Copyright reserved.—Reprinted from the Journal and Proceedings of t e Institute of Sewage
Purification, Part 2, 1963, pp. 167-173
Predicted and Observed Toxicities of Several Sewage
Effluents to Rainbow Trout
By R. LLOYD, B.Sc., M.I.Biol., and DOROTHY H. M. JORDAN, B.Sc., M.I.Biol.
( Water Pollution Research Laboratory, Stevenage)
Registered Office of the Institute :
10, Cromwell Place, South Kensington, London, S.W.7
Edwards The Printers Ltd., Coventry
2
PAPERS SUBMITTED FOR PUBLICATION
Paper No.
1
Predicted and Observed Toxicities of Several Sewage
Effluents to Rainbow Trout
By R.
LLOYD,
B.Sc., M.I.Biol., and DOROTHY H. M. JORDAN, B.Sc., M.I.Biol.
( Water Pollution Research Laboratory, Stevenage)
INTRODUCTION
In an extensive field and laboratory study on
the effect of a sewage effluent on a fishery', it was
shown that fish were sometimes killed by low
oxygen concentrations and at other times by a
combination of low, but not lethal, concentrations
of dissolved oxygen and dissolved toxic substances, of which ammonia was thought to be one
of the more important. The observed relation
between the survival of fish in the effluent and the
dissolved-oxygen content of the water was close
to that expected from laboratory studies in clean
water, but the toxicity sometimes observed when
the dissolved-oxygen concentration was at a safe
level was considerably greater than that expected
from laboratory studies of some of the poisons
present. However, recent work on the effect of
dissolved free carbon dioxide and oxygen in the
water on the toxicity of ammonia 2, 3 has indicated
that under the conditions prevailing in the effluent
this poison is more toxic than was originally
supposed, and experiments on the toxicity of
mixtures of poisons 4, 5, 6 suggest that the toxicity
of a sewage effluent might be the sum of the
toxicities of the individual poisons present.
Sewage effluents, especially those from works
treating industrial as well as domestic wastes, may
contain many toxic substances, although some
would be in concentrations too small to influence
the toxicity of the effluent. One of the objects of
the present work was to determine the extent to
which the toxicity of these effluents could be
accounted for from the concentrations of those
poisons commonly present—ammonia, phenol,
heavy metals, and perhaps cyanide—for which
the method of analysis is relatively simple and
whose toxicity to rainbow trout is known. Another
object was to determine whether the concentrations of dissolved toxic substances present in
sewage effluents were likely to be a potential
hazard to fisheries.
METHOD
Samples of effluent were taken from 11 sewage
works, designated A-K, which covered a range of
sewage-treatment processes. At works A, B, and
F the sewage treated was mainly of domestic origin,
while the remainder contained different types and
proportions of trade wastes. Spot samples (100
gal) were usually taken on Tuesday mornings, and
then settled for 2 h at the laboratory, filtered
through glass wool, and stored in 40-litre polythene
bottles at 9°C or lower. Preliminary tests showed
that effluents treated in this way could be stored
satisfactorily for at least a week.
Chemical analyses were made on samples of
the effluent as used on the first day of the toxicity
test, except those for heavy metals for which the
samples were further filtered through glass-fibre
pads (Whatman GF/C). Methods of analysis
recommended by the Ministry of Housing and
Local Government7 were used for the determination of ammonia (Method B), oxidized nitrogen,
biochemical oxygen demand, permanganate value,
heavy metals, and synthetic detergents. Nitrite
was determined by the method of Montgomery
and Dymock8 and monohydric phenols by the
4-amino antipyrene method9. Free cyanide was
determined by distillation with lead acetate,
followed by colorimetric analysis by the modified
Epstein method19.
