Wat. Res. Vol. 35, No. 12, pp.

2961–2969, 2001
# 2001 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
PII: S0043-1354(00)00591-1 0043-1354/01/$ - see front matter

PHYSICO-CHEMICAL CHARACTERISTICS OF CORROSION

SCALES IN OLD IRON PIPES
P. SARIN2, V. L. SNOEYINK1*, J. BEBEE1, W. M. KRIVEN2 and J. A. CLEMENT3
1
Department of Civil and Environmental Engineering, University of Illinois at Urbana-Champaign,
IL 61801, USA; 2 Department of Materials Science and Engineering, University of Illinois at
Urbana-Champaign, IL 61801, USA and 3 Black and Veatch, Cambridge, MA 02140, USA

(First received 1 July 2000; accepted in revised form 1 December 2000)

Abstract}Corrosion scales play an important role in modifying water quality in drinking water
distribution systems. The corrosion scales from old iron/steel pipes were analyzed for their structure and
composition. Scales were studied both before and after drying, and goethite, magnetite and lepidocrocite
were identified as the primary constituents of the dried samples. High concentrations of readily soluble
ferrous phases were detected in wet-scale samples. The corrosion scales had a shell-like, enveloping layer,
covering porous deposits of iron oxide phases. Our studies were able to identify important differences
between corrosion scales found in two different water distribution systems. Further studies are needed
to establish the role of corrosion scales in the mechanism of iron release from corroded pipes.
# 2001 Elsevier Science Ltd. All rights reserved

Key words}corrosion scales, corrosion, tubercles, iron oxides, enveloping layer

BACKGROUND Stumm, 1960; Obrecht and Pourbaix, 1967; Pisigan

Iron and steel pipes have been used in water and Singley, 1987). However, the interaction of
distribution systems for over five centuries. Since corroded iron (iron metal surface having corrosion
the thermodynamically stable form of iron in contact scales) with water has not been studied in depth. The
with atmospheric oxygen is ferric iron (Fe(III)), response of corroded iron to changes in water quality
exposure to aqueous conditions results in corrosion is expected to be a strong function of the physico-
of iron. Most water distribution pipelines have a chemical characteristics of the corrosion scales, and is
buildup of iron corrosion products inside the iron different from classical electrochemical corrosion of
pipes. Corrosion scales not only restrict the flow of iron.
water (Larson and Sollo, 1967), but also degrade the Several water quality parameters can influence the
water quality. These scales are porous deposits that formation of corrosion scales in iron/steel pipes in
are comprised of iron (hydr)oxide phases. Some ways water distribution systems. These include pH, alka-
in which corrosion scales can adversely affect water linity, buffer intensity, dissolved oxygen (DO), water
quality in drinking water distribution pipes are (a) as flow characteristics, temperature, water treatment
a source of iron, which when released into water can practices, application of an inhibitor, and fluctua-
result in ‘‘red water’’ or ‘‘colored water’’, (b) as a tions in water quality. Cornell and Schwertmann
source of high demand for chlorine and dissolved (1996) have categorized the formation mechanisms of
oxygen (Tuovinen et al., 1984; Geldreich, 1990; iron oxides and oxyhydroxides into (a) direct pre-
LeChevallier et al., 1993), (c) as an excellent breeding cipitation from solution, and (b) topotactic phase
ground for microbes, and (d) by adsorption and transformations from one solid phase into another by
accumulation of substances such as arsenic (Raven internal rearrangements of atoms.
et al., 1998), and radium (Field et al., 1995), which
can be released on modification of water quality. Scale structure and composition
Extensive research has been done to determine the
impact of water quality on the corrosion of new iron Corrosion scales originate at the pipe wall and
pipes (Eliassen et al., 1956; Larson and Skold, 1958; grow radially inwards towards the pipe center. The
pipe metal is the original source of iron and the scales
are porous deposits with a distinct, layered structure.
*Author to whom all correspondence should be addressed. Often a dense, shell-like layer has been observed near
Tel.: +1-217-333-4700; fax: +1-217-333-9464; e-mail: the surface of the scale that is in contact with water.
[email protected] Magnetite and goethite were reported as the
2961
2962 P. Sarin et al.

