Complexed Iron Removal from Groundwater

Rein Munter1; Heldi Ojaste2; and Johannes Sutt3

Abstract: The paper demonstrates an intensive work carried out and results obtained on the pilot plant of the City of Kogalym Water
Treatment Station 共Tjumen, Siberia, Russian Federation兲 to elaborate on a contemporary nonreagent treatment technology for the local
Downloaded from ascelibrary.org by Universidad Politecnica De Valencia on 05/11/15. Copyright ASCE. For personal use only; all rights reserved.

iron-rich groundwater. Several filter materials 共Birm, Pyrolox, hydroanthracite, Everzit, granulated activated carbon兲 and chemical oxi-
dants 共ozone, chlorine, hydrogen peroxide, oxygen, and potassium permanganate兲 were tested to solve the problem with complexed iron
removal from groundwater. The final elaborated technology consists of raw water intensive aeration in the gas-degas treatment unit
followed by sequential filtration through hydroanthracite and the special anthracite Everzit.
DOI: 10.1061/共ASCE兲0733-9372共2005兲131:7共1014兲
CE Database subject headings: Ground-water pollution; Iron; Oxidation; Adsorption; Ozone; Oxygen; Filtration.

Introduction
The growing demand placed on the world’s water in combination
with more stringent water quality regulations, has placed unprec-
edented demands on municipal water works to provide the public
with drinking water that is safe, reliable, and aesthetically pleas-
ing. This goal is in the case of groundwater frequently compli-
cated by contamination from hydrogen sulfide, acid water, in-
creased content of iron and manganese, volatile organic
compounds, and chlorine-resistant micro-organisms. Iron and
manganese are present in natural waters in their most reduced,
and most soluble, forms: Fe共II兲 and Mn共II兲. They enter ground-
water under reducing conditions existing in waters that have not
been exposed to atmospheric oxygen for long periods or in zones
having active anaerobic bacterial populations. Waters containing
iron can be divided into two main groups: Waters which separate
iron just after aeration and waters where iron remains in the so-
lution after aeration for endlessly long periods 共Sommerfeld
1999兲. In the latter case, we have groundwater in which the iron
behavior depends on organic types and concentrations. Organic
substances 共or silica兲 in water may interfere with the iron removal
process, forming stable complexes with iron, both Fe共II兲 and
Fe共III兲. The Fe共III兲-humic complexes are stronger and more
stable compared to Fe共II兲-humic complexes.
The soil scientists have been quite consistent in stating that
organics which cause these complexes, or peptized small metallic
colloids, are of relatively high molecular weight 共greater than
2,000兲, whereas water chemists generally concede that low
1
Professor, Dept. of Environmental and Chemical Technology, Tallinn
Technical Univ., 5 Ehitajate St., Tallinn 19086, Estonia 共corresponding
author兲. E-mail: rmunt@edu.ttu.ee
2
Dept. of Chemical Engineering, Tallinn Technical Univ., 5 Ehitajate
St., Tallinn 19086, Estonia.
3
Senior Engineer, CUEKS Ltd., 5 Silikaltsiidi, Tallinn 11216, Estonia.
Note. Discussion open until December 1, 2005. Separate discussions
must be submitted for individual papers. To extend the closing date by
one month, a written request must be filed with the ASCE Managing
Editor. The manuscript for this paper was submitted for review and pos-
sible publication on February 26, 2003; approved on December 8, 2003.
This paper is part of the Journal of Environmental Engineering, Vol.
131, No. 7, July 1, 2005. ©ASCE, ISSN 0733-9372/2005/7-1014–1020/
$25.00.
molecular weight organics are responsible for holding iron 共Old-
ham and Gloyna 1968兲.
Jobin and Ghosh 共1972兲 established that in the presence of
organic matter, the Fe共II兲-Fe共III兲 redox couple acts as a catalyst
for the oxidation of organic matter. In such systems, both com-
plexation of Fe共II兲 and reduction of Fe共III兲 by organic matter are
possible:
Fe共II兲 + 1/4O2 + organics → Fe共III兲 − organic complex 共1兲
Fe共III兲 + organics共reduced兲 → Fe共II兲 + organics共oxidized兲 共2兲
Fe共II兲 + organics → Fe共II兲 − organic complex 共3兲
In such cases, the ferrous-ferric system acts as an electron
transfer catalyst for the oxidation of organic material. Depending
on the relative rates of Fe共II兲 oxygenation and Fe共III兲 reduction
by organic material, a substantial retardation of the overall oxida-
tion of Fe共II兲 by organic material is possible 共Ghosh 1965兲. The
higher the pH, the more the oxidation process of the Fe共II兲-
organic complex is retarded. For example, at pH 8, a decrease in
rate constant by a factor of 10 results in doubling the half-life of
the Fe共II兲-organic complex with respect to oxidation 共Stumm
1990兲. Besides the pH, the temperature plays an important role in
chemical reactions kinetics. Unfortunately, there are no data
available on divalent iron or iron-complexes oxidation rates at
lower temperatures 共below 10°C兲.
Some researchers have expressed an opinion that utilization of
a strong oxidizing agent, such as ozone, to break the organic
molecules, thus reducing the color, can also make the organic
substance much less tenacious as a complexing agent 共Knocke et
al. 1992兲.
Cromley and O’Connor 共1976兲 compared ozone with oxygen
for the oxidation of iron in groundwater of high dissolved organic
carbon 共DOC兲 content. They found that ozone resulted in more
rapid precipitation of the iron, although it was not clear how
much of this iron was organically bound. These authors proposed
that high ozone doses resulted in the formation of very stable
iron-organic complexes which could not be oxidized subse-
quently, i.e., ozonation may not always be superior to simple
aeration in the removal of iron by filtration.
It was established 共Reckhow et al. 1991兲 that ozone oxidizes
iron very quickly in the absence of organic matter. However, in

