You are on page 1of 11

Review

Received: 3 September 2015 Revised: 23 November 2015 Accepted article published: 26 November 2015 Published online in Wiley Online Library: 21 December 2015

(wileyonlinelibrary.com) DOI 10.1002/jctb.4860

Applications of biofiltration in drinking water


treatment – a review
Onita D Basu,a* Sahil Dhawana and Kerry Blackb

Abstract
Biofiltration is a process in which an otherwise conventional granular filter is designed to remove not only fine particulates
but also dissolved organic compounds through microbial degradation. Biofiltration can reduce the need for chemicals in
drinking water treatment and thus improved applications of biofiltration in drinking water treatment can be viewed as green
or sustainable engineering technology. Recent trends in biofiltration technology for drinking water treatment have or have
attempted to extend the performance of biofilters through gaining a better understanding of operational constraints. This
review articles summarizes important operational parameters influencing biofiltration performance such as hydraulic loading,
empty bed contact time (EBCT), temperature, media type, and backwashing conditions. In addition, recent advancements
in biofiltration operations including, ozonation, ammonia removal and the influence of nutrient (nitrogen, phosphorous)
supplementation to facilitate carbon removal are explored.
© 2015 Society of Chemical Industry

Keywords: biofiltration; temperature impacts; hydraulic loading rate; filter acclimation; nutrient supplementation

INTRODUCTION organic matter prior to disinfection, and thereby helps minimize


Filtration is a key component to successful treatment of surface both the production of DBP and reduces the occurrence of bio-
water for the production of drinking water. Filtration is the removal logical regrowth in the distribution system; as such it has gained
of particles from the water phase by passing the water through a prominence in North America since the 1980s.6
granular or porous media and can be subcategorized as rapid gran- Given the growing prominence of biofiltration in drinking water
ular filtration, slow sand filtration and biological filtration. Rapid treatment, a recent review paper is needed. Several papers and ref-
granular filtration focuses on the removal of suspended particles erences on biofiltration have been published in the past, and the
whereas slow sand filtration and biological filtration remove both reader is directed to them for further information.3,5 – 9 However, a
suspended and dissolved organic particles from the water phase. recent review paper focused on drinking water treatment biofiltra-
One of the first filtration systems installed for a centralized system tion with the inclusion of important design parameters is needed.
This paper will also examine impacts of cold temperatures, oxida-
was in Paisley, Scotland in 1804 which was used to treat water for
tion pretreatment, biofiltration for membrane fouling control, as
both industrial purposes and to supply water to the local popula-
well as the capacity of biofilters to remove nitrogen compounds,
tion; at the time cleaned water was delivered by cart to residents.1
and the potential influence of available phosphorous on biofilter
In North America filtration was first implemented at Poughkeep-
performance.
sie, NY, USA in 1872.2 Early filtration systems utilized slow sand
filtration practices, which as the name may imply, have very low
filtration rates (ranging from approximate initial historical lows of
BIOFILTER DESIGN
0.1 m h−1 to more modern averages of 0.4 m h−1 ); rapid granular
Media selection for biofiltration
filtration has much higher filtration rates, ranging from approxi-
While conventional filters are commonly composed of numerous
mately 5 to 20 m h−1 .
types of granular material including sand, granular activated car-
Conventional water treatment processes typically involve some
bon (GAC), anthracite and expanded ceramics, biofiltration may be
level of coagulation and flocculation followed by filtration and
ideally suited with all or at least part of the media bed composed
disinfection.3 The first two processes can help remove organic
of GAC. GAC has been used extensively in drinking water treat-
matter in the water, however, the latter two do not. Residual ment processes due to its adsorptive capacity to remove specific
organic matter in the water combines with disinfectants to form
disinfection by-products (DBP) which have been shown to have
carcinogenic properties. Conventional treatment processes expe- ∗ Correspondence to: Onita D. Basu, Department of Civil and Environmental
rience difficulty in removing organics and disinfection by-product Engineering, Carleton University, Ottawa, ON, Canada.
precursors.4 In addition, residual organic matter that exits a drink- E-mail: onita.basu@carleton.ca
ing water treatment facility may cause biological regrowth in the
a Department of Civil and Environmental Engineering, Carleton University,
distribution system as a result of biodegradable organic matter Ottawa, ON, Canada
(BOM), NH4 + , Fe2+ , Mn2+ that may have been left untreated.5 Biofil-
585

tration, has the capacity to remove both suspended and dissolved b Department of Civil Engineering, University of Guelph, Guelph, ON, Canada

J Chem Technol Biotechnol 2016; 91: 585–595 www.soci.org © 2015 Society of Chemical Industry
www.soci.org OD Basu, S Dhawan, K Black

Table 1. BOM removals at 5 ∘ C and 20 ∘ C (adapted from Ref. 14)

GAC/Sand media (HLR = 7.5 m h−1 ; EBCT = 5.6 min) Anthracite/Sand media (HLR = 7.5 m h−1 ; EBCT = 5.6 min)
BOM Component % Removal 5 ∘ C % Removal 20 ∘ C % Removal 5 ∘ C % Removal 20 ∘ C

Acetate 85 90 45–85 85–90


Formate 85 90 30–80 90
Formaldehyde 85 95 95 10–85
Glyoxal 65–80 79 75–80 10–55

contaminants from solution. Conventionally, once the adsorptive surface for microbial growth and attachment that is less sensitive
capabilities of the GAC are exhausted, the media is removed and to changes in water temperature.
replaced with virgin GAC. However, once the adsorptive capabili-
ties of the GAC are exhausted, the media may be maintained in the Empty bed contact time (EBCT) and hydraulic loading rate
bed if the filter is intended to operate as a biofilter in which case it (HLR)
may be labeled biologically activated carbon (BAC) filtration. BAC is Empty bed contact time (EBCT) has been accepted as one of
a filtration system in which granular activated carbon (GAC) is used the single most common parameters for helping predict removal
as a growth medium to support beneficial microorganims, rather of organic matter in biofilters, although there is some discrep-
than for adsorption.10 GAC supports more dense microbial popu- ancy on the time frame involved which may indicate that other
lations than sand or anthracite (i.e. 4 to 8 times more biomass per operational parameters still need investigation to help predict
gram of media), due to a combination of factors including poros- removal of organic matter in biofilters. For instance, many ear-
ity, surface area, surface roughness, surface charge and adsorption lier studies found that removal of organic matter is directly pro-
capacity.11 portional to EBCT.21 – 23 LeChevallier et al.21 reported increases in
Presumably the higher microbial density will translate into TOC removal from approximately 30 to 50% when the EBCT was
improved dissolved organic carbon (DOC) removal within a biofil- increased from 5 to 20 min. However, other research has shown
ter. For instance, Thiel et al.12 conducted a pilot scale study to com- that the EBCT had little effect on TOC removal so long as the EBCT
pare the performance of GAC and anthracite. At an empty bed range was between 4 and 20 min.24 Similar optimal EBCT ranges
contact time (EBCT) of 8 min, DOC removals by GAC varied from have also been reported.17,25 Work by Hozalski et al. found that
approximately 11% to 14% whereas with anthracite DOC removals decreased operating temperatures resulted in increased empty
varied from approximately 1–3%. At an EBCT of 16 min, DOC bed contact time requirements to achieve steady state.26 Wert
removals by GAC varied from approximately 15–20%, whereas et al.27 studied the impact of EBCT on AOC removal with two par-
removals by anthracite varied from approximately 2% to 7%. In a allel anthracite/sand biofilters with EBCT of 3.2 min and 8.3 min,
full scale study, Emelko et al.13 found that at warmer temperatures respectively. Both columns demonstrated similar levels of removal
and in both cases, AOC removal stabilized at approximately 60%
(21–24 ∘ C), GAC and anthracite showed similar TOC removals,
after 175 days of operation.
however, at colder temperatures (1–3 ∘ C), median TOC removals
Hydraulic loading rate (HLR), also referred to as surface load-
were 23% by GAC filters and 14% by anthracite filters.
ing rate, is one of several parameters affecting biofiltration. How-
A bench scale study conducted by Liu et al.14 investigated the
ever, its exact impacts on biodegradation and removal still require
effects of media type and chlorinated backwash on biological
research. Early work by Galvis et al. found that removal efficiencies
organic matter removal. It was found that GAC/sand filters showed
varied widely depending on the influent loading rate and source
similar high levels of BOM removal at 5 ∘ C and 20 ∘ C whereas
water characteristics.28 Carlson and Amy22 found that the removal
the anthracite/sand filters demonstrated a much more variable
of DOC decreased as the filter loading rate increased. Carlson and
performance. A summary of removals obtained in this study has Amy29 further concluded that biomass utilization as a function
been provided along with a comparison with anthracite media in of EBCT did not depend on the HLR, indicating ‘that EBCT is an
Table 1. acceptable biofilter design parameter throughout a range of HLRs
Basu and Huck15 studied TOC removal by biofiltration at lab and filter bed depths.’ It is further possible that decreased removal
scale with anthracite/sand columns and an EBCT of 22 min, with of DOC at higher loading rates may be due to poor filter acclima-
TOC removal varying from 10 to 36%, with a water composition tization, suggesting that the available biomass generally acclima-
comprised of humic acids (65%) and readily biodegradable car- tizes at a lower hydraulic loading rate and were unable to assimi-
bons (35%). Price et al.16 reported removals of 17% and 9% for late the carbon as efficiently at the higher loading rate without a
GAC and GAC/sand dual media filters, respectively, at EBCT of sufficient new time period allowed for the biomass to re-acclimate
10 min. Wang et al.17 reported average removals of 29%, 16% and to the new biofilter conditions.29,30 In their study, Melin and Ode-
20% when using GAC, dual media filters and sand, respectively, at gaard found that the removal of oxidation by-products was not
EBCT of 9.2 min. Klevens reported 12% and 14% removal of TOC directly affected by hydraulic loading rate, but rather by influent
using BAC, when operated with 5 and 10 min EBCT, respectively.18 concentration and EBCT.30 They found that the removal of these
Seredynska-Sobecka et al.19 observed higher removal of humic oxidation by-products varied greatly. Melin and Odegaard sug-
acids during combined ozonation and biofiltration than with gested that the biomass and bacteria present at high loading rates
biofiltration alone, using GAC. Fan et al.20 found 54% removal of may prefer more easily biodegradable compounds.30 Cleary con-
humic acid using GAC, in a landfill leachate study. Overall, while cluded that the hydraulic loading rates were a critical parameter in
biofiltration can occur with several media types, it has generally achieving good effluent quality, particularly in a less biologically
586

