Filter backwash water treatment options

85
© IWA Publishing 2014 Journal of Water Reuse and Desalination
|
04.2
|
2014
Filter backwash water treatment options
S. Arendze and M. Sibiya
ABSTRACT
Filtration acts as the final step in the removal of suspended matter and protozoa. The accumulated
residue is removed during the backwash process and any subsequent recycling of filter backwash
water could potentially re-introduce these contaminants into the main treatment process. By
separating the filter backwash water from the main treatment process, factors that could interfere
with the integrity of the primary treatment barriers, will be eliminated. Treatment and recovery of the
filter backwash water would be beneficial in terms of water reuse, by replacing a proportion of the
S. Arendze (corresponding author)
M. Sibiya
Rand Water,
Process Technology Department,
Barrage Road,
Vereeniging,
1930 Gauteng,
South Africa
E-mail: sarendze@randwater.co.za
freshwater demand. The aim of this study was to investigate possible treatment options for the filter
backwash water at Rand Water. Treatment options for filter backwash water treatment plants usually
consist of a solids removal process and a disinfection process. Three solid removal processes for
filter backwash water from Rand Water’s filtration systems were selected for testing on an
experimental basis: (1) sedimentation without flocculation, (2) sedimentation with flocculation, and
(3) dissolved air flotation with flocculation. Flocculation with sedimentation produced the best results
when compared to the other two treatment options evaluated. It is a simple and effective option for
the treatment of filter backwash water.
Key words
| dissolved air flotation, filter backwash water, flocculation, recycling, sedimentation,
water reuse
INTRODUCTION
The concept of multiple-barriers is the cornerstone of safe
effective removal of contaminants, including protozoa. Fil-
drinking water (LeChavallier & Au ). The presence of
tration acts as the last physical unit process for suspended
multiple barriers means that failure of one barrier is com-
solid removal in conventional treatment systems.
pensated by effective operation of the remaining barriers,
During the filtration process, most of the residual sus-
thus minimising the likelihood of contaminants passing
pended matter and micro-organisms that may have passed
through the treatment system (LeChavallier & Au ).
through the sedimentation process are trapped by the filter
The barriers include factors such as source protection,
media (LeChavallier & Au ). The accumulated residue
optimisation of the water treatment plant processes, such
is then removed during the backwash process. Filters are typi-
as coagulation, flocculation, sedimentation, filtration and
cally backwashed by flushing them with water in the reverse
disinfection, and a properly maintained distribution system
direction to normal flow. Compressed air may also be used to
(Betancourt & Rose ).
aid this process. The resulting water is termed waste or filter
The design of the water purification process should pro-
backwash water (United States Environmental Protection
vide for a multi-barrier system during which chemical
Agency (USEPA) a). Filter backwash water is therefore
treatment in the form of coagulation and flocculation pre-
characterised as having a high concentration of suspended
cedes physical processes such as sedimentation and
solid residues of coagulants and flocculants, metals, inorganic
filtration. The incorporation of the successive treatment
matter, algae (which could cause taste and odour com-
steps, if well designed and operated, would ensure the
pounds), bacteria, viruses, invertebrates, and protozoan
doi: 10.2166/wrd.2013.131
86
S. Arendze & M. Sibiya
|
Filter backwash water treatment options
Journal of Water Reuse and Desalination
|
04.2
|
2014
parasites, such as Cryptosporidium and Giardia (Linde ).
could go from a water stressed country to a water scarce
Of all the processes that produce residual streams in water
country (United Nations Environment Programme (UNEP)
treatment, filter backwash typically produces the largest
). Due to these challenges, there has been a need to
volume of water at the highest rate (USEPA b).
improve the efficiency of water consumption and the need
to supplement existing sources of water in a more sustainable
The necessity for the treatment of filter backwash
manner, thus water reuse serves to protect freshwater
water
resources, as the direct use of wastewater streams can replace
a proportion of the freshwater demand (UNEP ; Depart-
Recycling of filter backwash water reintroduces all the dele-
ment of Water Affairs South Africa (DWA) ).
terious matter that was removed by filtration, usually to the
head of works, and thus back into the main treatment
scheme (Linde ). This could have an effect on the
USEPA filter backwash recycling rule (FBRR)
water treatment chemistry, and consequently the final
Although there is a lack of formal regulation of filter backwash
water quality, as conventional water treatment technology
water in South Africa, the importance of its management and
does not provide consistent removal of pathogens if present
control is recognised and legislated internationally. The most
in high numbers (Wilf & Pearce ). Filter backwash
prominent example of such regulation is the FBRR of the
water may also reintroduce these contaminants in more con-
United States Environmental Protection Agency (USEPA
centrated form into the main treatment process. Studies by
b). The FBRR is intended to reduce the opportunity for
Arora et al. () have shown that levels of protozoa in
recycle practices to adversely affect the performance of drink-
filter backwash water are higher than those in raw water;
ing water plants (USEPA b).
