See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/265990355

Electroflocculation in the Treatment of Polluted Water

Article · January 1999

CITATIONS READS
5 1,383

1 author:

Vivian Robinson
ETP Semra Pty Ltd
63 PUBLICATIONS   801 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Scanning Electron Microscopy View project

Electrolytic Water Treatment View project

All content following this page was uploaded by Vivian Robinson on 04 February 2015.

The user has requested enhancement of the downloaded file.

 1

Electroflocculation in the
Treatment of
Polluted Water
Dr Vivian Robinson,
Electropure Australia Ltd,
Wadonga, Victoria Australia

Abstract

Electrocoagulation and electroflocculation are two techniques involving the

electrolytic addition of coagulating metal ions directly from sacrificial
electrodes. These ions coagulate with pollutants in the water, in a similar
manner to the addition of coagulating chemicals such as alum and ferric
chloride, and allow the easier removal of the pollutants. The electrolytic
addition of these ions has a number of advantages over their addition
chemically. There is no addition of anions meaning no increase in salinity of
the treated water. The system produces half to one third of the sludge.
Greater activity means less metal ions required and a wider range of
pollutants can be removed. In electroflocculation, the pollutants are
removed by the bubbles which are generated during the process, capturing
the coagulated pollutants and floating to the surface. Appropriately
constructed a single unit can be used to coagulate the pollutants and float
them to the surface, from where they can be easily removed. Removal rates
in excess of 98% have been achieved from a single stage process. Results
from industrial installation confirm the theoretical benefits which can be
obtained from this process.

Key Words. Electroflocculation, electrodes, aluminium, clay water, fats, oils

and greases, sewerage treatment plant, faecal coliforms, removal.

1 Introduction

Wastewater is the largest disposal problem associated with waste and by

product production, being typically many tonnes of water per person per
year in industrialised countries. Yet the total amount of pollutant in the
water is often much less than 0.1% (1,000 ppm). Great quantities of water
are used to remove small amounts of pollutant. The magnitude of the waste
problem could be considerably reduced if techniques were used which
concentrated or removed the small amount of pollutant and left the majority
of water in a condition suitable for reuse. Many different techniques are
available, including a variety of filters, chemical dosing, reverse osmosis and
similar. Many of these are either very pollutant specific or more expensive
than dumping and using more water. These latter options are fast
disappearing and new methods need to be sought.
 2
For a technique to be successful it needs to satisfy a number of criterion,
including:-

i) It must be cost effective, producing results which are cheaper than

dumping and purchasing more water. With water costs in Australia of ≈
$0.70 c/kL for purchasing water and a similar price for disposing to sewer,
both the capital and operating costs must be low and the results as good as
tap water for the technique to be successful.

ii) It must remove the contaminants producing water which can be

reused in the same application, or as a replacement for water in
another application.

iii) It must not in itself produce great amounts of waste.

iv) It must operate reliably and have low downtimes.

Electrolytically based processes appear offer some of these advantages.

These involve the passage of an electric current via sacrificial electrodes.
Their operating costs are low because of the low cost of electricity and the
element materials used. Many attempts have been made in the literature to
take advantage of the potential associated with these capabilities, with
limited success. The process, usually known as electrocoagulation (Stuart,
1946; Bonilla, 1947, Matteson et al., 1995), involves the passage of an
electric current, usually using aluminium electrodes. Aluminium goes into
solution at the anode and hydrogen gas is released at the cathode. The
aluminium coagulates with the pollutants which can then be removed,
usually by either settling or filtration.

Some of the problems encountered with this technique have been associated
with element lifetimes (see for example Osipenko and Pogorelyi, 1977;
Nikolaev et al., 1982). This has resulted in high operating costs and made
the process uneconomic and the potential benefits were lost. Recently these
problems appear to have been overcome and at least one commercial
electrocoagulation unit is available. A variation of the electrocoagulation
process is one in which the gas generated by the process captures the
coagulated pollutants and floats them to the surface. Known as
electroflocculation, it offers the possibility of a one step process, in which
polluted water is pumped into a processing reactor, the pollutants are
floated to the surface and the cleaned water is pumped out at the end of the
process. Theoretically there are no chemicals to be added and no filters
which need cleaning, enabling this process to have the potential to achieve
the above goals.