been demonstrated that GAC provides a more robust media mature system.31

wileyonlinelibrary.com/jctb © 2015 Society of Chemical Industry J Chem Technol Biotechnol 2016; 91: 585–595
Review of water treatment biofiltration www.soci.org

Liu et al.14 studied the effects of hydraulic loading rate on the temperature, backwash, media type and type of organics which
removal of conventional pollutants (AOC, NH3 -N, TP, etc.) and resulted in filter acclimation periods ranging from 20 to 140 days.
microbial communities within a biofilter. Six biofilter units were Lower temperatures in particular took longer times for removal
operated in parallel. Each column had an effective height of with anthracite demonstrating times >140 days whereas GAC was
100 cm with an internal diameter of 10 cm and contained 2.3 L of 80 days at lower temperatures.
GAC and 1.1 L of sand. The columns were started on a hydraulic Although filter acclimation times can vary significantly it is
loading rate of 1 m h−1 (EBCT = 3 h). After 35 days of operation, important to note that GAC filters first operate under adsorption,
hydraulic loading in columns 2–6 was increased to 1.5 m h−1 and then transition to biofiltration as the adsorptive capacity is
(EBCT = 2 h), 2 m h−1 (EBCT = 1.5 h), 3 m h−1 (EBCT = 0.97 h), reached. Therefore it may be possible to maintain a high removal
5 m h−1 (EBCT = 35 min) and 8 m h−1 (22 min), respectively. of DOC during biofiltration colonization phase.3
During the initial 35 days, removal of AOC, NH3 -N and TP Some studies indicate that the amount of biomass decreases
remained fairly constant at 84.5% ± 0.5%, 73.6% ± 0.7% and with increasing depth and that the greatest level of removal occurs
76.9% ± 0.5%, respectively. As the HLRs were increased, a sharp at the top of the biofilters.17,35 Others demonstrated biomass
decline was observed in the removal of these contaminants which development through the biofilter column, for instance, Velten
was followed by a trend of increasing removal within 7 days with et al.36 measured biomass development throughout the depth of
steady state being reached within 14 days. The filter operating an O3 /BAC filter from start-up with virgin GAC media. Initially, 80%
at 3 m h−1 achieved highest AOC removal (85.2%) whereas high- of the influent DOC was removed through the filter column, which
est NH3 -N and TP removal 84.4% and 87.1%, respectively, were was directly attributed to the adsorptive capacity of the virgin
observed in the filter operating at 2 m h−1 . As HLR increased the GAC media by the study authors. This was followed by a transition
biomass concentration was noted to increase as well. Removals phase of decreasing DOC removal which then leveled off at 20%
for various pollutants ranged from 80% (HLR = 1 m h−1 ) to as low removal of DOC after 90 days which is ascribed to the resultant
as 20% (HLR = 8 m h−1 ); however, it should be noted that in all biomass development The authors studied biomass accumulation
cases as the HLR was adjusted the EBCT was altered as well. throughout the depth of the bed which reached steady state after
Using polymerase chain reaction-denaturing gradient gel elec- a period of 90 days and coincided with the steady state DOC
trophoresis (PCR-DGGE), the authors were able to determine the removal observed in the columns These steady state values ranged
dominant bacterial species in the columns. Shigella sp. was found from 1.17 (±0.2) × 10−6 gATP g−1 GAC at a depth of 10 cm versus
in all the biofilters. At low HLRs (1–3 m h−1 ), E. fergusonii was found 0.8 (±0.2) × 10−6 gATP g−1 GAC at a depth of 115 cm. Similarly,
to be more dominant than in the columns with higher HLRs. The E. Boon et al.37 monitored DOC removal per ATP count throughout
coli bands in the DGGE analysis for the filters operating at 1 m h−1 the depth of biofilter columns at 10, 45, 80, and 115 cm.
and 1.5 m h−1 were denser than those operating at higher HLRs.
Ko et al.32 conducted a study on a pilot scale system with 3
dual media GAC/sand biofilter columns operated in parallel with Backwashing
pre-ozonation (0.75 mg O3 mg−1 DOC). The filters each contained Proper backwashing influences the performance of biological
2 m of GAC followed by 20 cm of sand. HLRs for three filters were filters through various mechanisms, including detachment and
24 m h−1 (EBCT = 5 min) for filter 1, 12 m h−1 (EBCT = 10 min) for removal of biomass (bacteria, protozoa and other microorgan-
filter 2 and 6 m h−1 (EBCT = 20 min) for filter 3. Eight sampling ports isms), redistribution of media and associated fixed biomass,
were present on the filters, 25 cm apart. The removals for DOC adverse impacts of potential oxidants in the backwash water, and
were 31%, 26% and 19% for filters 3, 2 and 1, respectively. This elimination of accumulated particles. Chlorine and air scour can
would indicate a higher EBCT and lower HLR are advisable for DOC be implemented to supplement regular backwashing procedures.
removal. Biomass analysis was conducted on the media obtained Emelko et al.13 published a comprehensive paper examining the
from the eight ports on each filter. An HLR of 12 m h−1 in filter effects of air scour and subfluidized backwash on TOC and oxalate
2 gave the maximum biomass concentrations (75 nmol PO4 − g−1 removal at both warmer temperatures (21–24 ∘ C) and colder
GAC to 120 nmol PO4 − g−1 GAC) compared with 24 m h−1 in filter 1 temperatures (1–3 ∘ C), GAC/sand filters and anthracite sand filters
(70 nmol PO4 − g−1 GAC to 130 nmol PO4 − g−1 GAC) and 6 m h−1 in showed no significant change in TOC removal with and without
filter 3 (42 nmol PO4 − g−1 GAC to 100 nmol PO4 − g−1 GAC). air scour. Both TOC and oxalate removal remained unaffected by
collapse pulsing. The backwash conditions are summarized in
Table 2.
Filter acclimation and filter depth Liu et al.14 showed no significant impact on BOM removal for
Acclimation refers to the operation of a biofilter under steady backwash water chlorinated with 0.5 mg Cl2 L−1 at 20 ∘ C for
state conditions for particular contaminant removals it generally both GAC and anthracite/sand biofilters, although glyoxal removal
implies when removal of a DOC (or BOM for instance) has reached decreased from70% to 50% observed in the anthracite/sand filter
a maximum steady state removal value. Development of biomass only when exposed to chlorine versus no chlorine in the backwash
is important as some research has shown that removal of DOC while the GAC filter was unaffected. Similarly, at 5 ∘ C, a detrimental
during biodegradation is limited by biomass concentration or BOM effect was observed on BOM removal for the anthracite/sand biofil-
formation and not filter operating parameters.29 ter with chlorine (0.5 mg L−1 ) versus without chlorine; for example,
Acclimation periods required to reach steady state can vary glyoxal removal was 55% without chlorinated backwash water as
widely and depend largely on source water characteristics, opposed to 11% with chlorinated backwash water. Thus it appears
media selection and temperature. Acclimation periods have been that GAC biofilters are more robust to changes in both temperature
reported to range between 20 days and >16 months.14,15,17,25,33,34 and backwash conditions than anthracite/sand biofilters.
For instance, Basu and Huck15 demonstrated that for waters with Wert et al.27 examined the effects of various backwash proce-
a high humic content acclimation was close to 16 months, while dures (dechlorinated backwash water versus chlorinated water
587

