3233 © IWA Publishing 2020 Water Supply | 20.

8 | 2020

Design of up-flow aerated filters for the removal of iron

from groundwater
T. Thinojah, B. Ketheesan and G. B. B. Herath

ABSTRACT

Groundwater contains iron (Fe) that can be difficult to remove by singular physical–chemical T. Thinojah (corresponding author)
B. Ketheesan
methods. The aim of this presented study was to investigate the suitability of up-flow aerated filters Faculty of Engineering,
University of Jaffna,
as an upgrading step for the removal of total Fe from groundwater. Granular activated carbon (GAC) Kilinochchi,
Sri Lanka
was applied as a pollutant carrier and compared with Bio-Pac media (BPM) by conducting a column
E-mail: [email protected]
test to treat synthetic groundwater (SGW). Initially two adsorption kinetic studies were done for GAC.
G. B. B. Herath
The effect of influent Fe concentration, hydraulic retention time (HRT) and air supply to the system Faculty of Engineering,
University of Peradeniya,
were investigated using a column experiment. The sorption result follows pseudo-second-order Kandy,
Sri Lanka
kinetics with an adsorption amount of 1.36 mg/g. The column experiment results show that around
99% and 70% Fe removal efficiency can be achieved for the best condition of 3 mg/L influent Fe
concentration with seven hours HRT for GAC and BPM respectively. The air supply into the column
did not show any improvement in Fe removal efficiency. The performance of GAC is found to be
much better than that of BPM. All over, the new combined treatment method of aeration followed by
filtration was demonstrated to be an effective technique for Fe removal from groundwater and also
satisfied the World Health Organization (WHO) drinking water standard for Fe concentration.
Key words | Bio-Pac media, column test, granular activated carbon, groundwater, iron removal

HIGHLIGHTS

• Groundwater contamination.

• Fe removal from groundwater.

• Floc filtration technology.

• Upflow aerated filters for iron removal.

• Impact of influent quality parameters on iron removal.

INTRODUCTION

Water is an essential element of human life, its main The quality and the supply of water in the peninsula has
domestic uses being drinking, preparing food and been affected by groundwater contamination with salinity,
maintaining personal hygiene. Domestic, agricultural and nitrate and iron due to over-extraction of groundwater,
industrial activities in the Jaffna Peninsula located in the intrusion of seawater into the groundwater aquifers and
northern part of Sri Lanka depend highly on groundwater, unregulated use of fertilizers in the agricultural sector
which is the only foremost natural water source, with a (Saravanan et al. ). At present, several other factors
lower percentage of surface water because of its karstic such as changes in rainfall, indiscriminate water intake, sea
nature and the flat terrain (Senaratne & Dissanayake ). water intrusion, diversification or improper maintenance of

doi: 10.2166/ws.2020.229

Downloaded from http://iwaponline.com/ws/article-pdf/20/8/3233/812638/ws020083233.pdf

by guest
 3234 T. Thinojah et al. | Iron removal filter design for groundwater treatment Water Supply | 20.8 | 2020

water resources, improper maintenance of domestic sewage Peri-kinetic flocculation is dominant for particles up to
and population growth have further deteriorated the quality 1 μm size and, with Brownian motion as the driving force.
of groundwater and have resulted in a serious threat to Ortho-kinetic flocculation is dominant when the particle
human health (Mikunthan et al. ; Sivakumar ). size exceeds 1 μm, with artificial mixing (hydraulic or mech-
Fe is a common element in the Earth’s crust and dissol- anical) as the driving force (Oyegbile et al. ). The rate of
ution of Fe with groundwater can occur when it percolates ortho-kinetic flocculation is given by Equation (1):
through soil and rock (Dvorak et al. ). Settling of iron
(hydroxide) particles or post-treatment flocculation of dis- dn