Serial dilutions of all effluents were made with
an untreated hard borehole water (total hardness
320 p.p.m. as CaCO3). Since the tests were made
with a fixed volume of solution, a low temperature
was advisable to reduce further biological breakdown of the effluents during the test, but, as the
temperature could not be too far removed from
those which had been used for the laboratory tests
on the toxicity of individual poisons (10 to 20°C),
a temperature of 10°C was chosen. Dissolvedoxygen concentrations in a river below a sewageeffluent outfall are likely to be lower than in an
unpolluted river, and so an oxygen content of 50%
of the air-saturation value was maintained in the
test dilutions by aeration with equal volumes of air
and nitrogen ; this level of dissolved oxygen is well
above the lethal level for rainbow trout yet low
enough to affect significantly the toxicity of the
effluent. Finally, in chalk streams where the
dissolved-oxygen content is half the air-saturation
3
value, the pH value is usually in the range 7-5 to 80
and a value of 7-9 was chosen and obtained in the
test dilutions by enriching the gas mixture with a
controlled flow of carbon dioxide.
TLm value (median tolerance limit) : the
concentration of poison, in a specified dilution
water, which will kill 50% of a population of test
fish within a specified time.
The rainbow trout (3 to 5 in long) used in the
tests were fed on the day before the start of the test
and were acclimatized to the test temperature for
1 day. Batches of 10 fish were then exposed to
serial dilutions of the effluent and, since samples
of effluent were obtained on Tuesdays, it was
convenient to continue the tests for 3 days. The
solutions (40-litre volume) were changed daily and
the fish fed at the end of the first 24 h. The time
taken for each fish to die was recorded and median
periods of survival were calculated by a graphical
methodl 1. A control was run with each effluent
tested ; no control fish died.
Toxicity Index : the number of times an
effluent has to be diluted with a specified dilution
water to kill 50% of a population of test fish within
a specified time, i.e., a 3-day Toxicity Index is the
number of times an effluent has to be diluted to
give a 3-day TLm value.
Some of the experimental procedures were
modified in the case of Effluent J, which was taken
as a composite sample over 24 h. In one test, Ja,
the methods described above were used. In the
other, Jb, the effluent was diluted with an artificial
soft water (total hardness 45 p.p.m. as CaCO3), the
pH value was 7-2, and the rainbow trout used for
the test were previously acclimatized for 5 days to
the soft dilution water. The other conditions of
the test were the same as those for the main series.
RESULTS
The 3-day TLm value for each effluent used
in these tests was estimated by interpolation from
a straight line fitted by eye to the plot of the percentage killed at 3 days (on a probit scale) against the
logarithm of concentration.
Results of the chemical analyses, given in
Table 1, show that these samples covered a wide
range of effluent quality. It was thought that these
analyses would include the most important poisons
present, and it will be shown later that this was
probably correct for the majority of the effluents
tested.
There was considerable variation in the
shapes of the log concentration-log survival time
curves obtained for the different effluents ; four
are shown in Fig. 1. From the curve obtained for
Effluent H it appears that this effluent would not
have shown a greater toxicity even if the test had
been continued for more than 3 days. Curves for
Effluents A, C, E, and I were similar in shape to
The following terms to express the toxicity of
poisons or effluents are used in this paper.
Lethal threshold concentration : the concentration of poison which will kill only 50% of a population of test fish even after prolonged exposure.
TABLE
1.
ANALYSIS OF SEWAGE EFFLUENTS
From works treating
domestic sewage only
Analysis
A
B
1
From works treating domestic and industrial wastes
F
C
E
D
G
I
H
J
B.O.D. (p.p.m.) ..
..
15
12
4
13
>36
27
10
32
>50
9
22
P.V. (p.p.m.)
..
20
11
10
20
17
18
11
20
24
16
39
..
38.3
16.8
3.2
33.0
15.4
37.2
3.4
22.0
31.2
28-5
31.0
41-0
39.6
21-8
0.8
11.0
3.0
11.6
11.6
9.0
1.5
1.5
1.27
0.5
0.04
0.43
0-07
0-45
1.0
0-21
0-03
..
Ammonia (p.p.m. N)
Oxidized nitrogen (p.p.m. N) ..
Nitrite (p.p.m. N)
..
Synthetic detergent
(p.p.m. as Manoxol OT)
3.82
..
I
0-24
I
6.3
4.5
4.6
4-3
9.5
2-5
4.6
5-8
3.0
3-8
N.D.
N.D.
0.2
0.5
0.5
0.6
0.2
0.8
0.6
0.4
0.1
Cyanide (p.p.m. HCN) ..
..
N.D.