Liu et al.14 demonstrated that filter acclimation was impacted by with air scour) and varying filtration rates (4.8–14.6 m h−1 ) on

J Chem Technol Biotechnol 2016; 91: 585–595 © 2015 Society of Chemical Industry wileyonlinelibrary.com/jctb
www.soci.org OD Basu, S Dhawan, K Black

at lower operating temperatures.23,26,40 – 44 At colder temper-


Table 2. Backwash procedure details (adapted from Ref. 13)
atures, filters exhibit decreased microbiological activity and
Backwash protocol Details therefore exhibit decreased removal efficiencies. However,
some research has found limited impact on biofiltration at cold
Standard (Air scour) Air scour: 2 min; 3 scfm ft−2 temperatures.13,14,34,45
Settling time: 1 min Specifically, reductions in the removal of E. coli were impaired at
Chlorinated low rate wash: 9 min; 10.7 m h−1 lower temperatures, resulting in a decreased reduction of E. coli by
Chlorinated high rate wash: 11–12 min; 50% when influent temperatures were reduced from 20 to 2 ∘ C.40
42.7 m h−1
Similarly, Bellamy et al. found reduced removal of total coliforms
No air scour Chlorinated low rate wash: 9 min; 10.7 m h−1
when influent temperatures were reduced to 5 ∘ C, but found that
Chlorinated high rate wash: 11–12 min;
42.7 m h−1 removal of Giardia remained constant.41 Work by Huck et al. found
Collapse pulsing Chlorinated low rate wash: 9 min; 10.7 m h−1 that in the case of biofiltration, removal of organic matter from the
Subfluidization wash with air scour; 7 min; source water was significantly hindered at water temperatures that
water wash - 7.3 m h−1 ; air scour 3 scfm ft−3 dropped below 5 ∘ C.23 Similar results were observed by Hozalski
et al.26 In contrast, work by Melin et al. has shown that significant
biological activity can be achieved despite low temperature or
BOM removal. BOM (measured as AOC) was unaffected by back- nutrient deprived conditions.34 Their study also showed good
wash procedures and filtration rates. A 60% reduction in AOC was removal of colour and TOC, as well as 50% removal of BDOC during
observed after 165 days of filter operation. Formaldehyde removal filter operation under cold temperatures. In this case, raw water
was also monitored and not impacted with backwash conditions, temperatures were around 2–4 ∘ C during the winter months, and
formaldehyde removal reached 70% removal after 165 days of remained below 10 ∘ C during the summer months.
operation. Work by Wulfeck and Summers44 found a 34% decrease in TOC
Hozalski et al.26 observed that backwashing had very little effect removal and a 59% decrease in removal of disinfection by-product
on removal of biological organic matter, until the amount biomass precursors for a full-scale biofilter operated at less than 14 ∘ C,
lost during backwashing exceeded 60%. Their study also found compared with the performance at temperature above 14 ∘ C.
that filter acclimatization was relatively insensitive to backwashing Nishijima et al.45 showed that DOC removal was unaffected by
so long as the biomass loss during backwashing did not exceed temperature as long as the EBCT was above 24 min. Their work
60%. They concluded that backwashing did not significantly affect also showed that despite temperature changes on the order of
biofilter performance.26 Amburgey38 proposed extended terminal 5 to 30 ∘ C, seasonal changes in the removal of DOC and THMFP
subfluidizaation wash (ETSW) to backwash both biologically active were not observed. Mofidi et al. found that filter start-up time was
and conventional filters. ETSW is a process to backwash filters significantly longer for cold temperature filters.46 Their work found
under subfluidization conditions at the end of a regular fluidized that filters started in warmer temperatures achieve 64% removal
backwash to minimize the particle ripening stage of filters. after 3 months, versus 70% removal after 5 months for colder
In a study conducted by Liao et al.39 the impacts of backwashing temperature start-ups. Research performed by Moll et al. found
on DOC, dissolved organic nitrogen (DON) and N-nitrosamine pre- that biofilters operated at 5 ∘ C showed decreased biofiltration
cursor removal through a GAC filter system were investigated. The performance compared with operation at 20 and 35 ∘ C.47 In their
study also looked at the impact of backwashing on microbial com- work, AOC levels decreased on average by 23%, while removal of
munity structure and biomass concentrations. The study made use some DBPs decreased by >50%.
of a two-stage downflow BAC filtration system. Feedwater con- Work by Liu et al. found that temperature was an important fac-
sisted of lake water pretreated with pre-ozonation, rapid mixing, tor affecting steady-state removal of biological organic matter.14
flocculation, sedimentation, sand filtration and post ozonation and Further, their work found that biological filtration performed favor-
was fed at a hydraulic loading rate of 8 m h−1 . Before the study, ably at temperatures as low as 5 ∘ C provided that backwashes
the filters had been operating continuously for 18 months. Back- were performed in the absence of chlorine. In addition, GAC vs
washing consisted of a 2 min air scour (61 m2 s−1 ) followed by anthracite filters were found to reach steady state faster (20–80
a dechlorinated water only wash (81 m2 s−1 ) for 5 min and sam- days vs 20–140 days). Fonseca and Summers48 found that lower
pling took place over 5 days. Filter influent and effluent were col- operating temperatures decreased the overall removal of DOC,
lected immediately prior to backwash, immediately after back- and BDOC, but that there was no significant effect on the removal
wash and 2, 5 days post-backwash. At this time, GAC samples from of THM and HAAs. In addition, their work found that while tem-
the first downflow filter were also collected. Analysis indicated that perature affected the activity of the biomass, the concentration of
biomass concentrations decreased from approximately 32 nmol-P biomass remained relatively unchanged. Work by Emelko et al.13
g−1 dry media before backwashing to 15 nmol-P g−1 dry media found that neither GAC nor anthracite was significantly affected by
immediately after backwashing. This was followed by a recovery to temperature when TOC removal was investigated. GAC however,
approximately 25 nmol-P g−1 dry media 5 days after the backwash. consistently exhibited better TOC removal than anthracite under
Table 3 indicates the impact of backwashing on DOC, DON and cold temperature conditions.
N-nitrosamine precursor removal, further the data indicates that Persson et al.43 found that BDOC removal continued steadily at
it is possible that backwashing can have a transitionally favorable cold water conditions of 5 ∘ C. Their study also found no correlation
impact on DOC, DON and N-nitrosamine precursor removal. between AOC removal and temperature, suggesting that AOC
removal is independent of raw water temperature. However, their
research did find that the specific activity of biofilter biomass was
TEMPERATURE EFFECTS significantly impacted by temperature.43 Jasim et al.49 conducted
The majority of research has demonstrated that filtration per- a pilot scale study to investigate the effect of ozone on cold
588

formance with respect to biodegradation of organics decreases water coagulation. Treatment occurs through two parallel trains

wileyonlinelibrary.com/jctb © 2015 Society of Chemical Industry J Chem Technol Biotechnol 2016; 91: 585–595
Review of water treatment biofiltration www.soci.org

Table 3. Approximate percentage parameter removals immediately before, immediately after, 2 days and 5 days after backwash (adapted from
Ref. 39)

Parameter removal (%)


Parameter Influent range (mg L−1 ) Before backwash After backwash Day 2 Day 5

DOC 3.11 - 3.2 31.3 42.1 35 26.3


DON 0.15–0.2 25.3 64 39.6 19.6
N-nitrosamine precursor 100–130 × 10−6 25 62 40 20

with two filters: anthracite/sand followed by GAC/sand. In each


Table 4. Removal efficiencies for oxidation processes involving
train, pretreatment occurs through coagulation and flocculation. ozone*
An ozone dose of 1–2 mg L−1 was applied to train 2 only. Train 2
witnessed lower particle counts and lower turbidity for the filter Removal efficiency
effluent than train 1. Characteristic Ozone oxidation* Ozone + biofiltration
Limited research is available on the effect of sustained tempera-
ture regimes of less than 5 ∘ C; more recent research indicated that Colour 52–74% n/a
the impact of low temperatures can be somewhat mitigated by Dissolved Organic Carbon 2–33% 14–50%
appropriate operating conditions. Further research is needed to Biodegradable dissolved Increase of 5 increase of 10 to 55%
determine how to mitigate the potential adverse impacts of cold organic carbon (BDOC) to 50%
temperatures on biofiltration in order to ensure year round perfor- UVA (254 nm) 28–75% 60–80%
mance of biofiltration columns. THM precursors 5–55% 40–80%

*Adapated from Refs 71, 72, 50, 26, 73, 55, 74 and 54.

