A 30-year review of copper pitting corrosion and pinhole
leaks: Achievements and research gaps
| Bryan Karney
Department of Chemical Engineering
and Applied Chemistry, University of
Toronto, Toronto, Ontario, Canada
Department of Civil and Mineral
Engineering, University of Toronto,
Toronto, Ontario, Canada
John Gibson, Department of Chemical
Engineering and Applied Chemistry,
University of Toronto, 200 College St.,
Toronto, ON Canada.
Associate Editor: Vanessa L. Speight
Despite decades of research, the pitting of copper pipes is poorly understood.
This article summarizes the key research findings from 1990 to the present
and identifies research gaps. Several lines of evidence suggest that, for soft
waters, additional alkalinity or maintaining a pH <8.5 may reduce copper
pitting. Phosphate corrosion inhibitors, used to prevent lead corrosion, may
also moderate copper pitting in low-alkalinity waters. However, such inhibi-
tors could also exacerbate pitting due to insufficient coverage or could increase
biological growth. Biological processes are likely involved in pitting corrosion,
and practices to reduce biological growth may be beneficial. The research gaps
identified include a need for better benchmarking, the effects of chemical tran-
sients (e.g., chlorine burns) and of hydraulic transients (e.g., water hammer),
and changing the residual disinfectant from chlorine to chloramine.
copper, corrosion, pinhole leaks, pitting
Although copper pipes frequently provide long-term and
reliable service in building plumbing, pitting corrosion
leading to pinhole leaks remains a widespread challenge.
The estimated cost of prevention and repair of pinhole
leaks in the United States alone has been estimated at
approximately US$1 billion per year (Scardina, Edwards,
Bosch, Loganathan, & Dwyer, 2008), with slightly more
than half occurring in single-family homes. Approxi-
mately 750,000 pinhole leaks are estimated to occur in
the United States each year (Scardina et al., 2008). Partic-
ularly if undetected, pinhole leaks can result in expensive
water damage and mold growth inside the walls (Giani &
Hill, 2017). Homes with leaking pipes and mold are har-
der to sell and insure, which can have serious financial
repercussions for the property owner. Copper pitting cor-
rosion research, and lawsuits against the utility for pro-
viding alleged substandard water, can be traced back to
at least the mid-1960s (Rambow & Holmgren, 1966).
For practical reasons, this current review focuses on
work published between approximately 1990 and 2021.
The aim is to examine target publications, with an
emphasis on practical strategies for predicting, under-
standing, and mitigating copper pitting corrosion. Key
research gaps are suggested.
2|OVERVIEW OF COPPER
Copper pitting is traditionally divided into three types
(Giani & Hill, 2017). Type 1 pitting is sometimes called
cold-water pitting, since it has been observed primarily in
cold water pipes. Early work on Type 1 pitting associated
it with hard well waters and high carbon dioxide content
(Rambow & Holmgren, 1966). Type 1 pitting has been
reported as the most common type, accounting for 46% of
pitting failures (Scardina et al., 2008). Type 2 pitting is
sometimes called hot-water pitting, since it is often found
Received: 5 August 2020 Revised: 27 February 2021 Accepted: 5 March 2021
AWWA Wat Sci. 2021;e1221. wileyonlinelibrary.com/journal/aws © 2021 American Water Works Association. 1of9
in hot water pipes. It is worth noting that hot water pipes
are not hot during periods of stagnation, such as over-
night. Type 2 pitting is frequently observed in hot rec-
irculating water systems, and surface erosion by high
continuous flows may play a role (Roy, Coyne, Novak, &
Edwards, 2018; Scardina et al., 2008). Type 3 pitting is
called soft-water pitting and is associated with waters
with low hardness and alkalinity. Much of the current
published research is related to Type 3 soft-water pitting.
Oxidation, the process at the heart of corrosion,
involves the transfer of electrons. In corrosion, a metal
loses electrons (i.e., is oxidized) to produce positively
charged ions in solution. The electrons are accepted else-
where, often by oxygen or chlorine. Oxygen in surface
waters is dissolved from the atmosphere. However, disinfec-
tants like chlorine (HOCl or OCl
) have a higher affinity for
including pitting corrosion (Giani & Hill, 2017). The anode
is where electrons are lost and positively charged ions are
produced; the cathode is where electrons are accepted and
negative ions are produced (Snoeyink, 1980). Electrons are
often transferred through the metal from the anode to a
cathode. However, in copper pitting, it has historically been
believed that part of the pit cap structure acts as a cathode,
precipitated metal ions that forms over the anode (Cruse &
Pit initiation is complex and poorly understood yet is
believed to involve disruption of the passivating layer
(Giani & Hill, 2017). Copper is a passivating metal where,
in the presence of oxidizers, a scale layer forms on the
metal surface. This passivating layer tends to limit corro-
sion by protecting the metal from further chemical
attack. Pitting in many metals is believed to start with a
defect in the passivating layer, resulting in localized cor-
rosion (Giani & Hill, 2017).
Pitting occurs in other metals, including stainless and
carbon steels. In steel, some of the suggested causes for
pit initiation include manganese sulfide impurities in
stainless steels (Zhang & Ma, 2019), inclusions left over
from the steel-making process in stainless steel (Wang,
Cheng, Wu, & Li, 2018), and preferential adsorption of
chloride ions by impurities in carbon steels (Lin, Hu, Ye,
Li, & Lin, 2010). This suggests that impurities or imper-
fections resulting from the metal fabrication process
could play a role in pit initiation.
