Content uploaded by Bjørnar Eikebrokk
Author content
All content in this area was uploaded by Bjørnar Eikebrokk on May 24, 2016
Content may be subject to copyright.
NOM removal from drinking water by chitosan
coagulation and filtration through lightweight expanded
clay aggregate filters
Bjo¨ rnar Eikebrokk and Torgeir Saltnes
Bjo¨ rnar Eikebrokk (corresponding author)
SINTEF Applied Chemistry,
Dept of Water and Wastewater,
N-7465 Trondheim,
Norway
E-mail: bjornar.eikebrokk@sintef.no
Torgeir Saltnes
NTNU-Norwegian University of Science and
Technology,
Department of Hydraulic and Environmental
Engineering,
N-7491 Trondheim,
Norway
ABSTRACT
Recent innovations in Norway regarding coagulation-contact filtration for the removal of natural
organic matter (NOM) include chitosan as a natural and biodegradable coagulant, and filter media
based on lightweight expanded clay aggregates (Filtralite). The main advantages associated with
chitosan are: reduced solids production compared with conventional coagulants; and treatment and
disposal of natural, biodegradable sludge, which does not contain metal hydroxides from
metal-based coagulants. Filtralite can be produced with an inverse relationship between grain size
and density, thus allowing an approximation to the ideal situation of decreasing grain size in the
direction of flow.Traditionally, this important property of a filter bed is utilised in up-flow filters, or in
dual or multimedia down-flow filters with combinations of two or more filter media. This paper
presents experimental results from pilot-scale treatment of NOM-containing raw waters using
chitosan for coagulation and expanded clay aggregates as filter media. A dual media anthracite-sand
filter and alum coagulant was used as a reference for comparison with conventional process
configurations.
Key words |chitosan, coagulation, direct filtration, expanded clay aggregates, NOM removal
INTRODUCTION
Natural organic matter (NOM) in drinking water has
gained a lot of attention in many countries in recent years.
The main reasons for this are concerns regarding colour,
taste and odour, known and unknown disinfection
by-products (DBP), formation of biofilms in distribution
systems, and increased availability of micro-pollutants
associated with NOM. In 1998, the applied NOM removal
technologies in Norwegian treatment plants included
coagulation and filtration (74 plants), macro-porous
anion exchange (12), nanofiltration (63), and ozonation-
biofiltration (1).
Most coagulation filtration plants apply aluminium-
based coagulants and dual media anthracite-sand filters.
Recently, some plants have shifted to iron-based coagu-
lants and calcium carbonate (CaCO
3
) as a filter medium in
up-flow single medium alkaline filters, or in down-flow
triple media filters with anthracite, sand and CaCO
3
. The
main goal is to obtain NOM-removal and corrosion con-
trol, i.e. increased levels of calcium, alkalinity and pH. A
typical scheme of a coagulation-contact filtration plant for
NOM removal and corrosion control is shown in Figure 1.
Metal-based coagulants are used with conventional dual
media anthracite-sand filters, with lime and carbon diox-
ide for pH and corrosion control purposes. Due to the
loose nature of the metal–NOM coagulation products and
the high solids production from elevated coagulant doses
required for NOM control, the filter run lengths are short
compared with those achieved in turbidity removal. In
addition, some plants have problems complying with the
admissible residual metal concentration levels. Against
this background, optimisation of coagulation and filtra-
tion design and operation are important issues, including
testing and evaluation of alternative coagulants and types
of filter media.