1014 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JULY 2005

J. Environ. Eng. 2005.131:1014-1020.

 Table 1. Main Raw Water Quality Parameters
Parameter Unit Range
pH — 6.60–6.80
Conductivity ␮S / cm 178–208
Dry residue 共105°C兲 mg/ L 140–165
Color mg/ L Pt 26–55
Turbidity NTU 4–28
Fetotal mg/ L 2.7–6.0
Fe2+ mg/ L 2.4–4.0
CODMn mgO2 / L 4.8–6.6
Downloaded from ascelibrary.org by Universidad Politecnica De Valencia on 05/11/15. Copyright ASCE. For personal use only; all rights reserved.

CODCr mgO2 / L 8–13

TOC mg/ L 3.2–6.4
NH4+ mg/ L NH4+ 1.34–1.65
Mn mg/ L 0.09–0.16
Alkalinity mg-eq/ L 1.95–2.15
Note: COD⫽chemical oxygen demand; and TOC⫽total organic carbon.

the presence of humic substances at pH 6.3, the iron may be

completely protected from oxidation and precipitation, if the iron-
organic complexes were formed under reducing conditions.
Among the technologies now in use for dealing with these
challenges are the injection of air, oxygen, or ozone into water
streams for unwanted gases stripping, hydrogen sulfide, iron and
manganese oxidation, precipitate filtration, and postdisinfection
共Voorinen 1988兲. Iron and manganese can also be removed using
biological as opposed to physical/chemical means. However, bio-
logical treatment requires specific raw water qualities and condi-
tions, and not all groundwater or surface water can be treated
economically using this technique 共Mouchet 1992兲.
Waters with high concentrations of humic substances, in which
iron is bound primarily through complexation, will be very resis-
tant to precipitation of the metal through oxidation alone. Fortu-
nately, these high molecular weight organic substances can be
removed easily through chemical coagulation 共Reckhow et al.
1991兲. Coagulation also results in removal of much of the asso-
ciated iron. For this reason, treatment schemes employing coagu-
lation followed by ozonation have been especially effective for
iron removal in colored waters 共Paillard et al. 1989兲. However,
application of coagulation/clarification processes inevitably leads
to several problems with sludge handling and disposal.
The aim of the present study was to elaborate contemporary
treatment technology of high efficiency with minimal chemical
reagent consumption for the groundwater in Kogalym 共district of
Tjumen, Siberia, Russian Federation兲. The city of Kogalym
is supplied by groundwater from 42 wells of 170 m depth. The
water is quite cold 共0 – + 2 ° C兲, contains free dissolved carbon
dioxide 共50– 60 mg/ L兲, colloidal iron 共2.7– 6.0 mg/ L兲, some hy-
Fig. 2. General view of the pilot plant 共on the left the Stage I filter,
in the middle the dosing system, on the right ozonation column兲
drogen sulfide, organic compounds 共total organic carbon
3 – 6 mg/ L兲 and silica 共40 mg/ L SiO2兲. About 63% of all samples
taken from local groundwater in Kogalym does not meet the Rus-
sian standards on drinking water quality. Table 1 summarizes the
ranges of the main raw water quality parameters.
Experimental
Preliminary laboratory tests of raw water aeration and oxidation
clearly indicated that iron in raw water was complexed by organic
compounds; when oxidized by ozone up to the trivalent state, iron
stayed in the solution and was not precipitated. The total iron
guideline for the treated water 共drinking water兲 in Russia is
0.3 mg/ L 共in Estonia and EU countries this guideline is now
0.2 mg/ L兲.
For elaboration of optimal treatment technology for the City of
Kogalym raw water, the pilot plant with a maximum flow rate of
water up to 5 m3 / h was constructed 共Figs. 1 and 2兲. The main
tests were carried out at a flow rate of 1.58 m3 / h. The experimen-
tal setup consisted of a gas-degas treatment 共GDT兲 aeration unit
from the GDT Corp. 共Phoenix, Arizona兲 共GDT DS-100-316兲,
water collection reservoir 共D = 0.8 m; H = 0.5 m兲, contact column
共D = 0.61 m; H = 3 m兲, ozonation bubble column 共D = 0.14 m; H