In copper potable water systems, Rushing and
Edwards (2004) suggest that carryover of aluminum-
containing solids from the water treatment process may be
involved in pit initiation. Aluminosilicates from a cement
water storage tank have been implicated in copper pitting
failures in an apartment building, as suggested by the high
concentration of aluminum in the scale surrounding the
pits (Hassan, 2011). Improper soldering or excess soldering
flux and chlorides have also been implicated in copper
pitting (El Warraky, El Shayeb, & Sherif, 2004; Scardina
et al., 2008).
Pit initiation has been linked to aluminum carryover
(Rushing & Edwards, 2004), and aluminum is a widely
used water treatment chemical. In other metals, such as
steel, defects from the manufacturing process have been
suggested as pit initiators (Lin et al., 2010; Wang
et al., 2018; Zhang & Ma, 2019) and may also play a role
in copper pitting.
Water quality undoubtedly plays a role in pitting corro-
sion. Pitting corrosion may also be related to more uni-
form corrosion in which copper concentrations are
elevated at the tap or blue water occurs. Certainly, wall
thinning due to uniform corrosion will increase the
likelihood that pinhole leaks will form (Health
Canada, 2019). Uniform corrosion is only briefly con-
sidered here, but low pH or alkalinity exceeding
280 mg/L CaCO
in groundwaters likely plays a role
(Health Canada, 2019).
In terms of pitting corrosion, the pit cap may facilitate
the localized corrosion of copper by, for example,
maintaining low pH conditions. Lytle and Nadagouda (2010)
provide details of the chemical structure of cold-water pits
observed in the field. The pits consisted of a cap of green
basic copper sulfate (e.g., brochantite, Cu
a thin, perforated membrane of cuprite (Cu
O) containing a
pit of cubic cuprite crystals at low pH with chloride present.
Chloride, or other anions, may help to neutralize excess
positive charge in the pit. Similarly, Edwards, Ferguson,
and Reiber (1994) report that cold-water pit caps contain
(malachite), and hot-water caps contain
(brochantite). These minerals are likely
involved in creating the structure that maintains corrosive
conditions inside the pit and provide a link between water
chemistry and pitting.
Article Impact Statement
Approximately 750,000 pinhole leaks are esti-
mated to occur in the United States each year,
but we do not really know why.
2of9 GIBSON AND KARNEY
Well waters may have many features that may con-
tribute to pitting, including relatively high carbon diox-
ide, sulfides, and possibly sulfate-reducing bacteria
(Scardina et al., 2008). Free carbon dioxide has long been
associated with pitting (Rambow & Holmgren, 1966).
Sobue et al. (2003) provide experimental evidence that
generalized corrosion of copper increases with the carbon
dioxide content of the water and note that carbon dioxide
can form carbonic acid in water (H
). They suggest
that the resulting H
ions can act as electron acceptors in
place of the traditional chlorine or oxygen. The highest
rate of corrosion was observed when the free carbon diox-
ide content was near 20 mg/L. Jacobs et al. (1998) found
that when sulfides were added, the highest degree of
pitting was observed at highest pH (9.2), but interest-
ingly, generalized corrosion was found to be the lowest.
This observation implies that significant differences exist
between uniform and pitting corrosion.
Edwards et al. (1994) summarized the water chemistry
associated with observed pitting attack in field studies,
pointing out that these characteristics are more rules of
thumb than definitive associations. In general, pitting-
prone waters tend to be slightly basic pH (7.3) and have
high hardness (152 mg/L), high sulfate content (265 mg/
L), and high sulphate-to-chloride ratios (4.3). However,
there is wide variability in water chemistry where pitting
was observed. Sulfate concentration was high in some
cases, approaching 400 mg/L, but in a few cases, it was
undetectable. This suggests a complex relationship between
water chemistry and pitting attack in full-scale systems.
Harrison, Nicholas, and Evans (2004) studied soft-
water pitting in New Zealand and suggested that
maintaining a pH of 8 can aid in the formation of passivat-
ing cuprous oxide (Cu
O) and prevent the formation of
malachite, believed to be to be important for maintaining
pit growth. Pitting corrosion requires a fresh supply of
reactants for the pits to continue to grow. Cong and
Scully (2010) showed that the cathodic mass-transport fac-
tor, believed to control the growth rate of copper pits,
increased from pH 7 to 9.5, with a sharp increase as pH is
increased from 8.5 to 9.5 in soft water. In a follow-up
work, Cong and Scully (2010) showed that pitting factors
increased as the pH was increased from 7 to 9 in soft
waters in the presence of chlorine. Overall, faster pitting is
likely in pH >8.5, soft, low-alkalinity waters.
Evidence that high sulphate-to-bicarbonate ratios are
favorable for pitting is provided by Ha et al. (2011) in a
laboratory study. They found that the highest pit propa-
gation rates and deepest pits were formed in sulfate-
containing waters. In the sulfate-containing waters, the
initial pit growth rate was slower but did not repassivate
over time, leading to perforations. In chloride-containing
waters, repassivation occurred, halting pit growth. Ha
and Scully (2013) support this work with a lab-scale elec-
trochemical study. They show that HCO
tions could stifle the propagation of previously
established pits. They attribute this effect to the forma-
tion of highly resistive malachite (Cu
copper chloride layers inside the pit and pit cap. This sug-
gests that the addition of NaHCO
, a source of alkalinity,
may be a beneficial strategy to control pitting. Other
authors (Edwards et al., 1994; Harrison et al., 2004;
Lytle & Schock, 2008) have noted beneficial effects asso-
ciated with increased HCO
or alkalinity, which is fur-
ther supported by this work. Furthermore, the entire
category of soft-water pitting is often associated with low
alkalinity, suggesting that increased alkalinity may be
beneficial to slow or prevent pitting.
