View report page

10
5.—Variation in the condition of the crude sewage and its effect upon the effluent.
The sewage which has been subjected to the above treatment has varied much in character as
judged by the amount of suspended matter and of dissolved oxidisable matter which it contains.
The effluent derived from this sewage appears as a rule to have undergone a variation corresponding
to that of the sewage from which it has been produced. The results of the daily estimations of
oxidisable matter in the crude sewage and in the corresponding effluent will be found in the Diagrams
and Tables in Section III.
In order to render the variations in the degree of purification effected by the coke-beds more
evident, the percentage daily purification has been draNvn out as a curve in Diagram I., and in the
same diagram the average weekly percentage purification has been represented and the relative
amounts of oxidisable matter in the crude sewage and in the effluent also have been treated as curves.
It will be understood that the results deal only with the dissolved oxidisable matter in the sewage
and in the effluent. The solid suspended matters of the sewage wholly disappear in the coke-bed
and are therefore not taken into account.
The occasional diminution in the percentage purification does not appear to be due to an acid
reaction of the sewage hindering the bacterial action, since the sewage is always either alkaline or
neutral in reaction. Neither is this diminished purification apparently to be referred to the presence
in the sewage of undue proportions of chemical refuse derived from gas works and chemical works.
No evidence has been obtained of interference with the normal action of the coke-bed from such causes.
It appears, however, very probable that the purification of the raw sewage is most complete
when the sewage is in a fairly dilute and fresh condition. It is of interest to consider the rainfall
curve on the diagram in this connection. The least satisfactory results were obtained during the hot
dry weather of the recent summer, when the sewage arrived at Crossness in a less dilute condition,
owing to the absence of rain, and in an offensive and putrescent condition owing to the high
temperature of the air. Even when the sewage was poured upon the coke-beds in this offensive condition,
the effluent from the beds was not offensive in character, and it did not become offensive
when it was kept. It differed from better effluents only by containing a larger amount of
dissolved oxidisable matter.
6.—Relative purity of clear sewage, chemical effluent, coke-bed effluents, and lower
RIVER WATER

A comparison of these numbers with one another shows that by substituting a single coke
treatment for chemical treatment, the effluent sewage discharged into the river would be completely
free from suspended impurity, and would possess a purity, as regards dissolved putrescible matter,
of 51*3 as compared with 16*9 in the present effluent, representing an improvement of 67.1 per cent.
If discharged after double treatment in the coke-beds the percentage improvement on the chemical
effluent would be 75*6. The bacterial action continuing in the river would rapidly bring the purity
of such a liquid into a condition equalling that of the river water itself.
SECTION III.
TABULATION OF RESULTS OF CHEMICAL EXAMINATION OF THE RAW AND
THE BACTERIALLY-TREATED SEWAGE.
The results of the examination of the sewage before and after its passage through coke-beds were
obtained by estimating the amount of dissolved oxidisable matter in the crude sewage, which had been
clarified by passage through a paper filter, and then making a similar estimation in the same sewage
after it had passed through the coke-bed.
The oxidisable matter was estimated by exposing the liquid, which had been clarified by
filtration if necessary, to the action of acidified potassium permanganate, in closely stoppered bottles,
for four hours at a temperature of 80° Fahr.
The oxidisable matters are measured by stating the number of grains of oxygen per gallon of
the liquid which are required for their oxidation. When the coke-bed has been filled twice in a day, a
sample of each effluent is collected, and these equal samples are mixed for the chemical estimation.
This method of estimation is not appreciably affected by the small amount of nitrite which is
occasionally present in the effluent.
The Table which includes the numerical results will be found on pages 11 to 14. The results
obtained will be more readily appreciated by a consideration of the curvo3 in the Diagrams which
immediately follows page 41.