  many  of  these 
epidemics  have  been  directly  transmitted  through  drinking  water, 
it  is  impossible  to  know;  but  that  water  acted  as  a  carrier  to  a 
greater  or  less  extent  in  many  of  them  is  extremely  probable. 
In  the  period  from  1831  to  1873,  373,000  people  died  in  Prussia 
of  Asiatic  cholera,  and  in  1886  alone  114,000.  In  1892,  1,634 
persons  died  from  this  disease  in  Prussia,  and  from  the  Hamburg 
epidemic  in  this  same  year  8,616  deaths  resulted.  In  1910 
Germany  had  but  10  cases  of  cholera.^ 

Circumstantial  evidence  of  a  very  convincing  character  has 
been  collected,  which  proves  that  some  cholera  epidemics  were 
water-borne,  and  probablj^  no  discussion  of  this  subject  is  com- 
plete without  mentioning  the  disastrous  Hamburg  epidemic  which 
occurred  in  1892-93.  This  city  was  using  unfiltered  water  from 
the  River  Elbe,  which  was  contaminated  by  the  sewage  of  over 
800,000  people.  During  the  fall  of  1892,  17,000  cases  developed, 
resulting  in  8,600  deaths.     In  fact,  wherever  the  drinking-water 

^  Dr.   Arthuk  Ledekeh;  "The   Modern   Sewage  and  Water  Problem." 
Clinique)  August,  1912. 


TRANSMISSION  OF  DISEASES  29 

supply  was  either  filtered  or  obtained  from  some  source  other 
than  the  river,  few  or  no  cases  resulted. 

In  1887  the  city  of  Messina,  Sicily,  suffered  from  an  epidemic 
of  cholera,  during  which  5,000  cases  and  2,200  deaths  resulted. 
An  investigation  showed  that  water  purposely  diverted  from  a 
conduit,  conveying  water  to  the  city,  ran  into  pools,  which  were 
used  for  washing  soiled  clothing  by  the  Messina  washerwomen. 
Much  of  this  water  found  its  way  back  into  the  open  conduit, 
and  passed  into  the  city.  It  was  also  found  that  the  unglazed 
tile  used  to  distribute  the  water  in  the  city  were  broken,  and  that 
leakage  from  joints  was  common.  Sewers  laid  on  top  and  parallel 
with  the  water  mains  were  in  a  like  condition  and  offered  excel- 
lent opportunity  for  further  contamination.  After  a  supply 
of  pure  water,  carried  in  tank  ships  from  the  mainland,  was 
provided  for  drinking  water,  the  epidemic  ceased  at  once.^ 

A  similar  outbreak  of  cholera  in  1884  in  Cuneo,  Italy,  which 
resulted  in  3,344  cases,  was  traced  to  a  Uke  cause,  viz.,  washing 
infected  Hnen  in  a  brook  emptying  into  a  public  water  supply.  ^ 

Typhoid  Fever 

The  prevalence  of  typhoid  fever  in  civiKzed  countries,  where 
no  little  attention  is  paid  to  matters  of  sanitation,  seems  at  first 
thought  surprising.  But  not  until  1880  was  the  organism  which 
causes  this  disease  discovered  by  Eberth  in  the  spleen  of  persons 
dying  from  typhoid  fever.  Since  it  seems  doubtful  that  this 
disease,  as  it  develops  in  human  beings,  can  be  reproduced  in 
animals,  the  evidence  that  the  Eberth  bacillus  is  the  true  cause 
for  the  disease  has  been  only  slowly  accumulating.  Another 
factor  only  recently  discovered  is  that  persons  showing  no  cUnical 
symptoms  of  the  disease  are  genuine  "culture  factories"  for 
producing  the  bacillus  and  for  its  dissemination. 

Epidemics  resulting  from  these  "typhoid  carriers"  have  been 
satisfactorily  traced.  The  existence  of  such  persons  explains  in 
some  measure  the  continuance  of  the  disease,  and  its  persistence. 
The  transmission  by  direct  contact,  by  flies,  by  milk  and  by  water 
has  been  proven  in  scores  of  cases. 

Dr.  J.  F.  Anderson^  concludes  from  his  study  of  typhoid  fever 
epidemics  due  to  contaminated  water,  that  they  are  characterized 

by: 

'  W.  p.  Mason:  "Water  Supply." 

2  "A   Symposium    on    Typhoid    Fever."     Amer.    Jour.   Public   Hygiene, 

May,  1909. 


30 


WA  TER  P  URIFICA  TION 


(a)  A    general    distribution    of    cases    throughout    the    area 
supplied  by  a  particular  water. 

(b)  By  the  explosive  onset  of  the  outbreaks. 

(c)  By  the  trouble  occurring  in  the  late  winter  or  spring. 

(d)  By  the  comparative  freedom  from  the  disease  of  persons 
not  using  the  suspected  water. 

(e)  By  evidences  of  sources  of  infection  found  by  an  inspection 
of  the  watershed. 

(/)  By  the  outbreak  beginning  or  ending  after  a  change  in  the 
water  supply. 

(g)  And  by  indications  of  the  pollution  of  the  water  when 
analyzed. 


TYPHOID  FEVER 

DEATH   RATE 

PER 

100,000  OF  POPULATION 


RfC.Ar,_-a 


rmany 


Kng.  dc  Wulfs 


lyio 


.Ncthei'laaJs 


Switzerlaad 


Fic;.   1. — Typhoid  fever  death  rate  in  various  countries. 

In  the  following  table,  a  comparison  is  made  between  the 
typhoid  fever  death  rates  in  European  and  American  cities  for 
the  year  1910,  and  shows  even  at  this  period  how  much  more 
prevalent  this  disease  was  in  America  than  in  Europe. 


Unit  of  comparison 


Deaths    per 
Aggregate         100,000  from 
population        typhoid  fever, 
1910 


Thirty-three  principal  European  cities  in  Russia, 
Sweden,  Norway,  Austria-Hungary,  Germany, 
Denmark,  France,  Belgium,  Holland,  England, 
Scotland  and  Ireland 

Fifty  American  cities  of  100,000  inhabitants  or 
over 

Excess  of  deaths,  typhoid  fever  in  American  cities 
per  100,000  population 


31,590,000 
■  20,250,000 


6.5 
25.0 
18.5 


TRANSMISSION  OF  DISEASES 


31 


In  Figure  1  is  a  diagrainmatic  representation  of  the  relative 
prevalence  of  this  disease  in  several  European  countries  as  com- 
pared with  the  death  rate  in  the  United  States  for  the  year  1910. 

Some  of  the  more  important  typhoid  fever  epidemics  which 
have  been  traced  to  infected  water  are  Hsted  below: 


Place 


Number  of 


Year     ,  Population 


Cases  Deaths 


Caterham,  England 1879 

Plymouth,  Pa 1885 

Tees  River  VaUey,  England 

Lowell,  Mass '1890-91 

Lawrence,  Mass 1890-91 

1893 
1893-94 


Worthing,  England 

Grand  Forks,  N.  D 

Maidstone,  England I  1897 

Ithaca,  N.  Y 1903 

Butler,  Pa 1903 


5,000 

8,000 

1890-91251,976 

77,696 

44,654 

16,000 

6,000 

33.830 

18,000 

13,000 


352 
1,104 
1,330 

2,855 
1,792 
1,411 
1,245 
1,928 
1,350 
1,348 


21 
114 
100 

217 
137 
168  (Wells) 

96 
150  (Springs) 

82 
111 


Outbreaks  of  water-borne  typhoid  fever,  or  a  gradually  increas- 
ing prevalence  of  this  disease,  which  forced  the  authorities  to 
provide  remedial  measures,  have  occurred  at  Erie,  Pa.;  Niagara 
Falls,  N.  Y.;  Coates^-ille,  Pa.;  Ironton,  0.;  Winnipeg  Canada; 
Rockford,  111.;  Memphis,  Tenn.;  Council  Bluffs,  Iowa;  and 
Omaha,  Neb. 

Even  ground-water  supplies,  which  through  carelessness  or 
ignorance  are  not  protected  properly  from  pollution,  are  not 
infrequently  the  distributors  of  infectious  material.  An  instance 
of  this  character  occurred  some  years  ago  at  Lincoln,  Neb.,  where 
an  outbreak  of  typhoid  fever  resulted  and  severe  intestinal 
troubles  affected  several  thousand  people.  The  cause  was  found 
to  be  leakage  from  a  broken  sewer  which  found  its  way  through 
the  ground  and  an  abandoned  pipe  that  connected  directly  with  a 
well  from  which  the  public  water  supply  was  drawn. 

The  improvement  in  general  sanitation  during  the  past  25 
years  has  been  well  summarized  by  C.  A.  Holmquist  in  a  study 
of  typhoid  fever  statistics  for  the  State  of  New  York  as  a  whole, 
and  for  certain  cities  of  the  state.  He  finds  that  in  1900  the 
average  death  rate  from  typhoid  fever  in  the  registration  area  of 
the  United  States  was  31.3  per  100,000  of  population.  For  the 
same  area  for  the  period  from  1918  to  1922,  this  death  rate  had 


32  WATER  PURIFICATION 

fallen  to  5.3  per  100,000  of  population.     Comparing  these  rates 
with  those  for  the  State  of  New  York,  he  finds  a  similar  reduction. 

Death  R.\te  from  Typhoid  Fever  ix  the  State  of  Xew  York 

„    .    ,  _  ,  Average  Rate  per  100,000 

Period  Covered  of  Population 

1885  to  1906,  22-year  period 23 . 6 

1900                    l-j-ear  period 26 . 7 

1918  to  1923,    6-year  period 33 

1923                    1-year  period 2.9 

These  reductions  in  typhoid  fever  death  rates  are  due  in  part 
to  improved  water  supplies,  and  are  well  illustrated  in  the 
following  chapter. 

To  illustrate  the  decrease  in  typhoid  fever  between  1910  and 
1925  in  American  cities,  statistics  show  that  of  the  77  largest 
cities  in  the  Um'ted  States,  30  had  typhoid  fever  death  rates 
below  2  per  100,000  of  population  in  1925,  while  only  two  cities 
had  rates  over  20  per  100,000.  In  1910,  out  of  a  similar  group  of 
51  cities,  not  one  had  a  rate  below  2  per  100,000  of  population, 
and  17  had  rates  over  20  per  100,000. 

The  pollution  of  the  Great  Lakes  in  the  United  States  by  the 
cities  built  upon  their  shores  is  to  a  large  extent  local;  but  since 
these  communities  draw  their  water  supply  from  the  same  source, 
the  problem  of  preventing  the  drinking  water  from  becoming 
contaminated,  and  still  obtain  a  satisfactory  disposal  of  the 
sewage,  is  a  troublesome  one.  The  common  remedy  of  extend- 
ing the  water  supply  intakes  out  from  3  to  5  miles  from  the  shore 
has  been  resorted  to  with  fair  success,  and  with  marked  decreases 
in  the  typhoid  death  rate  in  most  cases.  The  City  of  Chicago 
has  diverted  a  large  part  of  its  sewage  through  a  drainage  canal 
into  the  lUinois  River,  thus  keeping  a  constantly  increasing 
volume  of  sewage  from  polluting  the  lake  water  farther  and 
farther  from  the  shore  line.  Other  lake  cities  are  contemplating 
partial  purification  of  their  sewage  in  order  to  conserve  the  purity 
of  their  water  supplies. 

The  effect  of  the  wind  on  these  large  bodies  of  water  in  causing 
currents,  the  influence  which  the  shore  lines  may  have  on  these 
currents,  and  the  movement  of  ice  polluted  with  sewage  from  the 
shore  out  into  the  lake  in  the  spring  months  are  all  factors  which 
may  at  times  be  the  cause  for  water-borne  epidemics. ' 

ID.  D.  Jacksox:  "Chlorination  at  Cleveland,   O."     Eng.  Record    vol 
65,  June  15,  1912. 


^ 


TRANSMISSION  OF  DISEASES  33 


References 

1.  "Report  of  the  Sanitary  Investigations  of  the  Illinois  River  and  Its 

Tributaries."     IlUnois  State  Board  of  Health,  p.  85,  1900. 

2.  Vitality  and  Isolation  of  Cholera  and  Typhoid  Organisms: 

{aj~~Jo2ir.  Infect.  Diseases,  vol.  1,  p.  641,  1904. 
(6)  Eng.  Record,  vol.  52,  Sept.  23,  1905. 

(c)  Eng.  Record,  vol.  65,  June  1,  1912. 

(d)  Zeit.  fur  Hygiene  und  Infect.  Krank.,  14. 

(e)  E.  O.  Jordan:  "General  Bacteriology." 

3.  D.    D.    Jackson    and    T.    W.    Melia:  "Differentiation    Methods    for 

Detecting    the    Typhoid    Bacillus    in    Infected    Water    and    Milk." 
Jour.  Infect.  Diseases,  vol.  6,  No.  2,  Apr.  1,  1909. 

4.  "A   Symposium   on    Tj^phoid    Fever"    (8  papers).     Am.   Jour.   Public 

Hygiene,  May,  1909. 

5.  "Epidemic    of    Typhoid    Fever    at    Columbus,    Ohio."     Jour.    Mass. 

Assoc.  Boards  of  Health,  vol.  14,  May,  1904. 

6.  Allan  J.    McLaughlin:  "Sewage   Pollution  of  Interstate  and   Inter- 

national Waters."     Public  Health  and  IMarine  Hospital  Service.     Hy- 
gienic Lab.  Bidl.  83,  March,  1912. 

7.  D.  D.  Jackson:  " Chlorination  at  Cleveland,  Ohio."     Eng.  Record,  vol. 

65,  June,  15,  1912. 

8.  "A  Polluted  Well  at  Lmcoln,  Neb."     Eng.  Record,  vol.  65,  June  1,  1912. 

9.  Typhoid  Fever  Epidemics  and  Statistics: 

(a)   Eng.  Record,  vol.  55,  p.  131,  February,  1907. 
(6)  Eng.  Record,  vol.  58,  p.  444,  October,  1908. 

(c)  Eng.  Record,  vol.  61,  p.  263,  677,  March,  1910. 

(d)  Eng.  Record,  vol.  62,  p.  630,  December,  1910. 

(e)  Eng.  Record,  vol.  63,  June,  1911. 

(/)  Eng.  Record,  vol.  65,  pp.  254,  300,  591  and  601. 
(g)  Eng.  Record,  vol.  66,  p.  95,  July,  1912. 
(h)  Eng.  News,  vol.  67,  June  13,  1912. 

10.  Charles  ~Bv  Boldrean:  "Typhoid   Fever  in   New  York  City,  etc."  /LC 

Am.  Jour.  PubliJ Health,  vol.  2,  1912. 

11.  H.  DE  B.   Parsons:  "Our  Typhoid  Streams."     Stevens  Inst.,  Janu- 

ary, 1911. 

12.  Nicholas  S.  Hill,  Jr.  and  Leon  R.  Whitcomb:  "The  Relation  of  a 

Pure   Water  Supply  to  Chronic  Intestinal  Tract  Infection."     Eng. 
News,  vol.  69,  No.  5,  Jan.  30,  1913. 

13.  George  A.  Johnson:  "The  Typhoid  Toll."     Jour.  Am.  Water  Works 

Assoc,  June,  1916. 

14.  C.  A.  Holmquist:  "Typhoid  Fever  and  Improvement  of  Water  Sup- 

pHes."     Eng.  Contr.,  p.  797,  Oct.  8,  1924. 

15.  Roger    G.    Perkins:  "Typhoid    Fever    in    Cleveland,     1873-1926." 

Jour.  Preventive  Medicine,  Sept,  1927. 


CHAPTER  IV 

THE  EFFECT   OF  IMPROVED   WATER   SUPPLIES  UPON 

HEALTH 

As  the  relation  between  impure  water  and  disease  becomes 
better  understood  and  appreciated,  more  attention  is  being  given 
to  the  quaUty  of  pubUc  water  suppHes.  Sometimes  the  method 
pursued  is  to  seek  a  purer  water  from  some  uncontaminated 
source,  or  to  render  a  polluted  supply  better  by  some  process  of 
purification.  By  substituting  a  pure  drinking  water  supply  for 
an  impure  one,  the  reduction  in  the  death  rate  from  water-borne 
diseases  has  been  notable,  and  the  improved  health  of  the  com- 
munity has  usually  been  demonstrated  beyond  question. 