N.D.
N.D.
N.D.
0.010
0.035
N.D.
0.040
0.005
0.002
0.10
..
0.29
0-06
0.005
0.02
0.09
0.31
0.035
0-003
0-12
0-01
0.20
0-023
0-010
0-010
nil
0.26
0-09
nil
0.03
0.04
1.10
0.09
0.03
0.22
0.09
0.48
0.03
0.01
0.01
0.03
0-50
0.06
0.03
0.25
0.08
0.50
0.04
0.02
0.19
0-02
N.D.
." 47 0p—• s,
I...■
..
0 0 000
Heavy metals (p.p.m.)
Zinc ..
..
Copper
..
Lead
Nickel
Chromium ..
IQ IQ I ,000
9.7
Monohydric phenols (p.p.m.) ..
not determined
33.5
0.12
0.02
0.31
0.05
2.65
0.065
0.015
0-19
0.05
4
that for H. Effluents D, J, and K, however, would
almost certainly have shown a greater toxicity had
the fish been exposed to lower concentrations for a
longer period. Table 2 shows the observed 3-day
Toxicity Indices for the effluents tested.
DISCUSSION
Prediction of Toxicity
Where the toxicity curve of a poison becomes
parallel to the survival-time axis after a certain
CONCENTRATION OF EFFLUENT (percent)
cent) LOG SCALE
Fig.
2.
20°C,
1.
Period of survival of rainbow trout in different concentrations of
Effluents D, H, Ja and K
TABLE
THE 3-DAY TOXICITY INDICES OBSERVED AND PREDICTED AT 10 °C, THE PREDICTED 3-DAY TOXICITY INDEX
AT
AND THE PREDICTED CONTRIBUTION OF SEVERAL POISONS TO THE OBSERVED 3-DAY TOXICITY INDEX
Effluent
A
B
F
C
D
E
G
H
I
ja
jb
K
3-day
Observed
10°C
1.6
N.T.1
N.T.
2-0
2.4
3.0
N.T.'
1.7
1.7
15.6
46.5
4-9
Predicted contribution
Toxicity Index
Predicted
10°C
1.7
0.8
0.3
1.8
1.5
2.5
0.4
2.0
1.7
12.7
37-3
3.6
Predicted
20°C
3.2
1.4
0.4
3.3
2.1
3.9
0.5
3.0
3.1
14.8
40-1
4.9
NII3
86
68
21
41
45
70
6
<1
22
Phenol
7
6
6
12
10
1
<1
I
( %) . to observed Toxicity Index
HCN
6
16
28
4
< 1
<1
29
N.T.' 20% mortality in undiluted effluent
N.T.2 No mortalities in undiluted effluent
Zn
6
4
13
12
8
10
72
77
19
Cu
8
10
16
6
7
5
2
3
3
Unknown
nil
11
38
19
nil
1
19
20
26
5
period of time, as does the curve shown for
Effluent H in Fig. 1, it is thought that a lethal
threshold concentration exists for that poison.
Recent studies on the toxicity of three pairs of
poisons (zinc and copper4 , ammonia and pheno15,
and ammonia and zinc6) have shown that the lethal
threshold concentration of a mixture is reached
when the equation
As/AT Bs/B T = 1
is satisfied, where A and B are the two poisons and
the suffixes S and T denote respectively the
concentration in solution and the lethal threshold
concentration of the poison for the conditions used
in the test.
In applying these results to the present work
two assumptions had to be made. First, it was
assumed the toxicity of a mixture containing more
than two poisons, including some which have not
been tested in mixtures, could be calculated by the
same method as shown above. Second, although
it is known that the lethal threshold concentrations
of cyanide, copper, and zinc are lower than those
required to kill 50% of a population of rainbow
trout in 3 days at 10°C, it has been assumed that
the 3-day TLm of a mixture containing these
poisons is obtained when the sum of their concentrations, expressed as a fraction of their 3-day
TLm value, is unity. If these assumptions are
valid, and the concentrations of the poisons in an
undiluted sewage effluent are expressed as fractions
of their individual 3-day TLm values, then the sum
of these fractions should represent the multiple
by which the effluent exceeds its 3-day TLm value,
and this number will thus be the predicted Toxicity
Index.