Fig. 1. Schematic of the pilot plant: 1—water pump; 2—flow meter; 3—injector Mazzei; 4—degassing separator; 5—feed tank; 6—reactor;
7—Stage I filter; 8—ozonation column; and 9—Stage II filter

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JULY 2005 / 1015

J. Environ. Eng. 2005.131:1014-1020.


Downloaded from ascelibrary.org by Universidad Politecnica De Valencia on 05/11/15. Copyright ASCE. For personal use only; all rights reserved.
Fig. 3. Schematic of the gas-degas treatment aerator
= 2.5 m兲, ozone generator GOC-2 共8 gO3 / h from oxygen兲, oxy-
gen generator AS-12 共5 LO2 / min兲, two sequential filters 共both
with D = 0.72 m; H = 3.96 m兲 with different filter media. The filter
media in the first filter were: Hydroanthracite 共diameter of par-
ticles 1.4–2.5 mm; h = 1.1 m兲 and sand 共0.4–0.8 mm; h = 1.0 m兲;
and in the second filter correspondingly Pyrolox 共0.425–0.750
mm; h = 1.2 m兲 or special anthracite “Everzit” 共0.6–2.0 mm;
h = 1.4 m兲 and sand 共0.4–0.8 mm; h = 0.5 m兲. The setup also in-
cluded a dosing system for different oxidants 关hydrogen peroxide,
chlorine 共NaOCl兲 and potassium permanganate兴 injection.
The pH and oxidation-reduction potential 共ORP兲 of water were
measured by Testo 230 共Testo Gmbh & Co兲, the temperature and
concentration of dissolved oxygen by electrode-analyzer model
MJ 97, the concentration of dissolved CO2 by the method of Hach
Comp. 共titration with 0.01 N solution of NaOH兲, color/turbidity
by spectrophotometer DR/2010 共Hach Comp.兲. The color of water
was measured at 455 nm 共Pt–Co scale兲 and turbidity at a wave-
length of 860 nm. The iron content 共ferrous and ferric兲 was de-
termined by spectrophotometer DR/2010 at the wavelength of
510 nm.
Two principally different hypotheses for iron removal mecha-
nisms were tested—external catalysis 共using manganese dioxide
containing filter materials 共Birm, Pyrolox etc.兲, and so-called in-
ternal catalysis, by which the reaction of the divalent iron oxida-
tion is accelerated by the FeO共OH兲 particles remaining in the
filter material 共hydroanthracite, Everzit兲 pores after the filter
backwash.
We were convinced by the first laboratory tests that to avoid
formation of the very strong and stable Fe共III兲-organic com-
plexes, the first step of the raw water treatment—aeration—must
proceed with maximum speed and minimum retention time of
water. This is why we chose the GDT for aeration. The GDT
aeration unit 共Fig. 3兲 consists of a patented Mazzei venturi injec-
tor to dynamically introduce gases 共air, oxygen, and ozone兲, a
patented centrifugal degassing separator to quickly and effec-
tively remove unwanted entrained gases, a degas relief valve to
release the captured gases for processing or venting, and a back-
pressure control valve on the water outlet pipeline from the sepa-
rator. The GDT process has proven to be extremely efficient in
dynamically transferring gases into pressurized solutions while
removing unwanted entrained or dissolved gases. In addition to
its technological benefits, the GDT process features a small foot-
print, low costs of capital, installation and operation, easy instal-
lation, and nonfouling pressurized operation.
For the pilot plant tests, the smallest GDT unit with water flow
rate of 1.1– 11.4 m3 / h was purchased 共Mazzei injector model
1016 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JULY 2005
J. Environ. Eng. 2005.131:1014-1020.
Fig. 4. Final flowsheet of groundwater treatment
885-X together with the degassing separator model DS-100-316兲.
The pilot plant was in operation with some short intervals for
more than 2.5 years 共December 1999–September 2002兲. During
the tests, the main operating parameters for determining the qual-
ity of filtered water were measured and varied: Detention time of
water in the GDT aerator, doses, and injection points of the oxi-
dants 共O2 , O3 , H2O2, and NaOCl兲, type of filtration medium, fil-
tration rate, etc.
Results and Discussion
In 1999–2000, several flowsheets 共schemes兲 for raw well water
treatment were tested on the pilot plant in Kogalym. Below, the
obtained results are summarized and the final flowsheet for full-
scale application is selected. As the main quality parameters of
treated water, the total and divalent iron content, color and turbid-
ity were chosen.
At the first stage of the pilot tests, our hopes were set mainly
on a new efficient catalytic filter medium for iron oxidation and
removal—Pyrolox 共American Minerals兲. Pyrolox is a very rare
form of almost pure 共100%兲 manganese dioxide. It can be pur-
chased in a natural form 共granules 0.750–2.36 mm兲 or crushed to
0.425–0.750 mm size. This material is extremely hard, like gar-
net, and heavy 共hardness 1–2, density 4.5– 5.0 g / cm3兲. It does not
need chemical regeneration with a solution of potassium perman-
ganate, such as Birm, Filox, and Manganese Greensand, but only
mechanical backwash at the rate of 60– 70 m / h in 24 h, and
dosing of a relatively small amount of active chlorine
共0.1– 1.0 mg/ L兲 at the inlet of the filter to keep the medium ac-
tive.
To break down the iron complexes formed in the aerated water
collection reservoir, ozone 共0.5– 1.6 mg/ L兲 and hydrogen perox-
ide 共0.45– 2.0 mg/ L兲 were injected into water before the Pyrolox
filter. The best quality parameters of the water filtered through
Pyrolox were obtained in the case of using preozonation: The
total iron content was reduced by 93% and divalent iron content
by 98.7%. The color of the filtered water was 1° 共Pt–Co兲 and
turbidity 0 mg/ L. It was also established that the hydrogen per-
oxide and °OH radicals formed in the combined oxidation system
共O3 / H2O2 = 1.6: 1兲 were too weak for iron complexes effective
oxidation: The divalent iron content in the filtrate increased up to
0.2– 0.25 mg/ L.
Using the activation of the Pyrolox filter with small doses of
sodium hypochlorite 共0.11 mg/ L兲, postozonation and a final pol-
ishing filter with granulated activated carbon 共GAC兲, even better
results were achieved: The total iron content was reduced by
95.3% and divalent iron content by 100%. The Pyrolox filter re-
moved 92% of total iron and 99% of divalent iron. The ferric iron
complexes which passed through the Pyrolox filter, were broken
down by ozone and the ferric hydroxide precipitate together with
organic byproducts of ozonation were adsorbed in the GAC filter.
However, by the end of this cycle of pilot tests Pyrolox had
lost about 50% of its initial activity due to the formation of a quite
 Table 2. Results of the Final Pilot Tests 共September 2, 2002兲
O2 CO2 Fe共II兲 Fetotal Color Turbidity CODMn
Sample 共mg/ L兲 共mg/ L兲 pH 共mg/ L兲 共mg/ L兲 共degrees兲 共NTU兲 共mgO2 / L兲
Raw water 0.1 40 6.63 1.90 4.40 17 0 6.40
After GDT 8.70 32 — — — — — —
After Filter I 0.70 28 6.56 1.50 2.40 15 0 —
After Filter II 5.40 26 6.55 0 0.1 8 0 4.40
Note: Time: 2.00 p.m.; temperature of raw water 共+2兲 ° C; and GDT⫽gas-degas treatment.
Downloaded from ascelibrary.org by Universidad Politecnica De Valencia on 05/11/15. Copyright ASCE. For personal use only; all rights reserved.

