William Phillips Dunbar.

Principles of sewage treatment online

. (page 18 of 36)
Online LibraryWilliam Phillips DunbarPrinciples of sewage treatment → online text (page 18 of 36)
Font size
QR-code for this ebook

and as much as 170,000 gallons of sewage was treated per acre per
day, or, after the introduction of artificial aeration, as much as
340,000 to 420,000 gallons. These experiments were repeated in
England with results which must now be considered as epoch-making,
since they gave rise to the so-called artificial biological processes. I have
attempted to draw a distinction between natural and artificial biological
methods by restricting the use of the word artificial to such filters as were
constructed not of naturally occurring soil, but which were built up
artificially. If this distinction is to be maintained, we must regard the
Massachusetts experiments, and even earlier English experiments, as the
starting-point of the development of these methods. It appears to me,
however, that a second characteristic feature should be noted. At.
Lawrence the material employed in the construction of the filters gradually
became so coarse that the usual method of applying the sewage failed.
The sewage did not distribute itself over the surface of the filter, but
simply passed downwards through the large pores of the gravel without
being purified. In order to obtain an even distribution, a layer of loamy
soil was first placed on the surface of the coarse gravel, but this did not

Digitized by



profluce a BatiBfactory result. Next, an automatic syphon was constructed
to discharge sewage on to the filter at regular intervals of twenty or thirty
minutes, and in this manner it was possible to treat larger volumes of
sewage than had hitherto been possible by biological processes. The
reports on these experiments attracted the attention of Sir Alexander
Binnie, at that time Chief Engineer to the London County Council, and
of J. Corbett, the Borough Surveyor of Salford. The experiments which
these two gentlemen commenced, independently of one another, have led
to a complete revolution in the domain of sewage purification. Quito
independently, Stoddart carried out a series of interesting experiments, to
which I shall again refer later.

The historical course of events has been variously described from time
to time. Some time afterwards Binnie declared that he had for years been
seeking a more rational method of disposing of the 1 70 million gallons of
sewage which London produces daily, and from which he obtained about
two million tons of sludge per annum. . The appearance of the
Massachusetts Report awakened in him hopes that he might be able to
achieve this object in the manner described in the Report, and he
accordingly commenced the experiments, which were carried out, under
his direction, by the engineer Santo Crimp and the chemist Dibdin
The position of Corlnjtt in Salford was worse than that of Binnie in London,
for the Manchester Ship Canal furnished the only means of disposing of
the sewage of a rapidly growing industrial population, and there was
no possibility of adopting either irrigation or land filtration. Corbett
intended his filters to deal with the largest possible vohune of sewage,
and regarded thorough aeration as a necessity. Hence, following the
Massachusetts example, the outlets from his filters were left open. All
his experiments were directed towards distributing the sewage over the
surface of the filters in the form of a rain. He first attempted to achieve
this by means of troughs placed a few feet above the surface ; he also
constructed filters in layers, with ventilating spaces between, so that the
sewage fell from one layer to the next like a shower of rain, (iradually
developing the technical details of d'stribution, in 1894 he tried experi-
ments with rotating sprinklers, and in the same year adopted fixed spray
jets. When Corbet! 's experiments with spray jets became known, most
experts shook their heads and regarded these fountain-like distributors as
expensive and worthless toys. At the present time some of the most
experienced engineers regard Corbett's method of distribution, which haa
in the meantime been further perfected by others, as the best, and as the
one among all the other methods which is specially suitable for adoption by
large towns. It affords me great satisfaction to place Corbett's experiments
in their proper light here, especially as he has been too modest to come
forward with any claim to priority.

The London experiments and the technical details developing out of
these have, during the last fifteen years, alone been in the foreground,

Digitized by



and it has generally been assumed that the experiments conducted at the
Barking works formed the basis of artificial biological methods. So
long as interest was centred in the contact process, to be described later,
and in so far as those experiments were a direct continuation of those of
Frankland and the Massachusetts Board, this assumption is warranted.
Since, however, for a considerable number of years the continuous method
of working biological filters has been more and more in the foreground,
it must be acknowledged that the development of technical detail is due
to the interpretation which Corbett placed upon the Massachusetts

Based directly upon Frankland's experiments, and not upon those of
the Massachusetts Board, Stoddart appears, as we shall see later, to have
had ideas upon the subject at an earlier date, but these did not attain to
the same practical importance as those of Corbett. The experiments of
the latter led to the development of the continuous process, whilst the
London experiments form the basis of contact methods.

Contact Beds.