¼ 4α0 GΩn=π (1)
solved Fe could lead to the degeneration of water quality dt
in a distribution system even when clear water meets the
drinking water standards. Higher concentration of Fe in where n ¼ number of particles per m3, t ¼ time (s), α0 ¼
distribution systems possibly increases the growth of fractions of collision which result in agglomeration due to
chlorine-tolerant microorganisms, which can raise the cost ortho-kinetic flocculation, Ω ¼ volume fraction of particles
of disinfection (Marsidi et al. ). According to WHO (m3/m3), G
(root mean square velocity gradient) ¼ P
Vμ
the maximum concentration of Fe in drinking water where P ¼ power input to the fluid (W), V ¼ volume of
should not exceed 0.3 mg/L. In order to avoid problems water in the vessel (m3), μ ¼ absolute viscosity of water
with iron particles and post-flocculation and subsequent (kg/m·s).
sedimentation and re-suspension of Fe in the distribution Also the removal process is affected by the different
network, a proper treatment technique should be suggested chemical and physical characteristics of the water including
to get very low Fe concentration. pH, temperature, total organic carbon (TOC) and concen-
Although conventional treatment processes (precipi- tration of dissolved oxygen (Zhang ). The form and
tation, electro-coagulation, filtration and sedimentation) concentration of Fe present in water also plays a role as
are widely used for the removal of Fe ions, they present a well. The oxidation rates are faster at high pH values and
number of drawbacks in terms of treatment capacity, slow at low pH values. Oxidation by aeration demands a
efficiency, stability, space requirements, and the generation pH greater than 6.3 for iron (Sharma et al. ). Carbonic
of large volumes of sludge, thus increasing maintenance acid lowers the pH value of water and hence slows down
and operational costs (Xia et al. ; Patil et al. ). the oxidation rate. Increased temperature also enhances
There are two main physicochemical mechanisms of iron the oxidation reactions while lower temperature slows
removal in filters, namely the adsorption–oxidation mechan- down the oxidation process. In general, the rate of reaction
ism (adsorptive filtration/catalytic iron removal) and the will double for every 10  C temperature increase. In the
oxidation–floc formation mechanism (floc filtration). Floc presence of higher concentrations of total organic carbon,
filtration is the common mode of iron removal in many the oxidation of iron is reduced due to less available
groundwater treatment plants, where soluble iron(II) is oxygen. The oxidation of iron(II) can take place even with-
first oxidized to insoluble iron(III) by aeration or chemical out external air supply at higher dissolved oxygen level
oxidation and the flocs formed are subsequently removed (>1–2 mg/L) in groundwater (Sharma et al. ).
in a rapid sand filter (Sharma et al. ). Lower removal Several studies have investigated the operating mechanism
efficiency due to colloidal iron (hydr) oxide or micro-flocs of up-flow aerated filters for pollutant removal, where aeration
passing through the filter is one of the main problems and precipitation (floc-formation) can take place simul-
associated with floc filtration mode. Formation of iron taneously inside the same filter and where the filter runs
hydroxide flocs of filterable size (flocculation) after the longer because of the use of more coarse media (Gage et al.
oxidation of iron(II) and subsequent hydrolysis is influenced ). These treatment techniques were used to remove Fe
by the quality of the water and the process conditions from drinking water and wastewater by early researchers,
applied (Sharma et al. ). Flocculation is of two distinct especially as biological treatment processes (Qin et al. ;
types, namely peri-kinetic and ortho-kinetic flocculation. Ma et al. ; Han et al. ; Lu et al. ; Toit et al. ).

Downloaded from http://iwaponline.com/ws/article-pdf/20/8/3233/812638/ws020083233.pdf

by guest
 3235 T. Thinojah et al. | Iron removal filter design for groundwater treatment Water Supply | 20.8 | 2020

This study is focused on investigating the suitability of times with a constant amount of distilled water until getting
up-flow aerated filters for Fe removal from groundwater as a clear colour and constant pH. Then the samples were
a drinking water treatment method considering the floc fil- dried in the oven at 105  C for 24 hours. Finally they were
tration technique. Two separate columns were designed stored safely to avoid contamination from the environment.
and operated with the same influent conditions of SGW These GAC particles have the specifications of a maximum
with the different filter media GAC and BPM for ease of 2% ash content and 6% moisture content with an iodine
comparison. In order to optimize the process of iron number of 600 g/g·min. The BPM having a height/diameter
removal from groundwater, the effect of initial Fe concen- ratio of 0.71 (height ¼ 10 mm, diameter ¼ 14 mm) and
tration, HRT and air supply to the system on Fe removal having a bio-surface area range of 300–850 m2/m3 was
efficiency was also assessed in this study. A kinetic exper- initially rinsed with distilled water and then dried at atmos-
iment was done for GAC to find out the adsorption ability pheric temperature for further use.
of GAC for Fe removal since the filter is designed to run
for a longer period, where adsorption removal mechanisms
Kinetic experimental studies
are also possible during the column experiment.