323 © IWA Publishing 2002 Journal of Water Supply: Research and Technology—AQUA |51.6 |2002
NOM-RELATED WATER QUALITY REGULATIONS
In order to maintain low concentrations of DBP, water
quality standards normally regulate the maximum admis-
sible concentration levels of NOM before chlorination. In
general, the Norwegian water quality standard has set the
maximum colour and organic carbon levels to 20 mg Pt
l
−1
and 5 mg TOC l
−1
, respectively. The maximum levels
of aluminium and iron are 0.2 mg l
−1
. However, when
used as coagulants in water treatment the maximum
admissible level is 0.1 mg l
−1
. In contrast to the maximum
colour and TOC levels established in Norway, the US
regulations on disinfectants and DBPs (USEPA 1998)
require 0–50% removal of TOC depending on raw water
TOC and alkalinity levels. In addition, many countries
also regulate the maximum admissible level of DBP, i.e.
trihalomethanes (THM) and haloacetic acids (HAA).
NOM COAGULATION AND FILTRATION
In the US the term ‘enhanced coagulation’ has evolved to
describe a coagulation process optimised for NOM-
removal, which was initially designed for turbidity
removal. Enhanced coagulation normally implies the use
of elevated coagulant doses and stricter pH control to
obtain the required 0–50% reduction in TOC. In high
alkalinity waters an inorganic acid or coagulant overdos-
ing is used to reduce the pH to optimum levels, normally
in the range of 5–6. In low alkalinity raw waters, however,
a base is normally required to prevent the pH from drop-
ping below the optimum level due to the acid reaction
of the metal coagulants. The elevated coagulant dose
requirements and sludge production rates normally
associated with increased NOM removal result in rela-
tively short filter runs due to early breakthroughs, more
rapid headloss development, increased backwash water
consumption, and the need for increased sludge process-
ing and disposal capacity. The importance of these factors
is even more pronounced in direct or contact filtration
systems. In this context it is important to evaluate alterna-
tive types of coagulants with respect to the potentials for
obtaining reduced coagulant doses and sludge production
rates, as well as a more easily disposable sludge. In
addition, new types of filter media need to be considered
that may be more adapted to or tailor-made for
coagulation-contact filtration processes primarily
designed for NOM removal. In Norway there is consider-
able interest in chitosan as a natural and non-toxic
Figure 1 |Schematic of a coagulation-contact filtration process with metal-based coagulants, dual media anthracite-sand filters, and lime and carbon dioxide for pH- and corrosion
control (Ødegaard et al. 1999).
324 B. Eikebrokk and T.Saltnes |NOM removal using chitosan and expanded clay aggregates Journal of Water Supply: Research and Technology—AQUA |51.6 |2002
alternative to metal-based coagulants, and in lightweight
expanded clay aggregates (Filtralite) as an alternative filter
media.
Chitosan as a coagulant
Chitosan (Figure 2) is a natural cationic biopolymer
obtained from full or partial deacetylation of chitin, the
structural polymer of the outer skeleton of insects and
crustaceans (shrimp and crab shells). The production pro-
cess involves processing of the shells, deacetylation, dry-
ing and milling. The degree of acetylation of the product
used here (Primex ChitoClear 90) is more than 90%. The
cationic charge stems from protonation of the amino
groups. In the experiments reported here the coagulant
was dissolved in hydrochloric acid, with a concentration
of the dosing solution of 10 g l
−1
. The coagulation abilities
have been known for a long time and, as it is virtually
non-toxic, this coagulant has undoubtedly a potential also
for the coagulation of NOM (Eikebrokk 1999). The disad-
vantage is that the current price level of chitosan makes
competition with metal-based coagulants and synthetic
polymers difficult. However, few results have been pre-
sented regarding important aspects of NOM-coagulation
and filtration performance with chitosan.
Expanded clay aggregates as filter media
Lightweight expanded clay aggregates (Filtralite) are pro-
duced by burning clay at high temperatures (1100°C).
Filtralite can be produced with different porosity, i.e. dry
density. In the experiments presented here, the qualities of
Filtralite used were high density crushed (F-HC) and
normal density crushed (F-NC), with approximate wet
densities of about 1,800 and 1,200 kg m
−3
, respectively.