Since the toxicity of ammonia varies with
slight differences in pH value, temperature, and
dissolved-oxygen content of the water, the
expected 3-day TLm value for this poison was
calculated for the average conditions of each test,
using the graphical method of Lloyd3. The 3-day
TLm values for the other four poisons at the
air-saturation value of dissolved oxygen are shown
in Table 3; the values for cyanide, copper, and
zinc were derived from data obtained at higher
temperatures and corrections were made to allow
for the lower temperature of the present series of
tests by assuming 14 a Q 10 of 2-35. Factors to
allow for the lower oxygen content of the test
15
solutions were obtained from Lloyd . With
these data, predictions have been made of the
toxicities of the effluents used in these tests from
the analyses shown in Table 1, and these values are
shown in Table 2. It can be seen that there is
reasonable agreement between the observed and
predicted Indices over a wide range of toxicities ;
the tendency is for the predictions obtained by
summing the toxicities of ammonia, phenol,
cyanide, copper, and zinc, to underestimate the
toxicity of an effluent and it is possible that other
substances were present which contributed slightly
to the toxicity. Table 2 also shows that the
percentage contribution of each poison to the total
toxicity varied considerably between effluents.
TABLE 3. 3-DAY TLm VALUE FOR MONOHYDRIC
PHENOLS, CYANIDE, COPPER AND ZINC SALTS AT THE
AIR-SATURATION VALUE OF DISSOLVED OXYGEN
3-day TLm (p.p.m.)
Poison
Hard water
Soft water
References
Monohydric phenols
(as CeHs0H)
4-4
4.4
6
Cyanide (as HCN) ..
0-09
0.09
12
Copper (as Cu)
..
0.59
0.11
Zinc (as Zn) ..
..
3.80
1.19 Jr-
13,14
Previous experiments on the toxicity of
synthetic detergents to rainbow trout16 showed
that detergent residues in sewage effluents were
distinctly less toxic than similar concentrations of
raw detergents in clean water. However, no
quantitative estimation could be made of the extent
to which the toxicity was reduced. In the present
series of tests, there was no correlation between the
unaccounted-for fraction of the toxicity of the
effluents and the concentration of detergent
residues present ; in fact, Effluent A contained the
highest concentration of detergent residues (9.7
p.p.m.), yet the toxicity of this effluent was fully
accounted for by the combined toxicities of
ammonia, copper, and zinc. It may be assumed
on this evidence that detergent residues make little,
if any, contribution to the toxicity of sewage
effluents.
In assessing the toxicity of the heavy metals
in the effluents, it has been assumed that they were
present in an ionic form and were not complexed
in any way which would reduce their toxicity. This
assumption may be incorrect for those effluents
which contained cyanide. Although Doudoroff 17
has shown that zinc cyanide is very toxic and may
therefore be almost completely dissociated, similar
tests with copper cyanide were inconclusive.
Nevertheless, the fair agreement between the
observed and predicted toxicities of the sewage
effluents described here gives some measure of
support to the validity of the predictive method
used, and thus suggests that both copper and zinc
might be present as ions. In all the effluents the
concentrations of lead, nickel, and chromium were
too small to make a significant contribution to the
total toxicity.
6
Indices of the effluents at 10°C can be adjusted to
allow for the higher temperature by the following
equation,
3-day Toxicity Index at 20°C =
observed Index at 10°C x predicted Index at 20°C
predicted Index at 10°C
Dilutions required of the effluents to reduce their
3-day Toxicity Index to 0.5 under these conditions
are shown in Table 4. It can be seen that the
minimum dilution calculated from the B.O.D.
analysis is less than that based on the 3-day
Toxicity Index for two of the 11 effluents (C, Ja)
and little difference exists between three others
(A, F, K). In the remaining six cases, the
dilutions based on B.O.D. values are distinctly
greater than those based on the 3-day Toxicity
Index.
ACKNOWLEDGMENTS
This paper is published by permission of the
Department of Scientific and Industrial Research.
REFERENCES
ALLAN, I. R. H., HERBERT, D. W. M., and
ALABASTER, J. S. A field and laboratory
investigation of fish in a sewage effluent.