iron-containing phases comprising this dense layer Under anoxic conditions, the amount of iron
(Kölle and Rosch, 1980; Sontheimer et al., 1981). released could depend upon the solid phase in the
Compounds usually found in iron corrosion scales scale that controls solubility and solution composi-
include goethite (a-FeOOH), lepidocrocite tion such as pH and the total carbonate concentra-
(g-FeOOH), magnetite (Fe3O4), siderite (FeCO3), tion. For example, when siderite is present and the
ferrous hydroxide (Fe(OH)2), ferric hydroxide total carbonate concentration is 103 M, it will
(Fe(OH)3), ferrihydrite (5Fe2O3
9H2O), green rusts control Fe(II) solubility below pH 10.5, and ferrous
(e.g., FeII III
4 Fe2 (OH)12(CO3)) and calcium carbonate hydroxide will control solubility above pH 10.5 at
(CaCO3) (Benjamin et al., 1996). Depending on the 258C (Snoeyink and Jenkins, 1980). Results from
composition of the water next to the pipe surface, Génin et al. (1998) showed that green rust controlled
corrosion of iron may result in the formation of iron solubility in soil solutions and this phase may
relatively soluble Fe(II) solids such as siderite and also be important in corrosion scales. A conceptual
ferrous hydroxide. Inside the scale, close to the pipe model of protective scale formation, proposed by
wall, iron phases in the oxidation state Fe(II) are Kölle and Sontheimer (1977) and further explained
expected. Ferrous hydroxide is formed under redu- by Sontheimer et al. (1981), highlighted the critical
cing conditions. However, in the presence of carbonic reactions that might occur within the scales. Later,
species, siderite (FeCO3) is the stable ferrous solid Kuch (1988) proposed a mechanism of non-steady-
(Singer and Stumm, 1970). Studies by Baylis (1926), state corrosion of iron by reduction of ferric oxide
Benjamin et al. (1996) and Sontheimer et al. (1981) scales with subsequent release of Fe(II) ions to the
have documented the presence of siderite as a solution.
corrosion product. The primary objective of this study was to
Green rusts are hydrated ferrous–ferric com- characterize the physical properties and chemical
pounds having Cl or SO2 2
4 or CO3 anions (Olowe composition of corrosion scales, and to identify the
et al., 1988; Drissi et al., 1994; Génin et al., 1997; differences between scales from dissimilar water
Refait et al., 1998). These compounds are formed distribution systems. The subjects of this study were
during the partial oxidation of ferrous hydroxide in iron/steel pipes that had been in service in drinking
the presence of other anions (Refait et al., 1997). water distribution systems for several decades.
Touvinen et al. (1980) found green rust as the The information obtained should be useful for
primary crystalline form of iron in cast iron understanding reasons for the success and failure
corrosion scales in the Columbus, Ohio, distribution of colored water and scale formation control
system. The degree of iron oxidation in green rust is procedures.
variable and an Fe(II)/Fe(III) ratio of 4.0–0.8 was
observed.
MATERIALS AND METHODS

Corrosion scales from two different water distribution

Corrosion and iron release systems were analyzed. A 70-year old galvanized steel pipe,
1.5-m long, and 3.2-cm diameter, was excavated from the
Corrosion is the oxidation of metallic iron, and Northern Illinois Water Corporation’s (NIWC) water
this process releases iron into solution, or results in distribution system in Champaign, IL, USA. The pipe had
been in the distribution system since the early 1920s. The
iron scales. Corrosion may uniformly attack a metal inner surface of the pipe had a semi-uniform corrosion layer
surface (uniform corrosion) or it may be focused at with a thickness of 2–5 mm. For comparison, corrosion
specific sites. While uniform corrosion results in the scales from 90 year old, 20.3-cm diameter cast iron pipes
development of uniform scales, localized corrosion from the Metropolitan Water Resources Authority
can produce growing mounds of corrosion products (MWRA), Boston, MA were also analyzed. The cast iron
pipes from MWRA were heavily tuberculated, occluding up
(Obrecht and Pourbaix, 1967), called tubercles to 76% of the cross-section originally available for flow.
(Larson and King, 1954; Larson and Skold, 1957). Table 1 compares various elements of water quality of the
Iron release refers to the transport of iron from the MWRA and NIWC distribution systems and the pipes that
corrosion scale or the metal surface to the bulk water, were studied from these systems. The quality of water in the
two distribution systems was different. Although the water
either in soluble form or as particulate matter. Iron chemistry data for these systems is incomplete for the
release from corroded iron surfaces can be brought duration that the pipes were in use, data from recent years
about by (a) corrosion of metal (b) hydraulic (1996–1998) show that MWRA water had an alkalinity of
scouring action of flowing water and (c) dissolution 10–35 mg/L as CaCO3, and a pH of 7–9. The NIWC water,
of corrosion scale components. In water distribution in comparison, had significantly higher alkalinity (135–
150 mg/L as CaCO3) and pH close to 9.0. The Ca and Si
systems iron release is a cumulative effect of all the content of NIWC water were significantly higher than
above. Whereas corrosion rate is usually measured MWRA water.
by weight loss from the metallic iron, iron release is Experimental pipe loops were set up using six sections of
measured as the concentration of iron in the bulk a cast iron pipe from MWRA, and a galvanized pipe from
NIWC. The influent water was carefully controlled, and the
water. The rate of iron release often bears no simple effluent water quality was monitored for both short and
relationship to the rate of iron corrosion (Benjamin long-term iron release studies. Corrosion scales were
et al., 1996). sampled for analysis from these pipes (see Table 2). The
 Corrosion scales in old iron pipes 2963

Table 1. Comparison of MWRA and NIWC distribution systems

MWRA distribution system NIWC distribution system

Water quality
pH 7–8 8.9
DO 7–8.5 mg/L 1–4 mg/L
Total alkalinity 10–35 (mg/L as CaCO3) 135–150 (mg/L as CaCO3)
Ca 3.6 mg/L 11–16.5 mg/L
Si 2.2–2.3 mg/L as SiO2 7–8 mg/L as SiO2
Cl 12.4 mg/L 14 mg/L
SO2
4 (mg/L) 5.6 mg/L 1–2 mg/L

