Water purification and sewage disposal

Survival, Water, Medical Field Manuals

Military Manuals

Tillmans, Josef, 1876 1935, Taylor, Hugh S. (Hugh Stott), 1890 1974. Tr

Document text


■mn 



MMUMm.- 




Cornell University 
Library 



The original of this book is in 
the Cornell University Library. 

There are no known copyright restrictions in 
the United States on the use of the text. 



http://www.archive.org/details/cu31924080794609 



In compliance with current 

copyright law, Cornell University 

Library produced this 

replacement volume on paper 

that meets the ANSI Standard 

Z39.48-1992 to replace the 

irreparably deteriorated original. 

1997 



T57 



TD 430 T57™*" ""'™'^'"^ '""'"^ 



W3*er purification and sewage disposal, 
3 1924 004 963 546 



BOUGHT WITH THE INCOME 
FROM THE 

SAGE ENDOWMENT FUND 

THE GIFT OF 



flettrs M, Sage 



1891 



a.^l^ss? 



i^L^. 



972^ 



The D. Van Nostrand Company 

intend this book to be sold to the Public 
at the advertised price, and supply it to 
the Trade on terms which will not allow 
of discount. 



WATER PURIFICATION 

AND 

SEWAGE DISPOSAL 



WATER PURIFICATION 

AND 

SEWAGE DISPOSAL 



BY 

DR. J. TILLMANS 

DIRKCTOR OK THE CHEMICAL DKPARTMENT OF THE 
MUNICIPAL INSTITUTE OF HYGIRNE, FRANKFORT-ON-MAINE 



TRANSLATED BV 

HUGH S. TAYLOR, M.Sc. 

UNIVERSITY OF LIVERPOOL 



With 1 1 llluitratiotis in the Text 




NEW YORK 

D. VAN NOSTRAND CO. 

TWENTY-FIVE PARK PLACE 

1913 



PREFACE 

TO THE ENGLISH TRANSLATION 

Since the recognition of their significance in the promotion of the 
pubhc health, methods of water purification and sewage disposal 
have earned a steadily increasing importance. Consequently, 
these branches of industry have enjoyed the most zealous atten- 
tion of the scientist and of the engineer. 

I have endeavoured in the present monograph to give a survey, 
short perhaps, but as complete as possible, of the present 
position in regard to Water Purification and Sewage Disposal. 

Owing to the wide range of this subject particular processes 
could only be treated shortly in the space at disposal. 

One chapter, which is very fully treated in the present volume, 
though slightly elsewhere, is the disposal of industrial sewage. 
Many of the processes of purification are of quite recent date, 
and fresh experience is being obtained daily and reported upon 
in the most diverse publications. I have endeavoured to collect 
such information. 

In the translation some small changes and additions have been 
made in the chapters on sand filtration, the removal of manganese, 
and Travis and Emscher wells. 

The extensive literature supplied in the German original, which 

served as the basis in the composition of the book, has not been 

printed in the English translation, as it treats in the main of 

German literature. 

THE AUTHOR. 

Frankfort-on-Maine, 
October, 191 2. 



TRANSLATOR'S PREFACE 

England, as the author points out, is the classic country for 
sewage disposal. As a consequence, the translation of a book 
on the subject of water purification and sewage disposal from the 
German point of view, might at first sight seem unnecessary. 
It is to be hoped, however, that a study of the present volume 
will prove that view to be incorrect. The careful attention which 
has been paid by the German authorities during the past few 
decades to the provision of suitable water supphes and the 
adequate disposal of sewage, renders the present critical survey 
of modern methods at once interesting and useful to the English 
reader. Especially should this be true of the chapter on the dis- 
posal of industrial sewage. 

I have to express my thanks to the author. Dr. Tillmans, for 
a revision of the present text, and to Messrs. Hubers and Mond. 
I desire also to record my indebtedness to my friends Mr. J. W. 
Yates, M.Sc, and Mr. A. Shacklady, B.Sc, for valuable guidance 
and assistance in the correction of the proofs. 

HUGH S. TAYLOR. 

Liverpool, 

October, 191 2. 



TABLE OF CONTENTS 

WATER PURIFICATION. 
GROUND-WATER, SPRING-WATER, SURFACE-WATER. 

PAGE 

A. Purification of Water for Drinking Purposes . . .2 

Bacteria and- disease germs in water ... 2 

I. Purification of Water for Drinking Purposes o^r a 

Large Scale . . . . . 4 

(i) Filtration Processes 4 

(a) Slow sand filtration ... 5 

Preparation, composition, and working of the filter. . 5 
Uniformity of the sand, filtration pressure, covered and 

uncovered filters . . . ... 6 

Washing and renewal of the sand . . 7 
Filter film, filtration pressure, Frankel and Piefke's 
fundamental investigations, double filtration according 
to Gotze, and preliminary filtration, Puech'-Chabal 
system . . . . . .8 

Cost ; control by estimation of the germ-content . . 9 

{Ji) Percolating sand filters . . . 10 

(c) Mechanical filters . ... 10 

Nature and control of the mechanical filler, old and new 

systems .11 

The Jewell Export Mechanical Filter . 1 1 

Experience with the Jewell Export Mechanical Filter 13 

(rf) Artificial preparation of ground-water from surface-water 14 

Natural bank filtration, infiltration . . 14 

Application of the process in various towns . . .16 

(e) Importance of water filtration in the maintenance of 

public health, and a critical examination of the value of 

sand and mechanical filtration . 16 

(ii) Sterilisation Processes 18 

(<0 The Ozone Process . . 18 

Siemens' Ozone plant at St. Petersburg 20 

Investigations on the action of Ozone . . 22 

{H) The Ferric Chloride Process . 24 



TABLE OF CONTENTS 



PAGE 



(c) The Chloride of Lime Process 25 

Nature and cost of the process 25 

Investigation and review . . 26 

{(i) Ultra-violet light . . . 27 

The investigations of Courmont and Nogier . . 27 

The investigations of the Konigl. Prufungsanstalt fiir 

Wasserversorgung (Imperial Institute of Tests for 

Water Supply) (Grimm and Weldert) . . 28 

Other authors on the process . 28 

Comparison of the costs of filtration and sterilisation . 29 

{e) Disinfection of water-mains and wells . . . 30 



(iii) Purification of Water in other Directions than of 
Health ... . . . 



30 



{a) Removal of iron .... 

Disadvantages of water containing iron ; nature of the 
processes for removing iron . ... 31 

The most important methods of iron-removal . 32 

Open and closed iron-removal plants . 33 

Method of working ; cost . . -34 

{b) Removal of manganese . . 35 

Disadvantage of water containing manganese ; the 
Breslau calamity ; different methods of removing 
manganese . . . . • • 35 

{c) Removal of free carbon dio.xide . . 36 

Corrosive action of water containing carbon dioxide ; 
removal by limestone ; description of the Frankfort 
plant . . 37 

Removal with caustic soda or sodium carbonate 



JO 



Removal by aeration ; vacuum process . . 40 

II. Purification of Drinking-Water on a Small Scale . 40 

Basic disadvantages of the purification of drinking-water 

on a small scale, as opposed to large-scale operations 40 

(i) Small or Household Filters 41 

More or less uncertain action of all small filters 41 

{a) Charcoal filters 41 

.{b) Stone filters 42 

(c) Asbestos filters 42 

(rf) Clay filters . 42 

{e) Earthenware filters 42 

(/) Kieselguhr filters 43 

(ii) Boiling the Water 45 

Apparatus for boiling larjje qu.uuiiics of water 45 

Advantages and disadvantages of boiling the water as 

compared with household filters 46 



TABLE OF CONTENTS xi 



PAGE 



(iii) Small Ozone Plants and Ultra- Violet Light Apparatus 47 

Stationary and transportable ozone plants, Siemens and 

Halske ... .... 47 

Experiments with household apparatus 48 

(iv) Eemoval of Iron from Single Wells 49 

Different methods of removing iron from single wells 49 

Bastard pump of Deseniss and Jacobi . 50 

B. Purification of Water for Technical Purposes . . 51 

Requirements of a water used for industrial purposes . 5 1 
Softening of boiler-feed water, temporary hardness, perma- 
nent hardness, degree of hardness ..... 52 
Softening by the lime-soda process : method, estimation of 

the amount to be added ... . . 53 

Description of a lime-soda softening plant (Voran system) 55 

Advantages and disadvantages of the lime-soda process . 58 
The Reisert Baryta Process . ... -59 

Advantages and disadvantages of the Baryta Process . 59 

The Permutit Process : method, advantages, disadvantages 60 
Importance of an intelligent supervision of softening plants, 

so-called boiler scale preventatives .... 61 

SEWAGE DISPOSAL. 



The significance of river pollution and the importance of 

sewage disposal ... . . 

Self-purification of rivers . . '. . 

Domestic and industrial sewage 

A. Purification of Domestic Sewage 

The different processes . 



63 
65 
67 

67 
68 

68 



I. Mechanical Purification of Sewage 

(i) Screens, Ba&es, and Sieves 68 

Coarse and fine screens, stationary and movable rakes 69 

Uhlfelder's revolving screen ; the Rien sieve . 69 

(ii) Grease Separators .72 

Kremer apparatus . 72 

Kremer septic wells ; other grease separators 73 

(iii) Grit Chambers 73 

(iv) Sedimentation Tanks and Wells 74 

Form of sedimentation tanks ; velocity of sedimentation 74 

Sludge disposal : addition of chemicals . 76 

Clarification effect, sedimentation wells, towers . 77 



xii TABLE OF CONTENTS 

PAGE 

(v) Septic Tanks 78 

Method ; sludge and removal of sludge ; advantages and 
disadvantages of septic tanks . . • 7^ 

(vi) Travis and Emscher Wells • • • 79 

Method ; Travis tank . . 79 

Emscher wells . • 82 

Critical review . • . • -83 

II. Degener's Coal-Pulp Process ... . . 84 

III. Biological Purification of Sewage ... 84 

(i) The Artificial Biological Process 85 

Method . . . . . ■ 85 

Sewage disposal in England, Preliminary purification, 
Material of the biological beds, Contact beds. Percolat- 
ing filters . -85 
Cost . . .... .89 

Theories as to the nature of biological purification of 

sewage (Brettschneider, Travis, Dunbar) . . 89 

(ii) Land Treatment ; Broad Irrigation 9° 

(a) Sewage Farming . . - . 9° 
Nature, good and bad soils for irrigation purposes, 

irrigation of sloping surfaces ... 90 

Irrigation of beds . . . . . 9° 
Eduardsfeld Process . . -91 

Action of the irrigation . . ... 92 

Importance of preliminary treatment . 92 

Various details on suitable methods of treatment . . 92 

Impregnation permissible, rent and cost ... 93 

IJj) Intermittent Sand Filtration . . . .94 

Historical distinction from broad irrigation . . 94 

Experience with the process in Massachusetts, U.S.A. . 94 

(iii) Purification of Sewage with Fish-ponds 



IV. Disposal and Profit from the Resltlting Residues 

(i) Grit-Chamber Residues, Residues from Screens, Sedi 
mentation Sludge 



96 

97 



97 

(ii) Drying of the Sludge 99 

Disposal on the land gg 

Contents of Emscher-well sludge . 100 

Drying in ditches 100 

Filter-presses 100 



TABLE OF CONTENTS xiii 



PAGE 



Schafer-ter-Mer Centrifuge . . loi 

Disposal in the sea 102 

Addition of nitrates 104 

Cost of the various processes . 104 

(iii) Profit from Sludge .104 

As manure . . 104 

Preparation of artificial manure . 105 

Recovery of grease . . 105 

Generation of electrical energy by burning . . . 107 

Frankfort sludge disposal . 107 

Gasification of sludge . . . . 108 

B. The Purification of Industrial Sewage ... .109 

I. General ... . 109 

Reception into the town's sewers . . 109 
Leakages, condenser waters, and effluents from particular 
manufacturing processes. Association of sewage from 

different processes . 1 10 

Reservoirs, filters . . .111 

Use of industrial sewage for laying dust on the roads . .111 



II. The Purification of Industrial Sewage in Detail 



112 



Classification of industrial sewage . . .112 

1. Cloth factories . 113 

2. Cardboard works . 113 

3. Straw-board works . . ..114 

4. Mines (coal-washing) . .114 

5. Sewage from works granulating slag . 1 14 

6. Paper and cellulose mills . . "5 

7. Breweries . . .117 

8. Tanneries . 118 

9. Dairies and margarine works ... .119 
10. Slaughter-houses, knackers' yards, glue works . 120 
I r. Sugar works 

12. Starch works 

13. Distilleries, yeast works 

14. Sour-krout works 

15. Dye works and print woiks 

16. Chemical works 

17. Bleach works 

18. Gas works 
ig. Ammonia works 

20. Potash works 

2 1 . Metal works and works manufacturing mordants 

22. Manufactories of photographic material and paper . 



120 
122 
123 
123 

123 
125 
126 
126 
127 
127 
128 

I2§ 



xiv TABLE OF CONTENTS 

PAGE 

23. Sewage containing cyanides .... 129 

24. Wool-scouring, wool-combing, and wool-finishing 130 

25. Petroleum refineries . 130 

26. Sewage containing soaps 131 

27. Oil works ... 132 

C. The Disinfectioti of Sewage 132 



List of Authors 135 

List of Localities . 137 

Indb;x of Subjects . 138 



TABLE OF EQUIVALENTS IN WEIGHTS 
AND MEASURES. 

I metre=io decimetres = 100 centimetres = 1 000 millimetres = 1*093 yards = 

3"28 feet = 39"37 ins. 
I cubic metre=iooo litres=i"3o8 cubic yards = 35"3 cubic feet = 22o gallons. 
I kilogram = 2 2046 lbs. looo kilograms = approximately i ton. i gram = 

1 5 43 grains. 



LIST OF ILLUSTRATIONS 



PAGE 



Fig. I. Diagrammatic Representation of a Sand Filter . 6 

„ 2. The Jewell Export Mechanical Filter . .12 

„ 3. Diagrammatic View of the Ozone Waterworks at St. Petersburg 19 

„ 4. Ozone Waterworks, St. Petersburg. Sterilisation Towers and 

Emulsifiers . . . 20 

,, 5. Ozone Waterworks, St. Petersburg. Batteries for the Ozone 

apparatus 21 

,, 6. Open Iron-removal Plant, " Voran" System 34. 

,, 7. Closed Iron-removal Plant, " Voran " System . . 34 

„ 8. Acid Neutralisation Plant in the Saxony Deep Reservoir 39 

„ 9. Berkefeld Filter . . 44 

„ 10. Apparatus for Boiling Water of the Firm of Aug. Schmidt and 

Sons, Hamburg-Uhlenhorst . . 46 

„ II. Bastard Pump of Deseniss and Jacobi . . . 50 

„ 12. Lime-Soda Softening Plant, " Voran " System . 56 

„ 13. Uhlfelder's Revolving Screen . 69 

„ 14. Rien Sieve . 70 

„ 1 5. Kremer Apparatus 7 1 

„ 16. Grease Separators of the " Sladtereinigung Firm," Berlin- 
Wiesbaden . . 72 

,. 17. Frankfort Sedimentation Tank (Cross Section) 75 

., 18. Clearing Well Installation at Norwich on the Travis Hydro- 
lytic System, with CoUoider hung in. From " Wasser und 
Abwasser," 1909-10, vol. II, p. 71 . .80 

„ 19. Emscher Wells . 81 

„ 20. Sludge Centrifuge : Schafer-ter-Mer System 102 

ij 21. 1, „- „ « . 103 



WATER PURIFICATION 



THE water present on the earth is engaged in a continuous 
circulatory process. 

It evaporates from the oceans, seas, rivers, etc., and passes into 
the atmosphere as water-vapour. In the higher strata it condenses 
to form clouds, and then returns to the earth's surface as rain, 
snow, hail, and dew. 

A part of this water evaporates immediately, another part 
flows away to the nearest surface reservoirs. A third part, on the 
contrary, penetrates into the earth's crust, and sinks deeper and 
deeper until it reaches an impermeable stratum, whereupon it 
collects and fills the pores and hollows of the overlying ground. 
This water is called ground-water. Volger considers ground- 
water as a product of the condensation of ground-vapours, and 
Mezger assumes that it is the vapours rising from the depths 
which condense. According to Novak it mainly results from 
the water of the oceans penetrating into the interior of the earth. 
Although ground-water may receive considerable additions from 
one or other of these sources, still it must originate in greater 
part in the first-mentioned manner, through infiltration. From 
the impermeable layer it passes along very slowly to the nearest 
lower-lying waters. 

Spring -water is a mountain ground -water which, owing 
to this movement, makes its appearance in a cleft of the 
motintain. 

In contradistinction to ground-water, all water which remains 
in contact with the outer air is designated as surface-water. 
Surface-water is therefore the water of seas, rivers, ponds, 
cisterns, dams, etc. 

Water is obviously a necessity to man. In the first place, it 



2 WATER PURIFICATION 

is used for drinking purposes. Moreo\'er, it is found to be an 
indispensable auxiliary' in almost every possible pursuit. 

Frequently it must be submitted to certain treatment before 
use. Tlie application proposed ser^'es as a method of differentia- 
tion. One can distinguisli, therefore, two processes : purification 
of water for drinking purposes, and for technical purposes. 

A. Purification of Water for Drinking Purposes. 

^^'ate^ used for human consumption must be free from sus- 
pended matter. It must be colourless, odourless, agreeable tp the 
taste, and must not ha\e too high a temperature. It should be 
of such a nature that it may be partaken of with pleasure. 
Further, and of chief importance, it must contain no disease 
germs. 

Surface-water and ground-water are both used for ^^•ater 
supply. Ground-water, taken from certain depths, is free from 
bacteria ; for, during its passage through the soil, it is freed 
from germs, which remain attached to the particles of sand in 
the ground. Surface-water, because it is easily polluted, is 
alwaj's more or less rich in bacteria. 

The germs of contagious diseases, typhoid, diarrhoea, cholera, 
and certain groups of the so-called ptomaines, are the most 
important disease germs disseminated through water and claiming 
consideration. Pus germs also are found. It is only in the rarest 
cases possible to demonstrate the existence of disease germs in 
infected water, since they only occasionally succeed in entering, 
and also between the date of their reception by the person and the 
development of the sickness a period of time always elapses (e.g. 
with typhoid, one to three weeks). If, therefore, developing 
diseases give rise to an in\'estigation, it is usuallj^ too late to 
trace the origin to water. Quite apart from this, also, the proof 
of the existence of disease germs amidst the many other harmless 
bacteria is a matter of considerable difficultj'. 

Proof of the absence of disease germs is not sufficient, and is 
no guarantee of the harmlessness of a water to health. Sanitary, 
opinion is based, however, upon the assumption that the strata 
of earth at slight depths are already free from anj' such germs, 
and that therefore a discovery of bacteria in water proves that 
it has come into contact A\-ith the outside world, that it has 



PURIFICATION FOR DRINKING PURPOSES 3 

received surface influxes. In the dissemination of typhoid and 
other contagious diseases, the possibihty that even secretions of 
such men or animals as harbour disease germs in themselves 
have been admitted to the water cannot be excluded. The more 
influxes of surface-water, the greater is the probability of con- 
tamination in a ground-water. From these considerations it 
follows that a river-water is much more dangerous than the water 
from a suitably situated reservoir. 

If a natural water containing bacteria be used for drinking 
purposes it must first be rendered innocuous to health. This 
can be done by methods of filtration or of bacteria destruction. 
But the harmfulness of a surface-water to health is not its only 
fault. Many surface-waters, such as river-waters, have also the 
disadvantage of an unappetising odour and taste, or a dis- 
agreeable temperature. Then, even the sanitary improvement 
of water by the destruction of bacteria or by filtration will only 
remove the gravest danger — the danger to health — but will 
not make of the inferior water a good one. It must furthermore 
be borne in mind, that when such measures of purification are 
demanded, failure must be expected either of the plant itself 
or failure due to excessive demands on the available material, of 
which the best-known example is the notorious water-main in 
the Ruhr district.^ As regards filtration, it must be remarked 
that removal of all the bacteria is not effected. Filtration 
only brings about a reduction of the number of germs. Since, 
however, for the origination of a contagious disease a definite 
amount of disease germs is always necessary, reduction of the 
germs, if it is at all considerable, signifies a considerable sani- 
tary improvement in the water. 

Ground-water from sufficient depths is therefore superior, 
for all the above-mentioned reasons, to surface-water, even if the 
latter be very carefully treated. But the former also may be 
doubtful for drinking purposes. There are two ways in which 
disease germs may reach a ground-water and make it unsafe as a 
water supply. The first manner in which it may be polluted, 
and the one coming into consideration in the greatest number of 
cases, is contamination from above. Through fissures in the 

' This water-main drew water from the River Ruhr. During- the year igi i the 
level of the river sank so considerably that the end of the pipe lay above the sur- 
face and no water could be drawn off. — Translator. 



4 WATER PURIFICATION 

ground, through leaking wells, and through a permeated ground- 
water, external streams may be incorporated without sufficient 
previous filtration. The second method of pollution with disease 
germs is through the so-called subterranean streams from strata 
in the earth rendered contagious by human refuse. This method 
of pollution has been frequently affirmed and disputed. There 
come into consideration here filled-up waste sewers, drains, 
depots for fjeces, etc., which, owing to underground hollows, for 
example, rat-nms and the like, are connected with wells, and 
through which the ground-water flows into the wells. Even 
without direct hoUows or runs existing, a certain amount of danger 
is assumed, owing to insufficiency of filtration. If the ground- 
water, in its passage from the suspected area to the place where 
it is drawn, flows through at least lo metres of ground free from 
objectionable features, that is regarded, in general, as adequate 
filtration. 

