In te rn a tio n a l P ip elin e C o n feren ce — V o lu m e 1

A S M E 1996

IPC1996-1852

CORROSION AND CATHODIC PRO TECTIO N AT DISBONDED COATINGS

J.H . Payer, K.M. Fink, J.J. Perdomo, R.E. Rodriguez, I. Song and B. T rautm an

Downloaded from https://asmedigitalcollection.asme.org/IPC/proceedings-pdf/IPC1996/40207/471/2506669/471_1.pdf by guest on 19 April 2020

Department of Materials Science and Engineering
Case Western Reserve University
10900 Euclid Avenue
Cleveland, Ohio 44106

ABSTRACT A deterministic approach to pipeline integrity analysis requires

knowledge of the potential and solution chemistry of steel at
The effectiveness of cathodic protection to control exposed areas. Significant progress has been made in this area
corrosion and the resulting corrosion rate of pipelines arc over the last several years; however, there is much work to be
determined by the chemical and electrochemical conditions at done to permit predictions on a sound engineering basis.
local areas along the pipeline. The disbonding of coatings and
tapes is also controlled to a large extent by the chemical and A major thrust of our present work is to extend the
electrochemical conditions. Processes that occur on the metal deterministic understanding to include a broader range of
surface and their effect on corrosion and cathodic protection are relevant pipeline operating conditions. There has only been
discussed with respect to real pipeline conditions. Disbonded limited data for the effects of prior corrosion products in the
coatings on steel can interfere with the current distribution from exposed steel area and the effects of cyclic environmental
cathodic protection. Shielding the current under disbonded conditions, e.g., wet/dry cycles and interruptions to cathodic
coatings can affect the level of protection, the corrosion current flow to the exposed steel. The interactions between
behavior and the disbonding of coatings. A major thrust in our coatings degradation, e.g., disbonding, and cathodic protection
laboratories has been the use of laboratory measurements and are not clearly understood.
computational models to determine the changes in the corrosive
environment that occur beneath disbonded coatings as a function In this paper, the chemical and electrochemical processes
of applied potential, disbonded area geometry, prior corrosion that occur at the pipeline steel surface are identified. Recent
products and wet/dry cycles. These results are summarized here. work on factors controlling the cathodic disbonding process are
described. Results to determine the potential and solution
conditions under disbonded coatings are summarized.
INTRODUCTION

Buried pipelines are protected from corrosion by a CHEMICAL AND ELECTROCHEM ICAL
combination of a protective organic coating and cathodic P R O C E SSE S
protection. The cathodic protection system is designed to
protect steel where the coating is damaged. Aging pipelines can The corrosion behavior of steel is commonly described by
have degraded coatings and sites of corrosion. A major issue the chemical composition of an aqueous solution in contact
facing the industry today is the assessment of current status and with the steel surface and the electrochemical potential of the
reliability of aging pipelines. steel. This is shown schematically in the figures below which
represent steel exposed to ground waters at a holiday in the
Evaluation of pipelines requires a combination of operating protective coating on a pipeline. Based upon the solution
experience, inspection technology, corrosion control composition and the potential of the steel in the solution, the
technology-coatings and cathodic protection, and reliability/life steel will exhibit immune, passive or active behavior. If the
prediction technology. The inspection technology to locate and steel is polarized to highly reducing potentials, the steel is
measure corrosion damage is improving. Improved empirical thermodynamically stable and will not corrode. If the steel is
and deterministic reliability methods for pipelines are required. exposed to potential (E) and pH combinations in the passive

Copyright © 1996 by ASME

range, the steel will not continue to conrode because the f A
corrosion process forms a protective layer of insoluble STEEL BEHAVIOR
corrosion products. If the potential and pH combinations are in ■ Im m u n e - I r o n is s ta b le , n o r e a c tio n
regions where soluble corrosion products are stable, the steel
• P a s s iv e - I n s o l u b l e , P r o te c t iv e F il m ;
will continue to corrode.
o x id e , F e j O ^ F e O O H ,

The corrosion rate of steel is determined by the combined p h o s p h a te , c h r o m a t e . . .

anodic and cathodic polarization behavior on the steel in the - S o lu b le C o r r o s io n P r o d u c t

• C o r r o s io n
holiday and beneath any disbonded coating. The solution
composition and potential will determine the combined effects A c id — F e++

of activation, concentration and ohmic polarizations. There are A lk a lin e — H F e 0 2 *

microcell and macrocell effects within the region due to current - N o n - P r o t e c t iv e O x id e : Fe^O j
distribution throughout the disbonded area.