Fish. Invest. Series 1, 6, 1958, No. 2.
2
LLovp, R., and HERBERT, D. W. M. The influence
of carbon dioxide on the toxicity of unionized ammonia to rainbow trout (Salmo
gairdnerii Richardson). Ann. appl. Biol.,
48, 1960, 399.
3
LLoYD, R. Effect of dissolved-oxygen concentrations on the toxicity of several poisons to
rainbow trout (Salmo gairdnerii Richardson).
J. exp. Biol., 38, 1961, 447.
4
LLovD, R. The toxicity of mixtures of zinc and
copper sulphates to rainbow trout (Salmo
gairdnerii Richardson). Ann. appl. Biol., 49,
1961, 535.
1
HERBERT, D. W. M. Freshwater fisheries and
pollution control. Proc. Soc. Wat. Treatm.
Exam., 10, 1961, 135.
6
HERBERT, D. W. M. The toxicity to rainbow
trout of spent still liquors from the distillation
of coal. Ann. app!. Biol., 50, 1962, 759.
7
MINISTRY OF HOUSING AND LOCAL GOVERNMENT.
Methods of Chemical Analysis as Applied to
Sewage and Sewage Effluents. H.M. Stationery
Office, London, 1956.
8
MONTGOMERY, H. A. C., and DYMOCK, J. F. The
determination of nitrite in water. Analyst,
86, 1961, 414.
9
0ciivivsKi, F. W. The absorptiometric determination of phenol. Analyst, 85, 1960, 278.
°AMERICAN PUBLIC HEALTH ASSOCIATION. Standard Methods for the Examination of Water
and Waste Water Including Bottom Sediments
and Sludges. New York, 11th Edition, 1960.
1
Buss, C. I. The calculation of the timemortality curve.
Ann. app!. Biol., 24,
1937, 815.
12
HERBERT, D. W. M., and MERKENS, J. C. The
toxicity of potassium cyanide to trout.
J. exp. Biol., 29, 1952, 632.
13
LLovp, R., and HERBERT, D. W. M. The effect
of the environment on the toxicity of poisons
to fish. J. Instn publ. Hlth Engrs, 61,
1962, 132.
14
LLovD, R. The toxicity of zinc sulphate to
rainbow trout. Ann. app!. Biol., 48, 1960, 84.
15
LLovp, R. The toxicity of ammonia to rainbow
trout (Salmo gairdnerii Richardson). Wat.
Waste Treatm. J., 8, 1961, 278.
16
HERBERT, D. W. M., ELKINS, G. H. J., MANN,
H. T., and HEMENS, J. Toxicity of synthetic
detergents to rainbow trout. Wat. Waste
Treatm. .7., 6, 1957, 394.
17
DouDoRoFF, P. Some experiments on the
toxicity of complex cyanides to fish. Sewage
industr. Wastes, 28, 1956, 1020.
(Received 21st August, 1962)
5
The Institute of Sewage Purification
FOUNDED AS THE ASSOCIATION OF MANAGERS OF SEWAGE DISPOSAL WORKS,
INCORPORATED
1901
1932
PAPER:
Predicted and Observed Toxicities of Several Sewage
Effluents to Rainbow Trout : A Further Study
By R.
LLOYD,
B.Sc., M.I.Biol., and
DOROTHY
H. M.
JORDAN
B.Sc., M.I.Biol.
( Water Pollution Research Laboratory, Stevenage)
Reprinted from :
The Journal and Proceedings of the Institute of Sewage Purification Part 2, 1964
Edwards The Printers Ltd. Coventry
3
Predicted and Observed Toxicities of Several Sewage Effluents
to Rainbow Trout : A Further Study
By R. LLOYD, B.Sc., M.I.Biol., and DOROTHY H. M. JORDAN, B.Sc., M.I.Biol.