OXIDATION PRETREATMENT
Combined oxidation and biofiltration systems have the potential
Zhang et al.67 sought to compare O3 /BAC and GAC (conventional
to result in the production of biologically stable water through
non-biologically active filter) processes for organics removal. Both
the removal of biodegradable organic matter (BOM), this has
processes showed similar removals of TOC (33.3% and 21.1%,
the benefit of minimizing the potential for bacterial regrowth
respectively). However, upon varying the ozone dose from
within the distribution system as well as removing disinfection
2 mg L−1 to 8 mg L−1 , the O3 /BAC process showed significantly
by-product precursors, reducing chlorine demand, and the poten-
higher removals of AOC. At an ozone dose of 3 mg L−1 , the AOC
tial removal/control of oxidation by-products.50,51
dropped from 380.9 μg acetate-C L−1 to 40.2 μg acetate-C L−1 in
the O3 /BAC versus 117.6 μg acetate-C L−1 in GAC.
Ozone A summary of the reported effects of ozonation and ozona-
During ozonation, natural organic matter (NOM) is oxidized and tion + biofiltration is provided in Table 4. This table represents a
transformed into intermediates still present as DOC.52,53 There- summary of the work by several researchers for ozone doses rang-
fore, very little reduction in DOC is observed during low-dose ing from 1 to 5 mg O3 mg−1 DOC.
ozonation. Typically, a shift from humic to non-humic, and Disadvantages to the use of ozone are the formation of ozone
high to low molecular weight NOM has been observed during by-products (OBPs) such as aldehydes (formaldehyde, acetalde-
ozonation.52,54 The reaction of ozone with the aromatic structures hyde, glyoxal, methyl glyoxal, etc.), ketoacids and carboxylic
and double bonds of NOM results in a significant decrease in UV acids.58,75 – 78 However, most ozonation by-products are highly
absorbance.4,55,56 For instance, one study found than an ozone biodegradable and can be removed through biofiltration prior to
dosage of 3 mg L−1 decreased UV260 absorbance by 50% [4], releasing the water into the distribution system, which is one rea-
another study reported a 50% decrease in UV254 at a 0.5 mg ozone son why ozone/biofiltration are complimentary processes.79,80
mg−1 DOC dosage and this value increased to 78% at a 1.5 mg For instance, the formation potentials of THMs, HAAs, and
ozone mg−1 DOC dosage.55 chloral hydrate were reduced by as much as 70–80% using
Ozonation can occur via two pathways; direct or indirect. The ozonation–biofiltration treatment.63
direct pathway favours reactions primarily with unsaturated Fonseca et al.81 compared microbial activity in biofilters treat-
double bonds and aromatic compounds, though reaction with ing ozonated water versus those treating non-ozonated water:
amines or sulphides being common.57,58 With the indirect path- three separate biofilter systems each containing three sand fil-
way, hydroxyl radicals are formed and react with NOM. Hydroxyl ters in series were set up and consisted of intermediate oxida-
radicals are strong, unselective oxidants. At higher pH, the indirect tion (1.3 ± 0.52 mg O3 mg−1 DOC), pre-oxidation (1.3 ± 0.35 mg O3
pathway is favored.59 mg−1 DOC) and no oxidation, respectively. For the pre-oxidation
Extensive work has shown that pre-ozonation doses of 1 to set of columns, analysis was carried out over two temperatures
2 mg O3 mg−1 TOC were optimal for enhancing biodegradation (12 ∘ C and 3 ∘ C). It was observed that ozonation of the water
of NOM in biofilters.6,26,34,60 – 65 Goel et al.66 found that an ozone increased BDOC levels by 30%. DOC results indicated that the
dose of 2 mg O3 mg−1 DOC was a reasonable compromise between pre-oxidation columns operating at 3 ∘ C performed better than
enhancing biodegradability of the NOM and ozone consumption, the non-oxidation set of columns.
but also found that sometimes up to 4 mg O3 mg−1 DOC was Yavich et al.25 sought to compare the biodegradation of NOM
589

needed, depending on the raw water chemistry. from three separate source waters under the influence of different

J Chem Technol Biotechnol 2016; 91: 585–595 © 2015 Society of Chemical Industry wileyonlinelibrary.com/jctb
www.soci.org OD Basu, S Dhawan, K Black

ozone dosages using BDOC as an indicator. Increasing the ozone


Table 5. TOC and DOC removals by biofilter and membranes under
dose at one of the sources from 0.6 mg O3 mg−1 C to 3 mg O3 mg−1 various conditions (adapted from Ref. 93)
C only resulted in the formation of 0.13 mg L−1 of BDOC leading
the authors to conclude that a majority of the organic matter Flux TOC DOC
from this source was refractory, i.e. not subject to biodegradation. (L m−2 h−1 ) Unit operation removal (%) removal (%)
However, increasing the ozone dose from 0 mg O3 mg−1 C to 1 mg
O3 mg−1 C resulted in an increase in BDOC from 1.15 mg L−1 (in 50 Biofilter 50 32
the raw water) to 2.78 mg L−1 . Alternatively, at the third site, BDOC Membrane + Biofilter 60 32
formation increased to 7.39 mg L−1 from 5.04 mg L−1 in the raw Membrane (without 48 10
pretreatment)
water upon increasing the ozone dosage from 0 mg O3 mg−1 C to
35 Biofilter 60 40
3 mg O3 mg−1 C
Membrane + Biofilter 90 45
Yang et al.82 compared different treatment processes
Membrane (without 55 10
post-conventional treatment (coagulation, flocculation, sedi- pretreatment)
mentation and rapid sand filtration) at pilot scale. Water from
the rapid sand filter was passed through GAC and ozone/BAC
processes with different EBCTs. The GAC column had an EBCT of Combined ozonation with biofiltration ahead of membrane
27.1 min while the BAC column had an EBCT of 39.1 min. Ozone filtration systems has been shown to be beneficial for mem-
dosage was 4 mg L−1 . DOC removal by rapid sand filtration process brane fouling control.65,80 Mosqueda-Jimenez et al. investigated
was 33.9%. The GAC process increased DOC removal efficiency by the effect of biofiltration on membrane performance.93 Two hol-
10% compared with the rapid sand filter whereas O3 /BAC process low fiber ultrafiltration membrane units were used to assess
increased removal efficiency by 30% compared with the rapid the impact of biofiltration pretreatment at two different perme-
sand filter. ate fluxes (50 L m−2 h−1 and 35 L m−2 h−1 ) with one membrane
installed prior to the biofilter and one membrane installed after
Hydrogen peroxide biofiltration.
Processes that use OH radicals as the main oxidant are called Table 5 lists the DOC removal with and without the biofilter
advanced oxidation processes (AOPs). These hydroxyl radicals are and membrane units; on their own the membranes show lim-
very short lived and extremely strong oxidizing agents. They react ited DOC removal (10%); when combined with biofiltration, the
with double bonds, or H-atom abstraction to form carbon centred removal increases to 45%. The authors attribute this to the molecu-
radicals. UV/ H2 O2 can be used to generate OH radicals. At high lar weight cutoff range of their membranes (500–30 000 Daltons).
doses, UV/ H2 O2 or UV/Ozone can mineralize NOM and decrease At this range, the membranes would be able to remove larger
the TOC concentration.9,56,57,68 – 70,83 The relatively high doses for particles easily, as is indicated by the percentage TOC removals.
UV/H2 O2 (in excess of 1500 mJ cm−2 and 20 mg L−1 H2 O2 ) can Their study on membrane permeability found that the higher flux
be economically unfeasible to achieve given the high energy (50 L m−2 h−1 ) resulted in a quicker loss of permeability compared
required. At the lower, more economically feasible doses (up to with the lower flux (35 L m−2 h−1 ) for any volume of water fil-
1000 mJ cm−2 , and less than 10 mg L−1 H2 O2 ) NOM is oxidized into tered. For instance, at a filtered volume of 260 L, the permeability
intermediate compounds that are less aromatic and have been reduction at the higher flux was 67% for the membrane without
shown to lower the tendency to form DBPs.56,84 – 87 pretreatment versus 59% for the membrane with pretreatment.
Reductions of 14.5 to 27% of TOC, and 36–55% of UVA were In contrast, at the lower flux, the permeability reductions were
observed following treatment by combined UV and peroxide.88 56% for the membrane without pretreatment versus 11% for the
In addition, Sarathy88 found that treatment led to a decrease in membrane with pretreatment. Thus, differences in permeability
hydrophobic fractions, and a shift from high molecular weight reduction were more evident at the lower flux. Biofiltration as a
NOM to low molecular weight NOM. Lauderdale et al.89 used pretreatment method to control membrane fouling has demon-
hydrogen peroxide at a dosage of 1 mg L−1 for a period of 10 strated promise, however there is limited research associated with
days. Pilot scale, dual media (GAC/sand) columns were used. DOC temperature affects and adverse water quality events for this type
removal was observed to be 15% which was approximately 50% of system.
higher than the control column. Supplementation of peroxide also
reduced terminal headloss by 60% compared with the control
column. Kozyatnik et al.90 investigated the effects of H2 O2 addition ORGANICS REMOVAL
on the efficiency of biofiltration to remove a solution of fulvic In drinking water treatment, natural organic matter (NOM) is
acid (50 mg O dm−3 ). Biofilters contained only GAC. The H2 O2 was responsible for imparting colour to the water. If left untreated
dosed at a concentration of 5 mg dm−3 . It was found that the throughout the treatment process, NOM is a potential disinfection
adsorption capacity of the GAC for fulvic acid increased by 82% by-product precursor in the presence of certain disinfectants;
with peroxide addition versus without. such as chlorine.94 – 97 Biofiltration with or without ozonation is
effective in reducing the NOM content in the water source through
mineralization of NOM by the microbes in the biofilm contained
BIOFILTER PRETREATMENT FOR MEMBRANES within the biofilter.25,98,99
Membrane fouling can affect membrane productivity, frequency Temperature has been found to be an important parameter influ-
of cleaning, membrane replacement rate, energy consumption, encing BDOC removal.47,100 Modeling results by Laurent et al.100
and overall operation and maintenance costs Strategies for min- indicate that BDOC removal is highly dependent on empty bed
imizing fouling have been investigated for many years, including contact time (EBCT) of the filters. It was found that for lower tem-
the investigation of pretreatment by biofiltration or activated car- peratures (5 ∘ C), an EBCT of 12 min would provide approximately
22% BDOC removal. A similar EBCT for water at 25 ∘ C could achieve
590