Pipe
Material Cast iron Galvanized steel
Diameter 20.32 cm (8 inch) 3.18 cm (1.25 inch)
Age in service 90 years (approx) 70 years (approx)

Table 2. Details of corrosion scale samples

Sl. No. Sample code Description of sample

1 NIWC-A As received sample of galvanized pipe from NIWC

2 NIWC-B Sample from the NIWC pipe after 1 year in pipe loop
3 Boston#1 Cast iron pipe scale sample from Boston pipe # 1
4 Boston#2 Cast iron pipe scale sample from Boston pipe # 2
5 Boston#3 Cast iron pipe scale sample from Boston pipe # 3
6 Boston#4 Cast iron pipe scale sample from Boston pipe # 4
7 Boston#5 Cast iron pipe scale sample from Boston pipe # 5
8 Boston#6 Cast iron pipe scale sample from Boston pipe # 6

NIWC-A scale sample was removed from the as-received remove the distinct shell-like layer from the MWRA scales
pipe, whereas the NIWC-B scale was taken from the same using a razor blade, and separate powder samples were
pipe after conducting short-term iron release studies for prepared for this layer. In spite of all the precautions taken
about 1 year in the pipe loop. The influence of stagnation during sampling, it was not possible to sample this dense
time, dissolved oxygen concentration, and pH on iron layer exclusively. Therefore, these powder samples include
release was observed before the NIWC-B sample was both the dense shell-like layer and the ‘‘top surface’’ layer
removed. and a suffix ‘‘-TSS’’ (top surface and shell) is used to identify
these samples.
Scale analyses Characterization of corrosion scale microstructure: Several
techniques were employed to study the physico-chemical
The scales were dried under vacuum at about 208C and characteristics of the corrosion scales.
were characterized by several experimental techniques. In (a) Physical structure: A table-top optical microscope
order to confirm if drying the scales modified any chemical (Nikon SMZ-2 T, Nikon, Inc., Japan) was used to study the
species present in the wet scales, the concentrations of Fe(II) peterographic slides and other cross-section samples. This
and Fe(III) in the scales were compared before and after analysis provides a ‘‘macro-view’’ of the corrosion-scale
drying. structure.
(b) Bulk density: The dried scale samples were first
Dry corrosion scales: Pipe sections with corrosion scales weighed and then their volume was determined. In order to
were first dried under vacuum and room temperature determine bulk density (mass/volume, including pores), the
conditions. The dried scales were then removed from the scale sample was covered with a very thin coat of paraffin
pipe wall and different types of samples were prepared. wax to prevent water from penetrating the scales. Scale
Cross-section samples: The dried scales were impregnated volume was calculated from the amount of water displaced
with an epoxy resin (Epo-Thin, Buehler Ltd., Lake Bluff, when immersed in water (Archimedes’ method). Mettler’s
IL) under vacuum, and sectioned by using a slow-action AE166 electronic weighing balance (Mettler Instrument
diamond blade. Epoxy impregnation was often repeated for Corp., Hightstown, NJ, USA), along with accessories for
these porous scales. The cross-sectioned samples were density measurement, was used for these experiments.
polished to a 1 mm surface finish on Buehler’s ECOMET (c) True density: True density, mass/volume of solid
III polisher/grinder (Buehler Ltd., Evanston, IL) using a excluding pores, of powdered samples of corrosion scales
diamond paste (Buehler METADI Aerosol Spray Diamond was determined using the AccuPyc 1330 Pycnometer
Compound 1 mm; Buehler Ltd., Lake Bluff, IL). Petro- (Micromeritics, Norcross, GA) to measure the volume of
graphic slides were also prepared for few specimens helium gas displaced by a known mass of powder sample.
(Wagner Petrographic, Provo, UT). (d) Surface area and pore size: The ASAP 2400 analyzer
Surface samples: A small part of the top surface of the (Micromeritics, Norcross, GA) was used to determine the
dried tubercle was studied using scanning electron micro- specific surface area of powder samples by N2 gas
scopy (SEM). This surface layer, referred to as the ‘‘top adsorption and desorption. The samples were degassed at
surface’’ of the scale, is present on top of the dense shell-like 1208C prior to analysis.
layer, and comes in direct contact with the flowing water. (e) Elemental composition: Powder, top surface and
Powder samples: Samples of the entire tubercle were polished cross-section samples were analyzed using SEM
powdered with an alumina mortar and pestle until they and energy dispersive spectroscopy (EDS) techniques. A
passed through a 150 mm a mesh sieve. It was possible to LaB6 filament, multipurpose, analytical scanning electron
2964 P. Sarin et al.