Cryptosporidium concentrations were found to be 61 times
The FBRR affects all water purification plants in the
higher, and Giardia concentrations were 16 times higher.
USA that:
The presence of protozoan cysts in high numbers could be
•
detrimental, as this challenges the effectiveness of the
multi-barrier treatment capability, and this could thereby
affect public health (Arora et al. ).
The separate treatment of filter backwash water in dedicated plants would reduce the potential risk of reintroducing
contaminants into the main treatment process. This would
•
•
use surface water or ground water under the direct influence of surface water;
treat water using conventional filtration processes; and
recycle one or more of the following: spent filter backwash water, thickener supernatant, or liquids from
dewatering processes.
avoid having to upgrade the primary treatment technology
The FBRR requires that recycle streams pass through all
and thus save on the capital and operational costs that
the unit treatment processes of a treatment system before
would be required to treat the entire volume of water with
recycling, in order to minimise the risk of contaminants not
advanced processes.
being contained by the system on recycling (USEPA ).
A further benefit is that the current volume of filter backwash water at Rand Water, estimated at 2% of the volume of
raw water treated, if treated separately, would increase the
METHODS
capacity of the plant by this margin (Linde ). Treatment
and recovery of the filter backwash water would be beneficial
Treatment options for filter backwash water are similar to
in terms of water reuse. South Africa is currently character-
those for raw water treatment in water purification, and
ised as a country with high water stress, due to low rainfall
usually consist of a solid removal process and a disinfection
volumes (Adewumi et al. ). Population growth, the
process. Table 1 shows a study done by Cornwell et al. ()
expanding economy, and high evaporation rates due to cli-
on different solid removal treatment options that were com-
mate change are all putting pressure on limited water
pared at different treatment plants for the treatment of filter
resources in the country, thus in the near future the country
backwash water. The turbidity and particle removal
87
S. Arendze & M. Sibiya
Table 1
|
|
Filter backwash water treatment options
Journal of Water Reuse and Desalination
Log reduction in turbidity and particles from spent filter backwash water for
|
04.2
|
2014
Sedimentation with flocculation
different treatment options and their relative cost ranking (USEPA 2002a)
Turbidity log
Particle log
Relative
Jar stirring tests were performed on 1 litre samples of this well
Treatment process
reduction
reduction
cost ranking
mixed filter backwash water. A high energy jar test method
Sedimentation without
flocculant
0.1 to 0.8
0.2 to 0.9
1
was used, which included a settling time of 15 minutes. This
Sedimentation with
flocculant
1.4 to 2.3
1.9 to 3.3
2
DAF with flocculant
1.7 to 2.7
1.9 to 3.5
3
Granular media
filtration with pretreatment
2.2 to 3.0
2.4 to 4.4
4
Membrane microfiltration
2.6 to 3.9
1.6 to 3.5
5
high energy jar test method, in conjunction with settling, is
a laboratory scale simulation of the flocculation and sedimentation processes (g-values, mixing velocity and settling rates)
on the full scale plant. This procedure was developed around
the conditions at Rand Water’s full scale plants and was
developed based on comparative data. A cationic polyelectrolyte, a diallyldimethyl ammonium chloride and polyamine
and polyaluminium chloride blended product, was dosed,
and turbidity readings were taken by drawing supernatant
efficiency, as well as the relative costs of each are also
from each beaker with syringes after settling.
shown. Relative cost ranking is rated as 1 for the lowest
treatment cost to 5 for the highest treatment cost.
