Water and Environment Journal.

Print ISSN 1747-6585

Implications of the corrosion index for the quality of flowing tap

water and the effects of added alkalinity on corrosion control
Yeong-Kwan Kim
Department of Environmental Engineering, Kangwon National University, Chuncheon, Gangwon-do, Korea

Keywords Abstract
alkalinity; carbon dioxide; CCPP; corrosion;
Langelier saturation index; lime; Ryznar index. Three corrosion indices, Langelier saturation index (LSI), Ryznar index (RI) and cal-
cium carbonate precipitation potential (CCPP), were determined to find whether
Correspondence the corrosive index of flowing tap water could indicate the quality of the water that
Yeong-Kwan Kim, Department of Environmental corresponded to the calculated index value. Water samples were collected from
Engineering, Kangwon National University,
tap water distribution pipe in buildings of Kangwon National University, Korea.
Chuncheon, Gangwon-do, 200-701, Korea.
Correlations among the LSI, RI and CCPP were also investigated. The effects of
Email: yeong@kangwon.ac.kr
alkalinity addition using lime and carbon dioxide on the progress of corrosion were
doi:10.1111/wej.12260 examined in a laboratory-scale simulated water distribution system (SWDS) for 5
months. In the SWDS study, corrosion rate in flowing tap waters was retarded by
12% with the alkalinity addition. The results of the current study confirmed that the
corrosion indices are not always the best indicators to predict the quality of the
flowing tap water.

Introduction of water. In water distribution systems, a drop in pH level

greater than 0.5 units can disrupt the effective passivation of
The primary concerns of water suppliers and customers are
corrosion surfaces. Therefore, it is important to maintain a
reliable distribution systems and the quality of tap water.
constant pH level in carrying out the corrosion study (Taghi-
The quality of the water in the distribution system changes
pour et al. 2012). However, the adjustment of pH alone is
because of unwanted physical, chemical and biochemical
often insufficient in water with low alkalinity because the lat-
reactions that accumulate over time (Edwards 2004). Corro-
ter limits the deposition of calcium carbonate and provides
sion in water distribution systems contravenes regulations little resistance to reductions in pH levels. The relation of low
and affects health, aesthetics and customers’ perceptions. It bicarbonate levels to the chloride and sulphate levels also
also causes premature pipe deterioration and has negative influences the corrosive tendency of water. Alternatives to
effects on both the economy and the environment (Roberge supplementation with alkalinity include lime with carbon
2008). The scale in corroded pipes is a principal cause of the dioxide, soda ash and sodium bicarbonate, which form a
deterioration of the quality of tap water, and corrosion is protective layer by providing adequate buffering intensity
often indicated by the occurrence of rusty water. Corrosion and consequently decreasing corrosiveness (Cheong et al.
can vary greatly within a single system because of the com- 2011). Carbon dioxide may be added in the form of water
plex electrochemical reactions occurring on the surfaces of saturated with CO2 or CO2 gas with little difference in the
the pipes. The amount of iron in rusty water is influenced by effect of each (Sohn et al. 2008). Corrosion control using
several factors, such as the microstructure and composition lime and carbon dioxide was shown to slow corrosion rates.
of the scale caused by the corrosion (Sarin et al. 2004). The rate in galvanised iron was higher than in carbon steel,
The physical factors affecting corrosion include water copper and stainless steel (Lee et al. 2008). However, in
flow and temperature. Corrosion tends to increase with high water distribution networks, because lime is normally added
temperatures and low-flow conditions. Chemical properties, after the filtration process, the amount of lime should be
such as pH, alkalinity, dissolved oxygen (DO) and total dis- minimal in order to reduce turbidity at the tap (Kim & Lee
solved solids (TDS), contribute significantly to the corrosive 2014).
tendency of water. According to Agatemor & Okolo (2008), Internal corrosion of the pipe wall material is frequently
low pH and moderate levels of DO, as well as low bicarbon- cited as contributing to deterioration of water quality and
ate levels, contribute significantly to the corrosive tendency loss of chlorine residual. However, some previous works