microscope (Zeiss DSM 960, Göttingen, Germany) and

SEM Hitachi S-530 ( Hitachi, Conroe, TX)) equipped with a
tungsten (W) filament were used for the SEM/EDS studies.
The EDS results were limited to elements with atomic
number 11 or greater.
(f ) Crystalline phase composition: A RIGAKU (D-MAX
II) X-ray powder diffractometer (Rigaku/USA Inc., Dan-
vers, MA) with a Cu Ka source (l ¼ 0:1540598 nm) and a
single crystal monochromator in the diffracted beam path,
was used for acquiring powder diffraction data. Multiple
scans were performed to improve resolution. The X-ray
diffraction (XRD) pattern was analyzed by the Rietveld
method (Rietveld, 1969) using the General Structure
Analysis System (GSAS) software from Los Alamos
National Laboratory (Von Dreele and Larson, 1997).
Crystalline-phase composition of the corrosion scales was Fig. 1. Optical microscopic image of the cross-section of the
quantitatively determined following the procedure of Hill corrosion scales from NIWC showing a dense exterior layer
and Howard (1987). and a porous interior.
(g) Fe(II)/Fe(III) content: The Fe(II)/Fe(III) ratio of the
powdered corrosion scale was measured by wet chemistry
methods. Samples were dissolved in an acidic solution
(20 mL of concentrated HCl (Fisher Scientific Certified
ACS) in 300 mL distilled and de-ionized (DDI) water.
The ferrous and total iron concentrations were deter-
mined using bathophenanthroline, following the
method proposed by Lee and Stumm (1960). A Beckman
DU 7500 Spectrophotometer (Beckman Coulter Inc., Full-
erton, CA) was used to analyze the absorbance of the
prepared samples.
Wet corrosion scales: Scales from three cast iron pipes
from MWRA being used in pipe loop studies were sampled
for these analyses. The water flow through these pipes was
stopped and small parts of the scales were scraped from top,
middle and bottom of the tubercles close to the end of the
pipes. Samples were immediately dissolved in an acidic
solution (20 mL of concentrated HCl in 300 mL DDI water),
avoiding exposure to atmospheric environment, and their
Fe(II)/Fe(III) content was analyzed. Scales from the NIWC
pipe section were sampled in a similar way at two locations.
Since the scales in NIWC pipes were much thinner in
comparison to MWRA scales, they were analyzed as an
entire scale only. Fig. 2. Optical microscopic image of the cross-section slide
of the corrosion scales from NIWC showing dense ‘‘shell-
like’’ layer, outermost layer at the water–scale interface, and
the porous interior usually present in corrosion scales found
in iron/steel drinking water distribution pipes.
RESULTS AND DISCUSSION

The physico-chemical characteristics of the cor-

rosion scales can determine the ensuing chemical observable by eye. An optical micrograph of NIWC
reactions when corroded pipe surfaces are exposed to corrosion scale is shown in Fig. 1. The dense shell-
aqueous conditions. The important features of the like layer isolated the porous interior of the scales
structure and composition of MWRA and NIWC from bulk water. The interface between the corrosion
scales are presented in the following sections. scale and water was not sharp (see Fig. 2). A
combination of Archimedes’ method and true density
measurements of powders (pycnometry) was used
Physical characteristics of corrosion scales to evaluate porosity of scales (see Table 3). The
Corrosion scales provide resistance to further measured bulk density was 2.2–2.3 and 1.7–2.2 g/cm3
corrosion of the underlying metal, and porosity is for NIWC and MWRA scales, respectively. The
one factor dictating their effectiveness. Porosity can calculated porosity values were 40–43 and 44–54%
determine the ease with which ions migrate within the for NIWC scales and MWRA scales, respectively.
scale layers (Benjamin et al., 1996). Corrosion scales are heterogeneous and variation in
The cross-section samples of MWRA and NIWC tubercle structure between two sampling sites can
scales were viewed under an optical microscope to explain the observed range in bulk density and
study the physical structure of the scales. A shell-like porosity for MWRA scales. The lower density and
layer was observed near the top of the tubercle, i.e., higher porosity of the MWRA scales in comparison
the section of the tubercle in close contact with water. to NIWC scales may have been caused by differences
The part of the corrosion scale near the pipe wall in water quality, different water flow rates, and
was very porous, and had voids as large as 1–2 mm, microorganism growth.
 Corrosion scales in old iron pipes 2965

Table 3. Density and porosity determination of the ‘‘corrosion scales’’

Sample ID True density (g/cm3) Bulk density in (g/cm3) Porosity (%)

NIWC-A 3.89 2.20 43.30

NIWC-B 3.93 2.34 40.51
Boston#1 3.72 1.75 53.02
Boston#2 3.92 1.80 54.18
Boston#3 3.78 1.73 54.33
Boston#4 3.60 2.01 44.13
Boston#5 3.59 1.99 44.43
Boston#6 3.67 2.22 39.56

Table 4. Surface area and porosity of powders from dry scale samples

Sample ID Surface area (sq. m/g) Volume in cm3/g (  103) Pore size in Å

BET SA of pores m-pore Vol. of pores m-pore Avg pore dia Avg pore dia
17–30008A area 17–30008A dia (BJHa) vol (BET) (BJHa)
dia (BJHa)

NIWC-A 154 85 28 90 12 32 42
NIWC-B 89 46 32 70 14 39 64
Boston#1-TSSb 60 54 8 140 4 93 106
Boston#1 43 47 1 160 0 130 138
Boston#2-TSS 50 46 8 110 4 91 99
Boston#2 32 36 0 120 0 142 140
Boston#3-TSS 95 73 4 160 1 63 87
Boston#3 43 45 1 180 1 125 163
Boston#4-TSS 79 64 12 190 5 89 122
Boston#4 82 70 9 240 4 97 139
Boston#5-TSS 49 41 7 140 3 98 139
Boston#5 53 57 0 240 0 160 165
Boston#6-TSS 41 33 3 130 1 114 132
Boston#6 54 49 1 170 0 117 140
a
Calculated using desorption data by Brunauer, Jura and Harkins (BJH) method.
b
Top layers of the Boston samples were analyzed separately also, and they have a suffix of ‘‘-TSS’’ (top surface and shell) at the end of their
code.