Pictures of Filtralite and anthracite grains are presented in
Figure 3, illustrating well the difference in porosity for the
anthracite and Filtralite NC grains. The settling velocity of
the filter grains following bed expansion during backwash
determines the stratification and the layer structure of a
filter bed. Thus, a correct combination of grain size and
filter material (i.e. density) in the different layers is essen-
tial to obtain the required decrease in grain size and
increase in grain settling rates with depth in a down-flow
filter bed (Ives 1979). As presented in Figure 4, initial
calculations on the settling rates at transient hydraulic
conditions (Newton) of the initially produced 0.5–1.6 mm
Filtralite HC and 1.6–2.5 mm NC grains showed a risk of
intermixing of the F-HC grains below 0.8 mm with the
largest F-NC grains (above 2.3 mm). Therefore, F-NC
grains below 0.8 mm in size were removed before the start
Figure 2 |Characterisation of chitin (top) and chitosan (Smidsrød & Moe 1995).
Figure 3 |Grains of anthracite (left) and Filtralite NC (right).
Figure 4 |Calculated settling rates of Filtralite HC and NC grains at transient hydraulic
conditions (5 and 10°C). Assumed wet densities: 1,850 and 1,240 kg m−3,
respectively.
325 B. Eikebrokk and T.Saltnes |NOM removal using chitosan and expanded clay aggregates Journal of Water Supply: Research and Technology—AQUA |51.6 |2002
of the experiments. In these calculations, the anticipated
wet densities were set to 1240 and 1850 kg m
−3
for F-NC
and F-HC.
In order to characterise the grains by settling velocity,
30 grains covering the given size range of the two Filtralite
fractions were measured and compared with anthracite
and sand grains (Figure 5). The settling velocities of the
grains were measured in a column of water as a function of
grain storage time in water. When soaked in water, grain
density and settling rates may increase because water is
penetrating the pores. Figure 5 shows that this is true for
the high porosity grains of F-NC where settling rates
appear to increase for a period of several weeks. However,
the settling rates of grains of F-HC, anthracite and sand
are constant during storage in water. The measured set-
tling rates are lower than those calculated in Figure 4, for
F-HC grains in particular, indicating that the actual wet
densities are lower than those used for the calculations.
However, no significant overlap in settling rates between
the grains of the different fractions was detected, and the
given Filtralite fractions were applied in pilot-scale
coagulation-contact filtration experiments.
COAGULATION-CONTACT FILTRATION
EXPERIMENTS
The range of raw waters tested was considered to cover
most Norwegian humic waters commonly treated by
coagulation and filtration. The raw waters had typical
colour levels from 15 (RW15) to 50 mgPt l
−1
(RW50). The
corresponding typical organic carbon levels were 2.3 and
5 mgNPOC l
−1
, respectively, and the range of specific
UV-absorption (SUVA) was 3.8–5 l m
−1
mgC
−1
.
Turbidity levels were in the range of 0.1–0.2 NTU. The
coagulants tested with this raw water included:
– alum (ALG)
– poly aluminium chloride (PAX14), and PAX with
high calcium content (Ca- PAX)
– ferric chloride sulphate (JKL)
– chitosan (Chi), a natural cationic biopolymer.
As an example of the performance of metal-based coagu-
lants, Figure 6 shows the minimum specific ALG-doses
required to comply with a residual aluminium standard of
0.1 mg Al l
−1
(< 0.1 res. Al). Also shown in Figure 6 are
the doses required to obtain 70–90% colour removal and
50–60% removal of organic carbon. Dose requirements
are related to the specific UV-absorption (SUVA) of the
tested raw waters.
It can be concluded from Figure 6 that residual alu-
minium controls the minimum dose requirement for raw
waters with SUVAs of 4.3 or higher unless the required
colour or organic carbon removal exceeds 90% or 60%,
respectively. As raw water SUVA decreases below 4.3,
indicating a shift towards lower molecular weight NOM,
colour and organic carbon in particular become more
important relative to residual aluminium with respect to
Figure 5 |Average settling rates (30 grains) and sieve size analysis for grains of Filtralite NC and HC, anthracite and sand.
