United States
                   Environmental Protection
                   Agency
Municipal Environmental
Research Laboratory
Cincinnati, OH 45268
                   Research and Development
EPA-600/S2-84-130 Sept. 1984
&ER&         Project  Summary
                   Seattle  Distribution  System
                   Corrosion  Control  Study:
                   Volume  VI.  Use  of  a  Rotating
                   Disc  Electrode to Assess
                   Copper  Corrosion
                   Ronald D. Hilburn
                     The  uniform corrosion of copper
                   tubing used for transport of Tolt River
                   water is characterized in this study as a
                   heterogeneous rate process composed
                   of metal oxidation and oxide film
                   growth, interfacial chemical reactions,
                   and mass transport in the liquid phase.
                   Quantitative rate expressions were
                   developed to characterize each of these
                   rate processes. Experiments designed
                   to measure the temperature and pH
                   dependence of corrosion under rate
                   control by each process were conducted
                   using steady-state  electrochemical
                   techniques.  The persistent and unex-
                   pected influence of solution  transport
                   of a reaction product, presumed to be
                   OH~, complicated characterization and
                   identification of underlying  rate pro-
                   cess. Surface pH could be characterized
                   empirically as a function of solution
                   temperature, pH, and diffusion layer
                   thickness.
                     This empirical correlation for surface
                   pH along with solution mass transport
                   models developed for turbulent and
                   laminar pipe flow were combined to
                   form a steady-state pipe flow model for
                   uniform copper corrosion. Predictions
                   made using  the model under stagnant
                   and low flow rate conditions show a sta-
                   ble and low corrosion rate of 0.2 mils
                   per year (MPY) in water of pH > 6.0. At
                   lower pH, predicted rates are substan-
                   tially increased as the pH is reduced and
                   temperature is increased. At high flow
                   rates, tremendous acceleration of cor-
rosion rate occurs, which again increas-
es with increasing temperature and de-
creasing pH. Only at pH > 8.0 are the
dramatic pH and temperature effects
dissipated so that the rate is stabilized
at a minimum value of approximately
0.2 MPY.
  Steady-state electrochemical tech-
niques gave rapid, reliable, and repro-
ducible corrosion rate  measurements
and provided the versatility necessary
to characterize quantitatively a hetero-
geneous rate process like aqueous cop-
per corrosion.
  This Project Summary was developed
by EPA's Municipal Environmental
Research 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

Background
  Corrosion of materials used to transport
drinking water is both a public health
concern and an  enormous economic
problem. In 1975,  the estimated national
cost for corrosion  in industrial, commer-
cial, and public establishments was $82
billion. This figure included monies spent
for replacement of deteriorated parts and
equipment, maintenance and repair, and
direct expenditures for corrosion control.
Costs as high as  $375 million per year

-------
have been estimated for  replacement
parts and repair of drinking water
distribution systems transmitting corro-
sive water, with an additional cost of $27
million per year for water treatment to
control  corrosion. The  Seattle Water
Department has estimated repair costs to
consumer plumbing systems to be 10 to
20 times  higher than costs associated
with the  distribution system. Annual
expenditures of $500,000  for corrosion
control in Seattle are estimated to reduce
consumer costs by $2 million.
  Corrosion occurs in the distribution or
plumbing system — the part of the drink-
ing  water cycle  closest  to the point of
customer use. Water quality degradation
and contamination occur because of the
release of metal cations and other
corrosion products into the water. Public
health  concerns involve ingestion and
bodily  accumulation of metal cations
from drinking water.  Dissolved  lead and
cadmium are two contaminants likely to
be present above permitted levels as a
result of aqueous corrosion in plumbing
systems using galvanized steel or copper
tubing with lead/tin solder.
  Adverse aesthetic effects often result
from the leaching of copper, iron, zinc,
and manganese from corroding pipelines.
For example,  drinking  water may be
undesirable because it has an unpleasant
taste or color,  or because it promotes
staining of porcelain bathroom and kit-
chen fixtures. All of  these effects arise
from a  chemical reaction between a
structural  material and a chemical com-
ponent of the transported natural water
As the reaction  proceeds,  the  metal  is
thinned and/or pitted, thus shortening its
useful lifetime according to the rate of the
overall reaction.  Reaction products may
also promote precipitation of solids,
which accumulate on the metal surface
and in some cases reduce water pressure
and pipeline carrying capacity. Corrosion
control efforts are then aimed at slowing
down the reaction to an acceptable rate or
stopping  it completely without creating
further ecological or water quality prob-
lems.