Laboratory prototype reactors had been constructed in which preliminary

work was undertaken to check the feasibility of the process and determine
its viability to treat different types of wastewater. These results also
enabled an understanding of the processes involved to be gained. Based
upon the success of laboratory experiments, industrial scale units were field
 3
trialled in a number of different applications. Our results have indicated our
ability to treat a variety of different types of water. Fats, oils and greases
(FOGs) removal rates in excess of 99.95% have been achieved. Turbidity of
rural clay water has been reduced to below 1 NTU, from values starting at
greater than 500 NTUs. Bacteria removal rates in excess of 99% have been
achieved from single stage reactors. The process appears to hold significant
advantages for the treatment of wastewaters.

2 Theory of Electroflocculation

Electroflocculation provides an alternative technique for the removal of

pollutants from wastewater. The process involves the application of an
electric current to sacrificial electrodes, usually aluminium, inside a
processing tank. The reactions at the anode and cathode respectively are
typically

1) Al - 3 e- Al3+
and
2) 2H2O + 2e 2OH- + H2

generating aluminium ions as a coagulating agent as well as gas bubbles.

The well known properties of the aluminium ions as a coagulating agent
causes them to combine with the pollutants. The gas bubbles generated can
capture the coagulated agglomerates, similar to the application in a
dissolved air flotation (DAF) unit, resulting in most of the pollutant being
floated to the surface. In theory, this electroflocculation process can be
used as an alternative to DAF processing, with the added advantage that
there are no chemicals to be added. As is the case with DAF, there are no
filters to be used for final pollutant removal.

In addition to the use of aluminium, other metals also have coagulating

properties and can be used in place of aluminium. The use of these in
conjunction with or independently of the use of aluminium, offers several
potential advantages associated with the electroflocculation process. Not
the least of these is that the aluminium/metal coagulating cation is added
without the addition of an anion. This means that the whole process is
carried out without any substantial effect upon the salinity of the water.
Additionally, the sludge produced by electrocoagulation is reduced by a
factor of 2 or 3 over the use of chemical coagulants( Musquere et al., 1983).
Also, experiments suggested that electrolytically added aluminium ions were
much more active than chemically added aluminium ions (Donini et al.,
1994), meaning that less aluminium was required and that this process
could be used to treat a number of different pollutants which could not be
handled by chemical flocculants such as alum (aluminium sulphate).

Based upon these and other perceived advantages, considerable

development work was undertaken with the objective of determining if the
above mentioned potential advantages could be converted into a practical
system.
 4

3 An Electroflocculation based water treatment system

In order to take advantage of the capabilities of the system, a number of

different features had to be incorporated into the operation. Perhaps the
most obvious of these is element lifetime. Considerable development work
was undertaken to determine the causes of element failure and take steps to
prevent said problems. This was done in a multi pronged approach. Firstly,
by evaluating the causes of element failure, we were able to take steps to
prevent the problems from occurring. Secondly, should the problem occur,
we were able to produce a design in which the elements were easily cleaned
and replaced, with low operating down times. By a combination of these
processes we were able to minimise the problems associated with the
element lifetimes, to the extent that these no longer represented the limiting
factor in a number of tested applications.

In order to use the produced hydrogen gas as a flotation medium, we

needed to design the elements such that this could be achieved effectively.
This has been done and the system now operates effectively as a flotation
system, with most of the pollutants floating to the surface, from where they
need to be removed. In our system, we have used a conical shaped top.
The pollutants float to the surface, where they are concentrated. From there
they are floated out the top by simply raising the water level by the injection
of further water to the bottom of the tank.
Provided this is done in a suitable manner, the water floats the pollutants
from the surface causing them to overflow into a suitable container.

Having developed a suitable system, the next step was to automate the
process so that it would operate continuously. In this regard, we decided to
operate under the batch process, involving the processing of the water one
batch at a time. This was primarily due to the different reactions which
occur at different times in the clarification of the water. It was determined
that there were far too many variables associated with the treatment of
water in a continuous process. During our development work, it was found
to be much easier if we retained control of the process, with the water being
treated for for the total reaction. We were also able to develop a system in
which any changes in the quality of the water would be automatically
compensated for by the electronics, throughout the whole of the process.