326 B. Eikebrokk and T.Saltnes |NOM removal using chitosan and expanded clay aggregates Journal of Water Supply: Research and Technology—AQUA |51.6 |2002
controlling the coagulant dose requirements. Even in that
case, removal requirements of more than about 65% and
50% with respect to colour and NPOC are needed
before NOM removal (i.e. colour or NPOC) controls the
minimum coagulant dose requirement.
The experiments were run in a pilot-scale coagulation
filtration plant shown in Figure 7, described in more detail
elsewhere (Østerhus & Eikebrokk 1994; Eikebrokk 1996).
The identical and parallel filter columns used in this
study, i.e. the anthracite-sand reference filter (F1) and the
Filtralite filter (F2) are shown in Figure 8. The Filtralite
NC and HC used in F2 had an effective size (d
10
)of1.65
and 0.84 mm, and uniformity coefficient (d
60
/d
10
)of1.29
and 1.54, respectively. Dried and sieved samples taken
from the middle sections of the anthracite and sand layers
in the dual media filter (F1), showed effective sizes (d
10
)of
0.82 and 0.50 mm, and uniformity coefficients (d
60
/d
10
)of
1.44 and 1.32, respectively.
By splitting the flow after coagulation and pH control,
possible differences in inlet water quality to the filters is
minimised. A direct comparison of the performance of the
two filters is then possible. The data from the flow meter,
on-line pH and turbidimeters (Hach Low Range 1720)
were continuously presented graphically on a PC-screen
and stored electronically.
The water sampling points (as shown in Figure 7)
were: (1) raw water (RW), (2) coagulated water (CW), and
(3) filter effluent water (FW). The raw water and filter
effluent water were characterised by:
–pH
– turbidity
– true colour
Figure 6 |Specific dose requirements as function of raw water SUVA when using ALG for coagulation of the tested raw waters (Eikebrokk and Saltnes 2000).
Figure 7 |Scheme of the pilot plant used in the coagulation-filtration experiments.
Figure 8 |Scheme of the parallel anthracite-sand reference (F1) and Filtralite (F2)
filters.
327 B. Eikebrokk and T.Saltnes |NOM removal using chitosan and expanded clay aggregates Journal of Water Supply: Research and Technology—AQUA |51.6 |2002
– total organic carbon as non-purgeable carbon,
NPOC (Dohrman DC-190)
– metal coagulant residues (according to the applied
coagulant).
Coagulated water samples were analysed with respect to:
– suspended solids (SS)
– zeta potential (Coulter Delsa 440 SX).
Water samples were stored at 4°C before analysis accord-
ing to the Norwegian Standard for Water and Wastewater
analysis. In addition, samples of raw water and filter
effluent water were analysed by Aquateam a.s with respect
to molecular weight fractionation (HPLC Sigmachrome
GFC-100) and biodegradable dissolved organic carbon
(BDOC) according to procedures described by Charnock
&Kjo¨ nno¨(2000).
Evaluation criteria
Coagulation and filtration performance was evaluated on
the basis of filter effluent quality and headloss data. From
the linear relationship between headloss and time, the
slope of the line is kC
0
v(McEwen 1998):
H
t
=H
0
+kC
0
vt (1)
where:
H
t
= headloss after filtration time t
H
0
= initial head loss at time 0 (‘clean’ bed)
k = constant
C
0
= concentration of particles in filter influent
water
Figure 9 |Obtained colour and organic carbon removal efficiencies when using chitosan
for coagulation of the raw waters RW15 and RW50. Water temperature:
6–11°C (Eikebrokk and Saltnes 2000).
Figure 10 |A typical example of process performance with chitosan coagulation and
filtration in the parallel anthracite-sand (A-S) and Filtralite (FMM) filters
(RW50, 7.5 mg Chi l−1, 10°C, 10 m h−1, pH 5.7–5.9).