Occasionally drinking-water must be purified for quite other 
reasons than the hygienic reasons pre%dously discussed. In such 
cases it is a question of remo\Tng substances, including salts, 
which give the water an unappetising appearance, or make it 
unsuitable as a drinking-water or for domestic use. There may 
be present substances which impart to the water a certain cor- 
rosive action on the walls of the pipes through which it is led 
or the vessels in which it is stored. 

^Methods of water purification may be divided into two classes, 
those on a large scale, which serve for central water supply, and 
those for the purification of water on a small scale, which are 
appUcable on a journey, in housekeeping, and in industry. 

I. PURIFICATION OF \^'ATER FOR DRINKING 
PURPOSES ON A LARGE SCALE. 

The purification of drinking-water on a large scale, in order 
to obtain a healthy and unobjectionable water, is effected either 
by filtration or sterihsation. In the first case the bacteria are 
mechanically removed, in the second case they are killed 

(i) FQtiatioii Processes. 

The greater part of the suspended matter can be removed in 
settling reservoirs, or by leading the water through clarifying 



FILTRATION PROCESSES 5 

basins in which the very small velocity of the water enables the 
greater part of the suspensions to settle to the bottom. 

Sedimentation basins for the purification of drinking-water 
are constructed and managed according to the same principles 
as those for the purification of eflfluent waters. On this account 
the question may be relegated from here to the chapter on the 
conduct of sewage purification. 

The most finely divided suspensions, the bacteria, are not 
removed from the water in this way ; for that purpose filtration 
is required. 

(a) Slow Sand Filtration. 

For the sand filtration of drinking-water we have to thank the 
Englishman James Simpson, who, in the year 1829, constructed 
the first filter of this type. In 1839 the London Water- 
works introduced the first of such filters for the purification of 
drinking-water. In the year 1853 Simpson sand filters were also 
constructed in Berlin, and shortly afterwards in many other 
towns. The inventor only had in view the removal of suspended 
matter and the clarification of the water by means of sand 
filtration. That it would remove the bacteria he could not have 
anticipated, as at that time such micro-organisms were still un- 
known. We know to-day, however, that the main importance of 
sand filtration lies in the elimination of bacteria. 

In case the water does not contain too large a quantity of 
suspended matter, a preliminary clarification is unnecessary, 
and the water flows immediately on to the filters. 

The filters (sefe=-Fig7=€-) consist of large, generally rectangular 
surfaces, surrounded by a wall. They are filled with gravel and 
sand. At the bottom there is a layer of stones 60 to 150 milli- 
metres in diameter. Above this rests a layer of gravel which 
serves to support the superimposed sand, being coarser below 
than above. The layers of gravel are generally set down with the 
first about the size of nuts (30 to 60 mm.), then one about the 
size of beans (20 to 30 mm.), one the size of peas (10 to 20 mm.), 
and a layer about the size of millet (3 to 5 mm.). Over this there 
rests the layer of sand upon which the raw water is placed to 
a certain depth. The sand and gravel layers can be set down 
together in varying amounts. The average height of water, sand, 
and gravel is about-O'60-4tLetre (2 feet) each. The filtered water 



6 



WATER PURIFICATION 



flows away underneath, and passes thence into the reservoirs 
for purified water. To supply a town with filtered water a large 
number of such sand-filters are needed. 

The plans of artificial sand filters must be so arranged that each 
individual fiJter can be separately filled, emptied and cleansed. 




Fig. I. Diagrammatic Representation of a Sand Filter. 

and that the purified water from each filter can be drawn upon 
independently. Only in this way is complete control of each sepa- 
rate filter possible. The water filters the more quickly through the 
sand the larger the size of the grains. Coarser sand does not yield, 
however, such pure, germ-free water as does the finer sand. With 
this finer sand the surface of the filter layer clogs up more 
quickly than with the coarser, which latter is easier to clean. 

Further, similarity of form in the sand is important. The 
more dissimilar the particles of sand are, the more erratically the 
filter works. In the cleaning of the filter the fine sand is in 
part washed away ; sand must therefore be added from time 
to time to the filter, according to the length of time it 
is in use, if it is .not to become continually coarser. Whilst 
the form of the filter must be adapted to the disposition of the 
available space, its size varies considerably. The size of a filter, 
according to Konig, varies in a series of large towns from 607 to 
7600 square metres, and is on the average about 2000 to 3000 
square metres. The water should be maintained in the filters 
as far as possible at the same height. The entry of the water 
to the filter takes place continuously from above, and the entrance 
of course lies opposite to the discharge-pipe. Frequently the 
inflow of water is automatically regulated by means of valves, 
to bring about as uniform an inflow as possible. ' A uniform 
discharge of water, also, is of no less importance. In consequence 
of the gradually increasing clogging of the filter the velocity of 
discharge would always become less if the head of water were 



FILTRATION PROCESSES 7 

not increased. This head of water is the difference of water-level 
in the filter and in the pure-water reservoir. To increase the 
head it is best to diminish the water pressure in the pure-water 
reservoir, since, as already mentioned above, it is not advan- 
tageous to alter the amount of water in the filter. 

In the Berlin Waterworks the head amounts to 60 or 65 
millimetres, in Altona to 1422 millimetres, in Kiel to 1000 milli- 
metres. 

In order to permit the air enclosed in the filter to pass out, 
so that it is not constrained to escape to the top, thus causing a 
breaking-up of the filter-bed, tubes for the removal of the air 
are let into the side-walls. 

Filters are either covered or uncovered, both systems having 
their advantages and disadvantages. 

Open filters have the disadvantage that during frosty weather 
cleaning is made very difficult, owing to the freezing of the moist 
sand. This objection disappears with covered filters. On the 
other hand, covered filters are far more costly. 

A further disadvantage of the covered filter is that with it 
the sediment layer is formed more slowly and more imperfectly, 
and the covered filters, consequently, do not yield a sufficiently 
germ-free water so quickly. This is readily explained, as the 
sediment layer is composed in part of organisms containing 
chlorophyll, and consequently needing light, which organisms 
cannot increase, or can only do so more slowly than is the case 
with open filters exposed to the full light. 

In practice the filters are arranged thus. They are filled with 
water from below to just above the surface of the sand ; then 
the impure water is allowed to flow in from above, and to remain 
at rest for a period of time during which the formation of the 
sediment layer is accelerated. The filtrate is allowed to flow away 
until the germ content has reached a certain limit, generally 100 
or less per cubic centimetre (1600 per cubic inch). Then the pipe 
to the pure-water reservoir is connected. 

After a certain time, when the sediment layer has become too 
strong, the filter works itself dead, and no more water passes 
through. It must then be purified. For this purpose a layer 
of sludge, generally about one inch thick, is first of all 
scraped off. The sand lying underneath likewise contains a lot 
of dirt, which is, however, of the greatest importance for the 



8 WATER PURIFICATION 

filter layer about to be constructed. After removal of the layer 
of sludge, therefore, the sand is loosened to a depth of about 
8 inches, and the filter is then allowed to remain unused 
for a day, to permit the access of fresh air. The so-called 
journey (running time) of a filter is likewise very varied. From 
eleven different waterworks, according to Konig, it amounts on an 
average to 25-5 days, during which the amount of water filtered 
per square metre averaged 69-3 cubic metres. 

From time to time the sand must also be replaced by fresh or 
washed sand. It is removed down to the layer of gravel, new 
sand filled in, and this is covered with a layer of the lower portion 
of used sand, which has a sticky nature and accelerates the 
formation of the sediment layer. Washed sand is only to be 
recommended in place of fresh sand in cases where the fresh 
material is dearer than the washed, since by washing more or less 
of the fine useful portion of the sand is removed. For this 
reason a sand which has been washed many times must be 
replaced by fresh sand. The washing of the sand takes place 
automatically in drums, or boxes, in which the sand comes many 
times into contact with fresh water. There are numerous different 
systems of sand washing. 

The more slowly filtration takes place, the purer, as a general 
rule, is the filtrate. The velocity of filtration, which varies largely 
in different waterworks, amounts on an average to about 100 
millimetres (4 inches) per hour. 

Very soon after the introduction of sand filtration it was 
recognised that the sediment layer would be difficult to control. 
This layer consists for the most part of organic suspended matter, 
displaying either living organisms or dead substances, while 
there is also present, in smaller degree, inorganic matter like clay, 
oxide of iron, etc. The most varied organisms are to be found 
in the composition of a sediment layer. 

C. Piefke demands a maximum filtration velocity of 100 milli- 
metres per hour, while other investigators could establish no 
variation in the bacteriological and chemical composition of the 
water with considerably greater velocities. 

C. Piefke found, further, that with an increase in the pressure 
of filtration the bacteria content of the pure water rises, more 
so, of course, the more bacteria the raw water contains. He 
also proved by investigation that a filter freed from the 



FILTRATION PROCESSES 9 

sediment layer yields water of smaU bacteria content more 
quickly than fresh sand. This is explained by assuming that the 
suspended matter penetrates into the sand a little, and this sand 
layer, containing suspended matter, takes part in the work of 
filtration. Of great importance is the answer to the question, 
whether the bacteria in the pure water have passed through the 
filter during the filtration of the raw water, and therefore are 
derived from the raw water, or whether they are washed away 
from the sediment layer or from the sand. By their investiga- 
tions on this question, Frankel and Piefke came to the con- 
clusion that the quantity of micro-organisms passing over into 
the filtrate is proportional to the bacteria content of the raw 
water. 

For towns which have to deal with very poor raw water, like 
Hamburg, Altona, Konigsberg, Warsaw, and others, double 
filtration, proposed by Gotze, is to be recommended. The 
most diverse hygienists express themselves very approvingly 
concerning this method. By double filtration, the raw water 
already passed once through a sand filter is sent again through a 
filter well clogged with sludge. Gotze showed that with double 
filtration a raw water with a germ content of 28,000 was purified 
to one of 780 bacteria per cubic centimetre in the preliminary 
filtration, and to 31 per cubic centimetre in the final filtrate. 
The preliminary filtration of Puech-Chabal is largely employed. 
In this system the water is purified by a series of coarse prelimin- 
ary filters. The fine filter then employed has only to further 
diminish the already considerably reduced quantity of bacteria, 
so that the water can be regarded as free, bacteriologically, from 
objection. 

The cost of sand filtration, exclusive of interest and repayment 
of loans, amounts on an average to o-7d. to 2-5d. per 1000 
gallons (Konig). 

The filtered water is submitted to a continuous bacteriological 
control. The constant estimation of the number of bacteria 
is an excellent means of settling whether the filter is working 
well. A crack arising in the sand, or any other abnormality, 
reveals itself immediately in a rise of the number of bacteria. 
The official rule runs that pure water should not contain above 
100 germs per cubic centimetre. 



10 WATER PURIFICATION 

(b) Percolating or Dry Sand Filters. 

In France, recently, the so-called percolating sand filters have 
come into use for water purification. They are based on the 
theory that oxygen plays a part in the removal of bacteria, 
especially of the pathogenic kind. The ordinary filter can contain 
no oxygen, since it is always covered by water. On this account 
the percolating sand filter is employed similarly to the per- 
colating filters in bacterial sewage purification, except that it is 
composed of fine material (sand). 

Miquel and Mouchet, in laboratory investigations, were unable, 
after using this method, to prove the presence of typhoid bacilli 
added to the raw water. 

This method was tested by Baudet, in Chateaudun (France), 
working on a large scale. The results seem very favourable. 
The number of bacteria fell from between 293 and 1498 in raw 
water, to 6 bacteria per cubic centimetre in pure water. Espe- 
cially remarkable is the consideration that in the pure water 
bacterium coli was never found. 

This filter likewise requires several months to build up, and 
the velocity of filtration cannot be raised indefinitely ; still, the 
percolating sand filters seem to be considerably more pro- 
ductive than the slow sand filter. 

Baudet maintains that with a clear but bacteriologically impure 
water, results are obtained with percolating sand filters which are 
better and less troublesome than those obtained by any other 
method. 

The introduction of the filter for barracks has been recom- 
mended by the French Army administration. In Germany the 
filters have not yet, to my knowledge, come into use. Further, 
there has been, up to the present, no German investigation of 
the method. 

(c) Mechanical Filters. 

The considerable cost of sand filters in relation to their pro- 
ductivity, and the great space necessary for a sand-filter installa- 
tion, shows clearly the advisability of replacing the sand filter 
in technical work by other apparatus which takes up less room 
and permits a greater velocity of filtration. From these considera- 
tions the mechanical filter originated. Since the natural, workable 
layer is first formed after a long interval of filtration, chemicals 



FILTRATION PROCESSES II 

are generally added to the water, especially with the newer filters 
of this type, in order to produce an artificial plankton and an 
artificial filter layer. With the mechanical filters of newer 
pattern, aluminium sulphate is almost always used for this 
purpose, as it reacts with the alkaline earths present in the water 
according to the following equation : 

Al2(SO,)3+3 Ca(HC03)2 = Al2(OH),+3 CaSO^+e CO,. 

Most of the flocculent, gelatinous aluminium hydroxide sinks 
to the bottom, and the suspended matter travels along with it. 
The flakes still remaining in the water form a sediment layer on 
the filter. The velocity of filtration exceeds that of the sand 
filter, generally being 60 or 70 times more rapid. 

There is a large number of different systems of the mechanical- 
filter type. 

Older patterns are, for example, the Anderson revolving 
purifier, the Warren Filter, the Krohnke Filter. 

Filters of newer construction are, amongst others, the Jewell 
Export Mechanical Filter, the Halvor Breda Filter, the Bell 
Filter, Reeves Filter, Candy Filter, Puech Filter, Sucro Filter. 

These filters are not only employed for removing bacteria from 
drinking-water, they frequently find application in technical work. 

As a type of this filter the Jewell Export Filter, which has been 
studied closely from different points of view, may be described 
more carefully. 

The Jewell Export Mechanical Filter. — The Jewell Filter, repre- 
sented in Figure 2, consists of the steel cylinder containing the 
filter-bed, which is encased in a second cylinder of somewhat 
larger diameter, so that between the two there is an annular space 
which is closed underneath. In this annular space, through the 
valve situated on the left-hand side, the raw water, which has 
been previously treated in sedimentation tanks, and also with 
aluminium sulphate, passes into the filter, in order to flow over 
the edge of the inner cylinder on to the filter-bed. After it has 
streamed through the filter-bed, composed of sand, it passes at 
the bottom through sieve heads in a system of outlet tubes which, 
in their turn, fit into a diametrically placed collector, and to 
which they are all rectangularly placed. From the collector 
the water passes through the regulator (Weston Controller), 
shown in the front of the diagram on the right side, into the 



152 WATER rUUIFIC'ATION 

jjure-watcr basin to be found underneath. This regukitor serves 
to keej) filtration constant. Tlie imj)ortance of a constant 
\'elocity of filtration has already been brought into prominence 
in the case of sand filters. With mechanical filters this importance 
is increased owing to the employment of chemical coagulants 
associated with it, and which can only be added in these cases 
in precisely estimated amounts. V>y means of a float working on 
a throttle-valve (shown on the left in the diagram) the inflow is 




Fig. 2. The Jewell Export Mechanical Filter. 



also regulated, and the water in the filter kept at a constant 
height. 

To clean the filter the water takes the reverse direction. It 
is allowed to stream in under the pressure of a pump or a high 
reservoir, through the cleaning-valve situated at the extreme 
right in the diagram. It is then allowed to flow through the 
collector, exit tubes, and filter-bed in the reverse direction from 
bottom to top, to pass o\-er the rim of the inner cylinder into the 
annular space, out of which it streams into the effluent tube, 
through a vah'e not \'isible in the diagram. At the same time the 
stirring arrangement shown in the diagram is set in motion, which 
thereby gets the whole of the filter-bed into a floating condition, 
so that each separate particle of sand is washed by the water 



FILTRATION PROCESSES 13 

and is consequently thoroughly purified. In the cleanmg process 
it is necessary, of course, that the cleansing water be distributed 
imiformly over the whole surface of the filter-bed, so that no 
stagnant comers or angles result. This important need is met 
in the Jewell Filter by keeping the throats of the strainers in the 
outlet tubes very narrow, so that the pressure and velocity of the 
cleansing water at these points are very large. 

After the first cleaning there comes a further cleansing process, 
in which the first water filtered is allowed to flow through the 
third of the three valves shown on the right of the diagram, in 
order to remove the muddy water stiU present in the filter. The 
cleaning and the subsequent operation take up about ten minutes, 
and are effected as a rule once daily. The starting of the filter 
is always performed mechanically without any manual labour 
on the filter-bed. 

Jewell Filters have come into use largely for the water supply 
of municipalities. I mention the towns of Alexandria, Trieste, 
Helsingfors (Finland), Annecy (France). Critical examinations 
on the basis of the tests previously put forward generally prove 
favourable. 

Bitter and Gottschlich ha\e obtained very favourable results 
with the Jewell Export Filter in Alexandria. 

Hilgermann has likewise made investigations with the JeweU 
Export Filter, and come to fewer favourable results. He holds, 
according to his experiments, that the JeweU Export Filter, like 
every other mechanical filter, is, in bacteriological respects, 
absolutely inferior to the sand filter. 

K. Schreiber has searchingly investigated the Jewell Filter, 
in a series of experiments with an experimental plant at the 
Berlin Waterworks, and comes to the conclusion that it is quite 
as good as the sand filter in bacteriological respects, in the 
removal of the turbidity and colour of the raw water, as well 
as in the method of washing, which takes place quite mechani- 
cally, without danger of contamination from the hands and 
clothes of the workmen. According to Schreiber, the method 
may, however, be far superior to the sand-filtration method. 

The amount of aluminium sulphate added is of great import- 
ance. Schreiber comes to the conclusion that with an addition 
of 33'6 g- of aluminium sulphate per cubic metre (5 oz. per 1000 
gallons), with a time of sedimentation lasting i hour 28 minutes. 



14 WATER PURIFICATION 

and at a velocity of filtration of 4 metres (13 feet) per hour, 
94-3 per cent of the bacteria in the raw water are removed by the 
Jewell Export Filter. 

The increase of sulphates occasioned by the addition of the 
chemicals is confined within such narrow limits that a deteriora- 
tion of the water for drinking and domestic purposes does not 
come into the question in practice. 

The increase of aluminium salts, apart from the consideration 
that this disappears completely with well-ordered management, 
is so small that hygienically it may be neglected. At all times in 
those places where in the water-supply plant there is no horizontal 
space at disposal for expansion, the Jewell Export Filter can be 
applied with advantage, as, for example, in cases of- water supply 
in besieged strongholds in time of war. 

Further, Friedberger has also carried out searching investiga- 
tions with the Jewell Filter, with the water of the town of Konigs- 
berg. He comes to fewer favourable conclusions than Schreiber. 
With water rich in bacteria, mechanical filtration does not 
guarantee so complete a reduction of the bacteria that one could 
be satisfied with mechanical filtration alone. 

A quite new work of Gottschlich and Bitter gives an account 
of over four years' practical experience of Jewell Filter man- 
agement for the town of Alexandria. The plant worked 
excellently during this time as regards removal of bacteria and 
turbidity, as well as the trustworthiness of the method. 

{d) Artificial recovery of Ground-water from Surface-water 
{Intermittent filtration) . 

For the reasons discussed on page 2 ground-water is to be pre- 
ferred to purified surface-water. As a consequence, municipalities 
which are in a position to do so are always attempting to supply 
themselves more and more with ground-water for drinking 
purposes. In order to obtain such water in sufficient quantity, 
many towns find it necessary to go a considerable distance from 
the town. Considerable expense thereby results in conveying, 
and also in superintending the water conduits, etc. Owing to the 
expense of bringing water from a distance, it has been attempted 
many times to increase artificially the ground-water in the 
neighbourhood of towns. 



FILTRATION PROCESSES 15 

The recovery of artificial ground-water was originated scienti- 
fically by Thiem. 

So-called natural sand or bank filtration comes into considera- 
tion here. It consists in the sinking of wells on the banks of a 
lake or river. The water in these wells is then pumped away, and 
its level thereby considerably lowered. As a consequence water 
enters the wells from the lake or the river through the sand or 
gravel strata, which act as the filtering medium. 

In its progress through the ground the water is freed from 
bacteria in a manner similar to ground-water. Bank filtration 
results in the suspended matter being gradually drawn through 
the sand. As opposed to artificial sand filtration in which fil- 
tration is vertical, bank filtration is horizontal, and this is really 
the cause of the observed passage of suspended matter through 
the filtering layer. 

The investigations of Scheelhaase with Maine River water 
showed that with wells at a distance of 25 metres from the river- 
bank, while the bacteria were of course removed, the water had 
become like ground-water in no other respects, since the tempera- 
ture was not sufficiently adjusted, nor was its odour nor taste 
improved. 

Intermittent filtration in a vertical direction has later been 
investigated in various ways. According to Richert there are 
the following methods : The surface-water is conducted to an 
irrigation field, where it is allowed to drain away. This method 
ought to have been tested on an experimental scale. It has proved 
of little use, as it is untrustworthy and difficult to control. 

It is better to lead the surface-water to an intermittent filtra- 
tion basin or well, which has been sunk to the ground-water level, 
or to a basin which lies over the ground-water. 

Success in producing a workable ground-water depends on the 
possibility of leading the stream sufficiently far in a horizontal 
direction, and on the water having time, apart from the removal 
of bacteria, to become a useful ground-water in respect of the 
temperature, colour, taste, and smell. In this direction, Scheel- 
haase, in Frankfort-on-Maine, has lately published important and 
interesting researches. 

River Maine water purified by means of a sand filter was allowed 
to flow to a double-branch irrigation bed laid out 3 metres deep, 
50 metres long, and constructed of gravel and drans. The 



16 WATER rURIFICATION 

infiltrate, finely distributed by this treatment, had to trickle 
in a vertical direction, to the natural ground-water level, through 
a layer of ground 13 to 14 metres deep, consisting of fine sand 
and gravel. Then it joined with the natural ground-water, and 
flowed along with it to the nearest pumping-station following 
the incline of the ground. The result of the investigation was that 
at 100 metres from the place irrigated, a distance which the in- 
filtrate flowed through in 190 days, Maine water, which is very 
dirty river-water, had been transformed as regards its bacterio- 
logical nature, its temperature, smell, taste, and colour, into a 
water equally as good as ground-water. 