Water and Environment Journal (2017) V

C 2017 CIWEM. 1
Implications of the corrosion index Yeong-Kwan Kim

have demonstrated the effect that the corroded pipe could examined in a laboratory-scale simulated water distribution
have on the maintenance of residual disinfectants in distribu- system (SWDS).
tion networks. The loss of disinfectant residual by reactions
on the scales produced on the inner pipe surface within dis- Materials and methods
tribution systems, commonly referred to as wall demand,
has been attributed to a number of factors such as pipe age Sampling and analysis
and material, pipe diameter, temperature and pipe rough-
This study was carried out at the campus of Kangwon
ness (Digiano & Zhang 2005; Clark et al. 2010). Pipe wall
National University in Korea. The tap water on the campus
demand for disinfectants is dominated by reactions on the
has been supplied from a nearby conventional water treat-
scales produced on the inner pipe surface. According to Dig-
ment plant that used coagulation, sedimentation, sand filtra-
iano & Zhang (2005), a zero-order overall kinetic model was
tion and chlorination. Water samples were collected from
well suited for describing the overall chlorine decay in a
each tap water distribution pipe located in 10 buildings con-
heavily tuberculated cast iron pipe.
structed from 1980 to 2007. In some distribution pipes that
Over a 1-year period, Volk et al. (1999) found that corro-
had been in place less than 10 years, corrosion problems,
sion rates were strongly related to seasons and water tem-
such as rusty water, were observed in taps that had not
perature. They suggested varying the concentrations of
been continuously used prior to this study. The preliminary
corrosion inhibitor according to the season. With respect to
study of the quality of tap on the Kangwon National Univer-
the characteristics water flow, Nawrocki et al. (2010)
sity campus showed a pH range that was not classified as
revealed that a layer of steady water surrounded corrosion
corrosive. Alkalinity levels were approximately 16 mg/L as
scale and could substantially influence the rate of corrosion.
CaCO3. Ten taps in several campus buildings were selected
They proposed that the occluded environment should be
based on the year that the pipe was installed and the mate-
taken into account in the formation of scale caused by corro-
rial of the pipe (copper, SUS and galvanised iron). Using
sion. In addition to physical factors, iron bacteria and
clean plastic bottles, 1 L of water was collected from each
sulphate-reducing bacteria can speed both the corrosion
tap and then stored in a refrigerator at a temperature below
and the formation of the by-products of corrosion. The
48C until the analyses. The samples were analysed for cati-
increased tendency to corrosion may also increase the
ons using ICP/OES (Perkin Elmer 4300DV). The amounts of
release of iron in the water (Agatemor & Okolo 2008). Corro-
sulphate and chloride were determined according to the
sion products that are attached to pipe surfaces or that accu-
Standard Methods (1998). The temperature, pH and TDS
mulated as sediments in the distribution system can shield
were measured in situ using a potable conductivity meter
microorganisms from disinfectants (Geldreich & Lechevellier
(VWR, USA). The alkalinity was determined by titration with a
1999). Although it is not yet clear whether corrosiveness is
0.02 N H2SO4 solution.
affected by a specific disinfection method, disinfection with
UV/Cl2 was shown to decrease iron release by rapid passiva-
Simulated water distribution system
tion compared to Cl2 alone (Zhu et al. 2014). An essential
property of water is that it dissolves most of inorganic sub- A laboratory-scale SWDS similar to that in Park & Kim (2008)
stances. Hence water contains a variety of impurities that was used to examine the short-term effects on corrosion
may lead to the formation of deposits in water lines. When rate of adding lime and carbon dioxide to the flowing tap
water is undersaturated, it can dissolve mineral deposits, water, as shown in Fig. 1.
such as calcium carbonate. Supersaturated water will precip- The system consisted of two sets of identical pipes used
itate calcium carbonate from water if it is allowed to rest. to distribute drinking water. The distribution pipe was 9.5 m
However, the assumption that water below saturation with long, with an internal diameter of 3 cm. In each set, 10
respect to calcium carbonate is corrosive is not always reli- removal test plugs equipped with SUS coupons were
able (Roberge 2008). In determining the saturation level of installed and sacrificed sequentially to compare the progres-
water in calcium carbonate, the Langelier Saturation Index sion of corrosion between the two sets. Prior to the supply
(LSI), Ryznar Index (RI) and calcium carbonate precipitation of water to the distribution system, a 200-L capacity plastic
potential (CCPP) are widely used indicators of the potential container was used to receive the tap water and to pump it
of water to cause scale and corrosion. into the SWDS. The pumping rate was controlled to maintain
The purpose of this work was to determine whether the the velocity of water within the pipe at 0.4–0.5m/s. One set
corrosive index of flowing tap water could indicate the qual- of the system (Set #1) received tap water alone, while the
ity of the water that corresponded to the calculated index other (Set #2) simultaneously received the additional water
value. Correlations among the LSI, RI and CCPP were also supplemented with lime and carbon dioxide. Distilled water
investigated. In addition, the effects of adding lime and car- saturated with CO2 was added to the lime solution, which
bon dioxide on the progress of corrosion were also had been previously prepared in a 20-L bucket with 1.5 mL