bon filtration and hybrid membrane processes.15,91,92

wileyonlinelibrary.com/jctb © 2015 Society of Chemical Industry J Chem Technol Biotechnol 2016; 91: 585–595
Review of water treatment biofiltration www.soci.org

Table 6. Summary of bulk organic carbon parameter percentage removal and filter specifications

Average removal (%)

Reference Filter specification BDOC AOC TOC

101 O3 /BAC, Filtration rate = 1.5 m3 h−1 , EBCT = 15 min 72 - -


102 O3 /BAC, Filtration rate = 8 m3 h−1 , EBCT = N/A - 51.2–60.6 -
103 O3 /BAC, Ozone Dose = 2–8 mg L−1 , EBCT = 11.5 min - 89 33.3
GAC, EBCT = 11.5 min - 69 21.1
47 Sand, loading rate = 3.6 m h−1 , EBCT = 7 min, T = 5 ∘ C 38 43 -
Sand, loading rate = 3.6 m h−1 , EBCT = 7 min, T = 20 ∘ C 60 55 -
Sand, loading rate = 3.6 m h−1 , EBCT = 7 min, T = 35 ∘ C 60 57 -

approximately 50% removal of BDOC. Moll et al.47 used sand filters Sodium azide was inhibitory to the ammonia oxidizing bacteria
operating at 7 min EBCT with a hydraulic loading rate of 3.6 m h−1 . even after its removal from the filters.
Influent water temperature varied from 5 ∘ C to 35 ∘ C, at 5 ∘ C, DOC, Wert et al.27 found that biofiltration was able to convert ammonia
BDOC and AOC removals were found to be 15%, 38%, and 43%, to 50% NO2 -N and 50% NO3 -N in a column with 183 cm anthracite
respectively. The corresponding removals at 35 ∘ C were 24%, 60%, and 20 cm sand after 290 days operation. In their study a high
and 57%. Thus, the results showed a decrease in removal of DOC, loading rate (14.6 m h−1 ) and low EBCT (8.3 min) were used and
BDOC and AOC at lower temperatures. conversion of ammonia to nitrite began after 130 days. Full nitri-
Mofidi et al.46 conducted studies on biodegradable organic mat- fication was observed after 170 days. In contrast, filters with 53 cm
ter (BOM) removal at the Mills Water Treatment Plant in South- anthracite and 25 cm sand operating at a loading rate of 4.8 m h−1
ern California. Their goal was to achieve 70% BOM removal. The (EBCT = 9.7 min) achieved complete ammonia removal by day 99.
plant was retrofitted with ozonation at the head of the treat- Andersson et al.105 present an interesting comparison between
ment process to make the BOM more readily biodegradable. Two ammonia removals at pilot scale and full scale level. The study
biofiltration systems (system 3 and system 4), each containing 16 made use of two different kinds of GAC media: closed superstruc-
anthracite/sand filters were present at the plant. BOM levels were tured (CS) and open superstructured (OS).
monitored by analyzing samples for short chain carboxylic acids At pilot scale, while they observed a decrease in nitrification
(CBXA). System 4 was initially set up for use with spent media at lower temperatures, a decrease in the quantity of nitrifying
with an almost immediate 70% reduction in BOM; the media for biomass attached to the media was not observed. OS media
system 4 was then replaced by virgin media and less than 11% showed greater reduction in ammonia than CS media (77% ver-
BOM removal was observed for a period of 2 months; however sus 55%) during the initial filter acclimation period. At full scale,
it must be noted that the media was replaced in winter with for temperatures greater than 10 ∘ C, OS media showed 98%
water temperatures about 8 ∘ C which likely had an impact on removal while CS media showed 90% removal. The increased
results. Following this 45% removal was observed after 2.5 months removals compared with the pilot scale filters were thought to
operation and recovery to 70% BOM removal was attained only be a result of greater EBCT for the full scale columns (20–30 min
after 5 months operation. System 3 was brought online with vir- vs 3.6 min).
gin media (water temperature 14 ∘ C) with 64% removal of BOM Stembal et al.107 evaluated the concentration profiles of NH3 -N,
observed after 3 months operation. This research may demon- NO2 -N and NO3 -N through the depths of four rapid sand fil-
strate that virgin GAC has properties which interfere with biofilm ters at varying filtration rates for groundwater sources. For filters
development and thus exhausted GAC is a better media for pre-acclimated with nitrifying bacteria, all the incoming ammo-
biofiltration. nia was converted to NO3 -N at a depth between 0.4 m and 0.8 m.
Table 6 provides a summary of various filtration conditions and For filters not acclimated, the ammonia removals were observed
organic removals levels with biofiltration. within 10 days of operation. Complete removal of ammonia was
observed within 25 days. This was accompanied by a spike in the
nitrite values which eventually decreased in another 10 days as
Nitrogen (ammonia) removal the nitrite was converted to nitrate. Utilization of biofilter columns
Nitrogen is found in surface water and groundwater in the form provides a cost effective method of converting ammonia to nitrate
of ammonia, nitrites and nitrates. All these forms are undesirable and nitrite with further conversion to nitrogen also feasible. How-
in drinking water. Traditionally, breakpoint chlorination has been ever, more research is needed on the microbial make-up of the
used to remove ammonia. However, this can be cost inhibitive104 system as well as the required HLR and EBCT.
and can also result in a decrease in short-term chlorine demand.105
Developing nitrification in biological filtration is a cost-effective
method to achieve simultaneous BOM and ammonia removal. Nutrient supplementation
In nitrification, Nitrosomonas bacteria convert ammonia to The molecular formula of a heterotrophic microorganism, such as
nitrite. The Nitrobacter bacterial species then converts nitrite to may be found in water systems, will of course vary; however a
nitrate.27 generalized molecular formula for the biomass may be considered
Olańczuk-Neyman and Bray106 studied the effects of chemical to be C60 H87 O23 N12 P.108 Thus, while we are targeting the removal
(15 mmol L−1 sodium azide addition) and physical (high temper- of carbon in biofiltration applications it may be of importance
591

ature treatment) inhibition on the growth of nitrifying bacteria. to keep in mind the expected, or required, C:N:P ratio to ensure