J. Braikowitz reports on the nature of artificial ground-water 
in different towns : "In Offenbach-on-Maine the water of the old 
waterworks is brought to the neighbourhood of the new works, 
which draws upon a well of water free from any objection. In 
Brunswick the condenser water from steam engines in the water- 
works is made to percolate through the ground, whereby the 
ground-water obtained only experiences a rise in temperature 
of 0-4° C. In the Ruhr Waterworks the river-water is led through 
ditches, or through a broken-up portion of the choked-up river-bed, 
to the layer of rubble-stone underneath. The Ruhrtalsperren Co. 
seeks to increase the lower waters of the Ruhr, and also to 
augment the ground- water, by constructing dams." 

According to Richert the ground-waterworks of the town of 
Schweinfurt is a beautiful example of natural filtration. 

Since the year 1875 the town of Chemnitz has employed inter- 
mittent filtration with the best results, in which, above a series 
of wells sunk in a ground-water area, an irrigation field for 
intermittent filtration was dug out to the ground-water level. 
The water introduced into the open ditches of the irrigation field 
unites directly with the ground-water, and flows along with it to 
the wells. 

In a similar manner the town of Gothenburg produces artificial 
ground-water. 

(e) The Importance of Water Filtration in Public Health Ad- 
ministration, and Critical Opinions on the value of Sand 
and Mechanical Filtration. 

The purification of surface-water for drinking purposes by 
means of filtration has become a question of great importance 



FILTRATION PROCESSES 17 

as affecting the health of towns and their economic improve- 
ment. 

According to Hilgermann, cholera and typhoid diseases 
especially have decreased in those places where sand filtration 
has been introduced and properly conducted. In those places 
where epidemics have appeared in spite of filtration, they have 
been caused by faulty arrangement of the sand filter or faulty 
management of the undertaking. During the cholera epidemic 
in Hamburg, in the year 1892, sand filtration worked splendidly, 
since, for example, the town of Altona, which used filtered Elbe 
water, was quite free from the epidemic although the filters 
received for their work Elbe water, rendered contagious by all 
the Hamburg discharges. 

According to Vincey, as recently as 1905 a number of town- 
ships around Paris employed raw Seine water. After the intro- 
duction of sand filters mortality in typhoid cases fell about 
42 per cent, cases of typhoid about 48 per cent. 

Hilgermann, in his work already mentioned, critically com- 
pared the newer American mechanical filters with the sand 
filter. 

European workers who have experimented in recent years 
with American mechanical filters have in general come to favour- 
able conclusions. American professional men who have had the 
opportunity of studying the mechanical filter on the spot for 
ten years are not especially favourable. 

According to Hilgermann, the main difficulty with the American 
mechanical filters which work by addition of aluminium sulphate, 
is that it is not possible, with the varying composition of the raw 
water, to add the right amount of chemicals. This causes faulty 
sedimentation and faulty formation of the filter layer. 

On the basis of his researches with the Jewell Filter, Hilger- 
mann comes to the following conclusions on the working of the 
mechanical filter : — 

1. With raw water containing a small number of bacteria 
the American mechanical filter yields good results. 

2. The efficiency of the whole method of filtration depends 
upon the sedimentation. 

3. The addition of aluminium sulphate at any time is dependent 
upon the amount of the substances suspended in the raw water. 

4. With poor raw water the mechanical filter fails if the addition 



18 WATER PURIFICATION 

of chemicals cannot be immediately increased with the increase 
of suspended matter in the raw water. 

5. There is no principle for such a regulation. The increase 
must be settled by experiment. 

6. As regards the removal of turbidity due to clay and also 
the removal of colour, and in respect of the hygienic method of 
cleaning, the mechanical filter is superior to the sand filter. 

Closer examination shows that as regards cost, sand and 
mechanical filtration are approximately alike. Of course, the 
cost of setting up the mechanical filter a second time is far smaller 
than with the sand filter, owing to its small dimensions and to 
the small space it requires. Still, the wear and tear of these 
machines is much greater than is the case with sand filters ; 
hence one has to allow for a greater depreciation. Further, the 
constant consumption of chemicals raises the cost of manage- 
ment considerably. 

(ii) Methods of Water Sterilisation. 

The removal of bacteria from water can take place by filtration 
and by sterilisation, i.e. the destruction of the bacteria. For 
this purpose a number of chemicals have been employed, e.g. 
ferric chloride, chrome iron alum, lime, hydrogen peroxide, 
calcium permanganate, chloride of lime, bromine, copper 
chloride, organic acids, ozone, etc. Of these very many must be 
rejected, since they either do not work with certainty in general, 
or they need a long time for the development of their sterilising 
action, or they alter the appearance, smell, and taste of the water 
too much. The only methods of any practical value are those in 
which ozone and chlorine (as chloride of lime or sodium hypo- 
chlorite) are used. 

{a) The Ozone Method. 

Ozone, the so-called active oxygen, which is formed from the 
oxygen of the air by the silent discharge of high-tension electric 
current, has proved to be a good medium for sterilising water. 
When dissolved in water it kills the greater part of the bacteria 
and then escapes again from the water, without influencing 
taste or smell in the slightest degree, since it decomposes to 
ordinary oxygen. For the sterilisation of drinking-water a number 
of ozone plants of varying design have been proposed ; for exampl e , 



20 



WATER PlTllIFICATION 



tlie system of Siemens and Halskc,^ Trindall, Al^raham Marmicr, 
Otto and Wjsmaer. Quite a number of to\\ns now purify tiieir 
water hv means of ozone, e.f;. Paderljorn-i.-W'.. St. Petersburg, 
Hermannstadt, Nizza, St. ]\rans, near Paris, and otliers. 

As an example of an ozone plant, that of the metropolis, St. 
Petersburg, may be more closely described here as one of the 
newest according to the system of Siemens and Halske. As 
can be seen from Figure 3, the raw water is taken direct from the 




Fig. 4. Ozone \\'.\ter\vorks of St. Petersburg. 

Neva by means of a pump, and pumped to the sedimentation 
reservoirs for purification. Before entering the purification 
reservoir tlie water is treated with aluminium sulpliate. It is 
then filtered througli tliirty-eight mechanical filters. These filters 
are on the Howatson system, which is similar in many respects 
to the pre\'iousl3? described Jewell Filter. To the filtration plant 



^ Recently the Berlin Ozone Compan\' has incorporated the firms of .Siemens 
and Halske and the General Electric Co., and lias taken o^er ;ill the patents 
of Siemens and Hcilske. 



STERILISATION PROCESSES 



21 



there is attached the actual ozone plant, which consists of two 
parts, the ozone batteries and the steriliser. 

In Fif^ure 4, on the left, the ozone batteries, consisting of 128 
])ieces of apparatus, are shown, and on the right the five sterilisers 
can be seen, (jne of which serves as a reserve. The individual 
apjKiratus arc Siemens and Halske ozone cylinder elements, as 
shown in I*"igiu-e 5. In this apparatus the oxygen of the air is 
converted into ozone by means of high-tension discharges. The 




l'"ic.. 5. OzoNi-: \V.\ I i:R\v<)iti;s, .St. Ti; 11 ksiu i;r.. Ozonk ISattkrijcs. 

concentration of ozone amounts to 2-5 grams in i cubic metre 
(i grain per cubic foot) of ozonised air. The air, before entering 
the ap])aratus, is dried in a cooling machine. The mo\'ement of 
till' air through the ozone batteries and pipes takes place by the 
aid of the so-called emulsihers (Otto's system). These cmulsifiers 
ari' injectors or water-jet air-pumjis, which, by means of a water 
jiressure of alxiut 4 nu'tres (i()o iu.), sucl<.s th(.' ozonised air out 
oi (he ozone batteries, and brings it mixed with water into the 
steriliser. The absorjition of the ozone and the consequent steri- 
lisation of the \vater takes place partly in the cmulsifiers placed 



22 WATER PURIFICATION 

near the sterilisers, and partly in the agitators, from the bottom of 
which the ozonised air rises to the top in a very fine state of divi- 
sion, and therefore in very intimate contact with the water. From 
the emulsifiers and sterilisers the water passes over a cascade for 
removing the air to a pipe which leads it to the pure-water reser- 
voir. From there it is pumped away into the town's mains. 

It is a necessary preliminary for the satisfactory working of an 
ozone plant that the water to be sterilised contain no suspended 
matter, and not too large an amount of organic matter, or ferrous 
oxide. In such cases the ozone is in great part consumed in the 
oxidation of the dissolved substances, or of the iron. The 
unsatisfactory working of a plant in Schierstein was attributed 
to the presence of a considerable amount of iron in the drinking- 
water. 

The researches undertaken by Erlwein, Ohlmiiller, and Prall on 
the "Auftrage des Kaiserlichen Gesundheitsamtes" (Commission 
of the Imperial Sanitary Board), and by Proskauer and Schuder,of 
the "Konigl. Institut fiir Inf ektionskrankheiten " (Royal Institute 
for Infectious Diseases), with the water of the Spree, and with 
water to which large quantities of pathogenic bacteria (typhoid, 
diarrhoea, cholera) had been added, are in agreement in proving 
that the bacteria are almost entirely destroyed, and that the 
pathogenic variety were in all cases destroyed without exception. 

The Pasteur Institute also obtained favourable results in its 
investigations on ozone processes. 

Halbertsma and Dolezalek prove that the opinion that no daily 
control is necessary in ozone processes as contrasted with sand 
filtration is wrong. 

Karl Schreiber, in the " Auftrage der Konigl. Prufungsanstalt 
fiir Wasserversorgung und Abwasserbeseitigung " (Commission 
of the Royal Test-Institution for Water supply and Sewage 
disposal), as a consequence of the unfavourable observations of 
Halbertsma and Dolezalek, undertook a searching examination 
of the ozone works in Paderborn, in which he established that the 
ozone process satisfied all demands. 

G. W. Chlopin and K. E. Dobrowolski report that in St. Peters- 
burg the bacteria are not completely killed off, but are decreased 
on an average about 98-8 per cent. Intestinal bacteria should 
be absolutely destroyed. The water should also undergo an 
improvement in taste and colour. The chemical composition 



STERILISATION PKOCESSES 23 

does not alter essentially, and the ozone dissolved in the water 
disappears after ten minutes. 

S. Rideal reports on his experience in the works at Paris. 
After ozonisation all bacteria, except the more resistible spores, 
were destroyed. The temperature of the water was not raised. 

According to Gartner the bacteriological action is as good with 
ozonisation as with sand filtration, and it is also more certain. 

Sauna made experiments in which ozone was used in quantities 
averaging about 4 milligrams per litre (1-75 grains per cubic 
foot. ' It gave the following results : Nitric acid is completely 
destroyed ; 15 to 43 per cent of organic matter is oxidised. 
The efficiency of ozone on the organic matter increases accord- 
ing as the water is more oxidisable. Ozone also oxidises 
ammonia present in water. Sulphates, carbonates, and chlorides 
are not altered. Nitrates and free oxygen show an increase. 
Within a few minutes the last traces of ozone disappear. 
Hydrogen peroxide is not formed By using suitable amounts of 
ozone, complete sterilisation of the water takes place. Ozone 
acts most effectively on pathogenic bacteria. 

The sterilising effect of ozone depends, according to Schreiber, 
on four factors, viz. — 

1. On the nature of the water. 

2. On the amount of water passing through the plant. 

3. On the concentration of the ozonised air. 

4. On the amount of the ozonised air used. 

For the right working and control of ozone plants, Schreiber 
recommends that a hygienist and an official experienced in 
electrical management should undertake a test of the four factors 
mentioned, and that they should then work out regulations for 
running the process in accordance with the test. The finished 
plan sliould be tested, as regards its sterilising action, with water 
in which bacterium coli and similar bacteria have been dis- 
seminated. The maintenance of the working regulations must 
be controlled from time to time by an electrical engineer. 

The cost of sterilising water in Paderborn ranges, according 
to Schreiber, from 0.43d. to i-5d. per 1000 gallons without pre- 
liminary filtration, and with filtration from o-66d. to i-94d. per 
1000 gallons. 

At the Parisian works, according to Rideal, 1-31 kilowatts 



24 WATER PURIFICATION 

current were necessary to sterilise loo cubic metres of water. The 
total cost, exclusive of the repayment of loans and payment 
of interest, amounted to o-33d. per looo gallons. 

In the St. Petersburg undertaking the cost of working is 
o-87d. to i-od., of which only half is due to ozonisation. 

The cost of sand filtration, according to Schreiber, amounts in 
comparable cases from i-o5d. to i-87d. per looo gallons. 

{b) The Disinfection of Drinking-W ater with free Chlorine. 

In the year 1894 it was shown by Traube that, by treatment 
with very small amounts of free chlorine, the bacteria in water 
could be destroyed. Later investigations by other authors 
showed that for certain destruction of the bacteria the amounts 
of chloride of lime must be greater. The process remained thus 
for a long time, but was eventually taken up again in practice 
several years ago. Recently, however, the method has been 
extensively applied, especially in America. 

The Duyk Ferric Chloride Process, Howatson System 
{Thumm and Schiele). 

In this process the raw water, either during or immediately 
after the sedimentation of the undissolved particles, is treated 
with a solution containing chloride of lime, and then with one 
of ferric chloride. The mixture is led on to a mechanical filter 
(Howatson system) without further sedimentation, or with an 
intermediate disposition of sedimentation arrangements ; the 
filtered water is then ready for use. 

By the addition of the chemicals the following reactions take 
place : 

6 CaOCl2+Fe2Cl6=6 CaClj+FeaOa+S CI2O. 
3 01^0=3 a +3 O. 

The resulting chlorine and oxygen act as disinfectants, whilst 
the voluminous ferric hydroxide and the calcium carbonate^ 
resulting from the decomposition of the calcium chloride, draw 
the suspended matter along with them to the bottom on settling. 

In Middelkerke, in Belgium, the method is in practical applica- 
tion ; there, and later in Paris, it has been tested. The con- 
clusions are favourable. The destruction of the bacteria must 
have been extensive. 



STERILISATION PROCESSES 25 

Thumm and Schiele, who inspected the Middelkerke under- 
taking, and base their view on the impressions obtained and the 
accounts received, express themselves as generally favourable. 
In the ferric chloride process it may be important to consider 
that the chlorine disappears of itself after some time, and that 
the water treated does not then contain any harmful chlorine 
compounds. 

(c) Sterilisation with Chloride of Lime or Sodium Hypochlorite. 

This method, according to Imhoff and Saville, has been intro- 
duced in considerably more than one hundred American towns. 
The method is especially suitable in those places where the water 
is physically good, but has the disadvantage of containing either 
permanent or transitory disease germs. 

The chloride of lime is generally added to the water in the 
proportion of i part to 350,000 parts of water (which means i part 
of active chlorine per 1,000,000 parts of water). 

The free chlorine acts upon the bacteria, more especially on 
those which communicate disease. 

The cost of the process is very moderate, being only o-o27d. 
per 1000 gallons, therefore only a twentieth the cost of other 
purification processes. 

The method is not applicable where much suspended matter, 
or even organic matter, or iron, is present in the water, for then 
the free chlorine is consumed in the oxidation of these substances. 

On account of the small amount of bleach added, it should not 
be necessary to displace the chloride of lime from the water 
subsequently ; the water, also, should not possess any appreciable 
taste or smell. 

Quite recently a series of reports upon the method have come 
from English and American towns. 

G. A. Johnson, who has taken up the process again in America, 
declares that sodium hypochlorite can be used as well as bleach. 
The following are mentioned as advantages of the process : 
the rapidity with which the bacteria, especially pathogenic 
bacteria, are quickly destroyed ; the ease with which the addition 
of chemicals can be adapted to any change in the water ; the 
absence of any harmful reaction product in the water ; the 
rapidity of the reaction. 



26 WATER PURIFICATION 

The bleach treatment is useless in the following points : bac- 
teria spores are not destroyed ; the bacteria embedded in the 
suspended solid particles remain unaffected. 

C. Walker reports on the chloride of lime water-purification 
process called the De Chlor process, which was tested for six 
months on an experimental scale. The De Chlor system removes 
the excess of chlorine by means of a preparation (charcoal) 
insoluble in water, added to the filter in granular form. The 
number of bacteria fell, after passing through the filter, to 
between 234 and 421 per cubic centimetre ; the purified water 
contains on an average 32 germs per cubic centimetre. Bac- 
terium coli was found in i cubic centimetre of raw water, in 
10 cubic centimetres of the first filtrate, and in the pure water 
was no longer found in 100 cubic centimetres. 

According to Craven, 0-3 to 0-4 milligram of chloride of lime 
is added to every litre (20 to 25 grains per 1000 gallons) of 
Ohama water, which is previously treated in clarifying basins. 
The number of bacteria is thereby reduced by about 97 per cent. 
Bacterium coli could only be found in isolated cases. In 
Minneapolis from 1-54 to 3-04 milligrams is added to a litre 
of water. The number of bacteria in the raw water sank from 
between 250 and 8000 to between 7 and 1200. Bacterium coli 
was not found in the water so treated. 

In the opinion of the author it is of advantage to use the 
method for a time with a certain amount of scepticism, as sus- 
picion almost prescribes that to be certain of the destruction 
of all the bacteria such an excess of bleach is necessary, that it 
is noticeable in smell and taste, or vice versa, that if there is 
no smell and taste in the purified water the bacteria are not 
destroyed with certainty. 

Add to this also that it is not yet established that the con- 
tinued use of small quantities of bleach, even should they be 
imperceptible by the senses, is not a matter for serious mis- 
givings. 

For this reason this method cannot be recommended for general 
imitation. It might possibly be of service in certain circum- 
stances, if there were temporary dangers with the drinking- 
water, as in times of epidemic, or it might perhaps be used with 
success in time of war. 



STERILISATION PROCESSES 27 

(d) Sterilisation of Drinking-Waier by means of Ultra-Violet 

Light. 

It is well known that if white light be analysed into its com- 
ponents by means of a prism, beyond the extreme violet end of 
the spectrum there can be indicated certain rays which cannot 
be perceived as light rays, and are therefore invisible, but to 
which there belong powerful chemical activities, e.g. towards a 
photographic plate. 

That these ultra-violet rays also possessed the power of killing 
bacteria has long been known. Downes and Blunt, two English 
investigators, observed this in the year 1877, and the power of 
sunlight to kill bacteria is attributed to these ultra-violet rays. 
Ultra-violet rays are to-day generated by means of the quartz 
mercury-vapour lamp. An electric current is sent through 
mercury vapour which is enclosed in an evacuated quartz-tube ; 
the mercury vapour thereby glows and sends out ultra-violet 
rays which have the property of passing through quartz though 
they are retained by glass. 

The real discoverers of the sterilisation of drinking-water 
by means of ultra-violet rays are the French workers, Courmont 
and Nogier. In their experiments a lamp was fastened in the 
axis of a cylinder of 60 centimetres diameter, so that the walls 
were not more than 30 centimetres distant from the source of 
the light. Only clear water, without turbidity or colour, is 
sterilisable in this manner. The lamp is purposely immersed in 
the water. In the first place the sterilising action is better when 
the lamp is immersed, as the water is in closer contact with the 
source of the rays, and all the rays are used up on every side. 
It seems, however, to be also necessary to immerse the lamp 
in order to cool it, and to prevent it varying on account of heat 
changes. It was formerly thought that the action of the ultra- 
violet rays rested on the formation of hydrogen peroxide or ozone. 
That is not the case ; such compounds could never be shown 
to be present. The taste, smell, temperature, and chemical 
properties of the water are in no way altered by the rays, and 
protracted experiments on animals have further demonstrated 
the complete harmlessness of the water so treated. 

The sterilising action, as communicated by Courmont and 
Nogier, is good, and not inferior to that in the water-purification 



28 WATER PURIFICATION 

methods previously mentioned. Also as regards economy, the 
process, according to the accounts of Courmont and Nogier, can 
bear comparison with the above purification methods. 

From experiments conducted by two members of the Konigl. 
Prufungsanstalt fiir Wasserversorgung und Abwasserbeseitigung 
(Royal Institute for the Testing of A\'ater Supply and Sewage 
Disposal) the results are not so favourable as those of Courmont 
and Nogier in all points. Grimm and \\'^eldert carried out their 
researches with the mercury-vapour lamp of the Quartz Lamp Co., 
Ltd., Hanau-on-Maine. They collect the results of their work as 
follows : — 

" (i) ^^'ith the apparatus tested clear water, containing few bac- 
teria, can be sterilised at the rate of 0-55 cubic metres (120 gallons) 
per hour under the conditions described above. Clear water, very 
rich in bacteria, can, on the other hand, only be rendered sterile 
at the rate of 0-45 cubic metre per hour, in which case it 
does not make any difference whether the bacteria are water 
bacteria or, with pathogenic bacteria, coli bacteria. (2) Turbidity 
of the water, even to a small extent, makes disinfection uncertain. 
With a high degree of turbidity the destruction of bacteria by 
the lamp is impossible, at all events within the limits which come 
into practical consideration. (3) Likewise the yellow colour of 
water due to colloids, such as bog-water shows, acts as a very 
great hindrance. Indeed, wdth only a slight amount of coloration, 
the hindrance is so great that it is practically impossible to 
accomplish disinfection by this method. (4) The water is not 
altered in any physico-chemical respects by its passage through 
the experimental apparatus, with the exception of an increase 
in temperature of a few tenths of a degree. With prolonged action 
of the rays, further increases of temperature result, as well as 
s3rmptoms of chemical decomposition. (5) The expenses of 
water purification by means of ultra-violet light, reckoned on the 
basis of the researches, are, comparatively speaking, very high, 
and cannot bear comparison with the cost of the methods of water 
purification employed at the present time on a large scale." 