Downloaded from https://asmedigitalcollection.asme.org/IPC/proceedings-pdf/IPC1996/40207/471/2506669/471_1.pdf by guest on 19 April 2020

C a th o d ic
The pipeline steel can exhibit three modes of behavior P ro te c tio n

shown in Fig. 1. The steel will be immune (no corrosion), C u rre n t

passive (protected by an insoluble film) or active (corrosion).

The corrosion proceeds when the corrosion products are soluble
or when the insoluble products are non-protective. Which of
the three modes of behavior will apply will depend upon the
composition of the solution in contact with the steel and the
potential of the steel surface. The potential-pH diagram for iron
identifies the relative regions over which each behavior will be
observed in pure water (Pourbaix, 1974). The presence of other
ionic species in the water can significantly affect the location of
the corrosion, immune and passive regions on the diagram.

Figure 2 illustrates how electrochemical processes in a

given chemical environment can affect the integrity of the
pipeline protection system. Failure at any interface or in any of
the interfacial regions of metal, oxide or coating will result in
loss of adhesion and disbondment of the protective coating. It
is important to recognize that a variety of processes contribute
to the behavior and resulting durability of each of these areas.
The effectiveness, and thus the usefulness, of the coating will
be determined by the weakest region, and useful life will be
extended only by strengthening the weakest region. Depending Fig. 1. Corrosion behavior of steel.
upon the exposure conditions, the controlling mechanism can
change from failure at one interface to failure at another. It is
not surprising, then, that a variety and apparently conflicting The two principal cathodic reactions on buried steel
set of observations and conclusions have been reported from structures are oxygen reduction and hydrogen evolution. Both
laboratory and field studies of cathodic disbonding. In Figs. 2 of these reactions result in the solution at the holiday becoming
and 3, the processes that affect the behavior at each interface in more alkaline (higher pH). An additional consideration
the metal/coating system are described. identified in the work at Case is the damage that can result from
other products generated during cathodic polarization, e.g.,
The electrochemical polarization processes on steel at a peroxide and similar species (Gervasio and Payer, 1992; Payer
holiday or within a disbonded area are shown in Fig. 3. These et al., 1993a-c). The detrimental effects of alkaline solutions on
polarization processes describe the rate of cathodic and anodic organic coatings is well recognized, but the generation of higher
reactions and are determined by the inherent reactivity alkalinity solutions at the holiday also results in a beneficial
(activation polarization), concentration and transport of reaction effect. The benefit is that steel corrosion is greatly reduced in
species (concentration polarization) and electronic or ionic alkaline solutions (Payer et al.. 1994; Perdomo and Payer,
conductivity (ohmic polarization). Each type of polarization 1995).
will depend upon the solution composition and potential of the
steel. Important effects occur over large areas (macrocells) or in The reactions that occur and the rates of those reactions
local regions (microcells). The access to and distribution of will depend upon the transport processes into the holiday and
cathodic current over the disbonded area depends on the along the disbonded region. The transport of gases, water and
magnitude of current and the geometry of the disbonded area. ionic species are important in the overall process. Tire
transport will be controlled by the geometry of the disbonded 1992, 1993; Trautman, 1994; Rodriguez, 1996). The extent of
area, concentration gradients, potential gradients and convection. cathodic disbonding is strongly dependent upon the chemical
The effects of wetVdry cycles are an area that is clearly important and electrochemical conditions. Selected results from our work
and yet an area that has received little serious study. are shown in Figs. 4-6. The important factors of starting
solution composition, level of cathodic protection, level of
aeration and wet/dry cycles greatly affect the rate of disbonding
and the total amount of disbonding. None of these factors are
addressed in standard cathodic disbondment tests, and this results
f -------------------------------------------- \
POSSIBLE DISBO N DIN G PROCESSES in poor correlation between the standard tests and field
performance.
• M e t a l o x id e d i s s o l u t io n
• In te r fa c ia l p o ly m e r a tta c k
• P o l y m e r / o x i d e a d h e s io n lo s s
• M e t a l d i s s o l u t io n
• M e t a l / m e t a l o x id e a d h e s io n lo s s ---------------------------------------------N