However, it has often proved difficult to replicate lab
findings at larger scales. Several lines of evidence suggest
that high sulfate concentration (Duthil et al., 1996) or high
sulfate-to-bicarbonate ratios (Edwards et al., 1994; Ha
et al., 2011) contribute to pitting. However, in the pilot-
scale work of Lytle and Schock, (2008), increasing the sul-
fate concentration from 0 to 150 mg/L did not result in
observable pitting in this water at pH 7, 8, or 9 in pipe loop
trials. The role of sulfate in pitting attack at the pilot scale
In summary, water chemistry inevitably plays a role in
pitting, but the results can be difficult to generalize and
replicate, especially at larger scales. Overall, this suggests
that some other factors, which may be independent of
traditional water chemistry parameters, may be important.
Boulay and Edwards (2001) suggested that natural
organic matter (NOM) stimulates biological growth that
depletes oxygen and reduces pH, both of which are
believed to contribute to copper corrosion. Bremer, Web-
ster, and Wells (2001) provided a review of microbially
influenced corrosion (MIC) of a copper pipe. Several field
studies noted that pitting occurred in close proximity to
biofilms. Bremer et al. (2001) also examined the role of
organic substances in creating soluble copper complexes,
elevating the copper concentration at the tap. The
authors reported that this type of pitting occurred in soft
waters with a high pH and high assimilable organic car-
bon content. The authors suggested that stagnation and
loss of chlorine residual may also play a role.
The role of MIC in pitting was further explored by
Keevil (2004). The water was soft (hardness, 25–40 mg/L
), poorly buffered (alkalinity, 10–20 mg/L), and
below 60 C. The proposed mechanism for pitting involved
the attachment of humic substances from the source water
GIBSON AND KARNEY 3of9
to the pipe surface. Bacterial cells were observed at the bot-
tom of the pits. Nam, Lee, and Kim (2018) provided several
lines of evidence suggesting that biological processes con-
tribute to copper corrosion and may be linked to pitting.
Some evidence of the role of microbial activity on uniform
copper corrosion was provided by Pavissich, Vargas,
González, Pastén, and Pizarro (2010), who observed higher
copper concentrations in pipes inoculated with microbe-
containing water. Vargas et al. (2017) provide a detailed
review of copper corrosion, with an emphasis on microbial
processes. They discussed how biofilms can create differ-
ences in oxygen concentration at the pipe surface, with the
potential to create anodic regions that may result in pit ini-
tiation. The authors reported that traditional chlorine disin-
fectants reduce planktonic bacteriabutarenotexpectedto
affect biofilms on the pipe surface, where corrosion is
occurring. Burleigh, Gierke, Fredj, and Boston (2014) pro-
vided additional evidence for a role of microbes in copper
pitting. Scanning electron microscopy revealed long tubu-
lar structures resembling actinobacteria deep inside copper
pits. The authors speculate that these bacteria may be
involved in acid production that contributes to pit growth.
Seawater has a higher concentration of sulfate than
drinking water. Several researches have observed
sulphate-reducing bacteria and copper sulfide formation
in copper pipes carrying seawater (Chen & Zhang, 2018;
Chen, Wang, & Zhang, 2014). This suggests that sulfate-
reducing bacteria can play a role in copper corrosion.
Heterotrophic bacteria require a carbon source. As
such, efforts to reduce the organic carbon content of
waters may be beneficial in preventing MIC. This may
have the additional benefit of removing disinfection
byproduct (DBP) precursors as well.
Biological processes are definitely involved in copper
corrosion (Bremer et al., 2001; Keevil, 2004; Nam
et al., 2018) and likely play some role in pitting corrosion
(Burleigh et al., 2014). In general, reducing the organic
carbon content of the water will reduce the food available
for microorganisms to grow. Where possible, there is evi-
dence that higher temperatures (i.e., 70 C) may help to
reduce MIC, including pitting (Montes, Hamdani, Creus,
Touzain, & Correc, 2014). Disinfectant residuals may
help, but it is important to remember that corrosion
occurs at the surface and not the bulk water (Vargas
et al., 2017). A mass transfer process is likely required to
transfer the disinfectants to the metal surface, which may
be especially difficult inside pits.
Roy et al. (2018) reviewed the physical processes involved
in the deterioration of copper pipe, including concentration
cell corrosion, cavitation, particle/bubble impingement,
and high-velocity impingement. Fundamentally, erosion
appears to remove the passivating layer on the pipe sur-
face, allowing corrosion to proceed more rapidly, particu-
larly if the chemical conditions are favorable. The
authors reported that erosion corrosion is often observed
in hot recirculating water systems, where the water is
continuously in motion and temperatures are high.
Erosion corrosion can result in both wall thinning and
pinhole leaks (Roy et al., 2018).