Table 3. Results of the Final Pilot Tests 共September 3, 2002兲

O2 CO2 Fe共II兲 Fetotal Color Turbidity CODMn
Sample 共mg/ L兲 共mg/ L兲 pH 共mg/ L兲 共mg/ L兲 共degrees兲 共NTU兲 共mgO2 / L兲
Raw water 0.2 38 6.63 3.20 4.00 27 0 4.80
After GDT 9.00 34 — — — — — —
After Filter I 0.00 26 6.52 0.80 1.00 10 0 —
After Filter II 5.80 26 6.49 0.00 0.10 5 0 3.60
Note: Time: 8.30 a.m.; temperature of raw water 共+2兲 ° C; and GDT⫽gas-degas treatment.

Table 4. Results of the Final Pilot Tests 共September 6, 2002兲

O2 CO2 Fe共II兲 Fetotal Color Turbidity CODMn
Sample 共mg/ L兲 共mg/ L兲 pH 共mg/ L兲 共mg/ L兲 共degrees兲 共NTU兲 共mgO2 / L兲
Raw water 1.4 38 6.60 2.88 4.00 27 0 5.54
After GDT 7.6 34 — — — — — —
After Filter I 1.1 32 6.58 0.12 0.21 12 0 3.96
After Filter II 0.5 30 6.50 0.07 0.12 7 0 2.77
Note: Time: 9.00 a.m., Temperature of raw water 共+2兲 ° C, GDT⫽Gas-degas treatment.