Erlwein reports on the experience of the firm of Siemens and 
Halske in this respect. The energy required is greater than in 
sterilisation with ozone. With regard to the other factors, a com- 
parative estimate of the cost of working, under the conditions 
operative in a central waterworks imdertaking, is wanting in the 



STERILISATION PROCESSES 



29 



case of the ultra-violet light method, and especially the most 
important, viz. a more exact knowledge of the deterioration 
and cost of repairs of the still very expensive quartz lamps. 

Buywid is of the opinion that the ultra-violet light method is 
more promising than the ozone method. 

To the knowledge of the author the method has not been to 
the present applied on a large scale. Mercury- vapour quartz 
lamps are supplied by different firms, e.g. The Ultra- Violet Firm, 
22 Rue Chanchat, Paris ; Westinghouse, Cooper, Hewitt and 
Co., 131 Wilhelmstrasse, Bferlin, the Quartz Lamp Co., Ltd., 
Hanau-on-Maine. 

Taking into consideration the fact that the water is not in any 
way changed as regards its nature, and that extended researches 
with animals have shown that water acted upon by the rays even 
for a long time is not harmful, the method should have a future 
if the cost can be reduced. 

In addition to their researches on the ultra-violet light method, 
Grimm and Weldert collected together the costs of the principal 
water-purification methods. Nos. i to 4 of the following table 
are the averages of estimates of various undertakings on a large 
scale. No. 5 is reckoned on the basis of the experiments of 
Grimm and Weldert. 



(I) 



(2) 



(3) 



(4) 



(5) 



List of the various Processes. 


Cost 
of 


per 1000 gallons 
purified water. 


Slow Sand Filters. 






Cost of working 
Total cost 




o-6id. 
5-2d. 


Mechanical Filters. 






Cost of working 
Total cost 




2-8d. 
5-8d. 


Ozone Plants. 






Cost of working 




2-8d. 


Total cost 




8.2d. 


Chloride of Lime Plants. 






Cost of working (Johnson) . 
Total Cost (Johnson, Imhoff, and 
Ultra-violet light plant. 
Cost of working : 
100% Bacteriological Effect 
99 to 99-9% „ 


Saville) 
5s. 


o-o7d. 
o-34d. 

14s. 6d. 
to 7s. 3d. 



30 WATER PURIFICATION 

(e) Disinfection of Water-Mains and Wells. 

If an epidemic reigns in a town, and there are grounds for 
believing that the contagion is to be attributed to drinking- 
water, the disinfection of the water-mains is to be recommended, 
especially in those cases in which a fresh drinking-water, free 
from any objection, must be led into the infected main. Fliigge 
and Bischoff, during a typhoid epidemic in Beuthen (Upper 
Schleswig), employed for this purpose a o-z per cent solution of 
sulphuric acid. The acidified water remained standing in the 
main many hours. The strength of the sulphuric acid in the water 
is controlled at the stop-cocks. In an epidemic of typhoid fever 
in Gelsenkirchen, in the year 1901, the mains were likewise dis- 
infected with sulphuric acid. 

With wells which in general are suitably situated and lie in 
surroundings free from objection, but which seem to be infected 
from above, it is possible, according to a proposal of M. Neiszer, 
to introduce compressed steam (5 atmospheres) ; the whole 
contents of the well are thereby brought to the boiling-point. 

The same method, or disinfection with " Carbolic-Sulphuric 
acid " (Frankel), which is subsequently pumped away, can be 
applied in the preliminary works of a central water supply for 
disinfecting borings, so as to be able to take away samples 
bacteriologically free from objection. 



(iii) Purification of Water in directions other than that of Health. 

As already mentioned, substances occasionally appear in water 
which, without being detrimental to health, still cause great 
trouble, since, as with iron and manganese, they may discolour 
the water, or, as with free carbonic acid, may attack the walls 
of pipes and reservoirs. 

{a) Removal of Iron. 

Ground-waters of the diluvial and alluvial strata of the North 
German Lowland often contain iron irf greater or lesser quantities, 
as ferrous carbonate or the ferrous salt of humic acid. When 
freshly drawn the water is generally clear, but becomes turbid 
after some time owing to the separation of a brown precipitate, 



OTHER PROCESSES 31 

since the oxygen of the air converts the ferrous salt into in- 
soluble ferric hydroxide with evolution of carbon dioxide. 

Although this turbidity due to iron is objectionable not so 
much from the point of view of health as from that of appearance, 
still it frequently causes trouble. The flocculent hydroxide, 
settling in the pipes and reservoirs, renders their frequent cleans- 
ing necessary. Further, the presence of iron in water favours the 
appearance of numerous micro-organisms which store up iron, 
especially the iron bacteria, which decompose and evolve odours 
of sulphuretted hydrogen and other decomposition products, 
and gives to the water a specific metallic taste. 

Such water is not suitable for most technical purposes. 

A few tenths of a milligram of iron in a litre of water may make 
an iron-removal plant necessary, since iron bacteria thrive best 
in waters weak in iron. 

The methods of removing the iron are based on three physico- 
chemical actions. 

First, by contact of dissolved ferrous iron with air, oxygen 
converts the ferrous compound into insoluble ferric compounds. 
Consequently the water may be aerated, and subsequently filtered. 

Secondly, since carbonic acid keeps the iron in solution, the 
precipitation of the iron may be effected by neutralisation of the 
carbonic acid with lime. 

Finally, if the iron is present in the water in colloidal form 
(organically combined) , a coagulant, such as aluminium sulphate 
or ferric chloride, is employed. 

The IcLst method is often used in America, whilst in Germany all 
processes for the removal of iron are based upon aeration and 
filtration. 

The method of effecting this aeration and filtration varies 
considerably with different systems. 

Aeration is effected by allowing water to fall through the air, 
cascades (Elbing), by means of raining devices using perforated 
troughs (Wismar), in coke-towers (Piefke), over wood (Berlin), 
over clinkers (Delitzsch), over glazed brick (Sternberg), and in 
other ways. 

Similarly, the method of filtration varies. In general, sand and 
gravel have proved the best media for filtration. The size of the 
grains of sand is an important factor in the formation of the filter 
layer, and in the complete retention of the ferric hydrate particles. 



32 WATER PURIFICATION 

The method of cleaning the filter and aerator is also different 
in the various systems. 

In the following review of the most important methods for 
removing iron are also included those which are used mainly 
for industrial purposes, since they are based on the same principles 
and are conducted in precisely the same way as are the larger 
undertakings for central water supply. The removal of iron 
from single wells, on the other hand, will be specially treated on 
page 49. 

The most important systems for the removal of iron are 
characterised as follows (Schwers) : — 

1. Piefke system : One of the oldest and best systems. The 
water containing iron is brought to the coke-tower, a cylindrical 
upright vessel, filled with pieces of coke the size of a man's fist. 
The water flows slowly over the coke and passes into a settling- 
tank situated underneath, whence it flows on to a sand filter, 
which is arranged on the usual lines. The coke-tower is cleaned 
by flushing in the reverse direction. It should only be necessary 
to refill the tower once or twice each year. These Piefke towers 
are largely used, and have proved excellent. 

2. Oesten system : A raining arrangement (2 m. rainfall), 
with roses, and filtration through gravel the size of wheat-grains. 

3. Kurth system : Violent rainfall and gravel filtration, worked 
by the stroke of a piston ; for small plants. 

4. Bieske system ; Analogous to the previous one. 

5. Thiem system : Raining arrangement with perforated trays. 

6. Reichling system : Raining arrangement with centrifuge 
or sieve, closed filtration upwards, under pressure, through layers 
of sand, gravel, and wood-wool ; used in industry. 

7. Koerting system : Similar to the previous one. 

8. Pfeiffer system : Simple aeration and sand filtration. 

9. Wingen system : Waterfalls, sand filtration. 

10. Taacks system : Waterfalls, sand filtration. 

11. Krohnke system : Coke-tower, rotating cylindrical filter 
that is alternately filled with sand and water. 

12. Lanz system : Filtration through natural sandstone. 

13. Fischer system : Filtration through porous artificial stone 
(Wormser Kunststein). 

14. Agga system : Filtration through pipes of artificial stone 
in the sand filter, purification by a reverse stream of water. 



OTHER PROCESSES 33 

15. Reisert system : Filtration in the open air through gravel, 
with or without previous aeration over coke ; cleansing by a 
reverse stream of water. 

16. BoUmann system : Gravel filtration under pressure without 
previous aeration, purification by a stream of water in the 
reverse direction. 

17. Breda system : Filtration under pressure through " Ton- 
koks " (clay-coke) and gravel of various sizes, after preliminary 
aeration in a mixer ; purification by a reverse stream of water. 

18. Helm system : Filtration through brown iron-ore slag 
in which aeration takes place by means of occluded oxygen ; 
purification by reversing the water current. 

19. Biihring system : Filtration, as in the previous case, 
through bone charcoal ; purification with dilute hydrochloric 
acid ; for domestic purposes. 

20. Biittner system (von der Linde and Hesz) : Filtration under 
pressure through wood shavings, impregnated with tin oxide 
without any special aeration ; purification by reversal of stream. 

21. Bock system : Filtration analogous to the previous one, 
through wood-wool. 

22. Sellenscheidt system : Raining arrangement, filtration 
through plant-fibre ; used especially in breweries. 

23. Dehne system ; Aeration by raining with an injector, 
addition of milk of lime, filtration under pressure through felt 
discs ; used in industrial concerns. 

24. Jewell system : Rapid sand filtration with or without 
addition of milk of lime and sulphate of alumina, unaccompanied 
by any aeration. This system was introduced in Posen, in 1909, 
producing 30,000 cubic metres per day. No chemicals are added ; 
the water is freed from iron by simple filtration. 

25. " Voran " system : Frankfort-on-Maine. Aeration in 
the closed system by compression through nozzles, in the open 
system by sprinkling over a cataract-like erection of coke, bricks, 
etc. An open and a closed " Voran " plant are pictured in 
Figures 6 and 7. The figures are intelligible without further 
explanation. 

Plants for the removal of iron are built both open and closed. 
Closed systems offer greater protection, naturally, against 
infection of the water by bacteria ; still, according to Schwers, 
plants with open apparatus have proved as good as the closed 



3i 



WATER PUIUFICA'IION 



fn.im the bacteriological point ol \-iew. In recent years a \-ehe- 
ment dispute has been in jirogrcss with reterence to the greater 
or lesser suitability of the open anci closed systems. Both systems 
have been equally praised and criticised. On the \\hole they may 
be regarded as of equal \'alue. 

The efficiency of iron-renio\-al plants depends upon the height 
of the aerator and filter, the velocity during aeration and filtra- 
tion, and other technical points. 



Hi 




Fig. 6. C^pi:n Plant for 

TIIK REMOVAL OF n^OX. 
■• \OKAN " SVSTF.M. 







e LOSI.n 1 LANT r (Ik TIIL kLMO\AL01 
1kO.\. " \'l>KA.\ " SV.-irLM. 



The height of the aerator is generally about 3 to 7 metres ; 
the ^'elocity of filtration in open s^'stems seldom amounts to more 
than I metre, in closed systems it is often 10 metres (per hour). 

The pressure is generally produced bj- the difference in level 
between the surface of water in the filter and in the pure-^^•ater 
reservoir. 

Good iron-remo^'al plants yield a water which at most still 
contains o-i milligram iron per litre. 

The cost of remo\-ing iron from water amounts in a series of 
German towns from 0-04 to o-qd. per 1000 gallons. 



OTHER PROCESSES 35 

(b) Removal of Manganese. 

As a result of the water calamity in Breslau in the year 1906, 
general attention has been turned to the occurrence of manganese 
salts in water. Although only present in small amounts, they 
deposit a weak scum which makes the water insipid, stains linen 
and paper, pollutes the reservoirs in breweries, injures the com- 
plete action of yeast, etc. Certain micro-organisms absorb it 
to a still greater degree than iron. Manganese almost always 
accompanies iron, but the amount is generally so small that its 
presence in the majority of cases does not approach practical 
importance at all. 

Proskauer wels the first to point out the occurrence of man- 
ganese salts in water. 

Manganese is removed by aeration, like iron, but it separates 
out with greater difficulty, since, during aeration, mangani- 
manganous compounds result, which are soluble to a considerable 
extent in water. 

The calamity in Breslau originated in consequence of the 
existing geological conditions. Iron and manganese sulphides, 
in the humous layers situated over the ground-water, were 
converted to sulphates by oxidation and then passed into the 
ground-water owing to a flood which inundated the whole tract 
of country. 

For the removal of manganese, and also of iron, Permutit has 
recently been recommended. Permutit is an artificially prepared 
product (aluminium silicates), principally employed in the 
softening of water (see page 60) . 

Permutits have the property of withdrawing from water, 
lime, magnesia, iron, and manganese, in exchange for sodium, 
if the water be allowed to flow over them. Liihrig and Becker, 
Gaus and Noll, have carried out experiments on the removal of 
manganese by means of calcium permutit. The manganese is 
thereby exchanged for calcium. Whilst Liihrig' and Becker 
obtained good results in laboratory experiments, a trial on a large 
scale proved a failure. The water took from the permutit sub- 
stances which imparted to it an alkaline reaction, causing a 
precipitation of manganese as oxyhydrate. This precipitate 
choked up the filter. Noll found on an experimental scale that 
manganese was quantitatively removed from water by calcium 



36 WATER PURIFICATION 

permutit, so long as the content of the permutit in manganese 
was less than 2 per cent. The cost of removing manganese with 
calcium permutit is estimated by Noll to be o-ogd. per looo 
gallons. 

Quite recently the Permutit Filter Company, Berlin, have recom- 
mended a new method for removing manganese and iron by means 
of permutit. According to Kriegsheim, the method is as follows : 
From a suitable permutit, e.g. sodium permutit, a manganese 
permutit is prepared by treatment with a solution of manganese 
chloride. This manganese permutit is then treated with potassium 
permanganate. The permanganic acid is thereby easily combined 
with the manganese oxide of the permutit to form highly oxidised 
manganese compounds. If a water containing manganese be now 
filtered through a permutit filter so treated, these higher oxides 
of manganese can very rapidly effect complete removal of the 
manganese, even with rapid velocity of filtration. The action 
is based upon the fact that the oxygen necessary for the oxidation 
process is presented to the manganese separating out during the 
filtration of the water in the solid condition and easily split off 
from the highly oxidised manganese oxides. The oxidation 
causes the precipitation of the manganese in the water in an 
insoluble form. Should the action diminish, the filter can be 
regenerated by means of a 2 to 3 per cent solution of perman- 
ganate. The cost of this process should be very small. 

As already mentioned, the process is not only recommended for 
removal of manganese, but also for the removal of iron. The 
principle is the same in the latter case also. 

(c) Removal of free Carbon Dioxide. 

If water contains much free carbon dioxide, it exercises a very 
deleterious action on the various materials required in water- 
works. Thus it was observed, for example, at Frankfort-on- 
Maine, that a newly constructed deep reservoir made of reinforced 
concrete was strongly corroded by the water. Similarly the iron 
pipes were vigorously attacked. 

Whilst the action of the water on these materials is more a 
matter of economy than of hygiene, the corrosion of lead pipes, 
which are used in most cases for the conveyance of water inside 
houses, is in the highest degree serious from the point of \ iew of 



OTHER PROCESSES 37 

health, since lead passes into the drinking-water as a result and 
is, even in the smallest quantities, inimical to health. 

In practice three methods are used for the removal of free 
carbonic acid. 

I. The water is allowed to flow through limestone (Heyer- 
Scheelhaase) . The free carbonic acid is thereby converted into 
calcium bicarbonate according to the following equation : 

CaCOg +H2O +C02= Ca(HC03)2. 

By this method, therefore, hardness due to carbonates is in- 
creased. Where the water is of itself very soft this increase in 
hardness is not harmful, and may even be desirable. If, however, 
the water is already fairly hard, the increase in hardness might 
be a disadvantage. Besides, the removal of carbon dioxide by 
means of limestone is not accomplished in such cases, or only 
incompletely. The method is in use on a large scale in Frankfort- 
on-Maine for the removal of the free carbonic acid from Stadt- 
wald water of 1-5° hardness, and containing 30 milligrams CO2 
per litre. It had caused great trouble in the new deep reservoirs 
of Sachsenhaus, and also in the mains. The hardness of the water, 
when freed from acid, amounts to about 5° 

As the Frankfort plant for removing carbonic acid has proved 
excellent, it may be shortly described as follows (see Fig. 8) : — 

In the years 1906-7 the plant for removing carbon dioxide 
was erected at a cost of 78,000 marks (about £3900) in the 
Chamber A of the deep reservoirs of the Sachsenhaus plant. 

Seven of the ten passages of the chamber were employed for 
the neutralising process. The first serves as an inlet chamber, 
the second and third as sand filters, and the four following as 
limest6ne scrubbers. 

From the inlet chamber the water is distributed through per- 
forations in the partition-wall on to the sand filter, which only 
serves to retain the particles of ferric oxide carried along out of 
the pressure tubes. The velocity of filtration amounts to 80 m. 
in 24 hours. The filtered water passes through openings to the 
base of dividing-wall, and with the help of distributors passes 
under the limestone scrubber. The limestone scrubber is com- 
posed of a layer of flints at the bottom, over which rests a layer 
of gravel, and then four layers of limestone in different states of 
division, namely, about the size of walnuts, beans, peas, and 



38 WATER PURIFICATION 

coarse sand. Each of the first three layers is 3 inches high, the 
height of the fourth being 2 feet. The velocity of water in the 
scrubber reaches 40 metres (130 feet) in 24 hours. The water, 
when neutralised, flows partly through an opening in the partition- 
wall between sections 7 and 8 to the last three sections of Chamber 
A, and partly through a special pipe to Chambers B, C, and D 
of the reservoir. About every three months the sand filter has to 
be cleaned, by being washed through with water in the reverse 
direction from the bottom upwards. 

With about 5,000,000 gallons of water neutralised daily, 
approximately i| tons of limestone are dissolved by the water 
and carried away. 

2. Caustic soda, or sodium carbonate, is added to the water 
in calculated amount, according to the estimation of carbonic acid 
(Heyer). The free carbonic acid is converted to the bi-carbonate. 

CO2+H2O +Na2C03^2 NaHCOs. 

This method was introduced by Heyer to neutralise the drinking- 
water in Dessau. In the eighth decade of last century a large 
number of persons became ill through lead poisoning. Heyer 
perceived that the capacity to dissolve lead was due to the 
amount of free carbonic acid present. After various other 
experiments he finally proposed the neutralisation of the water 
with caustic soda and sodium carbonate. 

There are many types of apparatus for measuring the amount 
of chemicals to be added. It is stated and agreed concerning 
this method that it makes the carbonic acid quite harmless. The 
general public, however, has generally an instinctive aversion 
towards a drinking-water treated with chemicals. This fact is, 
doubtless, the main objection to this otherwise good system. 

3. If a water containing carbon dioxide is allowed to rain 
down in a fine state of division, or allowed to trickle slowly over 
coke, glass, gravel, etc., it is freed from carbon dioxide. 

With larger amounts of carbon dioxide present, the raining 
process has to be repeated many times to remove the carbonic 
acid. Further, the height of the fall has an influence. The 
removal of carbon dioxide in this way is disadvantageous for the 
reason that the water is greatly enriched with oxygen. A water 
rich in oxygen is also not good as, in its turn, it causes severe 
rusting of the iron pipes. 




< 

is 

O 
I 

O 

b 

< 



k; 






9U0:fsauji-] Ajeuiuii/s^ Durfne^ilqsin 



40 WATER rUllIFICATION 

H. Wehner has now advanced the idea of effecting the trickhng 
process in a vacuum. According to the accounts of the inventor, 
not only ought the carbon dioxide to be removed more rapidly, 
and with less height of fall necessary, but also, simultaneously, 
the amount of oxygen should be considerably diminished. 

Wehner's vacuum process is to be used in some English and 
German towns. 



II. PURIFICATION OF DRINKING-WATER ON A 
SMALL SCALE. 

The greatest advantage of a central water supply is this, that, 
owing to the size of the project, there are more extensive pre- 
liminaries for securing a water free from any objection, greater 
certainty for carrying out the process according to conditions, 
and greater possibility of expert supervision of the process, 
so that disturbances are easily recognised and prevented. 

Over and above this there exists in all small-scale purification 
processes the main weakness that the points of preference named 
above with regard to large-scale operations are either non- 
existent or are not sufficiently assured. 

Isolated houses, estates, establishments, etc., would do well, 
therefore, to choose for their water supply, water free as far as 
possible from any objection, and not needing any purification. 

Should unforeseen circumstances intervene, making the water 
of a central water supply, or even a single supply, seem doubtful, 
then methods of purification on a small scale can find application. 
Of first importance is boiling of the water before use. 

Moreover, the employment of purification apparatus on a small 
scale is to be regarded as a makeshift, since it should be con- 
stantly controlled to determine whether it works well and 
continuously. This, naturally, in the majority of cases, cannot 
be accomplished. 

Circumstances like epidemics, war, journeys in unknown or 
quite uncivilised lands, constitute an exception. In such cases 
it is often impossible to procure water free from objection. The 
small amount of water necessary, comparatively speaking, on 
such occasions can be purified with sufficient certainty by the 
aid of purification methods on a small scale. This is all the more 



SMALL OR HOUSEHOLD FILTERS 4.1 

true when, as, for example, in wars, or in times of epidemic, 
the necessary personnel is at one's disposal. The observance of 
the regulations can then be supervised. 

This review of apparatus for water purification on a small 
scale naturally refers only to methods which aim at the removal of 
bacteria. Of apparatus for removing substances prejudicial to 
the appearance of the water, such as iron, etc., there are also 
many on a small scale, as already set forth above, and it appears 
that the small apparatus has the same action as the corresponding 
apparatus on the large scale. 

(i) Purification by means of Small or Household Filters. 