In  the  United  States  the  typhoid  fever  death  rates  undoubtedly 
furnish  the  best  indicators  of  the  quality  of  public  water  supplies. 
The  following  table  taken  from  Dr.  Geo.  M.  Kober's  paper  on 
the  "Conservation  of  Life  and  Health  by  Improved  Water  Sup- 
ply" summarizes  the  statistics  of  61  cities  in  the  United  States  for 
the  years  1902-06.1 

Mean   Typhoid   Fever   Death    Rate    from    1902-06    per    100,000    of 
Population  for  Cities  Using  Various  Classes  of  Water 

4  cities  using  ground  water  from  large  wells 18. 1 

18  cities  using  impounded  water  and  conserved  rivers 

or  streams 18.5 

8  cities  using  water  from  sm.all  lakes 19.3 

7  cities  using  water  from  the  Great  Lakes 32 . 8 

5  cities  using  both  surface  and  underground  water. .  .  45 . 7 

19  cities  using  polluted  river  water 61 . 1 

From  the  same  paper  is  reproduced  a  diagram  (Fig.  2)  showing 
in  more  detail  the  typhoid  fever  death  rates  in  different  cities 
according  to  the  character  and  the  source  of  their  water  supply. 

Spring  waters,  ground  waters  from  wells,  and  filtered  waters 
evidently  furnish  the  safest  supplies.  Surface  waters,  whether 
from  streams  or  lakes,  may  furnish  safe  drinking  water,  but  they 
are  much  more  likely  to  be  polluted. 

Probably  the  most  striking  effects  in  reducing  typhoid  fever 
have  come  from  the  purification  of  polluted  supplies.     By  filter- 

^Eng.  Record,  vol.  57,  June,  1908. 

34 


IMPROVED  WATER  SUPPLIES 


35 


ing  an  impure  water  supply  marked  reductions  in  water-borne 
diseases    have    almost    invariably    resulted.     Even    where    the 


0  10  20  30  40  50  CO  TO  80  90100U0120 


70 


Co 


60 


55 


50 


45 


40 


35 


30 


20 


15 


10 


PURE  MOUNTAIN  SPRINGS 


FILTERED    WATERS 
EUROPEAN    CITIES 


FILTERED  WATERS 
AMERICAN    CITIES 


GROUND  WATERS 
LARGE  WELLS 


J_ 


IMPOUNDING  RESERVOIRS 
PROTECTED  WATER  SHED3 


3P: 


4_l        PROTECTED  RIVER 
OR  STREAM  SUPPLY 


u. 


m 


n 


SMALL  LAKES 


I    I 


GREAT  LAKES 
SUBJECT  TO  POLLUTION 


H 


MIXED  SURFACE  AND 
UNDERGROUND  WATERS 


/Munich 

*"\  Vienna 

r  Berlin 

Zurich 

O.GS  Hamburg 

I  Paris 

LLondon 

Paterson 
Binghamton 


18.1 


ia.5 


18.3 


RIVER  WATER 
SUBJECT  TO 
POLLUTION 


IQ,'2\  Albany 

J ;  Lawrence 

1        ;  Watertown 

Richmond  Borough 

Queens  Borough 

Camden 

Lowell 

Fitchburg 

Cambridge 

Somerville 

Worcester 

Bridgeport 

Hartford 

Maiden 

Boston 

Chelsea 

New  Bedford 
.Waterbury 

Holyoke 

Bronx  Borough 

Manhattan  Borough 

Fawtucket 

Newark 

Jersey  City 

Baltimore 

Eochester 

Syracuse 

Fall  Eiver 

Brockton 

Taunton 

Haverhill 

Portland 

Salem 

Milwaukee 

Detroit 

Chicago 

83.1  Bu  Halo 

J ,  Erie 

J Cleveland 

J iDuluth 

St. Paul 

Canton 

Brooklyn  Borough 

Columbus 

.McKeesport 

Minneapolis 

Seattle 

New  Orleans 

Toledo 

Evansville 

Springfield 

Covington 

Grand  Rapids 
Wilmington 

Richmond 

Cincinnati 

Louisville 

Philadelphia 

Lancaster 

Atlanta 

Harrisburg 

Wheeling 

Allegheny 
Pittsburgh    


45.7 


Cl.U 


0  10  20  30  40  50  60  70  80  90  100110120 

Fig.  2. — Typhoid  fever  death  rate  according  to  water  supply. 

community  has  been  only  partially  supplied  with  purified  water, 
the  effect  on  the  typhoid  death  rate  has  been  noticeable. 


36 


WATER  PURIFICATION 


Typhoid  Death  Rates  per  100,000  of  Population  for  Cities  Changing 
FROM  Polluted  TO  Purified  Water  Supplies 


1907 


190S 


Columbus,  Ohio,, 
New  Orleans,  La, 
Louisville,  Ky . .  . 
Pittsburgh,  Pa. .  . 
Philadelphia,  Pa. 


The  filter  plant  for  the  City  of  Columbus,  Ohio,  was  started 
in  August,  1908;  the  high  rate  for  this  year  was  due  to  an  epidemic 
in  the  early  part  of  the  year.     The  filter  plants  in  New  Orleans 


no 

130 

120 

110 

100 

90 

80 

70 

60 

50 

40 

30 

\ 

K 

1 : 

> 
O 

3 
O 

\ 

\ 

A 

^ 

A 

o 

■ii 

a 

3 

a, 

V 

\ 

\   i 

/ 

\  -S 

1  u 

\/ 

n 

CT    3 

V 

1 

1 

a 

u 
a 

3 

' 

FROM  HYGIENIC  LABORATORY 
BULLETIN   No.  83 
BY 

ALLAN  J.  McLaughlin 

U.S.) 

1       o: 

1 

1    ^;,t 

i 

a 
a 

i 

a 
3 

\  o 

\ 

1 

■ 

V 

-- 

in 

J900  J901  1902  1903  1904  1903  191)6  1907  1908   1909  1910 

Fig.  3. — Typhoid  death  rate  by  years  for  city  of  Pittsburgh,  Pa. 

and  Louisville  were  started  in  1909.  In  Pittsburgh  and  Philadel- 
phia filter  plants  were  placed  in  operation  in  1907  and  1908, 
respectively;  but  in  neither  city  was  the  entire  population  sup- 
plied with  the  purified  water. 


IMPROVED  WATER  SUPPLIES 


37 


Figure  3  is  of  special  interest  in  this  connection,  in  showing  how 
pronounced  a  reduction  in  the  death  rate  followed  the  introduc- 
tion of  even  a  limited  volume  of  purer  water. 

From  the  typhoid  fever  statistics  of  Cincinnati,  Ohio,  a  most 
convincing  argument  for  the  purification  of  a  polluted  water 
supply  can  be  presented. 

Number  of  Cases  and  Deaths  from  Typhoid  Fever 


Unfiltered  water  from  old  works 


Filtered  water  from  new  works 


Year 

1904 

1905 

1906 

Total 
for  3 
years 

1908 

1909 

1910 

Total 
for  3 
years 

Cases 

Deaths 

1,646 
270 

746 
155 

1,940 
239 

4,332 
664 

235 

67 

218 
45 

183 
21 

636 
133 

The  figures  for  the  year  1907  are  omitted  because  water  from 
both  the  old  and  the  new  works  was  supplied  to  the  city. 

Tf  the  above  figures  are  expressed  as  cases  and  deaths  per 
100,000  of  population,  a  better  comparison  may  be  made  with 
other  statistics. 


Number  of  Cases  and  Deaths  per  100,000  of  Population 


For  3  years  before  int 
filtered  watei 

roducing 

For  3 

years  after  introducing 
filtered  water 

Average                   1908                       1909 

1910 

Cases 

Deaths 

Cases 

Deaths 

417 

64 
Percentage  n 

67 
19 
jduction  from 
84 
70 

62 
13 

the  average. 
85 
80 

50.0 
5.7 

88.0 
91.0 

In  a  report  on  the  purification  of  the  Montreal  water  supply 
Messrs.  Hering  and  Fuller  present  a  table  showing  the  effect  of 
purification  by  filtration  on  the  death  rate  from  typhoid  fever  in 
a  number  of  American  cities. 

The  relation  between  impure  water  supplies  and  certain  intes- 
tinal diseases  other  than  typhoid  fever,  is  more  or  less  obscure. 


38 


WA  TEH  P  URIFICA  TION 


Death  Rates  from  Typhoid  Fevek  per  100,000  Population  in  American 
Cities   Using   Filtered   Water 


City 


Before 
Year  filtra- 

plant  was       tion 

com-     ! 

pleted    '' 


Before 
filtra- 
tion 


After 
filtra- 
tion 


Years  averaged 


Death  rate 


Sand  filters 


Albany,  N.  Y. . . 
Lawrence,  Mass. 
Pittsburgh,  Pa. . 


1 

1899 

10 

9 

90 

1893 

7 

15 

114  ! 

1907 

S 

1 

133 

1 

22 
25 

471 


Mechanical  filters 


Bingham  ton,  X.  Y 
Cincinnati,  Ohio . . 
Columbus,  Ohio... 
Paterson,  X.  J. .  .  . 
Watertown,  X.  Y. 

York,  Pa 

Hoboken,  X.  J 


1907 

5 

5 

1908 

4 

1 

1908 

11 

1 

1902 

5 

7 

1904 

5 

5 

1899 

2 

8  i 

1905 

/ 

4 

47 
50 

78 
32 
100 
76 
19 


15 
16 
20 
10 
38 
22 
14 


Infant  mortality  from  diarrhea  and  enteritis  is  probably  both 
directly  and  indirectly  the  result  of  drinking  impure  water, 
although  other  causes  contribute  more  frequently  to  death. 
About  85  per  cent,  of  the  deaths  listed  under  "diarrhea  and 
enteritis"  in  the  United  States  census  mortality  statistics  occur 
in  children  under  2  years  of  age.  Dr.  A.  J.  McLaughlin-  in 
discussing  this  subject  says: 

"Instead  of  one  disease  designated  under  different  names  we  are 
probably  considering  several  diseases  with  common  factors  of  trans- 
mission. Whatever  the  real  relation  between  typhoid  fever  and  enteritis 
or  diarrhea  of  children  may  be,  one  fact  is  clear,  the  same  causes  operate 
to  cause  excessive  prevalence  of  both.  It  is  probable  that  cases  of 
typhoid  in  children  under  2  years  in  man}^  cities  are  often  incorrectly 
diagnosed  as  enteritis.  It  must  be  remembered,  however,  that  the 
causative  agent  of  bacillary  dysentery  is  transmitted  in  the  same  waj'^ 
and  by  the  same  media  as  that  of  tj-phoid.  There  are  too  manj^  cases 
of  fatal  illness  in  children  under  2  years  classed  as  diarrhea  and  enteritis, 
and  an  exhaustive  investigation  should  be  made  to  establish  the  real 

1  Including  Allegheny,  supplied  with  unfiltered  water. 
*  Public    Health    and    Marine    Hospital  Service.     Hygienic  Lab.,  Bull. 
83,  March,  1912. 


IMPROVED  WATER  SUPPLIES 


39 


cause  of  death  in  enteritis  and  diarrhea  of  children.  Without  such  an 
investigation  it  is  impossible  to  assign  the  real  cause  of  the  excessive 
child  mortaUty  from  diarrhea  and  enteritis.  In  cities  of  less  than  50,000 
population  without  slums  and  which  are  not  'mill'  towns  an  enteritis 
rate  in  children  under  2  years  above  100  deaths  per  100,000  indicates 
prevalence  of  an  acute  intestinal  disease  preventable  by  the  same 
measures  that  prevent  typhoid  fever.  It  is  probable  that  in  such  cities 
proper  enforcement  of  prophylactic  measures  against  typhoid  fever 
would  reduce  the  enteritis  rate  below  40  deaths  per  100,000.  Enforce- 
ment of  prophylactic  measures  would  include  the  installation  of  pure 
water  supplies  and  proper  sewerage  systems,  coupled  with  a  vigorous 
campaign  against  the  insanitary  outdoor  privy  and  the  equally  dangerous 
shallow  well.'" 

Dr.  McLaughlin  gives  the  following  table  which  emphasizes 
the  complexity  of  the  problem. 


City 

Typhoid 

death  rate 

per 

100,000, 

average  for 

10  years, 

1900-1909 

Character 

of  water 

supply 

Death  rate 

enteritis, 

average 

for  5  years, 

1904-1908 

Keiiiarks 

Rochester,  N.  Y 

Syracuse,  N.  Y 

Albany,  N.  Y 

Binghamton,  N.  Y.  .  .  . 

Utica,  N.  Y 

Schenectady,  N.  Y.  .  .  . 

Amsterdam,  N.  Y 

Yonkers,  N.  Y 

Cohoes,  N.  Y 

13.7 

14.8 

21.9 

20.9 

17.3 

22.4 

18.6 

9.5 

83.8 

129.1 

148.5 

27.0 

Good 

Good 

Good 

Good 

Good 

Good 

Good 

Good 
Polluted 
Polluted 
Polluted 
Polluted 

89.5 

105.5 
80.0 
104.7  . 
133.7 
164.7 
150.7 
207.7 
170.9 
173.2 
175.0 
151.6 

Sanitary  conditions 
good. 

Mill    and   factory 
towns;  bad  sanitary 
conditions. 

Sanitary    conditions, 

Niagara  Falls,  N.  Y . .  . 

Ogdensburg,  N.  Y 

Buffalo,  N.  Y 

exclusive    of   water, 
good.    Mill  and  fac- 
tory towns. 

Some  striking  evidence  of  the  effect  of  purer  water  supplies  in 
cities  in  the  State  of  New  York  has  been  brought  out  by  C.  A. 
Holmquist  in  a  study  referred  to  in  the  previous  chapter.  The 
cities  of  Albany,  Cohoes,  Niagara  Falls  and  Binghamton  suffered 
severely  from  impure  water  supplies  until  measures  were  taken 
to  purify  their  drinking  waters.  The  following  table  summarizes 
this  evidence  in  a  concise  form. 


40 


WA  TER  P I  ^RIFICA  TION 


City 

Period 

Death  rate 

per  100.000  of 

population 

Albany^ 

Before  slow  sand  filters  were  used 

89   4 

First  j-ear  after  filters  were  used  (1899) 

40.0 

Five  years  1919  to  1923 

4.1 

Cohoes 

Prior  to  filtration  and  disinfection 

S.T     ^ 

Six  years  1918-1923 

3.7 

Niagara  Falls 

Prior  to  filtration,  1912 

Year  preceding  filtration  and  disinfec- 

131.8 

tion 

179.1 

First  year  following  filtration  and  disin- 

fection 

65.2 

Second  year  following  filtration  and  dis- 

infection 

26.6 

Six  years,  1918-1923 

5.0 

Bingliaiiitom 

Before  filtration 

56  2 

Six  years,  1918-1923 

4.0 

1  At  Albany  added  pre-filters  to  slow  sand  filter  plant  in  1908.     Chlorina- 
tion  of  this  supply  commenced  in  1909. 


In  a  paper  by  H.  Burdett  Cleveland  published  in  the  Natiori's  Health 
in  October,  1924,  a  graphical  representation  of  typhoid  fever  statistics  for  a 
number  of  cities  in  the  State  of  New  York,  including  several  of  the  above- 
mentioned  cities,  gives  visible  evidence  of  the  effect  of  improved  water 
supplies.  Mr.  Cleveland's  statements  in  his  paper  are  given  in  part  in 
order  to  explain  the  charts. 

"Typhoid  mortality  curves  for  eight  of  the  larger  cities  have  been 
plotted  on  semi-log  paper  with  a  circle  interposed  to  show  when  pasteuri- 
zation of  milk  was  required  by  ordinance.  Sanitary  control  of  milk 
supplies  cannot  be  so  definitely  fixed,  either  in  point  of  time  or  degree, 
and  has  not  been  indicated.  The  periods  during  which  each  water 
supply  was  untreated  and,  later,  treated  together  with  the  treatment 
employed  are  shown  by  these  curves." 

"A  curve  of  the  typhoid  rate  for  the  combined  population  of  all  the 
cities  for  the  25-year  period  is  also  shown  on  a  cross-section  chart  as  well 
as  a  curve  for  this  period  showing  the  percentage  of  population  served 
with  untreated  water.  Supplies  have  been  classified  as  treated,  if 
filtered,  sterilized,  or  both." 

"It  is  not  to  be  claimed  that  all  of  this  reduction,  more  particidarly 
tluring  the  past  10  to  15  years,  has  been  due  to  water  purification. 
Much   of   the    decrease   in   typhoid    mortality   has   undoubtedly   been 


IMPROVED  WATER  SUPPLIES 


41 


a     o    o     o 
o     o    o     o 

v£>        ^      CJ       O 


cs  o  o  o 

C3  C3  O  <=> 

C3  oO  >XJ  "^ 

ro  C^  CV  CSJ 


CD  O  O  O  O  O  g 

CD  O  C>  C3  O  —   Si 

CO  O  CO  tO  ^ 

C^J  oJ  — 


'='   s.    <^ 

C71  oo  r— 


o  o 


5 


S  R  o  o 

rO  CO  —  '-' 


■OCOCVOilOJOJ      —     —       —       I  I 


o 


C<3 


42 


WATER  PURIFICATION 


no 


90 
80 

70 

60 

50 


Unfreafed 

F,7-hred 

Filtered  and Sier'iUzed  = 

Sferi/ized 


16981900 


1920 1922 


Fig.  36. — Mortality  rate  from  typhoid 
and  water  supply  conditions  in  Albany 
and  Utica  during  the  twenty-five  year 
period  covered.  In  all  the  charts  the 
official  health  records  supply  the  vital 
statistics  and  in  most  instances  the  water 
conditions  are  drawn  from  reports  of 
inspection  bv  the  State  Department  of 
Health. 