328 B. Eikebrokk and T.Saltnes |NOM removal using chitosan and expanded clay aggregates Journal of Water Supply: Research and Technology—AQUA |51.6 |2002
v= rate of filtration (m h
−1
)
t= time of filtration (h)
In a particular case of filtration with constant conditions
regarding inlet water quality, type and dose of coagulant,
rate of filtration, etc., the equation can be simplified:
H
t
=H
0
+Kt(2)
where K is a ‘constant’ depending on the type of particles
to be removed and properties of the filter bed. In the
experiments presented here comparing two filter beds that
are tested in parallel and receiving the same water, differ-
ent K-values illustrate differences in filter bed properties
and different distribution patterns of deposited particles
within the beds.
Table 1 |Process performance and filtration characteristics when using chitosan or ALG in anthracite-sand (F1) or Filtralite (F2) filter beds. (Raw water 50 (RW50) indicates a raw water
colour level of 50 mg Pt l−1)
Run
no
Raw
water
Coag
type
Dose
(mg l−1)
Filtration
rate (m h−1)
C0
(gSS m−3)
Anthracite-sand (F1) Filtralite (F2)
H0
(mH2O)
K
(mH2Oh
−1)
k= K/vC0
(cmH2O/
(gSS m−2))
H0
(mH2O)
K
(mH2Oh
−1)
k= K/vC0
(cmH2O/
(gSS m−2))
201 50 Chi 7.5 10 9 0.45 0.232 0.26 0.08 0.056 0.06
205 50 Chi 5 10 6 0.34 0.144 0.23 0.09 0.031 0.05
206 50 Chi 5 7.5 7 0.30 0.114 0.22 0.06 0.028 0.05
Avg Chi 5.8 9.2 7 0.36 0.163 0.24 0.08 0.038 0.05
204 50 ALG 3.1 7.5 15 0.38 0.070 0.06 0.06 0.019 0.02
208 50 ALG 3.1 12.5 16 0.59 0.160 0.08 0.13 0.041 0.02
211 50 ALG 3.1 7.5 16 0.30 0.061 0.05 0.07 0.019 0.02
Avg ALG 3.1 10.0 16 0.42 0.097 0.06 0.09 0.026 0.02
Table 2 |Average effluent water quality data from filters F1 and F2 with ALG and Chi as coagulants (RW 50, three runs with every coagulant as shown in Table 1)
Coag-
Filter
Filtr. rate
(m h−1)pH
Turb.
(NTU)
Colour
(mgPt l−1)
Org. carbon
(mgNPOC l−1)
Removed fractions
H0
(mH2Oh
−1)
K
(mH2Oh
−1)
k
(cmH2O/
gSS m−2)Colour
Org.
carbon
ALG-F1 10 6.5 0.42 3 1.7 0.94 0.65 0.42 0.097 0.06
ALG-F2 10 6.5 0.93 3 1.9 0.95 0.62 0.09 0.026 0.02
Chi-F1 9.2 6.5 0.09 9 3.1 0.80 0.36 0.36 0.163 0.24
Chi-F2 9.2 6.6 0.10 9 3.2 0.79 0.34 0.08 0.038 0.05
329 B. Eikebrokk and T.Saltnes |NOM removal using chitosan and expanded clay aggregates Journal of Water Supply: Research and Technology—AQUA |51.6 |2002
RESULTS OF EXPERIMENTS WITH CHITOSAN AND
FILTRALITE FILTERS
The relationship between chitosan dose and removal effi-
ciencies with respect to colour and TOC is shown in
Figure 9 (Eikebrokk 2000). It was demonstrated before
that Filtralite may serve as an efficient substitute for
anthracite in dual media filters when ALG, JKL and chi-
tosan are used for coagulation (Eikebrokk & Saltnes 2000;
Eikebrokk 2000).