Water hammer may play a role in transforming pits
into leaks. Water hammer is the sudden change in pres-
sure resulting from flow adjustment. Modern dish-
washers, refrigerators, washing machines, and other
equipment often use fast-acting solenoid valves that sud-
denly stop and start water flow (Duncan, 2001). The max-
imum pressure increase from water hammer can be
estimated using the Joukowsky equation:
where ΔPisthepressurechange;ρis the water density;
ais the wave speed, which depends on pipe material and
wall thickness; and ΔVis the change in velocity. The pres-
sure change can be both positive or negative depending on
if the flow suddenly starts or stops. Assuming a flow veloc-
ity of 1.0 m/s (3 ft/s) and a wave speed of 1,200 m/s for a
½00 copper pipe, the maximum expected pressure increase
due to water hammer is 1,200 kPa (175 psi). Though the
pipes may be designed to withstand these transient pres-
sures, pits are not. The impact of surge protection has been
largely unexplored in the published literature related to
copper pitting, with the exception of Roy et al. (2018), who
mention water hammer briefly.
Copper pitting tends to form small holes, and it may be
possible to fill them without replacing the pipe. Lytle and
Nadagouda (2010) provided field evidence that corrosion
products outside the pipe reduced the leakage rate. Pipe
remediation strategies could involve the addition of bulking
agents and the use of corrosion products, such as those
formed by corrosion inhibitors, to fill holes (Tang, Tri-
antafyllidou, & Edwards, 2015). However, the authors point
out that this is more difficult as water pressure increases.
Current regulations related to DBPs have prompted many
utilities to switch from chlorine to chloramine as their
secondary disinfectant (Norton & LeChevallier, 1997).
Although chloramines can form fewer regulated DBPs, they
are subject to biological growth in the form of nitrification
4of9 GIBSON AND KARNEY
(Krishna & Sathasivan, 2010). To combat high bacterial
counts at the tap when using chloramine, high concentra-
tions of chlorine are used intermittently to curtail biological
growth (i.e., chorine burns) (Carrico et al., 2008). Relatively
little is known about how the switch to chloramine, or chlo-
rine burns, will affect copper corrosion in general and
pitting corrosion in particular.
Lead corrosion is definitely affected by the choice of
disinfectants. Lead concentrations were shown to increase
by a factor of 10 when chloramine was used, compared
with free chlorine (Woszczynski, Bergese, & Gagnon,
2013). Vasquez, Heaviside, Tang, and Taylor (2006)
argued that the higher redox potential of chlorine
resulted in the formation of insoluble lead dioxide, which
slowed the overall rate of corrosion. When it comes to
copper, Treweek, Glicker, Chow, and Sprinker (1985)
observed consistently higher copper concentrations in
water samples in the presence of chloramine when com-
pared with chlorine. High copper concentrations in water
samples are consistent with uniform corrosion, where
much of the pipe surface contributes to copper, rather
than pitting corrosion. The role of chloramine in copper
pitting is not clear and is suggested as an area for addi-
There is evidence that high chlorine concentration
burns may exacerbate copper pitting. Lytle and
Schock (2008) observed that elevated chlorine concentra-
tion accelerated the rate of pitting and increased the pH
range where pitting attack occurred in soft water. In
static tests in hot water, Montes et al. (2014) found evi-
dence of more highly oxidized tenorite (CuO) in the pres-
ence of 100 mg/L of sodium hypochlorite but less
oxidized cuprite (Cu
O) when the disinfectant was
absent. This suggests that high chlorine concentration
can facilitate copper oxidation. In a dynamic pipe loop
test by the same authors, at 50 C, pitting was observed
when 25 mg/L of chlorine was present after 4 weeks, but
when the chlorine concentration was 1 mg/L, the corro-
sion was uniform. Similarly, Fujii and Baba (1984) found
evidence of Type 2 copper pitting in hot water with a
chlorine concentration of 2–3 mg/L but none when chlo-
rine was absent or the sample was aerated. At ambient
temperature (20 C), pipe loop testing of Boulay and
Edwards (2001) showed an increase in copper release,
usually associated with uniform and not pitting corro-
sion, when the chlorine residual was increased from 0.7
to 2 mg/L. Overall, this suggests that higher free chlorine
concentrations may stimulate pitting. No systematic stud-
ies of the effects of chlorine burns were found, and this is
suggested as an area of future research. Furthermore, the
effect of increased biological growth associated with nitri-
fication of chloramines and its impacts on corrosion are
In the United Kingdom, benzotriazole and tri-
ethanolamine are used as corrosion inhibitors. Cyclic
voltammetry showed that, once applied, the corrosion
inhibition abilities of these compounds persisted even after
they were no longer present in the water (Burstein, Bi, &
NOM has been shown to reduce the rate of uniform
copper corrosion from roughly 15 to <0.05 mil per year
(mpy) (Edwards et al., 1994). However, NOM has also
been shown to increase the corrosion of lead, possibly by
disrupting the passivating layer (Korshin, Ferguson, &
Lancaster, 2005). Sarver and Edwards (2012) suggested
that humic substances may be beneficial to slow pitting.
NOM is a complex assemblage of organic molecules
affected by local geography. The characteristic effects of
NOM are likely site-specific or seasonal (Bhatnagar, 2017;
Weiss et al., 2013), so it is difficult to generalize.
Phosphate corrosion inhibitors are often recommended
to reduce lead and copper concentrations at the tap
(Giani & Hill, 2017; Scardina et al., 2008). A possible disad-
vantage of phosphate addition is increased biological
growth, since phosphorus is a nutrient. Miettinen, Var-
tiainen, and Martikainen (1997) suggested that phosphate
was a growth-limiting nutrient in their boreal water source,
and phosphate addition stimulated growth. Sathasivan and
Ohgaki (1999) suggested that either phosphate or carbon
can be growth-limiting depending on the water source.