190 


140 


g.  90 
S.  80 
o    70 

§,  60 
o 
°    50 

J.  40 

at 

%    30 


20 
15 

10 


I 


.  K 

j[ 

WA&ARA  FAIlA     i 

f\ 

V    ' 

V 

/\ 

"^ 

-Bi/F/A 

t^^ 

^ 

1 

1 

1898 1900        1905 


1910 


1915 


1920  1922 


Fig.  3c. — The  correlations  are  quite 
as  marked  between  water  suppHes  and 
typhoid  fever  incidence  in  Niagara 
Falls  and  Buffalo.  Circle  shows  when 
pasteurization  of  milk  was  required  by 
ordinance. 


no 

100 

-p    90 

i    60 

^    70 

§.  fco 

£■    50 
o 

§   40 
-   30 

J 

1      1 

i 

l\ 

J 

I 

v 

\ 

\ 

'A\A 

1  1 

\V  A  ^j^"-- . 

^  20 
1    15 
1    10 

o 

\ 

n 

^VATvr 

/^s 

f- 

IJ 

,         \ 

/^    \ 

BINGHA 

MTON  fV; 

V 

i 

1 

Xl 

\ 

1 

\! 

1898 1900       1905 


I9J0 


19:5         1920 1922 


Fig.  3d. — Further  proof  that  sani- 
tary measures  do  control.  Note  the 
downward  plunge  at  each  introduction 
of  new  sanitary  requirements. 


Q- 
O 


90 
80 
70 
60 

SO 

40 

30 

20 
15 
10 


1          I 

1    i 

k  ' 

\ 

* 

A 

V 

A- 

V 

\ 

r^OCHESTER      yELM/RA     a 

r 

^ 

/Vad./\ 

A 

V 

\ 

A 

V> 

r 

\ 

1900        1905        1910 


1915 


1920  1922 


Fig.  3  e . — R  o  c  h  ester  and  Elmira 
afford  further  proof  of  the  general  prop- 
osition. 


IMPROVED  WATER  SUPPLIES 


43 


brought  about  by  better  sanitation  of  watersheds  and,  in  some  of  the 
larger  cities  and  more  progressive  communities,  by  sanitary  control  of 
milk  supplies  and  especially,  by  milk  pasteurization." 

"To  an  indeterminate  degree,  also,  typhoid  vaccination  since  the 
World  War,  as  well  as  closer  control  by  public  health  authorities  of  the 
typhoid  carrier  problem  and  of  infected  food  transmission,  and  a  gradual 


o 
a. 

-o 

a> 

c 

E 
o 
o 


100 

90 
80 
70 
60 
50 
40 
30 
ZO 
10 


Percemage  oftoialWaferSupp/y 
Unfreahd(Encluiive  offJ.Y.CiiySupply) 


50 


45 


40  :i 

35.1 

E 


30 


Z5  4! 
o 

ex 


20   ~ 


15  -o 
o 


10 


1898  1900 


1920  1922 


Fig.  3/. — A  curve  of  the  tvphoid  rate  for  the  combined  population  of 
all  the  cities  for  the  twenty-five  year  period.  Shows  also  the  percentage 
of  population  served  with  untreated  water. 


improvement  in  general  sanitary  conditions,  have  all  contributed  to  this 
noteworthy  decline  in  the  typhoid  fever  death  rate." 

"Beyond  question,  however,  the  effect  of  filtration  and  sterilization 
of  water  supplies  in  reducing  typhoid  rates  has  been  very  much  greater 
than  that  of  all  other  preventive  measures.  This  is  clearly  indicated 
by  the  table  and  curves  presented  herewith," 


44 


WATER  PURIFICATION 


z 

o 

< 


D 

O 

u 
a 


o 

§ 

o 
;;  O 

I   « 

a  P3 

n 

o 

z 

o 

<; 


o 
< 

(1. 
o 


o  £ 
o 

o  e 

S  z 

a  ^ 

^  « 

o  < 


K 

> 


3     . 
C   O 


&s  ■ 

c  :'  =3 


IH 


^^0^ 

'^^ 


Q^' 


Q=0 


TO  O  "O 

•  X  O 

-H  X  Ci 

t^  —  ^ 


I 

t^  o  » 

•    O  —1 

M    O  C-. 


I  (M  ir;  ~3  •1' 

O  — I  •  .—  rt 

X  c;  -*  c:  o 

X  '-  -1  ^  -1 


N  o  X 

•  X  o 

(N  X  OJ 

CO  »-<  »-< 


lO    o    C") 

C:    X   C-. 


0  0  05 

•  o  o 

X  oi  m 

lO  rt  ■-! 


u7    C*- 


o  o  o  o 


t^  X  c; 
M    --i    -^ 


O  I 

■  O  X 

lO  X  o 

O  X  C5 


O  Ci 

•  o 

TO  03 

C^  1-1 


O  O  1^ 

■  X  o 

CO  X  o; 

IC  -<  t-H 


X 

I.-;    O 

.  o 

X    — ' 


I 

O   to  M    0-. 

•   X  — <       • 

O   X  O;    03 

(N   >-i  ^ 


t,  o  •-I  — I  c^  r^ 

-rti     X     1-H  •     l-H     ^H 

.   X   03   X  05  o: 


1 

U5   O  t^ 

•  X  o 

>o  X  o 

lO   -H  1-1 


e  X  X 

™  §  s 


^H  *-<  »0  05  O  lO 

■  X  ^  ■  1-1  Oi 

o  X  05  lO  c;  05 

CO  -<  ,-.  ■-.  rt 


•o 


I 

o  o  o  to  ^ 

•     X  rt  ■  -H     rt 

O   X  03  CO  o>  o 

Tf    .-1  i-<  <N  rt    i-( 


CO   05    '-'5 

1.0  c;  O! 


CD    »0    »-0 

LO  c:   o 


t^    05    lO 

•    O    C^I 

t^   O    03 


"^    CO    lO 
lO    C:    O 


O   •*   10 
O   o:    o: 


~f   —•   ">> 

■  c-.  c. 

O    -H    -I 


:^.  ri^.    ^  ,-. 


^'    '—'    »^ 
LO  o;  oi 


c^  CO  lO 

CO    C-.    C-. 


X    LO 

— •    !M 

O    03 


I 

C3    LO 

CO   c:   c; 


IN 


CO    CO    "O 

■  .-1  l» 

t^  c:5  o 


o 
-a 


«•    £ 


>>     'P 


•V    a 


-o 
c 

03 


i    c    fl 

—     GO 

-a  •-  -5 

c  c:   a 

too 

CJ 


a 

CO 

u 

a) 
o 

OJ 

1-4 

U4 

— . 

r> 

a 

rt 

CC 

o 

3 

o 

a> 

3 

•— ' 

«> 

p 

Cl 

-< 

1—1 

- — ■ 

1— * 

o 

O^ 

a 

o 

•-^ 

*^ 

•n 

0 

c 

(U 

rt 

V 

d 

-^ 

3 

CI 

■^ 

^ 

Ti 

0) 

O 

d 

o 

TJ 

t.. 

o 

n 

c 

o 

O: 

3 

o: 

(-^ 

M 

& 

o 

^ 

CJ 

13 

^ 

O 

5J 

■T3 

U 

03 

o 

(U 

^ 

a 

- 

cn 

Q 

r1 

in 

>. 

■-^ 

o 

a 

a 

»— 1 

3 

c 

cc 

CO 

,^ 

O 

o 

r 

~-l 

,...( 

C 

c: 

'-^ 

CJ 

'^ 

■a 

ui 

j! 

c; 

CJ 

ca 

>, 

a 

a 

-h:> 

o 

LO 

0 

3 
O 

".' 

X 

-n 

^ 

o 

S3 

S 

J= 

c 

I— < 

3 

n 

r 

•ii 

E 

-^ 

& 

m 

1—1 

>. 

T) 

o 

r! 

c 

>o 

it 

o 

C.) 

o 

I~. 

c 

T1 

C 

3 
T3 

CJ 

C 

CJ 

cn 

n 

T) 

C^ 

tM 

-^ 

'  j: 

0' 

T! 

■o 

c; 

CJ 

o 

c. 

-*^ 

0^ 

c 

a 

a 

s 

CJ 
CJ 

CD 
CO 

S 

ni 

c 

0, 

o 

e 

i-H 

o 

O: 

H 

'I* 

^ 

o 

1- 
0 

■a 

^C 

-«J 

Z 

*j 

c 

o 

o 

1 

c 

CJ 
J2 

c 

IMPROVED  WATER  SUPPLIES 


45 


The  follovdng  mortality  statistics  taken  from  reports  of  the 
Health  Department  of  Cincinnati,  Ohio,  point  pretty  plainly 
to  the  fact  that  a  change  from  an  impure  water  supply  to  a 
purified  supply  made  a  marked  difference  in  the  number  of 
deaths  from  dysentery,  diarrhea  and  enteritis  in  persons  over  2 
years  of  age. 

The  following  figures  show  a  reduction  in  deaths  from  dysen- 
tery for  3-year  periods  before  and  after  the  introduction  of 
filtered  water  of  64.3  per  cent,  and  for  diarrhea  and  enteritis 
of  55.1  per  cent. 


Unfiltered   Water  from  Old 
Works. 


Filtered  Water  from 
New  Works. 


Year 


1904  1905 


1906 


Total 
for  3 
years 


1908 


1909 


1910 


Total 
for  3 
years 


Dysentery 

Diarrhea  and  enteritis  over  2 
years  of  age 


27      21      22 
152    167  174 


70 
493 


9      11        5 
90      60  !  71 


25 
221 


N.  B.  The  year  1907  is  omitted  because  water  was  supplied  to  the 
city  both  from  the  old  works  and  from  the  new  works. 

THE   MILLS-REINCKE   PHENOMENON   AND   HAZEN'S   THEOREM 

The  available  evidence  clearly  proves  that  cholera,  typhoid 
fever,  dysentery  and  gastro-intestinal  troubles  are  commonly 
transmitted  by  infected  drinking  water,  and  that  there  is  reason 
to  believe  that  other  diseases  are  also  conveyed  in  the  same 
manner.  The  effect  of  purifying  a  polluted  water  supply,  in 
decreasing  the  deaths  from  diseases  other  than  cholera  and 
typhoid,  has  been  studied  recently  by  W.  T.  Sedgwick  and  J. 
Scott  MacNutt.  They  point  out  that  in  1893-94,  IVIessrs. 
Hiram  F.  Mills,  C.  E.,  of  Lawrence,  Mass.,  and  Dr.  J.  J.  Reincke 
of  Hamburg,  Germany,  respectively,  noted  independently  a 
decHne  in  the  general  death  rate  of  each  of  these  cities  as  a  result 
of  improving  their  water  supplies.  Professor  Sedgwick  and  his 
associate  have  collected  numerous  mortality  statistics  in  a  paper ^ 
on  this  subject,  and  have  termed  the  coincidence  between  a 

1  W.  T.  Sedgwick  and  J.  Scott  MacNutt:  "On  the  Mills-Reincke 
Phenomenon  and  Hazen's  Theorem  Concerning  the  Decrease  in  Mortality 
from  Diseases  Other  than  Typhoid  Fever  Following  the  Purification  of 
Public  Water  Supplies."     Jour.  Infect.  Diseases,  vol.  7,  1910. 


46 


WA  TER  P  URIFICA  TION 


lowered  death  rate  and  a  purified  water  supply  as  the  "Mills- 
Reincke  Phenomenon."  In  1904,  Allen  Hazen  in  a  paper  read 
before  the  International  Engineering  Congress  held  in  St.  Louis, 
gave  a  quantitative  expression  for  the  phenomenon  by  stating 
that:  "Where  one  death  from  typhoid  fever  has  been  avoided 
by  the  use  of  better  water  a  certain  number  of  deaths,  probably 


38 


36 


34 


32 


30 

a 

S28 

a 


^  ? 


u 
t-, 


-26 


^S 


g 


t-  22 

o 
c 

O20 

e 
» 
£18 

a 
u 


O  o 


=  12 


\^ 

K 

"^ 

■--\ 

k 

W 

^ 

Y' 

\ 

\ 

\ 
\ 

\ 

\ 

V  \ 
i  \ 

V 

y  ■ 

^  \ 
\  \ 
\  \ 

\\ 

\ 

X 

/ 

\ 

> 

/ 
/ 

A 

^ 

.^ 

LEGEND 
-— Filtered  Wafer 

Population 
—•—Typhoid  Fever 

Death  Rates 

^— '*' 



-^ 

— 

"^ 

38 
36 
3t 
32 
30 
28 
26 
21 
22 
20 
13 
16 
1* 
12 
10 
8 
6 
4 
2 


3       S       S 


Fig.  4. 


-Growth  of  water  filtration  and  decrease  in  typhoid  fever  death 
rate  in  the  registration  cities  of  the  United  States. 


two  or  three,  from  other  causes  have  been  avoided."  Sedgwick 
and  MacNutt  have  called  this  expression  ''Hazen's  theorem," 
and  have  concluded  from  the  studies  which  they  have  made, 
that  the  statement  appears  sound  and  conservative,  but  not 
necessarily  precise.  The  ratios  they  worked  out  varied  widely. 
For  example,  at  Hamburg,  Germany,  for  every  death  less  from 


IMPROVED  \VATER  SUPPLIES  47 

typhoid  fever  after  filtration,  there  were  15.8  deaths  less  from 
other  causes;  at  Lawrence,  Mass.,  the  ratio  was  1  to  4.4;  at 
Lowell,  Mass.,  1  to  6.0;  at  Albany,  N.  Y.,  1  to  4.1;  and  at  Bing- 
hamton,  N.  Y.,  1  to  1.5. 

From  the  data  which  they  compiled  with  respect  to  diseases 
of  the  respiratory  organs,  the  evidence  in  many  cases  is  striking 
and  warrants  a  more  extended  study.  Factors  not  specifically 
related  to  an  improved  water  supply  may  have  been  potent  in 
reducing  the  death  rate  in  the  cities  studied.  Greater  knowledge 
of  methods  for  the  prevention  of  disease  must  have  played  its  part 
in  diminishing  mortality,  and  in  extending  the  length  of  the 
average  life.  On  the  other  hand,  the  real  causes  for  many  dis- 
eases are  still  unknown,  and  the  probability  that  some  of  them 
can  be  transmitted  through  polluted  drinking  water  is  by  no 
means  unreasonable. 

The  elimination  of  disease  germs  from  a  water  supply  would  be 
a  direct  cause  for  a  decrease  in  the  death  rate,  while  an  indirect 
cause  could  come  from  an  increased  vital  resistance  produced  by 
the  use  of  purer  drinking  water.  Probably  both  of  these  factors 
are  instrumental  in  bringing  about  a  lowered  death  rate. 

The  accompanying  diagram  (Fig.  4)  from  George  A.  Johnson's 
paper  "The  Typhoid  Toll,"  in  the  Journal  of  the  American 
Water  Works  Association  for  June,  1916,  is  a  striking  graphic 
presentation  of  the  effect  which  water  purification  has  had  in 
reducing  typhoid  fever  in  the  LTnited  States. 

References 

1.  Geo.    M.    Kober:   "Conservation    of    Life    and    Health    by    Improved 

Water  Supply."     Eng.  Record,  vol.  57,  June,  1908. 

2.  Arthur    Lederer:  "The     Modern     Sewage    and     Water    Problem." 

Clinique,  August,  1912. 

3.  E.  BoxjEAx:   "Les  eaux  d'alimentation  publique  observations  generales 

sur  leur  role  epidcmiologique;  leur  choix;  etat  actuel  de  I'epuration." 
Rev.  Scient,  49,  An.  5,  pp.  138-146. 

4.  W.  T.  Sedgwick  and  J.  Scott  MacNutt:  "On  the  Mills-Reincke  Phe- 

nomenon and  Hazen's  Theorem  Concerning  the  Decrease  in  Mortality 
from  Diseases  Other  than  Typhoid  Fever  Following  the  Purification  of 
Public  Water  Supplies."     Jour.  Infect.  Diseases,  vol.  7,  1910. 

5.  Geo.   A.   Johnson:   "The   Typhoid   Toll."     Jour.   Amer.   Water   Works 

Assoc,  June,  1916. 