A typical example of the performance of chitosan
coagulation and filtration in anthracite-sand and Filtralite
filters is presented in Figure 10. Data on process perform-
ance and filter effluent quality obtained in the experimen-
tal runs with chitosan and Filtralite are summarised in
Table 1. Table 2 presents the average values obtained with
ALG and chitosan in filters F1 and F2. Removal efficien-
cies with respect to colour and TOC are in the range of
79–80% and 34–36%, respectively. Corresponding effi-
ciencies with ALG are 94–95%, and 62–65%. The solids
production with Chi is in the range of 6–9 mg SS l
−1
with
15–16 mg l
−1
with ALG at the given dose levels of
5–7.5 mg Chi l
−1
and 3.1 mg Al l
−1
. The benefits of the
Filtralite filter (F2) are obvious in terms of lower initial
headloss and rate of headloss increase compared with
anthracite-sand (F1). This is true regardless of the coagu-
lant used. The typical range of optimum pH values is
5.8–6.6 with ALG and 5.0–6.5 with Chi, with typical
optimum zeta potentials of − 10 to + 20 mV. The increase
in zeta potential with increasing coagulant dosage seems
more consistent with chitosan. The filter run lengths
obtained until the occurrence of breakthrough or headloss
termination are considerably longer with chitosan than
with ALG.
Although chitosan is able to meet the TOC and colour
removal requirements in most cases, ALG turns out to be
more efficient in this respect, especially with regard to
TOC removal. When ALG is used for coagulation, the
results show that effluent turbidity is higher from filter F2
than F1. This is probably due to the increased minimum
Figure 11 |MW distribution in untreated raw waters RW15 and RW50, and in filter effluent water from anthracite-sand (F1) and
Filtralite (F2). Coagulation of RW15 (upper) and RW50 (lower) with 1.0 and 2.9 mg Al l−1 as ALG (left), and 1.6 and 7.3 mg l−1
as Chi (right).
330 B. Eikebrokk and T.Saltnes |NOM removal using chitosan and expanded clay aggregates Journal of Water Supply: Research and Technology—AQUA |51.6 |2002
grain size used in F2 (0.8 mm versus 0.4 mm in F1). Thus,
the relatively loose metal–NOM floc particles have a ten-
dency to penetrate the filter F2. However, when using
chitosan, effluent turbidities are low in both filters.
Figure 11 shows the distribution of TOC present in the
different molecular weight fractions in raw waters and
treated water samples, demonstrating how the different
fractions of NOM are removed in the process. It can be
seen that a large part of the total TOC is in the 5,000–
20,000 molecular weight fraction. There is a considerable
difference between the results from the two filters,
especially when coagulating the most concentrated raw
water (RW50) with ALG. There is a tendency for ALG to
be able to remove more of the lower MW fractions than
Chi, probably due to adsorption to aluminium hydroxide
products.
From the data presented in Figure 12 it is evident that
although chitosan does not reduce TOC by more than
about 35%, the reduction in BDOC is close to 50%. This is
also indicated by the decrease in the BDOC to TOC ratio
as a result of water treatment. Hence, both ALG and Chi
are capable of removing the BDOC to a larger extent than
the DOC. It should be noted here that chitosan adds TOC,
and probably also BDOC, to the water.
CONCLUSIONS
The following conclusions can be drawn from this study:
1. Chitosan is able to remove colour and organic
carbon from NOM-containing raw waters. At doses
Figure 12 |Distribution of TOC in different molecular weight fractions in untreated and ALG or Chi coagulated waters (left).
Total TOC, BDOC and TOC/BDOC ratios for untreated and coagulated waters (right).Anthracite-sand filter (F1) only.
331 B. Eikebrokk and T.Saltnes |NOM removal using chitosan and expanded clay aggregates Journal of Water Supply: Research and Technology—AQUA |51.6 |2002
up to 7.5 mg l
−1
, about 80% of the colour is
removed. However, the removal of organic carbon is
only 35–40%.