Seasonal variation of iron concentration in the study area

throughout the year was found to be a limitation to collect-
MATERIALS AND METHODS ing groundwater for the experiment. For a better-controlled
experiment SGW was simulated by adding groundwater
Preparation of filter media with laboratory reagent FeSO4.7H2O and the physico-
chemical properties of the groundwater are given in
The filter media, GAC produced from coconut shell, which Table 1. The batch-scale kinetic experiment was performed
can remove certain chemicals that are dissolved in water, with SGW having 1 mg/L Fe. Conical flasks having
and BPM made from high-grade poly propylene, which 100 mL capacity were shaken at 150 rpm at room tempera-
can act as a pollutant carrier, were selected for this study. ture 25  C. The pre-set time range of 10–300 minutes was
It was chosen based on the resistance to corrosion and attri- used to check the variation of adsorption amount of GAC
tion, high specific surface area, durability, possession of at different equilibrium times. The experimental data were
appropriate specific weight, chemical dependency, being then described using two types of sorption kinetic models,
easy to manage and having the ability not to clog the reactor namely pseudo-first-order in Equation (2) and pseudo-
and was purchased from Enviroequip (Pvt) Ltd in Colombo, second order in Equation (3) for the prediction of adsorp-
Sri Lanka. Activated carbon particles granular in size having tion rate and the best-fitting equations for the experimental
a sieve range of 8 × 20 were used in this study. Then the par- data were determined by calculating the coefficient of
ticles were sieved and a mean particle size of 1.2 mm was determination.
calculated from the particle size distribution curve. This The linear form of the pseudo-first-order equation is
size range is commonly used in drinking water treatment expressed as:
plants and this grade of material meets the requirement of
international standards for water treatment (Hasan et al.
). The sieved particles were separately rinsed several ln (Qe
Qt ) ¼ ln Qe
Kt (2)

Table 1 | Physico-chemical parameters of tap water used to prepare SGW

Parameter pH TDS EC Salinity DO Zn2þ Mg2þ Ca2þ Cu2þ

Unit _ mg/L μS mg/L mg/L mg/L mg/L mg/L mg/L

Value 7.05 836 1,177 579 5.64 0.17 6.83 87.01 0.04

Downloaded from http://iwaponline.com/ws/article-pdf/20/8/3233/812638/ws020083233.pdf

by guest
 3236 T. Thinojah et al. | Iron removal filter design for groundwater treatment Water Supply | 20.8 | 2020

where Qe is the mass of Fe adsorbed at equilibrium (mg/g), ferrous iron into insoluble ferric iron, which could easily
Qt (mg/g) is the mass of Fe adsorbed at time t and K is the precipitate inside the filter. The top of the filter allowed a
first-order reaction rate constant (L/min). 30 cm buffer zone to stop media being washed out in the
The linear form of the pseudo-second-order equation is process of backwashing. The widely used up-flow configur-
expressed as: ation was used in this study. A peristaltic pump was used
to supply SGW into the column. The filter media were
t 1 t packed partially with 80 cm media height starting 40 cm
¼ þ (3)
Qt K2 × Q2e Qe from the bottom with the support of polypropylene mesh
having a pore size less than 1 mm.
where K2 is the second-order reaction rate constant (g min/
mg), t is the reaction time (min), Qt is the mass of Fe
Operation of the up-flow aerated filter system
adsorbed at time t (mg/g), Qe is the mass of Fe adsorbed
at equilibrium (mg/g).
The in situ porosity of the filter bed was measured using the
volumetric method. Initially the distilled water was sent up
Description of the up-flow aerated filter system to the filter media height from the bottom. Then the
system was allowed to become saturated for 20 minutes.
The schematic and the experimental setup are given in Finally the water was drained from the bottom and the
Figure 1. A laboratory-scale filter made from acrylic tube drained volume was measured for both filter media separ-
having dimensions of 15 cm diameter (D) and 150 cm ately to calculate the volume of voids in the filter media
height (H) was designed with an effective working volume column. The procedure was continued three times to make
of 14 L by considering the excess wall effects from filter sure of the full saturation of the GAC media by conforming
media, channelling effects and H/D ratio to avoid short cir- to the same volume of drained water. The average value of
cuit. The sampling ports were placed throughout the column in situ porosity was used to calculate the relevant media
height at 20 cm intervals for sample collection. The air and flow rate. The flow rate was measured by dividing average
influent synthetic groundwater were fed at 20 and 10 cm drained water volume by hydraulic retention time. The
from the bottom respectively for mixing before entering ratio between drained water volume and total bed volume
the media. An aquarium air pump with 2.5 L/min airflow represents the bed porosity and was found to be 0.28 and
rate was used to supply ambient air (which contained 0.79 for GAC and BPM respectively.
oxygen) to the column, which would increase the dissolved The effects of initial Fe concentration, HRT and air
oxygen level of the influent. This helped in oxidizing soluble supply to the system were tested in three different phases