From those listed in Table 1, the following treatment pro-
DAF with flocculation
cesses were selected and evaluated on filter backwash water
Jar tests were again performed on 1 litre samples using a
from Rand Water’s Vereeniging treatment plant: (1) sedimen-
high energy jar test method, including 15 minutes of float-
tation
without
flocculation,
(2)
sedimentation
with
flocculation, and (3) dissolved air flotation (DAF) with floccu-
ing after the DAF was applied. The same cationic
polyelectrolyte was used as described above. As soon as
lation. The processes were chosen based on economic
stirring was complete, water saturated with dissolved air
viability, least complex methodology, potential to implement
(‘white water’) was added. The ‘white water’ was added
and acceptable turbidity removal. Investigations into disinfection practices were not done in this study.
at 10% of the 1 litre sample volume. The saturated water
was produced using a saturator. Figure 1 shows a schematic of the experimental set-up of the saturator. Figure 2
Sampling
shows the scum-float that was formed with the flocculated
Filter backwash water samples (100 litres) were sampled from
Rand Water’s Vereeniging treatment plant, while filters were
being washed. All the samples were then added to a large
200 litre tank, and mixed well using a submersible pump.
Sedimentation without flocculation
This method was tested to see if existing sedimentation
tanks could be used to settle the filter backwash water.
One litre samples were put into settling cones and left to
settle for 4 h, i.e. the approximate retention time in a
normal sedimentation tank, and longer, up to 12 h. Turbidity
readings were then measured. Different retention times were
tested on filter backwash water samples from different days,
due to the manual analysis involved.
Figure 1
|
The experimental set-up of the DAF saturator.
88
S. Arendze & M. Sibiya
|
Filter backwash water treatment options
Journal of Water Reuse and Desalination
|
Table 2
|
04.2
|
2014
Results for the sedimentation of filter backwash water without flocculation
Settling
period
Average
turbidity before
Average
turbidity after
Average percentage
decrease in turbidity
(h)
(NTU)
(NTU)
(%)
4
335
7.97
96.9
5
274
1.40
99.5
6
433
4.40
99.0
7
496
2.30
99.6
9
206
0.90
99.6
12
745
3.80
99.5
Sedimentation with flocculation
Table 3 shows the average turbidity results after settling;
using progressively increasing dosages of polyelectrolyte
on filter backwash water. The table also shows the average
percentage decrease in turbidity. The average initial turbidFigure 2
|
The scum-float that was formed with the flocculated particles once DAF was
applied.
ity was 385 NTU. The settled turbidity decreases with an
increase in polyelectrolyte concentration. A turbidity of
less than 5 NTU is an acceptable operational specification
matter once DAF was applied. Samples for turbidity
for settled water before filtration, thus a concentration of
measurements were drawn from taps on the side of the
3 mg/L is considered the optimum amount of polyelectro-
jars (the bottom tap on Figure 2), at 2 cm below the surface
lyte to be dosed to settle the filter backwash water. The
of the water, so as not to disturb the float. Jar test runs for
sedimentation with flocculation, and DAF with floccula-
average percentage decrease in turbidity is greater than
99% at dosages higher than 4 mg/L polyelectrolyte.
tion, were done on the same filter backwash water, on
the same day.
RESULTS
DAF with flocculation
The turbidity results for the treatment of filter backwash
water with increasing concentrations of polyelectrolyte in
Sedimentation without flocculation
Table 3
ity and percentage decrease in turbidity for the settling
periods listed. As mentioned before, different backwash
waters were used. The turbidity values before indicate the
variation of the different filter backwash water samples.
The average percentage decrease in turbidity for 4 h, i.e.
the approximate retention time in a normal sedimentation
tank was 96.9%. For periods greater than 5 h, the average
decrease in turbidity was greater than 99.0%.
Results for the sedimentation of filter backwash water with flocculation using a
polyelectrolyte
Table 2 shows the results from settling the filter backwash
water for periods of 4–12 h. It shows the average final turbid-
|
Concentration of
Average turbidity
polyelectrolyte
after settling
Average percentage
decrease in turbidity
(mg/l)
(NTU)
(%)
1
11.10
97.1
2
6.54
98.3
3
4.35
98.9
4
3.74
99.0
5
3.06
99.2
6
2.81
99.3
89
|
S. Arendze & M. Sibiya
Filter backwash water treatment options
Journal of Water Reuse and Desalination
|
04.2
|
2014
conjunction with DAF are shown in Table 4. The average
turbidity removal of 98.9%. The average percentage
percentage decrease in turbidity is also shown. The average
decrease in turbidity was greater than 99% at dosages
initial raw water turbidity was also 385 NTU. The final tur-
greater than 4 mg/L polyelectrolyte. The use of a chemical
bidity varied with polyelectrolyte dose. The optimum
coagulant produced better final turbidity values when com-
dosage point was found to also be at 3 mg/L, which gave
pared with settling alone.
a minimum turbidity of 11.5 NTU, and turbidity removal
DAF in conjunction with a flocculant produced a mini-
of 97.0%. In Figure 2, it can be seen that the scum-float
mum turbidity of 11.5 NTU at a dosage of 3 mg/L
that is formed is fairly thick which suggested that DAF
polyelectrolyte. This turbidity is higher than that of sedimen-
could be a viable option in the treatment of filter backwash
tation with a flocculant. This related to a turbidity removal
water.
of 97.0%. The turbidity removal from DAF did not go
below 5 NTU.
