United States
Environmental Protection
Agency
Water Engineering
Research Laboratory
Cincinnati, OH 45268
Research and Development
EPA/600/S2-87/031 June 1987
f/EPA Project Summary
Evaluation of Silicate and
Phosphate Compounds for
Corrosion Control
Shankha K. Banerji, John E. Bauman, and John T. O'Connor
Various dosages of selected silicate
and phosphate compounds were evalu-
ated for their ability to inhibit corrosion
of cast iron, copper, lead, and galva-
nized steel specimens in drinking
water. The compounds selected for
study were zinc polyphosphate (Calgon
C-39*), zinc orthophosphate (Virchem
V-931), sodium metasilicate, and glassy
silicate. The effectiveness of these com-
pounds for corrosion inhibition were
studied under different water quality
conditions using gravimetric and elec-
trochemical corrosion tests.
Study results indicate that some cor-
rosion inhibitors provide better protec-
tion for some metallic systems than
others. Utilities should therefore use
either a gravimetric or an electrochemi-
cal corrosion test to evaluate the corro-
siveness of their water systems and the
effectiveness of any corrosion-inhibit-
ing compounds used.
This Project Summary was devel-
oped by EPA's Water Engineering Re-
search Laboratory, Cincinnati, OH, to
announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
Corrosion in potable water distribu-
tion systems is a continuous problem
faced by water utilities. The problems
created as a result of corrosion can be
grouped into three categories: econom-
ics, aesthetics, and health. Corrosion
•Mention of trade names or commercial products
does not constitute endorsement or recommenda-
tion for use.
may result in the deterioration of water
quality and may significantly decrease
the hydraulic capacity of water mains by
promoting pipe wall pitting and the
growth of tubercles. If corrosion is not
inhibited, costly main replacements are
inevitable. Excessive dissolution of iron
and copper from plumbing and distribu-
tion systems can cause aesthetic prob-
lems with respect to taste, color, or
staining characteristics. Furthermore,
excessive lead levels can cause health
problems.
Silicate and phosphate compounds
have been used for corrosion control in
numerous water systems, but their use-
fulness and effectiveness in various sys-
tems has not been well documented.
Furthermore, the mechanisms by which
these compounds prevent corrosion
has not been clearly defined.
The general objectives of this re-
search were to study the factors that in-
fluence the corrosion protection offered
by selected silicate and phosphate com-
pounds. Specifically, the research ob-
jectives were to examine the effective-
ness of various dosages of selected
silicate and phosphate compounds ap-
plied to test water for corrosion inhibi-
tion of cast iron, copper, lead, and gal-
vanized steel coupons in a closed-pipe
loop system. The study examined effec-
tiveness of these compounds for corro-
sion inhibition under differing water
quality conditions using gravimetric
and electrochemical methods.
Chemical speciation studies were
also performed for different silicate and
phosphate compounds within the nor-
mal ranges of pH, alkalinity, chloride,
sulfate, and hardness concentrations
-------
encountered in the water supply sys-
tem. The corrosion products and coat-
ings on the metal coupons were also
characterized with and without the addi-
tion of corrosion-inhibiting compounds.
The corrosion inhibitors tested were
zinc polyphosphate (Calgon C-39), zinc
phosphate (Virchem V-391), sodium
metasilicate, and glassy silicate. The
test water was University of Missouri
tap water.
Methods and Materials
Gravimetric corrosion experiments
were conducted using a corrosion test
apparatus with four parallel polyvinyl
chloride (PVC) pipe loop systems of a
recirculating type similar to that pro-
posed in the Method B ASTM D2688-83.
The loop system had five locations
where duplicate metal coupons could
be inserted into the test water flow path
and selected coupons could be re-
moved for examination after a set expo-
sure period. The circulating test water in
the loop system was treated with
specific doses of the silicate or phos-
phate compound that was to be evalu-
ated for corrosion control.