2 Water and Environment Journal (2017) V

C 2017 CIWEM.
Yeong-Kwan Kim Implications of the corrosion index

Fig. 1. Schematic diagram of the

simulated water distribution system
(SWDS). Unit: cm.

of a 20% commercial liquid lime (by wt.). The lime concentra- increased corrosion tendency, were nearly negligible at less
tion was 15 mg/L, and the final pH of the supplemented than 0.1 mg/L in all samples. The concentration of copper
water was between 7.0 and 7.5. was also low, but it was higher in the samples taken from
Two coupons were removed to measure the rate of corro- copper pipes (samples #2 and #5). However, it was noted
sion at 2, 3, 4 and 5 months following the start of the opera- that, the old copper pipes (sample #2) released more copper
tion. Prior to the measurement, the detached test coupons ions than the newer pipes did (samples #5 and #6).
were dried overnight in a drying oven. The rate of corrosion According to Ferguson et al. (1996), the corrosion of
of the test coupon was measured by comparing the differ- metallic tubing is often strongly related to the concentra-
ence between the initial weight and the final weight as tions of chloride and sulphate ions relative to bicarbonate
follows: ion. In all the samples analysed in this study, the ratios of
alkalinity relative to chloride and sulphate ions were less


Rate mg=dm2 2day 5 than 5.0, which suggested that CaCO3 film had not been pro-
Initial weightðmgÞ – final weightðmgÞ duced within the distribution pipes (Merrill 1978). As indices

of the corrosion tendency of water, the LSI, RI and CCPP
Coupon surface area dm2 3 time of operationðdayÞ
were calculated. The results are presented in Table 2. All
(1)
indices indicated that the water had a slightly corrosive
tendency.

Results and discussion

LSI and RI
Corrosion indices
The LSI is based on the effect of pH on the equilibrium solu-
As shown in Table 1, the analytical results showed that the bility of CaCO3. LSI simply indicates the driving force of scale
major influential environmental factors, including pH, alkalin- formation and growth in terms of pH. The pH at which water
ity and calcium concentration, did not significantly differ in is saturated with CaCO3 is saturation pH or pHs. The LSI is
all 10 samples. The pH values were about 7, indicating that calculated as follows:
the water was not corrosive. Neither aluminium nor manga-
nese was detected in all 10 samples. A high level of TDS level LSI 5 pH – pHs (2)
is usually associated with a high concentration of ions, which
increases the conductivity of water. The conductivity has a When the LSI value is negative, the water has a corrosive
direct relationship with the water’s ability to conduct a cor- tendency. The levels of iron concentration in the samples in
rosive current (Schock 1999). However, there was little dif- this study were low. The regression of LSI on the concentra-
ference among the samples regarding TDS concentration. tion of iron did not indicate any correlation, as shown in
Iron levels, which are known to be directly related to Fig. 2.