Table 5. Results of the Final Pilot Tests 共September 18, 2002兲

O2 CO2 Fe共II兲 Fetotal Color Turbidity CODMn
Sample 共mg/ L兲 共mg/ L兲 pH 共mg/ L兲 共mg/ L兲 共degrees兲 共NTU兲 共mgO2 / L兲
Raw water 0.10 — 6.50 2.97 4.48 14 0 5.12
After GDT 8.50 — 6.67 — — — — —
After Filter I 0.00 — 6.51 0.33 0.34 3 0 —
After Filter II 3.30 — 6.44 0.01 0.10 1 0 2.60
Note: Time: 4.30 p.m.; temperature of raw water 共+2兲 ° C; and GDT⫽gas-degas treatment.

Table 6. Results of the Final Pilot Tests 共September 30, 2002兲

O2 CO2 Fe共II兲 Fetotal Color Turbidity CODMn
Sample 共mg/ L兲 共mg/ L兲 pH 共mg/ L兲 共mg/ L兲 共degrees兲 共NTU兲 共mgO2 / L兲
Raw water 0.2 46 6.80 2.28 3.76 28 0 5.84
After GDT 7.6 40 6.90 — — — — —
After Filter I 0.1 38 6.82 0.70 0.80 10 0 3.76
After Filter II 2.90 34 6.76 0.01 0.05 6 0 2.57
Note: Time: 2.00 p.m.; temperature of raw water 共+2兲 ° C; and GDT⫽gas-degas treatment.

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JULY 2005 / 1017

J. Environ. Eng. 2005.131:1014-1020.

 tight yellowish precipitate on the surface of its particles. The ac-
tivity of the Pyrolox filter was restored by washing the medium
with chlorinated water. The chemical analysis of this precipitate
revealed the presence of hydrophosphates of iron and aluminum
共the phosphates PO43− content in the raw water was about
0.6 mg/ L兲. It was decided to add to the process scheme a prefilter
to protect the Pyrolox filter from possible precipitate formation
and inactivation.
In 2000, the scheme of the pilot plant was modified: As the
protecting prefilter, a filter with hydroanthracite 共40% sand 0.35–
1.0 mm兲 was connected to the scheme, the aerated water was led
Downloaded from ascelibrary.org by Universidad Politecnica De Valencia on 05/11/15. Copyright ASCE. For personal use only; all rights reserved.

directly to the hydroanthracite filter without the water reservoir,

and to the Pyrolox filter about 35% of sand was added to improve
the filtration properties. At the second stage of the pilot tests, the
impact of different oxidants, postozonation, prefiltration, and
GAC filtration was studied. The tests with hydroanthracite prefil-
ter clearly indicated that in the case of using both filters the total
iron content in filtered water was 1.7 times and color 1.5 times
lower.
However, when analyzing the load of both filters by total iron
removal, it appeared that the hydroanthracite prefilter removed
about 90–94% of total iron and the Pyrolox filter 6–10%, i.e., the
total iron average removal efficiency of the hydroanthracite was
almost the same as that of the Pyrolox filter 共92%兲 when used as
a one-stage filter. Thus, on the basis of these results it is impos-
sible clearly to prefer one hypothesis 共external catalysis of iron
oxidation using Pyrolox兲 to another 关internal catalysis by the
FeO共OH兲 film in filter media兴.
It was also established that postozonation had a quite positive
additional impact on the filtered water quality: The CODMn was
reduced by 20%, the color was reduced further by 10°, and in-
stead of a total 22% of color reduction in the scheme, 54% was
achieved. However, there was no impact of postozonation on the
degree of total iron reduction 共92–94%兲.
When a filter with GAC was connected after the postozonation
column, the total iron content in the filtered water was
0 – 0.07 mg/ L 共reduction by 98–100%兲, color 2–5°, turbidity less
than 0.58 mg/ L, and pH 7.0–7.8. The excellent quality of treated
water can be explained by the impact of the GAC filter. It is well
known that GAC is a very good adsorbent for heavy metals, or-
ganics, residual oxidants, and that on its surface several oxidation
processes proceed, especially when ozone is added at the inlet to
the filter. However, due to the quite high price of the GAC, it was
decided to give it up, and to replace Pyrolox in the second filter
by a new special anthracite—Everzit.
Everzit-Special is a rather new filtration material which effi-
ciently removes THM and other chloro-organic compounds, am-
monia, and phosphates from water. Its great advantage is that this
material is able to reduce iron and manganese contents in water
significantly, including also complexed iron. At the same time,
there is no need to use strong oxidants, such as chlorine, ozone,
hydrogen peroxide, and potassium permanganate. It is only
needed to have enough dissolved oxygen in water to form the
self-regenerating catalytic film of FeO共OH兲 on the surface of the
Everzit particles.
Giving up on ozone also eliminated problems with ozonation
byproducts 共aldehydes, ketones, etc.兲 control. Everzit-Special has
particle diameter of 0.6–2.0 mm and specific interfacial area
about 350 m2 / g. It is used in open filters at filtration rates of
5 – 10 m / h with the water backwash rate of 30– 35 m / h.
Taking into consideration the experience collected during the
previous tests, we modified the pilot setup to the following very
simple scheme which served as a basis for the design of the
Fig. 5. General view of the big gas-degas treatment aerators at the
Kogalym water treatment plant 共above raw water collection tank,
injectors, degassing separators below兲
full-scale plant. The final flowsheet diagram of the pilot plant is
described by the scheme in Fig. 4. The raw water is quickly and
intensively aerated in the GDT unit and led directly to the first
anthracite filter. The filtered water is enriched with oxygen and
then led to the second polishing filter with Everzit.
Tables 2–6 give the results of the final pilot tests in September
2002. The GDT unit was operated at the inlet pressure of 2.5 bar
and at the outlet pressure of 0.6 bar. The less the back-pressure,
the more entrained gas 共CO2 and H2S兲 will be desorbed and the
more oxygen from air will be dissolved. However, for efficient
iron removal the concentration of dissolved oxygen as well as
ORP is limited. For example, our pilot plant tests revealed that for
noncomplexed iron removal the best conditions at pH= 7.0 are
+300 mV⬍ ORP⬍ + 500 mV. If ORPⰆ 300 mV, then iron is
present in water as divalent cations Fe2+. If ORPⰇ 500 mV, then
iron will dissolve in water in the form of cations Fe共OH兲2+.
Given values of ORP are valid only for the pure system
Fe– H2O. Our ORP measurements on the pilot and full-scale plant
in the conditions of iron-organics complexes showed that the iron
removal process proceeds successfully when the average ORP of
water after the first filter is +共96– 100兲 mV, and +共200– 205兲 mV
after the second filter. The raw water ORP was in the range of
−共56– 90兲 mV. The concentration of dissolved oxygen after the
GDT was in the range of 7.6– 9.0 mg/ L. If the concentration of
dissolved oxygen after the first filter is ⬍1 mg/ L, then the water
before the second stage filter must be enriched with oxygen from
the oxygen generator.
Fig. 6. General view of the Stage I filters with hydroanthracite
共above兲 and Stage II filters with Everzit 共below兲 at the Kogalym
water treatment plant