The physical characteristics of powders prepared methods. The elemental composition of the scales as
by grinding dry scales are important for comparing identified by EDS measurements (see Table 5)
scales from different systems. Results of the nitrogen showed that they were primarily comprised of iron
adsorption/desorption studies (see Table 4) show that (Fe). The presence of Zn in the NIWC scales was
(a) NIWC scale powders had very high micropore possibly from the galvanization layer on the steel
surface area (28–32 m2/g) in comparison with pow- pipe. However, copper (Cu) was not present in the
ders from MWRA scales (negligible to 1.5 m2/g), pipe metal or the galvanization layer, and the water
(b) NIWC scale powders had a higher Brunauer, was the only possible source. Therefore, the appear-
Emmett and Teller (BET) specific surface area (89– ance of Cu in iron corrosion scales may only be
154 m2/g) than did the MWRA samples (32–54 m2/g), explained by absorption from water. The high
and (c) the average pore diameter (by BET) of NIWC concentration of silicon (Si) in the top surface layer
scale powders (32–39 Å) was less than one-third of of the NIWC scales is most likely due to its high
MWRA scale powders (129–160 Å). The Boston #4 concentration in the NIWC water supply (Table 1).
sample showed the highest micropore area, highest Although elemental analysis provides useful in-
surface area, and smallest average pore size of all the formation about the distribution of elements in
MWRA samples. This was probably due to an corrosion scales, it is the compounds which they
experimental error. For MWRA samples, powders form that determine their reactivity. Since iron oxides
from the ‘‘top surface and shell’’ (shaded rows and readily form stable crystalline phases, X-ray diffrac-
Sample ID with suffix ‘‘-TSS’’) samples had higher tion was the preferred method for compound
surface areas, higher micropore areas, and smaller identification. The XRD patterns were acquired for
average pore diameters when compared to powders powder samples and the crystalline phases present in
prepared from the entire corrosion scale from the scales were identified. Additionally, quantitative
same MWRA sample. phase composition information was also extracted
from these patterns by using the Rietveld method
Chemical composition of corrosion scales with GSAS software (Von Dreele and Larson, 1997).
The chemical composition of corrosion scales was In this method, the experimental X-ray diffraction
characterized using EDS, XRD and wet-chemistry pattern of the sample was least-squares-fitted with a
2966 P. Sarin et al.

Table 5. EDS analysis of elemental composition of corrosion scale samplesa

Sample Analysis of Al Si P S Cl Ca Fe Cu Zn

NIWC-A Surface } 17 1 } } 5 65 3 6
Bulk 1 8 } 1 } 1 7- 4 16

NIWC-B Surface 1 12 1 } } 3 68 6 9
Bulk 1 7 } 2 } 1 72 5 12

Boston#1 Surface 2 8 1 1 } } 87 } }
TSSb 1 2 } 1 } } 93 } 2
Bulk 1 1 1 3 1 } 92 } 1

Boston#2 Surface 1 4 1 1 } 1 89 } }
TSS 1 2 } } } 1 92 } }
Bulk 4 1 } 2 } } 9- 1 }

Boston#3 Surface 3 6 1 1 } } 88 } 1
TSS 1 2 } 1 } 1 93 } 1
Bulk 1 1 1 2 } } 93 } 1

Boston#4 Surface 4 11 1 1 } } 78 } 2
TSS 2 4 1 1 } } 88 1 }
Bulk 1 1 1 3 } } 92 } }

Boston#5 Surface 2 6 } 1 5 1 78 } 2
TSS 1 2 } 2 3 } 91 } }
Bulk 1 1 } 3 3 } 91 } }

Boston#6 Surface 2 4 } 1 1 } 91 } }
TSS 1 3 } 1 } } 93 } }
Bulk 1 1 } 3 } } 92 } }
a
All values are in atom% with an error of  5%. The reported values are normalized and values less than 1 atom% are not reported.
b
Top layers of the Boston samples were analyzed separately and they have a suffix of ‘‘-TSS’’ (top surface and shell) at the end of their code.

Table 6. Rietveld analysis of powder XRD data from MWRA and NIWC scales

Sample ID w2 Magnetite (Wt.%) Geothite (Wt.%) Lepidocrocite (Wt.%)

NIWC-A 2.76 24.3 75.7 0.0

NIWC-B 2.79 20.7 79.3 0.0
BOS-1 TSS 2.41 37.2 43.9 18.9
BOS-1 2.19 34.0 59.4 6.5
BOS-2 TSS 2.38 46.2 33.6 20.2
BOS-2 2.67 35.5 60.6 3.9
BOS-3 TSS 2.39 40.5 43.4 16.0
BOS-3 2.22 35.7 59.8 4.4
BOS-4 TSS 2.53 31.0 49.9 19.1
BOS-4 2.41 23.3 71.5 5.2
BOS-5 TSS 2.33 44.4 41.9 13.7
BOS-5 1.99 30.5 59.0 10.5
BOS-6 TSS 3.61 31.6 58.5 9.9
BOS-6 2.41 27.6 67.3 5.1