Appenzeller (2001) found that phosphate had no effect on
biological growth in new pipe and significantly decreased
growth in corroded cast-iron pipe. Gouider, Bouzid, Sayadi,
and Montiel (2009) found no evidence of increased biologi-
cal growth after addition of 1 mg/L of phosphate. In con-
trast, Jang, Choi, Ro, and Ka (2012) did see evidence of
increased turbidity and biological growth when 5 mg/L of
phosphate was added to a ductile-iron pipe reactor. Overall,
it appears that the effects of phosphates on biological
growth depend on the organic content of the water and
may be regional or site-specific.
In terms of the efficacy of phosphate addition, Lytle,
Schock, Leo, and Barnes (2018) provide ample evidence
from a large number of field studies that phosphates can
lower copper concentrations at the tap. However, high
concentrations at the tap are often associated with uni-
form, rather than pitting, corrosion. Schock and
Fox (2001) showed that 3 mg/L of phosphate could
reduce copper and lead concentrations at the tap in field
studies. Lytle and Schock (2008) observed that phosphate
inhibitors slowed corrosion, including pitting corrosion,
in their soft, low-alkalinity water. There is evidence that
phosphates may affect new and old pipes differently as the
passivating scale moves toward more thermodynamically
GIBSON AND KARNEY 5of9
stable forms (Schock & Sandvig, 2009). Yohai, Schreiner,
Vázquez, and Valcarce (2013) provided evidence that phos-
phate corrosion inhibitors can slow uniform copper corro-
sion, likely through the formation of a passivating layer of
on the metal surface.
The influence of phosphates on pitting corrosion is
less clear. After a series of tests in which high voltages
are applied to generate pitting, the only sample that did
not show evidence of pitting corrosion did not have
phosphate added. Water alone, water and inhibitor, or
inhibitor and chlorine together, all showed evidence of
pitting in polarization tests and visual observations
(Yohai et al., 2013). Zhang and Andrews (2013) showed
that both phosphate addition and low-flow velocity
could reduce copper concentrations. In one of the rela-
tively few studies dedicated to phosphate addition and
pitting corrosion specifically, Lytle and White (2014)
observed that phosphates tended to collect in the pit
cap. They suggest that this could be related to the flow
of anions (e.g., PO
) toward the pit to balance the
excess positive charge at the anode (e.g., Cu
). This pro-
vides important information about the structure and
mechanism of pitting corrosion inhibition; however, no
information was provided about the efficacy of this
treatment. Montes et al. (2014) showed that adding
phosphate corrosion inhibitors to soft water at 50 Cdel-
ayed the initiation of pitting. Sarver and Edwards (2012)
showed that 1 mg/L of phosphate completely prevented
the initiation of pits, and 5 mg/L silica significantly
decelerated pitting in a pH 8.3, low-alkalinity (34 mg/L
) water. However, much lower doses of these
inhibitors had little benefit and actually accelerated the
rate of attack. This is consistent with traditional pitting
theory, in which insufficient coverage of the inhibitor
can promote localized corrosion. This is supported by
Feser and Schewe (2016), who showed that partial cov-
erage of a copper sheet with a silicate film resulted in
large potential differences and the formation of an
anode at the boundary. Ha and Scully (2013) showed
that 10 mg/L of phosphate could slow the rate of copper
pitting in a synthetic water, particularly if the pits were
shallow (<20 μm deep). Since it appears that corrosion
inhibitors can be both beneficial and detrimental to
pitting, additional lab-scale, pilot-scale, and full-scale
testing of inhibiters is recommended.
Taken as a whole, the literature clearly indicates the
challenge of identifying consistent trends in copper
pitting. This may suggest that either there are important
factors that are not currently well understood or the
interaction of known factors is particularly complex.
The following factors are identified as contributing to
pitting to a degree and in ways that are not yet fully
1. Effects of phosphate corrosion inhibitors.Sincethe
water-related events in Flint, Michigan, there has
been an increasing incentive to add phosphate-
based corrosion inhibitors to reduce the lead con-
tent in drinking water, especially in North America.
Several lines of evidence suggest that these corro-
sion inhibitors may be beneficial to prevent pitting
corrosion, particularly in low-alkalinity waters. How-
ever, inadequate surface coverage can stimulate corro-
sion (Feser & Schewe, 2016), and phosphates have the
potential to increase biological growth in some cases
(Gouider et al., 2009; Miettinen et al., 1997). Additional
research on pitting and corrosion inhibitors is
2. Chemical transients. Unlike in the laboratory, the
chemical conditions in water distribution networks
are constantly changing. For example, high chlorine
concentrations (“burns”) are sometimes used to con-
trol biological growth in summertime. Disinfectant
residual concentrations are inherently variable as flow
and chlorine demand change. The role these events
play in pitting corrosion is not clear, but several
authors suggest that higher chlorine levels can facili-
tate pitting (Fujii & Baba, 1984; Lytle & Schock, 2008;
Montes et al., 2014).