1018 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JULY 2005

J. Environ. Eng. 2005.131:1014-1020.

 Table 7. Results of the Full-Scale Tests 共September 8, 2002兲
O2 CO2 Fe共II兲 Fetotal Color Turbidity CODMn
Sample 共mg/ L兲 共mg/ L兲 pH 共mg/ L兲 共mg/ L兲 共degrees兲 共NTU兲 共mgO2 / L兲
Raw water 1.7 40 6.48 3.10 3.88 42 0 6.0
After GDT 7.40 32 6.28 2.79 3.84 22 0 —
After Filter 4-I 6.70 30 6.58 2.39 3.12 31 0 —
After Filter 4-II 4.00 22 6.68 0.40 0.70 14 0 —
After Filter 4-III 5.90 34 6.59 1.87 2.55 15 0 —
After Filter 4-IV 5.20 32 6.62 1.52 1.97 34 0 —
After Filter 4-V 6.30 32 6.61 1.84 2.42 24 0 3.80
Downloaded from ascelibrary.org by Universidad Politecnica De Valencia on 05/11/15. Copyright ASCE. For personal use only; all rights reserved.

Note: Time: 9.00 a.m.; temperature of raw water +共1.8兲 ° C; and GDT⫽gas-degas treatment.

Using the above-described system of two sequential filters,
dissolved and complexed iron was quite efficiently removed. The
raw water flowrate was 1.58 m3 / h. Both filters were operated at
the filtration rate of 3.89 m / h with air/water backwash in 24 h.
The first-stage filter 共hydroanthracite兲 removed 69–96% of diva-
lent iron and 75–95% of total iron. The second-stage filter
共Everzit兲 worked as a good polishing filter reducing the total iron
content 共including iron complexes兲 at the outlet of the filter to
0.05– 0.12 mg/ L which is much less than the required Russian
guideline 0.3 mg/ L, and easily meets the EU guideline
共0.2 mg/ L兲. The manganese content in the filtered water was
0.1 mg/ L 共Russian guideline 0.1 mg/ L, EU guideline
0.05 mg/ L兲, phosphate content 0.2 mg/ L, and ammonia content
1.2 mg/ L.
Due to the reduced amount of air sucked into the injector at
elevated backpressure, the CO2 removal degree of the GDT unit
was much less 共15–20%兲 than possible 共60–70%兲 when operating
at minimum backpressure. The pH of the filtered water was in the
range of 6.40–6.76 which means that poststabilization of treated
water with soda 共Na2CO3兲 is really needed.
More detailed studies of the dissolved organic compounds in
Kogalym groundwater using exclusion chromatography revealed
the presence of two fractions, with high molecular mass
共370,000–416,000兲, and low molecular mass 共1,500–2,240兲. It
was also established that 59.5% of iron and 6.7% of manganese
was complexed with organic matter.
On the basis of the pilot plant tests, the full-scale water treat-
ment plant for the City of Kogalym was designed by the Finnish-
Estonian engineering company PIC ESTONIA Ltd. 共Tallinn, Es-
tonia兲 and constructed by the company CUEKS Ltd. 共Tallinn,
Estonia兲. The maximum capacity of the plant is 2,500 m3 / h of
water. Fifty percent of the plant’s capacity was successfully put