Purpose and Scope of Work
  This study  characterizes aqueous
copper corrosion as a heterogeneous rate
process by combining principles from
various disciplines to develop quantitative
rate expressions for each component rate
process involved. Principles taken from
solid-state, electrochemical, and corro-
sion sciences are combined with those of
water chemistry and environmental
engineering to describe  and explain the
overall corrosion process as a composite
result of fundamental rate processes.
Although the component rate  processes
are coupled,  each  is influenced  by a
distinct  set of environmental variables
that affects the rate at which it proceeds.
Thus this work presents an overview of
copper corrosion  in  drinking  water  as
a composite result of several fundamental
rate processes, and it determines through
laboratory experiments which rate pro-
cesses exert the greatest rate-controlling
influence on the overall process. Environ-
mental variables most important in influ-
encing the rate-controlling processes are
also evaluated.
  Corrosion rate was measured with
standard steady-state electrochemical
techniques augumented  with  special
instrumentation necessary for measure-
ment in  natural waters of low conductiv-
ity. Measurements were  made under
varying  conditions  of rate  control  to
evaluate the dependence of component
processes on temperature, fluid motion,
and chemical composition of the system.

Results

General Observations
  The likelihood of rate control by each of
the component rate processes is shown
in a comparison of the magnitude  of the
three  types of data  as a function  of pH
(Figures 1 and 2) at 25°C. The following
general  observations are made:
  a) Corrosion rates measured at  pH >
    7.0 on electrodes without oxide films
    were much greater than those mea-
    sured on oxide-covered electrodes.
  b) Corrosion rates  measured in stag-
    nant solution  were considerably
    lower than those made at 3OOO RPM.
These observations demonstrate both the
influence of oxide film growth and solu-
tion mass transport on the overall rate
and, thus, suggest  that a  model of the
overall  process  include  the coupled
effects  of  these two rate processes.
Kinetic data may be important, but only at
sites  where  the oxide film  has been
damaged or removed. The fact that solu-
tion mass transport  control data were of
smaller  magnitude  than the  other two
sets of  data at corresponding pH and
temperature values indicates  that some
species  exist in solution whose diffusion
affects the rate of the overall corrosion
process
  The rate of the process  thus depends
not only on the chemical  reactions
involved and the growth of oxide film, but
also on the hydrodynamic flow regime of
the corroding system since it inevitably
influences the  rate  of solution  mass
transport. A definite solution  mass
transport influence  exists,  and the  A
species exerting that influence may very  "
well be OH" diffusing away from the sur-
face into bulk solution. In situations
involving a reaction product that must
diffuse away from the reacton site, a per-
sistent influence is exerted on the overall
process rate (even at very high Reynolds
Numbers) to  the  point of masking the
effects of the underlying process. A slow
diffusion of OH" away from the surface
increases the pH of the solution adjacent
to the oxide surface and reduces the oxide
film growth rate. As the diffusion rate is
increased (at  higher Reynolds Numbers),
the surface pH can drop, allowing a faster
rate of oxide film growth  and greater
corrosion rate. The  high pH values that
arise at the  oxide surface may promote
precipitation  of such solids as Cu(OH)2,
which  have low solubilities. If such is the
case, actual  rate control of the corrosion
process  may occur  in  the solution
adjacent to the oxide surface,  not within
or on the oxide itself. In addition, a variety
of Cu(OH)n"n complexes may form in the
solution adjacent to the  oxide surface,
reducing the "free OH"" concentration and
lowering the  surface pH.

Pipe Flow  Model
  The principal value of corrosion research   A
performed in  the laboratory  lies in its   \
application to real-life corroding systems,
such as copper tubing used for cold water
plumbing. A simulation model is presented
for  predicting the rate at which  copper
tubing  will corrode under a given set of
environmental conditions. The model is
based on (1) the presumption that mass
transport of OH"  is the rate-controlling
process and (2) the combined results of
laboratory studies with  quantitative
models for mass transport in laminar and
turbulent pipe flow. The result is an ability
to predict the rate of uniform  copper
corrosion in cold water plumbing systems
under varied conditions of flow, tempera-
ture, and pH of the water.
  Results of  the  calculation for varying
Reynolds  Numbers at 25°C  appear  in
Figure 3. The rates increase  with flow
rates at Reynolds Numbers greater than
2000. Effects of pH < 6 are apparent at all
flow rates. Rates become very high at low
pH  and high flow rate.
  The  influence of temperature appears
in Figure 4, which shows corrosion rates
versus pH  at a  high  Reynolds Number
(50,000). The corrosion rate is much
more pH dependent at 15°C  and 25°C
than at 5°C.
  Model predictions  provide valuable
insight into  identification of principal  A
variables affecting  the overall process,  ™

-------
    1.5
                                      T-25°C
                                                                 * *
      6.0
                 6.5
                                                              8.5
                            7.0         7.5

                                     pH

Figure 1.    Comparison of corrosion rates measured with and without Oxide film at 25°C.
the extent to which each variable exerts
an influence, and the range of magnitude
in which the maximum effect is mani-
fested. Possible strategies for corrosion
control (or at least corrosion rate reduc-
tion) are also suggested.