6.  F.  H.  Waring:  "The  Typhoid  Fever  Epidemic  at  Xenia,  Ohio."  Jour. 

Amer.  Water  Works  Assoc,  p.  167,  June,  1919. 

7.  M.   C.   Sjoblom:   "Typhoid   Fever  at   Moline,  111."  Jour.  Amer.  Water 

Works  Assoc,  p.  230,  June  1919. 


48  WATKR  PURIFICATION 

8.  M.   C.  Sjoblom:  ''The  Outbreak  of  Typhoid  Fever  and  Dysentery  at 

the   Chicago  and   Alton   Shops  at    Bloomington,   III."     Jour.  Anier. 
Water  Works  Assoc,  July,  1920. 

9.  ''Comparative    Death    Rates   and    Sanitary   Conditions."     The    forty- 

eighth  Semi-annual  Report  of  the  Sewerage  and  Water  Board  of  N«w 
Orleans,  La.     {Ahst.)  (Eng.)  Contr.,  p.  820,  Oct.  8,  1924. 

10.  J.  W.  Ellms:  "A  Sanitary  Survey  of  Lake  Erie  Opposite  Cleveland, 

Ohio,  in  1920."     Jour.  Water  Works  Assoc,  IMarch,  1922. 

11.  Russell  Suter:  "Stream  Pollution  in  Terms  of  Sanitary  Quality  of 

Drinking  Water."     Jour.  Amer.  Water  Works  Assoc,  March,   1923. 

12.  Abel   Wolmax:  "Cooperative   Research   in   Water-purification  Prob- 

lems."    Jour.  Amer.  Water  Works  Assoc,  July,  1920. 

13.  E.  Sherman  Chase:  "Quality  of  Water  Supply."     Engr.  Contr.,  May 

9,  1923. 

14.  H.  W.  Streeter:  "Some  Recent  Developments  in  Stream  Pollution." 

Proc    Seventeenth    Annual    Convention    of   Indiana   Sanitary   Water 
Supply  Association. 

15.  "Changes  in  Water  Treatment."     Editorial,  Engr.  News-Record,  Aug. 

21,  1924. 


CHAPTER  V 
OBJECTS  AND  METHODS  OF  WATER  PURIFICATION 

The  purpose  of  purifying  a  drinking  water  is  obviously  to  make 
it  more  pleasant  and  more  wholesome  to  drink.  The  removal  of 
suspended  sediment  and  of  vegetable  coloring  matter,  in  order  to 
convert  a  dirty  water  into  a  clear  and  colorless  one,  commends 
itself  if  for  no  other  reason  than  that  the  water  becomes  more 
acceptable  to  the  senses  of  taste,  smell  and  sight.  If  in  addition 
to  improving  its  appearance,  it  is  possible  to  reduce  materially 
or  eliminate  entirely  the  bacteria,  some  of  which  may  be  capable 
of  producing  disease,  a  second  important  reason  for  the  purifica- 
tion of  a  drinking  water  can  be  advanced.  By  rendering  a  water 
hygienically  safe  for  drinking  purposes,  sickness  and  death,  as 
well  as  economic  losses  of  no  small  magnitude  are  avoided.  It  is 
not  difficult  to  show  that  a  pure  and  wholesome  water  supply  is  a 
sanitary  investment  which  pays  in  dollars  and  cents,  as  well  as  in 
health  and  happiness. 

For  industrial  purposes  many  waters  are  entirely  unfitted  for 
use  without  purificarion.  Usually  the  dissolved  salts  which  pro- 
duce hardness  make  a  water  unsuitable  for  use  in  steam  boilers. 
The  softening  of  hard  waters  is  a  method  of  purification  which 
has  long  been  practised,  and  which  could  be  even  more  exten- 
sively employed  today  to  the  financial  advantage  of  users  of 
boiler  scale-forming  waters.  In  general,  soft  waters  are  more 
desirable  for  most  industrial  purposes.  In  such  industries  as 
brewing,  distilling,  starch  and  sugar  manufacturing,  the  bacterial 
content  of  the  water  is  of  some  importance,  since  certain  organ- 
isms may  produce  fermentations  which  are  undesirable.  Waters 
containing  iron  in  solution  or  suspension  are  especially  objection- 
able for  domestic  purposes  and  for  such  industries  as  dyeing, 
paper  making  and  bleaching. 

General  Methods  of  Purification. — The  impurities  in  water 
which  it  is  necessary  to  remove  are  in  suspension  and  solution,  and 
purification  methods  fall  naturally  into  two  classes  in  consequence. 
For  the  removal  of  matter  in  suspension  two  processes  are  em- 
ployed, namely,  sedimentation  and  filtration.     These  processes 

49 


50  WATER  PURIFICATION 

occur  normally  in  nature,  but  being  uncontrolled  produce  re- 
sults of  varying  degrees  of  excellence.  When  artificially  controlled, 
sedimentation  and  filtration,  properly  carried  out,  are  able  to 
purify  waters  of  very  poor  quality,  and  with  a  degree  of  certainty 
that  have  established  these  methods  upon  a  scientific  basis. 
Mineral  matter  in  suspension  can  be  entirely  removed,  as  well  as 
practically  all  of  the  organic  matter,  which  latter  includes  the 
bacteria.  Chemical  reagents  may  or  may  not  be  employed  to 
facilitate  either  or  both  of  these  processes.  The  chemical  com- 
pounds used  in  this  connection  are  mechanical  agents  exclusively, 
and  merely  assist  and  hasten  the  bringing  together  of  finely  divided 
suspended  particles  into  larger  masses,  which  will  settle  and  filter 
out  with  greater  ease. 

For  the  removal  of  dissolved  impurities,  such  as  salts  of  lime 
and  magnesia,  chemical  reagents  are  usually  resorted  to.  A  pre- 
cipitation of  the  bases  in  an  insoluble  form,  and  their  subsequent 
removal  by  sedimentation  and  filtration,  constitute  the  usual 
methods  of  purification.  In  the  case  of  easily  oxidizable  soluble 
salts,  such  as  those  of  iron,  aeration  is  effective  in  breaking  down 
the  ferrous  compounds  by  the  removal  of  dissolved  carbon  dioxide, 
and  the  introduction  of  sufficient  oxygen  to  bring  about  the  oxida- 
tion of  the  ferrous  to  the  insoluble  ferric  compounds. 

Filtration  usually  follows  the  process  of  sedimentation,  and  is 
carried  out  in  various  ways.  Sand,  gravel,  coke,  charcoal,  coal, 
porous  tile  and  sponges  are  some  of  the  materials  which  have  been 
and  are  still  used  as  filtering  media.  Onh^  the  first  two  media, 
viz.,  sand  and  gravel,  are  commonly  employed  in  large  filtration 
plants. 

Purification  methods  which  do  not  fall  under  any  of  the  above- 
mentioned  classes,  are  those  of  distillation  and  sterilization  by 
chemical  or  physical  agents.  In  the  case  of  distillation  all  dis- 
solved and  suspended  impurities  are  removed,  as  well  as  the 
bacteria  being  killed.  It  is  obviously  the  most  perfect  method  of 
purification,  but  not  necessarily  the  most  desirable.  Its  cost 
makes  it  prohibitive  for  all  practical  purposes  so  far  as  large 
supplies  are  concerned. 

Disinfection  and  sterilization  present  the  most  advanced 
phases  of  water  purification,  and  furnish  methods  of  great  prac- 
tical value  for  supplementing  other  methods,  as  well  as  in  some 
cases,  furnishing  themselves  all  the  purification  needed. 

The  action  of  these  agents  is  primarily  upon  the  readily  oxidiz- 


WATER  PURIFICATION  51 

able  matter  which  may  be  in  a  water,  whether  it  be  of  organic  or 
of  mineral  origin.  The  bacteria  being  a  low  form  of  plant  life  are 
thus  attacked,  and  are  either  destroyed  or  their  vitahty  so  im- 
paired that  they  no  longer  are  able  to  reproduce  themselves. 
The  effect  of  these  agents  is  not  selective,  except  in  so  far  as  they 
react  first  with  the  more  easily  oxidized  material.  If  not  ex- 
hausted on  the  latter,  they  will  react  slowly  upon  the  more  re- 
sistant matter,  but  may  fail  to  effect  a  complete  oxidation  or  to 
hinder  vital  processes.  Those  bacteria  which  are  capable  of  form- 
ing spores,  and  which  may  be  in  this  stage,  are  sometimes  not 
affected,  on  account  of  their  well-known  resistance  to  injurious  con- 
ditions. The  effect  of  these  disinfectants  upon  the  pathogenic 
forms  is  the  same  as  upon  the  non-disease-producing  bacteria 
which  may  be  present  in  a  water.  If  practically  all  the  bacteria 
are  destroyed,  therefore,  the  water  is  rendered  hygienically  safe. 

The  agents  used  for  disinfecting  water  are  the  hypochlorites  of 
calcium  and  sodium,  pure  chlorine  gas,  ozone,  copper  sulphate, 
and  the  rays  of  ultra-violet  light.  Other  chemical  compounds 
have  been  suggested  and  experimented  with,  and  the  use  of  the 
electric  current  has  frequently  been  tried.  None  of  these  agents 
is  of  much  practical  value  excepting  chlorine  and  its  compounds 
in  the  form  of  the  hypochlorites,  ozone  gas  and  the  ultra-violet 
light.  All  of  the  chemical  disinfectants  are  used  in  very  minute 
amounts  in  water-purification  work;  nevertheless,  they  are  ex- 
tremely effective  when  properly  applied.  The  relative  merits  of 
these  various  agents  will  be  discussed  in  a  later  chapter. 

The  use  of  copper  sulphate  as  an  agent  for  killing  off  growths 
of  algse  and  diatoms  in  water  supplies  is  of  practical  importance, 
and  of  much  value.  The  troublesome  character  of  these  growths 
in  the  operation  of  water-woi'ks,  and  even  of  purification  plants 
will  be  considered  in  detail  later. 

Statistics  of  Purification  Plants. — Sixty  years  ago  in  the  United 
States  no  plants  for  the  purification  of  a  public  water  supply  were 
in  existence.  In  Europe  plants  which  clarified  the  water  were 
in  use,  and  incidentally  effected  a  certain  degree  of  bacterial  puri- 
fication. They  were  operated  upon  an  empirical  basis  and  with 
no  true  understanding  of  the  scientific  principles  underlying  the 
art.  In  fact,  the  modern  theory  of  the  cause  for  many  diseases 
was  unknown,  and  the  part  played  by  minute  vegetable  and 
animal  organisms  in  the  dissemination  of  disease  was  undreamed 
of.     By  the  wonderful  investigations  and  deductions  of  Pasteur 


52 


TT'^  TER  P  URIFICA  TION 


and  Koch,  the  relation  between  polluted  waters  and  the  trans- 
mission of  certain  diseases  became  apparent  and  furnished 
a  rational  basis  upon  which  methods  of  purification  could  be 
constructed. 

Statistics  compiled  by  Prof.  Geo.  C.  Whipple  in  1911  show 
the  growth  of  water  purification  plants  from  1880  to  1910.  The 
figures  are  given  in  the  accompanying  table. 

Populations  Supplied  with  Filtered  Water  at  Different  Dates 


Total  urban  popu- 
lation in  U.  S., 
places  of  more 

than  2,500  inhabi- 
tants 

Population  supplied  with  filtered  water 

Per  cent,  of 

Year 

Sand  filters 

Mechanical 
filters 

Total 

urban  popu- 
lation sup- 
plied 

1870 
1880 
1890 
1900 
1910 

None 

None              None 

0.00 

13,300,000 
21,400,000 
29,500,000 
38,350,000 

30,000 

35,000 

360,000 

3,883,221 

275,000 
1,500,000 
6,922,361 

30,000 

310,000 

1,860,000 

10,805,582 

0.23 

1.45 

6.30 

28.20 

Since  these  statistics  were  compiled  the  filtration  of  public 
water  supplies  has  steadily  increased,  and  must  continue  to 
increase  as  the  population  grows  and  the  pollution  of  the  sources 
of  water  supply  becomes  greater. 

From  figures  compiled  in  1924  by  C.  G.  Gillespie,  Director  of 
the  Bureau  of  Sanitary  Engineering  of  the  State  Board  of  Health 
of  California,  it  has  been  found  that  there  are  634  cities  in  the 
United  States  which  have  filter  plants  of  one  million  gallons,  or 
over,  capacity  daily.     His  summarized  statistics  are  as  follows: 


Type  of  plant 


Number  of 

cities  having 

plants 


Total 
installed  ca- 
pacity m.g.d. 


Total  popu- 
lation served 


Slow  sand  filters.  . 
Rapid  sand  filters 

Total 


47 
587 

634 


916.0 
4,048.4 

4,964.4 


5,054,000 
18,610,000 

23,664,000 


Statistics  Relating  to  Disinfection  of  Water  Supplies, — The 
phenomenal  increase  in  the  number  of  disinfected  water  supplies 
during  the  past  15  years  is  evidence  of  the  value  of  this  form  of 
treatment  for  rendering  drinking  water  safe  from  a  sanitary 
standpoint.  It  is  estimated  by  Linn  H.  Enslow  that  in  1911  over 
800  million  gallons  of  water  per  day  were  being  treated  with 


WATER  PURIFICATION  53 

bleaching  powder  in  a  score  of  the  larger  cities  of  the  United 
States  and  Canada.  In  1913,  liquid  chlorine  began  to  replace  the 
use  of  bleaching  powder,  although  over  1,700  millions  of  gallons 
of  water  per  day  were  being  treated  with  the  latter  reagent. 
By  1915,  approximately  one  billion  gallons  of  water  per  day  were 
being  treated  with  bleaching  powder  and  a  like  volume  with 
liquid  chlorine.  In  1918,  more  than  a  thousand  cities  and  towns 
in  North  America  were  employing  disinfection;  and  by  1924, 
3,750  million  gallons  per  day  were  being  treated  with  liquid 
chlorine,  and  50  million  gallons  per  day  with  bleaching  powder. 
It  is  estimated  that  6,000  installations  of  apparatus  are  required 
for  this  purpose,  of  which  12  are  electrolytic-cell  plants  where 
the  chlorine  is  produced  at  the  point  of  application  to  the  water, 
25  hypochlorite  plants,  and  the  balance  liquid  chlorine  plants. 

The  above  figures  presumably  include  filter  plants  using 
chlorine  for  disinfecting  the  water  after  filtration  as  well  as  those 
public  supphes  which  do  not  have  the  water  filtered. 

The  investment  represented  by  purification  plants  mounts  into 
many  millions  of  dollars,  but  the  conservation  of  life  effected  by 
them,  measured  in  dollars,  is  very  many  times  more  than  their 
first  cost.  Efficiency  of  these  plants  is  of  the  utmost  importance 
both  from  the  hygienic  and  economic  standpoint,  and  is  becoming 
more  and  more  appreciated  by  communities  owning  them.  To 
the  technical  problem  of  purification  much  study  has  been  given 
in  the  past,  and  is  still  being  given  at  present;  but  the  art  as  a 
whole  is  on  a  thoroughly  scientific  and  practical  basis,  as  the 
beneficial  results  of  water  purification  amply  testify. 

References 

1.  Allen  Hazen:  "Filtration  of  Public  Water  Supplies."     Appendix  IV. 

2.  "Water  Purification  in  the  United  States."     E7ig.  Neivs,  vol.  47,  p.  310. 

3.  Rideal:   "Water  and  Its  Purification." 

4.  G.    C.    Whipple:   "The    Present   Status  of   Water   Purification   in   the 

United  States,  etc."     Proc.  Congress  of  Technology,  Boston,  1911. 

5.  Eng.  Record: 

(a)   Geo.   W.   Fuller:  "Importance  of  Proper  Operation  of  Water 
and  Sewage  Purification  Plants."     Vol.  58,  p.  498,  1908. 
(6)   George  C.  Whipple:  "Policy  of  Water  Filtration."     Vol.  60,  p. 
718,  1909. 

(c)  Rudolph  Hering  and  George  W.  Fuller:  "History  of  Devel- 
opment of  Water  Purification — Report  on  Montreal  Supply."  Vol. 
62,  p.  539,  1910. 

6.  C.    G.    Gillespie:  "Filtration    Plant    Census."     Jour.    Amer.    Water  ^-^i 

Works  Assoc,  August,  1925. 


CHAPTER  VI 

SEDIMENTATION 

The  removal  of  suspended  mineral  and  organic  matter  from 
water  supplies  is  most  easily  accomplished  by  settling  in  reser- 
voirs. The  clarification  and  purification  effected  by  settling  in 
this  manner  is  usually  spoken  of  as  plain  sedimentation.  By  the 
addition  of  certain  chemical  compounds  to  naturally  turbid 
waters,  an  artificial  flocculation  or  coagulation  of  the  suspended 
particles  is  produced  whereby  they  settle  out  more  rapidly. 
This  process  is  termed  "sedimentation  after  coagulation"  or 
"precipitation  with  chemicals." 