3. Hydraulic transients and physical stresses. Pits can
weaken the pipe wall. It is possible that these weak
spots become leaks as a result of physical stress. For
example, water hammer may play a role in the devel-
opment of leaks. Water surge protection devices may
play a role in preventing leaks caused by pitting
4. Chloramine. Many utilities are moving toward chlora-
mine instead of chlorine to limit the formation of
DBPs. The effect this will have on pitting corrosion is
unclear. Conventional wisdom is that chloramine is
not as powerful an oxidizer as chlorine and is better at
penetrating, and thus controlling, biofilms. This may
make it less likely to promote traditional electrochem-
ical corrosion and MIC, yet there is little proof.
5. Benchmarking. Copper pipes in a home can be in ser-
vice for many decades. It is not clear how many pits
should develop into leaks after, say, 30 years.
Benchmarking can help indicate what is a high or low
level of pitting. A standard measure, such as number
of leaks per home per year of service based on obser-
vational studies, would be helpful.
6of9 GIBSON AND KARNEY
The interface between the water and the pipe wall
contains billions of molecules that interact. There can be
millions of surface imperfections, changes in copper
grain structure, and microbiological communities that
can find a home in this environment. In addition, the
complex history of pipeline material, water quality, tem-
perature, pipe stresses, and installation quality may also
play a role. Thus, it is hardly surprising that many uncer-
The study of pitting corrosion in full-scale systems
often involves a kind of forensic research, in which the
factors leading up to pitting attack are often unknown. It
is customary to try to link something that can be easily
measured, such as the current water quality, to the occur-
rence of pits that may have developed over many years
under possibly different conditions. Sometimes these
water quality characteristics are supported by evidence of
pitting in bench-scale and pilot-scale research, but consis-
tent patterns are hard to discern. In addition, there may
have been short-lived events in the past that encouraged
pitting. Full-scale systems have many potential factors
that may contribute to pitting corrosion, and such factors
are seldom considered in the lab. These may include fab-
rication issues, chemical and physical transients, biology,
and changing disinfectants.
The reviewed literature identifies several trends,
though there is no guarantee that these trends will apply
to all situations:
•For low-alkalinity and soft waters, additional alkalinity
may to be beneficial to slow or prevent pitting corrosion
(Edwards et al., 1994; Harrison et al., 2004; Lytle &
Schock, 2008). The mechanism may involve stabilization
of the pH inside the pits. In addition, maintaining a pH
below approximately 8.5 in these waters may also be
beneficial (Cong & Scully, 2010; Harrison et al., 2004).
•Phosphate corrosion inhibitors used to comply with the
Lead and Copper Rule may be beneficial to slowing
pitting corrosion in some cases. This is supported by
research in primarily low-alkalinity (<150 mg/L) waters
(Ha & Scully, 2013; Lytle & White, 2014; Sarver &
Edwards, 2012). However, phosphates can also contrib-
ute to biological growth or promote localized corrosion
if the dose is insufficient (Appenzeller, 2001; Feser &
Schewe, 2016; Jang et al., 2012; Sathasivan &
•Biological processes are likely involved in pitting corro-
sion (Burleigh et al., 2014; Pavissich et al., 2010). Prac-
tices to reduce biological growth may be beneficial to
slowing pitting. This may include reducing the assimi-
lable organic carbon content of the water (e.g., through
enhanced coagulation or biofiltration), maintaining
disinfectant residuals, and preventing stagnation
(Vargas et al., 2017). Reducing the organic carbon con-
tent may also reduce the potential for DBP formation.
However, some sources of organic carbon, such as
humic substances, may be natural corrosion inhibitors,
so the effects are likely site-specific.
CONFLICT OF INTEREST
The authors declare no potential conflict of interest.
John Gibson: Investigation; methodology. Bryan
Karney: Conceptualization; resources.
DATA AVAILABILITY STATEMENT
Data sharing not applicable to this article as no data sets
were generated or analyzed during the present study.
John Gibson https://orcid.org/0000-0003-0915-8130
Bryan Karney https://orcid.org/0000-0001-9154-8722
Appenzeller, B. (2001). Effect of adding phosphate to drinking
water on bacterial growth in slightly and highly corroded pipes.
Water Research,35, 1100–1105. https://doi.org/10.1016/S0043-
Bhatnagar, A. & Sillanpää, M. (2017). Removal of natural organic
matter (NOM) and its constituents from water by adsorption—
a review. Chemosphere,166, 497–510. https://doi.org/10.1016/j.
Boulay, N., & Edwards, M. (2001). Role of temperature, chlorine,
and organic matter in copper corrosion by-product release in
soft water. Water Research,35, 683–690. https://doi.org/10.
Bremer, P. J., Webster, B. J., & Wells, D. B. (2001). Biocorrosion of
copper in potable water. Journal American Water Works Associ-
Burleigh, T., Gierke, C., Fredj, N., & Boston, P. (2014). Copper tube
pitting in Santa Fe municipal water caused by microbial
induced corrosion. Materials,7, 4321–4334. https://doi.org/10.
Burstein, G. T., Bi, H., & Kawaley, G. (2016). The persistence of inhibi-
tion of copper corrosion in tap water. Electrochimica Acta,191,
Carrico, B. A., Digiano, F. A., Love, N. G., Vikesland, P. J.,
Chandran, K., Fiss, M., & Zaklikowski, A. (2008). Effectiveness
of switching disinfectants for nitrification control. American
Water Works Association Journal,100, 104–115.