Figure 1 illustrates an electroflocculation based water treatment system

reaction tank. Water is pumped into the tank until it reached the pre
determined level. The electrodes are activated, generating the coagulating
agent at the anode and hydrogen gas at the cathode. The coagulating agent
captures the pollutants and combines them into larger assemblies, the same
as occurs with electrocoagulation. But there the similarity ends. In our
electroflocculation reactors, the hydrogen gas captures the coagulated
pollutants and floats them to the surface. There the conditions in the tank
assure that they remain at the surface. At appropriate times, the water
level is raised and the pollutants are forced out of the top of the tank.
 5

Processing times depend upon the nature of the pollutant to be removed and
how quickly it is required to remove them. Typical processing times are
between 1 and 3 hours. The water is pumped in and processed. The
pollutants are removed from the top by periodically adding water from the
flush water reservoir, the amount of water added being about 5% of the
volume of the water being processed. The cleaned water is then pumped
out via the flush water reservoir, refilling it every time it is emptied. In this
manner, the process is completely self contained and will repeat itself for as
long as there is contaminated water to be processed and the cleaned water
reservoir has space to take more water. The only difference between this
and a continuous process is the existance of water storage tanks before and
after the processing tank. If these are provided, the combination can
receive water at a given rate, process it and pass it on again at the same
rate. This automated operation feature gives it the properties of a
continuous system.

The process adds aluminium to the water as well as raising the pH slightly
due to the generation of hydroxide radicals through reaction 2. The amount
of aluminium added depends upon the nature of the water being treated.
However typical amounts added, as determined from calculations based
upon the passage of current and the reaction at the anode, reaction 1,
suggest that amounts of between 10 and 50 mg/L are used. Beyond that
amount, the process starts to be too expensive for many applications. The
insolubility of aluminium and the nature of the coagulating reaction mean
that most of the aluminium is used in the coagulating process and the
aluminium levels in solution are not often beyond the 2 mg/L level.

Figure 1 Schematic illustration of an electroflocculation based water

treatment system reaction tank.
 6
Figure 2 Schematic illustration of the design of an electroflocculation
process which concentrates the pollutant to produce a sludge containing as
little as 1% of the water treated.

After the water has reacted fully it is pumped out of the reactors.
Aluminium levels at this stage are typically less than 1 mg/L. Measurements
have shown that this level of aluminium drops as the water is allowed to
stand further. Typically values of less than 0.1 mg/L are reached after 12 -
24 hours standing, that being less than the levels of aluminium often found
in drinking water.

The removed water containing the pollutant flows into a storage tank. In
most applications, the pollutant settles out of the water upon standing for a
short time, usually by sinking to the bottom. This concentrates the pollutant
into a sludge, while the water can be drained off and returned to the waste
stream for further processing, as is illustrated in figure 2. Water from the
top of the tank flows into the slurry receival and settling tank. There the
pollutants settle to the bottom. As this tank fills, water overflows back to
the raw water source. When the slurry has built up too much, it is pumped
into the sludge receival tank, where further settling also occurs. Periodically
the sludge at the bottom of the processing reactor is also pumped into the
sludge settling tank. There the sludge settles to the bottom and excess
water flows from the top to the raw water source. By this mechanism, the
pollutant is concentrated into as little as 1% of the original water. The
resultant sludge is pumped away and or otherwise disposed of as required.
This has several benefits. Firstly the pollutant has been highly concentrated
and the volume of material to be disposed is considerably reduced.
Secondly, it is possible to reuse the cleaned water, sometimes in the original
application, at other times in the secondary applications. Alternatively, the
removal of the sludge makes the costs of disposal to sewer much lower
because of the reduced levels of pollutant in the water.
 7
4 Some applications