The household filter serves a double purpose : firstly, it removes 
suspended matter and makes the water clear ; secondly, it is 
used for retaining bacteria also. Household filters are much 
less germ-tight than sand filters. The bacteria are washed 
through the filter if it is not very frequently cleaned, and not 
infrequently does it happen that the filtered water is richer in 
bacteria than the raw water. Small filters, therefore, are to be 
used with great caution. 

The number of different forms of domestic filters is legion. 
They are all based on the filtration of the water through some 
porous material with or without the employment of increased 
pressure. Von Esmarch divides them into the following groups : — 

(a) Charcoal Filters. 

These are the oldest form of apparatus of this kind. They are 
composed of plastic retort carbon or finely sieved charcoal, 
powdered coke, and compounds of the most varied preparations. 
They generally cost between 30 and 70 shillings each. 

Charcoal filters are scarcely used nowadays, since their bacterio- 
logical effect is practically nil. Indeed, if a charcoal filter is in use 
for some time, the filtrate is often worse than the raw water, since 
the bacteria which have remained behind in the charcoal have 
increased considerably, and are then carried along with the 
stream of water passing through. Also, as regards the removal 
of turbidity these filters are of little use. 



42 WATER PURIFICATION 

(b) Stone Filters of Sandstone, Pumice, and the like. 

The stone is burnt from coarse or fine sand, quartz, lime, and 
magnesium silicates. For filters that work without pressure 
coarse material is used, whilst for pressure filters extremely fine 
material is employed. 

The clarification of the water may be rather good with such 
apparatus. The bacteria, on the other hand, go through the 
filter fairly readily, at the latest after two or three days. The 
productivity of the filters is generally small, usually about i litre 
per hour. Frequently the yield falls off quickly also, and cleaning 
then becomes necessary. 

(c) Asbestos Filters. 

Asbestos of very fine fibre is used as the filtering medium, and 
is employed as a pulp, or compressed or mixed with other 
materials. Asbestos filters are also used in technical work for the 
filtration and clarification of turbid liquids (beer, oil, wine, etc.), 
and for the retention of ferric oxide in iron-removal plants. 

The filter of C. Piefke and the micro-membrane filter of Fried- 
rich Breyer are capital asbestos filters. Asbestos filters retain 
bacteria fairly well ; but in general they choke up very quickly ; 
this necessitates frequent cleaning and sterilisation of the appara- 
tus, which takes up much time. 

According to Gartner, Breyer's micro-membrane filter ought 
to be germ-tight. 

Good asbestos filters are supplied by the firm of Arnold and 
Schirmer, 123 Grosze Frankfurter Strasze, Berlin, N.O., or by 
H. Jensen and Co., 20 Reichenstrasze, Hamburg. 

(d) Clay Filters. 

Of these filters there are likewise a number of different speci- 
mens as, for example, those of Olschewski and Hesse. They only 
filter free from bacteria for a short time. The greater the yield 
of water the shorter time are they germ-tight. 

(e) Porcelain Filters. 

The principal example of this type of filter is the Chamberland 
Filter. It consists of a metal cylinder, which can be screwed on 



SMALL OR HOUSEHOLD FILTERS 43 

to the water-tap. Inside the cylinder there is fastened an inner 
hollow mould, made of fine porous kaolin and attached to the 
cylinder so as to be water-tight. The water enters between the 
outer shell and the porous mould, penetrates inwards through 
the porous cylinder, and flows away through the opening situated 
underneath. 

The filtrate is bacteria-free and so the yield of the filter is 
small, amounting after a few days to but a few litres per day. 

The apparatus is cleaned by boiling and thoroughly heating 
the dried apparatus. The purified filter then regains its original 
productivity. 

The Chamberland Filter can be obtained from the firm of 
Lautenschlager, Dramenburger Strasze, Berlin. 

(/) Kieselguhr Filters. 

The chief representative of these filters is the Berkefeld Filter, 
which is constructed on the same principles as the Chamberland 
Filter, but is composed of baked infusorial earth (see Fig. 9). 
They filter free from bacteria for various lengths of time, as a rule 
several days. They generally yield from | to 2 litres per minute 
at a pressure of i to 2\ atmospheres. Gradually the yield 
diminishes, but it can be raised again to just the original amount 
by brushing the cylinders clean. They can be sterilised by simply 
warming slowly on the water-bath. 

The filters are, as far as freedom from bacteria and 3'ield are 
concerned, certainly the best small filters. With daily sterilisa- 
tion one can rely with tolerable certainty on a filtrate continuously 
free from bacteria. 

The price of a filter for attaching to the water-tap in houses 
is from 30 to 35 shillings ; smaller types from 13 to 16 shillings, 
and as pump filters from 46 to 200 shillings. The price of a repair 
cylindrical mould is 4s. 6d., of an army fUter (yielding J litre 
per minute) 30 shillings, and a transportable pmnp £%. The 
filters can be obtained from the Berkefeld Filter Co., Celle 
(Hanover). 

P. Schmidt has studied the mechanism of bacterial filtration 
with Berkefeld Filters. He finds that the effective size of the 
pores in Berkefeld Filters (with Liliput moulds) is probably about 
0-5 yu. By finely grinding Berkefeld Filters choked up with 



44 



WATER PURIFICATION 



bacteria, it was shown that the choking only takes place on the 
surface of the filter, so that complete cleaning is possible in the 
mechanical way by reversal of the water-current. 

Staphylococcus and organisms of about that size never pass 
through the filter, but, on the other hand, a number of small 
bacteria pass through in measurable amount if they are washed 




Fig. 9. Berkefeld Filter. 



along in the water in very large quantities. Schmidt is of the 
opinion that for the passage of bacteria through a filter their size 
is of primary importance, and their mobility next in importance. 
In general, according to Konig, the capacity of a domestic 
filter to retain bacteria depends: (i) on the nature of the filtering 
medium ; it must be uniform, with the pores not too big ; (2) on 
the water-pressure ; this should not be more than i to 2 atmo- 
spheres, and should not act jerkily, since such pressure assists 



BOILING THE WATER 45 

the passage of bacteria through the filter ; (3) on the amount of 
pollution in the water ; the more suspended matter a water 
contains, the more quickly does the filtrate cease to be free 
from bacteria ; high temperature also assists the passage of 
bacteria. 

Generally speaking, the more germ-tight a filter is, the less 
productive it is also. 

(ii) Boiling the Water. 

All vegetative forms of bacteria are killed by long-continued 
heating of water at the boiling-point. Those bacteria which form 
endogenous spores withstand, of course, for the most part, the 
process of boiling, but even they are considerably weakened. 

The pathogenic varieties coming mainly into consideration — 
cholera and typhoid — form no endogenous spores. For this 
reason boiling has been employed with advantage for the sterilisa- 
tion of drinking-waters on a small scale. 

If larger quantities of water have to be boiled — for families, 
hospitals, schools, factories, ships, etc. — and if this sterilisation, 
as for example in times of epidemic, has to be undertaken for 
a long time, then the following apparatus, according to von 
Esmarch, are to be recommended : — 

{a) The apparatus of the German Continental Co. (Deutschen 
Continentalgesellschaft), of Dessau, for gas-heating, yields 
6| gallons per hour, using 10-5 cubic feet of gas (price £3 15s. od.). 

(b) An apparatus by Grove, Friedrichstrasze, Berlin, also for 
heating by gas an attachment to the water-pipe. This apparatus 
yields 15 to 22 gallons per hour with 14 cubic feet consumption 
of gas (price £15). The water flowing out is about 5° warmer than 
that flowing in. 

(c) Apparatus by Siemens and Co., Berlin, yields about 8 gallons 
of water hourly ; 100 gallons require 80 cubic feet of gas. The 
price is £2 5s. od., or, with a control apparatus which is to be 
recommended, £3 15s. od. The effluent water is from 5 to 10° 
warmer than the inflow. 

{d) Apparatus by Schafier and Walcker, Berlin, for gas-heating, 
yielding 6 to 8 gallons per hour. 

{e) Apparatus of Pape and Henneberg, Hamburg, for burning 
gas, petroleum, or coal with automatic self-regulation of the 
inflow. The smaller apparatus yields 55 gallons hourly, and costs 



46 



WATER PURIFICATION 



£38 ; 220 gallons of water require about 300 cubic feet of gas, or 
28 lb. of coal. 

(/) Apparatus by C. Aug. Schmidt and Sons, Hamburg-Uhlen- 
horst. The firm supplies apparatus for houses, hospitals, etc., 
also with automatic self-regulation of the inflow. For a yield 
of 20 to 30 gallons per hour the apparatus costs £37 to £60 
(see Fig. 10). 




Fig. 10. Apparatus for Boiling Water of the Firm of 
Aug. Schmidt and Sons, Haimburg-Uhlenhorst. 

(g) Transportable apparatus of Rietschel and Henneberg, 
Berlin, for the use of armies and in case of epidemics. The plant 
yields about 66 gallons per hour; smaller portable apparatus 
can be had, yielding 20 gallons per hour. 

As against filtration, purification of drinking-water by boiling 
has the following disadvantages : — 

(i) It only kills the bacteria, and does not remove the suspended 
matter. 



SMALL OZONE PLANTS 47 

(2) The upkeep of the apparatus is expensive. 

(3) As a result of loss of gases and salts, and as a consequence 
of the rise in temperature, boiled water is generally much less 
palatable than filtered water. 

As opposed to these disadvantages there are the following 
points of preference : — 

(i) The productivity remains continuously the same, whilst 
with filters it falls away. 

(2) Boiling the water renders the destruction of disease germs 
absolutely certain. It is, on this account, the best of water- 
purification methods in this direction. 

(iii) Small Ozone Plants and Ultra-Violet Light Apparatus. 

The ozone method of sterilising drinking-water, as set forth 
previously, has proved itself good in central water supplies. 
Lately, various firms, especially the firm of Siemens and Halske, 
have constructed small ozone plants for sterilising smaller 
quantities of water for communal and private industrial concerns, 
as well as for the purpose of supplying drinking-water to troops 
in the field. 

Both stationary and transportable small ozone plants have 
been designed. 

They are used in the Munich Brewery, for the purification of 
the water used for cleansing vessels and for similar purposes. 
Transportable ozone plants are employed for the supply of water 
to troops in the field. Such plants were used by the Russian 
authorities in the Russo-Japanese war, and their use for military 
purposes is said to be under consideration by various other 
Governments. The whole apparatus is lodged on two waggons, 
the machine-waggon and the sterilisation-waggon. 

In practice the raw water is brought, by means of a water-pump 
on the machine-waggon, through a thick suction and pressure pipe 
to a filter, and from there to the sterilising-tower. The filter and 
tower are on the sterilising- waggon. Through a thinner air- 
suction and pressure pipe the air passes from bellows on the 
machine-waggon into the ozone apparatus on the sterilising- 
waggon, and from here passes to the base of the sterilising-tower. 
By means of a cable the primary current of an alterpating-current 
machine on the machine-waggon is conducted to the transformer 



48 WATER PURIFICATION 

on the sterilising-waggon, "which is set up directly underneath the 
ozone apparatus to generate the requisite working-tension. 

Each waggon requii"es one horse, and weighs about i ton. 

The plant yields 400 to 600 gallons per hour, and requires for 
the process about 2 horse-power. 

For safety the ozone is generated in such excess that with 
very dirty water only one-third to one-half is consumed. 

The apparatus was tested by Proskauer, the Russian hygienist 
and bacteriologist, K. Kressling, and a Russian military com- 
mission prior to its despatch to the seat of war in Manchuria. The 
results of these tests were very satisfactory. 

M. Neiszer made a report on experiments with two ozone 
plants for work on a small scale, which were placed at his disposal 
by the firm of Felten and Guilleaume-Lahermeyer-werken A.G. 
of Frankfort-on-Maine. 

In both apparatus the ozone was generated by current 
from the supply for lighting purposes. The mixing of the ozone 
with the water was effected by means of an aspirator attached 
to the water-tap. 

Opening the water-tap started both the generation of ozone 
and its mixing with the water. The ozone generator consists 
of a 120 to 5000 volt high-tension transformer enclosed in a 
protecting box, with plate condensers. The ozone generator 
is attached to the town supply (120 volts, 45 periods) and con- 
sumes 0-55 ampere. The amount of ozone generated is about 
5 to 6 milligrams per litre of air, with about 2 litres per minute. 
The experiments were conducted with staphylococcus {Staph, 
■pyog. Aur.) and with bacterium coli. They showed that with 
suitable water (no suspended matter, no iron, and little organic 
matter), thousands of bacteria per cubic centimetre can be killed 
with certainty, provided they have a resistibility intermediate 
to that of staphylococcus and bacterium coli. It is important to 
observe that when mixing was intimate, and when sufficient 
quantity of ozone was used, momentary contact was sufficient 
to kill the germs. 

Further, the firm of Siemens and Halske have designed, as a 
result of their previous experience, a sterilising apparatus, making 
use of ultra-violet light rays, for affixing to the domestic supply 
and also for single water-supplies. Their contrivance is character- 
ised essentially by a mechanism whereby, on turning a single 



REMOVAL OF IRON FROM WELLS 49 

handle, the lamp is set working through being tilted, and at the 
same time the water-tap is opened. An electro-magnet is further 
provided which only allows a flow of water through the apparatus 
when the lamp is properly alight. Also, if the lamp goes out 
during the process the water-inflow tap immediately closes, so 
that water which has not been sterilised cannot be drawn from 
the apparatus. 

(iv) The Removal of Iron from Single Wells. 

The methods of iron removal, using small plants, whether it 
be for the purposes of a central water supply or for industrial 
application, have already been discussed on page 32 and 
the following pages. There remain still to be mentioned 
those methods which serve especially to remove iron from 
single wells. 

The oldest and simplest method of removing iron from well- 
water is to pour into the well iron-free water which has taken up 
oxygen (generally by standing in the air). The oxygen precipi- 
tates the iron as hydroxide, which then settles to the bottom. 
Water can in this way be freed for many days from the iron it 
contains. The chief disadvantage lies in the fact that the 
precipitated iron remains in the well. In addition to this the 
water poured into the well should, of course, be hygienically 
free from any objection, which it never is. 

For this purpose the Dunbar Filters are very suitable. Of such 
filters the simplest and cheapest construction consists of a tub 
filled with sand, having a tap in the side at the bottom. The 
water is allowed to flow out of the pump on to the sand. The iron- 
free water is removed through the tap underneath. About every 
three months the sand must be freed from the precipitated 
oxide of iron by washing. Such an arrangement is not suitable 
to wells situated in the open, since they are liable to freeze. 
Still, this objection can be overcome, as the whole plant can 
be housed against the frost in a special small hut. 

The Deseniss and Jacobi system of so-called bastard pumps 
(see Fig. 11) differs from all other systems in this, that by means 
of a pump the water is taken from the well already freed from 
iron. All intermediate apparatus, such as scrubbers, clarifying 
reservoirs, etc., are therefore not required. 

E 



5U WATER PURIFICATION 

The following is the arrangement and method of working such 
pumps : To the pump employed for raising the water there is 
attached a second cylinder of double the circumference of the 
cylinder of the water-pump. In both cylinders there move 




F[<;. [I. Till-: Bastard Pi'.mi' 
OF Deseniss and Jacobi. 

tightly closing pistons, provided with valves and fastened to 
a piston-rod. The water streaming in from the lower cylinder 
to the one above, since the latter has double the capacity of the 
former, is thus mixed with an equal volume of air, sucked in 
through a valve laterally placed in the higher cylinder. After 
being mixed with air the water is forced on to a forged-iron filter, 
carrying sand of 0-5 millimetre grain as filtering material. It is 
thereby freed from the hydroxide of iron that has been formed. 
The filter can be set up in a small shaft standing in close proximity 
to the pump, or it may e\'en be placed in the pump uprights 
itself. 

It is not necessary to renew the filter layer at all ; the purifica- 



PURIFICATION FOR TECHNICAL PURPOSES 51 

tion of the filter is effected by temporary automatic flooding 
in the reverse direction, and is attained by simply reversing the 
stop-cocks. 

Schreiber has tested this method in an experimental plant, set 
up in the chief pumping-station of the Charlottenburg ground- 
water works. He finds that the pump removes even the last 
traces of iron, or at most leaves an inconsiderable amount behind. 
It further fulfils all the demands made upon a hand-pump as 
regards simplicity in construction, slight supervision, and pro- 
tection against pollution. 

Moreover, the bastard pump is not only supplied for hand- 
work, but also for working by machinery. 

According to the declarations of the firm, the removal of the 
iron should be alike complete with a hand-pump of usual size, 
or with a large-scale plant yielding many thousands of litres per 
hour. 

B. Purification of Water for Technical Purposes. 

Besides the use of natural water for drinking purposes, water 
is needed for almost all human occupations. In most cases with 
a natural water purification is not necessary for technical pur- 
poses, since the demands made upon the quality of such a water 
are naturally less than with a drinking-water. 

For cleansing public urinals and lavatories, for fountains, and 
for watering the streets and gardens, untreated river-water is 
frequently used. Where the river-water is very dirty, at most 
a coarse gravel filter is employed, and this simply serves to retain 
suspended matter, but not, as is the case with drinking-water, 
the bacteria. Many towns, like Frankfort-on-Maine, have along- 
side the drinking-water pipe, one for river-water, which serves 
the purposes mentioned. 

For many industrial purposes the water, however, must undergo 
treatment. For example, in paper manufacture, the presence of 
iron is troublesome, since iron forms a compound with cellulose 
and makes the paper spotty. The removal of iron in such cases 
is effected in small iron-removal plants, which have already been 
described on page 32 and the following pages. In many industries 
— for example, with laundries, dye works, and others — hard water 
is a great disadvantage, as a part of the working material is 



52 WATER PURIFICATION 

precipitated by the constituents causing hardness, and is thereby 
rendered useless. Hence, for such purposes the water must be 
softened. Quite generally, however, the softening of water 
plays a large part in industry, owing to the use of water in boilers. 

The Softening of Boiler-Feed Waters. 

The hardness of a water is due to the calcium and magnesium 
salts dissolved in it. A distinction is drawn between temporary 
hardness, or hardness due to carbonates, and permanent hardness, 
or hardness due to mineral acid salts. Carbonate hardness is 
caused by the bicarbonates of calcium and magnesium which, 
unlike the normal salts, are soluble in water. On boiling the water 
these bicarbonates give up half their carbon dioxide in the free 
state, change to the normal salt, and therefore become insoluble. 
The observation that a part of the hardness of water disappears 
on boiling has earned for such hardness the title of temporary 
hardness. 

Permanent hardness is due to the presence of the salts of cal- 
cium and magnesium with sulphuric, hydrochloric, and nitric 
acids, and therefore to calcium and magnesium sulphates, 
chlorides, and nitrates. 

Hardness is measured in German or French degrees of hardness. 
1° on the German scale is equivalent to i mg. CaO or 0-74 mg. 
MgO per 100 CCS. of water. 1° on the French scale is equivalent 
to I mg. CaCOg or 0-84 mg. MgCOg per 100 c.cs. water. ^ 

If hard water is used for feeding boilers, in process of time there 
settles on the boiler plates a scale, consisting of precipitated 
calcium and magnesium salts. With a large amount of tem- 
porary hardness, there further results in the boiler a big sludge of 
loose, precipitated normal calcium carbonate. The troublesome 
effects of this boiler scale are well known. In the first place, once 
a boiler scale has reached a certain thickness, the conductance of 
heat to the water is greatly hindered, and so coal is consumed 
very uneconomically. Boiler scale can also lead, however, to 
the direct deterioration of the boiler plates, since the scale 
possesses a different coefficient of expansion from that of the 
boiler plates. In this way fissures and cracks result. The 
chlorides of lime and magnesium, when present in water in fairly 

1 1° on the English scale is equivalent to i grain CaCOj per gallon. — Trans. 



PURIFICATION FOR TECHNICAL PURPOSES 53 

large amounts, are very troublesome as they are hydrolysed 
with formation of free hydrochloric acid, which passes away with 
the steam and attacks the equipment. 

At the present time three methods are principally employed 
for softening water, namely, the Lime-soda method, the Reisert 
Baryta process, and the Permutit process. 

The lime-soda method, as the name implies, consists in adding 
to the \vater lime and sodium carbonate. The temporary hard- 
ness is rernoved, by the addition of lime, in the form of calcium 
carbonate and magnesium hydroxide, according to the following 
equations : — 

1. Ca(HC03),+Ca(OH)2=2 CaC03-f2 Ufi. 

2. (a) Mg(HC03)2 4-Ca(OH)2=CaC03+MgC03+2H20. 
(b) MgC03+Ca(OH)2=Mg(OH),+CaC03. 

Permanent hardness is removed by means of sodium carbonate 
according to the following equations : — 

1. CaSOi +Na2C03 =CaC03 +Na2S04. 

2. (a) MgCla +Na2C03 =MgC03+2NaCl. 

lb) MgC03+Ca(OH)2=CaC03 +Mg(0H)2 (Pfeiffer). 

All calcium salts in the water are therefore precipitated in the 
form of the normal carbonate ; all magnesium salts, no matter 
in what form they are present (whether as temporary or per- 
manent hardness), are finally precipitated as the completely 
insoluble hydroxide. Hardness due to carbonates is therefore 
entirely removed from water by this method, but in place of the 
precipitated salts causing permanent hardness, an amount of 
sodium equivalent to the calcium and magnesium originally 
present passes into the water as sulphate, chloride, or nitrate. 