Note: The samples from the top surface have ‘‘-TSS’’ as a suffix in the Sample ID.

calculated diffraction pattern. The diffraction pattern (see Table 6). A sample Rietveld fitting of an XRD
of standard crystal structures of compounds sus- pattern is shown in Fig. 3. The plus (+) symbols
pected of being present were input to the program, represent the experimental data and the grey line,
and the amount of each phase was varied until the almost overlapping with the experimental data, is the
calculated diffraction matched the measured pattern simulated diffraction pattern. The three rows of
with a desired accuracy. The powder samples were markers under the experimental and simulated
found to be crystalline and no amorphous phase was patterns show the peak positions for lepidocrocite,
detected. Since the corrosion scales were formed goethite and magnetite, in that order from top row to
under natural conditions, there was a strong possi- the bottom row. The curve at the bottom of the
bility that the compounds present in corrosion scales graph is the difference between the experimental and
were impure. The use of diffraction patterns for the simulated patterns.
standard crystal structures to calculate the simulated The quantitative phase composition of MWRA
diffraction pattern was the probable reason for large and NIWC scales is summarized in Table 6. Our
goodness of fit parameter (w2 ) values obtained results confirmed that goethite, magnetite and
 Corrosion scales in old iron pipes 2967

Fig. 3. Rietveld fitting of powder XRD studies on the corrosion scales from NIWC, Champaign, IL, for
quantitative analysis of phase composition. The three rows of markers correspond to the peak positions
for lepidocrocite (top row), goethite (middle row), and magnetite (bottom row), respectively.

lepidocrocite were the dominant constituents of iron pipes have been in service for several decades and
corrosion scales. This is in agreement with reports by have presumably witnessed a wide variation in water
other researchers. Stumm (1960), Baylis (1926) and quality during use, it is not possible to give the
Larson and King (1954) have also reported siderite as conditions that were responsible for the formation of
a component mineral phase, but the scales from lepidocrocite in MWRA pipes and its absence in
MWRA and NIWC did not contain any siderite. NIWC system pipes.
Moreover, lepidocrocite was conspicuously absent in
the NIWC scales. Higher concentrations of magnetite
in the top layers of MWRA scales is in accordance Dry vs. wet scales
with the finding of Benjamin et al. (1996) that the The analyses reported above were for dried
percentage of magnetite increased with distance from samples. In order to determine if sample preparation
the pipe wall. (vacuum drying at room temperature) had modified
The shell-like layer was present in all of the the scale composition, small amounts of wet scale
corrosion scales analyzed during this study. Its role samples were dissolved in 1 M HCl, and the solution
has been examined in considerable depth by others was analyzed for Fe(II) and Fe(III) concentrations by
(Kölle and Sontheimer, 1977; Kölle and Rosch, 1980; the bathophenanthroline method. Results showed
Sontheimer et al., 1981). Goethite and magnetite that wet scales had a relatively higher Fe(II) content
have been reported as the phases comprising this than the dry scales (Tables 7 and 8). Analyses of three
dark, relatively hard and brittle layer (Benjamin MWRA samples showed a decrease in Fe(II) content
et al., 1996). Powder samples from top layers (surface towards the top of the tubercles, reconfirming that
and shell) from MWRA scales were comprised of the oxidation state of iron increased with distance
magnetite, goethite and lepidocrocite. The top layers from the pipe wall (Table 7). These samples also
come into contact with waters having high DO and showed a high variability from sample to sample,
chlorine concentrations. The presence of lepidocro- indicating non-uniformity of scale composition.
cite in top layers is consistent with this mineral phase Powder samples used for XRD studies were also
being formed by rapid oxidation of ferrous hydro- analyzed for their Fe(II)/Fe(III) content to check the
xides (Cornell and Schwertmann, 1996). correctness of the percentage phase composition as
Synthesis studies have shown that goethite forma- obtained by the Rietveld method. While goethite and
tion predominated over lepidocrocite formation at a lepidocrocite have iron in the Fe(III) state only,
slow oxidation rate (Benjamin et al., 1996). A high magnetite (Fe3O4) has one Fe(II) and two Fe(III) in
pH and a slower rate of oxidation favored formation each molecule. Calculation of Fe(II) content from the
of magnetite over lepidocrocite (Cornell and phase composition determined from XRD results
Schwertmann, 1996). Since the MWRA and NIWC correlated well with the results of the wet chemical
2968 P. Sarin et al.

