Expanded
Summary
Effect of connection type on galvanic corrosion
between lead and copper pipes
Brandi C lark, Cl é m ent C a rti e r, Ju stin St. C l ai r , S i mo n i T r i an taf y l l i d o u , M i ch èl e P r év o s t,
and Marc Edwa rd s
http://dx.doi.org/10.5942/jawwa.2013.105.0113
Pipe connectors can influence galvanic corrosion between
lead and copper pipes by distancing the lead from copper
pipe, introducing a third metal, and forming crevices. In
this study, the effects of distance, connector material, and
crevices on galvanic corrosion were examined, and benchscale comparison testing of commercial connectors was
conducted using real tap waters. Brass connectors only
slightly decreased (< 25%) the galvanic current that
sacrifices lead pipe, with higher reductions for new brasses
with higher zinc content. Crevices in brass connectors
contained water with extremely high levels of lead (up to
9.4 × 106 μg/L), and in bench-scale tests, connections with
crevices caused approximately four times more lead release
to water than did direct connections.
The practice of partial lead service line replacement
(PLSLR), which involves replacing a portion of a lead
service line with new copper, has come under scrutiny
because of concern over elevated lead in water in both
the short and long term (Cartier et al, 2012; Giammar et
al, 2012; Brown et al, 2011; Triantafyllidou & Edwards,
2011; USEPA, 2011). Longer-term problems can arise
from direct galvanic corrosion between lead and copper
pipe or from deposition corrosion from copper onto lead
(Giammar et al, 2012; Triantafyllidou & Edwards, 2011;
Britton & Richards, 1981).
interpret longer-term comparison studies of lead contamination of water using several commercially available
connectors in two different laboratories and water supplies in Blacksburg, Va., and Montreal, Que.
STUDY BACKGROUND
Research is needed to better understand how different
types of connectors might influence the resulting galvanic
corrosion by
• breaking the electrical connection between the copper
cathode and the lead anode via an insulator or a dielectric,
• changing the distance between the lead anode and
copper cathode of the galvanic cell,
• introducing a third metal if a conductive connector
is used, and
• introducing a crevice.
The authors analyzed commercially available connectors and connectors used in previous research (Cartier et
al, 2012; Hu et al, 2012; Wang et al, 2012; Triantafyllidou & Edwards, 2011) and found wide variation in
connector material (e.g., brass, plastic), connector length
(0–5.5 in. [0–14 cm]), and the presence or absence of a
crevice or broken electrical connection.
Study objective. The goal of the current study was to
develop a mechanistic understanding of the factors influencing galvanic corrosion arising from PLSLRs using
short-term bench-scale studies. Results were used to
MATERIALS AND METHODS
Mechanistic tests. The effects of distance and connector
material were investigated by connecting 6 in. (15.2 cm)
of copper pipe to 2.5 ft (76.2 cm) of lead pipe. The connector distance was varied, and plastic connectors were
compared with red, yellow, and dezincified yellow brass.
Galvanic currents were measured using a digital multim­
eter when pipes were filled with fresh Blacksburg tap
water (chloramines; alkalinity of 40 mg/L; pH of 7.8; zinc
orthophosphate corrosion inhibitor). Crevice corrosion
was tested in several commercially available connectors
during a seven-week dump-and-fill experiment in Blacksburg tap water using 1.5-ft (45.7-cm) lengths of lead and
copper pipe. Galvanic currents were measured using
external wires, and rig construction allowed for measurement of crevice pH using a microelectrode, as well as
sampling for analyses of anions by ion chromatography
and total metals by inductively coupled plasma/mass
spectrometry (ICP/MS).
Comparison testing. To extend the mechanistic insights
to practical situations, a 26-week bench-scale study of
simulated PLSLRs with copper was conducted. Dumpand-fill pipe rigs were constructed using commercially
available brass connectors and 2.5-ft (76.2-cm) sections
of lead and copper pipe. Rigs were filled with Blacksburg tap water or aerated Montreal tap water (free
chlorine < 0.3 mg/L; alkalinity of 100 mg/L; pH of 8.5).
Differences in corrosion rate were assessed via galvanic
current and lead release to water analyzed by ICP/MS.
2013 © American Water Works Association
C LA R K ET A L | 105: 10 • JO U R NA L AWWA | O C TO B ER 2013
69
CONCLUSIONS
Mechanistic studies and short-term bench-scale studies
of simulated PLSLRs with copper in two real tap waters
yielded the following conclusions.
• As distance separating lead and copper increases,
galvanic current tends to decrease. In this study, at a
separation distance of 12 in. (30.5 cm), 80% reduction
in current was achieved.