Batch experiments were conducted to
evaluate the effectiveness of zinc
polyphosphate (Calgon C-39) for con-
trolling corrosion of cast iron in a gal-
vanic cell configuration under different
pH conditions. Cast iron and copper
coupons were connected by a copper
wire to form a galvanic cell and then
immersed in the test water. The test sys-
tem water quality parameters such as
total iron, alkalinity, hardness, or-
thophosphate, total phosphate, and pH
were measured at different times. At the
end of the test period (28 days), the cor-
rosion rates of the cast iron coupons
were measured by the gravimetric pro-
cedure.
Electrochemical corrosion of metal
samples was determined under differ-
ent environmental and water quality
conditions by measuring the polariza-
tion resistance of the sample.
Speciation studies for the phosphate
systems were done initially by using
equilibrium equations and constants
from the literature. Simultaneous equa-
tions were solved by an iterative
method on an Apple He microcomputer.
A more sophisticated approach was
later offered by REDEQLEPAK, an
aqueous chemical equilibrium pro-
gram. The program was set to allow for
precipitation of solids, to balance both
charge and mass, to keep the pH at 8.0,
and to be open to the atmosphere with
respect to CO2. Dilute aqueous solu-
tions of silicic acid were studied with the
interfaced calorimeter to look at en-
thalpies of protonation for various solu-
tions. Computer modeling of these so-
lutions was done using a program that
solves simultaneous equations, SEQS,
to determine the species present. All sil-
ica solutions were prepared upon dilu-
tion of a stock 0.1 M Si02 solution from
solid hydrated metasilicate
(Na2Si03-9H2O).
The corrosion products on the cast
iron coupons were analyzed by X-ray
diffraction and scanning electron micro-
scope (SEM) analysis. A nuclear mag-
netic resonance (NMR) spectrometer
was used to determine the extent of py-
rophosphate hydrolysis in the test
water.
Results and Conclusions
Gravimetric Corrosion
Evaluation
1. At doses S4.36 mg/L as P in test
water, zinc polyphosphate (Calgon C-
39) effectively controlled the corrosion
rate of cast iron after 21 days of expo-
sure, at a temperature of 30" ± 2.5°C and
a flow rate of 19.5 cm/sec. At lower tem-
peratures (20° ± 2.5°C) and flow rates
(9.5 cm/sec), it was also effective in con-
trolling the corrosion of cast iron at 4.36
mg/L as P. The system with 13.09 mg/L
as P had a higher corrosion rate com-
pared with the control after 26 days of
exposure. Zinc polyphosphate was
more effective in controlling the corro-
sion of cast iron at pH 5.0 than at higher
pH values.
2. At 2.18 mg/L as P in test water, zinc
orthophosphate (Virchem V-391) was
marginally more effective than the con-
trol in inhibiting the corrosion of cast
iron after 38 days of exposure time.
Higher doses of zinc orthophosphate
did not effectively control the corrosion
of cast iron.
3. At a dose of 30 mg/L as Si02 in tap
water, sodium metasilicate was most
effective in controlling the corrosion of
cast iron compared with other doses
tested.
4. At a dose of 15 mg/L as Si02 in test
water, glassy silicate controlled the cor-
rosion of cast iron more effectively than
other doses tested.
5. Among the four corrosion in-
hibitors tested, zinc polyphosphate was
the most effective in controlling the cor-
rosion of cast iron at comparable tem-
peratures and flow velocities.
6. At doses of 4.36 mg/L and 13.09
mg/L as P in test water, zinc polyphos-
phate effectively controlled the corro-
sion of copper.
7. At 13.09 mg/L as P, zinc polyphos-
phate had the lowest lead corrosion
rate, but the difference between lead
corrosion rates at 4.36 and 8.72 mg/L as
P was not much higher than at 13.09
mg/L as P after 28 days of exposure.