Prom the equations given it can easily be calculated that, for 
each degree of temporary hardness, lo g. of lime (ioo% CaO) are 
needed to soften i cubic metre of water. Also it is obvious from 
the equation that for each milligramme of magnesia pi-esent 
in a litre of water, no matter what the form, 1-4 g. of lime (100% 
CaO) must be added to a cubic metre of the water. For each 
degree of permanent hardness 19 g. of sodium carbonate (100% 
Na.^C03) are needed per cubic metre. In the measurement of the 
quantity added, it is to be borne in mind that natural waters 
generally contain some free carbon dioxide. Since carbonic acid 



54 WATER PURIFICATION 

combines with lime to form bicarbonate, which then, of course, 
combines with more hme to form the normal carbonate, for every 
milligram of free carbon dioxide per litre an addition of 1-27 g. 
of 100% CaO must be made per cubic metre, in order to neutralise 
this free carbonic acid. The addition of lime, therefore, must be 
increased by such an amount for every milligram of free carbon 
dioxide present. 

From various investigations, however, the amount of lime to 
be added, thus theoretically reckoned, proves to be too high in 
practice. It must be cut down by about 20 to 25 per cent. Vari- 
ous reasons for this have been put forward. 

Instead of calcium hydrate, caustic soda can be used. From 
the bicarbonate there results sodium carbonate which, in its 
turn, serves to remove permanent hardness. 

Ca(HC03)2+2NaOH=CaC03+Na2C03+2H20. 
Mg(HC03)2+4 NaOH=Mg(OH)2+2 Na2C03+2 H^O. 

If precisely so much temporary hardness is present that the 
sodium carbonate resulting according to the above equations 
is just sufficient to remove the permanent hardness, then the 
water can be completely softened with sodium hydrate. If 
more permanent hardness is present, then the sodium carbonate 
still required has to be added. If, on the other hand, there is too 
little permanent hardness, only so much sodium hydrate should 
be added as can form the amount of sodium carbonate requisite 
for the removal of the permanent hardness. The remainder of 
the unprecipitated carbonates must then be removed by means 
of lime. 

Formulae for reckoning the necessary addition of caustic soda, 
sodium carbonate, and lime have frequently been given. If a 
is the combined carbonic acid, h the total lime, and c the total 
hardness expressed in equivalents of hardness, then the following 
formula of Kalmann give the necessary additions of caustic 
soda, lime, and sodium carbonate : — 

m=2a — b ; n=c—a. 

If m and n are positive, the water contains more temporary 
than permanent hardness ; m is then the lime to be added, n the 
caustic soda. 

If 2a — b = o, there is present in the water as much temporary 



PURIFICATION FOR TECHNICAL PURPOSES 55 

as permanent hardness. No lime is then needed, but only n units 
of caustic soda. 

If 2a— b is negative, the water contains more permanent than 
temporary hardness. There must then be added m parts of 
sodium carbonate and c—a=n parts of caustic soda (Wehren- 
pfennig). 

The caustic soda is generally prepared by mixing sodium 
carbonate solution with lime and allowing the precipitated 
calcium carbonate to settle. 

In ascertaining the necessary addition, if lime and sodium 
carbonate be reckoned, and not caustic soda, precisely the same 
amounts require to be added as were given above, on page 53, 
for lime and sodium carbonate. Therefore, with caustic soda 
as the softening agent, the numbers given with respect to lime 
and sodium carbonate can be adhered to. 

The lime-soda process is the oldest, and doubtless, also, has 
the widest practical application at the present time. Softening 
with lime and sodium carbonate never attains 0° hardness, but 
generally only about 3° to 4°, which, however, is quite sufficient 
for general requirements . There are numerous firms who construct 
softening plants of the lime-soda type, which are used with the 
greatest success in practice. Among others are the firm of 
Humboldt in Kalk, Dehne in Halle a. S., Reisert in Cologne, and 
the " Voran " firm of Frankfort-on-Maine. Figure 12 represents 
a lime-soda softening plant, " Voran " system. 

This plant may be briefly described as an example of a lime-soda 
softener. 

The automatic water-purification apparatus of the " Voran " 
System (Model BI) consists essentially of reservoir A, divided into 
three parts by means of partition- walls, the clarifying tank B, 
the lime saturator C, the sodium carbonate regulator D, and the 
filter of wood shavings and wood-wool arranged under the 
clarifying tank B. 

The raw water which is led into the apparatus through the 
pipe E next flows through the discharge pipe F into the part of 
the tank A set aside for the raw water. 

In this section, at equal heights, are placed two stop-cocks 
G and H, through which the raw water can flow away. The 
larger portion goes, through the stop-cock G and the pipe Gj, 
directly to the funnel-shaped central portion I, in which it 



56 



WATER PURIFICATION 



takes up a violent whirling motion owing to its tangential 
entry. 

The smaller quantity of raw water, arranged exactly according 
to the amount of lime-water required, flows through the stop- 
cock H and the discharge tube H^ to the lime saturator C, to the 



. — .>ia^ 




Fig. 12. Lime-soda Water Softener. " Voran " System. 



bottom of which it is led through the central funnel-pipe Hg in 
order that it may pass into the saturator equally distributed on 
all sides. 

From here on its way to the overflow K the raw water is com- 
pelled to percolate through the milk of lime which, since it is 
specifically heavier than water, has settled to the bottom. The 



PURIFICATION FOR TECHNICAL PURPOSES 57 

water is thereby itself completely saturated with caustic lime 
and flows away at the top completely saturated and clear lime- 
water, through the overflow-pipe K to the mixing-trough. 

If the ratio of clear saturated lime-water to raw water is once 
fixed, then it will always remain the same, whether much or little 
raw water is flowing in, for the amounts of raw water flowing 
through the two stop-cocks G and H should always be pro- 
portional. 

The'^§ioda solution necessary flows from its particular part of 
the reservoir A through a hair-sieve M and the attached pipe M^ 
to the sodium carbonate regulator D, in which the inflow is 
regulated by the float N. From the regulator D the sodium 
carbonate solution passes to the mixing-trough through the 
exit pipe Q, so arranged that it can be revolved, and which is 
attached in its turn to the float R in the raw-water reservoir. 
When the float R sinks the discharge-pipe Q is raised and the 
discharge reversed. 

If the raw-water section of the reservoir becomes empty at 
any time, the float R will sink to the bottom, thereby raising the 
discharge-pipe so high that no more sodium carbonate can flow 
out. 

On the other hand, the three streams of raw water, lime-water, 
and sodium carbonate solution commence to flow simultaneously, 
as soon as raw water flows into its particular section of the 
reservoir. 

By this means the amount of chemicals added always bears a 
fixed ratio to the amount of raw water consiuned, that is to say, 
the apparatus is automatically self-regulating. 

If too much raw water be led into the raw-water reservoir it is 
run off by means of an overflow-pipe. 

The lime-water and sodium carbonate solution, after mixing, 
flow from the mixing-trough to the central funnel-shaped pipe 
I, and of course meet the circular motion of the raw water there. 
In this way an intimate mixing with the water is ensured, and 
the commencement of the reaction is hastened. It is especially 
advantageous for the rapid progress of the reaction to have the 
cross section of the pipe I wide at the top and decreasing down- 
wards, since with a constantly increasing velocity of water a 
stronger whirling motion is produced and a more intimate mixing 
ensues. 



58 WATER PURIFICATION 

After leaving I the water is reversed in direction and then 
distributed uniformly in the reaction space and clarifying portion 
B. Owing to the comparatively largie cross section of the chamber 
the water rises to the top with very slow velocity, and on this 
account is freed as far as possible from fine suspended matter. 

The greater part of the sludge does not alter its direction with 
the water, but in consequence of its higher specific gravity sinks 
to the bottom. It settles there in the sludge tank T, from which 
it can be withdrawn as required through the stop-cock T. 

The water rising to the top of the clarifjdng space B, by the 
time it has reached the discharge-pipe U, has allowed the greater 
part of the insoluble precipitated components to settle out. 

Consequently through the discharge-pipe M, water carrying 
only floating particles passes to the underneath side of the filter. 
These remaining particles are retained during the passage of the 
water through the filter, with the result that the water leaves 
the apparatus at Ui softened and filtered, to be led away to the 
place of consumption. 

The wood-shaving or wood-wool filter will only need cleaning 
or renewing at long intervals of time, perhaps after six to twelve 
months, since it is only very slightly used. Renewal should 
easily be effected at a cost of a few shillings. 

The disadvantage of the lime-soda process is simply this, that 
in place of permanent hardness, as has been already mentioned, 
an equivalent amount of sodium salts passes into solution. 
These sodium salts are, of course, quite neutral and readily 
soluble in water, but in course of time attain to such a strong 
concentration in the boiler that they cause inconvenience. The 
salts crystallise at the valves, irregular boiling of the water 
occurs, and so on. The lye must then be blown off. A further 
disadvantage of the process is this, that with the addition of too 
much lime and sodium carbonate the boiler plates may be 
attacked. Generally, however, with frequent checking of the 
pure water produced, the process can be worked in such a way 
that just the right amount of substances is added. Blacher 
rightly calls attention to the fact that it is not sufficient to control 
only the softened water, but that the control must extend also 
to the water in the boiler. For, even if only a slight excess of 
the reagents be present, these excesses are concentrated in the 
boiler in such a way that the boiler plates will be attacked. 



PURIFICATION FOR TECHNICAL PURPOSES 59 

According to Blacher it is considered permissible for the water 
to contain up to 3° of permanent hardness and excess of sodimn 
carbonate up to 57 milligrams Na2C03 per Htre. 

The Reisert Baryta process for softening water removes 
temporary hardness by means of lime exactly as in the first- 
named process. The most troublesome permanent hardness in 
boiler-water is due to gypsum, as this jdelds hard boiler scale, 
which adheres very tenaciously to the boiler plates. The Reisert 
method is therefore confined to the removal of that permanent 
hardness due to gypsum, and this is effected, of course, by the 
addition of barium carbonate. Barium carbonate is practically 
insoluble in water. It is deposited in the water, care being taken 
that it remains in contact with the water for a period of time, 
during which the following decomposition with the calcimn 
sulphate takes place : — 

CaSO^ + BaCOs =BaS04 + CaCOs- 

Gypsum and barium carbonate are therefore converted into 
insoluble calcium carbonate and completely insoluble barium 
sulphate. The lime necessary is naturally the same as in the 
lime-soda process, no matter what the amount of barium car- 
bonate is. Also, the whole of the magnesia is finally precipitated 
as hydroxide, so that it holds in this case also that each milli- 
gramme of MgO present in the water requires double the amount 
of hme. Therefore, for each milligramme of MgO per litre the 
amount of lime to be added per cubic metre must be raised to 
1-4 grammes of 100 per cent CaO, according to the methods of 
estimation on the basis of hardness discussed previously. 

The chief point of preference in the baryta process lies in the 
fact that the sodium carbonate, which when present in con- 
siderable excess in the softened water acts very corrosively on 
the equipment, is done away with. Since barium carbonate as 
mentioned is practically insoluble in water, the baryta process 
possesses the further advantage that the addition of carbonate 
may be liberal. An excess of barium carbonate is thrown into the 
water, which is then allowed after a certain time to settle free 
from the undissolved matter, whereupon it can be employed 
immediately. An excess of barium carbonate is impossible in 
the water softened. Further, in the baryta process, another 
inconvenience mentioned in the lime-soda process disappears. 



60 WATER PURIFICATION 

In that process in place of the permanent hardness sodium 
salts enter the water. By the baryta method the temporary 
hardness and that due to calcium sulphate is removed practically 
quantitatively. It can therefore be said of this process that there 
is less risk of corrosion in the boiler than -is the case with the 
lime-soda process. 

Against these advantages of the baryta process there stand a 
number of disadvantages. Only sulphate hardness and -not that 
of chlorides or nitrates is removed with barium carbonate. It 
is consequently only applicable to water which does not contain 
these salts in appreciable amounts. Waters containing alkali 
sulphates cannot be softened by the baryta process, since in this 
case sodium carbonate would be formed by the interaction of the 
barium carbonate with the alkali sulphates, according to the 
following equation : 

Na^SO^ + BaCOg = Na^COg + BaSOi. 

Sodium carbonate would therefore be present in the softeaed 
water, would get into the boiler and ruin it. 

The most recent method used in practice is the Permutit 
method. Permutit is a complex compound of sodium, aluminium, 
and silicic acid. Such compounds occur in nature as zeoliths. 
Permutit is the name for the artificial product obtained by 
melting aluminium silicate with sodium carbonate, and has 
the peculiar property of removing lime, magnesia, manganese, 
or iron from water when such water is slowly filtered over it. 
At the same time equivalent amounts of sodium are given up to 
the water. The process would scarcely have come into practical 
consideration at all if it had not been discovered simultaneously 
that permutit or zeolith, through which water has been filtered 
for a long time, and which has quite lost its softening capacity, 
can be easily regenerated by washing it with a solution of common 
salt. Thus the process can be reversed. The bases removed 
pass back into solution, and in their stead sodium from the 
common salt passes into the compound. The product so washed 
is precisely of the same utility as the original, and after working 
itself out again can be regenerated anew. 

The advantages of the process are manifest. Above all, the 
process is extraordinarily simple to manage. It is simply a 
question of filtering the water to be softened through a layer of 



PURIFICATION FOR TECHNICAL PURPOSES 61 

permutit of a certain depth at a velocity determined by tests. 
Unlike the two softening processes previously mentioned, which 
both reduce the hardness only to about 3 to 4°, with methodical 
control the softening in this process readily reaches o". A further 
advantage is naturally that the necessity for the addition of 
definite amounts of reagent is obviated. The main disadvantage 
of the process lies in the fact that with water rich in carbonates 
(and most hard waters are rich in these salts), in place of the 
substances removed an equivalent amount of sodium bicarbonate 
enters. In the boiler the sodium bicarbonate gives up half its 
carbon dioxide and is converted to the normal salt. There will 
therefore be constantly present in the boiler water containing 
sodium carbonate which may cause corrosion as already men- 
tioned. 

The method is employed, as previously noted, not only for 
softening, but also, especially of late years, for removing iron 
and manganese from waters. 

Softening plants of the Permutit type are supplied by the 
Permutit Filter Co., Berlin. 

For the correct working of a softening plant, expert super- 
vision, by examination of both the softened and the boiler water, 
is of the utmost importance. Blacher a short time ago suggested 
a very subtle method, by the aid of which exact insight can be 
obtained into the existing conditions in boiler management. 
The method consists in finding out the acid required to neutralise 
the boiler water by titration with jn acid, first using phenol 
phthalein and then methyl-orange as indicator. In addition the 
total hardness is estimated with potassium stearate and phenol 
phthalein. From these three values, which can be obtained in 
quite a short time, one is in a position to suggest how the water 
is constituted. Professor Blacher has constructed a box for 
testing raw water, softened water, and boiler-water, which with 
the accompanying instructions can be obtained at a cost of 
25 shillings from the Vereinigten Chemischen Fabriken fiir 
Laboratoriumsbedarf of Berlin. Every boiler user is warmly 
recommended to provide himself with one. The necessary experi- 
mental work can be learned by an intelligent foreman. 

In conclusion it may further be pointed out that the so-called 
boiler-scale preventatives which are often recommended are 
very much advertised and overrated. In the majority of cases 



62 WATER PURIFICATION 

they are crude swindles, for it is self-evident that substances like 
sugar, starch, and others are not such as would prevent boiler 
scale. Of late, remedies have come into the market whose action 
depends on simultaneous deposition of the medium along with 
the boiler scale, whereby the scale is of a softer nature and more 
easily removed. It is not impossible that such a remedy might 
actually be successful, but a certain amount of scepticism exists 
in regard to them. The most rational method of protecting 
oneself against damage due to the formation of boiler scale is 
always a properly managed softening plant. 



SEWAGE DISPOSAL 

THE SIGNIFICANCE OF RIVER POLLUTION AND THE 
IMPORTANCE OF SEWAGE DISPOSAL. 

ALL water, whether for drinking or domestic purposes generally, 
L or whether employed in cleansing, washing, and industrial 
processes, when not consumed through leakage, evaporation, or in 
some such manner, eventually becomes sewage. Unlike purified 
water, or even a natural unpurified surface-water, sewage varies 
considerably in composition. It contains suspended and dis- 
solved organic and inorganic matter in more or less large quanti- 
ties, and on this account is characterised by a muddy appearance 
and frequently also by a foul odour. 

The purification of sewage is quite recent in its origin. Even 
up to a few decades ago short work was made of effluents. 
Generally speaking, the water used in dwelling-houses was 
conducted in ordinary open sewers to the nearest river. 

The vast increase of large towns and the unexpected progress 
of industry led to the most serious nuisances in this disposal of 
sewage into rivers. The evil, which is every day becoming more 
apparent, is both hygienic and economic. 

Many large towns are dependent on river-water for their 
water supply. As was explained in the previous section, the 
bacteria are generally removed by means of a sand filter. Apart 
from the consideration that the filter naturally cannot remove 
dissolved filth, substances which owing to their origin may be 
harmless seem to leave a water unappetising as a drinking-water. 
Also a greater pollution of surface-water with sewage produces a 
larger number of bacteria in the filtered water, since no filters 
are completely germ-tight. This is all the more serious, as, 
naturally, in sewage waters which contain all human excreta, 
disease germs may very easily be present. 

In those places where surface-waters are not used for the 

63 



64 SEWAGE DISPOSAL 

water supply, river-water is nevertheless often used for baths, 
washing, watering the streets, and similar purposes. All these 
processes ought to be carried out without any danger to health. 

It was shown by the Director of the Pasteur Institute in 
Constantinople that cases of cholera that occurred in the year 
1908 could be attributed to infection of the hands with the water 
of the Bosphorus and the Golden Horn. Similarly it was medi- 
cally established with absolute certainty that a number of typhoid 
cases in the French army had resulted through bathing in 
polluted waters. 

These examples show that river pollution may be very serious 
both as regards washing and bathing. Apart, however, from 
these direct dangers it is greatly to be regretted if river baths, 
which in the view of experts are decidedly superior from the 
standpoint of health to baths in closed rooms, are avoided as a 
result of the foul and unattractive nature of the water. 

Besides, a river polluted with sewage, choked with mud, or 
even smelling badly, will excite in every normal man feelings of 
repugnance and displeasure which may have economic results, 
for the banks of such a river are avoided as places of residence. 
Where a polluted river flows through or by a town, there is 
always a danger of the river-banks depreciating in value. 

In addition to the great economic disadvantages of the pollu- 
tion of rivers, harm also results in the destruction of fish. Often 
the fish are ruined in large quantities by the introduction of 
sewage into the river. The delicate and precious varieties 
especially are very sensitive to pollution of the water, and very 
soon disappear entirely with continued pollution. With certain 
kinds of sewage containing substances of specific taste or smell, 
the fish assume similar taste and smell. An embarrassing point 
of social significance relative to this is, that lower middle-class 
people, like fishermen and proprietors of baths, the protection 
and maintenance of whom in these days of the capitalist on the 
one hand and the proletariat on the other seems to be demanded, 
are either annihilated or grievously injured economically. A 
disadvantage too of river pollution, by no means insignificant 
economically, is demonstrated also in this, that as a result of the 
deposition of mud on the river-bed, expenditure has to be in- 
curred in dredging operations to remove the mud and in similar 
processes of purification. 



RIVER POLLUTION 65 

Consequently the most diverse and weighty reasons have 
constrained authorities in most civilised lands to face the problem 
of excessive river pollution, and to require of towns and factories 
adequate purification of sewage before disposal. 

Self -purification of Rivers. 

Up to a certain point, of course, rivers are able to purify 
themselves from the filth that becomes incorporated. It is by 
no means seldom that just beyond a town a river-course is seen 
to be very dirty, but that a few miles further on the water has 
already reassumed a good condition. This so-called self-puri- 
fication of rivers is explained as being due to a wonderful co- 
operation of physical, chemical, and, above all, biological forces. 
The suspended matter deposited in a river by the sewage gradu- 
ally settles to the bottom as mud. Here it is disposed of by 
snails, mussels, insect-larvae, beetles, worms, and other organisms, 
which consume it, or by continually loosening it prepare it for 
a gradual washing away. Since the organisms mentioned 
represent the chief source of nourishment for the fish feeding on 
the river-bed, it follows that they can never get the upper hand. 
The dissolved organic matter which an effluent water carries to 
a river, especially the more molecularly complex portions, and 
therefore above all the albumens and their decomposition 
products, are mainly decomposed by bacteria. There then 
appear Ciliata, Rotatoria, and Infusoria, which in their turn 
live on these bacteria. Green Alg£e also join in. These and the 
previously mentioned groups are able to absorb dissolved 
organic matter directly from the water. In the main, however, 
the green algs provide for their own nourishment by assimilating 
carbonic acid, nitrates and nitrites which again originate from 
the activities of the bacteria, and from them with the help of 
their green colouring matter, chlorophyll, they build up the 
necessary albumens. During these processes of assimilation 
oxygen is liberated, and this produces powerful aeration of the 
water. This aeration accounts for the existence of the inhabitants 
of the water already mentioned, and also for that of the more 
highly organised parts. The green algae, rotatoria, ciliata, and 
infusoria, etc., serve in their turn as food for the small water- 
crabs or water-fleas, and these again play quite a considerable 



66 SEWAGE DISPOSAL 

part as food for fash. So, finally, the sewage deposited in the 
water is converted into useful fish. 

Industrial sewage containing heavy metals deposits these as 
insoluble sulphides or basic carbonates on the river-bed. Owing 
to the bi-carbonates present in river-water many river-waters have 
a considerable capacity for neutralising acid. The Maine, for 
example, at Frankfort, shows an average alkalinity of about 
3 cubic centimetres of normal acid per litre of water. Since with 
the Maine at average height, 80 cubic metres of water pass 
through its cross section per second, it follows that the amount 
of water passing one point in twenty-four hours would be able 
to take up in round numbers 830,000 litres or 500 tons of 
concentrated 100 per cent sulphuric acid without the water 
giving an acid reaction. For this it is, of course, presumed that 
these amounts of sulphuric acid could mix thoroughly with the 
amount of water under consideration, which naturally is not 
possible. It always happens in the introduction of acids into 
river-water that only a small portion of the river-water at dis- 
posal will mix with the acid effluent, and the consequence is that 
the river-water becomes acid over a certain stretch. 