• As a connector material, new brass may offer slight
benefits over direct connection to copper in terms of
decreasing galvanic corrosion. These benefits are greatest
when the zinc content of the brass is high, and may dissipate with time as brass dezincifies.
• Creation of a crevice involves the outer wall of the
lead pipe in galvanic corrosion and lead leaching to
potable water and also creates a small volume of water
with extremely high lead levels (as high as 9.4 × 106 μg/L)
and elevated anion concentrations (as much as 8.5 times
higher chloride, 6.5 times higher sulfate, and 25 times
higher phosphate than in the bulk water). In some cases,
when the outermost lead pipe wall was coated with
epoxy, lead concentrations in the crevice dropped by a
factor as high as 150. Connectors with crevices can markedly worsen lead release to water, compared with direct
connections or brass connectors without crevices.
• Some rubber materials present in brass connectors
do not function as a true dielectric because their resistance is on the order of 1–5 kΩ.
• In both Blacksburg and Montreal tap water during
a 26-week dump-and-fill study, all commercial brass connectors used in partial pipe replacements exhibited higher
lead release than both externally bridged connectors used
in previous laboratory studies and direct connection
between lead and copper (Figure 1). This is believed to be
a result of crevice corrosion. Although dielectrics are the
best connector choice in terms of lead release, in Mon-
treal some of the brass connectors released only one third
of the lead released by others. Therefore, if dielectrics are
not viable or allowed, utilities conducting PLSLRs should
consider using (to the extent practically reasonable) the
connectors that released less lead.
REFERENCES
Britton, A. & Richards, W.N., 1981. Factors Influencing Plumbosolvency
in Scotland. Journal of the Institute of Water Engineers and Scientists, 35:349.
Brown, M.J.; Raymond, J.; Homa, D.; Kennedy, C.; & Sinks, T., 2011.
Association Between Children’s Blood Lead Levels, Lead Service
Lines, and Water Disinfection, Washington, D.C., 1998–2006.
Environmental Research, 111:1:67. http://dx.doi.org/10.1016/
j.envres.2010.10.003.
Cartier, C.; Arnold, R.B. Jr.; Triantafyllidou, S.; Prévost, M.; & Edwards,
M., 2012. Effect of Flow Rate and Lead/Copper Pipe Sequence on
Lead Release from Service Lines. Water Research, 46:13:4142.
http://dx.doi.org/10.1016/j.watres.2012.05.010.
Giammar, D.E.; Welter G.J.; & Cantor, A., 2012. Review of Previous
Water Research Foundation Projects on Galvanic Corrosion.
www.waterrf.org/resources/Lists/PublicProjectPapers/
Attachments/3/4349_LiteratureReview.pdf (accessed July 2012).
Hu, J.; Gan, F.; Triantafyllidou, S.; Nguyen, C.K.; & Edwards, M., 2012.
Copper-induced Metal Release From Lead Pipe Into Drinking
Water. Corrosion, 68:11:1037. http://dx.doi.org/10.5006/0616.
Wang, Y.; Jing, H.; Mehta, V.; Welter, G.J.; & Giammar, D.E., 2012. Impact
of Galvanic Corrosion on Lead Release From Aged Lead Service
Lines. Water Research, 46:5049. http://dx.doi.org/10.1016/
j.watres.2012.06.046.
Triantafyllidou, S. & Edwards, M., 2011. Galvanic Corrosion After Simulated Small-scale Partial Lead Service Line Replacements. Journal
AWWA, 103:9:85.
Corresponding author: Brandi Clark is a doctoral
candidate at Virginia Polytechnic Institute and State
University, 408 Durham Hall, Blacksburg, VA 24061;
[email protected].
Connector comparison testing in Blacksburg, Va., (A) and Montreal, Que., (B) tap water
FIGURE 1
A
B
14,000
Pb concentration—µg/L
12,000
19x
20x
27x
10,000
8,000
14x
6,000
10x
4,000
5x
2,000
0
Cu
Pb
Direct
Connection
Union
Corporation
Valve
Cu
Plastic
Plastic
Disconnected Connected
Union
Corporation
Valve
Cu—copper, Pb—lead
Pooled lead concentration data are for the last eight weeks of the 26-week dump-and-fill studies conducted in phase 4. In Blacksburg
(part A), direct end-to-end connection, a brass union, and a corporation valve were compared with pure lead pipe. In Montreal (part B), an external
galvanic connection, a brass union, and a corporation valve were compared with a disconnected flexible plastic laboratory connector (simulated
dielectric). Error bars represent 95% confidence intervals. Red arrows and numbers indicate the increase in lead concentration compared
with the control condition.
2013 © American Water Works Association
70
OCT OBE R 2 0 1 3 | J O U R N A L AW WA • 1 0 5 :1 0 | C L A R K E T AL