Barking. — In following the practical development of artificial biological
processes, it is convenient first to give a description of the London experi-
ments. In June 1890 four wooden tanks were constructed, each having
a superficial area of ^^^acre. The first of these was filled with burnt
clay, the pieces being about the size of peas ; the second contained stones
of the same size ; the third coke breeze ; and the fourth was filled with
sand and gravel of varying size. The fourth filter was used to give the
sewage a preliminary treatment before passing it on to a polarite filter.
The effluent pipes from the tanks were conducted so high that the filters
could be filled to the surface with sewage and kept full whilst the sewage
was flowing through. Sewage was passed through for eight hours each
day, and then the effluent pipe was lowered and the filters emptied.
During three months working in this manner, the filters produced a non-
putrescible effluent when dosed with previously precipitated London
sewage. This arrangement was at that time regarded as a direct imitation
of that adopted in Miissachusetts. Santo Crimp has declared that he had
been considering for some time the possibility of imitating the Massachusette
method of feeding filters automatically by means of syphons on a large
scale, and had come to the above described arrangement. In Massachusetts
the feed was discontinued every twenty or thirty minutes, and hence the
pores of the filter were never completely filled with sewage. In London,
however, the filters were completely full of sewage for eight hours, and the
results were by no means ecjual to those produced in Massachusetts. It
was, however, considered quite sufficient for the requirements of Ix)ndon
if the effluents did not putrefy nor the filters become clogged.

In November 1892 it was decided to construct a similar filter, one acre

Digitized by



in area and filled tx) a depth of 3 feet with coke breeze, having a 3-inGh
layer of gravel on the top in order to keep the coke down. When operated
in the same manner as the smaller filters, this large filter did not produce
the required results. The filter jsoon became clogged, and after six weeks'
operation the effluent was putrescible. After twelve weeks the filter would
scarcely allow the sewage to pass through, although it had only received
chemically precipitated sewage. Finally, the sewage stood six inches deep
on the surface of the filter.

The official report by Dibdin states that it had been learnt that the
filter should be worked intermittently, and should at first be worked at a
slower rate than had been the case. The filter, which had been worked
to death, was allowed to stand for three and a half months. After throe
months the putrescent odour of the filter began to disappear. From
November 17, 1894, the filter was filled and allowed to stand full for two
hours, after which it was emptied. In this manner excellent results were
produced. The capability of the filter was so promising that Dibdin was
able to recommeud the town of Sutton to undertake experiments with
sewage which had not previously been chemically treated.

Sutton.— Preliminary experiments with clarified sewage carried on at
Sutton during 1894 had yielded results so satisfactory that Dibdin's
recommendations were carried out. Biological filters were constructed of
burnt clay and fed with sewage from which only the coarser suspended
solids had been removed by screening. The filters were operated in the
same manner as the London coke-breeze filter, by alternately filling and
emptying, a method which is now known as the contact method. At
Sutton 750,000 gallons of sewage were treated per acre daily by this
process, so as to yield a non-putrescible effluent. The oxygen absorbed
was reduced by 86*5 per cent. The effluent from this filter was submitted
to a second treatment in a similarly constructed and operated filter
containing coke breeze. The effluent from this second filter was clear,
colourless, and without smell. It was calculated that during seventy-six
days, 77 tons of sludge had been retained by the biological filter, and it
was believed that this sludge would be decomposed, Le, liquefied and
gasified. At any rate the results were so promising that the town of
Sutton abandoned the methods of chemical precipitation and irrigation
and definitely adopted the new process. Similar experiments were soon
commenced at Manchester, Leeds, and other English towns, with the
result that the favourable observations made at London and Sutton were

The results obtained in the English experiments were a decided advance
upon those obtained in Massachusetts, so far as the quantities of sewage
treated per acre were concerned. In Massachusetts the practical result
of intcnnittent filtration had generally been the treatment of a layer of
sewage 3 cm. deep daily; in the Lawrence experiments 116 cm. had been
treated daily, or 40 cm. after preliminary chemical treatment. In Sutton

Digitized by



a filter only 70 cm. deep had treated a layer of crude sewage 86 cm. deep
daily, and had converted it into a non-putrescible effluent. Later, experi-
ments at Leeds gave 45 cm. for crude sewage, over 80 cm. for the effluent
from septic tanks, and as much as 130 cm. for chemically precipitated
sewage; whilst at Manchester, after treatment in septic tanks, a layer
62*5 cm. deep was efficiently treated daily.

These results at once attracted attention in the widest circles. They
were discussed not only in technical journals, but also in the daily press.
In England the time was ripe for the adoption of such methods The
English rivers were scarcely any better than a quarter of a century before,
in spite of the steps which had been taken by the authorities. The sewage
problem was at a standstill ; no advance was being made with irrigation,
and chemical methods had proved unsuccessful.