Table 8. Water Quality Parameters of the Kogalym Plant in December 12–15, 2002
O2 Fe共II兲 Fetotal Color
Date Time pH 共mg/ L兲 共mg/ L兲 共mg/ L兲 共degrees兲
12.12.02 10.30 a.m.
Raw water 6.99 0.0 2.27 3.72 37
After GDT 6.90 9.7 1.29 3.50 13
After Filter 4-III 6.90 0.1 0.13 0.32 11
After Filter 5-III 6.72 0.3 0.09 0.18 10
12.12.02 5.00 p.m.
Raw water 6.86 0.0 2.95 3.64 35
After GDT 6.91 9.7 2.48 3.62 14
After Filter 4-III 6.89 0.8 0.22 0.28 7
After Filter 5-III 6.76 0.1 0.10 0.20 5
12.13.02 9.30 a.m.
Raw water 6.81 0.5 2.61 3.66 33
After GDT 6.88 9.5 2.41 3.14 24
After Filter 4-III 6.80 0.5 0.06 0.24 13
After Filter 5-III 6.68 0.2 0.03 0.14 8
12.14.02 9.00 a.m.
Raw water 6.81 0.1 2.27 3.72 37
After GDT 6.88 9.7 1.29 3.54 30
After Filter 4-III 6.82 0.8 0.06 0.19 8
After Filter 5-III 6.68 0.2 0.03 0.14 6
12.15.02 9.00 a.m.
Raw water 6.81 0.0 3.12 3.58 19
After GDT 6.58 9.5 2.48 3.45 16
After Filter 4-III 6.87 1.2 0.05 0.21 13
After Filter 5-III 6.68 0.1 0.02 0.16 9
Note: GDT⫽gas-degas treatment.

JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JULY 2005 / 1019

J. Environ. Eng. 2005.131:1014-1020.

 into operation in September 2002. The plant flow sheet is the
following: The raw water is aerated in five parallel GDT aeration
units 共Fig. 5兲 共each with maximum capacity up to 750 m3 / h兲, and
then led to ten filters of the first stage with hydroanthracite and
sand. From the filters of the first stage, the filtered water moves
by gravity flow to the ten filters of the second stage with a special
anthracite Everzit and sand 共Fig. 6兲.
Before the filters of the second stage, the water is enriched
with oxygen from the oxygen generators to help the formation of
the catalytic film FeO共OH兲 on the surface of the Everzit. After
that, the treated water is pumped into two reservoirs of drinking
Downloaded from ascelibrary.org by Universidad Politecnica De Valencia on 05/11/15. Copyright ASCE. For personal use only; all rights reserved.
water and disinfected with chlorine. It is also planned to put into
operation the unit of drinking water stabilization soon to avoid
pipeline corrosion 共increasing pH up to 7.5 by dosing of a 15%
solution of soda ash 共Na2CO3兲.
On September 7, 2002, the full-scale test with the GDT unit
and first-stage filters was started. Two GDT units from the five
units 共Fig. 5兲 and five first-stage parallel filters 共from 4-I up to
4-V兲 from ten were put into operation with the total flow rate of
water 1,150 m3 / h. Table 7 summarizes some results of the full-
scale tests on September 8, 2002. The dissolved oxygen after the
GDT unit was 7.40 mg/ l and CO2 was removed by 20%.
From the five parallel first-stage filters 共4-I–4-V兲 outlet, the
following average values were obtained: pH 6.60; Fe共II兲
1.6 mg/ L; Fe共total兲 2.15 mg/ L; color 24°; turbidity zero. This
means that the divalent iron was reduced only by 48% and the
total iron only by 44.5%. Comparing these values of divalent and
total iron content here with the corresponding values obtained on
the pilot plant at the outlet of the first-stage filter, one can see that
the full-scale filters had not yet reached the optimum regime 共the
samples were taken 17 h after the backwash of the five first stage
filters兲. Formation of the internal catalytic layer FeO共OH兲 inside
of the anthracite filters obviously took time.
By December 12, 2002 the Kogalym Plant had been in con-
tinuous operation with the first- and second-stage filters already
about three months. The iron removal efficiency of both stages
had improved significantly. Table 8 illustrates the water quality
parameters registered on December 12–15, 2002. On the basis of
the divalent and total iron concentrations after the first-stage filter
4-III and after the second-stage filter 5-III, it can be concluded
that the full-scale plant behavior in general has been quite similar
to the pilot plant one. The prevailing amount of total iron 共90–
94%兲 is separated in the first-stage anthracite filters. The second-
stage polishing Everzit filters reduce the divalent iron content to
0.02– 0.1 mg/ L and total iron content to 0.14– 0.20 mg/ L which
are higher than the values obtained on the pilot plant but still meet
the Russian as well as the EU guidelines. Oxygen is intensively
used for the catalytic FeO共OH兲 film buildup in both filters as its
remaining concentration at the outlet of the filters is low
共0.1– 1.2 mg/ L兲. Between the two filters, water is enriched with
pure oxygen from the oxygen generator. An important role of the
FeO共OH兲 film in the iron removal process was just occasionally
proved when one of the first-stage hydroanthracite filters was, by
accident, thoroughly backwashed with chlorinated water. The iron
removal efficiency of this filter dropped to zero, and it took al-
most 4 days to restore the activity of the filter.
Conclusions
Successful removal of divalent and trivalent iron from groundwa-
ter under the conditions when iron is complexed with organic
material has always been a real challenge to water treatment spe-
1020 / JOURNAL OF ENVIRONMENTAL ENGINEERING © ASCE / JULY 2005
J. Environ. Eng. 2005.131:1014-1020.
cialists. This problem may be solved in two different ways: First,
using almost all available current water treatment operations
共aeration, coagulation, flocculation, clarification, oxidation, and
filtration兲 together with treatment problems of significant amount
of sludge, or, second, using a new modern nonreagent treatment
technology 共fast intensive aeration together with the selection of
appropriate, and new filtration media capable of separating iron-
organics complexes兲. In this study, the second solution was elabo-
rated: Intensive aeration in the GDT unit together with the se-
quent two-stage filtration through hydroanthracite and special
anthracite Everzit. The elaborated technology does not include
any strong oxidants other than the usual gaseous oxygen to keep
the concentration of dissolved oxygen in water at the optimal
level between the stage I and II filters.
Almost the same iron removal efficiency can be achieved
using the manganese dioxide containing external catalytic filter
materials such as Pyrolox, together with postozonation and a
GAC polishing filter. However, this option will be significantly
more expensive. Postozonation is needed to break down the ferric
iron complexes which passed through the Pyrolox filter as, differ-
ent from Everzit Special, Pyrolox does not remove complexed
iron.
Acknowledgments
The financial support of the Estonian Science Foundation 共Grant
No. 4669兲 is gratefully acknowledged. The authors also thank
M.Sc. Ergo Rikmann and B.Sc. Mait Soost for their help and
cooperation in carrying out the experiments at the pilot plant.
References
Cromley, J. T., and O’Connor, J. T. 共1976兲. “Effect of ozonation on the
removal of iron from groundwater.” J. Am. Water Works Assoc.,
68共6兲, 315–319.
Ghosh, M. M. 共1965兲. “Filtration as related to the removal of iron from
groundwaters.” Sanitary Eng. Series, No. 32, Univ. of Illinois, Ur-
bana, Ill.
Jobin, R., and Ghosh, M. M. 共1972兲. “Effect of buffer intensity and or-
ganic matter on the oxygenation of ferrous iron.” J. Am. Water Works
Assoc., 64共9兲, 590–595.
Knocke, W. R., Shorney, H. L., and Bellamy, J. 共1992兲. “Reactions be-
tween soluble iron, dissolved organic carbon, and alternative oxidants
during conventional water treatment.” Proc., AWWA Annual Confer-
ence, Denver, Colo.
Mouchet, P. 共1992兲. “From conventional to biological removal of iron and
manganese in France.” J. Am. Water Works Assoc., 84共4兲, 158–167.
Oldham, W. K., and Gloyna, E. F. 共1968兲. “Effect of colored organics on
iron removal.” Proc., Canadian Section Meeting, Toronto, Ont.,
Canada.
Paillard, H., Legube, B., Bourbigot, M. M., and Lefebre, E. 共1989兲. “Iron
and manganese removal with ozonation in the presence of humic sub-
stances.” Ozone: Sci. Eng., 11共1兲, 93–114.
Reckhow, D. A., Knocke, W. R., Kearney, M. J., and Parks, C. A. 共1991兲.
“Oxidation of iron and manganese by ozone.” Ozone: Sci. Eng.,
13共6兲, 675–695.
Sommerfeld, E. O. 共1999兲. Iron and manganese removal handbook,
American Water Works Association, Denver, Colo.
Stumm, W., ed. 共1990兲. Aquatic chemical kinetics, Wiley-Interscience,
New York.
Voorinen, A. 共1988兲. “Chemical, mineralogical, and microbiological fac-
tors affecting the precipitation of Fe and Mn from groundwater.
“Water Sci. Technol. 20共3兲, 249–259.