( Water Pollution Research Laboratory, Stevenage)
INTRODUCTION
In a previous survey of the toxicity of sewage
effluents to rainbow trout', samples of effluent
were obtained from sewage-disposal works which
rqnged from those treating only domestic sewage
to those treating sewage containing a high proportion of industrial wastes. Some of these effluents
were not toxic tv rainbow trout under the conditions of the test, while others were rapidly lethal;
the most toxic effluent had to be diluted 46 times
with clean water to obtain only a 50% mortality
within 3 days. The toxicity of these effluents as
determined by bio-assay agreed closely with the
toxicity predicted from an analysis of their contents
of ammonia, monohydric phenols, cyanide, zinc,
and copper. However, it was thought that predictions of toxicity based on this limited number
of poisons could sometimes be inadequate far
effluents from works at which a high proportion of
the sewage treated consisted of industrial discharges. Accordingly, a further survey was made
with greater emphasis on the toxicity of effluents
from sewage-disposal works serving large industrial
towns and cities.
MATERIALS AND METHODS
Twelve samples of effluents were obtained on
Tuesday mornings and, on return to the Laboratory, were filtered through glass wool and stored in
full, stoppered containers in a cool place. Effluent
M was from a pilot-scale filter treating domestic
sewage; Effluents G2 and G3 were from the works
from which Effluent G had been taken in the
previous survey but to which an increasing volume
of gas liquor was being discharged. The remaining nine effluents (N—V) were obtained from works
serving a variety of industrial areas. The survey
was carried out from November 1962 to March
1963, and during this time some of the sewage
works were adversely affected by the abnormal
cold weather, so that the samples obtained may
not have been typical of the normal effluent quality.
The chemical analytical methods were those
used in the previous survey, except that the
concentration of free cyanide in Effluent 0 was not
determined by neutral distillation followed by
colorimetric analysis, but by using the modified
Epstein's colorimetric method2 for thiocyanate and
cyanide, followed by determination of thiocyanate
alone on a sample from which the cyanide had been
removed by heating and aeration in the presence of
boric acid. In addition, total phenols were
measured by the method of Scott3.
The bio-assay procedure used was identical
with that used in the previous series. All dilutions
of effluent were made with a hard water (320 p.p.m.
as CaCO3) and aerated with a controlled mixture
of air, nitrogen, and carbon dioxide.
Average
dissolved-oxygen concentrations ranged from 43 to
47% of air saturation and mean pH values ranged
from 7.73 to 7.95, although for individual tests
these values were more constant. Since Effluents
U and V had a very low alkalinity, sodium bicarbonate was added to the dilutions to raise the
pH values to the accepted range when aerated with
the standard gas mixture. Temperatures were
maintained at 10°C.
Ten rainbow trout (3-5 in long) were tested
in each dilution of effluent, and from the survival
or death of these fish at the end of the experiment
the number of times the effluent would have to be
diluted to kill 50% of the fish in three days was
calculated. This value is termed the 3-day
Toxicity Index.
RESULTS
Analyses of the effluents are shown in Table 1.
From the ammonia, monohydric phenol, free
cyanide, zinc, and copper contents, and from data
on the toxicity of these substances when tested
individually, predictions of the 3-day Toxicity
Index of each effluent (supposing the substances
named to be the only toxic materials present) were
made by the procedure described in the earlier
paper. This predictive method is based on the
assumption that the toxicity of a mixture of poisons
is equal to the sum of their individual toxicities.
These predicted 3-day Toxicity Indices are corn-
4
TABLE I. ANALYSIS OF SEWAGE EFFLUENTS
After filtration through glass wool
Analysis
B.O.D. (p.p.m.)
P.V. (p.p.m.)
G2
..
..
..
G3
42
13
25
19
M
N
16
9
27
13
92
7-8
0
P
19
12
24
11
13-5
21.0
Q
62
R
S
10
17
37
20
23
75-5
39-5
42.2
T
U
16
V
75
188
22
33
131
17-4
15.0
48.5
Ammonia (p.p.m. N) ..
18.5
7.5
Oxidized nitrogen
(p.p.m. N) ..
37.5
22-5
14.0
34-0
6.5
8-0
4.5
1-5
1.8
4-0
9-5
9.0
Nitrite (p.p.m. N)
0-7
0-5
2-3
0.6
0.3
1-7
0.3
0-1
0-4
0-9
0-4
1.7
Synthetic detergent
(p.p.m. as Manoxol OT)
4.4
2.5
8-6
3.3
2-8
3.2
6.7
5.2
4-6
1-0
1.3
7.4
Monohydric phenols
(p.p.m.)