PLAIN  SEDIMENTATION 

Turbid  surface  waters  flowing  into  impounding  reservoirs  be- 
come partially  clarified  in  passing  slowly  through  or  in  standing 
undisturbed  in  them.  This  result  is  produced  because  the  veloc- 
ity of  flow  of  the  water,  which  has  maintained  these  minute  par- 
ticles in  suspension,  is  lessened  or  reduced  to  practically  zero. 
The  reduction  in  the  velocity  of  flow  permits  the  sediment  to 
gravitate  toward  the  bottom  of  the  reservoir,  and  to  be  finally 
deposited  thereon.  Provided  the  wind  does  not  set  up  surface 
currents,  nor  temperature  changes  produce  vertical  currents,  the 
sooner  will  the  deposition  of  the  sediment  be  effected.  For  ob- 
vious reasons  a  perfectly  quiescent  condition  of  the  water  can  not 
be  produced,  and  complete  clarification,  therefore,  is  not  usually 
practicable. 

The  ponds  and  lakes  are,  as  a  rule,  natural  setthng  reservoirs, 
in  so  far  as  the  turbid  surface  waters  which  may  flow  into  them 
are  concerned.  The  water  in  them  is,  therefore,  usually  quite 
clear.  The  factors  tending  to  disturb  the  water  and  to  retard 
settlement  are  of  course  active  at  times. 

Artificial  settling  reservoirs  and  coagulation  basins  form  a 
most  important  part  of  purification  plants  treating  clay-bearing 
waters.  As  a  preliminary  treatment  for  filtration  they  are 
absolutely  essential  for  efficiently  and  economically  handling 
these  turbid  waters.     They  may  be  also  of  much  assistance  in 

.54 


SEDIMENTATION  55 

purifying  waters  which  are  only  turbid  during  portions  of  the 
year.  In  water  softening,  settHng  basins  are  essential  to  obtain- 
ing a  well-clarified  effluent. 

Theory  of  Sedimentation.- — The  very  practical  importance  of 
sedimentation  in  water  purification  has  caused  the  conditions 
which  influence  the  phenomena  to  be  carefully  studied.  Much  of 
the  experimental  work  has  of  necessity  been  empirical  in  charac- 
ter, because  of  the  number  of  factors  which  are  involved.  Some 
investigators,  however,  have  approached  the  problem  from  a 
theoretical  standpoint,  and  many  of  their  deductions  are  valuable 
and  suggestive.  The  early  work  of  Brewer,  Durham,  Hunt, 
Barus  and  Seddon  is  of  interest  in  connection  with  this  general 
subject.  The  more  recent  paper  of  Allen  Hazen,^  however, 
has  a  more  practical  bearing  on  sedimentation,  although  discuss- 
ing in  a  theoretical  manner  certain  important  factors. 

From  the  simple  assumption,  "that  whenever  a  particle  of  sus- 
pended matter  hits  the  bottom,  it  remains  where  it  strikes,  and 
is  never  carried  forward  on  the  bottom,  or  picked  up  again; 
second,  that  all  the  sediment  in  the  water  is  of  the  same  hydraulic 
value,  that  is  to  say  that  every  particle  settles  through  the  water 
at  the  same  rate  as  every  other  particle,"  Hazen  proceeds  to 
develop  mathematically  fifteen  special  propositions.  He  dis- 
cusses the  velocity  with  which  particles  of  sediment  settle 
through  still  water,  and  the  effect  of  temperature,  flocculation 
and  coagulation  upon  the  rate  of  settlement. 

The  minute  size  of  the  particles,  their  very  slow  subsiding 
velocities,  and  the  mixing  action  produced  by  surface  and  vertical 
currents,  all  tend  to  maintain  the  finest  particles  in  practically 
constant  suspension.  Hazen  gives  a  table  of  the  subsiding 
velocities  of  various-sized  particles  which  is  reproduced  on  the 
following  page. 

Hazen's  resume  of  the  deductions  and  conclusions  at  the  close 
of  the  paper  is  clearly  and  concisely  stated  as  follows: 

Resume 

"The  fundamental  proposition,  in  clearing  water  by  sedimentation, 
seems  to  be  that  every  particle  of  sediment  moves  downward  through 
the  water  at  a  velocity  depending  upon  its  size  and  weight  and  upon 
the  viscosity  of  the  water.  Particles  of  sediment  are  generally  so  far 
apart  that  they  do  not  influence  each  other;  and,  while  there  is  no  doubt 

1  Trans.  Am.  Soc.  C.  E.,  vol.  53,  p.  45,  1904. 


56  WATER  PURIFICATION 

Velocities  at  which    Particles  of  Sediment  Fall  in  Still  Water 


T-..         ,        r        .-1        Hydraulic   value  in 
Diameter  of  particles   .^ifii^eters  per  second 
in  millimeters  ^qoq    =  50°Fahr. 


Remarks 


1 . 0000 

100.0 

0.8000 

83.0 

0.6000 

63.0 

0.5000 

53.0 

0.4000 

42.0 

0.3000 

32.0 

0.2000 

21.0 

0.1500 

15.0 

0.1000 

8.0 

0.0800 

6.0 

0.0600 

3.8 

0.0500 

2.9 

0.0400 

2.1 

0.0300 

1.3 

0.0200 

0.62 

0.0150 

0.35 

0.0100 

0.154 

0.0080 

0.098 

0.0060 

0.055 

0.0050 

0.0385 

0.0040 

0.0247 

0.0030 

0.0138 

0.0020 

0.0062 

0.0015 

0.0035 

0.0010 

0.00154 

0.0001 

0.0000154 

Experiments  by  the  writer. 

Experiments  by  the  writer. 
Experiments  by  the  writer. 
Experiments  by  the  writer. 
Experiments  by  the  writer. 
Experiments  by  the  writer. 
Experiments  by  the  writer. 
Experiments  by  the  writer. 
Experiments  by  the  writer. 
Interpolated  from  connecting  curve. 
Interpohitcd  from  connecting  curve. 
Interpolated  from  connecting  curve. 
Interpolated  from  connecting  curve. 
Interpolated  from  connecting  curve. 
Wiley's  formula. 
Wiley's  formula. 
Wiley's  formula. 
Wiley's  formula. 
Wiley's  formula. 
Wiley's  formula. 
Wiley's  formula. 
Wiley's  formula. 
Wiley's  formula. 
Wiley's  fornuila. 
Wiley's  formula. 
Wiley's  formula. 


Note. — These  values  are  not  given  as  being  precise,  but  they  arc  believed 
to  be  sufficiently  accurate  for  the  purpose  of  this  discussion. 

that  they  do  sometimes  collect  in  groups  and  thus  change  the  condi- 
tions, it  seems  to  be  generally  true  that  each  particle  will  settle  as  if  no 
other  particle  were  present. 

"If  the  water  in  a  basin  were  absolutely  quiet  there  would  be  a 
regular  sequence  of  clearing  beginning  at  the  top.  The  coarsest  particles 
would  go  down  fastest,  but  at  any  given  point  there  would  be  a  gradual 
clearing,  and  this  clearing  would  take  place  most  rapidly  at  the  top, 
and,  after  longer  intervals,  at  lower  points  in  the  basin. 

"Seddon  started  out  with  this  theory,  but  found  it  to  be  not  in 
accordance  with  the  facts.  His  observation  showed  that  while  the 
amount  of  sediment  in  the  water  in  the  top  was  a  little  less  than  in  the 
water  in  the  bottom,  the  distribution  was  nearly  equal  throughout 
the  mass,   a  condition  of  affairs  inconsistent  with  the  theory.     He 


SEDIMENTATION  o7 

accounted  for  this  distribution  of  sediment  by  the  constant  mixing 
of  the  water  from  top  to  bottom,  and  to  the  sustaining  power  of  vortex 
motions  in  the  water.  These  motions  he  thought  arose  from  the  internal 
motion  of  the  water  at  the  time  of  entrance,  and  from  wind,  and  from 
temperature  changes. 

"The  writer  has  taken  Seddon's  development  of  the  case  as  his 
starting  point,  and  has  carried  the  discussion  further.  He  believes  that 
while  the  internal  motions  keep  the  water  mixed,  and  with  nearly  the 
same  density  of  sediment  from  top  to  bottom,  the  tendency  of  the 
particles  of  sediment  to  settle  is  nevertheless  an  unbalanced  force  always 
acting  to  take  the  particles  to  the  bottom,  and  the  number  of  particles 
that  hit  the  bottom  in  a  given  time  is  proportional,  first,  to  the  velocity 
at  which  the  individual  particles  settle,  and  second,  to  the  density  of 
sediment  in  the  water  immediately  above  the  bottom. 

"With  these  fundamental  relations  in  mind,  it  is  easy  to  compute 
and  to  express  by  simple  formulas  the  proportions  of  particles  of  sedi- 
ment of  a  given  hydraulic  value  which  will  hit  the  bottom  under  given 
conditions  and  which,  therefore,  presumably  will  be  removed. 

"The  fundamental  propositions  may  be  very  concisely  expressed. 
They  are:  First,  that  the  results  obtained  are  dependent  upon  the 
area  of  bottom  surface  exposed  to  receive  sediment,  and  that  they 
are  entirely  independent  of  the  depth  of  the  basin;  and  second,  that  the 
best  results  are  obtained  when  the  basins  are  arranged  so  that  the 
incoming  water  containing  the  maximum  quantity  of  sediment  is  kept 
from  mixing  with  water  which  is  partially  clarified.  In  other  words, 
the  best  results  are  obtained  where  any  given  lot  of  water  goes  through 
the  basin  with  the  least  mixing  with  the  water  which  enters  after  it. 
Tills  is  practically  accomplished  by  dividing  the  basins  into  consecutive 
apartments  by  baffles  or  otherwise. 

"Thus  far,  the  discussion  is  easy  and  apparently  certain.  The  next 
step  is  a  more  difficult  one.  It  relates  to  bottom  velocities,  and  has  to 
do  with  the  question  whether  these  velocities  are  such  as  to  allow  the 
particles  to  remain  on  the  bottom  when  they  get  there,  or  whether 
they  will  be  taken  up  again  and  be  kept  in  motion  with  the  body  of  the 
water.  This  is  a  point  upon  which  further  experimental  data  are 
needed.  The  problem  of  securing  such  data  seems  to  be  difficult. 
The  observations  must  be  made  at  the  bottom  of  a  layer  of  liquid  of 
considerable  thickness,  where  the  conditions  of  observation  are  not 
favorable.  The  observations,  further,  must  be  made  on  very  low 
velocities  and  on  particles  so  small  as  to  be  practically  microscopic. 

"Whatever  view  may  be  taken  of  the  second  part  of  the  problem,  or 
whatever  researches  upon  it  may  show,  the  arrangements  of  basms 
most  favorable  to  taking  particles  to  the  bottom  should  stand." 


58  WATER  PURIFICATION 

Coagulation  of  Sediment.— The  collection  of  minute  particles 
of  suspended  matter  into  larger  aggregates  is  a  process  which 
appears  to  occur  only  to  a  limited  extent  in  natural  waters,  al- 
though it  may  be  artificially  produced  by  the  addition  of  certain 
chemical  compounds.  Since  the  deposition  of  these  finely  divided 
particles  greatly  facilitates  sedimentation  and  subsequent  filtra- 
tion, when  the  latter  forms  a  part  of  the  purification  process,  the 
phenomenon  is  of  much  practical  interest  and  importance  in 
water  purification. 

Colloidal  Character  of  Sediment. — The  analogy  between  the 
physical  properties  of  liquid  suspensions  of  certain  chemical  com- 
pounds, and  those  of  turbid  clay-bearing  waters  is  so  marked  that 
some  description  of  their  properties  will  be  of  interest.  In  a  paper^ 
published  some  years  ago  by  the  author  attention  was  called  to 
Dr.  A.  A.  Noyes'2  classification  of  colloidal  mixtures,  in  which  he 
defined  non-viscous,  non-gelatinizing,  but  readily  coagulable 
mixtures  as  ''colloidal  suspensions."  He  regarded  them  as 
really  suspensions  of  minute  particles  and  not  true  solutions.  He 
differentiates  further  by  designating  mixtures  in  which  the  par- 
ticles may  be  visible  under  the  microscope  as  "microscopic  sus- 
pensions," reserving  the  term  "colloidal  suspensions"  for  those 
containing  particles  beyond  the  Hmit  of  microscopic  visibility. 

Physical  and  Electrical  Properties.— Natural  waters  furnish 
many  examples  of  colloidal  suspensions.  Most  turbid  waters  in 
their  natural  condition  may  be  considered  as  mixtures  of  "mi- 
croscopic and  colloidal  suspensions."  A  beam  of  light  passed 
through  Ohio  River  water,  even  after  the  water  has  stood  for 
many  weeks,  is  plainly  visible,  in  the  same  manner  as  when  a  sun- 
beam passes  through  dusty  air.  This  is  a  fairly  good  proof  of  the 
presence  of  minute  particles  in  suspension,  which  reflect  the  light 
from  their  surfaces.  Colloidal  suspensions  of  gold  and  arsenious 
sulphide  artificially  prepared  act  in  a  similar  manner  to  a  ray  of 
light. 

Another  property  of  colloidal  and  microscopic  suspensions, 
appai-ently  depending  on  the  presence  of  electric  charges  upon 
them,  is  seen  in  the  migration  of  the  colloidal  particles  by  the 
passage  of  an  electric  current  through  the  mixture.  Thus  the 
particles  of  a  colloidal  suspension  of  ferric  hydroxide  or  aluminum 

i"The  Coagulation  and  Precipitation  of  Impurities  in  Water  Purifica- 
tion."    Eng.  Record,  vol.  51,  May  13,  1905. 
'  Jour.  Am.  Chem.  Soc,  vol.  27,  No.  2. 


SEDIMEN  TA  TION  59 

hydroxide  migrate  with  the  positive  current  toward  the  cathode, 
while  kaoHn  and  other  similar  colloidal  or  microscopic  particles 
migrate  with  the  negative  current  toward  the  anode.  The  col- 
loidal suspension  of  clay  particles  characteristic  of  turbid  Ohio 
River  water  act  in  a  precisely  similar  manner  to  those  of  kaolin. 
Experiments  by  the  author  with  slightly  turbid  Ohio  River  water 
produced  an  average  rate  of  travel  of  the  particles  toward  the 
anode  of  0.29  cm.  per  hour  for  a  potential  gradient  of  4  volts  per 
centimeter.  These  results  are  of  the  same  order  of  magnitude  as 
were  obtained  by  Whitney  and  Blake  for  the  migration  of  par- 
ticles in  a  colloidal  suspension  of  silicic  acid,^  Silica  doubtless 
formed  the  larger  proportion  of  the  particles  in  the  sample  of 
Ohio  River  water  with  which  the  above  experiments  were  made. 
The  clearing  of  the  water  at  the  cathode  and  the  movement  of  the 
particles  toward  the  anode  indicated  them  to  be  negatively 
charged. 

Coagulation  with  Chemicals. — Turning  now  to  another  impor- 
tant property  of  colloidal  suspensions  in  general,  viz.,  their  coagula- 
tion, we  find  that  turbid  natural  waters  possess  similar  properties 
to  colloidal  suspensions  artificially  prepared.  Non-electrolytes 
do  not  have  the  power  of  coagulating  colloidal  suspensions; 
but  all  electrolytes  will  do  so  with  a  proper  degree  of  concentration 
of  the  latter,  and  a  sufficient  length  of  time.  Acids,  bases  and 
salts,  therefore,  will  coagulate  colloidal  suspensions,  all  being 
more  or  less  dissociated  in  aqueous  solution  and  capable  of  con- 
veying an  electric  current.  In  a  similar  manner  hydrochloric  acid, 
caustic  soda,  caustic  lime  or  an  ordinary  salt  solution  will  each,  if 
of  the  proper  concentration,  coagulate  the  colloidal  claj^  of  a  natu- 
rally turbid  water  like  that  in  the  Ohio  River.  It  should  always 
be  borne  in  mind  in  dealing  with  natural  waters  that  we  are  work- 
ing with  extremel}^  dilute  solutions  of  salts,  even  in  those  waters 
which  we  commonly  speak  of  as  high  in  soluble  compounds.  In 
other  words  the  concentration  of  the  electrolytes  is  very  low. 

In  a  number  of  modern  purification  plants  in  which  lime  is  used 
to  soften  the  water,  in  addition  to  the  employment  of  sulphate  of 
iron  or  aluminum  sulphate  to  effect  clarification,  the  action  of  the 
caustic  alkali  is  virtually  that  of  a  coagulant,  especially  if  the  lime 
is  added  in  sufficient  amounts  to  produce  a  slightly  caustic  con- 
dition. Some  quite  extensive  experiments  undertaken  on  a  large 
scale  some  years  ago,  clearly  proved  that  by  rendering  the  Ohio 

*  Jour.  Amer.  Chem.  Soc,  October,  1904. 