Table 7. Fe(II)/Fe(III) analyses of wet corrosion scales CONCLUSIONS

Sample details Ferrous (%) Ferric (%)

The complex microstructure of the corrosion scales
Sample ID Sampling position from old iron/steel water distribution pipelines was
NIWC-A Entire scale 30.7 69.3
analyzed using several analytical techniques. Corro-
29.2 70.8 sion scales were identified as porous deposits of iron
oxide or oxyhydroxide phases. A distinct ‘‘shell-like’’
Boston#1 Top 25.9 74.1
Middle 22.3 77.7 dense layer was present near the scale–water interface
Bottom 52.6 47.4 in all the scales analyzed. An apparent pattern was
Boston#3 Top 3.5 96.5
observed in the distribution of phases within the
Middle 94.5 5.5 scales. The top layers of the corrosion scales had
Bottom 92.2 7.8 higher concentrations of magnetite and lepidocrocite
Boston#6 Top 28.3 71.8 phases. Fe(II)/Fe(III) analyses of wet and dry scales
Middle 79.9 20.1 from the two systems illustrated that wet scales had
Bottom 80.6 19.4 significantly higher amounts of Fe(II). Important
differences were established between the corrosion
scales from MWRA and NIWC. The high concen-
tration of Si in the top layer of NIWC scales was
Table 8. Comparison of calculated Fe(II) and Fe(III) content from
probably due to the higher concentration of this
XRD studies with wet chemistry results for dry powder samples element in the NIWC water. Magnetite, goethite and
Sample ID Fe2+ and Fe3+ (wt.%)
lepidocrocite were the primary crystalline phases
present in MWRA corrosion scales, while only
Calculated from XRD Wet chemistrya magnetite and goethite were found in NIWC scales.
Fe(II) Fe(III) Fe(II) Fe(III) MWRA scales were more porous than NIWC scales.
NIWC-A 9.0 91.0 14.2 85.8
Comparison of the physical characteristics of the
NIWC-B 7.7 92.3 15.5 84.5 powdered scales showed that particles from MWRA
BOS-1 TS 13.5 86.5 15.6 84.4 scales had lower surface areas and insignificant
BOS-1 12.4 87.6 11.6 88.4
BOS-2 TS 16.6 83.4 13.9 86.1 microporosity in comparison to the NIWC scale
BOS-2 12.9 87.1 13.6 86.4 particles. While we do not understand the exact
BOS-3 TS 14.6 85.4 13.1 86.9 reasons effecting such particle formations in the two
BOS-3 13.0 87.0 12.9 87.1
BOS-4 TS 11.3 88.7 16.9 83.1 systems, we believe that the differences are note-
BOS-4 8.6 91.4 12.6 87.4 worthy.
BOS-5 TS 16.0 84.1 14.0 86.0
BOS-5 11.2 88.8 10.3 89.7
Our studies have been able to identify measurable
BOS-6 TS 11.6 88.5 11.8 88.2 physico-chemical characteristics of corrosion scales.
BOS-6 10.2 89.9 9.3 90.7 These characteristics are thought to be important
a
Values averaged for three measurements factors that affect the release of iron and the
degradation of water quality in drinking water
distribution systems. However, research is needed
to show the mechanism of this release and to develop
analysis on the same powders (Table 8). This verified efficient water quality control strategies.
the accuracy of the Rietveld analysis of XRD
Acknowledgements}The authors acknowledge the use of
patterns. facilities of the Center of Microanalysis of Materials at the
The observance of higher Fe(II) concentrations in Materials Research Laboratory, UIUC and the NSF Center
wet scales compared to dry scales indicated that for Advanced Cement Based Materials at UIUC. Prof. G. P.
Fe(II) phases were converted to Fe(III) phases during Wirtz and Prof. H. H. Chen at Department of Materials
drying. The dry samples had ferrous ions present Science and Engineering at UIUC are thanked for some
insightful discussions, as is C. J. Johnson (UIUC) for
only in the form of magnetite (Fe3O4), and the helping with the wet chemistry analyses, and Mr. M Hayes
conversion of a phase such as Fe(OH)2, to magnetite (Black and Veatch, Inc.) for operating the MWRA pipe
by oxygen according to the following reaction, could loops. This research was supported by the American Water
account for the observed change: Works Association Research Foundation (AWWARF) and
the Metropolitan Water Resources Authority (MWRA),
1 Boston, MA.
2 O2 þ 3FeðOHÞ2 ! Fe3 O4 þ 3H2 O;

DG8 ¼ 272 kJ=mol

REFERENCES
This explanation is reasonable because the oxidation
reaction is thermodynamically favored at ambient Baylis J. R. (1926) Prevention of corrosion and red water.
conditions and Fe(OH)2 may exist between the shell- J. Am. Water Works Assoc. 15(6), 598–631.
Benjamin M. M., Sontheimer H. and Leroy P. (1996)
like layer of the tubercle and the pipe wall. Other Corrosion of iron and steel. In Internal Corrosion of
similar reactions involving green rust cannot be ruled Water Distribution Systems. Cooperative Research Re-
out, however. port, AWWA Research Foundation, Denver, CO.
 Corrosion scales in old iron pipes 2969