The average percentage decrease in turbidity from DAF
treatment at a flocculant dosage of 3 mg/L was 97.0%,
DISCUSSION
which is lower than that of sedimentation with a flocculant
Based on the results obtained, sedimentation without a flocculant showed good turbidity removal. The average
percentage decrease in turbidity was 96.9% for 4 h of
settling, the average settling time in a normal sedimentation
tank, and greater than 99% for more than 5 h of settling.
These values are indicative that the filter backwash water
consists of a high concentration of suspended solid residues
of coagulants which have already been destabilised and
settle very readily. Sedimentation without a flocculant
could be considered for the removal of solids, given adequate settling infrastructure and sufficient time.
Sedimentation with flocculation using a polyelectrolyte
showed decreased turbidity with increasing polyelectrolyte
dosage. Low turbidity values in the supernatant were
observed, with values less than 5 NTU at dosages of
3 mg/L and greater. A turbidity of less than 5 NTU was
achieved at a concentration of 3 mg/L, which related to a
which was 98.9% at the same dosage. This could be due to
factors inherent to small scale experiments, such as disturbance of the scum float while taking turbidity samples.
However it could be that the particle settling rate was
higher than the rise rate, thus some of the flocculated
material did not float.
In terms of turbidity removal, the methods tested rank,
from best, as follows:
1. flocculation with sedimentation;
2. sedimentation without flocculation; and
3. DAF with flocculation.
Flocculation with sedimentation produced the best
results compared to the other two treatment options evaluated. It is a simple option, yet could be just as effective as
the more complicated and expensive treatment options
that are available. After dosing with polyelectrolyte and
settling, water with a turbidity of 4.35 NTU was obtained,
Table 4
|
Results for the flocculation of filter backwash water using a polyelectrolyte in
conjunction with DAF
this adheres to a turbidity of 5 NTU after sedimentation.
Based on these preliminary results, the recommended
Concentration of
Average turbidity
Average percentage
polyelectrolyte
after settling
decrease in turbidity
(mg/l)
(NTU)
(%)
1
16.65
95.7
2
16.00
95.8
3
11.50
97.0
4
12.30
96.8
5
13.60
96.5
6
13.00
96.6
separate treatment for filter backwash water is as follows.
Cationic polyelectrolyte can be used as a coagulant, to flocculate particles. The concentration however will have to be
optimised through regular jar tests, as the filter backwash
water quality will change depending on season and changes
in raw water quality. After settling, water with a turbidity of
less than 5 NTU will be obtained. This water could possibly
be disinfected with ultraviolet light (UV) and fed back into
the main stream to enter the system just prior to filtration.
90
S. Arendze & M. Sibiya
|
Filter backwash water treatment options
Journal of Water Reuse and Desalination
|
04.2
|
2014
A separate dedicated filtration plant will therefore not be
2. sedimentation without flocculation, and 3. DAF with
necessary.
flocculation. Flocculation with sedimentation produced
The incorporation of the filter backwash water will
the best results compared to the other two treatment
increase the filtration rate in each filter by 2%, if all filters
options evaluated. It is a simple option, yet could be
are in commission; however it could be more if filters are
just as effective as the more complicated and expensive
being backwashed or if a filter is out of commission. This
treatment options that are available. This water could
should have little effect on filtration in the main stream
then possibly be disinfected with UV, and then fed back
(Letlape ).
into the main stream to enter the system just before fil-
The treated filter backwash water will then undergo primary disinfection with chlorine along with the rest of the
outgoing, treated water. This will decrease the raw water
intake by about 2%, with minimal capital expenditure
tration, which may have little effect on filtration in the
•
apart from a dedicated UV process.
main stream.