J Chem Technol Biotechnol 2016; 91: 585–595 © 2015 Society of Chemical Industry wileyonlinelibrary.com/jctb
www.soci.org OD Basu, S Dhawan, K Black

that all elements are available for carbon uptake. A study con- distribution system and contribute to bacterial regrowth. In
ducted in Finland found that phosphorus is an important com- addition, removal of organics by biofiltration helps control dis-
pound to regulate microbial growth. Miettinen et al.109 determined infection by-product formation in water systems and may be a
heterotrophic plate counts for treated water from three surface useful step in controlling membrane fouling. Advanced oxidation
water sources, two artificially recharged groundwater sources and preceding biofiltration has demonstrated good correlation with
one groundwater source post phosphorus addition. Phosphorus assisting in organics removal in combination with biofiltration.
dosages varied from 1 μg P L−1 to 50 μg P L−1 . Strong correlations Research examining microbial speciation in biofilters under vary-
were determined between phosphorus and AOC availability to ing conditions seems lacking and there is a lack of understanding
microbial growth. It was determined that AOC alone was not the of the effects of bacterial growth within the biofilter column
driving catalyst but rather its availability in the presence of phos- and associated removal of organics. Finally, there is a lack of
phorus that drove the process of growth. research on cold water temperature effects and novel methods
Nishijima et al.110 carried out phosphorus supplementation to help overcome limitations to compensate for decreased bac-
experiments in a 220 mL continuously mixed reactor containing terial growth rates under cold water temperatures with drinking
110 mL of GAC. For an initial DOC concentration of 9.5 mg L−1 , the water applications. Nutrient supplementation has demonstrated
average effluent DOC value from the filter was 3.9 mg L−1 for GAC the potential to facilitate indigenous bacterial growth in the
with adsorbed phosphorus (3.2 mg P g−1 GAC) and 8.7 mg L−1 for biofilter to maximize organics removals and the need to assess
GAC without adsorbed phosphorus. After 47 days, the influent to incoming water quality to ensure it is suitable for biofiltration
the GAC without adsorbed phosphorus was spiked with 0.1 mg P applications.
L−1 . This resulted in a dramatic decrease in DOC from 8.7 mg L−1
to 2.2 mg L−1 .
Xin et al.111 studied the effects of phosphorus supplementation REFERENCES
on two pilot scale GAC/sand filters. Each filter had a hydraulic 1 Symons G, Water treatment through the ages. J Am Water Works Assoc
loading between 7.5 and 8.5 m h−1 . This corresponded to an 98:87–98 (2006).
2 Logsdon G, Horsley MB, Freeman SDN, Neemann JJ and Budd GC, Fil-
EBCT in the range 6.4–7.2 min. A 54% increase was observed tration processes – a distinguished history and a promising future.
in the heterotrophic plate counts after phosphorus addition. An J Am Water Works Assoc 98:150–162 (2006).
increase in the permanganate index removal (CODMn ) was also 3 Simpson DR, Biofilm processes in biologically active carbon water
observed (20.56% removal with P addition versus 14.54% with- purification. Water Res 42:2839–2848 (2008).
4 Kim WH, Nishijima W, Shoto E and Okada M, Pilot plant study on
out P addition). Lehtola et al. showed that phosphorous can be a ozonation and biological activated carbon process for drinking
growth-limiting nutrient in ozonated waters.112 They found that water treatment. Water Sci Technol 35:21–28 (1997).
ozonation increased the content of phosphorous that can be easily 5 Chaudhary DS, Vigneswaran S, Ngo H, Shim WG and Moon H,
assimilated. Their work showed that lack of phosphorous inhib- Biofilter in water and wastewater treatment. Korean J Chem Eng
ited microbial growth in ozonated drinking waters, rather than 20:1054–1065 (2003).
6 Uhl W, Biofiltration processes for organic matter removal, in Biotech-
organic carbon and that even a minor increase in phosphorous nology, 2nd completely revised edn, Vol 11c, ed. by Rehm HJ
resulted in significant increases in microbial growth. These find- and Reed G. Environmental Processes III. Wiley-VCH, New York,
ings however were restricted to phosphorus-limited waters. Laud- pp. 457–478 (2000).
erdale et al.89 supplemented phosphoric acid (∼20 μg L−1 PO4 -P) 7 Raghuvanshi S, Majumder S and Gupta S, Biofiltration: essentials,
research and applications, in Bioremediation and Sustainability:
in three pilot scale dual-media (GAC/sand) biofilter columns to
Research and Applications, ed. by Mohee R and Mudhoo A. John
increase the C:N:P ratio to 100:10:1 on a mol basis. Supplemen- Wiley & Sons, Inc., Hoboken, NJ, USA (2012).
tation was found to decrease biofilter headloss by greater than 8 Cohen Y, Biofiltration – the treatment of fluids by microorganisms
15%. Sang et al.113 used two identical ceramic biofilters (biofil- immobilized into the filter bedding material: a review. Bioresource
ters with porous ceramic media). One of the filters was dosed Technol 77:257–274 (2001).
9 [9]Randtke SJ and Horsley MB, Water Treatment Plant Design, 5th
with phosphorus (25 μg of PO3− 4 -PL−1 ) for 60 days while the other edn. AWWA, American Society of Civil Engineers, McGraw-Hill
which was nutrient limited was not, average TOC removals of (2012).
22–26.6% were observed in the phosphorous supplemented fil- 10 American Water Works Association Research Foundation (AWWARF),
ter while 17.1–21.3% TOC removals were observed in the control. Ozone and biological treatment for DBP control and biological
Granger et al.114 investigated the effects of phosphorus supple- stability, ed. by Price ML. AWWA Research Foundation, Denver, CO
(1994).
mentation on manganese and DOC removal within bench scale 11 Wang G-S, Liao C-H, Chen H-W and Yang H-C, Characteristics of nat-
GAC/anthracite columns. Phosphorus was dosed at 20 μg L−1 to ural organic matter degradation in water by UV/H2 O2 treatment.
200 μg L−1 . It was found that phosphorus was not a limiting nutri- Environ Technol 27:277–287 (2006).
ent for manganese removal. However, on average, 23% of DOC 12 Thiel P, Zappia L, Franzmann P, Warton B, Alessandrino M, Heitz
A et al., Activated carbon vs anthracite as primary dual media
was removed in columns with phosphorus supplementation ver- filters – a pilot plant study. In 9th Annual Water Industry Engineers
sus 12% in the control columns. and Operators Conference, Bendigo Exhibition Centre (Vol. 5, pp.
Research into nutrient supplementation for biofilter optimiza- 8–14) (2006).
tion is a necessary field of study, and would be of particular use 13 Emelko MB, Huck PM, Coffey BM and Smith EF, Effects of media,
when examined in conjunction with other operating parameters backwash and temperature on full-scale biological filtration. J Am
Water Works Assoc 98:61–73 (2006).
such as temperature and hydraulic loading rate. 14 Liu X, Huck PM and Slawson RM, Factors affecting drinking
water biofiltration. J Am Water Works Assoc 93:90–101 (2001).
15 Basu OD and Huck PM, Integrated biofilter-immersed membrane
CONCLUSIONS system for the treatment of humic waters. Water Res 38:655–662
(2004).
Biofiltration has proven to be a useful technology in drinking 16 Price ML, Bailey RW, Enos AK, Hook M and Hermanowicz SW, Evalua-
water treatment systems. Biofiltration has the capacity to decrease tion of ozone/biological treatment for disinfection byproducts con-
592

organic matter content that would otherwise enter into the water trol and biologically stable water. Ozone Sci Eng 15:95–130 (1993).

wileyonlinelibrary.com/jctb © 2015 Society of Chemical Industry J Chem Technol Biotechnol 2016; 91: 585–595
Review of water treatment biofiltration www.soci.org