60  WATER  PURIFICATION 

River  water  caustically  alkaline  with  lime,  the  suspended  clay 
could  be  coagulated,  and  could  be  subsequently  removed  by 
rapid  filtration  through  sand.  The  explanation  of  the  phenom- 
enon seems  to  be  that  the  introduction  of  the  base  calcium  hy- 
drate in  excess,  furnished  an  electrolyte  of  sufficient  concentration 
in  the  water  to  effect  coagulation  of  the  colloidal  clay. 

The  similarity  between  artificiall}^  prepared  colloidal  suspen- 
sions and  the  verj^  small  suspended  particles  characteristic  of  tur- 
bid waters  as  above  described,  seems  to  warrant  the  classification 
of  the  latter  mixtures  with  the  former.  If  we  seek  further  analo- 
gies in  the  co-precipitation  or  absorption  by  colloids  of  other  sub- 
stances in  solution  and  suspension  with  them  when  these  colloids 
are  coagulated,  the  true  character  of  turbid  waters  seems  even 
more  apparent. 

When  a  turbid  water  containing  carbonates  of  lime  and  mag- 
nesia in  solution  is  treated  with  sulphate  of  alumina,  for  example, 
a  reaction  results  setting  free  aluminum  hydroxide.  The  liber- 
ated aluminum  hydroxide  absorbs  or  mechanically  traps  the 
finely  divided  material  in  suspension,  and  in  the  course  of  time 
the  finer  particles  come  together  to  form  larger  particles  which 
settle  out  readily.  This  process  of  coagulation  of  the  colloidal 
clay  suspension  and  co-precipitation  of  the  larger  suspended  par- 
ticles, appear  to  be  started  by  the  coagulation  of  the  aluminum 
hydroxide.  It  is  evident,  as  Dr.  Noj^es  points  out  in  the  paper 
referred  to,  "that  the  mechanism  of  this  coagulation  is  not  j^et 
understood,"  although  it  appears,  "that  it  is  the  ion  with  a  charge 
opposite  to  that  of  the  colloid  particles  that  is  mainly  responsible 
for  their  coagulation."  It  is  also  probable  that  it  is  the  electric 
charge  upon  the  particle  which  tends  to  hold  it  in  suspension. 

Time  Required  for  Coagulation. — The  time  factor  in  the 
coagulation  of  turbid  waters  is  an  important  one.  The  length  of 
time  required  to  bring  about  the  proper  degree  of  coagulation 
varies  with  different  waters.  It  is  apparently  influenced  by  the 
character  of  the  suspended  colloids,  by  the  kind  and  amount  of 
salts  in  solution,  by  the  quantity  of  the  coagulating  chemical 
applied,  by  the  temperature  of  the  water,  and  by  the  agitation  to 
which  the  mixture  is  subjected.  The  presence  of  much  organic 
matter  may  retard  the  formation  of  the  floe.  An  ordinary  clay 
suspension  is  usually  quickly  coagulated.  Complete  flocculation 
and  partial  clarification  may  take  place  under  favorable  condi- 
tions in  as  short  a  time  as  10  or  15  min.;  on  the  other  hand,  the 


SEDIMENTATION  61 

author  has  seen  a  water  which  was  not  affected  at  the  end  of  4 
his.  This  water  had  only  a  slight  turbidity.  Two  hours  after 
applying  the  coagulating  chemical  to  this  water  it  could  be  passed 
through  a  sand  filter  without  removing  the  suspended  matter  or 
all  of  the  coagulant.  The  large  amount  of  soluble  organic 
matter  present  in  the  water  evidently  produced  a  colloidal  sus- 
pension with  the  aluminum  hydroxide,  formed  by  the  decomposi- 
tion of  the  aluminum  sulphate  which  was  added,  and  was  only 
slowly  precipitated. 

Under  ordinary  conditions  a  period  of  3  to  5  hrs.  is  sufficient  to 
properly  coagulate  and  settle  a  turbid  water.  Of  course  a  proper 
arrangement  of  basins,  inlets,  outlets  and  baffles  is  requisite  for 
bringing  about  effective  coagulation  and  clarification.  The  con- 
struction of  typical  coagulation  and  settling  basins  will  be  de- 
scribed in  a  succeeding  chapter. 

Loss  of  Chemical  by  Adsorption.  —  Another  phenomenon 
common  to  the  coagulation  of  turbid  waters  is  the  adsorption  of 
the  applied  chemical  compound  by  the  flocculent  material  pro- 
duced by  the  reaction  and  the  associated  suspended  matter.  The 
particular  significance  of  this  fact  in  water  purification  was  pointed 
out  by  George  W.  Fuller  in  his  report  on  the  experiments 
on  Ohio  River  water  made  at  Louisville,  Ky.,  some  years  ago. 
While  adsorption  or  concentration  of  a  liquid  on  the  contact- 
surface  of  the  particles  is  not  alone  pecuHar  to  colloids,  but  is 
characteristic  of  all  solid  precipitates,  it  has  a  practical  bearing 
on  coagulation  in  that  there  is  some  loss  of  chemical  on  account 
of  this  property.  The  amount  of  chemical  thus  lost  is  probably 
proportional  to  the  surface  area  of  the  particles  of  the  precipitate, 
and  is  a  function  of  the  nature  of  the  solid  and  dissolved  bodies, 
and  of  the  concentration  of  the  latter. 

Natural  Colloids.— The"  schmutzdecke,"  which  forms  on  the 
top  of  slow  sand  filters,  is  doubtless  a  true  colloid,  and  by  its 
adsorptive  power  removes  the  organic  and  inorganic  suspended 
matter.  The  difference  between  the  appearance  of  the  sand  m 
the  beds  of  slow  sand  filters  receiving  water  carrying  considerable 
dissolved  and  suspended  organic  matter,  and  those  which  receive 
water  holding  more  or  less  clay  in  suspension,  as  pointed  out 
by  George  W.  Fuller  in  his  discussion  of  Allen  Hazen's  paper 
"On  Sedimentation,"  previously  quoted  from,  may  possibly  be 
explained  by  a  difference  in  the  character  of  the  colloids  pro- 
duced or  carried  by  the  two  types  of  water.     The  colloid  formed 


62  WATER  PURIFICATION 

by  the  soluble  and  suspended  organic  matter  in  a  water  has  prob- 
ably the  characteristic  of  a  true  colloid,  i.e.,  a  viscous,  gelatinizing 
compound  not  coagulated  by  salts,  like  gelatine  for  example.  On 
the  other  hand  clay-bearing  waters  produce  or  carry  a  colloid 
which  is  non-viscous  and  non-gelatinizing,  but  which  may  be 
readily  coagulated  by  compounds  like  aluminum  sulphate  or 
sulphate  of  iron.  Bacterial  activity  very  likely  aids  in  the  forma- 
tion of  the  natural  scums  found  on  slow  sand  filters. 

In  the  absence  of  any  artificially  applied  coagulant  like  alumi- 
num hydrate  the  ''granular  appearance"  of  sand  beds  receiving 
clay-bearing  waters  indicates  the  formation  of  a  colloid  whose 
ability  to  prevent  the  passage  of  finely  divided  suspended  matter 
is  limited.  The  colloidal  particles  pass  deeper  into  the  sand  than 
would  a  true  gelatinous  colloid,  and  their  power  of  adsorption  is 
soon  exhausted.  The  "ripening"  of  the  sand  beds  of  both  slow 
sand  and  rapid  sand  filters,  is,  therefore,  probably  a  process  of  the 
formation  and  deposition  of  colloidal  particles  upon  the  sand 
grains,  by  which  their  adsorptive  power  is  slowly  increased. 

It  is,  therefore,  in  the  character  of  the  colloid  formed  that  one 
must  seek  for  the  explanation  of  the  phenomena  connected  with 
the  coagulation  and  filtration  of  natural  waters.  It  is  evident 
that  much  more  light  is  needed  on  a  great  many  of  these  prob- 
lems, but  that  some  of  them  are  associated  in  some  way  with  the 
colloidal  state  of  the  clay,  of  the  organic  matter  and  of  the  artifi- 
cially appHed  coagulants,  is  self-evident.  While  the  above  points 
may  at  first  sight  appear  to  be  only  of  scientific  interest,  they 
are  in  reality  intimately  associated  with  the  most  efficient  de- 
sign of  purification  plants.  The  size  of  coagulation  basins,  the 
mixing  and  the  agitation  of  the  raw  water  with  the  coagulant,  the 
ability  to  vary  the  period  of  coagulation  for  any  given  water,  the 
best  chemical  compounds  to  use  for  the  proper  coagulation  of  the 
suspended  matter  in  the  water,  the  advisability  of  providing 
for  plain  sedimentation  before  attempting  coagulation  and  the 
size  and  character  of  the  sand  grains  composing  the  filter  bed  arc 
all  factors  entering  into  the  practical  design  of  water-purifica- 
tion plants.  The  poor  results  often  obtained  in  the  operation 
of  plants  is  quite  as  frequently  the  result  of  improper  design  as 
of  faulty  methods  of  handling  the  plant. 


SEDIMENTATION  63 

Character  of  Forces  Affecting  Coagulation  and  Filtration. — It  is 

cbvious  that  the  nature  of  the  forces  in  play  in  coagulation  and 
filtration  phenomena  is  but  imperfectly  understood,  and  that 
their  magnitude  is  so  small  as  practically  to  place  them  in  the 
same  class  with  those  affecting  molecules.  Nevertheless,  some 
speculation  regarding  these  microforces,  as  Hannan^  has  so 
aptly  termed  them,  may  be  profitable.  He  discusses  the  prob- 
able influence  of  forces  of  "low  potential  energy"  only  capable 
of  acting  on  masses  of  practically  molecular  dimensions,  and 
within  exceedingly  small  ranges.  In  the  hquid-liquid,  gas- 
liquid  and  Hquid-solid  interfaces,  Hannan  believes  that  the 
molecules  of  liquid  "marshall  themselves  in  layers,  each  layer 
differing  probably  to  some  slight  extent  from  its  neighbors." 
This  orientation  of  the  molecules  in  fluids  (gas  or  liquid)  is 
probably  replaced  by  minute  stresses  in  the  case  of  solids.  "For 
brevity  the  whole  may  be  referred  to  as  an  interfacial  system." 
At  the  intersection  of  interfacial  systems  and  particularly  at  the 
solid-liquid-gas  junction,  Hannan  believes  that  these  micro- 
forces  are  especially  manifest. 

Hannan  regards  water  surfaces  as  layers  of  oriented  molecules 
probably  thousands  deep.  "When  the  water  surface  takes  a 
curved  form,  the  electromagnetic  field  thereby  developed  is 
such  that  the  following  laws  obtain:  (a)  convex  attracts  convex, 
(b)  concave  attracts  concave,  and  (c)  convex  and  concave  are 
mutually  repellent.  Not  only  so,  but  the  degree  of  attraction  or 
repulsion  is  seen  to  vary  with  the  curvature."  If  curvature, 
fineness  of  division  or  degree  of  dispersion  are  all  virtually  synony- 
mous terms,  we  have  the  stage  set  for  the  play  of  forces  in  inter- 
facial phenomena  that  must  produce  the  characteristic  results 
observed  in  and  around  the  capillaries  formed  by  the  deposits 
of  coagulum  about  the  sand  grains  of  a  filter  bed. 

Hannan  again  in  this  extremely  suggestive  paper  quotes  a 
statement  of  Bancroft'-  in  which  the  latter  refers  to  mordanting 
with  alumina  as  follows:  "The  substance  adsorbed  and  held 
firmly  is  colloidal  alumina.  Coagulated  alumina  may  be 
absorbed  to  some  extent,  but  it  easily  rubs  off  the  material." 
Hannan  goes  on  to  state  that  "by  colloidal  alumina  is  here  under- 

1  Frank  Hannan:  "Microforces:  with  Reference  More  Especially  to 
Orientation  and  Curvature."     Jour.  Amer.  Water  Works  Assoc,  Nov.,  1923. 

2  W.  D.  Bancroft:  "Mordants-alumina."  Jour.  Phys.  Chem.,  vol. 
26,  pp.  .501   536. 


\ 


64  WATER  PURIFICATION 

stood  the  hydrosol.  Water  filtration  is  quite  closely  analogous 
to  dyeing.  Complete  purification  of  the  water,  the  main  object 
sought,  corresponds  to  absolute  exhaustion  of  the  dye  bath.  The 
fabric  in  our  case  is  the  filter  bed;  and  fastness  to  washing  is,  for 
us,  undesirable.  In  these  two  short  sentences  of  Bancroft  may 
be  discerned  a  complete  vindication  of  the  empirically  developed 
mechanical  filtration  process.  The  preliminary  intimate  mixture 
of  alum  and  water  yields  colloidal  {i.e.,  hydrosol)  alumina  which 
adsorbs,  or  is  adsorbed  by,  the  impurities  to  be  removed;  a 
very  rapid  process.  Coagulation  follows;  the  alumina  is  left 
upon  the  sand  in  that  condition  in  which  it  "rubs  off  easily." 
"Massive  alumina  is  found  to  adsorb  color  with  great  readi- 
ness; but  bacteria  less  readily,"  Gore^  observes;  "the  act  of 
catching  impurities  by  the  alum  is  the  fundamental  point  of 
coagulation,  and  must  not  be  confounded  with  the  precipitation 
which  follows  it  .  .  .  Unless  the  impurity  is  caught  before 
filtration  very  little  prospect  remains  of  its  being  caught  at  all." 
It  is  not  to  be  doubted  that  bacteria  were  the  impurities  which 
Gore  had  in  mind,  and  Clark's-  results,  while  not  bearing  him  out 
to  the  letter,  yet  show  that,  so  far  as  bacteria  are  concerned. 
Gore's  statement  really  conveys  the  gist  of  the  matter.  An 
ingenious  plan,  in  principle  akin  to  Clark's  was  developed  by 
Maddock^  at  Oshkosh,  Wis.  His  results  should  throw  further 
light  on  this  problem.  It  is  desirable  we  should  know  to  what 
extent  the  hydrosol  form  of  alumina  is  indispensable  for  effective 
purification.  Whenever  it  can  safely  be  dispensed  with,  the 
way  to  marked  economies  in  coagulation  is  clear." 

""  References 

Theories  of  Sedimentation: 

1.  Brewer:  Proc.  Nat.  Acad.  Sciences,  November,  1883. 

2.  Durham:  Chem.  News,  vol.  30,  p.  57,  1874;  ibid.,  vol.  37,  p.  47,  1878. 

3.  Hunt:  Proc.  Boston  Soc.  Natural  History,  February,  1874. 

4.  Barus:  Bull.  U.  S.  Geological  Survey,  No.  36,  1886. 

5.  Seddon:  Jour.  Assoc.  Eng.  Soc,  p.  477,  1889. 

6.  Hazen:  Trans.  Amer.  Soc.  C.  E.,  vol.  53,  p.  45,  1904. 

7.  Longley:  Eng.  Record,  vol.  57,  pp.  793,  813. 

Coagulation  : 

1.  Ellms:  Eng.  Record,  vol.  51,  p.  552,  1905. 

1  William  Gore:  Eng.  Conir.,  vol.  58,  No.  6,  pp.  134-135,  1923. 

'  H.  W.  Clark:  Eng.  Contr.,  vol.  58,  No.  2,  pp.  16-17. 

3R.  A.   Maddock:  Eng.  Contr.  vol.  54,  p.  23,  Dec.  8,   1920,  ibid.,  vol. 
59,  No.  2.  p.  .381,  1922. 


SEDIMEN  TAT  I  ON  65 

2.  Eng.  Record,  vol.  54,  pp.  439,  475,  1906. 

3.  Eng.  Record,  vol.  58,  p.  292,  1908. 

4.  Eng.  Record,  vol.  61,  pp.  176,  215,  247,  599,  1910. 

5.  Eng.  Record,  vol.  64,  p.  476,  1911. 

6.  J.    W.    Ellms:  "Coagulation    and    Sedimentation    with    Chemicals." 

Jour.  Amer.  Water  Works  Assoc,  May,  1922. 

7.  Malcolm  Pirnie:  "Application  of  Colloid  Chemistry  to  Study  of  Filter 

Effluents."     Jour.  Amer.  Water  Works  Assoc,  p.  247,  March,  1922. 

8.  John  R.  Baylis:  "The  Solution  of  Corrosion  and  Coagulation  Problems  -^ 

at  Montebello   Filters,   Baltimore,   Md."     Jozir.  Amer.   Water   Works  '■^-^ 

Assoc,  p.  408,  May,  1922. 