Conclusions
  Over the pH range of 6.0 to 8.5, the
rate of copper corrosion is reduced as pH
is increased. Over the temperature range
of 5°C  to 25°C,  the  rate of copper
corrosion  is reduced as temperature is
reduced. The presence of a Cu20 film on
the copper surface reduces the corrosion
rate at all  pH and temperature  values
studied.
  The transport of a reaction product
(presumed  to  be OH~)  away from the
oxide/solution interface is the principal
process controlling the overall corrosion
rate. At low pH (6.0), rate control by mass
transport is nearly complete. At higher pH
                                        values (>8.0), the influence of the under-
                                        lying rate process exerts a greater
                                        influence on the overall rate.
                                          In pipe flow under stagnant or low rate
                                        (laminar flow) conditions, the corrosion
                                        rate of copper is stabilized for pH > 6.0 at
                                        a value of 0.2 MPY. Only when the pH is
                                        reduced below 6.0  do the  accelerating
                                        effects of low pH and high  temperature
                                        manifest themselves. At high (turbulent)
                                        flow rates in pipe flow, corrosion rates are
                                        accelerated dramatically with reduced pH
                                        and increased temperature.  Under these
                                        conditions,  only at  pH >8.0  is the
                                        corrosion rate stabilized to an acceptably
                                        low value of 0.2 MPY.
                                          The use of steady-state electrochemical
                                        techniques gave reliable and reproduci-
                                        ble corrosion rate measurements, even in
                                        water of low  conductivity.  In  addition,
                                        they  provide the versatility needed for
                                        experimental design of adequate sophis-
                                        tication to provide data that  can be used
in mechanism determination and model
development. These qualities greatly
enhance the research capabilities of the
investigator.

Recommendations
  The  ability to characterize complex
metal-oxide-solution  systems as hetero-
geneous rate processes is an important
step in determining  the dependence of
the corrosion rate on system variables.
Crucial to the development of aqueous
corrosion science is the gathering of
appropriate multidisciplinary information
to develop quantitative rate expressions
for processes involved in the general
corrosion of  metals  other than copper
used in the transport of drinking water.
  The  model presented  for  aqueous
copper corrosion should be extended and
refined. Extension to a broader range of
aqueous species and concentrations to
include species involved in water treat-
ment processes  such as chlorination
seems advisable.  Also in order are (1)
more sophisticated mathematical model-
ing of the oxide/solution interf acial inter-
actions controlling the solution pH just
next to the interface, and (2) precipitation
studies involving Cu2+ and OH~ species.
An evaluation could then be made of the
effect of precipitated hydroxide and
carbonate solids in altering the corrosion
rate by depositing on the metal surface.
  The  steady-state methods used  here
should be further  applied to the study of
other  metal-oxide-solution systems
along with  the testing  and development
of transient electrochemical techniques
such as the A.C. impedance method. The
latter, because of its possibility for instan-
taneous measurement, can  provide in-
valuable insight into the coupled rate pro-
cesses involved in aqueous corrosion.
  The full report was submitted in fulfill-
ment of Contract No.  R806686-010 by
the University of Washington, under the
sponsorship of the U.S. Environmental
Protection Agency.

-------
  0.5
  0.4
I
„• 0.3

I
.1
8-
o
  o.i
    o  t
      6.0
                                               T-2S°C
                Oxide film growth
               Stagnant diffusion
                  6.5
7.0
                                           7.5
                                                       8.0
                                                                    8.5
Figure 2.   Comparison of corrosion rates measured with and without solution mass transport
            effects at 25°C.
  I.OO-i
       2.500
                       3.000
                           4.000
4.500
                                        3.500

                                           Log RE

Figure 3.    Corrosion rate estimates for 1/2-in. copper tubing at varied flowrates and 25°C.

-------
   1.00
   O.80
 ti 0.60
i
fc 0.40
o
   0.20
                              RE-50000
       5.O                   6.O                  7.0                   8.0



Figure 4.    The pH dependence of corrosion rates estimated for turbulent flow conditions.
                                                                                                          *USGPOs   1984-759-102-10697

-------
      Ronald D. Hilburn is with University of Washington. Seattle. Washington 98195.
      Marvin C. Gar dais is the EPA Project Officer (see below).
      The complete report, entitled "Seattle Distribution System Corrosion Control
        Study: Volume VI. Use of a Rotating disc Electrode to assess Copper Corrosion,"
        (Order No. PB 84-229 707; Cost: $19.00, subject to change) will be available
        only from:
              National Technical Information Service
              5285 Port Royal Road
              Springfield.  VA 22161
              Telephone: 703-487-4650
      The EPA Project Officer can be contacted at:
              Municipal Environmental 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 PAID
        EPA
  PERMIT No. G-35
Official Business
Penalty for Private Use S300
                              ACCT  -  C06324C
                              LCD  W  TiLLrY
                              REGION  V  EPA
                              LIBRARIAN
                     LAB  -  M€RL
                                                    ST
                                                             60604

-------