Naturally, the heart of every tow^n councillor leapt with joy when he
was assured by experts holding responsible positions that the sewage
problem was solved ; that expensive palliatives, like chemical treatment,
upon which even small towns like Sutton, with 18,000 inhabitants,
were spending £1000 or more annually, could be done away with;
and that a method had been discovered which would cost practically
nothing, the adoption of which, for even only half of the sewage in
England, would mean an annual saving of at least sixty million pounds

Hamburg Experiments. — In the spring of 1897, when such wonderful
news was being circulated, the Hamburg authorities had to deal with a
most difficult problem in connection with sewage disposal. The large
Alster Lake has a world-wide reputation for its beauty, but at this date
its purity and, since it lies in the heart of the town, its very existence
were threatened by settlements which were taking place with increasing
rapidity in its gathering ground. These settlements were largely outside
the Hamburg boundaries, in another of the Federal States, and were
gradually becoming town-like in character and adopting the water-carriage
system of sewerage. The authorities were sanctioning the discharge of
the sewage, after simple chemical treatment, into the tributaries of the
Alster. If these schemes had been carried out, Hamburg would have
been forced to fill ih the Alster within a few years. I cite this example,
not because I consider it the most important, but in order to show that
at that time there was a serious gap in the practice of sewage purification
which could not be bridged, under the conditions existing in Hamburg, by
the application of land filtration. A visit to various English works con-
vinced me that sewage was being treated by the new method, at rates
which up to that time had been considered quite impossible. The works
I saw and the methods of operation were so very similar to those for land
filtration that I inquired of the experts in charge what there was new
about the method. The newspapers reported that the method was new
in that the sewage was no longer allowed to flow continuously through

Digitized by



the filter, but was fed intermittently, and that the filters were inoculated
with living organisms which digested the sludge retained by the filter
and thus prevented clogging, a difficulty with which earlier attempts at
sewage filtration had htid to fight. Of this destruction of sludge it was
said : ^* This is secured by first putting in organisms which have the nasty
taste to feed and thrive on refuse." The idea became prevalent that the
filters were inoculated with bacteria which were able to consume the
organic matters out of the sewage. It was assumed that this .was
accomplished whilst the filters were full of sewage. The object of allowing
a period of rest after discharging the purified sewage was to wait " until
the filth destroyers again get hungry and ready to perform their office."
These sludge-consuming bacteria possessed considerable interest for me,
especially as my own experiments to discover bacteria which would be
able to accelerate the decomposition of organic matters in sewage had
remained without result. It then appeared, however, that no more favour-
able results had been obtained in London from experiments carried out with
this object in view. Pure cultures of bacteria were not employed, and
the whole matter rested upon a theoretical assumption. The experiments
were simply attempts to imitate on a larger scale those carried out at
Lawrence with Filter No. 16a, using the modified method of operation
above described, and also coarser filtering material of coke, clinker, and
broken bricks.

Just as in Massachusetts it had been believed that the bacteria decom-
posed the dissolved organic matters whilst the sewage was passing through
the soil, so in England it was thought that this decomposition took place
whilst the sewage was standing in the filter. The coke and other materials
were assumed simply to form a nidus for the bacteria, and at the same
time to distribute the sewage in drops through the pores of the filter, so
that each drop could be attacked by the micro-organisms. From this view
arose the name of " bacteria bed " for this form of filter, and the method
was termed "bacterial purification." It was assiimed that the energy of
the bacteria was so great as to decompose the substances to such an extent
that no clogging of the filter would ever occur. In some quarters it was
even asserted that sand and mineral detritus would be destroyed by the
bacteria, so that even from these substances no clogging was to be feared.
Such views could naturally only be regarded as exaggerations. The
questions as to whether the decomposition of the organic matters occurred
whilst the filters were standing full of sewage, and whether the decomposi-
tion could take place without endangering a gradual clogging of the filters,
appeared, however, to be of considerable importance. We have therefore
conducted a long series of experiments in order to elucidate these questions.
The practicjil importance of contact beds has been considerably diminished
by the development of percolating filters, which possess greater capabilities,
but the results of the Hamburg experiments are of equal scientific and
practical importance for both these methods of treatment. A short de-

Digitized by




scription of these experiments therefore follows here,
out (luring the years 1897 to 1900.

They were carried

Are the Dissolved Organic Matters in Sewage Directly Attacked, De-
composed, and Mineralised during the Period when the Contact
Beds are Standing FuUI

If a tank filled with pieces of clinker is allowed to stand for a few hours
filled with sewage to the surfjice of the clinker, and is then emptied, the
effluent is non-putrescible and contains nitrates, even although the sewage
originally run into the tank was free from nitrates. If these changes are
due to a direct decomposition of the dissolved organic matters by means of
micro-organisms, it should be possible to show that the decomposition takes
place gradually. In order to test this, six filters were constructed of exactly
similar size and of the same material, and they were charged simultaneously
every day with the same sewage. The first filter was discharged after
standing full for half an hour, the second after an hour, and so on. The
results are shown in the following table : —

Sudden Reduction in Oxygen Absorbed effected bv Contact Beds.


absorbed (parta per


Time Bed i

standing .