The same limitation operates as regards what was said above 
with respect to the biological self-regulating process. The 
wonderful co-operation of all these forces and the accompanying 
correct self-purification, are only attained in a satisfactory 
manner when the amount of sewage bears a certain relation to 
the purifying forces. If this relation is unsuitable the river-bed 
becomes covered with mud and with foul suspended matter. 
In the water, large amounts of putrefying bacteria appear, the 
above-mentioned organisms do not get their requirements of 
life and disappear. When this occurs important links in the 
chain are wanting and the orderly co-operation of the whole is 
destroyed. The water at these stages will assume a foul nature 
and cannot sufficiently purify itself. 

As a rule, good limiting values cannot be given as regards the 
amount of sewage that a river can absorb. The amount of 
effluent with which a river may be burdened without being 
harmed is a question which cannot be decided generally, but only 
from case to case with tests of all the factors coming into con- 
sideration. In the first place, for example, the composition of 
the sewage varies greatly. Also, every river possesses its own 



PURIFICATION OF DOMESTIC SEWAGE 67 

specific biological activity, and with that its own specific 
capacity for self-purification which varies from river to river. 
Quite generally it may be said that a river can absorb more 
sewage the more water there is present, and the greater the 
velocity of that water. Demands as to purification of sewage 
must be correspondingly stricter the smaller the flow of the river 
into which it is to be conducted. On the contrary, large rivers, 
containing much water affected by the tide, and having a swift 
current, can fully absorb sewage that is only quite superficially 
purified, '_ 

Sewage can be divided into domestic and industrial sewage. 
Domestic sewage comprises, in the main, waters from dwelling- 
houses, containing therefore faeces, cleansing and washing waters, 
kitchen refuse, and similar substances. It is therefore rich in 
easily decomposable organic substances which very soon become 
foul. Many towns, however, now receive large amounts of 
industrial effluents into their sewage, and consequently in many 
industrial towns the domestic sewage is often mixed with very 
considerable quantities of industrial sewage. 

The sewage from many industrial concerns is characterised, 
as with domestic sewage, by a large content of dissolved organic 
matter which causes it to putrefy readily. Many industrial 
effluents have, however, quite a different composition. Definite 
information on this point cannot be given, since it is dependent 
wholly on the nature of the manufacturing process. 

Although the particular methods of purification are capable 
of equal application in their main features to both classes of 
sewage, the kind of purification necessary for normal domestic 
sewage is always the same, while for industrial effluents, quite 
frequently, variations in the processes occur, adapted to the 
individuality of the sewage coming at the moment into question. 
A distinction between the methods of purification for domestic 
and industrial sewage is therefore to be recommended. 

A. Purification of Domestic Sevs^age. 

The different methods for the purification of domestic sewage 
may be thus distinguished : — 

Mechanical purification (rakes, screens, sieves, grease extractors, 
grit chambers, settling-tanks, settling-wells, septic tanks, Travis 
and Emscher wells). 



68 SEWAGE DISPOSAL 

Coal-pulp processes. 

Artificial biological purification (contact beds and percolating 
filters) . 

Land treatment (broad irrigation or sewage farming and 
intermittent filtration). 

And lately also fish-ponds. 

Mechanical methods of purification aim solely at the separation 
of suspended matter. They have no action on dissolved sub- 
stances. 

The coal-pulp method, on the contrary, has a small influence 
on the dissolved matter, whilst the artificial biological processes 
and land treatment remove dissolved substances so largely that 
the water loses its capacity to putrefy. There is consequently 
this distinction between sewage purified mechanically and that 
purified biologically, that the former becomes foul on standing, 
whilst the latter is incapable of so doing. Mechanical methods 
of purification are therefore applied in those cases where, owing 
to the nature of the river into which it is run, the sewage need only 
be partially purified. Dilution with large quantities of river- 
water prevents a mechanically purified sewage from becoming 
foul. Where the ratio of river-water to sewage is small and the 
dilution does not guarantee that the purified sewage will not 
putrefy, biological treatment must be adopted, in which case land 
treatment is generally superior to artificial biological processes. 

I. MECHANICAL PURIFICATION OF SEWAGE. 

(i) Rakes, Screens, and Sieves. 

Apparatus of this nature serves to separate the undissolved 
substances over a certain size. Of first importance are the 
coarser floating substances such as paper, vegetable leaves, 
orange-peel, match-boxes, corks, etc., which are removed from 
sewage by this treatment. According to Fruhling, one under- 
stands by screens all the particular varieties of such devices for 
removing coarse suspended matter, such as bear the particular 
names of bar-screens, sieve-screens, grating-screens, etc. Bar- 
screens consist of bars set parallel to one another, grating-screens 
of wires crossing each other, and sieve-screens of a network of 
wires or of perforated metal plates. 



RAKES, SCREENS, AND SIEVES 



69 



There are both coarse and fine screens. Coarse screens are 
mostly iron rods about lo to 20 centimetres apart, obhquelj^ 
placed at an obtuse angle to the surface of the water. These 
only serve to keep back the coarsest suspensions, such as tin 
bo.xes, the larger pieces of wood, large pieces of entrails, large 
rags, and the like. 

Fine screens are generally stationary, or automatically moving 
apparatus. Stationary screens are mostly bar-screens, also 




Fig. 13. Uhlfelder's Revolving Screen. 



consisting of bars laid alongside one another, and placed at an 
obtuse angle to the surface of the water. The distance of the 
bars apart is, however, considerably less than with coarse screens, 
usually between 10 and 25 millimetres. The water flows against 
these bars and leaves there substances of larger size. Movable 
screens differ from the stationary type in that they are driven 
through the water and fish out the floating material. ' Of the best 
constructed and most frequently used movable bar-screens in 
municipal clarifying plants (Frankfort, Elberfeld) is the Uhlfelder 
Revolving Screen (Fig. 13). It is a circular revolving screen. 



70 



SEWAGE DISrOSAI. 



composed of five single screens, \\'liich rotate uniformly against 
the current of water. The coarser matter floating upon or 
suspended in the water is remo^'ed by the screens while in motion 
and is raised out of tlie water. An automatic brush follows 
which, pressing through the screen, brushes it outwards and casts 
the material on to a platform underneath. This is then tipped 
up by the motion of the screens and empties the contents on to 
a travelling platform which carries the material away. With such 




Fig. 14. KiEN Sievk. 



screens the work is purely automatic and the purification me- 
chanical. Other kinds of screens may be cleaned by hand. 

Among automatically working sieve-screens the Rien has 
frequently been employed. It is used to screen Dresden sewage, 
which is then run into the Elbe without further treatment. 
The Rien apparatus, as is shown in Figure 14, consists of a 
metal disc set obliquely in the water and composed of sheet- 
metal sieving or sie\'e plates. In the centre of the disc there is 
placed a truncated cone of the same material. The water flows 
against the disc, which is continually turning. As a consequence 
the suspended filth remains on the disc, whilst the water passes 



RAKES, SCREENS, AND SIEVES 



71 



through the openings. Owing to the motion of the disc, the 
sludge is removed from the water, and then scraped away from 
the screens, and from the truncated cone also, by means of 
brushes. 




Fig. 15. Kremer Apparatus. 



a. Inlet channel, b, Lateral inflow pipes, c. Space for 
attaining- an upward thrust in the water, d. Grease ex- 
tractor, e, Floating- layer rich in fat. f. Clarifying space. 
g. Reversal of the direction of the water at the edge of the 
circular partition, h. Circular overflow channel, an outlet 
for the clarified water. i. Sludge cylinder. k, Sludg-e 
drain. I, Cleansing pipe. 



SEWAGE DISrOSAL 



(ii) Grease Extractors. 



Many effluents, such as those from slaughter-houses, kitchens, 
and other similar places, contain a large amount of grease. 
Since the fats are lighter than water they cannot be separated 
in grit-chambers, settUng-basins, and other contrivances. They 
can, however, be removed from sewage in grease extractors, and 
in certain circumstances these allow of a recovery of the grease. 

There is a whole series of plants for the recovery of grease. 
The best known of these is the Kremer apparatus. This apparatus 




J 

Fig. i6. Grease Extractors of the "Stadtereinigung 
Firm " of Berlin-Wiesbaden. 



can also be used independently as a clarifying plant, since it also 
separates from the sewage any suspended matter that is heavier 
than water. The apparatus (Fig. 15) separates the undissolved 
substances by means of a peculiar motion of the stream working 
in such a way that the substances are divided into two layers of 
sludge, according to their specific gravity. Grease particles 
adhering to light organic matter form the upper floating layer, 
whilst the lower layer is composed of the heavy particles of 
suspended matter. According to Vogelsang, the Kremer ap- 



GRIT CHAMBERS 73 

paratus is to be recommended strongly for preliminary purifi- 
cation with the biological processes, since it separates much of 
the light suspended matter, such as grease, paper, straw, and the 
hke, which cannot be removed by sedimentation processes, and 
which consequently choke up the contact beds and percolating 
filters with mud. The " Gesellschaft fiir Abwasserklarung" (Sew- 
age Clarif5dng Co.), of Berlin-Schoneberg, supply this apparatus. 
The firm also employ them in combination with Emscher wells 
(see page 82 in this book) . They call the combination the Kremer- 
Faulbrunnen (Kremer septic well), and warmly commend its 
great utUity. 

A grease extractor of another type is that of the " Stadter- 
einigung und Ingenierbau A.-G." firm of Berlin- Wiesbaden. 
As shown in Figure 16, the greasy water enters through the 
funnel c into the space d, where by means of a deflector e it is 
uniformly divided and diverted in direction upwards. The 
particles of grease have time to separate out in the upper 
parts of d, whilst a quantity of water in the space d, corresponding 
to the inflow from above, passes out below through the circular 
opening F. The sludge sinks lower and falls eventually into the 
sludge-pail g, whilst the water, freed from grease and mud, rises 
to the brim of the reservoir and, overflowing at i, passes into the 
outlet pipe at z. 

Of other systems the grease extractors of Kaibel, Darmstadt, 
of Kremer-Schilling, and of Heyd, Darmstadt, may also be named. 

(iii) Grit Chambers. 

Many plants for preUminary purification also make use of 
grit chambers. Whilst screening removes the coarser floating 
matter, grit chambers are intended to separate the heavier and 
coarser sinking suspensions. This is attained by allowing the 
pipe conveying the water to the clarifjang plant to suddenly 
discharge its contents into a larger space known as the grit 
chamber. As a result, the velocity of the sewage, owing to the 
increase in cross section of the chamber, is quite considerably 
reduced. Consequently the coarse suspensions, like rags, bones, 
sand, and such-like, settle to the bottom. The cross section of grit 
chambers varies greatly. They may be funnel-shaped or rect- 
angular, or they may have arched bottoms, etc. The mud which 



74 SEWAGE DISPOSAL 

settles is removed either by hand or mechanically, and the 
removal may take place either after emptying or during the 
process. The mechanical sludge-remover generally consists of 
a dredging-machine, which dredges the sludge and casts it on to 
a travelling platform. With grit chambers built funnel-shaped, 
the dredger can remain standing in the one place. With other 
forms the dredger can often be moved from one end of the 
chamber to the other, as at Frankfort-on-Maine. According 
to the investigations of the author in Frankfort, with a good grit 
chamber and a system of fine screens, about one-fifth of the 
total suspension in a sewage is removed. 

(iv) Sedimentation Tanks, Wells, and Towers. 

These contrivances serve for the deposition of the finer sus- 
pended matter. The sewage is simultaneously introduced into 
a number of such tanks or wells, so that, owing to the increased 
cross section, the velocity of water is very considerably di- 
minished, and consequently the finer particles of suspended 
matter can settle to the bottom. 

Sedimentation tanks are generally elongated rectangular 
chambers, constructed either open or covered, though they are 
most frequently open. They are generally about 40 metres long, 
but there are, however, sedimentation tanks considerably shorter 
and longer. The inflow generally occurs through openings 
situated below the surface of the water, in order that the already 
precipitated sludge shall not be disturbed by the motion of the 
water. The beds of the tanks are built both rising and falling 
towards the outlet. According to the researches of Steuernagel 
and Grosze-Bohle, the rising bed has proved satisfactory for 
sedimentation work. Generally at either inlet or outlet, or even 
at both places, sump-pumps are placed to pump away the 
deposit which collects there. The Frankfort sedimentation tanks 
are 40-6 metres long and have pumps at 7-5 metres distance from 
the inlet and outlet. The chamber bottom is so constructed 
that in the middle it is raised and falls away with a fall of i in 10 
to the sumps. There is the same fall from both inlet and outlet. 
This arrangement of the chamber bottom serves to lead the mud 
of itself to the pumps, through which it is removed. This renders 
superfluous the cleaning of the tank bottoms with shovels or 



SEDIMENTATION TANKS, ETC. 



75 



similar contrivances. In Figure 17 the form of the Frankfort 
sedimentation tank is shown in cross section. 

It was formerly thought that to effect good sedimentation the 
chambers must be built so large that the velocity of the water 




Fig. 17. Frankfort Sedimentation Tank (Cross Section). 



would be extraordinarily small (2 to 4 mm. per second). Ex- 
periments of Grosze-Bohle in Cologne have shown, however, that 
with a velocity of 20 to 40 millimetres the same sedimentation 
is attained. Also in Frankfort it could be demonstrated that 



76 SEWAGE DISPOSAL 

the abnormally small velocit\- of 2 to 4 millimetres per second 
gave no better results than one of 12 millimetres per second. 
\Mth higher velocities than 12 millimetres per second considerable 
diminution of the sedimentation was obtained, so that one 
cannot generalise on the results from Cologne ^\^thout further 
details. According to Schiele, in England they still keep quite 
generally to the small \-elocity of 2 to 4 millimetres. 

In the settling tanks there also separate out a number of floating 
substances which are composed firstly of the fatty matter in the 
sewage, and secondly, especially when the tanks ha^"e been in 
use for a long period, of particles of sludge rising up on account 
of fermentation. In order to avoid admitting these floating 
particles into the outlet, a board is generally partially immersed 
in the water and set obliquely a short distance in front of the 
outlet. The particles collect in front of this. 

The mud is generally disposed of first b}- running off all the 
water. The sludge is then pimiped away from the slime deposit. 
In some sedimentation plants sludge-pumps working mider 
water might be tried ; these naturally would be ad\antageous 
in that the sludge could then be removed during the sedimenta- 
tion process without the chamber that is to be purified being 
emptied. In general, however, sludge-pumps do not appear to 
have worked well. The main difficulty lies in this, that the 
sludge does not slip along sufficiently towards the pumps ; on 
the contrary, a crater is often formed in the sludge and the 
sewage is suddenly sucked over instead of the sludge. 

The settling process is frequenth' assisted by the addition of 
chemicals. Alum and lime, or ferrous sulphate and lime, and 
other substances are used. The principle upon which the use 
of these substances is based is as follows : The chemicals form 
gelatinous and voluminous precipitates (alum -|- lime = gypsum 
and aluminium hydroxide), which on sinking to the bottom carry 
the suspended matter along with them. The chemicals ha\e 
also a slight action on the dissolved organic matter. This action, 
according to some, \\ill be insignificant in itself, the main action 
being the assistance which the chemicals render towards settling 
out the suspensions. With domestic sewage the action of 
chemicals is frequently denied. Thus, parallel experiments ha\"e 
been carried out by Lepsius and later by Freund in Frankfort- 
on-Maine with chemical precipitation and with purely mechanical 



SEDIMENTATION TANKS, ETC. 77 

sedimentation, by which it was estabUshed that the influence of 
chemicals on sedimentation was practically insignificant. Accord- 
ing to Schiele, however, in opposition to this view, the addition 
of chemicals to domestic sewage is still practised in England. 
In the author's opinion chemicals certainly have an effect on 
the dissolved substances, and also act in the case of domestic 
sewage on the pseudo-dissolved substances, the colloids, to no 
inconsiderable extent. Aluminium hydroxide and ferric hy- 
droxide are, as is well known, also colloids, and have an adsorbent 
action on the colloids in the sewage. Frequently, addition of 
chemicals to industrial effluents must be made. The main 
drawbacks to the addition of chemicals are the considerable 
increase in cost of the sedimentation process, and the difflcvilty 
of disposing of the sludge. The sludge obtained when chemicals 
are used is, in the first place, far more aqueous, and so takes up 
much more room, while secondly, on account of the sludge 
containing chemical substances, which at least are indifferent from 
the manurial point of view, it is considerably depreciated in value 
as a manure. The amount of chemicals added varies greatly 
according to the composition of the sewage to be purified. It 
varies between about 43 and 950 grammes per cubic metre (7 to 
150 oz. per 1000 gallons) of sewage. The chemicals are added 
either in the liquid or solid state. In the liquid form it is allowed 
to flow into the sewage in a thin stream. When applied in solid 
form the chemicals are added to the water in wire baskets, so 
that the water flowing through dissolves out the chemicals. For 
these to work properly it is of great importance that care be taken 
to get good mixing between chemicals and sewage. If there be 
sufficient space between the place of addition and the settling 
tanks, the water itself ensures adequate mixing. Now and again, 
however, various devices are employed to attain good mixture. 
Thus, tongues are placed in the sewage-pipe, by which means the 
water is broken up. 

The amount of suspended matter separated out, expressed as 
a percentage of the total content of suspended matter in the 
crude sewage, amounts, with good sedimentation tanks and not 
too high a velocity (12 mm.) in the basins, to 60 or 70 per cent. 

Sedimentation wells and towers are generally cylindrical in 
form and are distinguished essentially from tanks in that the 
water flows through them upwards and in a vertical direction^ 



78 SEWAGE DISPOSAL 

whereby the suspended matter settles out underneath. In such 
sedimentation reservoirs the velocity of the sewage is generally 
considerably less than in sedimentation tanks. With towers and 
wells the sludge can be removed more easily while the process is 
in operation than is possible with tanks, since it is in such cases 
more conveniently concentrated into a certain space. To guard 
against the water breaking through (see page 76), various con- 
structions have been proposed, based on the closing of the water 
space whilst the sludge is being pumped off. 

(v) Septic Tanks. 

Septic tanks are only employed as preliminary purifiers in 
biological purification plants ; they do not come into considera- 
tion as separate clarifying plants. They are employed in a 
manner similar to the usual sedimentation tanks, but with the 
difference that the sludge which separates out is allowed to lie 
for a long period. It is thereby converted into a foul state of 
fermentation, which is then communicated to the sewage. The 
sewage is coloured black, owing to the presence of ferric sulphide, 
and assumes a very foul odour. As a result of the activity of 
the fungi setting up putrefaction a portion of the organic matter 
is decomposed. There forms on the septic tank after a short 
time a scum consisting of grease mixed with particles of sludge 
raised by the fermentation. This scum shuts off the air from the 
sewage, and is of great importance in the putrefactive action 
which is caused by anaerobic bacteria. In some English septic- 
tank installations the floating layer, according to Schiele, now 
and then becomes so strong that one can actually walk upon it. 

Moreover, these septic tanks act like sedimentation tanks as 
regards removal of suspended matter. With the slow passage of 
the water through the septic tanks the suspensions settle out. 
A certain amount of time always elapses before a septic tank is 
in good working order, that is, until it yields a thoroughly 
putrefied effluent. Septic tanks are built both open and covered. 
The open tanks have the drawback that they cause annoyance 
from their smell. Septic tanks, according to Schiele, must be 
capable of holding at least half a day's dry-weather drainage. 
Smaller plants are generally built large enough to hold from 
three to six times the dry-weather discharge. 



TRAVIS AND EMSCHER ^\TSLLS 79 

Naturally the removal of sludge takes place less frequently in 
this case than with the usual settling tanks. It is useful never 
to remove the whole of the sludge, but always to leave some stUl 
l5'ing ; the tank then gets into working order afresh more quickly, 
because the putrefactive fungi from the thoroughly putrefied 
sludge can more easUy communicate themselves to the sewage. 

The main advantages of septic tanks may be thus noted. In 
the first place, the sludge need not be removed so often, which 
naturally means cheaper working ; and secondly, the sludge from 
septic tanks, especially according to the results of recent work, 
is by no means so disagreeable as the sludge from the usual 
sedimentation tanks. It is more easily freed from water and 
contains a considerably larger bulk of dry substances. These 
relations will be more carefully dealt with in the chapter on 
sludge. As a further advantage of septic tanks, there has to be 
considered the decomposition of a portion of the organic sub- 
stances which, from the observations of Dunbar and his pupils, 
takes place both in the sewage and the sludge. 

With septic tanks there is the disadvantage that the eflfluent 
is not so completely freed from suspended matter as with sedi- 
mentation tanks. Owing to the fermentation set up in the sludge, 
fine particles are propagated upwards and so pass into the dis- 
charge. Since suspended matter vitiates bacterial action in 
contact beds and percolating filters to a considerable degree, and 
also causes their more rapid pollution, sieves are often placed in 
the discharge-pipe to check this annoying feature. 

(vi) Travis and Emscher Wells. 

These apparatus are distinguished from sedimentation and 
septic tanks by the separation of the sludge from the actual 
sedimentation space. 

According to Collins (" Surveyor," 1909), the Travis plant in 
Norwich consists of rectangular tanks, whose lower parts are 
wedge-shaped in cross section (see Fig. 19). Lengthways in each 
tank a roof is built, which is cut open at the coping in the long 
axis of the tank and is carried upwards above such axis. Where 
the roof is joined to the tank-walls and to the coping, gaps are 
left. Through these gaps the three compartments of the tank, 
arising out of the mode of construction, are in communication 



80 



SEWAGE DISPOSAL 



with one another. Both the outer compartments serve as sedi- 
mentation compartments. The sludge on settHng out slides 
down the sloping surface formed by the roof and falls through 
the gaps into the middle section of the tank, the reduction or 
sludge chamber. In the central portion of the settling tank there 
are placed grating-like frames (colloiders) about lo inches from 
each other. These devices should serve to retain and to exert a 
certain coagulating influence on the colloids ; 80 per cent of the 
sewage enters through the two outer settling tanks, whilst 20 per 
cent enters the reduction chamber or sludge area through the 
openings at the coping of the roof. The water passes out over 




Fig. iS. Clearing Well Installation at Norwich on the Travis 
Hydrolytic System, with Colloider hung in. 