Downloaded from https://asmedigitalcollection.asme.org/IPC/proceedings-pdf/IPC1996/40207/471/2506669/471_1.pdf by guest on 19 April 2020

v ______________ _______________ / POLARIZATION PROCESSES

•A ctivation: A n o d ic/O xidation

C a th o d ic Ca thodic/R educ tion
P ro te c tio n
•C oncentration: Oxygen
C u rre n t

•O hm ic (iR): Soil, Solution, Films

Soil
\_________________________ /
FBE

H o lid a y C a t h o d ic
P r o te c t io n
•Steel Substrate:: C u rre n t

• Polymer chemical
resistance
Cathodic
• Oxide stability
Disbonding Metal Oxide
• Solution composition
• Potential
¡Metal Substrate':

Fig. 2. Possible disbonding processes.

C A TH O D IC D ISBOND ING PR O C ESS

Organic coatings are all permeable to water, gases and ions

to varying degrees. They are not true barrier films. The
permeabilities will depend upon coating composition and
thickness. Some coatings will transport sufficient water and Fig. 3. Electrochemical polarization processes.
oxygen to support corrosion reactions beneath an intact coating.
This leads ultimately to coating disbonding and failure. The
chemical stability and adhesion of the coating to the pipeline is Figure 4 demonstrates the effect of starting solution
affected by its formulation. The surface preparation and coating chemistry and level of cathodic protection on the extent of
process are critical to behavior and performance. disbonding. Starting tests with a strongly alkaline solution
resulted in more severe disbonding, and disbonding was much
Disbonding of coatings and pipeline corrosion have been more severe as the level of cathodic protection was increased in
examined in our laboratories for several years (Gervasio et al.. all three starting solutions. The level of aeration was also found
to have a major affect oil caihodic disbonding. Starting with a Potential and solution chemistry have been experimentally
neutral environment and the same level of cathodic protection, measured to simulate the effects of highly insulating disbonded
over a 4 month period, the extent of disbonding was 25 times coatings on buried pipelines under cathodic protection (Perdomo
more severe for an air-saturated solution (25 cm2 disbonded area) et al., 1996a; Perdomo and Payer, 1996). Figure 7 shows the
than for a deaerated solution (less than 1 cm2). potentials in various locations within a disbonded coating in
laboratory test cells. Potentials within a disbondmeni shift
from a protection range (i.e., more cathodic than -0.85 V vs.
C u/C uS 04, CCS) to a corrosion range (i.e., toward the
corrosion potential of steel) after each of the five drying cycles.
When rewetting occurs, the potential shifts back to the
protection range as long as CP (Cathodic Protection) is
maintained. Figure 8 shows the effect of caihodic protection
interruption on solution pH. After short limes (hours) the pH

Downloaded from https://asmedigitalcollection.asme.org/IPC/proceedings-pdf/IPC1996/40207/471/2506669/471_1.pdf by guest on 19 April 2020

rises to levels (pH-11) where corrosion of steel is expected to
be low. When the CP current is interrupted, the pH shifts to
lower levels and corrosion occurs, but when CP is re­
established by rewetting, the protective pH is re-established and
corrosion is controlled by the change of the environment.

Fig. 4. Disbonding distance as a function of starting solution

and level of cathodic protection.