8. At the doses tested (4.36 to 13.09
mg/L as P), zinc polyphosphate did not
effectively control the corrosion of gal-
vanized steel.
9. The alkalinity and hardness de-
creased with time in all the systems
tested. The largest decrease was ob-
served in the blank systems. A good
correlation generally existed between
the alkalinity and hardness decreases in
the different systems. This result indi-
cates some precipitation of CaC03 and/
or calcium silicate or calcium phos-
phate, depending on the system.
Decreases in alkalinity and hardness in
the zinc-poly phosphate-dosed systems
were higher with cast iron than with
copper, lead, or galvanized steel sys-
tems. Precipitation of CaCO3 in the
blank systems did not sufficiently in-
hibit corrosion compared with systems
having inhibitors. A positive saturation
index did not protect the metal from
corrosion. Generally, little correlation
existed between corrosion rate and sat-
uration index.
Electrochemical Corrosion
Testing
1. The corrosion rate of cast iron was
significantly lower in the absence o
oxygen. Systems with zinc salts pro
vided better corrosion protection fo
cast iron than did zinc salt combinec
with pyrophosphate in oxygenated sys
terns in distilled water experiments.
2. Up to a point, orthophosphate in
creased the effectiveness of the sodiurt
hexametaphosphate corrosion inhibito
for cast iron in test water systems; bu
larger doses of these chemicals raise*
the corrosion rate.
3. The presence of zinc salt increase*
the effectiveness of sodium pyrophos
phate corrosion inhibitor for cast iron ii
test water. Adding more zinc salt im
proved the inhibitor's effectivenes
more than adding a correspondin
amount of pyrophosphate.
4. Systems with a zinc polyphos
phate (Calgon C-39) dose of 13.09 mg/
as P in test water produced the lowe;
cast iron corrosion rate.
-------
5. Zinc orthophosphate (Virchem V-
931) at comparable doses did not con-
trol the corrosion rate of cast iron in test
water as effectively as zinc polyphos-
phate (Calgon C-39).
6. At nominal doses of sodium
metasilicate (up to 50 mg/L as Si02), the
cast iron corrosion rate in test water
was not measurably affected compared
with the blank system. At doses up to 30
mg/L as Si02, glassy silicate did not ef-
fectively inhibit the cast iron corrosion
rate.
7. A low dose of zinc polyphosphate
(2.18 mg/L as P) effectively controlled
the corrosion rate of copper, but sodium
silicate was not as effective at compara-
ble doses.
8. Zinc polyphosphate controlled the
corrosion of lead and galvanized steel
more effectively than sodium silicate,
but the overall corrosion rates were
quite low, and meaningful conclusions
were difficult to obtain.
Comparison of Gravimetric and
Electrochemical Corrosion
Testing
1. The electrochemical corrosion
rates were generally higher than the
gravimetric corrosion rates measured
on the same metal under similar test
conditions. Note, however, that the
electrochemical corrosion test meas-
ures the corrosion rate on a clean metal
surface, whereas the gravimetric
method measures the average corro-
sion rate for the exposure period during
which the metal surface is subjected to
corrosion and deposition of compounds
produced from inhibitor reactions.
2. In many instances, similar corro-
sion inhibition trends were measured
by the two methods in the presence of
various corrosion inhibitors for different
metals (e.g., cast iron, lead, and galva-
nized steel in the presence of zinc
polyphosphate). However, the corro-
sion rate data produced by the two
methods did not correlate well for some
systems (e.g., cast iron in the presence
of sodium silicate, and copper in the
presence of zinc polyphosphate). Selec-
tion of the corrosion rate measurement
method will therefore depend on the
type of system being studied.