2. Sludge solids production with chitosan is less than
50% compared with alum.
3. Lightweight expanded clay aggregates (Filtralite) can
be used successfully as a substitute for anthracite in
dual media filters. Filtralite in two size and density
fractions (F-NC and F-HC) can also be used alone
as the only filter material. The size fractions of
1.6–2.5 mm F-NC and 0.8–1.6 F-HC used in this
study performed well although the grains of F-NC
increased in density and settling rate due to water
uptake during the first 2–3 months.
4. Because the minimum grain size used in the single
medium two-layer Filtralite filter was 0.8 mm in
order to avoid mixing of the layers, the effluent
turbidity was higher for this filter than from the
reference anthracite-sand filter when alum was used
for coagulation. However, with respect to colour and
TOC removal, no major differences were observed
between the Filtralite and the anthracite-sand
reference filters when ALG or chitosan were used as
coagulants.
5. When chitosan was used for coagulation, a similar
effluent quality was obtained from the Filtralite and
the anthracite-sand reference filter. However,
Filtralite was advantageous in terms of initial
headloss and rate of headloss increase.
6. Both chitosan and alum were able to remove
biodegradable organic carbon more efficiently than
dissolved organic carbon. The dominating molecular
weight fractions in the raw waters used were in the
range of 5,000–20,000.
7. Although the best TOC removal with ALG and
chitosan was achieved in the molecular weight range
of 5,000–20,000, TOC was removed to some extent
over a wide range of MW.
REFERENCES
Charnock, C. & Kjo¨ nno¨, O. 2000 Assimilable organic carbon and
biodegradable dissolved organic carbon in Norwegian raw and
drinking waters. Wat. Res. 34(10),2629–2642.
Eikebrokk, B. 1996 Removal of humic substances by coagulation. In
Chemical Water and Wastewater Treatment V (ed. H. H. Hahn,
E. Hoffmann & H. Ødegaard), pp. 173–187. Springer-Verlag,
Berlin.
Eikebrokk, B. 1999 Coagulation-direct filtration of soft, low
alkalinity humic waters. Wat. Sci. Technol. 40(9),55–62.
Eikebrokk, B. 2000 Pilot scale testing of Filtralite as an alternative
to anthracite in dual media filters treating coagulated humic
waters. SINTEF report STF22 F00311, Trondheim, Norway.
Eikebrokk, B. & Saltnes, T. 2000 Removal of natural organic matter
(NOM) using different coagulants and lightweight expanded
clay aggregate filters. Proceedings IV International Conference:
Water Supply and Water Quality, 11–13 September, Krakow,
Poland.
Ives, K. J. 1979 The basis for the application of multiple layer filters
to water treatment. Z. Wass. Abwass. Forsch. 12(3/4),
107–110.
McEwen, J. B. (ed.) 1998 Treatment Process Selection for Particle
Removal. AWWARF/IWSA-report, Denver, Colorado.
Ødegaard, H., Eikebrokk, B. & Storhaug, R. 1999 Processes for the
removal of humic substances from water—an overview based on
Norwegian experiences. Wat. Sci. Technol. 40(9),37–46.
Østerhus, S. W. & Eikebrokk, B. 1994 Coagulation and corrosion
control for soft and coloured drinking water. In Chemical
Water and Wastewater Treatment III (ed. R. Klute & H. H.
Hahn) pp. 137–153. Springer-Verlag, Berlin.
Smidsrød, O. & Moe, S. 1995 Biopolymerkjemi. Tapir, Trondheim,
Norway (in Norwegian).
USEPA 1998 Disinfectants and Disinfectants Byproducts. Final
Rule. Fed. Reg. 63:241:69478, USA.
First received 13 August 2001; accepted in revised form 29 November 2001
332 B. Eikebrokk and T.Saltnes |NOM removal using chitosan and expanded clay aggregates Journal of Water Supply: Research and Technology—AQUA |51.6 |2002