In a standard cathodic disbonding test, the positioning of

the anode during the test was shown to have a great effect on
the amount of disbonding for an FBE (Fusion Bonded Epoxy)
coating. When the anode was in the same test compartment as
the cathodic disbonding specimens, rapid and extensive
disbonding occurred. When the anode was placed in a separate
compartment, the extent of disbonding was greatly reduced.
These standard tests were run in an aqueous chloride solution,
and gaseous chlorine was generated at the anode throughout the
test. If this chlorine can migrate to the specimens, accelerated Fig. 5. Comparison of disbonding rates for specimens exposed
disbonding is observed. The effect of this was also observed in to continuous wetting and cyclic wetting-and-drying.
the "bleaching'“ of color from the specimen coatings. While
this condition accelerates disbonding in the laboratory, it is not Recent work (Perdomo et al., 1996a; Perdomo and Payer.
relevant to field conditions. Yet, cathodic disbonding tests are 1996) has suggested that somewhat caihodic potential values
still run by some using this test procedure. can be obtained even in the absence of current as long as the
environment directly in contact with the steel surface is alkaline
Figures 5 and 6 show the dramatic increase in disbonding (pH>9) and deaerated. This has been corroborated by modeling
with wet/dry cycles compared to that observed for continuously the changes within a disbonded coating crevice solution
wet conditions. Under most field conditions, alternate wetting (Perdomo et al., 1996b). The model used a one dimensional
and drying cycles are experienced with time, and none of the (length > width > thickness) crevice where no current tlow was
standard cathodic disbonding test protocols address this effect. assumed to occur directly through the coating. Even though the
current flow to occluded areas will depend on tire resistivity of
the medium and thickness of the disbondments. chemical
C O N D IT IO N S UNDER D IS B O N D E D C O A T IN G S changes (high pH and low oxygen) can be provided in occluded
areas (where conditions of stagnation prevail) by diffusion and
migration of species created by both the concentration and
potential gradients. coating as well as through macroscopic defects, e.g., holidays
or macrocracks. Furthermore, these microcracks allow less
Another important issue arisen from these results (Perdomo restricted transport of species to and from the steel surface.
et al., 1996a; Perdomo and Payer, 1996) is that the surfaces of Since the microcracks allow current to be easily passed, the
the steel pipes that have been exposed to real environments are environment at the steel surface can be readily modified to an
not in bare metallic conditions but have oxides such as alkaline condition where corrosion is minimized. However,
magnetite on the surfaces. These oxide films will thus affect interruption of CP could result in a more rapid loss of this
the corrosion rate of the pipe. Previous work (Perdomo et al., chemical protection to the surroundings than with an
1996a,b; Perdomo and Payer, 1996; Whitman et al., 1923, impermeable coating because of the microcracks.
1924) has looked at the determination of corrosion rates using
"fresh” steel samples where the presence of previous corrosion
products has been ignored. Because the kinetics of formation
and transformation of oxide films depends strongly on the -0.60
characteristics of the steel surface and the environment, the

Downloaded from https://asmedigitalcollection.asme.org/IPC/proceedings-pdf/IPC1996/40207/471/2506669/471_1.pdf by guest on 19 April 2020

electrochemical behavior of the steel surface with pre-existing
oxides (e.g., magnetite) needs to be studied. This will also help -0.70
to clarify the uncertainties in our current understanding of the
oxidation scheme of magnetite to FeJ+ species, as to whether
-0.80
the Fe3+ species will be soluble or insoluble in a given
environment (Evans, 1960, 1969; Evans and Taylor, 1972).
Corrosion rates determined from a specimen with an oxide of -0.90
interest (e.g., magnetite) on the surface could be more realistic
in terms of predicting the useful, remaining life of existing
on 1.00
pipelines. (J
-

- 1.10

- 1.20
Dry
Dry

Rewet -1.30
6.0 -
Dry
-1.40
Rewet
0 5 10 15 20 25
Dry
S 4.0 / Days
Dry / Wet/Dry
Rewet
Rewet y y/
y
2.0 Fig. 7. Effect of cathodic protection interruption on potential of
V
steel surface along a simulated disbonded cell in 10 mM
Continuous wetting- Na2S 0 4. CP at -1.08 V vs. CCS in a 0.89 mm (35 mil)
crevice. Arrows indicate the points of drying.
0.0 *£■------1--------i--------i--------i-------- i-------- U J ____ l_