— _ —

in Oxygen



1 Absorbed on

6th day.






Grade sewage



9 07





Effluent from bed 1

0-6 1





' 6873


1 1



4 07


1 72-48








>i 1, 4







i> „ 6













1 86-21

The table shows clearly that the reduction in the oxygen absorbed,
which may be taken as a measure of the changes which have occurred, is
not so great on the first as on the following day. On the sixth day the
filters were so far matured that they effected a considerable reduction in
the oxygen absorbed. Even with only half an hour's contact the oxygen
absorbed was reduced from 11 '42 to 3-57, i,e. by 68-73 per cent., and the
effluent was non-putrescible. \^y a longer contact the oxygen absorbed
was further reduced, but not by any means at the same rate. The main
portion of the purification had taken place, therefore, during the first half
hour. The experiment was repeated, allowing the first filter to stand full
of sewage for five minutes, the second for thirty minutes, and so on. The

Digitized by




filters were now more mature, and within the first five minutes the oxygen
absorbed was reduced from 13*87 to 2*34, ie. by 83*2 per cent. The main
portion of the purification had thus been achieved during the first five
minutes. The separation of the putrescible matters in solution does not
therefore occur gradually, as would be the case if it were due to the direct
decomposing action of bacteria, but quite suddenly. This sudden reduction
of the organic matters must therefore have some other cause.

The result of the following experiment also cannot be explained by the
assumption of a direct bacterial decomposition of the dissolved organic
matters. A clinker bed was filled with sewage, emptied after standing full
for an hour, and then continuously dosed with a volume of sewage sufiicient
to fill the bed five times, the bed remaining full to the surface. The results
are given in the following table : —

Dbcrkasb in the Purification eppectbd by Contact Beds
when worked continuously.


Oxygen absorbed (parts

per 100,000) .
Percentage redaction
Odoar ....





one Hour

in Bed.



After Passage of Sewage Equivalent
to (Fillings)


3*54 3*40 ; 6-90
66-0 66-5 I 320
musty musty slightly



8 47

By remaining for one hour in the bed the sewage was converted into a
non-putrescible product, smelling somewhat musty, but devoid of any faecal
odour. The same result was obtained from the second filling ; the effluent
was non-putrescible, although it had only been in the filter a few minutes.
During the next filling the purification effected was not so great and the
effluent had a slight faecal odour. Such results cannot be explained by the
assumption of a direct decomposition by means of the bacteria in the
filters, but may be explained by the assumption of absorptive action.
It is known that absorption is a rapid process, taking place, as in this
case, within a few minutes, and being able to separate considerable
quantities of dissolved organic substances from a liquid, and also that the
process takes place repeatedly with the same rapidity until the absorptive
powers are exhausted. These powers must then be regenerated by the
action of micro-organisms, i.e. the retained organic matters must be
decomposed and mineralised before the filter can again resume its

Digitized by



If a solution of albumen, containing about as much organic matter as is
present in ordinary domestic sewage, is placed in a sterile clinker filter,
and the liquid examined every few minutes, it will be found that in this
case a separation of the organic matter from the solution takes place.
During the first few minutes 50 per cent, or more of the organic matter is
removed, whilst later the action takes place much more slowly (see fig. 88).
The same action takes place, therefore, in the absence of bacteria as occurs
with sewage in biological filters.

This absorptive action may be easily demonstrated by the use of
colouring matters, such as methylene blue. A deep blue solution of this
colouring matter assumes a much lighter greenish colour by simply being


Mlnutcfc «»o<»M»i#l
Hour*. ]

Fig. 88.— Absorption of Albumea by Sterile Clinker Filter.

poured through a mature clinker filter ; if the solution is allowed to stand
for two hours in the filter it is almost completely decolourised.

If sewage is coloured with methylene blue, the blue colour disappears in
a day or two, but on shaking with air the blue colour reappears, and these
processes may be repeated for weeks. The solutions which have been
decolourised in a clinker filter do not, however, behave in this manner ; on
shaking with air the blue colour is not restored, because the colouring
matter has not been reduced by the action of bacteria, as in the case of
sewage, but has been retained by absorption in the filter. Experiments
with fuchsin, litmus, and other similar colouring matters gave exactly the
same results.

By far the largest part of the purification which is effected whilst
sewage is standing in a clinker filter is undoubtedly due to absoq^tion
(see p. 142). At the same time, to a certain extent, biological processes
are occurring in the full filter. This may bo concluded from the formation


Digitized by CjOOQlC



of carbon dioxide which takes place in the full filter. A sewage containing
no free carbon dioxide, after being placed in a contact bed, contained 6*38
parts per 100,000, and, four and a half hours later, 11*51 parts. In order to

Online LibraryWilliam Phillips DunbarPrinciples of sewage treatment → online text (page 18 of 36)