Water and Environment Journal (2017) V

C 2017 CIWEM. 3
Implications of the corrosion index Yeong-Kwan Kim

Table 1 Characteristics of the water sample collected from each distribution pipe in 10 buildings
Samples 1 2 3 4 5 6 7 8 9 10
pH 7.00 6.91 7.05 7.05 7.06 7.01 7.01 6.99 7.13 7.09
Temp (8C) 21.6 21.9 21.7 21.5 21.5 21.7 21.5 21.4 21.5 21.6
Ca 7.23 7.15 7.39 7.16 7.24 7.19 7.15 7.09 7.23 7.24
Cu 0 0 0 0.185 0.14 0.65 0 0.01 – 0.182
Fe 0.132 0.002 0 0 0 0.012 0 0.353 0 0.025
Alkalinity 15.4 16.0 15.6 15.3 17.2 15.3 15.4 15.2 15.5 14.6
SO224 4.2 4.5 4.5 3.9 4.8 5.0 4.4 4.7 4.7 5.1
Cl2 6.2 6.8 6.4 6.3 7.0 7.2 6.8 6.5 6.0 7.4
TDS 55.5 55.3 54.8 55.4 55.5 55.4 56.3 55.3 54.6 55.5
I (3103) 1.036 1.031 1.023 1.034 1.036 1.034 1.051 1.033 1.019 1.014

Alkalinity: mg/L as CaCO3.

I: Ionic strength.
mg/L otherwise.

The RI is empirical and applies only to flowing systems however, was an important factor affecting corrosiveness
where the environment at the pipe wall is quite different (Atasoy & Yesilnacar 2010). Calculated LSI and RI values are
from that of the bulk water. The RI can indicate the dissolu- listed in Table 2.
tion of metals into the water during distribution (Agatemor &
Okolo 2008). The RI is calculated as follows:
CCPP
RI 5 2pHs – pH (3) While the LSI indicates the tendency either to precipitate or
to dissolve CaCO3, CCPP indicates the tendency and quantity
An RI value greater than 7.0 is interpreted as corrosive of CaCO3 in dissolution or precipitation. However, CCPP
(NALCO 1988). The high values of RI founded in this study
does not indicate the ability of the CaCO3 film to provide pro-
supported the LSI values, indicating the corrosive tendency
tection against corrosion (McNeil & Edwards 2001). Although
in the water samples. However, low iron concentrations
CCPP should not be the sole criteria of corrosion control, it is
suggested that the occurrence of the iron was continuously
a valuable index used in conjunction with other corrosion
flushed out during the test. Consequently, similar to the LSI,
indices. CCPP values from 25 to 210 mg/L as CaCO3 are
no relation between RI and iron concentration was
considered mildly corrosive. Those less than 210 mg/L are
observed, as shown in Fig. 3.
considered corrosive or aggressive (Gebbi 2000).
In water with medium-range alkalinity, the chlorides and
One method of determining the CCPP value is a graphical
sulphates content can interfere with natural formation of film
procedure using the Caldwell and Lawrence (C–L) diagram.
(Roberge 2008). As shown in Table 1, the alkalinity levels and
The C–L diagram for water at 258C and a TDS of 40 mg/L was
concentrations of chlorides and sulphates in the tap water
used in this study. In determining the CCPP values, the acid-
samples in this study were low, which suggested that
ity of the water was calculated using the following equation,
chlorides and sulphates were not significant contributors to
corrosion. High sulphate concentration in groundwater,

Table 2 Three corrosion indices determined for each sample

Samples LSI RI CCPP Pipe material Year of installation
1 22.57 12.13 29.9 Cast iron 1981
2 22.62 12.15 211.1 Copper 1981
3 22.49 12.02 28.5 SUS 1980
4 22.52 12.10 28.1 Cast iron 1989
5 22.45 11.96 29.9 Copper 1994
6 22.54 12.09 29.5 Copper 1994
7 22.46 12.11 29.1 Copper 1996
8 22.56 12.11 29.3 Cast iron 1982
9 22.42 11.97 28.4 SUS 2007
10 22.49 12.06 28.9 Cast iron 1989
Fig. 2. Plot of LSI on Fe concentration. [Colour figure can be viewed at
CCPP: mg/L as CaCO3. wileyonlinelibrary.com]