Chen, S., Wang, P., & Zhang, D. (2014). Corrosion behavior of cop-
per under biofilm of sulfate-reducing bacteria. Corrosion Sci-
ence,87, 407–415. https://doi.org/10.1016/j.corsci.2014.07.001
Chen, S., & Zhang, D. (2018). Effects of metabolic activity of sul-
phate-reducing bacteria on heterogeneous corrosion behaviors
of copper in seawater. Materials and Corrosion,69(8), 985–997.
Cong, H., & Scully, J. R. (2010a). Effect of chlorine concentration
on natural pitting of copper as a function of water chemistry.
GIBSON AND KARNEY 7of9
Journal of the Electrochemical Society,157, C200. https://doi.
Cong, H., & Scully, J. R. (2010b). Use of Coupled Multielectrode
Arrays to Elucidate the pH Dependence of Copper Pitting in
Potable Water. Journal of the Electrochemical Society,157, C36.
Cruse, H., & Pomeroy, R. D. (1974). Corrosion of copper pipes. Jour-
nal AWWA,66(8), 479–483. https://doi.org/10.1002/j.1551-
Duncan, J. (2001). Plumbing technology (2nd ed.). Rosemont, Ill:
American Society of Plumbing Engineers.
Duthil, J.-P., Mankowski, G., & Giusti, A. (1996). The synergetic
effect of chloride and sulphate on pitting corrosion of copper.
Corrosion Science,38, 1839–1849. https://doi.org/10.1016/
Edwards, M., Ferguson, J. F., & Reiber, S. H. (1994). The pitting cor-
rosion of copper. Journal American Water Works Association,
sion of copper in chloride solutions. Anti-corrosion Methods and
Feser, R., & Schewe, S. (2016). Pitting corrosion of copper tubes for
drinking water applications due to silicate films. ECS Transac-
tions,75, 137–144. https://doi.org/10.1149/07501.0137ecst
Fujii, T., & Baba, H. (1984). The effect of water quality on pitting
corrosion of copper tube in hot soft water. Corrosion Science,
Giani, R. E., & Hill, C. P. (2017). M58—internal corrosion control in
water distribution systems (2nd ed.). Denver, CO: American
Water Works Association.
Gouider, M., Bouzid, J., Sayadi, S., & Montiel, A. (2009). Impact of
orthophosphate addition on biofilm development in drinking
water distribution systems. Journal of Hazardous Materials,
167, 1198–1202. https://doi.org/10.1016/j.jhazmat.2009.01.128
Ha, H. M., & Scully, J. R. (2013). Effects of phosphate on pit stabili-
zation and propagation in copper in synthetic potable waters.
Ha, H., Taxen, C., Williams, K., & Scully, J. (2011). Effects of
selected water chemistry variables on copper pitting propaga-
tion in potable water. Electrochimica Acta,56, 6165–6183.
Harrison, D. B., Nicholas, D. M., & Evans, G. M. (2004). Pitting cor-
rosion of copper tubes in soft drinking waters: Corrosion mech-
anism. Journal American Water Works Association,96,67–76.
Hassan, S. F. (2011). Cement particle induced failure of cold potable
water copper plumbing. Engineering Failure Analysis,18,
Health Canada. (2019). Guidelines for Canadian drinking water.
Quality guideline technical document: Copper. Ottawa, Ont:
Minister of Health.
Jacobs, S., Reiber, S., & Edwards, M. (1998). Sulfide-induced copper
corrosion. Am. Water Works Assoc,90,62–73. https://doi.org/
Jang, H.-J., Choi, Y.-J., Ro, H.-M., & Ka, J.-O. (2012). Effects of
phosphate addition on biofilm bacterial communities and water
quality in annular reactors equipped with stainless steel and
ductile cast iron pipes. Journal of Microbiology,50,17–28.
Keevil,C.W.(2004).Thephysico-chemistry of biofilm-mediated pitting
corrosion of copper pipe supplying potable water. Water Science
and Technology,49,91–98. https://doi.org/10.2166/wst.2004.0096
Korshin, G. V., Ferguson, J. F., & Lancaster, A. N. (2005). Influence
of natural organic matter on the morphology of corroding lead
surfaces and behavior of lead-containing particles. Water
Research,39, 811–818. https://doi.org/10.1016/j.watres.2004.
Krishna, K. C. B., & Sathasivan, A. (2010). Does an unknown mech-
anism accelerate chemical chloramine decay in nitrifying
waters? Journal American Water Works Association,102,82–90.
Lin, B., Hu, R., Ye, C., Li, Y., & Lin, C. (2010). A study on the initia-
tion of pitting corrosion in carbon steel in chloride-containing
media using scanning electrochemical probes. Electrochimica Acta,
Lytle, D. A., & Nadagouda, M. N. (2010). A comprehensive investi-
gation of copper pitting corrosion in a drinking water distribu-
tion system. Corrosion Science,52, 1927–1938. https://doi.org/
Lytle, D. A., & Schock, M. R. (2008). Pitting corrosion of copper in
waters with high pH and low alkalinity. Journal American
Water Works Association,100, 115–129. https://doi.org/10.1002/
Lytle, D. A., Schock, M. R., Leo, J., & Barnes, B. (2018). A model for
estimating the impact of orthophosphate on copper in water.
Journal American Water Works Association,110,E1–E15.
Lytle, D. A., & White, C. P. (2014). The effect of phosphate on the
properties of copper drinking water pipes experiencing local-
ized corrosion. Journal of Failure Analysis and Prevention,14,
Miettinen, I. T., Vartiainen, T., & Martikainen, P. J. (1997). Phos-
phorus and bacterial growth in drinking water. Applied and
Montes, J. C., Hamdani, F., Creus, J., Touzain, S., & Correc, O.