0 4 6 12 16 20
Time (days) Determination of the chemical composition and structure of
corrosion products after controlled exposures provides insight
Fig. 6. Disbonding distance for wet/dry and continuously wet into the chemical and electrochemical processes underway
conditions within corroding regions. An array of microscopies and
spectroscopies are used to examine the corrosion products and
deposits from the laboratory exposures. Comparisons of these
As pipelines age, microcracks may develop and the results with existing knowledge regarding the transformation
coatings may become permeable. Therefore, it is necessary to and growth of oxides and other corrosion products can be used
design laboratory tests to simulate the conditions of such to develop protocols to determine conditions of active and
existing, aging pipelines and the effects of CP on these inactive corrosion. One such scheme for the structural
pipelines. This is an important consideration because the CP transitions in iron oxide/hydroxide is given by Bernal et al.
current can reach the steel surface through microcracks in the (1959). For example, in their report, transitions from hydrated,
ferrous to anhydrous, alpha Fe203 are shown as a function of • Is the present cathodic protection system giving effective
oxidation and dehydration processes. These processes can be corrosion control?
used to rationalize the effects of wet/dry cycles and changing • Were prior cathodic protection upgrades effective in
levels of polarization on corroding areas along pipelines. controlling corrosion?
• What will be the benefit of upgraded cathodic protection on
corrosion rate?
• How does the current condition of the coating influence
cathodic protection?
• Does the cathodic protection lead to further coating
degradation?
• Is rehabilitation necessary?

A deterministic approach to pipeline integrity analysis

requires knowledge of the potential and solution chemistry of

Downloaded from https://asmedigitalcollection.asme.org/IPC/proceedings-pdf/IPC1996/40207/471/2506669/471_1.pdf by guest on 19 April 2020

steel at exposed areas. Significant progress has been made in
this area over the last several years; however, there is much
work to be done to permit predictions on a sound engineering
basis. A major thrust of present work is to extend the
deterministic understanding to include a broader range of
relevant pipeline operating conditions.

ACKNOW LEDGM ENTS

Work in this area at Case has been supported from several

sources: (a) the Gas Research Institute; (b) the Alyeska
Pipeline Service Co.; and (c) Pipeline Research Committee of
American Gas Association. The authors also wish to
acknowledge 3M Austin Center for donating the epoxy resins.

0 5 10 15 20 25

Days REFERENCES

Bernal, J.D., Dasgupta, D.R., and MacKay, A.L., 1959,

Fig. 8. Effect of cathodic protection interruption on pH of the “The oxides and hydroxides of iron and their structural inter­
crevice solution along a simulated disbonded cell in 10 mM relationships,” Clay Miner. Bull. 4, 15.
Na-,S04. CP at -1.08 V vs. CCS in a 0.89 mm (35 mil) Evans, U.R., 1960, "The economic advantages of a sound
crevice. Arrows indicate the points of drying. painting scheme," Trans. Inst. Metal Finish. 37, 1.
Evans, U.R., 1969, "Mechanism of rusting", Corr. Sci.,
9, 813.
SUM M ARY Evans, U.R., and Taylor, C.A., 1972, "Mechanism of
atmospheric rusting", Corr. Sci. 12, 227.
The effectiveness of cathodic protection to control Gervasio, D„ and Payer, J.H., 1992, “Cyclic potential ring
corrosion and the resulting corrosion rate of pipelines are measurements (CPRM) for studying oxygen reduction on steel
determined by the chemical and electrochemical conditions at in basic aqueous solution with relevance to disbonding of
local areas along the pipeline. The condition of protective coatings from cathodically protected steel,” Abstract No. 128,
coatings and tapes on the pipeline greatly influences the Extended Abstracts of 182nd Meeting of the Electrochemical
response to cathodic protection and effectiveness of corrosion Society in Toronto. The Electrochemical Society, Pennington,
control. The results of an on-going laboratory-based program NJ.
to better understand these processes have been summarized here. Gervasio, D., Song, I., Trautman, B., and Payer, J.H.,
The results are relevant to the following important questions for 1992, “Fundamental research on disbonding of pipeline
operating pipelines: coatings,” GRI Report No. GRI-92/0166, Gas Research
Institute, Chicago IL.
• Is the corrosion active now? Gervasio, D„ Song, I., Trautman, B., and Payer, J.H.,
• When did the corrosion occur? 1993, “Fundamental research on disbonding of pipeline
• What is the corrosion rate if the pipeline continues to coatings,” GRI Report No. GRI-93/0265, Gas Research
operate as is? Institute, Chicago, IL.
 Payer, J.H., Trautman, B„ and Gervasio, D„ 1993a,
“Methods to determine the role of electrochemical reduction
products on coating disbonding from cathodically protected
steel,” pp. 109-110, in Proc. of the Am. Chem. Soc. Spring
Mtg.. Denver. CO. Div. of Polymeric Materials: Science and
Engineering. Vol. 68, American Chemical Society. .
Payer, J.H., Trautman, B., Gervasio, D„ and Song, I„
1993b, “Role of coating/oxide/steel interfaces on cathodic
disbonding of pipeline coatings,” pp. 21-26, Mat. Res. Soc.
Svmp. Proc.. Vol. 304, Polymer/Inorganic Interfaces, Opila,
R.L., Boerio, F.J., and Czandema, A.W., Eds., Materials
Research Society, Pittsburgh.
Payer, J.H., Trautman, B„ and Gervasio,. D„ 1993c,