4 Water and Environment Journal (2017) V

C 2017 CIWEM.
Yeong-Kwan Kim Implications of the corrosion index

Fig. 3. Plot of RI on Fe concentration. [Colour figure can be viewed at

Fig. 5. Correlation between LSI and CCPP. [Colour figure can be
wileyonlinelibrary.com]
viewed at wileyonlinelibrary.com]

which is valid for water at 258C with a TDS of up to 200 mg/L

(Gebbi 2000): pipe material might influence the degree of corrosion to
some extent on the release of metals. The concentrations of


Acidity5Alkalinity 114:2453106 3102pH (4) iron and copper (Table 1) were found low enough to exclude
the possible effect of the pipe material. For example, iron
The corrosive tendency determined by the CCPP values levels in the samples taken from cast iron pipe were negligi-
indicated that the all the water samples were mildly corro- ble, except #8 at 0.353 mg/L. Copper levels in the samples
sive, except sample #2. Calculated CCPP values are listed in taken from copper pipe were also negligible. The only
Table 2. noticeable copper concentration was observed in sample
#6, which had been taken from a pipe installed in 1994. Even
Correlations among the indices though all three corrosion indices indicated corrosiveness in
the water samples, the magnitude of the indices did not
Correlations between LSI versus RI (Fig. 4) and LSI versus
reflect the level of metal concentrations in water. Therefore,
CCPP (Fig. 5) were determined yielding coefficients of deter-
despite the corrosiveness determined by the indices, the
mination (R2 value) at 63.6 and 31.55%, respectively (Minitab
actual level of metals present in flowing waters should not
ver. 16). As shown, the RI appeared to be more consistent
be determined by the corrosion index only. In this study, the
with the LSI than the CCPP in estimating the aggressiveness
age of the pipe installation appeared to be of greater influ-
of corrosion.
ence on metal release. The indices derived from samples 1,
Effects of pipe age and material 2 and 8 indicated the effects of pipe aging on corrosive index
values. The three distribution pipes had been installed in the
In addition to water characteristics, the age of the pipe might early 1980s.
be responsible for greater tendencies to corrosion, and the

Effect of lime and carbon dioxide addition on

corrosion control
The tap water samples in this study exhibited low calcium
content, and lime and carbon dioxide was supplemented to
promote a protective (oversaturated) condition in the SWDS.
The SWDS was operated over a period of 5 months from Feb-
ruary to June. The average water temperature was 8.18C in
February, and it increased to as high as 19.88C in June. Tap
water pH, alkalinity and TDS were maintained, in a range of
7.0 to 7.5, 15.1 to 16.8 mg/L as CaCO3 and 45.3 to 61.5 mg/L,
respectively. In the alkalinity-supplemented water, the pH
was between 7.2 and 7.6, and TDS increased by about 2%.
Fig. 4. Correlation between LSI and RI. [Colour figure can be viewed at The average calcium concentrations were 7.3mg/L and
wileyonlinelibrary.com] 9.5 mg/L, respectively, in Set #1 and Set #2.

Water and Environment Journal (2017) V

C 2017 CIWEM. 5
Implications of the corrosion index Yeong-Kwan Kim

some insights into corrosion tendencies, but correlations

among the indices could not be identified. The RI appeared
to be more consistent with the LSI (R2 value of 63.6%) than
with the CCPP (R2 value of 31.5%) in estimating the aggres-
siveness of the water in the samples. Over the experimental
period of 5 months, the corrosion rate of the tested coupon
was retarded by 12% with lime and carbon dioxide addition.
The individual effects of the addition of lime or carbon diox-
ide on the retardation were beyond the scope of this study.
The results of the current study confirmed that corrosion
indices are not always the best indicators of the quality of
flowing tap water.