9.  James  W.   Armstrong:  "Some   Observations  and   Experiences  in   the 

Operation  of  Coagulating  Basins."     Jour.  Amer.  Water  Works  Assoc, 
p.  160,  June,  1919. 

10.  F.  A.  Dallyn  and  A.  V.  Delaporte:  "Preparation  of  Water  for  Filtra- 

tion."    Eng.  Contr.,  p.  271,  Aug.  8,  1923. 

General  Experimental  Data: 

1.  Fuller:  "Report  on  Water  Purification  at  Louisville,  Ky." 

2.  Fuller:  "Report  on  Water  Purification  at  Cincinnati,  Ohio." 

3.  Weston:  "Report  on  Water  Purification  at  New  Orleans,  La." 

4.  "Report  of  Filtration  Commission  of  Pittsburgh,  Pa." 

5.  George  F.  Catlett:  "Colloidal  Theories  Applied  to  Colored  Water." 

Eng.  Record,  June  3,  1916. 

6.  Thorndyke  Saville:  "The  Nature  of  Color  in  Water."     Eng.  Contr., 

January  18,  1917. 

7.  Abel  Wolman:  "Cooperative  Research  in  Problems  of  Water  Purifica- 

tion."    Jorur.  Amer.  Water  Works  Assoc,  p.  572,  July,  1920.  C'-U^J' 

8.  Report  of  Committee  on  Colloidal  Chemistry,  Council  of  Standardiza-    C 

tion,  p.  273,  Jour.  Ayner.  Water  Works  Assoc,  March,  1923.  / 


CHAPTER  VII 

TYPES  OF  SETTLING  RESERVOIRS  AND  COAGULATION 

BASINS 

The  design  of  reservoirs  and  basins  for  efficient  sedimentation 
has  received  considerable  attention  in  water-purification  work, 
but  not  as  much  as  the  subject  deserves.  The  complexity  of  the 
problem  has  not  always  been  appreciated,  and  too  frequently  this 
portion  of  the  plant  has  been  made  to  accommodate  itself  to  other 
features  of  the  design  instead  of  being  properly  coordinated  with 
them. 

A  certain  amount  of  flexibility  in  the  design  of  these  reservoirs 
is  quite  possible  without  materially  sacrificing  efficiency.  Never- 
theless, it  is  the  author's  opinion  that  much  more  of  the  work  of 
clarification  should  be  thrown  on  this  part  of  the  purification 
plant  than  is  now  the  practice.  They  are  by  far  the  least  com- 
plicated portions  of  the  works,  and  the  least  expensive  to  operate. 
Their  tendency  to  make  more  uniform  the  operation  of  the  filters 
strongly  commends  them  to  the  operators  of  such  plants.  It  is 
believed  that  in  designing  settling  reservoirs  of  all  kinds  more  con- 
sideration should  be  given  to  obtaining  efficient  sedimentation 
and  to  making  them  as  large  as  good  engineering  practice  will 
permit,  due  regard  being  had  for  other  portions  of  the  plant,  in 
order  to  produce  a  well-balanced  design. 

Donaldson,  i  in  a  resume  of  the  need  for  plain  sedimentation 
prior  to  coagulation,  states  that  in  plants  handling  from  20  to  40 
tons  of  suspended  dry  solids  per  million  gallons,  such  as  are  found 
in  those  waters  met  with  in  the  ^Middle  West  and  Southwest, 
the  advantages  of  preliminary  sedimentation  have  been  well 
recognized  for  some  time.  He  considers  the  use  of  grit  chambers 
for  settling  out  the  coarse  material,  even  before  plain  sedimenta- 
tion is  begun,  entirely  justifiable  in  most  cases.  Mechanical 
clarifiers  have  been  employed  in  a  few  plants  to  permit  of  the 
continuous  removal  of  sediment  which  accumulates  with  great 
rapidity  in  the  handling  of  some  very  muddy  waters. 

'  Wellington  Donaldson:  "The  Trend  of  Purification  Plant  Design 
and  Operation."     Eng.  Contr.,  Aug.  12,  1925. 

66 


RESERVOIRS  AND  COAGULATION  BASINS        67 

Impounding  Reservoirs. — Impounding  reservoirs  built  pri- 
marily for  storing  water  are,  of  course,  to  be  considered  from  a 
somewhat  different  standpoint  than  those  especially  designed 
for  sedimentation  purposes.  The  depositing  of  suspended  sedi- 
ment in  storage  reservoirs  is  an  incident  rather  than  the  object  of 
their  operation.  However,  considerable  purification  is  effected 
in  them,  and  the  influences  which  lead  to  such  results  will  be 
discussed  in  another  section. 

Settling  Reservoirs. — Storage  of  turbid  waters  for  purposes  of 
plain  sedimentation,  in  reservoirs  especially  constructed  for  this 
purpose,  is  undoubtedly  good  practice,  but  is  not  always  provided 
for.  Usually  such  supplies  are  drawn  from  turbid  streams  in 
which  the  water  may  at  times  be  loaded  with  sediment.  Pro- 
\'ided  the  reservoirs  are  relatively  large  in  proportion  to  the  con- 
sumption of  water,  quite  a  number  of  days  of  settlement  may  be 
possible  before  the  water' is  drawn  off.  The  relative  positions  of 
the  inlet  and  outlet  of  these  reservoirs  are  obviously  of  great 
importance  in  even  approximating  theoretical  displacement  of 
the  water  if  they  are  used  continuously.  The  tendency  of  the 
water  currents  to  seek  the  shortest  path  between  the  inlet  and 
outlet,  and  thereby  render  more  or  less  ineffective  certain  parts 
of  the  reservoir,  is  a  commonly  observed  condition.  Baffling 
undoubtedly  has  the  effect  of  breaking  up  "short-circuiting 
currents  of  water,"  but  is  not  always  employed  in  plain  sedi- 
mentation basins. 

Cleaning  out  the  sediment  deposited  in  this  class  of  reservoirs 
is  greatly  facilitated  by  smooth  and  hard  linings  in  the  reservoirs, 
by  sufficient  slope  to  the  sides  and  bottom  and  by  an  adequate 
system  of  gutters  and  drains.  Proper  facilities  for  flushing  out 
the  mud  by  streams  of  water  under  pressure,  or  by  scraping  the 
mud  to  the  gutters  assisted  by  a  flow  of  water  which  is  not  under 
pressure,  are  necessary  in  all  well-designed  basins. 

In  all  settling  basins  the  sand  and  heavier  portions  of  the  silt 
will  be  found  deposited  close  to  the  inlet.  In  some  cases  the 
amount  of  sediment  dropped  near  the  inlet  is  very  great,  espe- 
cially where  the  turbid  stream  supplying  the  water  carries  a  good 
deal  of  sand  and  silt,  and  where  the  pumping  from  the  stream  is 
practically  continuous. 

Cincinnati  Storage  and  Settling  Reservoirs. — Combined  stor- 
age and  settling  reservoirs  are  in  operation  at  Cincinnati,  Ohio, 
(Fig.  5)  in  which  Ohio  River  water  undergoes  plain  sedimentation 


68 


WA  TEE  F I  'RIFICA  TION 


before  being  coagulated  with  chemicals  and  filtered.  These  two 
reservoirs  hold  approximately  392,000,000  gal.  of  water,  and 
were  designed  with  the  idea  of  their  being  operated  on  the  fill-and- 


i 


Fig.  5. — Cincinnati  settling  reservoirs. 


i 


Fig.  6. — Cincinnati  settling  reservoirs,  effluent  float  tubes. 


draw  plan.  This  method  of  operation,  however,  has  never  been 
followed.  The  water  flows  continuously  through  them  in  parallel. 
The  inlets  to  the  reservoirs  are  between  500  and  600  ft.  from  the 
outlets.     The  latter  consist  of  movable  pipes  (Figs.  6  and  7) 


RESERVOIRS  AND  COAGULATION  BASINS 


69 


with  their  mouths  held  about  4  ft.  under  the  surface  of  the  water 
by  means  of  floats.     The  water  is  thus  continuously  skimmed 


5SSSS5JS^^^ 


Fig.  7. — Cincinnati  settling  reservoirs,  profile  showing  float  tubes. 


PLAN 


Slope  3-1      ^.         ._,  nigh  Water  Slope  10-1  Slope  6-1 


SECTION   ON   LINE  A-B 


Fig.  8. — Cincinnati  settling  reservoirs,  plan  and  cross  section. 


from  the  surface.     The  depth  of  water  in  the  reservoirs  varies 
from  35  to  50  ft. 


70  WATER  PURIFICATION 

The  irregular  shape  of  these  reservoirs  (Fig.  8)  is  accounted  for 
by  the  desire  in  construction  to  make  the  excavations  equal  the 
embankments  as  nearly  as  possible.  As  it  was  intended  that 
they  should  be  operated  by  first  filling  with  water,  then  allowing 
the  latter  to  stand  and  deposit  its  sediment  for  a  day  or  so,  and 
finally  drawing  off  the  settled  water,  it  was  of  no  particular  conse- 
quence if  the  inlet  was  close  to  the  outlet.  Since  they  are  not 
vitilized  in  this  manner,  but  are  operated  continuously,  their 
efficiency  is  undoubtedly  somewhat  reduced  because  of  their 
irregular  outline,  and  the  short  distance  between  the  inlet  and 
outlet.  The  apparently  dead  spaces  in  these  reservoirs,  however, 
are  by  no  means  useless,  since  experience  shows  that  diffusion  of 
the  sediment  causes  a  much  more  uniform  deposition  of  the  latter 
than  would  be  supposed.  The  depth  of  mud  in  the  lobes  of  these 
reservoirs  will  probably  run  from  35  to  40  per  cent,  of  the  average 
depth  in  the  portions  of  the  reservoirs  where  more  active  sedimen- 
tation is  in  progress. 

Louisville  Settling  Reservoirs. — The  two  Crescent  Hill  reser- 
voirs at  Louisville,  Ky.,  perform  a  similar  service  to  those  at 
Cincinnati.  They  cover  an  area  of  750,000  sq.  ft.,  and  hold,  when 
filled  to  a  depth  of  20  ft.,  a  little  over  100,000,000  gal.  of  water. 
Originally  the  water  as  it  was  pumped  from  the  Ohio  River  was 
dehvered  into  the  gate  house  at  the  end  of  the  division  wall  sepa- 
rating these  two  basins.  It  passed  over  a  weir  above  the  20-ft. 
level  into  the  first  basin,  and  was  drawn  off  at  a  point  diagonally 
opposite  at  the  bottom  through  a  conduit  5  ft.  in  diameter. 
Through  this  conduit,  the  water  passed  to  the  corresponding  cor- 
ner of  the  second  basin,  where  it  entered  at  the  bottom.  It  was 
drawn  off  at  the  diagonally  opposite  corner  through  the  gate 
house. 

Delivering  the  muddy  water  at  the  surface  of  these  reservoirs 
and  drawing  off  at  the  bottom  has  been  changed,  so  that 
the  water  now  enters  from  the  gate  house  at  the  bottom  of  the 
first  basin,  and  is  withdrawn  over  a  weir  tower  built  to  enclose  the 
inlet  end  of  the  conduit.  After  passing  through  the  latter  to 
the  bottom  of  the  second  basin,  it  is  skimmed  off  at  the  top  of  this 
basin  at  the  gate  house  from  which  it  flows  to  the  coagulation 
basins. 

George  W.  Fuller  observed,  when  these  reservoirs  as  origi- 
nally arranged,  were  drawn  off  to  be  cleaned,  that  with  the 
exception  of  the  coarser  material  piled  up  near  the  inlet,  the 


RESERVOIRS  AND  COAGULATION  BASINS        71 

depth  of  sediment  was  substantially  uniform  over  the  entire 
bottom.  Allen  Hazen  has  noted  the  same  condition  in  other 
reservoirs,  and  accounts  for  it  by  the  mixing  action  by  which 
water  in  all  parts  of  the  reservoir  is  made  substantially  of  the 
same  quality. 

New  Orleans  Grit  Reservoirs. — For  removing  the  heavy  silt 
and  sand  of  a  normally  turbid  water,  relatively  small  grit  reser- 
voirs have  been  used.  In  the  New  Orleans  purification  plant,  two 
such  reservoirs  are  in  service.  Each  reservoir  is  75  by  150  ft.  in 
area,  and  has  outer  walls  20  ft.  high.  These  reservoirs  are  pro- 
vided with  a  center  baffle  which  causes  the  water  to  travel  up  one 
side  of  the  reservoir  and  down  the  opposite  side  to  the  outlet.  At 
the  normal  capacity  of  the  plant,  a  sedimentation  period  of  about 
1  hr.  is  obtained  with  one  basin  in  service.  The  heaviest  sedi- 
ment settles  nearest  the  inlet,  and  grows  much  lighter  as  the  out- 
let is  approached  (Fig.  9). 

Kansas  City,  Mo. — In  this  city,  four  so-called  clarifiers  of  the 
Dorr  type  are  used  as  presedimentation  units.  These  clarifiers 
are  circular  tanks,  each  200  ft.  in  diameter,  containing  a  slowly 
rotating  system  of  rake  arms,  and  which  in  this  case  are  driven 
from  the  periphery  of  the  tank.  A  3-hour  detention  period 
for  a  flow  of  100  million  gallons  of  water  a  day  is  provided.  The 
sludge  may  be  removed  continuously  from  the  bottom  of  the 
tank  at  the  center  by  pumps.  It  is  expected  that  these  tanks 
will  be  capable  of  removing  from  80  to  90  per  cent,  of  the  sus- 
pended matter  in  the  raw  Missouri  River  water. 

COAGULATION  BASINS 

The  bringing  together  of  finely  divided  particles  of  suspended 
sediment  in  natural  waters  by  means  of  the  action  of  certain 
chemical  compounds  is  termed  coagulation.  Basins,  in  which 
this  action  may  take  place  and  the  flocculated  sediment  settle  out, 
are  commonly  provided  in  connection  with  rapid  sand  filter 
plants.  They  may  also  form  a  part  of  slow  sand  filter  plants 
where  the  sediment  in  the  water  settles  out  slowly  and  is  liable  to 
clog  the  sand  in  the  filters. 

Mixing  Channels.^ — The  introduction  and  uniform  distribution 
in  the  water  to  be  treated  of  chemical  solutions  of  the  strengths 
usually  employed  in  water  purification  offer  some  mechanical 
difficulties.  Since  the  efficiency  of  the  coagulating  compound 
depends  upon  the  formation  of  a  large  and  well-defined  floe,  the 


72 


WATER  PURIFICATION 


RESERVOIRS  AND  COAGULATION  BASINS        73 

size  and  character  of  which  is  directly  influenced  by  the  tempera- 
ture of  the  water  and  the  nature  of  the  salts  and  colloids  in  solu- 
tion and  suspension,  a  thorough  mixing  of  the  water  with  the 
solution  is  of  the  utmost  importance. 

By  sending  the  water  through  relatively  narrow  channels  at  a 
velocity  of  1.5  to  2.0  ft.  per  second,  and  allowing  the  travel 
through  these  channels  to  take  about  1  hr.,  a  thorough  mixing  and 
a  complete  coagulation  is  effected,  which  produces  rapid  sedimen- 
tation in  the  coagulation  basins  into  which  the  water  discharges. 
The  eddies  formed  in  the  channels  by  reversing  the  direction  of 
flow  or  by  baffles  or  by  projecting  portions  of  the  concrete  con- 
struction in  the  channels,  assist  materially  in  bringing  about  a 
thorough  mixing. 

In  New  Orleans,  mixing  channels  are  provided  for  applying  a 
5  per  cent,  milk  of  lime  in  the  manner  described  above.  In  this 
plant  dupHcate  reservoirs,  75  by  320  ft.  each  with  outer  walls  19 
ft.  high,  are  divided  into  16  rectangular  double-decked  passages 
with  a  cross-sectional  area  of  over  60  sq.  ft.  each,  and  an  aggre- 
gate length  of  about  5,120  ft.  When  passing  40,000,000  gal.  of 
water  in  24  hrs.,  the  flow  of  the  water  through  the  channels 
requires  about  1  hr. 

Iron  sulphate  for  coagulation  purposes,  and  soda  ash  for 
reducing  the  permanent  hardness  of  the  water  may  be  both  intro- 
duced in  these  channels,  but  provision  is  also  made  for  applying 
a  coagulant  at  various  points  in  the  flow  of  the  water  through  the 
coagulation  basins  proper. 

Obviously,  mixing  channels  may  become  settUng  chambers 
unless  the  velocity  of  flow  is  sufficient  to  retain  the  flocculated 
sediment  in  suspension.  This  does  take  place^  but  corrects  itself 
by  reducing  the  cross-section  of  the  channel  by  deposition  of 
sediment  until  the  velocity  is  sufficient  to  maintain  a  scouring 
action  through  the  passage.  On  the  other  hand,  too  great  a 
velocity  may  break  up  the  floe  and  thereby  diminish  its  ability  to 
settle  rapidly,  as  well  as  failing  to  entrap  the  very  fine  suspended 
sediment  for  which  the  coagulating  chemical  was  added.  Pro- 
vided a  very  thorough  mixing  action  is  obtained  in  the  channels, 
it  is  evident  that  the  coagulation  basins  themselves  may  be  made 
smaller  than  they  otherwise  could  be. 