0.7
0-9
0.1
Nil
1.7
0.2
3.4
0-3
0.4
0.2
0.3
5.7
Total phenols (p.p.m.) ..
1-3
7.7
0.5
0-4
2-7
2.1
12-4
4-8
3-7
3.3
4.0
27.0
0.01
N.D.
Nil
0-08
0.03
0-06
0.06
0.12
0-02
0-008
0-03
0.08
0.03
0.007
0.07
0.04
0-50
1-00
Nil
2-18
0.40
0-87
0-09
0-02
0.40
0.10
0-22
0-01
0-003
0-32
0-04
0-18
0-03
0-007
0-23
0.10
0.24
0-26
0-055
0-12
0.03
0-23 .
0.05
0-02
0-09
0.15
0.30
0-05
0-12
0.44
0.18
0'40
0-10
0-14
1.20
0-10
Free cyanide (p.p.m. HCN) <0-005
Heavy metals (p.p.m.)
„
Zinc ..
Egger..
::
Nickel
Chromium
0.27
0-05
0-006
0-03
0.04
..
0-25
0-03
0-01
0-20
0-02
0-07
0-03
0.003
0-01
0.02
N.D. not determined
TABLE
2.
OBSERVED AND PREDICTED 3-DAY TOXICITY INDICES OF 12 SAMPLES OF SEWAGE EFFLUENTS TO RAINBOW TROUT
Effluent ..
..
G2
G3
M
N
o
P
Q
R
S
T
1
U
V
Observed
..
<1.0
I <1-0
4-5
<1.0
1-5
1-2
3-6
2.6
2-6
1-2
I
1-1
8.9
Predicted
..
0-9
0-9
4.0
4-5
1-7
4-7
2-7
3-9
1.0
I
1.0
4.3
..
I
0-35
pared with the observed toxicities of the effluents
in Table 2, where it can be seen that there is
reasonable agreement between the observed and
predicted toxicities with all but four of the
effluents-0, P, S, and V.
DISCUSSION
Sewage effluents must often contain a large
number of toxic substances, many of which are at
sub-lethal concentrations, and the accurate prediction of the toxicity of such effluents, based on a
knowledge of the toxicities of all the individual
poisons and their interactions, would be a most
difficult task. One of the objects of these two
surveys has been to find out how far the toxicity
of these effluents can be ascribed to the sum of
the toxicities of a few poisons which are known
often to be present in sewage effluents, when
standard analytical procedures are used to determine their concentrations. The results of both
surveys, given in Fig. 1, show that in many
instances the simple method for predicting toxicity
is reasonably accurate. Out of the 24 samples
examined, 6 were correctly predicted to be not
toxic, and the predicted toxicities of 13 of the
remaining 18 samples were within + 30% of the
observed values.
Where the toxicity has been under-predicted
it is very probable that poisons other than those
analysed were present in sufficient concentration
to increase substantially the toxicity of the effluent,
and it is of interest that the two effluents whose
toxicities were under-predicted by more than 30%
were from works receiving a high proportion of
industrial discharges in the sewage. An increase
in accuracy of prediction in these instances might
have been obtained by a more detailed investigation
of the nature of the industrial discharges and of the
poisons which they contained.
Over-prediction is more difficult to account
for, since it implies that a proportion of the poison
found by analysis was present in a less toxic form.
Of the five poisons considered to be important in
sewage effluents, the heavy metals are the most
likely to contribute to this error, although in the
previous survey it was concluded that the concentrations of zinc and copper in those effluents were
as toxic as similar concentrations in clean water,
particularly since the toxicity of one effluent (J) was
almost wholly correlated with its high zinc content.