4.1 Clay water/suspended solids

It can be appreciated that there are a large variety of different types of

wastewater available for processing and determining results which are
typical for all of these is not feasible. Instead we embarked upon tests with
known amounts of a constant pollutant. In this case, we used clay in
Sydney tap water. The conductivity varied between ≈ 150 µS/cm to ≈ 250
µS/cm. We used 1.0 g clay per litre of water, which gave a turbidity of ≈
400 NTUs. The water was processed until the reaction was complete, at
which stage it was pumped out. It should also be noted that this also
represents conditions typical of processing fresh water found in inland dams
and rivers in Australia and many other countries. This water would
otherwise be suitable for domestic use, were it not for the presence of the
clay and sometimes other contaminants such as bacteria and algae, etc. In
this case, the turbidity of the processed water is typically less than 10 NTUs
at discharge, with an aluminium content of ≈ 1 - 2 mg/L. Upon standing,
the turbidity reduces to less than 3 within a few hours and is often less than
0.5 NTUs within 24 hours.
The turbidity is due to at two factors, residual micro bubbles remaining in
solution and residual aluminium in suspension. The aluminium settles out
with time, producing a white precipitate at the bottom of the container. As
this occurs, the turbidity and aluminium content reduce. The slight increase
in pH, typically ≈ 0.5 - 1.5 pH units depending upon the starting pH and the
amount of the reaction, is not sufficient to alter the condition of the water.
The final conductivity of the water is usually within 10% of the starting
conductivity, meaning that the system neither reduces nor increases the
salinity of the water. The observed improvements in the quality of the water
strongly suggest this system can be used to comply with the Australian
Drinking Water Guidelines.

Despite the complexities of the reactions inside the reactor, the effect of the
process is that water is pumped into our reactor, the impurities are floated
off the top and the clear water is pumped out the bottom. The process
works on suspended solids, down to the size of and including large dissolved
molecules. Included in suspended solids are bacteria, algae, clay (from
dam/river water), spores, some BODs. Included in large dissolved
molecules are decayed organic matter (humus), dye molecules, some
detergent molecules and a whole range of other materials. Typical removal
rates are 98% + of what the process can remove.

4.2 Fats, Oils and Grease (FOGs)

By the appropriate choice of electrodes and operating conditions, which vary

from industry type to industry type, we have been able to obtain satisfactory
results from a wide range of pollutants. One of the most successful has been
the removal of fats, oils and/or greases (FOGs) from emulsion in the water.
These are associated with wastewaters from grease traps and oil traps.
 8
The application of a voltage and the subsequent passage of an electric
current through water containing emulsified FOGs, causes them to be
released from emulsion. The gas bubbles produced combine with the
natural buoyancy of the FOGs and the coagulating effect of the metal, to
raise the FOGs to the surface, from where they can be floated off and
removed. Removal rates of greater than 99% have been achieved from a
single stage process. When combined with a two stage process, removal
rates of greater than 99.95% have been consistently achieved in industrial
applications treating up to 10,000 litres per day. Table 1 gives an indication
of results achieved from grease trap water. As well as reductions in the FOG
concentrations, the COD level has been reduced by over 96%. In the case
of this particular application, the residual FOG after treatment, 20 mg/L, was
less than the 60 mg/L associated with sewer water and the water could be
disposed directly into the STPs EAT unit. Several methods were available for
disposal of the removed material. As the volumes of water now containing
the pollutants was considerably reduced, the costs associated with disposal
were also reduced, representing a considerable saving.

Figure 3 gives an illustration of the results achieved from oil trap water,
where it can be seen that almost black oil containing water has been reduce
to a turbidity of 20 NTUs. With further settling and/or filtration, a final
turbidities of less than 1 NTU has been achieved. Again, the concentration
of the oils means that the volume of water to be disposed is likewise
considerably reduced, giving large reductions in disposal costs. No attempt
has yet been made to determine if the removed oils can be used as a fuel or
for recycling purposes. If either of these applications were available, this
would considerably reduce the costs of operating the system by producing a
by product which was saleable. The above results were achieved from the
installation shown in figure 4.

4.3 Sewerage Treatment Plant (STP) Effluent Aeration Treatment (EAT) Unit

Water in an EAT unit at a sewerage treatment plant was passed through one
of these systems to determine the effects achievable using an
electroflocculation treatment. Table 2 summarise some of the results
obtained. Typically turbidity was reduced to below 2 NTUs, which, after
settling, was further reduced. Bacteria counts were reduced by ≈ 98 - 99%
in the primary treatment tank. This reduced the count sufficiently that only
minor antiseptic treatment was required to further reduce the levels to zero.
This could be done by the use of ultra violet light, or the electrolytic addition
of silver, or both. This could eliminate the need for chlorination of the
water. It should also be noted that the removal of the suspended solids was
achieved without the addition of alum or other traditional coagulating
agents.
 9
Table 1. Before and after values obtained from treating grease trap water
and oil trap water in a system with a capability of up to 10,000 litres per day

Water source Pollutant Before mg/L After mg/L Percent Removal

Grease Trap FOG 81,440 20 99.97
COD 145,500 5,320 96.30
Oil Trap FOG 6.091 36.80 99.40
COD 16,360 763 95.30

Figure 3 Photographs of oil trap water before, LHS and after, RHS,
treatment in the electroflocculation process.