A  more  recent  design  is  found  in  the  Grand  Rapids,  Mich., 
plant,  where  the  water  after  passing  through  a  grit  chamber  hold- 
ing about  a  26.5-min.  supply  at  a  normal  rating  of  20,000,000  gal. 


74  WATER  PURIFICATION 

per  day,  enters  a  mixing  chamber  44  ft.  wide  bj'  160  ft.  long.  The 
chamber  holds  732,000  gal.,  or  about  a  o3-min.  supply  at  normal 
rating.  The  basin  is  provided  with  wooden  baffles  of  "around  the 
end  type,"  spaced  3  ft.  apart  for  the  full  length  of  the  chamber. 

J.  W.  Armstrong  who  designed  this  plant,  as  well  as  the  plant 
at  New  Orleans,  states  this  type  of  baffle  permits  the  operation 
of  the  plant  with  varying  heads  of  water,  and  offers  reasonably 
good  facilities  for  cleaning  and  inspection. 

Baffled  mixing  chambers  of  either  the  "over  and  under," 
"around  the  end,"  or  combinations  of  these  two  types  are, 
according  to  Donaldson,  "admittedh'  lacking  in  flexibility  to 
meet  variations  in  w^ater  flows.  The}^  have  never  been  very 
popular  in  the  western  country  on  account  of  the  large  amount  of 
suspended  solids  carried  by  the  water,  which  tends  to  deposit  and 
break  down  the  baffles.  There  seems  to  be  a  tendency  to  get 
away  from  this  type  of  mixing  device." 

Mixing  by  Means  of  the  Hydraulic  Jump.^ — Experiments  by 
the  author  at  Cleveland,  Ohio,  resulted  in  the  development  of  a 
flume  in  which  a  hj-draulic  jump  could  be  produced.  The 
application  of  the  chemical  to  be  mixed  with  the  water  to  be 
treated  is  made  just  prior  to  the  entrance  of  the  water  to  the 
flume.  The  water  acquires  a  velocity  of  10  to  12  ft.  per  second 
in  passing  down  the  sloping  floor  of  the  flume  before  it  strikes 
the  pool  of  slowly  moving  water  on  or  near  the  bottom  of  the 
slope.  As  a  result,  there  is  an  abrupt  rise  in  the  surface  of  the 
moving  stream  in  the  region  of  impact,  and  is  accompanied  by 
a  great  deal  of  turbulence  in  the  moving  mass  of  water  (Fig.  9a). 

The  hydraulic  jump,  may  be  scientifically  defined  as  that 
phenomenon  in  which  a  part  of  the  kinetic  energy  of  a  stream  of 
water  flowing  in  an  open  channel,  at  less  than  the  critical  depth, 
is  converted  into  potential  energy  in  the  attempted  turbulent 
passage  from  the  lower  to  the  alternative  flow  level.  These 
alternative  levels  are  the  only  two  possible  levels  for  steady  flow 
under  given  energy  conditions;  and  the  critical  depth  is  that 
depth  at  which  alternative  flow  levels  coincide. 

The  hydraulic  jump,  while  a  common  phenomenon  in  nature, 
has  been,  so  far  as  is  known,  only  utilized  for  the  dissipation  of 
energy  in  flowing  water.  In  the  mixing  apparatus  devised  by 
the  author,  this  energy  has  been  utilized  to  mix  a  small  volume 
of  chemical  solution  with  the  large  volume  of  water  to  be  treated. 

1  United  States  Patent  1362611. 


RESERVOIRS  AND  COAGULATION  BASINS        75 

One  method  by  which  the  effect  may  be  produced  is  by  causing 
the  water  to  flow  through  an  open  flume,  which  widens  uni- 
formly from  the  throat,  and  which  slopes  downward  from  the 
end  of  the  throat  to  the  end  of  the  flume.  The  hydrauhc  jump 
will  occur,  if  the  flume  is  correctly  designed,  on  the  slope  of  the 
expanding  section  of  the  flume. 

Since  a  comparatively  simple  relation  exists  between  the 
critical  depth  and  the  known  cross-section  of  the  throat  of  the 
flume,  it  is  possible  to  develop  a  meter  to  measure  the  flow,  and 
Avhich    could    take  the  place  of  a  Venturi  tube  and  register. 


Fig.  9a. — Hydraulic  jump  in  one  of  the  three  mixing  flumes  at  the  Baldwin 

filter  plant  at  Cleveland,  Ohio. 


Doubtless  a  mechanism  could  also  be  devised  by  which  the  feed- 
ing of  a  chemical  solution  would  automatically  be  proportioned 
to  the  flow  of  the  main  body  of  water  through  the  flume. 

The  advantages  of  this  apparatus  for  mixing  may  be  sum- 
marized as  follows: 

1.  The  mixing  effect  is  produced  with  great  rapidity  and  with 
remarkable  thoroughness. 

2.  The  chemical  reactions  cannot  take  place  until  the  chemical 
or  chemicals  are  in  solution  and  are  diffused  throughout  the 
water  to  be  treated,  which  latter  is  so  many  times  larger  in  volume 


76 


WATER  PURIFICATION 


than  the  reagent  solutions  being  applied,  that  the  value  of  a 
rapid  and  thorough  method  of  mixing  is  obvious. 

3.  The   energy  relations  of  the   hydraulic   jump  have   been 
studied  by  several  investigators,  and  the  principles  involved  are 


fx/ra  Si  rang  8  'threaded  f/an^e 
jl      riveted  io  bofh>m  of  Tank         '     '^ 


II  _  /_  // 


5  ExiraSfrong  Pipe  2-0'' Ig. 
thread  one  end  5. 563  "O.  D. 

'  8  ioS  C.I.  Screw  Reducer 

'- 8  Nipple, 6' Ig.  fhrd.boih  ends, Ex. Sirong Pipe__ 

'  -     -  'A 


Scale:  ^'=3-0 

Fig.  9&. — Section  through  mixing  tank  showing  motor-driven  agitators  for 
mi.xing  chemical  solutions  with  water. 


sufficiently  well  understood  so  that  it  is  possible  to  design  and 
construct  a  flume  in  which  the  "jump"  may  be  produced  and 
controlled  within  comparatively  narrow  limits. 

4.  The  loss  of  head  through  such  a  "jump''  may  be  made  as 
low,  if  not  lower,  than  is  produced  in  any  of  the  conventional, 


RESERVOIRS  AND  COAGULATION  BASINS        77 

baffled  mixing  chambers;  hence  its  practicability  and  economy 
from  the  hydraiiHc  standpoint. 

5.  The  structure  in  which  the  "jump"  may  be  produced  is 
simple,  requiring  comparatively  Uttle  space  and  a  small  amount  of 
material  for  its  construction. 

6.  The  hydraulic  jump  flume  conveniently  lends  itself,  from 
the  structural  standpoint,  to  being  placed  at  the  end  of  raw 
water  influent  pipe  lines  or  conduits,  and  preceding  conduits 
leading  to  the  coagulation  and  sedimentation  basins. 

This  method  of  mixing  has  been  employed  at  quite  a  number 
of  plants  with  very  satisfactory  results.  The  largest  plant  using 
this  method  of  mixing  is  the  Baldwin  Filtration  Plant  of  Cleve- 
land, Ohio.  This  plant  has  a  capacity  of  165  m.g.d.  Three 
flumes  side  by  side  are  provided  at  this  plant  for  mixing  the  alum 
solution  with  the  raw  water. 

Other  Methods  of  Mixing. — Mechanical  stirring  devices 
located  in  a  mixing  tank  have  also  been  utilized  for  diffusing  the 
chemicals  in  the  water  to  be  treated.  These  devices  have 
usually  taken  the  form  of  paddles  slowly  rotated  in  a  tank  by 
motors,  whereby  the  chemical  solution  and  water  to  be  treated 
are  thoroughly  mixed.  This  is  the  type  of  mixer  used  in  the 
Sacramento,  Calif.,  filter  plant,  and  also  in  some  water-softening 
plants  like  those  at  Newark,  Ohio,  and  So.  Pittsburgh,  Pa.  The 
Topeka,  Kan.,  plant  utihzes  a  combination  of  "around  the  end 
baffles"  and  mechanical  agitators;  while  the  Kansas  City,  Mo., 
plant  employs  circular  mixing  and  reaction  tanks  without 
mechanical  agitators. 

In  rotating  paddle  mixing,  H.  N.  Jenks  concludes  that  in  order 
to  obtain  a  uniform  velocity  of  0.6  ft.  per  second  for  the  water 
being  treated,  as  recommended  by  C.  P.  Hoover,^  it  was  necessary 
to  have  the  area  of  the  paddles  25  per  cent,  of  the  area  of  the 
vertical  cross-section  of  the  mixing  tank. 

Settlement  After  Coagulation. — The  deposition  of  coagulated 
sediment  should  take  place  rapidly  if  the  proper  amount  of 
chemicals  have  been  added  to  the  water,  a  thorough  mixing  of  the 
water  with  the  chemicals  has  been  effected,  and  temperature  con- 
ditions are  favorable.  The  same  general  principles  governing  the 
deposition  of  sediment  in  plain  sedimentation  basins,  as  previously 
discussed,  apply  here,  except  that  the  suspended  particles  are 

1  Charles  P.  Hoover:  "Mixing  Devices  and  Reaction  Times."  Jour. 
Am.  Water  Works  Assoc,  May,  1924. 


78  WATER  Pl'lUFICAriON 

larger  and  should  gravitate  toward  the  bottom  more  rapidly. 
A  velocity  of  flow  of  1.0  to  1.5  ft.  per  minute  can  probably  be 
safely  maintained  in  most  cases  without  carrying  too  much  of 
the  finest  material  forward  to  the  outlets  of  the  basins.  Baffles 
will  probably  be  found  advantageous  in  most  basins  of  this  type. 
A  uniform  and  even  distribution  of  flow  across  the  entire  cross- 
section  of  the  basin  is  desirable  between  the  inlets  and  outlets. 
Skimming  from  the  surface  at  the  outlets  is  the  proper  method 
for  the  withdrawal  of  the  settled  water.  Provision  for  secondary 
application  of  the  coagulating  chemicals  as  the  water  flows 
through  the  basins  for  the  purpose  of  correcting  the  original 
treatment,  should  never  be  omitted  in  any  well-designed  plant. 
If  possible,  an  auxiliary  basin  for  getting  the  benefit  of  this 
secondary  treatment  ought  to  be  installed,  otherwise  too  much 
coagulated  material  is  liable  to  be  carried  through  the  outlets  to 
the  filters. 

The  primary  coagulation  basins  are  usually  constructed  in 
duplicate  in  order  to  be  able  to  clean  one  of  them  at  a  time 
without  interrupting  the  operation  of  the  whole  plant.  An  ade- 
quate number  of  sumps  and  drains  ought  to  be  provided,  so 
that  the  deposit  when  ready  to  be  cleaned  out  does  not  have  to 
be  mov'ed  too  far.  The  size  of  the  drains  should  be  ample  to 
prevent  clogging,  and  plenty  of  water  should  be  available  for 
flushing  purposes. 

St.  Louis  Settling  Basins. — The  six  original  settling  basins  of 
this  water-works  plant  (Fig.  10)  were  constructed  as  plain  sedi- 
mentation basins.  For  a  number  of  years  since  the  chemical 
coagulants  sulphate  of  iron  and  caustic  lime  have  been  used  to 
clarify  the  water,  they  have  become  coagulation  basins.  They 
originally  consisted  of  six  rectangular  masonry  tanks  670  by  400 
ft.  in  plan,  and  placed  side  by  side.  The  walls  are  of  masonry  but 
the  bottoms  are  of  concrete.  They  were  intended  to  be  operated 
in  parallel,  but  since  coagulants  have  been  used  the  walls  dividing 
adjoining  basins  have  been  cut  down,  so  that  the  water  now 
passes  through  them  in  series,  entering  through  four  3-ft.  by  3-ft. 
openings  at  the  end  of  one  of  the  basins.  The  water  is  withdrawn 
through  a  masonry  conduit  at  the  end  of  the  sixth  basin  after 
having  passed  through  the  preceding  five  basins.  Sluice  gates  in 
each  basin,  connected  with  both  the  inlet  and  outlet  conduits, 
make  it  possible  to  withdraw  basins  from  service  for  cleaning, 
and  to  provide  for  emergency  conditions. 


RESERVOIRS  AND  COAGULATION  BASINS        79 

Two  additional  basins,  covering  together  an  area  400  ft.  wide 
by  1,660  ft.  long  have  been  constructed  of  concrete.  The  two 
basins  are  separated  by  a  division  wall,  and  are  each  400  by  800 
ft.  in  plan,  and  have  an  average  depth  of  about  21  ft.  Their 
combined   capacity   is   75,000,000    gal.     This   makes   the   total 


Fig.   10. — St.  Louis  settling  basins. 

capacity  of  the  eight  basins  255,000,000  gal.  These  are  the 
largest  coagulation  basins  known  to  the  author,  and  they  handle 
effectively  a  very  turbid  water.  The  effluent  from  these  basins 
is  now  filtered. 


Fiu.  10a. — St.  Louis  filtration  plant. 

Cincinnati  Coagulation  Basins. — A  good  example  of  coagulation 
basins  which  are  effective,  but  which  are  unprovided  with  baffles, 
are  the  three  basins  of  the  Cincinnati  plant  (Figs.  11  and  12).  A 
lined  basin  or  reservoir  400  by  400  ft.  in  plan  is  divided  by  a  con- 
crete wall  into  two  basins,  each  of  which  may  be  operated  inde- 
pendently, but  which  are  usually  operated  together  in  parallel. 
The  depth  of  these  basins  is  about  21  ft.     The  water  enters  the 


80 


WA  TER  P  URIFICA  TION 


two  basins  through  sluice  gates  at  the  bottom  of  a  gate  house  at 
one  end  of  the  division  wall.  It  flows  into  the  bottom  of  the 
basins  through  two  reinforced-concrete  conduits  on  either  side  of 
the  above-mentioned  gate  house.     These  conduits  are  7  ft.  in 


i 


Fig.   11. — -Cincinnati  coagulation  basins. 


h_ 

\ 

X 

Coai^ulation 
Basia.No.1 

J— 

\ 

/ 

/ 

\ 

-B 

Coagulation 
Basin  Xo.2 

) 

Coag-ulation 

( 

Basin  No.3 

E    I 


Filter  House 


PLAN 


Slope  lX-1 


SECTION   ON   LINE  A-B 

Fig.  12. — Cincinnati  coagulation  basins,  plan  and  cross  section. 


diameter  and  have  set  in  the  top  of  each  of  them  twenty-one  20-in. 
tees.  Each  of  these  tees  discharges  through  the  two  openings  in 
line  with  each  other,  and  in  parallel  with  the  line  of  flow  through 
the  body  of  the  conduit.     In  this  manner  the  velocity  of  the 


RESERVOIRS  AND  COAGULATION  BASINS        81 

counter-currents  of  water  tend  to  oppose  each  other,  and  to 
produce  a  uniform  and  quiet  inflow.  The  water  then  passes 
across  to  the  opposite  side  of  the  basins,  and  is  skimmed  off  at 
the  surface  by  openings  into  a  steel  conduit  which  conveys  the 
settled  water  to  either  a  third  basin  for  secondary  treatment, 
or  directly  to  the  filters.  Two  mud  valves  hydraulically  oper- 
ated are  placed  in  the  bottom  of  each  basin.  The  slope  of  the 
bottom  toward  the  sumps  of  these  valves  is  approximately  2  per 
cent.  The  sides  have  a  slope  of  1.75  to  1.  These  basins  have  a 
concrete  lining  covered  with  hard-burned  brick  which  are  grouted 
in  place  so  as  to  form  a  smooth  hard  surface  for  flushing  out  the 
deposited  mud. 


Fig.  12a. — Cincinnati  filtered  water  reservoir. 


The  third  basin  is  400  by  80  ft.  and  about  16  ft.  deep.  It  is 
used  in  series  with  the  two  larger  basins,  and  although  operated 
continuously  to  obtain  the  maximum  period  of  sedimentation,  was 
originally  intended  for  settling  the  water  after  a  secondary  appli- 
cation of  chemicals,  when  the  first  application  had  been  found 
insufficient.  Whenever  it  is  necessary  to  make  a  second  applica- 
tion of  chemicals,  which  is  rarely  the  case,  the  third  basin  is  used 
as  originalh'  intended. 

The  two  large
…[truncated]