Figure 1 | Schematic and laboratory-scale up-flow aerated filter system, sampling ports: influent, H1, H2, H3, H4, and effluent.

Downloaded from http://iwaponline.com/ws/article-pdf/20/8/3233/812638/ws020083233.pdf

by guest
 3237 T. Thinojah et al. | Iron removal filter design for groundwater treatment Water Supply | 20.8 | 2020

and the column was operated for 48 hours continuously in in Equation (6).
each phase. The operating conditions in each phase for
both filter media are described in Table 2. The filters were (C0
Ce )
Fe removal efficiency ¼ × 100% (4)
operated continuously throughout the study and back- C0
washed after each scenario using clean water with a flow v
Qt ¼ (C0
Ce ) × (5)
rate of 1,200 mL/min for 30 minutes in the upward direction m
for uncomplicated operation. The samples were collected
at four-hour intervals from the sampling ports including where Qt is the amount of Fe adsorbed from GAC in mg/g,
influent and effluent using 100 mL polypropylene sample C0 and Ce are the initial Fe concentration and the Fe con-
bottles in each phase. centration at equilibrium in mg/L respectively, v is the
volume of solution in L and m is the mass of adsorbent in g:

Analytical methods PN
(yi
^yi )2
R2 ¼ 1
PNi¼1 2
(6)
The analysis of pH, temperature and total dissolved solids i¼1 (yi
〈y〉)

(TDS) was performed via a PC 2700 multimeter. DO and

total Fe were measured using a Cyber Scan Series 600 in which y represents the experimental data and
DO meter and DR 900 HACH calorimeter respectively. yi ≡ y(t ¼ ti ) and ^yi ≡ ^y(t ¼ ti ) are the values of y and ^y,
These parameters were tested immediately after the respectively, at time ti and 〈y〉 stands for the average value
sample collection in the laboratory. At the end of the of the yi (i ¼ 1,…,N).
column test 5 g of GAC samples from top and bottom
were collected and soaked in 0.5 M HCL for one hour,
and then samples were filtered through 0.45 μm filter
RESULTS AND DISCUSSION
paper before measurement using an atomic adsorption
spectrophotometer to find out the totally removed Fe con-
Kinetic study
centration on GAC media. All the measurements were
triplicated and the average value was used for the data
To find out the mechanism of Fe adsorption and the poten-
analysis using an Excel spread sheet.
tial rate-controlling steps, including diffusion mass transport
Fe removal efficiency for both filter media was calcu-
and chemical reaction processes, it is required to character-
lated using Equation (4) and plotted for each media height
ize the adsorption mechanism. In order to understand the
for comparison. The adsorption amount of Fe on
kinetics of Fe removal using GAC as an adsorbent,
GAC media was calculated using Equation (5) and the
pseudo-first- and second-order models were analysed with
scattered plots were created for kinetic models in order to
the experimental data. This said that the experimental data
understand the relationship between kinetic parameters by
best fits to the linear model for pseudo-second-order kinetics
obtaining the coefficient of determination, which is defined
and it is plotted in Figure 2 together with the adsorption
curve. The pseudo-second-order kinetics model fits the
Table 2 | Operating conditions of both GAC and BPM columns in each phase
experimental data with a coefficient of determination of
Influent Fe 0.96. The parameters K2 and Qe were calculated as
concentration HRT Aeration
0.01 g.min/mg and 1.36 mg/g from the linear plot results.
Phase 1 increased from 7 hours no air supply The adsorption curve depicts the effect of contact time on
1 mg/L to 3 mg/L the adsorption amount of Fe for GAC media. In the adsorp-
Phase 2 3 mg/L reduced from 7 no air supply tion test, the rate of Fe removal was relatively fast at the
to 3.5 hours
beginning due to the greater availability of the GAC surface
Phase 3 3 mg/L 7 hours 2.5 L/min
area. For GAC the hardness sorption shows a quick