SEM and X-Ray Diffraction
Studies
1. Application of scanning electron
microscopy (SEM) to the cast iron corro-
sion products showed the presence of
iron, silica, phosphorus, aluminum, and
zinc. Phosphorus was found in samples
from tests that had used phosphate in-
hibitors for corrosion control. A sample
of corrosion products from a blank sys-
tem (test water, cast iron, and no in-
hibitor) showed significant amounts of
calcium; otherwise, the calcium content
of the corrosion product was quite
small. The SEM data on cast iron cou-
pons (from which corrosion products
had been removed) indicated the pres-
ence of zinc, phosphorus, and calcium
on the protected areas of the coupon
compared with the unprotected part.
2. The X-ray diffraction data on the
cast iron corrosion product showed the
presence of zinc phosphate and calcium
phosphate when zinc orthophosphate
inhibitor was used. In the presence of
sodium metasilicate inhibitor, the corro-
sion products contained iron silicate
compounds. After removal of corrosion
products. X-ray diffraction of the cast
iron coupon surface that had been in-
hibited with phosphate compounds
showed very little residual phosphate.
But systems using sodium metasilicate
inhibitor indicated the presence of vari-
ous silicate compounds on the coupon
surface.
Speciation Studies
1. A model solution of test water con-
taining zinc and pyrophosphate indi-
cated that 20 mg/L Na4P207 reduced the
corrosion of cast iron significantly. Ad-
dition of 10 mg/L ZnS04 in the presence
of 20 mg/L Na4P207 further reduced the
corrosion rate. This additional corrosion
reduction was due to the formation of
zinc silicate (silicates are present natu-
rally in the test waters). At doses of
ZnSOa >30 mg/L, the corrosion rate
drop was due to the formation of zinc
pyrophosphate.
2. In the experiments with sodium
metasilicate as an inhibitor (<60 mg/L
as SiO2 and pH 8), the occurrence of cal-
cium and magnesium silicate should
theoretically be negligible, since most
of the silicate will occur as Si(OH)4. The
extent of silicate polymerization should
be negligible. The silicate solid phase
compounds are very difficult to identify
because of the large number of possibil-
ities. More kinetic and thermodynamic
data are needed to define the com-
pounds formed.
NMH Studies
At pH 10, there was no hydrolysis of
the pyrophosphate compound tested.
At pH 8.0, which is nearer that of the test
water, pyrophosphate hydrolysis was
negligible during the electrochemical
corrosion tests. Some hydrolysis could
probably occur during the course of the
gravimetric studies, which can last a
month or more.
Recommendations
1. Study results indicate that some
corrosion inhibitors provide better pro-
tection for some metallic systems than
others. To evaluate the corrosion inhibi-
tion effectiveness of an inhibitor com-
pound or to evaluate the corrosiveness
of a water system, one must perform
either a gravimetric or an electrochemi-
cal corrosion test. In most cases, the
two tests give parallel results. Utilities
should use such corrosion-testing pro-
cedures to validate their use of
corrosion-inhibiting compounds.
2. The saturation index is of minimal
value in monitoring the corrosivity of
water to various metals. Even a system
with a positive saturation index and
CaCO3 precipitation does not prevent
corrosion of the metal under consider-
ation. A better corrosion-monitoring
index is needed for the utilities to evalu-
ate their water supplies.
3. More research needs to be done to
determine the kinetic and thermody-
namic data for the solid phase and com-
plexed silicate systems.
The full report was submitted in fulfill-
ment of Cooperative Agreement No.
CR-809759 by the University of Missouri
under the sponsorship of the U.S. Envi-
ronmental Protection Agency.
-------
Shankha K. Banerji, John E. Bauman. and John T. O'Connor are with the
University of Missouri, Columbia, MO 65211.
Marvin C. Gardels is the EPA Project Officer (see below).
The complete report entitled "Evaluation of Silicate and Phosphate Compounds
for Corrosion Control," (Order No. PB 87-180 972/AS; Cost: $18.95, subject
to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Water Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
BULK RATE
POSTAGE & FEES P/
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/600/S2-87/031
0169064 WERL
TraBTircr
US cP4
LIBRARY
CHICAGO
IL 60604
------- |