Cornell R. M. and Schwertmann U. (1996) The Iron Oxides, corrosion and disinfection of biofilm bacteria. J. Am.
573pp, VCH Publishers, Weinhein. Water Works Assoc. 85(7), 111–123.
Drissi H., Refait P. and Genin J. M. R. (1994) The Lee G. F. and Stumm W. (1960) Determination of ferrous
oxidation of Fe(OH)2 in the presence of carbonate ions: iron in the presence of ferric iron with Bathophenanthro-
structure of carbonate green rust one. Hyperfine Inter- line. J. Am. Water Works Assoc. 52, 1567–1573.
actions 90, 1–4. Obrecht M. F. and Pourbaix M. (1967) Corrosion of metals
Eliassen R., Pereda C., Romeo A. J. and Skrinde R. T. in potable water systems. J. Am. Water Works Assoc. 59,
(1956) Effects of pH and velocity on corrosion of steel 977–992.
water pipes. J. Am. Water Works Assoc. 48(8), 1005– Olowe A. A., Génin J. M. R. and Bauer P. (1988) Hyperfine
1018. interactions and structures of ferrous hydroxide and
Field R. W., Fisher E. L., Valentine R. L. and Kross B. C. green rust II in sulfated aqueous media. Hyperfine
(1995) Radium-bearing pipe scale deposits}implications Interactions 41, 501–504.
for national waterborne radon sampling methods. Am. J. Pisigan Jr. R. A. and Singley J. R. (1987) Influence of buffer
Public Health 85(4), 567–570. capacity , chlorine residual and flow rate on corrosion of
Geldreich E. E. (1990) Internal corrosion and deposition mild steel and copper. J. Am. Water Works Assoc. 79(2),
control. In Water Quality and Treatment. McGraw Hill, 62–70.
New York. Raven K. P., Jain A. and Loeppert R. H. (1998) Arsenite
Génin J. M., Refait P., Olowe A. A., Abdelmoula M., Fall I. and arsenate adsorption on ferrihydrite}kinetics, equili-
and Drissi S. H. (1997) Identification of green rust brium, and adsorption envelopes. Environ. Sci.Technol.
compounds in the aqueous corrosion processes of steels; 32(3), 344–349.
the case of microbially induced corrosion and use of 78 K Refait P., Abdelmoula M. and Génin J. M. (1998)
CEMS. Baltzer. Hyperfine Interactions 112, 1–4. Mechanisms of formation and structure of green rust
Génin J. -M. R., Bourrié G., Abdelmoula M., Jaffrezic A., one in aqueous corrosion of iron in the presence of
Refait P., Maitre V., Himbert B. and Herbillon A. (1998) chloride ions. Corro. Sci. 40(9), 1547–1560.
Thermodynamic equilibria in aqueous suspensions of Refait P., Drissi S. H., Pytkiewicz J. and Génin J.-M. R.
synthetic and natural Fe(II)–Fe(III) green rusts: occur- (1997) The anionic species competition in iron aqueous
ences of the mineral in hydromorphic soils. Environ. Sci. corrosion: role of various green rust compounds. Corro.
Technol. 32(8), 1058–1068. Sci. 39(9), 1699–1710.
Hill R. J. and Howard C. J. (1987) Quantitative phase Rietveld H. M. (1969) A profile refinement method for
analysis from neutron powder diffraction data using the nuclear and magnetic structures. J. Appl. Crystallogr. 2,
Rietveld method. J. Appl. Crystallogr. 20, 467–474. 65–71.
Kölle W. and Rosch H. (1980) Untersuchungen an Singer P. C. and Stumm W. (1970) The solubility of ferrous
Rohrnetz–Inkrustierungen unter mineralogischen Ge- iron in carbonate-bearing waters. J. Am. Water Works
sichtspunkten. Vom Wasser 55, 159–178. Assoc. 68, 198–202.
Kölle W. and Sontheimer H. (1977) Untersuchugen zur Snoeyink V. L. and Jenkins D. (1980) Water Chemistry.
Schutzschichtbildung in Güssrohren. Vom Wasser 49, Wiley, New York, NY.
277–294. Sontheimer H., Kolle W. and Snoeyink V. L. (1981) The
Kuch A. (1988) Investigations of the reduction and re- siderite model of the formation of corrosion-resistant
oxidation kinetics of iron(III) oxide scales formed in scales J. Am. Water Works Assoc. 73(11), 572–579.
waters. Corros. Sci. 28(3), 221–231. Stumm W. (1960) Investigation on the corrosive behaviors
Larson T. E. and King R. M. (1954) Corrosion by water at of waters. J. Sanitary Eng. Division ASCE 86(11),
low flow velocity. J. Am. Water Works Assoc. 46, 1–9. 27–45.
Larson, T. E. and Skold, R. V. (1957) Corrosion and Tuovinen O. H., Button K. S., Vuorinen A., Carlson L.,
tuberculation of cast iron. J. Am. Water Works Assoc. 49, Mair D. and Yut L. A. (1980) Bacterial, chemical,
1294–1301. and mineralogical characteristics of tubercles in distribu-
Larson T. E. and Skold R. V. (1958) Laboratory studies tion pipelines J. Am. Water Works Assoc. 72(11),
relating mineral quality of water to corrosion of steel and 626–635.
cast iron. Corrosion 14(8), 285t–288t. Tuovinen O. H., Mair D. M. and Banovic J. (1984) Chlorine
Larson T. E. and Sollo Jr. F. R. (1967) Loss in water main demand and trihalomethane formation by tubercles from
carrying capacity. J. Am. Water Works Assoc. 59(12), cast iron water mains. Environ. Technol. Lett. 5, 97–108.
1565–1572. Von Dreele R. B. and Larson A. C. (1997) GSAS}General
LeChevallier M. W., Lowry C. D., Lee R. G. and Gibbon Structure Analysis System. The Regents of the University
D. L. (1993) Examining the relationship between iron of California, Los Alamos, NM.