Downloaded from http://iwaponline.com/ws/article-pdf/20/8/3233/812638/ws020083233.pdf

by guest
 3238 T. Thinojah et al. | Iron removal filter design for groundwater treatment Water Supply | 20.8 | 2020

Figure 2 | Pseudo-second-order kinetic model for GAC and the curve of adsorption amount.

tendency to reach equilibrium starting at 40 minutes with a to 99% when increasing the influent Fe concentration
maximum present removal attained after about 240 minutes. from 1 to 3 mg/L for the GAC filter. Increased Fe concen-
tration in the water resulted in increasing the collision
Performance of GAC and BPM up-flow aerated filters probability and enhanced the subsequent agglomeration of
flocs, which can also be acknowledged by Equation (1).
Phase 1: GAC and BPM filters were operated with flow And also, for both initial conditions for GAC media, effi-
rates of 10 mL/min and 25 mL/min for 7 hours HRT ciency of Fe removal rises with the increment of media
respectively. The rates of total Fe removal from GAC and height. However considering BPM, the Fe removal effi-
BPM filters at different initial Fe concentrations (1, 3 mg/L) ciency of effluent shows an oscillating variation between
are shown in Figure 3. It can be observed that effluent Fe 20% and 80% for both influent Fe conditions. Influence of
removal efficiency increases approximately from 95% initial pollutant Fe concentration does not show any

Figure 3 | Removal efficiency of total Fe at phase 1 for GAC filter with (a) 1 mg/L influent Fe and (b) 3 mg/L influent Fe, and BPM filter with (c) 1 mg/L influent Fe and (d) 3 mg/L influent Fe.

Downloaded from http://iwaponline.com/ws/article-pdf/20/8/3233/812638/ws020083233.pdf

by guest
 3239 T. Thinojah et al. | Iron removal filter design for groundwater treatment Water Supply | 20.8 | 2020

significant improvement of Fe removal efficiency on BPM. concentration did not meet the WHO drinking water stan-
Irregular increment in Fe removal efficiency through BPM dards at any condition. Finally, it is clear that reducing the
height also proves that a filter run time of 48 hours is not HRT has an uncooperative impact on Fe removal efficiency
enough to reach a steady stage of Fe removal using BPM. for both GAC and BPM filters. However, by comparing
While GAC showed a reduction in effluent Fe concen- both filters, GAC showed better performance than BPM.
tration, BPM showed an increment in effluent Fe Phase 3: Figure 5 compares the effect of air supply to the
concentration when the influent Fe concentration was filtration system which was tested for 3 mg/L Fe contami-
increased. Overall by comparing both filter media, GAC nated SGW with seven hours HRT. The SGW entered in
media provided better performance than BPM and increas- the air–water mixing area from the bottom of the upflow
ing the initial Fe concentration improved the efficiency of filter to be mixed with the air using an air pump with
the GAC filter in Fe removal due to higher media packing 2.5 L/min flow rate. Aeration to the GAC filter showed a
density compared with BPM. sudden increment of Fe concentration of about 0.08 mg/L
Phase 2: At first, two different HRT (HRT ¼ 3.5 hrs, at the beginning and then dropped to 0.05 mg/L after eight
HRT ¼ 7 hrs) were selected to monitor the Fe concentration hours of continuous run of the GAC filter. Then the value
of the outlet water in both filters, and thus the best HRT of was varying between 0.05 and 0.08 mg/L, which is compara-
the effluents was determined. The filters were operated at tively higher than the system without any air supply. Air
flow rates of 10 and 24.5 mL/min namely for GAC and binding was also observed during the GAC column run.
BPM filters at seven hours HRT and these were doubled The BPM filter illustrated a reduction in the effluent Fe
when the HRT was reduced by half. In the steady operation, from approximately 1.5 to 1 mg/L with the effect of air
the treatment effect of Fe with different HRT is compared and then showed a rise in effluent Fe concentration through-
for both filters in Figure 4. The GAC filter shows an increased out the column run with the effect of air. The air bubbles
Fe concentration when the HRT is reduced by half for 3 mg/L stored at the bottom of the BPM were observed in the oper-
influent Fe level. It can be seen that the average of 0.03 rose ation of the filter. The steady stage was not observed for
to 0.05 for GAC media with the reduction of HRT. The incre- either filter during the 48 hours of the filter run. Oxygen
ment in flow rate led to the reduction in contact time of transfer efficiency is the main factor determining the effec-
pollutant with filter media which reduced the Fe removal effi- tive performance of the filter and which depends on the
ciency. Considering the BPM, it also showed an increment in air/water ratio. Previous studies done by researchers say
effluent Fe concentration with reduction of HRT. The effluent that 2–10:1 of air/water ratio gives a better removal of pollu-
concentration increased nearly from 1 mg/L to 1.5 mg/L tant for aerated filters, and also increasing the air–water
with the reduction of HRT. Although BPM showed the ratio can push down the air/water interface, where an
better performance at seven hours HRT, the effluent Fe uneven distribution of air is also possible (Kamarden et al.