Figure 4 Photographs of installation used to treat grease trap and oil trap
water

In previous applications, this type of water has often been clarified by the
addition of alum. As a result of this, the treated water has too high a
salinity to be reused for agricultural purposes. However at the end of this
treatment, it is seen that there has been no increase in salinity and as such,
provided the other criterion are met, the water appears to be suitable for
agricultural reuse. Again, the pollutants can be removed and concentrated
as illustrated in figure 2. An indication of the removal rates of several
 10
parameters involved in the treatment of STP’s EAT unit water is given in
Table 2. Please note that these figures are a compilation of measurements
made at different times and do not relate to individual sets of
measurements. However, they serve to indicate the types of results which
have been achieved. The addition of silver as a disinfectant, typically 0.2 -
0.4 mg/L can considerably enhance the ability of the system to remove all
faecal coliforms and the 0 faecal coliforms per 100 ml results were obtained
after either ultra violet light or the addition of silver. Residual silver was
0.11 mg/L The pH of the treated water was typically in the range 8 - 9. The
low turbidities were achieved with the conductivity remaining well below
1,000 µS. Tests for pH, conductivity, turbidity and faecal coliforms were
performed according to EL12, EL5, EL2 and EL28A respectively, Standard
Methods for the Examination of Water and Wastewater, 19th Edition, 1995,
APHA. Cryptosporidium and giardia were tested according to G.C.W. 13.

Table 2 Summary of some measured values obtained after treating EAT unit
water using the electroflocculation based system.

EAT Decant Electropure Treatments Plus Ultra violet light

Faecal Coliforms 360,000per 100ml 94/100ml 78/100ml 0/100ml

Turbidity 1.5 NTU 2.4 NTU 1.5 NTU
Conductivity 395 µS/cm 508 µS/cm 700 µS/cm
Cryptosporidium 750/50L 12/50L 10/50L <2/50L
Giardia 7500/50L 70/50L 55/50L <2/50L

5 Discussion

From the above, it is easily seen that the results obtained from small scale
industrial trials of a commercial electroflocculation system appear to be quite
promising. We have been able to reproduce the results of laboratory tests
across to these small scale industrial units, with scale up factors of x 100, at
least for these waters. In laboratory tests, we have examined a large
variety of different pollutants. Here we have found some interesting
features relating to the removal of pollutants. In most cases, the larger the
physical size of the pollutant particles, the easier it is to remove them. This
means that micron size and larger suspended solids are generally the first to
come out. After that comes the smaller suspended solids and the larger
dissolved molecules. Dye molecules were generally easy to remove and
removal rates in excess of 98% have been achieved. Indeed, different dye
molecules, which are often different size, are removed at different rates, as
is seen when they are removed - molecules of one colour are often removed
before those of another colour start to come out. (It is acknowledged that
this may not necessarily be due only to size.)

The operation of the system depends heavily upon the conductivity - which
varies from between ≈ 50 µS/cm (µS) for contaminated rain water, to over
 11
10 mS/cm (mS) for some water heavily contaminated by salts - and pH of
the water. Best results were achieved with a starting pH of between ≈ 4.5
and 8. Where water to be treated had a pH outside that range, it was
usually adjusted back inside that range or the results were not satisfactory.
pH dosing and control is readily available so this does not represent a
problem. Variations in conductivity are handled by the design of the
elements and power supply. This is done at the time of installation, from a
knowledge of the conductivity of the water. Once this has been set,
variations in conductivity by a factor of 2 up or down can be easily
accommodated by the system.