(From " Wasser und Abwasser," 1909-10, II, 71.) 



a weir which occupies the whole breadth of the tank, and which, 
corresponding to the three tanks, is three-sided. The effluent 
from the liquefying chamber is subsequently purified in a special 
tank provided with " colloiders " like the sedimentation tank. 
The removal of sludge, necessary from time to time, is con- 
veniently attained by having the bed of the sludge chamber 
composed of a number of funnels. The walls of these funnels are 
so inclined that, on opening the outlet situated at the lowest 
point of the shoulder, the sludge is pressed out by means of the 
pressure of water above. 

Travis tanks have come into use considerably, especially in 
England, and presumably have proved excellent. The main 
difference between the Travis tank and Emscher wells consists 



Plan 




^ITi'lM'I'IMM'I'l'!/ ^ 






'•\'Z'\-y'--'\ .■ff/udoe Space':]:r'.\^ ■■'■''■; —m 



[^:-:,'r;:i' ''■ 




Fig, 19. Emscher Wells. 

Arrangement of several wells with settlingf channel in common. Re- 
moval of sludge by excess water pressure to the lower lying sludge 
storage grounds. 

A, Inlet. S, Screening chamber, e, Channel for screenings «, Cir- 
culating channel. P, Outlet. D, Fume Chamber. G, Partitions. 
u. Waste-weir. i^, and F2, Liquefying chambers. J?, Sludge pipe. 
S, Cleansing pipe. 

G 



82 SEWAGE DISPOSAL 

in this, that with the former from time to time fresh water is 
purposely led through the liquefying chamber in order to remove 
the liquid constituents of the sludge generated by the putre- 
faction. With Emscher wells the principle is to avoid any 
flushing of the liquefying chamber. 

Emscher wells (Fig. 19) consist in the main, according to 
Middledorf , of deep wells which are intended for the reception of 
the sludge. In the upper part of these wells by means of a partition- 
wall a settling tank or well is apportioned off, and through this the 
water flows. As soon as it touches the sloping bottom the sludge, 
settling out in the tank, flows of itself through gaps at the lowest 
parts of the sedimentation space into the sludge wells. The 
water flows only through the sedimentation area and, in agree- 
ment with the principle, not through the putrefactive area. The 
putrefaction is thereby confined to the sludge alone, and the 
water flowing away is obtained as fresh as possible, unmixed with 
any of the polluted water. The amount of water from the 
sludge which gains access to the sewage should amount to about 
one part in a thousand. With variation of temperature also there 
should be little tendency for water from the sludge to rise into 
the sedimentation tank, since it is specifically heavier than fresh 
sewage. As the sludge from the settling tanks is withdrawn 
uninterruptedly and automatically, there is absolute certainty 
that it is always removed at the right time and does not remain 
lying in the settling tanks, injuring the clarifying process by its 
foulness. The particles of mud propagated upwards by fer- 
mentation cannot pass through the clefts in the partition-wall, 
but are forced against the wall underneath the sewage and sink 
again from there to the bottom. The sedimentation in Emscher 
wells should be equally as good as with other good settling plants 
with the same period of sedimentation. The sludge putrefies in 
the liquefying chamber, and thereby becomes more suitable for 
disposal (see under Sludge). The fact that water is not allowed 
to flow through the chamber has not proved a hindrance to the 
putrefaction of the sludge. It is because of this that the freshness 
of the water to be purified is maintained, and it may be reckoned 
an advantage in the process. With Emscher wells, in contra- 
distinction to other liquefying chambers through which the water 
is allowed to flow, scarcely any trace of sulphuretted hydrogen is 
evolved. According to the investigations of Spillner, the gases 



TRAVIS AND EMSCHER WELLS 83 

which escape from the Hquefying chamber and pass both through 
the sewage and around the sedimentation tanks consist mainly 
of methane and carbon dioxide. Odourlessness of the sludge 
should be attained equally well with purely domestic sewage 
purifiers, and with plants receiving industrial effluents. The 
sludge is led through pipes from the liquef jdng chamber to the 
sludge drying-place. The end of the pipe reaches to the lowest 
point of the funnel-shaped bed of the well. Only the under- 
neath portion of the most thoroughly putrefied sludge is led away. 
If the area for drying the sludge can be situated from 1-5 to 2 
metres lower than the water-level in the wells the sludge can 
then be removed by the natural fall, since for this purpose a head 
of I metre is generally sufflcient. If this head cannot be obtained 
the sludge is removed either by a Wagner suction apparatus, 
with a hand or air-pressure pump, or with a vacuum plant 
driven by an air-pressure pump. In all plants of the Emscher 
Co. the sludge should be displaced through the sludge-pipes 
without any difficulty. The reason for this is largely because the 
sludge during the septic action loses its felt-like nature and 
forms a black, pulpy, fluid mass, in spite of its low per- 
centage of about 70 to 80 per cent water. The removal of the 
sludge takes place perhaps once every two or three months. The 
liquefying chamber should be so large that it can hold the sludge 
falling in during this length of time. 

Meanwhile, the question of Emscher wells has been vigorously 
debated in the literature. In the main the following points have 
been contested : Whether the plant is practically odourless ; 
whether the sedimentation is as good in Emscher wells as in 
other good sedimentation plants ; whether liquefaction actually 
occurs in the sludge ; whether the water remains at rest in the 
liquefying chamber, or whether rather, just so much of the 
polluted water from the liquefying chamber enters the sedi- 
mentation tank as sludge separates out. 

It must be pointed out, however, that Emscher wells are com- 
ing more than ever into application, both as separate sedimenta- 
tion plants and also as plants for the preliminary purification of 
sewage about to be bacterially treated. 

The main advantage of Travis and Emscher wells over the 
usual sedimentation tanks is therefore the simplicity with which 
the sludge problem is solved. 



84 SEWAGE DISPOSAL 

" There is this great advantage in a joint plant working with 
automatic separation of sludge. A small velocity of sedimenta- 
tion can be chosen without the process becoming more difificult 
as regards the handling and disposal of large amounts of thin 
sludge containing high percentages of water, such as generally 
settle out under such conditions " (Schmidtmann, Thumm, and 
Reichle) . 

. II. DEGENER'S COAL-PULP PROCESS. 

As regards the nature of its action, this method stands midway 
between the mechanical and the bacterial methods, for like the 
former its main action depends on the removal of the suspended 
matter, and like the latter it has an undoubted action on the 
dissolved substances. 

I to 2 kilograms of ground-peat or 2-5 to 4 kilograms of turf 
are added to a cubic metre of the sewage to be purified. Chemicals, 
such as aluminium or ferric sulphates, are simultaneously added. 
Generally, the sewage is then allowed to settle in sedimentation 
towers. The sludge is fairly aqueous, but is usually freed from 
water in filter-presses, and in this case is then no longer capable 
of putrefaction. The air-dried sludge, on account of its high 
carbon content, is a valuable burning material. Lately it has been 
rendered more valuable by gasification (see under Sludge) . 

The whole process should work without smell, and on this 
account, and because the troublesome sludge question with its 
dangers is obviated, it is a method of great importance. The 
further advantage of a reduction in the dissolved organic sub- 
stances has been referred to previously. Against these advan- 
tages there is the main disadvantage of cost, which is very 
considerable. It is because of this that the number of plants 
working the process on a large scale remains so limited, in spite 
of the method being so superior. 

III. BACTERIAL PURIFICATION OF SEWAGE. 

As already mentioned, the dissolved organic matter is un- 
changed in a mechanically purified sewage. If such an effiuent 
be allowed to stand for some time it very soon begins to putrefy 
and, from the plentiful development of putrefactive bacteria, 
assumes a stinking condition. 



ARTIFICIAL BIOLOGICAL METHOD 85 

The task of bacterial sewage purification is so far to remove 
the dissolved organic matter from the sewage that putrefaction, 
with its accompan5dng disagreeable phenomena, is avoided. As 
the name indicates, in bacterial purification processes bacterial 
agencies come into play. 

In particular two processes are distinguishable, viz. artificial 
biological sewage treatment, and broad irrigation or sewage 
farming. 

(i) The Artificial Biological Method. 

Artificial biological sewage purification is the designation 
applied to that treatment whereby, after suitable preliminary 
purification by mechanical methods as detailed in Part I, the 
sewage is led over large pieces of slag, coke, or similar materials. 
The organic matter is by this means so far removed from the 
sewage as to prevent subsequent putrefaction in the effluent. 

Biological processes for the treatment of sewage arose in 
England, where they were first used and tested on a large scale. 
That this method was first improved in England is quite reason- 
able, since England is the classic country for sewage disposal. 
In consequence of the peculiar circumstances operating there, 
its position as the premier country in the world, England was 
compelled to work out good methods of sewage purification. 
The largest English rivers have not the capacity and velocity 
even of our Spree or Havel (tributaries of the Elbe, Trans.). With 
industry flourishing in the middle of the last century there went 
hand in hand alarming pollution of the whole river system of 
England. 

Thorough preliminary purification is of vital importance for 
the continued good working of biological processes. Grit cham- 
bers, screening, and the mechanical purification involved in 
sedimentation and septic tanks, all come into consideration. 
For domestic effluents the sewage is generally first purified in 
sedimentation or septic tanks. If much industrial effluent be 
admixed with the household sewage, chemical treatment is 
generally indispensable. Moreover, it is advisable to find out in 
each case, by special experiments, the most suitable preliminary 
treatment for the sewage. 

Many substances are used as material for the beds employed 
in these biological processes. Clinkers from refuse destructors, 



86 SEWAGE DISPOSAL 

slag, earthenware, bricks, coke, pieces of slate, and such-like 
materials have all been tried. According to Schiele, the main 
requirements in such a material are great hardness, capacity to 
withstand the action of sewage, roughness, and firmness. 

Large beds are laid down composed of lumps of these materials 
loosely collected together. Distinction must be made between 
contact beds and percolating filters. The distinction between 
the two lies in the manner of their impregnation with sewage. 
Contact beds are filled with sewage ; the sewage remains for a 
period of time lying in the beds and is then led away again, by 
which means air is sucked into the beds. With percolating 
filters the sewage is continually percolating through the beds. 
The size of the pieces of material used should be about 3 to 8 
millimetres (j^th to ^rd in.) in the case of contact beds, whilst 
with percolating filters it is considerably larger, amounting to 15 
to 75 miUimetres (^ to 3 in.), and there are even larger in use 
(Schiele). With contact beds there are two processes, known 
as single-contact and double-contact treatment. In the single- 
contact process treatment is complete after leaving the bed. 
In the double-contact process the sewage from the primary bed 
is delivered to a secondary bed. The material of a primary bed 
is coarser in grain than that of the secondary bed. Percolating 
filters are generally worked with single contact. 

As with septic tanks, so here the beds have to be prepared for 
the work in order that in course of time they cease to yield 
effluents capable of putrefaction. The time for this process to 
take place may amount to some months. It is generally shorter 
with percolating filters than with contact beds, and depends on 
the gradual formation of a slimy layer on each piece in the bed, a 
discussion on the nature of which layer will be detailed below (see 
page 8g). Contact beds are so worked that they are filled with 
sewage, which is then allowed to stand some time in contact 
with the beds, and then to drain away. The sewage is generally 
led on to the contact beds from above, and is drawn off from under- 
neath. After each impregnation contact beds must be allowed 
to stand some time, in order to work up the substances taken 
from the sewage. In England good results have been obtained 
with the beds full for two hours and standing empty four hours 
(Schiele). In practice single-contact beds can be impregnated 
twice per day, double-contact beds three times daily. According 



ARTIFICIAL BIOLOGICAL METHOD 87 

to this, I cubic metre of single-contact bed can be treated with, 
at most, 0-66 cubic metre of sewage, double contact with 0-5 
cubic metre per day. 

Percolating filters can be continuously impregnated with 
sewage without the purifying action falling off. 

The beds are either built in the ground or on the groimd. It is 
important for good working that they should be maintained at 
uniform temperature. They should therefore be protected under 
certain circumstances against cold in winter. 

Contact beds are generally rectangular ; percolating filters are 
either rectangular, octagonal, or circular in shape. 

In the case of contact beds the water is distributed on to the 
beds in a very simple manner. It takes place through feed 
channels, perforated clay, or stoneware pipes, open, perforated 
drains, perforated gutters, and other devices. With contact beds 
good uniform distribution is not by any means so important as 
with percolating filters. In the case of contact beds the main 
concern is the quickest possible filling. They are generally filled 
and emptied by sluices and valves worked by hand. Frequently 
there are also automatic devices for filling and emptying. Uni- 
form distribution is, however, a matter of very special importance 
in the case of percolating filters. The number of devices invented 
to attain this is almost legion. There are sprinklers, movable 
revolving sprinklers, stationary perforated pipes, portable 
sprinklers, etc. 

In general, according to Schiele, with regard to distributors 
for percolating filters, the following can be said : The system 
employed must always be adapted to local circumstances. It is 
most suitable to carry out experiments beforehand. In Birming- 
ham the following requirements were demanded of a good 
distributor : — 

(i) Uniform distribution, so that each portion of the surface 
of the bed contains precisely the same quantity of water. 

(2) Distribution in the form of drops. 

(3) Absolute control over the distributor. 

(4) Limited cost of plant. 

(5) Small working expenses. 

(6) As few movable parts as possible. 

-(7) Small consumption of power in distribution. 

With regard to the water-pressure necessary for working, 



88 SEWAGE DISPOSAL 

fixed sprinklers are at a disadvantage when compared with 
the automatic revolving sprinklers. The revolving sprinklers 
are found in English experience to be limited to a bed of 
30 metres diameter, or at the most 35 metres. When they 
are to be worked mechanically, large rectangular beds with 
movable or fixed sprinklers prove more useful and advan- 
tageous than circular beds with revolving sprinklers. In 
Germany automatic revolving pipe sprinklers, with a bed 20 
metres diameter and perforations every 10 millimetres, have 
proved the best for withstanding strong cold in winter. In 
England at the present time stationary sprinklers, especially 
the Fiddian type, are being increasingly employed. 

Contact beds and percolating filters are on the whole to be 
regarded as of equal value. Both have their advantages and 
disadvantages. The advantages of contact beds are : Simple 
distribution of the sewage on to the beds, small trouble due to 
smell, no plague of flies, little suspended matter in the discharge ; 
small need for a head of sewage on account of the small depth of 
the bed, greater certainty of working in the cold weather. As 
opposed to these the main advantages of percolating filters are : 
Greater mechanical power than with contact beds ; since the 
beds can be built higher smaller space is necessary ; coarse- 
grained, cheap material ; little attention. Whilst contact beds 
are always gradually getting choked with sludge, and on this 
account need purifying more frequently, percolating filters do 
not get choked with sludge, or at most with very little. Per- 
colating filters also withstand an overload of sewage in rainy 
weather, since the admittance of air is never quite cut off. On 
the contrary, percolating filters always contain a great deal of sus- 
pended matter, and the discharge must be subsequently freed from 
this. In percolating filters small flies establish themselves in 
extraordinary numbers. Especially is the Psychoda (the butterfly 
gnat) to be noted. Further, smells are never quite avoided in the 
treatment of sewage with percolating filters. After the invention 
of percolating filters one often heard the opinion expressed that 
contact beds would soon disappear. This view has not proved 
correct. For the purification of sewage in large towns it is true 
that percolating filters have proved themselves superior to con- 
tact beds, since all the above-named advantages of percolating 
filters mean cheaper cost of construction and of upkeep. For 



ARTIFICIAL BIOLOGICAL METHOD 89 

the centralised clarification plants of towns, therefore, percolating 
filters are nowadays generally erected. Still, even in the most 
recent years in England, in large towns like Manchester, contact 
beds have been put down and have proved excellent. They are 
especially suitable where the proximity of places of residence, as, 
for example, with small domestic or manufacturing plants, 
compels one to avoid smell or flies. 

As has been already mentioned, the percolating filter effluents 
always contain large amounts of suspended matter which are 
washed out of the beds. These substances differ greatly from 
those in the raw sewage, in that they are not capable of putre- 
faction. They impart to the purified water an objectionable 
appearance, and would lead to the deposition of sludge in the 
river ; consequently percolating filter effluents are always sub- 
sequently clarified. In England this is frequently effected by 
land irrigation (i acre to looo inhabitants). Generally, 
however, the discharges are subsequently treated in simple 
sedimentation tanks or wells. Since the sludge is no longer 
capable of putrefaction, it is easily freed from water and its 
disposal causes no difficulty. 

The cost of sewage treatment by the biological process varies 
greatly. The price of material for the beds varies, accord- 
ing to Schiele, from 3 to 12 marks per cubic metre (2s. 6d. to 
los. per cubic yard), whilst i cubic yard of ready - prepared 
material costs from 5 to 35 shillings. The cost of the bed in 
England cannot, as a rule, be less than 15 shiUings per cubic yard. 

As to the nature of biological purification of sewage there exist 
two views. According to one, which is put forward by Baurat- 
Brettschneider of Charlottenburg, and which in its essentials 
coincides with the so-called Hampton doctrine championed by 
Travis, the whole so-called biological sewage treatment consists 
of nothing more than a filtration of the organic filth, which, in 
the main, would be present in the colloidal form in sewage 
(together with the greater part of the so-called dissolved sub- 
stances) . This theory is opposed to that of Dunbar and his pupils, 
according to whom biological sewage purification is to be explained 
in the following manner : Each single portion of the bed is 
gradually covered with a slimy surface of bacteria and other 
organisms. During impregnation the greater part of the organic 
matter is adsorbed by this covering. The film also adsorbs 



90 SEWAGE DISPOSAL 

oxygen in large quantities. With the help of this oxygen and of 
the oxygen otherwise entering the bed, the adsorbed substances 
are decomposed by the organisms in the intervals of rest between 
each impregnation. This theory is supported by numerous 
experimental data, so that there can be no doubt as to its truth. 

(ii) Land Treatment. 

Treatment of sewage on the land is the oldest form of sewage 
treatment. Within the memory of man it has always been known 
that the surface of the earth was able to remove the evil smell of 
sewage and to make polluted water pure. 

Two methods of land treatment must be distinguished, namely, 
broad irrigation or sewage farming, and the intermittent sand 
filtration of Frankland. 

(a) Sewage Farming. 

Broad irrigation or sewage farming was first practised in 
England. To-day there is still in existence an irrigation bed that 
has been in use two hundred years. As is well known, sewage 
contains a number of substances which are valuable as nutri- 
ment for plants (nitrogen, phosphates, potash). The irrigation 
process makes use of this fact, and on the farms all kinds of useful 
plants are grown. Thus by this method of sewage disposal, the 
sewage is purified and use is made of the nutritious substances 
present in it. 

In England alluvial soil is considered to be the best for sewage 
farming. Soil lying over high waters generally in river valleys, 
or sandy loam over gravel or gravelly sand, or even gravelly, 
sandy subsoil with light or medium mainsoil about 40 centimetres 
deep, are all good. Gravel without a finer covering layer serves 
also as regards permeability (Schiele). The poorest earth is clay, 
loam, clayey soil and turf. Clayey soil, since it is not permeable, 
can only be used for the so-called surface or rude irrigation. 

According to Dunbar, there are two main kinds of irrigation 
processes. In the first kind the water flows down from the highest 
point of the land. After trickling over the surface of one field 
it is collected into a ditch, from which it is again uniformly dis- 
tributed over the next field. With a sharply inclined tract of 
land, dams must be thrown up to retard the flow of the sewage 



LAND TREATMENT 91 

and to distribute it afresh. When the track is not so steep the 
operation is carried out in the following manner: The sewage 
flows from the distributors, passing transversely from the irriga- 
tion plant into smaller trenches arranged perpendicularly to the 
larger. These are dammed up at the ends so that the sewage 
must overflow from the trenches at the sides. It then flows over 
the sloping surfaces of meadow into lower l3ang ditches. Then 
from these second distributing ditches, in precisely the same 
manner, the water is again distributed over a second series of 
meadows. This process is generally used for surface irrigation 
only, and in those cases where the land lies so low that it cannot 
be drained. A good drainage is not usual, therefore, with the plant. 

In Germany and also in France, as far as possible, it is en- 
deavoured to lay out sewage farms on the principle known as bed 
irrigation. In this process surface treatment with sewage is done 
away with, as the water passes through the grovmd. It is therefore 
true irrigation. The sewage is led on to the farm in ditches. 
The distributing canals are only allowed to fill so far with sewage 
that this has to pass into the bed sideways and below the surface. 
Wetting the stems and leaves of the plants is thus avoided. The 
beds are usually only i metre broad by 20 to 40 metres long, as 
otherwise uniform distribution of the sewage cannot be attained. 
In this process many distributing ditches are required, and also 
paths from which the beds can be attended to. This means, 
therefore, considerable loss in working space. 

Fvurther, a process may also be briefly described here which, 
up to the present, has not been widely employed, but which 
might render good service in those places where no suitable land 
for sewage farming is to be had in the neighbourhood of towns 
and manufactories. The method consists in conducting the 
sewage to farms, and here by means of hoses it is squirted on to 
the fields. In the year 1897 it was first used by Nobel in Eduards- 
feld, near Posen, ^\'ith 5^ million gallons of sewage yearly from 
the town of Posen. The process is named after the inventor, and 
is known as " Benobelimg " (The Nobel Treatment), or the 
Eduardsfeld process, and also as hose irrigation. 

In most cases the subsoil is drained. The purified sewage is 
conveyed to the river in collecting ditches. Under loam or clayey 
soils drainage is naturally not effected, since it would serve no 
purpose. The 
…[truncated]