Figure 4 | Effluent Fe concentration in phase 2 for GAC and BPM filters.

Downloaded from http://iwaponline.com/ws/article-pdf/20/8/3233/812638/ws020083233.pdf

by guest
 3240 T. Thinojah et al. | Iron removal filter design for groundwater treatment Water Supply | 20.8 | 2020

Figure 5 | Effluent Fe variation in GAC and BPM filters with the effect of 2.5 L/min air supply.

). In this study the air/water ratio reached 100:1, which REFERENCES
is found to be not suitable for this study.
Dvorak, B. I., Skipton, S. O. & Woldt, W.  Drinking Water:
Iron and Manganese, G1714. University of Nebraska–
CONCLUSIONS Lincoln, Lincoln, NE, USA.
Gage, B., O’Dowd, D. H. & Williams, P.  Biological iron and
manganese removal: pilot and full scale applications. In:
In this research a combined use of oxidation and floc-filtration
Ontario Water Works Association Conference, 3 May,
experiment was done by developing GAC and BPM columns Toronto, ON, Canada.
and then compared for the total Fe removal from SGW Han, M., Zhao, Z., Cui, F., Gao, W., Liu, J. & Zeng, Z.  Pre-
while examining at three varieties of phases. A positive result treatment of contaminated raw water by a novel double-layer
biological aerated filter for drinking water treatment.
of increased Fe removal efficiency was found for GAC filter Desalination and Water Treatment 37 (1–3), 308–314. doi:
media when increasing the initial Fe concentration in 10.1080/19443994.2012.661287.
phase 1. The highest water recovery was achieved by increas- Hasan, H. A., Abdullah, S. R. S., Kamarudin, S. K. & Kofli, N. T.
 A review on the design criteria of biological aerated filter
ing the HRT from 3.5 to seven hours for both filter media. An
for COD, ammonia and manganese removal in drinking
optimum condition of 3 mg/L influent Fe with seven hours water treatment. Journal of Institution of Engineers, Malaysia
HRT leading to minimum impurity contamination in the efflu- 70 (4), 25–33.
ent were determined. Introducing an air supply to the system Kamarden, H., Hassan, M. A. A., Noor, Z. Z., Ibrahim, R. K. R. &
Evuti, A. M.  Effect of temperature and air flow rate on
did not show any improvement in either filter performance
xylene removal from wastewater using packed column air
due to the effect of air scouring and improper air/water stripper. Jurnal Teknologi 67 (4), 41–44.
mixing ratio. Overall, the GAC filter media illustrated better Lu, S., Liu, J., Li, S. & Biney, E.  Analysis of up-flow aerated
performance compared with BPM. This is due to the behav- biological activated carbon filter technology in drinking
water treatment. Environmental Technology 34 (16),
iour of BPM, which is loosely packed inside the column and 2345–2351.
has the highest in situ porosity compared with GAC. Hence, Ma, X., Shang, J., Piao, F. & Wu, W.  Biological removal
an up-flow system with co-current air and SGW flow for of iron, manganese and ammonia nitrogen from low
temperature groundwater using biological aerated filter.
GAC filter media was demonstrated to be the most cost-
In: International Conference on E-Product E-Service and
effective treatment method for Fe removal from groundwater. E-Entertainment, 7–9 November, Henan, China.
Marsidi, N., Abu Hasan, H. & Sheikh Abdullah, S. R.  A
review of biological aerated filters for iron and manganese
DATA AVAILABILITY STATEMENT ions removal in water treatment. Journal of Water Process
Engineering 23, 1–12. doi: 10.1016/j.jwpe.2018.01.010.
Mikunthan, T., Vithanage, M., Pathmarajah, S., Arasalingam, S.,
All relevant data are included in the paper or its Supplemen- Ariyaratne, R. & Manthrithilake, H.  Hydrogeochemical
tary Information. Characterization of Jaffna’s Aquifer Systems in Sri Lanka.