(2014). Impact of chlorinated disinfection on copper corrosion
in hot water systems. Applied Surface Science,314, 686–696.
Nam, G., Lee, J., & Kim, K. (2018). Failure analysis of pitted copper
pipes used in underground water and preventive measures.
Metals and Materials International,24, 496–506. https://doi.
Norton, C. D., & LeChevallier, M. W. (1997). Chloramination: Its
effect on distribution system water quality. Journal American
Water Works Association,89(7), 66.
Pavissich, J. P., Vargas, I. T., González, B., Pastén, P. A., &
Pizarro, G. E. (2010). Culture dependent and independent ana-
lyses of bacterial communities involved in copper plumbing
corrosion: Bacteria in copper piping corrosion. Journal of
Applied Microbiology,109, 771–782. https://doi.org/10.1111/j.
Rambow, C., & Holmgren, R. (1966). Technical Legal Aspects of
Copper Tube Corrosion. Journal AWWA,58(3), 347–353.
Roy, S., Coyne, J. M., Novak, J. A., & Edwards, M. A. (2018). Flow-
induced failure mechanisms of copper pipe in potable water
systems. Corrosion Reviews,36, 449–481. https://doi.org/10.
8of9 GIBSON AND KARNEY
Rushing, J. C., & Edwards, M. (2004). Effect of aluminium solids
and chlorine on cold water pitting of copper. Corrosion Science,
46, 3069–3088. https://doi.org/10.1016/j.corsci.2004.05.021
Sarver, E., & Edwards, M. (2012). Inhibition of copper pitting corro-
sion in aggressive potable waters. International Journal of Cor-
Sathasivan, A., & Ohgaki, S. (1999). Application of new bacte-
rial regrowth potential method for water distribution
system—A clear evidence of phosphorus limitation. Water
Research,33, 137–144. https://doi.org/10.1016/S0043-1354
Scardina, P., Edwards, M., Bosch, D. J., Loganathan, G. V., &
Dwyer, S. K. (2008). Assessment of non-uniform corrosion in
copper piping (p. 200). Denver, Colo: AWWA Research
Schock, M.R., Fox, J.C., 2001. Solving copper corrosion problems
while maintaining lead control in a high alkalinity water using
Schock, M. R., & Sandvig, A. M. (2009). Long-term effects of ortho-
phosphate treatment on copper concentration. Journal Ameri-
can Water Works Association,101,71–82. https://doi.org/10.
Snoeyink, V. L. (1980). Water chemistry. New York, NY: Wiley.
Sobue, K., Sugahara, A., Nakata, T., Imai, H., & Magaino, S. (2003).
Effect of free carbon dioxide on corrosion behavior of copper in
simulated water. Surface and Coatings Technology,169-170,
Tang, M., Triantafyllidou, S., & Edwards, M. A. (2015). Autogenous
metallic pipe leak repair in potable water systems. Environmen-
tal Science & Technology,49, 8697–8703. https://doi.org/10.
Treweek, G. P., Glicker, J., Chow, B., & Sprinker, M. (1985). Pilot-
plant simulation of corrosion in domestic pipe materials. Jour-
nal American Water Works Association,77,74–82. https://doi.
Vargas, I., Fischer, D., Alsina, M., Pavissich, J., Pastén, P., &
Pizarro, G. (2017). Copper corrosion and biocorrosion events in
premise plumbing. Materials,10, 1036. https://doi.org/10.3390/
Vasquez, F. A., Heaviside, R., Tang, Z. J., & Taylor, J. S. (2006).
Effect of free chlorine and chloramines on lead release in a dis-
tribution system. Journal American Water Works Association,
98, 144–154. https://doi.org/10.1002/j.1551-8833.2006.tb07596.x
Wang, Y., Cheng, G., Wu, W., & Li, Y. (2018). Role of inclusions in
the pitting initiation of pipeline steel and the effect of electron
irradiation in SEM. Corrosion Science,130, 252–260. https://doi.
Weiss, W. J., Schindler, S. C., Freud, S., Herzner, J. A., Hoek, K. F.,
Wright, B. A., …Becker, W. C. (2013). Minimizing raw water
NOM concentration through optimized source water selection.
Journal American Water Works Association,105, E596–E608.
Woszczynski, M., Bergese, J., & Gagnon, G. A. (2013). Comparison
of chlorine and chloramines on Lead release from copper pipe
rigs. Journal of Environmental Engineering,139, 1099–1107.
Yohai, L., Schreiner, W. H., Vázquez, M., & Valcarce, M. B. (2013).
Phosphate ions as inhibiting agents for copper corrosion in
chlorinated tap water. Materials Chemistry and Physics,139,
Zhang, B., & Ma, X. L. (2019). A review—Pitting corrosion initiation
investigated by TEM. Journal of Materials Science and Technol-
Zhang, H., & Andrews, S. A. (2013). Effects of pipe materials, ortho-
phosphate, and flow conditions on chloramine decay and
NDMA formation in modified pipe loops. Journal of Water Sup-
ply Research and Technology-Aqua,62, 107–119. https://doi.
How to cite this article: Gibson J, Karney B. A
30-year review of copper pitting corrosion and
pinhole leaks: Achievements and research gaps.
AWWA Wat Sci. 2021;e1221. https://doi.org/10.
GIBSON AND KARNEY 9of9