17 Wang JZ, Summers RS and Miltner RJ, Biofiltration performance: 42 Fogel D, Isaac-Renton J, Guasparini R, Moorehead W and Ongerth
Part 1. Relationship to biomass. J Am Water Works Assoc 87:55–63 J, Removing giardia and cryptosporidium by slow sand filtration.
(1995). J Am Water Works Assoc 85:77–84 (1993).
18 Klevens CM, Collins MR, Negm R and Farra MF, Characterisation of 43 Persson F, Heinicke G, Uhl W, Hedberg T and Hermansson M, Per-
NOM removal by biological activated carbon, in Advances in Slow formance of direct biofiltration of surface water for reduction of
Sand and Alternative Biological Filtration, ed. by Graham N and biodegradable organic matter and biofilm formation potential.
Collins R. John Wiley and Sons, Chichester, pp. 79–88 (1996). Environ Technol 27:1037–1045 (2006).
19 Seredynska-Sobecka B, Tomaszewska M and Morawski A, Removal 44 Wulfeck WM Jr and Summers RS, Control of DBP Formation using
of humic acids by the ozonation-biofiltration process. Desalination retrofitted GAC filter-adsorbers and ozonation. In Proceedings of the
198:265–273 (2006). 1994 AWWA Water Quality Technology Conference, San Francisco, CA.
20 Fan H, Chiu T, Yang H, Chen W and Furuya E, Removal of humic American Water Works Association (1994).
acids, fulvic acids and non humic substances from landfill leachate. 45 Nishijima W, Kim WH, Shoto E and Okada M, The performance of
J Environ Eng Manage 17:325–331 (2007). an ozonation-biological activated carbon process under long term
21 LeChevallier MW, Becker WC, Schorr P and Lee RG, Evaluating the operation. Water Sci Technol 38:163–169 (1998).
performance of biologically active rapid filters. J Am Water Works 46 Mofidi A, Johnston R, Coffey BM, Gerringer FW and Krasner SW, Per-
Assoc 84:136–146 (1992). formance of large-scale biological filtration for removal of particles
22 Carlson KH and Amy GL, The relative importance of HLR and EBCT in and BOM produced by ozonation. In Proceedings from the Water
biofiltration, in Proceedings of 1995 AWWA Water Quality Technol- Quality and Technology Conference, Quebec City, 2005. American
ogy Conference, New Orleans, LA. Water Works Association (2005).
23 Huck PM, Finch GR, Hrudey SE and Peppler MS, Design of biological 47 Moll DM, Summers RS, Fonseca AC and Matheis W, Impact of temper-
processes for organics control. AWWA Research Foundation, Denver, ature on drinking water biofilter performance and microbial com-
CO (1998). munity structure. Environ Sci Technol 33:2377–2382 (1999).
24 Hozalski RM, Goel S and Bouwer EJ, TOC removal in biological filters. 48 Fonseca AC and Summers RS, Evaluation of different ozonation strate-
J Am Water Works Assoc 87:40–54 (1995). gies and of temperature effects on biological filter performance. In
25 Yavich AA, Lee KH, Chen KC, Papa L and Masten SJ, Evaluation of Proceedings of the AWWA Water Quality and Technology Conference.
biodegradability of NOM after ozonation. Water Res 38:2839–2846 Philadelphia, PA (2003).
(2004). 49 Jasim SY, Ndiongue S, Borikar D, Johnson B and Schweitzer L, The
26 Hozalski RM, Bouwer EJ and Goel S, Removal of natural organic effect of ozone on cold water coagulation. Ozone: Sci Eng 30:27–33
matter from drinking water supplies by ozone-biofiltration. Water (2008).
Sci Technol 40:157–163 (1999). 50 Cipparone LA, Diehl AC and Speitel Jr GE, Ozonation and BDOC
27 Wer EC, Neemann JJ, Rexing DJ and Zegers RE, Biofiltration for removal: effect on water quality. J Am Water Works Assoc
removal of BOM and residual ammonia following control of bro- 89:84–97(1997).
mate formation. Water Res 42:372–378 (2008). 51 Wu JJ, Wu C and Chuang W, Evaluation of oxidation by products
28 Galvis G, Visscher JT and Lloyd B, Multi-stage surface water treatment and control of organic matters using advanced oxidation processes
for community water supply in Columbia. Waterlines 10:26–29 (AOPs) combined with biological fluidized bed for the treatment
(1992). of eutrophicated raw water. In Proceedings of the 16th International
29 Carlson KH and Amy GL, BOM removal during biofiltration. J Am Water Ozone Congress, Las Vegas, NV (2003).
Works Assoc 90:42–52 (1998). 52 Owen DM, Amy GL, Chowdhury ZK, Paode R, McCoy G and Viscosil
30 Melin ES and Odegaard H, The effect of biofilters loading rate K, NOM characterization and treatability. J Am Water Works Assoc
on the removal of organic ozonation by-products. Water Res 87:46–63 (1995).
34:4464–4476 (2000). 53 Fahmi NW and Okada M, Improvement of DOC removal by multistage
31 Cleary SA, Sustainable drinking water treatment for small communi- 22 AOP-biological treatment. Chemosphere 50:1043–1048 (2003).
ties using multistage slow sand filtration. Masters thesis, University 54 Bérubé PR, Mohseni M, Chin A and Toor R, Comparing UV-Ozone
of Waterloo (2005). and UV- hydrogen peroxide for the reduction of the disinfection
32 Ko YS, Lee YJ and Nam SH, Evaluation of a pilot scale dual media by-product formation potential. Proceedings of the 2004 AWWA
biological activated carbon process for drinking water. Korean J Water Quality Technology Conference, San Antonio, Texas (2004).
Chem Eng 24:253–260 (2007). 55 Kleiser G and Frimmel FH, Removal of precursors for disinfection
33 Servais P, Billen G and Hascoet M, Determination of the biodegrad- by- products (DBPs) - differences between ozone- and OH radical-
able fraction of dissolved organic matter in waters. Water Res induced oxidation. Sci Total Environ 256:1–9 (2000).
21:445–450 (1987). 56 Thomson J, Parkinson A and Roddick FA, Depolymerization of
34 Melin E, Skkog R and Odegaard H, Ozonation/biofiltration with cal- chromophoric natural organic matter. Environ Sci Technol
cium carbonate as biofilter media, in Recent Progress in Slow Sand 38:3360–3369 (2004).
and Alternative Biofiltration Processes, ed by Gimbel R, Graham NJD 57 Gottschalk C, Libra J and Saupe A, Ozonation of Water and Wastew-
and Collins MR. IWA Publishing, London, 406–413 (2006). ater: A Practical Guide to Understanding Ozone and its Application.
35 Urfer D and Huck PM, Measurement of biomass activity in drink- Wiley-VCH, New York (2000).
ing water biofilters using a respirometric method. Water Res 58 von Gunten U, Ozonation of drinking water: Part I. Oxidation kinetics
35:1469–1477 (2001). and product formation. Water Res 37:1443–1467 (2003a).
36 Velten S, Boller M, Köster O, Helbing J, Weilenmann HU and Hammes F, 59 Hoigne J and Bader H, Ozonation of water - role of hydroxyl radicals
Development of biomass in a drinking water granular active carbon as oxidizing intermediates. Science 190:782–784 (1975).
(GAC) filter. Water Res 45:6347–6354 (2011). 60 Werner P and Hambsch B. Investigations on the growth of bacteria in
37 Boon N, Pycke BF, Marzorati M and Hammes F, Nutrient gradients drinking water. Water Supply 4:227–235 (1986).
in a granular activated carbon biofilter drives bacterial com- 61 van der Kooij D, Hijnen WAM and Kruith JC, The effect of ozonation,
munity organization and dynamics. Water Res 45:6355–6361 biological filtration and distribution on the concentration of easily
(2011). assimilable organic carbon (AOC) in drinking water. Ozone: Sci Eng
38 Amburgey JE and Amirtharajah A, Strategic filter backwashing tech- 11:297– 311 (1989).
niques and resulting particle passage. J Environ Eng 131:535–547 62 Murphy BM, Ozone-induced conversion of DBP precursors (DOC) to
(2005). biodegradable by-products (BDOC). Masters dissertation, Depart-
39 Liao X, Chen C, Zhang J, Dai Y, Zhang X and Xie S, Operational per- ment of Civil, Environmental, and Architectural Engineering,
formance, biomass, and microbial community structure: impacts University of Colorado, Boulder, CO (1993).
of backwashing on drinking water biofilter. Environ Sci Pollut Res 63 Siddiqui MS, Amy GL and Murphy BD, Ozone enhanced removal
22:546–554 (2015). of natural organic matter from drinking water. Water Res
40 Huisman L and Wood WE, Slow Sand Filtration. World Health Organi- 31:3098–3106 (1997).
zation, Geneva, Switzerland (1974). 64 American Water Works Association Research Foundation (AWWARF),
41 Bellamy WD, Hendricks D and Logsdon GS, Slow sand filtration: Advanced oxidation and biodegradation processes for the destruc-
influences of selected process variables. J Am Water Works Assoc tion of TOC and DBP precursors. AWWA Research Foundation, Den-
593

77:62–66 (1985). ver, CO (1999).

J Chem Technol Biotechnol 2016; 91: 585–595 © 2015 Society of Chemical Industry wileyonlinelibrary.com/jctb
www.soci.org OD Basu, S Dhawan, K Black