However, this conclusion may not apply when the
concentration of copper is high. In a recent
paper4, McDermott, Moore, Post, and Ettinger
presented data on the removal of copper sulphate
from sewage by the activated-sludge process and
5
Even the value of 0.08 p.p.m. free HCN found in
Effluent 0 by the procedure described previously
may be too high because copper cyanide complexes
can also interfere with this method. Thus, where
an effluent contains a relatively high content of
heavy metals and heavy-metal cyanide complexes,
the methods of chemical analysis employed in this
survey might well lead to an over-prediction of its
toxicity to fish. Although more refined techniques
for the analysis of free cyanide are available5 , it
would be extremely difficult to obtain an accurate
showed that only about half of the total soluble
copper in the effluent would react in the cuprethol
test; this implies that the remainder was cornplexed. Effluents 0, P, and S contained a
relatively high concentration of copper, and
Effluent 0, where the over-prediction of toxicity
was particularly high, had the highest copper
content. It is possible, therefore, that some of the
copper was present in the form of less toxic complexes. Also, too high a value for the concentration of free cyanide may have been obtained from
01961-62
X1962-63
50 —
0/
I
20
II
-
a
X/
>,
' 20-
•
/0
in
a
LU
i-10
LU
EX
a.
0.5
I
02
1
05
1. 0
1
1 1 1 1
1
2.0
20
5.0
10
OBSERVED 3-DAY TOXICITY INDEX
I
1
I
I
I
50
Fig. 1. Predicted and observed 3-day Toxicity Indices of sewage effluents
from the surveys of 1961-62 and 1962-63. Broken lines are limits of ± 30% error
the two analytical methods used in this survey.
An analysis of the free cyanide content of Effluent 0
by neutral distillation gave a value of 0.67 p.p.m.
HCN ; if this concentration was present, rainbow
trout should have been killed much more rapidly
in the undiluted effluent than was observed to be
the case. The high value of free cyanide was
probably the result of the breakdown of copper and
nickel cyanide complexes during distillation, and
these complexes are relatively non-toxic to fish.
estimation of the concentration of heavy metals
existing in the toxic, ionized, form in a sewage
effluent.
Although the results obtained from these two
surveys indicate that, for about two-thirds of the
sewage effluents investigated, the toxicity could be
predicted with an accuracy of ± 30%, from the
observed concentrations of a limited number of
common industrial poisons, it does not follow that
6
bio-assays of such effluents are unnecessary. Only
bio-assays can demonstrate with any certainty
whether the effluent is toxic or harmless. But if a
chemical analysis, similar to that used in these
surveys, is made on the sample of effluent used for
the bio-assay, and the results of the bio-assay are
subsequently shown to be in agreement with the
prediction based on the concentrations of poisons
present, the advantages gained are considerable.
Not only can the relative importance of each poison
present be assessed but also the probable effect of
changes in the composition of the effluent and of
the chemical and phyaical condition of the river
water to which the effluent is discharged can be
allowed for. These advantages have been discussed more fully in the previous paper.
In the previous survey it was shown that there
was no correlation between the B.O.D. of the
effluents and their 3-day Toxicity Indices when
recalculated for a temperature of 20°C. The
greater inaccuracies of the predictions in the
present survey do not allow such a recalculation
to be made for these effluents, although it is
reasonable to expect that their toxicities would be
greater than those observed at 10°C. Nevertheless, inspection of the present data again shows
clearly that there is no correlation between the
B.O.D. of the effluents and their toxicity caused by
dissolved poisons.
ACKNOWLEDGMENT
This paper is published by permission of the
Department of Scientific and Industrial Research.
REFERENCES
LLovD, R., and JORDAN, D. H. M. Predicted
and observed toxicities of several sewage
effluents to rainbow trout. yinst.Sew.Purif.,
1963, (2), 167.
1
AMERICAN PUBLIC HEALTH ASSOCIATION. Standard
Methods for the Examination of Water and
Waste Water Including Bottom Sediments and
Sludges. New York, 11th Edn, 1960.
2
3
ScoTT, R. D. The detection of phenols in water.
Industr. Engng Chem., 13, 1921, 422.
4
McDERmoTT, G. N., MOORE, W. A., POST, M. A.,
and ETTINGER, M. B. Effects of copper on
aerobic biological sewage treatment. J. Wat.
Pollut. Contr. Fed., 35, 1963, 227.
5
SCHNEIDER, C. R., and FFtEUND, H. Determination of low level hydrocyanic acid in solution
using gas-liquid chromatography.
Anal.
Chem., 34, 1962, 69.
(Received, 9th October, 1963)