Downloaded from https://asmedigitalcollection.asme.org/IPC/proceedings-pdf/IPC1996/40207/471/2506669/471_1.pdf by guest on 19 April 2020

“Chemical and Electrochemical processes of cathodic disbonding
of pipeline coating,” Paper No. 579, CORROSION 93. NACE
annual conference in New Orleans, National Association of
Corrosion Engineers, Houston, TX.
Payer, J.H., Trautman, B., Fink, K„ and Song, I., 1994,
“Mechanism of cathodic disbonding of pipeline coatings,”
presented at 6th annual Pipeline Monitoring and Rehabilitation
Seminar. Houston, TX.
Perdomo, J.J., and Payer, J.H., 1995, “Chemical and
electrochemical conditions on steel at disbonded coatings,”
AGA/PRC Report No. PR75-9310.
Perdomo-Diaz, J.J., Chabica, M.E., and Payer, J.H.,
1996a, "Chemical and electrochemical conditions on steel under
disbonded coatings: The effect of applied potential, solution
resistivity, crevice thickness and holiday size", to be submitted
to Corrosion.
Perdomo-Diaz, J.J., Cawley, J.D., and Payer, J.H., 1996b,
"Chemical and electrochemical conditions on steel under
disbonded coatings: Mathematical modeling of mass transport
and chemical changes in crevices", to be submitted to
Corrosion.
Perdomo-Diaz, J.J., and Payer, J.H., 1996, "Chemical and
electrochemical conditions on steel under disbonded coatings:
The effect of previously corroded surfaces and wet and dry
cycles", to be submitted to Corrosion.
Pourbaix, M„ 1974, Atlas of electrochemical equilibria in
aqueous solutions, pp. 307-321, NACE, Cebelcor, Brussels.
Rodriguez, R.E., 1996, “Influencing factors in cathodic
disbonding of fusion bonded epoxy coatings from a steel
substrate,” M.S. Thesis, Case Western Reserve University,
Cleveland, Ohio.
Trautman, B.L., 1994, “Cathodic disbonding of fusion
bonded epoxy coatings,” M.S. Thesis, Case Western Reserve
University, Cleveland, Ohio.
Whitman, W„ Russel, R., Welling, C., Cochrane, J.,
1923, "The effect of velocity on the corrosion of steel in
sulfuric acid,” Ind. Eng. Chem. 15, 672.
Whitman, W., Russel, R., and Aliteri, V., 1924, "Effect
of hydrogen-ion concentration on the submerged corrosion of
steel", Ind.. Eng. Chem. 16, 665.