65 Kim HC, Yu MJ, Koo JY and Leo S, Application of O3 /GAC process 88 Sarathy S, Effects of UV/H2 O2 advanced oxidation on physical and
for advanced and selective removal of NOM from conventionally chemical characteristics of natural organic matter in raw drinking
treated water. Water Sci Technol 16:101–108 (2006). water sources. Doctoral dissertation, University of British Columbia,
66 Goel S, Hozalski RM and Bouwer EJ, Biodegradation of NOM: effect of Vancouver, BC (2009).
NOM source and ozone dose. J Am Water Works Assoc 87:90–105 89 Lauderdale C, Chadik P, Kirisits MJ and Brown J, Engineered
(1995). biofiltration: enhanced biofilter performance through nutrient
67 Zhang XX, Zhang ZY, Ma LP, Liu N, Wu B, Zhang Y et al., Influences of and peroxide addition. J Am Water Works Assoc 104:298–309
hydraulic loading rate on SVOC removal and microbial community (2012).
structure in drinking water treatment biofilters. J Hazard Mater 90 Kozyatnik IP, Klimenko NA, Savchina LA and Vrubel TL, The use of
178:652–657 (2010). hydrogen peroxide for raising the efficiency of filter operation
68 Shukairy HM and Summers RS, DBP speciation and kinetics as affected with biologically activated carbon. Water Chem Technol 30:72–79
by ozonation and biotreatment disinfection by-products in water (2008).
treatment: the chemistry of their formation and control, In Disin- 91 Peldszus S, Benecke J, Jekel M and Huck PM, Direct biofiltration
fection By-Products in Water Treatment, ed. by Minear RA and Amy pretreatment for fouling control of ultrafiltration membranes. J Am
GL. CRC Press Inc, Boca Raton, FL (1995). Water Works Assoc 104:430–445 (2012).
69 Speitel GE Jr, Symons JM, Mialaret JM and Wanielista MME, 92 Stoquart C, Servais P, Bérubé P and Barbeau B, Hybrid membrane
AOP/Biofilm process for DOX precursors. J Am Water Works processes using activated carbon treatment for drinking water
Assoc 92:59–73 (2000). production: a review. J Membr Sci 411–412:1–12 (2012).
70 Digiano FA, Singer PC, Parameswar C and Lecourt TD, Biodegradation 93 Mosqueda-Jimenez DB, Huck PM and Basu OD, Fouling characteristics
kinetics of ozonated NOM and aldehydes. J Am Water Works Assoc of an ultrafiltration membrane used in drinking water treatment.
98:92–104 (2001). Desalination 230:79–91 (2008).
71 Greaves GF, Grundy PG and Taylor GS, Ozonation and slow sand 94 Chen M, Ohman K, Sinclair J, Petkau D, Yau R, Deng JF et al., Disin-
filtration for the treatment of coloured upland waters - pilot plant fection by-products in drinking water- a case study on Calgary,
investigations, in Slow Sand Filtration. Recent Development in Water Alberta, Canada. Water Quality Res J Canada 46:200–210 (2011).
Treatment Technology, ed. by Graham NJD. Ellis Horwood, UK, 95 Chae S, Evaluation of drinking water treatment processes focusing
pp. 153–162 (1988). on natural organic matter removal and on disinfection by-product
72 Cable CJ and Jones RG, Assessing the effectiveness of ozonation fol- formation. Water Supply 2:459–464 (2002).
lowed by slow sand filtration in removing THM precursor material 96 Sohn J, Amy G and Yoon Y, Process-train profiles of NOM through
from an upland raw water, in Advances in Slow Sand and Alternative a drinking water treatment plant. J Am Water Works Assoc
Biological Filtration, ed. by Graham NJD and Collins R. John Wiley & 99:145–153 (2007).
Sons, UK, pp. 29–37 (1996). 97 Bekbolet M, Uyguner CS, Selcuk H, Rizzo L, Nikolaou AD, Meric S et al.,
73 Westerhoff P, Debroux J, Aiken G and Amy G, Ozone induced Application of oxidative removal of NOM to drinking water and
changes in natural organic matter (NOM) structure. Ozone Sci Eng formation of disinfection by-products. Desalination 176:155–166
(2005).
21:551–570 (1999).
98 Chaiket T, Singer PC, Miles A, Moran M and Pallotta C, Effectiveness of
74 Galapate R, Baes A and Okada M, Transformation of dissolved organic
coagulation, ozonation, and biofiltration in controlling DBPs. J Am
matter during ozonation: effects on trihalomethane formation
Water Works Assoc 99:81–95 (2002).
potential. Water Res 35:2201–2206 (2001).
99 Yavich AA and Masten SJ, Use of ozonation and FBT to control THM
75 Singer PC, Formation and control of disinfection by-products in drinking
precursors. J Am Water Works Assoc 95:159–171 (2003).
water. AWWA Publishing, Denver, CO (1999).
100 Laurent P, Prévost M, Cigana J, Niquette P, and Servais P, Biodegrad-
76 von Gunten U, Ozonation of drinking water: Part II. Disinfection and
able organic matter removal in biological filters: evaluation of the
by- product formation in presence of bromide, iodide or chlorine.
CHABROL model. Water Res 33:1387–1398 (1999).
Water Res 37:1469–1487 (2003b). 101 Xiang H, Lu X, Yin L, Yang F, Zhu G and Liu W, Microbial commu-
77 Huang W-J, Chen L-Y and Peng H-S, Effect of NOM characteris- nity characterization, activity analysis and purifying efficiency in a
tics on brominated organics formation by ozonation. Environ Int biofilter process. J Environ Sci 25:677–687 (2013).
29:1049–1055 (2004). 102 Liao X, Chen C, Wang Z, Wan R, Chang CH, Zhang X et al., Changes of
78 Hammes F, Salhi E, Koster O, Kaiser HP, Egli T and von Gunten U, Mech- biomass and bacterial communities in biological activated carbon
anistic and kinetic evaluation of organic disinfection by-product filters for drinking water treatment. Process Biochem 42:312–316
and assimilable organic carbon (AOC) formation during the ozona- (2013).
tion of drinking water. Water Res 40:2275–2286 (2006). 103 Kong LY, Zhang XJ and Wang ZS, Pilot plant study on ozonation
79 Krasner SW, Sclimenti MJ and Coffeey BM, Testing biologically active and biological activated carbon process for drinking water treat-
filters for removing aldehydes formed during ozonation. J Am Water ment. J Environ Sci 18:232–235 (2006).
Works Assoc 85:62–70 (1993). 104 Wert EC, Rosario-Ortiz FL, Drury DD and Snyder SA, Formation of
80 Karnik BS, Davies SHR, Chen KC, Jaglowski DR, Baumann MJ and Mas- oxidation byproducts from ozonation of wastewater. Water Res
ten SJ, Effects of ozonation on the permeate flux of nanocrystalline 41:1481–1490 (2007).
ceramic membranes. Water Res 39:728–734 (2005). 105 Andersson A, Laurent P, Kihn A, Prévost M and Servais P, Impact of
81 Fonseca AC, Summers RS and Hernandez MT, Comparative measure- temperature on nitrification in biological activated carbon (BAC)
ments of microbial activity in drinking water biofilters. Water Res filters used for drinking water treatment. Water Res 35:2923–2934
35:3817–3824 (2001). (2001).
82 Yang J, Dong-Xing Y and Weng T, Pilot study of drinking water 106 Olańczuk-Neyman K and Bray R, The role of physico-chemical
treatment with GAC, O3 /BAC and membrane processes in Kinmen and biological processes in manganese and ammonia nitro-
Island, Taiwan. Desalination 263:271–278 (2010). gen removal from groundwater. Polish J Environ Studies 9:91–96
83 Toor R and Mohseni M, UV/H2 O2 based AOP and its integration (2000).
with biological activated carbon treatment for DBP reduction in 107 Štembal T, Markić M, Ribičić N, Briški F and Sipos L, Removal of
drinking water. Chemosphere 66:2087–2095 (2007). ammonia, iron and manganese from groundwaters of north-
84 Oppenlander T, Photochemical Purification of Water and Air. ern Croatia - pilot plant studies. Process Biochem 40:327–335
Wiley-VCH, New York (2003). (2005).
85 Tuhkanen TA, UV/H2 O2 processes, in Advanced Oxidation Processes for 108 Metcalf and Eddy Inc, Wastewater Engineering, Treatment and Reuse.
Water and Wasterwater Treatment, ed. by Parsons S. IWA Publishing, McGraw-Hill, New York (2003).
London, pp. 86–110 (2004). 109 Miettinen IT, Vartiainen T and Martikainen PJ, Phosphorus and bacte-
86 Chin A and Bérubé P, Removal of disinfection by-product pre- rial growth in drinking water. Appl Environ Microbiol 63:3242–3245
cursors with ozone-UV advanced oxidation process. Water Res (1997).
39:2136–2144 (2005). 110 Nishijima W, Shoto E and Okada M, Improvement of biodegra-
87 Sarathy S and Mohseni M, The Impact of UV/H2 O2 advanced oxidation dation of organic substance by addition of phosphorus in
on molecular size distribution of chromophoric natural organic biological activated carbon. Water Sci Technol 36:251–257
594

matter. Environ Sci Technol 41:8315–8320 (2007). (1997).

wileyonlinelibrary.com/jctb © 2015 Society of Chemical Industry J Chem Technol Biotechnol 2016; 91: 585–595
Review of water treatment biofiltration www.soci.org

111 Xin, Y, Xiao-jian Z, Xiao-jian Z, Xiao-dong Z and Zhan-shen W, Phos- 113 Sang J, Zhanga X, Lib L and Wang Z, Improvement of organics removal
phorous limitation in biofiltration for drinking water treatment. by bio-ceramic filtration of raw water with addition of phosphorus.
J Environ Sci 15:494–499 (2003). Water Res 37:4711–4718 (2003).
112 Lehtola MJ, Miettinen IT, Vartiainen T, Myllykangas T and Martikainen 114 Granger HC, Stoddart AK and Gagnon GA, Direct biofiltration
PJ, Microbially available organic carbon, phosphorus and microbial for manganese removal from surface water. J Environ Eng
growth in ozonated drinking water. Water Resources 35:1635–1640 140:04014006-1-04014006-8 (2014).
(2001).

595

J Chem Technol Biotechnol 2016; 91: 585–595 © 2015 Society of Chemical Industry wileyonlinelibrary.com/jctb

You might also like