The nature of the materials to be removed also varies greatly, from the
FOGs situation mentioned above, to suspended solids, dissolved molecules,
bacteria, algae, heavy metal cations, anions, BODs, spores, nutrients and
more. This range is so large that it is not possible for a single system to
handle all these different conditions. However it is possible to modify the
system to meet these different requirements. Fortunately the water quality
generally does not change too much within a particular industry and
installation. By tailoring the system to the needs of that installation, it is
possible to use the basic system in a wide range of applications. We believe
this system can be used to handle a wider range of water and pollutant
types than any other single system. As mentioned above, the system has
been used to handle suspended solids (clay) in water, where it has reduced
turbidity from over 400 NTUs to less than 1 NTU. Removal rates of FOGs
from oil and grease trap water have been well over 99.9%, with a COD
removal rate of greater than 95%. Bacteria removal rates have been
greater than 98%,. Parasites have been totally removed. All of these have
been achieved using the basic reaction tank and varying the operating
parameters. In our laboratory experiments we have also been able to
remove a large assortment of other materials, with removal rates typically in
excess of 98%. It is perhaps worth pointing out again that this is done
without the need to add any chemicals, apart from pH dosing.

One of the industries in which we have achieved successful industrial

installations has been the poultry industry. Here we have been able to
remove suspended solids from dam water to be used for drinking purposes
as well as spraying for humidity control. Without the fine clay in the water,
the spray heads no longer blocked up and humidity control was much more
reliable. Units have also been installed to remove bacteria and other
contaminants from water which comes into contact with humans and where
it is needed to be removed for horticultural reuse.

The lack of chemicals and the low overall maintenance requirements make
this well suited for use in remote locations. The process adds metal cations
in the same manner as the electrocoagulation purposes. As such all the
advantages of the electrocoagulation process are available in this process.
This includes the ability of the system to remove heavy metal cations, with
copper and lead having been successfully removed.
 12
6 Conclusions

An electroflocculation based water treatment system has been developed. It

uses sacrificial electrodes to add coagulating metal cations to the water and
then uses the hydrogen gas generated at the cathode to float these
coagulated pollutants to the surface. From there they are removed by
raising the level of the water slightly, to flush them out from the top of a
conical container. Although this process uses about 5% of the water volume
to achieve this result, most of this can be recovered if required. The one
basic system can be modified to handle a wide range of water conductivities
and pollutant types to be removed. Removal rates are typically 98% of
those which it can remove, with rates over 99.9% having been achieved in
industrial applications. Because the complete reaction occurs in the reaction
chamber, there is no need for any ancillary equipment, such as filters, to
remove the coagulated pollutants. The system has demonstrated an
effectiveness in removing faecal coliforms and parasites which can represent
a threat to human health.

Acknowledgements

The author gratefully acknowledges the assistance of Mr Geoff Hardaker, Mr

Geoff Newton and Mr John Ball.

References

Bonilla, C. F., (1947), Possibilities of the electronic coagulator for water treatment, Water
and Sewage, Vol. 85, March, pp. 21,22,44,45

Donini, J.C, Kan, J.; Szynkarczuk, J.; Hassan, T.A.; Kar, K.L., (1994), Operating cost of
electrocoagulation , Canadian Journal of Chemical Engineering , Vol. 72, No. 6 (December),
pp. 1077-1012.

Matteson, Michael J., Dobson, Regina L.; Glenn, Robert W. Jr; Kukunoor, Nagesh S.; Waits,
William H. III; Clayfield, Eric J., (1995), Electrocoagulation and separation of aqueous
suspensions of ultrafine particles, Colloids and Surfaces A: Physicochemical and Engineering
Aspects , Vol. 104, No. 1 (November), pp. 101-109.

Musquere, P, Ellingsen, F. and Vik, E.A., (1983), Electrotechnics in drinking and wastewater,
Water Supply, 8 Special Subjet, (2-3), pp. SS 8-1 to SS 8-25.

Nikolaev, N.V., Kozlovskii, A. S.; Utkin, I. I., (1982), Treating natural waters in small water
systems by filtration with electrocoagulation, Soviet Journal of Water Chemistry and
Technology, Vol. 4, No. 3, pp. 244-247.

Osipenko, V.D., Pogorelyi, P.I., (1977), Electrocoagulation neutralization of Chromium

Containing effluent, Metallurgist (English translation of Metallurg), Vol. 21, No. 9 to 10, pp.
44-45.

Stuart, F.E, (1946), Electronic water purification; Progress report on the electronic
coagulator - a new device which gives promise of unusually speedy and effective results,
Water and Sewage, Vol. 84, May, pp. 24-26.

View publication stats