Dedicated filter backwash water treatment plants are also
a financially sound option, as the need to upgrade the
entire primary treatment technology would be avoided;
Recycling of filter backwash water could challenge the
the concentration of contaminants in the filter backwash
effectiveness of the multi-barrier treatment capability, by
water could affect the raw water quality through recy-
reintroducing these contaminants in more concentrated
cling, thus costs to treat the entire volume of water with
form, which could affect public health. The separate treat-
advanced processes would be evaded. Water treatment
ment of filter backwash water in a dedicated plant would
plants should incorporate filter backwash water treat-
reduce the potential risk of reintroducing contaminants
ment plants into their budgets.
into the main treatment process. This would also be a financially sound option in the long run, as this would avoid
having to upgrade the entire primary treatment technology,
•
Treatment of filter backwash water is also a benefit to the
environment through the concept of water reuse, as a proportion of the freshwater demand will be replaced.
due to concentrated contaminants found in a waste stream;
this would save on the capital and operational costs that
would be required to treat the entire volume of water with
advanced
thus
look
processes.
at
Water
incorporating
treatment
filter
plants
backwash
should
water
treatment plants into the budget. Lastly treatment of filter
backwash water will also add value in terms of the
environment through water reuse, as it would serve to
augment the supply and thus the freshwater demand will
be reduced.
CONCLUSIONS
•
Recycling of filter backwash water could challenge the
effectiveness of the multi-barrier treatment capability,
which could eventually affect public health. The separate
treatment of filter backwash water is important as it
would reduce the potential risk of reintroducing contami-
•
nants into the main treatment process.
In terms of turbidity removal, the methods tested rank,
from best, as follows: 1. flocculation with sedimentation,
REFERENCES
Adewumi, J. R., Ilemobade, A. A. & Van Zyl, J. E.  Treated
wastewater reuse in South Africa: Overview, potential
and challenges. Resources, Conservation and Recycling 55,
221–231.
Arora, H., Di Giovanni, G. & LeChavellier, M.  Spent Filter
Backwash Water: Contaminants and Treatment Strategies.
AWWA, Denver, Colorado, USA.
Betancourt, W. Q. & Rose, J. B.  Drinking water treatment
processes for removal of Cryptosporidium and Giardia.
Veterinary Parasitology 126, 219–234.
Cornwell, D., MacPhee, M., McTigue, N., Arora, H., DiGiovanni,
G., LeChevallier, M. & Taylor, J.  Treatment Options for
the Removal of Giardia, Cryptosporidium, and other
Contaminants in Recycled Backwash Water. AWWARF,
Colorado, USA.
DWA – Department of Water Affairs, South Africa 
Development of a Reconciliation Strategy for the Olifants
River Water Supply System – Future Water Reuse and other
marginal Water use Possibilities. Report number: P WMA
04/B50/00/8310/4, DWA, Pretoria, South Africa.
LeChavallier, M. W. & Au, K.  Water Treatment and
Pathogen Control: Process Efficiency in Achieving Safe
91
S. Arendze & M. Sibiya
|
Filter backwash water treatment options
Water. World Health Organisation (WHO), IWA Publishing,
London, UK.
Letlape, M.  Unpublished Evaluation of Increased Filtration
Rates on Filtrate Quality and Total Volume of Water
Produced Per Filter Cycle. Rand Water, Process Technology,
Vereeniging, South Africa.
Linde, J. J.  Evaluation of A Filter Backwash Recovered Plant
to Establish Guidelines for Design and Future Operation.
WRC, 920/1/03, Pretoria, South Africa.
UNEP – United Nations Environment Programme  Water and
Wastewater Reuse. Division of Technology, Industry and
Economics, UNEP, Japan.
UNEP – United Nations Environment Programme  Water
Profile: Republic of South Africa. UNEP, Paris, France.
Journal of Water Reuse and Desalination
|
04.2
|
2014
USEPA – United States Environmental Protection Agency 
National Primary Drinking water; Filter Backwash Recycling
Rule; Final Rule. USEPA, USA.
USEPA – United States Environmental Protection Agency a
Filter Backwash Recycling Rule: Technical Guidance
Manual. Office of Groundwater and Drinking Water,
USEPA, USA.
USEPA – United States Environmental Protection Agency b
Filter Backwash Recycling Rule: Technical Guidance
Manual. Office of Groundwater and Drinking Water,
USEPA, USA.
Wilf, M. & Pearce, G.  Reclamation of sand filter backwash
effluent using capillary ultra-filtration. Membrane Technology
2003 (3), 6–9.
First received 12 July 2013; accepted in revised form 8 November 2013. Available online 16 December 2013