Downloaded from http://iwaponline.com/ws/article-pdf/20/8/3233/812638/ws020083233.pdf

by guest
 3241 T. Thinojah et al. | Iron removal filter design for groundwater treatment Water Supply | 20.8 | 2020

International Water Management Institute, Colombo, Sharma, S. K., Petrusevski, B., Owore, D. & Schippers, J. C. 
Sri Lanka. Optimisation of floc formation in iron removal from
Oyegbile, B., Ay, P. & Narra, S.  Flocculation kinetics and groundwater. In: Asian Waterqual 2003: IWA Asia-Pacific
hydrodynamic interactions in natural and engineered Regional Conference, 20–22 October, Bangkok, Thailand.
flow systems: a review. Environmental Engineering Research Sharma, S. K., Petrusevski, B. & Schippers, J. C.  Biological
21 (1), 1–14. doi: 10.4491/eer.2015.086. iron removal from groundwater: a review. Journal of Water
Patil, D. S., Chavan, S. M. & Oubagaranadin, J. U. K.  Supply: Research and Technology – AQUA 54 (4), 239–247.
A review of technologies for manganese removal from doi: 10.2166/aqua.2005.0022.
wastewaters. Journal of Environmental Chemical Sivakumar, S. S.  Reclamation of land and improve water
Engineering 4 (1), 468–487. doi: 10.1016/j.jece.2015.11.028. productivity of Jaffna Peninsula of northern Sri Lanka by
Qin, S., Ma, F., Huang, P. & Yang, J.  Fe (II) and Mn (II) improving the water quality of the lagoons. Research Journal
removal from drilled well water: a case study from a biological of Science and IT Management 2 (8), 20–27.
treatment unit in Harbin. Desalination 245, 183–193. Toit, G. du, Blignaut, H. & Theunissen, B.  Biological
Saravanan, S., Mikunthan, T. & Gunaalan, K.  An overview of Filtration for Sustainable Treatment of Groundwater with
a water quality monitoring in water supply scheme: a case High Iron and Manganese Content
A Case Study from
study on iron present in groundwater wells in Vallipuram, Overstrand Municipality. South Africa.
Jaffna. In: 3rd Water Resources Management Workshop, Xia, S., Nan, J., Liu, R. & Li, G.  Study of drinking water
‘Importance of Groundwater Monitoring in Sri Lanka’, 17 treatment by ultrafiltration of surface water and its
July, BMICH, Colombo, Sri Lanka. application to China. Desalination 170 (1), 41–47. doi: 10.
Senaratne, A. & Dissanayake, C. B.  Palaeogeographic 1016/j.desal.2004.03.014.
reconstruction of Jaffna Peninsula, Sri Lanka. Journal of the Zhang, Y. H.  Filter Media Rids Iron from Groundwater.
Geological Society of India 23, 545–550. American Water Works Association.

First received 6 January 2020; accepted in revised form 8 September 2020. Available online 21 September 2020

Downloaded from http://iwaponline.com/ws/article-pdf/20/8/3233/812638/ws020083233.pdf

by guest