Acknowledgements
Fig. 6. Experimental results of the simulated water distribution system.
[Colour figure can be viewed at wileyonlinelibrary.com]
The author gratefully acknowledges the financial support
from Kangwon National University, Korea in completing this
study.
Figure 6 shows the results of the simulated distribution
system operation. With the addition of lime and carbon diox- To submit a comment on this article please go to http://mc.
ide, the corrosion rate was retarded by 12% on average over manuscriptcentral.com/wej. For further information please see the
the 5-month experimental period. The rate increased stead- Author Guidelines at wileyonlinelibrary.com
ily toward the end of the operation, which might have been
partly due to the increased temperature in June (McNeil &
Edwards 2002).
References
For corrosion control in water distribution pipes, Germany Agatemor, C. and Okolo, P. (2008) Studies of Corrosion Tendency
and Netherland established the level of LSI as a control of Drinking Water in the Distribution System at the University
standard along with other parameters such as alkalinity and of Benin. Environmentalist, 28, 379–384.
selected anions (AWWARF 1996). National Environmental APHA, AWWA and WEF. (1998) Standard Methods for the
Research Institute of Korea did a survey on eight major drink- Examination of Water and Wastewater (20th edn). American
Public Health Association, Washington, DC.
ing water sources and proposed to establish the LSI as a cor-
Atasoy, A.D. and Yesilnacar, M.I. (2010) Effect of High Sulphate
rosion index in Drinking Water Quality Criteria. Among the
Concentration on the Corrosivity: A Case Study from
water industries of Korea that supply drinking water to pub- Groundwater in Harrin Plain, Turkey. Environ. Monit. Assess.,
lic, K-Water is the only one that set up the level of pH over 166, 595–607.
7.0 and LI above 21.5 on yearly average basis. Results of AWWARF. (1996) Internal corrosion of water distribution
this work, however, suggested that LSI should not be used systems. AWWARF and DVGW-TZW Cooperative Research
alone as a control measure of corrosion tendency. In addi- Report, Denver, CO.
tion to the LSI, other measure such as Larson’s modified ratio Cheong, W.S., Kim, J.K., Park, D.J., Kim, S.W. and Jeong, S.G.
(LMR) which incorporates hydraulic detention time would be (2011) A Study on the Corrosion Control of Tap Water by Lime
necessary as a supplemental index. and Carbon Dioxide. J. Kor. Soc. Water Wastewater, 25 (2),
193–199.
Clark, R.M., Yang, H., Impelliteri, C.A., Haught, R.C., Schupp, D.A.,
Conclusions Pangluri, S. and Krishnan, E.R. (2010) Chlorine Fate and
This work was carried out to determine whether the corro- Transport in Distribution Systems: Experimental and Modeling
sive index of flowing tap water corresponded to the calcu- Studies. J. Am. Water Works Assoc., 102 (5), 144–155.
lated index value of its chemical characteristics. The LSI, RI Digiano, F.A. and Zhang, W. (2005) Pipe Section Reactor to
and CCPP were determined in flowing tap water samples Evaluate Chlorine-Wall Reaction. J. Am. Water Works Assoc.,
97 (1), 74–85.
from 10 buildings on the Kangwon National University cam-
Edwards, M. (2004) Contolling Corrosion in Drinking Water
pus in Korea. The actual levels of the dissolved metal in the
Distribution Systems: A Grand Challenge for the 21st Century.
samples were nearly negligible regardless of the corrosive- Water Sci. Technol., 49 (2), 1–8.
ness determined by the indices, which was possibly due to Ferguson, J.L., von Franque, O. and Schock, M.R. (1996)
the frequent use of tap water. Under similar environmental Corrosion of Copper in Potable Water Systems. In Internal
conditions, the age of the pipe appeared to be a factor influ- Corrosion of Water Distribution Systems (2nd edn), AWWA,
encing metal release. The corrosion indices might provide Denver, USA. pp. 231–268. AWWARF/DVGW-TZW.

6 Water and Environment Journal (2017) V

C 2017 CIWEM.
Yeong-Kwan Kim Implications of the corrosion index

Gebbi, P. (2000) Water stability-what does it mean and how do Park, S.K. and Kim, Y.K. (2008) Effect of Chloramines
you measure it?. Paper presented at the 63rd Annual Water Concentration on Biofilm Maintenance on Pipe Surfaces
Industry Engineers and Operators Conference, 6–7 September Exposed to Nutrient-Limited Drinking Water. Water SA, 34 (3),
2000, Warrnambool. 373–380.
Geldreich, E.E. and Lechevellier, M. (1999) Microbiological Quality Roberge, P.R. (2008) Corrosion Engineering, Principles and
Control in Distribution Systems. In AWWA (ed) Water Quality Practice. McGraw-Hill, New York.
and Treatment: A Handbook of Community Water Supplies, Sarin, P., Snoeyink, V.L., Bebee, J., Jim, K.K., Beckett, M.A.,
pp. 18–49. McGraw-Hill, London. Kriven, W.M. and Clement, J.A. (2004) Iron Release from
Kim, J.K. and Lee, J.H. (2014) Improvement of Tap Water Corroded Iron Pipes in Drinking Water Distribution Systems:
Corrosivity by Lime and Carbon Dioxide. J. Kor. Soc. Water Effect of Dissolved Oxygen. Water Res., 38, 1259–1269.
Wastewater, 28 (6), 725–733. Schock, M.R. (1999) Internal Corrosion and Deposition Control. In
Lee, D.J., Kim, Y.I., Song, Y.I. and Park, H.A. (2008) Corrosion AWWA (ed) Water Quality and Treatment: A Handbook of
Control in Water Distribution System Using Lime and Carbon Community Water Supplies, pp. 1–17. McGraw-Hill, London.
Dioxide(II)-Evaluation on the Characteristics of Corrosion as a Sohn, B.Y., Byun, K.S., Kim, Y.I. and Lee, D.J. (2008) Corrosion
Function of Pipe Material. J. Kor. Soc. Water Wastewater, 22 Control in Water Distribution System Using Lime and Carbon
(3), 379–388. Dioxide(I)-Determination of Optimum Operational Conditions
McNeil, L.S. and Edwards, M. (2001) Iron Pipe Corrosion in in Lime Adding Process. J. Kor. Soc. Water Wastewater, 22 (3),
Distribution Systems. J. Am. Water Works Assoc., 93 (7), 88–100. 373–378.
McNeil, L.S. and Edwards, M. (2002) The Importance of Taghipour, H., Shakerhatibi, M., Pourakbar, M. and Belvasi, M.
Temperature in Assessing Iron Pipe Corrosion in Water (2012) Corrosion and Scaling Potential in Drinking Water
Distribution Systems. Environ. Monit. Assess., 77 (3), 229–242. Distribution System of Tabriz, Northwestern Iran. Health
Merrill, D.T. (1978) Chemical Conditioning for Water Softening Promot. Perspect., 2 (1), 103–111.
and Corrosion Control. In Sanks, R.L. (ed) Water Treatment Volk, C., Dundore, E., Schiermann, J. and Lechevallier, M. (1999)
Plant Design. Ann Arbor Science, Ann Arbor, MI. 497–565 Practical Evaluation of Iron Corrosion Control in a Drinking
NALCO Water Handbook. (1988) Water Chemistry and Water Distribution System. Water Res., 34 (6), 1967–1974.
Interpretation of Water Analysis. McGraw-Hill, London. Zhu, Y., Wang, H., Li, X., Hu, C., Yang, M. and Qu, J. (2014)
Nawrocki, J., Raczyk-Stanislawiak, U., Swietlik, J., Olejnik, A. and Characterization of Biofilm and Corrosion of Cast Iron Pipes in
Sroka, M. (2010) Corrosion in a Distribution System: Steady Drinking Water Distribution System with UV/Cl2 Disinfection.
Water and Its Composition. Water Res., 44, 1863–1872. Water Res., 60 (1), 174–181.

Water and Environment Journal (2017) V

C 2017 CIWEM. 7